BLUK077-Stein
April 14, 2007
15:29
Pulmonary Embolism
i
BLUK077-Stein
April 14, 2007
15:29
Pulmonary Embolism Second Edition by
Paul D. Stein,
MD
Director of Research Education St. Joseph Mercy-Oakland Pontiac, Michigan, USA Professor, Full Time Affiliate Department of Medicine Wayne State University School of Medicine Detroit, Michigan, USA Adjunct Professor of Medical Physics Oakland University Rochester, Michigan, USA
iii
BLUK077-Stein
April 14, 2007
15:29
C 1996 by Williams & Wilkins, Maryland C 2007 by Paul D. Stein
Published by Blackwell Publishing Blackwell Futura is an imprint of Blackwell Publishing Blackwell Publishing, Inc., 350 Main Street, Malden, Massachusetts 02148-5020, USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Blackwell Science Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia All rights reserved. No part of this publication may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without permission in writing from the publisher, except by a reviewer who may quote brief passages in a review. First published 1996 Second edition 2007 1
2007
ISBN: 978-1-4051-3807-9 Library of Congress Cataloging-in-Publication Data Stein, Paul D. Pulmonary embolism / by Paul D. Stein. – 2nd ed. p. ; cm. Includes bibliographical references and index. ISBN 978-1-4051-3807-9 (alk. paper) 1. Pulmonary embolism. I. Title. [DNLM: 1. Pulmonary Embolism. WG 420 s819p
2007]
RC776.P85s74 2007 616.2 49–dc22 2007005029 A catalogue record for this title is available from the British Library Commissioning Editors: Steve Korn and Gina Almond Development Editor: Beckie Brand Editorial Assistant: Victoria Pittman Set in 9/12 Minion and Frutiger by Aptara Inc., New Delhi, India Printed and bound in Singapore by Markono Print Media Pte. Ltd. Cost for publication of this book was supported in part by unrestricted grants from Diatide, Inc., Londonderry, New Hampshire and Dupont Pharmaceuticals Co., Wilmington, Delaware. For further information on Blackwell Publishing, visit our website: www.Blackwellmedicine.com The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards. Blackwell Publishing makes no representation, express or implied, that the drug dosages in this book are correct. Readers must therefore always check that any product mentioned in this publication is used in accordance with the prescribing information prepared by the manufacturers. The author and the publishers do not accept responsibility or legal liability for any errors in the text or for the misuse or misapplication of material in this book.
iv
BLUK077-Stein
April 14, 2007
15:29
Contents
Preface, ix Part I Prevalence, risks, and prognosis of pulmonary embolism and deep venous thrombosis
1 Pulmonary embolism and deep venous thrombosis at autopsy, 3 2 Incidence of pulmonary embolism and deep venous thrombosis in hospitalized patients, 16 3 Case fatality rate and population mortality rate from pulmonary embolism and deep venous thrombosis, 19 4 Prognosis in acute pulmonary embolism based on right ventricular enlargement, prognostic models, and biochemical markers, 24 5 Changing risks of untreated deep venous thrombosis and acute pulmonary embolism, 31 6 Resolution of pulmonary embolism, 35 7 Upper extremity deep venous thrombosis, 37 8 Thromboembolic disease involving the superior vena cava and brachiocephalic veins, 41 9 Venous thromboembolic disease in the four seasons, 44 10 Regional differences in the United States of rates of diagnosis of pulmonary embolism and deep venous thrombosis and mortality from pulmonary embolism, 47 11 Venous thromboembolism in the elderly, 52
12 Pulmonary thromboembolism in infants and children, 66 13 Venous thromboembolism in men and women, 68 14 Comparison of the diagnostic process in black and white patients, 72 15 Pulmonary thromboembolism in Asians/Pacific Islanders, 76 16 Pulmonary thromboembolism in American Indians and Alaskan Natives, 83 17 Venous thromboembolism in patients with cancer, 85 18 Venous thromboembolism in patients with heart disease, 93 19 Venous thromboembolism in patients with ischemic and hemorrhagic stroke, 98 20 Pulmonary embolism and deep venous thrombosis in hospitalized adults with chronic obstructive pulmonary disease, 101 21 Pulmonary embolism and deep venous thrombosis in hospitalized patients with asthma, 107 22 Deep venous thrombosis and pulmonary embolism in hospitalized patients with sickle cell disease, 109 23 Venous thromboembolism in pregnancy, 113 24 Air travel as a risk for pulmonary embolism and deep venous thrombosis, 119 25 Estrogen-containing oral contraceptives and venous thromboembolism, 122
v
BLUK077-Stein
April 14, 2007
15:29
vi
26 Obesity as a risk factor in venous thromboembolism, 125 27 Hypercoagulable syndrome, 128 Part II Diagnosis of deep venous thrombosis
28 Deep venous thrombosis of the lower extremities: clinical evaluation, 139 29 Clinical model for assessment of deep venous thrombosis, 144 30 Clinical probability score plus single negative ultrasound for exclusion of deep venous thrombosis, 147 31 D-dimer for the exclusion of acute deep venous thrombosis, 149 32 D-dimer combined with clinical probability assessment for exclusion of acute deep venous thrombosis, 158 33 D-dimer and single negative compression ultrasound for exclusion of deep venous thrombosis, 160
Contents
42 The history and physical examination in all patients irrespective of prior cardiopulmonary disease, 192 43 Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes, 197 44 Clinical assessment in the critically ill, 203 45 The electrocardiogram, 206 46 The plain chest radiograph, 216 47 Arterial blood gases and the alveolar–arterial oxygen difference in acute pulmonary embolism, 221 48 Fever in acute pulmonary embolism, 229 49 Leukocytosis in acute pulmonary embolism, 232 50 Alveolar dead-space in the diagnosis of pulmonary embolism, 234 51 Neural network computer-assisted diagnosis, 236
34 Contrast venography, 161
52 Empirical assessment and clinical models for diagnosis of acute pulmonary embolism, 239
35 Compression ultrasound for the diagnosis of deep venous thrombosis, 164
53 D-dimer for the exclusion of acute pulmonary embolism, 243
36 Impedance plethysmography and fibrinogen uptake tests for diagnosis of deep venous thrombosis, 168
54 D-dimer combined with clinical probability for exclusion of acute pulmonary embolism, 250
37 Computed tomography for diagnosis of deep venous thrombosis, 171
55 D-dimer in combination with amino-terminal pro-B-type natriuretic peptide for exclusion of acute pulmonary embolism, 253
38 Magnetic resonance angiography for diagnosis of deep venous thrombosis, 175 39 P-selectin and microparticles to predict deep venous thrombosis, 179
56 Low tissue plasminogen activator plasma levels and low plasminogen activator inhibitor-1 levels as an aid in exclusion of acute pulmonary embolism, 254
Part III Diagnosis of acute pulmonary embolism
57 Echocardiogram in the diagnosis and prognosis of acute pulmonary embolism, 255
40 Clinical characteristics of patients with no prior cardiopulmonary disease, 183
58 Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism, 260
41 Relation of right-sided pressures to clinical characteristics of patients with no prior cardiopulmonary disease, 190
59 Techniques of perfusion and ventilation imaging, 262
BLUK077-Stein
April 14, 2007
15:29
vii
Contents
60 Ventilation–perfusion lung scan criteria for interpretation prior to the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED), 267 61 Observations from PIOPED: ventilation–perfusion lung scans alone and in combination with clinical assessment, 271 62 Ventilation–perfusion lung scans in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior cardiopulmonary disease, and patients with chronic obstructive pulmonary disease, 278 63 Perfusion lung scans alone in acute pulmonary embolism, 280 64 Probability interpretation of ventilation–perfusion lung scans in relation to largest pulmonary arterial branches in which pulmonary embolism is observed, 282 65 Revised criteria for evaluation of lung scans recommended by nuclear physicians in PIOPED, 284
72 Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries, 318 73 Quantification of pulmonary emboli by conventional and CT angiography, 319 74 Complications of pulmonary angiography, 321 75 Contrast-enhanced spiral CT for the diagnosis of acute pulmonary embolism before the Prospective Investigation of Pulmonary Embolism Diagnosis, 325 76 Methods of PIOPED II, 340 77 Multidetector spiral CT of the chest for acute pulmonary embolism: results of the PIOPED II trial, 348 78 Outcome studies of pulmonary embolism versus accuracy, 355 79 Contrast-induced nephropathy, 357 80 Radiation exposure and risk, 359 81 Magnetic resonance angiography for the diagnosis of acute pulmonary embolism, 364
66 Criteria for very low probability interpretation of ventilation–perfusion lung scans, 288
82 Serial noninvasive leg tests in patients with suspected pulmonary embolism, 371
67 Probability assessment based on the number of mismatched segmental equivalent perfusion defects or number of mismatched vascular defects, 294
83 Predictive value of diagnostic approaches to venous thromboembolism, 373
68 Probability assessment based on the number of mismatched vascular defects and stratification according to prior cardiopulmonary disease, 298 69 The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation–perfusion lung scans, 304 70 Single photon emission computed tomographic perfusion lung scan, 310 71 Standard and augmented techniques in pulmonary angiography, 311
84 Diagnostic approaches to acute pulmonary embolism, 376 Part IV Prevention and treatment of deep venous thrombosis and pulmonary embolism
85 New and old anticoagulants, 389 86 Prevention of deep venous thrombosis and pulmonary embolism, 405 87 Treatment of deep venous thrombosis and acute pulmonary embolism, 414 88 Withholding treatment of patients with acute pulmonary embolism who have a high risk of bleeding provided and negative serial noninvasive leg tests, 422
BLUK077-Stein
April 14, 2007
15:29
viii
89 Thrombolytic therapy in acute pulmonary embolism, 425 90 Thrombolytic therapy for deep venous thrombosis, 437 91 Inferior vena cava filters: trends in use, complications, indications, and use of retrievable filters, 444 92 Catheter-tip embolectomy in the management of acute massive pulmonary embolism, 454
Contents
93 Pulmonary embolectomy, 459 94 Chronic thromboembolic pulmonary hypertension and pulmonary thromboendarterectomy, 464 Index, 467
BLUK077-Stein
April 14, 2007
15:29
Preface
Since the first edition of Pulmonary Embolism was published in 1996, major strides have been made in many aspects of pulmonary embolism and its immediate cause, deep venous thrombosis. The purpose of this second edition is to provide detailed, in depth information on pulmonary embolism in a format that is readily usable by practicing physicians, and at the same time, provide fully referenced data that can serve as a resource for physicians with a deeper interest in the field. Many associates contributed to the investigations upon which much of the information in this book is based. The PIOPED I investigators supplied data for an impeccable database used for much of what we know about the accuracy of clinical assessment as well as ventilation–perfusion lung scans. The PIOPED II investigators supplied an equally impeccable database upon which much of what we know about the accuracy of multidetector CT angiography is based. This database is just beginning to be examined for useful information on a variety of additional subjects. The PIOPED III investigators are just starting to evaluate magnetic resonance angiography, and hope to acquire an equally impressive database. Grants from the National Institutes of Health, National Heart Lung and Blood Institute (NHLBI) made all of this possible. Guidance by representatives of the NHLBI contributed to the success of the study. Many of the investigators who participated in PIOPED I continued through PIOPED II and into PIOPED III. Some even participated in the Urokinase-Pulmonary Embolism Trial, published in 1973, which contributed a huge amount of information about pulmonary embolism, beyond that which was learned about thrombolytic therapy. Investigators with preferences for diverse approaches to the diagnosis and management of pulmonary embolism subverted their personal interests to collabo-
rate on scientific levels. This resulted in collaborations with only one goal: advancement of the field. The investigators, including physicians, statisticians, nursecoordinators, and technicians, often contributed long and hard hours with little reward. I am thankful for their efforts, and for the deep friendships with many that have resulted from these collaborations. Another database that was used extensively for epidemiological information related to pulmonary embolism and deep venous thrombosis was The National Hospital Discharge Summary. This database is available to the general public. Many details about its correct use, however, required consultation with representatives of the National Center for Health Statistics, and this assistance was graciously given. Several bright young men have worked with me over the years. Jerald W. Henry, MD, worked for several years on obtaining data from PIOPED I before going to medical school. He is now a practicing radiologist. More recently, Kalpesh C. Patel, MBBS, and Neeraj K. Kalra, MD, assisted. They are now finishing subspecialty training. Fadi Kayali, MD, did dedicated and superior work. He has been accepted into fellowship training. Afzal Beemath, MD, is a brilliant former fellow. He not only contributed importantly to several investigations, but also helped in a major way in the completion of this book. Nikunj Patel, MD, also worked long hours in assisting in the preparation of this book, as did his brother Hiren who assisted for a few months. Fadi Matta, MD, and Abdo Yaekoub, both of whom recently started with me, have done a sensational job. Finally, thanks to Steven Korn, formerly of Futura Publishing Company and Beckie Brand of Blackwell Publishing Company for encouraging me to complete this labor.
ix
BLUK077-Stein
March 12, 2007
I
14:45
PART I
Prevalence, risks, and prognosis of pulmonary embolism and deep venous thrombosis
1
BLUK077-Stein
March 12, 2007
1
14:45
CHAPTER 1
Pulmonary embolism and deep venous thrombosis at autopsy
Prevalence of pulmonary embolism at autopsy The prevalence of pulmonary embolism (PE) at autopsy varies according to the age and morbidity of the population studied. Dalen and Alpert in 1975 estimated that 15% of deaths in acute general hospitals and 25% of deaths in nursing homes or chronic hospitals were due to PE [1]. In more recent years, with more extensive use of antithrombotic prophylaxis, PE at autopsy was shown with similar prevalences among patients who died in acute care hospitals (24%) and patients who died in chronic care hospitals (22%) [2]. Outpatients, however, had a lower prevalence of PE at autopsy (5%) [2]. The prevalence of PE at autopsy of patients in general hospitals and in entire communities, with one exception, ranged from 9 to 28% and has not changed in over 60 years [2–21] (Table 1.1). One study, however, reported gross PE in 55% of patients at autopsy [10]. On average, PE at autopsy occurred in 7031 of 55,090 patients (13%) (Table 1.1, Figure 1.1).
Small PE at autopsy In an autopsy study that employed postmortem pulmonary arteriography as well as gross dissection and microscopic examination, gross dissection showed PE in 34 of 225 (15%) of autopsied patients [4]. Among these, PE was limited to muscular pulmonary artery branches (0.1–1 mm diameter) in 26 of 34 patients (76%) and PE was in elastic pulmonary artery branches (>1 mm diameter) in 8 of 34 patients (24%) [4]. Microscopic examination showed PE in pulmonary arterioles in 13 of 34 patients (38%) with grossly visible PE. The smallest PE that have been identified in living patients were with wedge pulmonary arteriography, which showed PE in 1–2-mm-diameter pulmonary artery branches [24] (see Chapter 71). Fibrous bands, webs, and intimal fibrosis have been interpreted as the final state of organization of PE and these have been reported by some to indicate old PE at autopsy [7]. Meticulous dissection and microscopic examination for minute and barely visible fragments showed traces of fresh or old PE at autopsy in 52% and 64% of patients [7, 8].
Large or fatal PE at autopsy Large or fatal PE in patients at autopsy in general hospitals or communities from 1939 to 2000 occurred in 2264 of 54,364 patients (4%) (range 0.3–24%) [2, 3, 8–11, 13–19, 21, 22] (Table 1.1, Figure 1.1). In most studies, the prevalence of large or fatal PE ranged from 3 to 10%. In elderly institutionalized patients, the rate of fatal PE at autopsy was within that range, 18 of 234 (8%) [23]. Data on institutionalized patients are not included in Table 1.1. A sudden increase in the rate of PE at autopsy was observed in London in 1940 due to cramped conditions in air raid shelters [22]. These rates also are not included in Table 1.1.
Unsuspected PE at autopsy Pulmonary embolism was unsuspected or undiagnosed antemortem in 3268 of 3876 patients in general hospitals or communities who had PE at autopsy (84%) (range 80–93%) [3, 5, 8, 11, 12, 16, 18] (Table 1.2, Figure 1.2). Remarkably, even in patients with large or fatal PE at autopsy, the majority, 1902 of 2448 (78%), were unsuspected or undiagnosed antemortem [2, 11, 12, 14–16, 18, 19, 25] (Table 1.2, Figure 1.2). In our experience, PE at autopsy caused death in 5%, contributed to death in 0.5%, and was incidental in 9.2% of 404 autopsies, and the distribution, according
3
BLUK077-Stein
March 12, 2007
14:45
4
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 1.1 Prevalence of pulmonary embolism at autopsy in general hospitals and communities. Any PE/No autopsies (%)
Fatal or large PE/No autopsies (%)
Study years
First author, year [Ref] Simpson, 1940 [22]
4/242 (2)
1939
606/4391 (14)
—
1945–1954
Coon, 1959 [3]
34/225 (15)
—
1960–1961
Smith, 1964 [4]
118/981 (12)
—
1956–1960
Uhland, 1964 [5]
17/61 (28)*
—
1951–1959
Freiman, 1965 [6]
55/263 (21)†
—
1964–1965
Morrell, 1968 [7]
567/4600 (12)
202/4600 (4)
1964–1974
Coon, 1976 [8]
—
319/1350 (24)
1976
Schwarz, 1976 [9]
280/508 (55)‡
92/508 (18)
1969–1970
Havig, 1977 [10]
216/1455 (15)
54/1455 (4)
1973–1974
Goldhaber, 1982 [11]
389/2398 (16)
—
1966–1976
Dismuke, 1984 [12]
—
105/1133 (9)
1966–1970
Dismuke, 1986 [13]
—
53/1124 (5)
1971–1975
””
—
43/1128 (4)
1976–1980
””
—
44/1276 (3)
1980–1984
313/2388 (13)
239/2388 (10)
1979–1983
Rubenstein, 1988 [14] Sandler, 1989 [15]
1934/21,529 (9)
67/21,529 (0.3)
1960–1984
Karwinski, 1989 [16] Linblad, 1991 [17]
161/766 (21)
68/766 (9)
1957
250/1117 (22)
93/1117 (8)
1964
””
346/1412 (25)
83/1412 (6)
1975
””
260/994 (26)
93/994 (9)
1987
59/404 (15)
20/404 (5)
1985–1986
Stein, 1995 [18]
””
—
92/2427 (4)
1985–1989
Morgenthaler, 1995 [19]
288/3334 (9)§
—
1966–1974
Mandelli, 1997 [20]
182/1144 (16)§
—
1989–1994
431/2356 (18)
178/2356 (8)
1987
Nordstrom, 1998 [2]
525/3764 (14)
221/3764 (6)
1980–2000
Pheby, 2002 [21]
””
* An additional 22/61 (36%) showed traces of residual pulmonary embolism (PE), fibrous bands, or webs. †
An additional 31% had had fibrous bands or intimal fibrosis indicative of old PE. An additional 72 of 508 (14%) were visible only by microscopy. § Massive and submassive PE.
PE, DVT at autopsy (%)
‡
50 45 40 35 30 25 20 15 10 5 0
43
13 4
Any PE
Large or fatal PE
Any DVT
Figure 1.1 Prevalence of pulmonary embolism (PE) and deep venous thrombosis (DVT) at autopsy.
BLUK077-Stein
March 12, 2007
CHAPTER 1
14:45
5
PE and DVT autopsy
Table 1.2 Unsuspected pulmonary embolism at autopsy. Any unsuspected or
Unsuspected or
undiagnosed PE
undiagnosed minor or
Unsuspected or undiagnosed fatal or large PE
[unsuspected PE/total
small PE [unsuspected
[unsuspected large PE/total
PE (%)]
small PE/total PE (%)]
PE (%)]
Study years
First author, year [Ref]
563/606 (93)
—
—
1945–1954
Coon, 1959 [3]
91/107 (85)
—
—
1955–1960
Uhland, 1964 [5]
514/567 (91)
—
—
1964– 1974
Coon, 1976 [8]
199/217 (92)
161/162 (99)
38/54 (70)
1973–1974
Goldhaber, 1982 [11]
310/389 (80)
219/244 (90)
91/145 (63)
1966–1976
Dismuke, 1984 [12]
—
—
30/44 (68)
1980–1984
Rubenstein, 1988 [14]
—
—
186/195 (95)
1979–1983
Sandler, 1989 [15]
1619/1934 (84)
436/484 (90)
1183/1450 (82)
1960–1984
Karwinski, 1989 [16] Stein, 1995 [18]
52/59 (88)
36/37 (97)
14/20 (70)
1985–1986
—
—
47/92 (51)
1985–1989
Morgenthaler, 1995 [19]
—
—
189/279 (68)
1987
Nordstrom, 1998 [2]
—
—
124/169 (73)
1995–2002
Attems, 2004* [25]
* All patients ≥70 years old. PE, pulmonary embolism.
to whether diagnosed and treated, suspected but not diagnosed or treated, or unsuspected is shown in Table 1.3 [18]. Many patients with unsuspected large or fatal PE had advanced associated disease [18]. Patients who suffer sudden and unexplained catastrophic events in the hospital are a group in whom the diagnosis might be suspected more frequently if physicians maintain a high index of suspicion [18].
Rate and sequence of organization of thromboemboli
Unsuspected PE/total PE (%)
A thrombus contains extensive regions of masses of agglutinated platelets [26]. Platelets are deposited first,
Figure 1.2 Prevalence of unsuspected pulmonary embolism (PE) at autopsy.
90
followed by leukocytes, followed after a variable period of time by fibrin with trapped red cells and a few scattered leukocytes [26]. The rate of organization of thromboemboli has been assessed in rabbits [27, 28]. The following results were shown [27, 28]: 8 minutes. Thrombus covered by an eosinophilic rim of platelets. Small amounts of fibrin were interspersed among the platelets at the edge of the thrombus [28]. 3 days. Thrombi contained masses of red cells, fibrin, platelets, and white cells together with a number of macrophages. Parts of the surface not in contact with the vessel wall were covered by flattened cells and in places these were buttressed by a layer of elongated cells beneath. Platelets were particularly
84 78
75 60 45 30 15 0
Patients with any unsuspected PE
Patients with unsuspected large or fatal PE
BLUK077-Stein
March 12, 2007
14:45
6
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 1.3 Autopsy patients ≥18 years old (n = 404). PE caused death (%)
PE contributed to death (%)
PE incidental (%)
PE total (%)
Diagnosed and treated
3 (0.7)
0 (0)
1 (0.2)
4(1.0)
Suspected but not
3 (0.7)
0 (0)
0 (0)
3 (0.7)
Unsuspected
14 (3.5)
2 (0.5)
36 (8.9)
52(12.9)
Total
20 (5.0)
2 (0.5)
37 (9.2)
59 (14.6)
diagnosed or treated
Modified from Stein and Henry [18] and reproduced with permission. PE, pulmonary embolism.
Table 1.4 Deep venous thromboses; autopsies with full limb dissection. DVT n/N (%)
Site (number of thrombi)
95/324 (29)
Site (number of patients)
First author, year [Ref]
Thighs or pelvis 7
Rossle, 1937 [29]
Thighs and Calves 38 Calves only 50 100/165 (61)
Thighs 22
Neumann, 1938 [30]
Calves 87 Ankle 17 Foot 71 88/200 (44)
Thighs only 3
Hunter, 1945 [31]
Thighs and Calves 28 Calves only 57 35/130 (27) 32/100 (32)*
Raeburn, 1951 [32] Thighs only 18
McLachin, 1962 [33]
Thighs and Calves 10 Calves only 4 149/253 (59)
Thighs only 24
Gibbs, 1957 [34]
Thighs and Calves 39 Calves only 86 13/27 (48)
IVC 1
Thighs only 1
Pelvic 1
Thighs and Calves 7
Thigh 23
Calves only 5
Stein, 1967 [35]
Calves 35† 540/1350 (40)
Pelvic 41‡
Schwarz, 1976 [9]
Thigh 21 Calves 74 161/261 (62)
IVC 8 Pelvic 31 Thigh 129 Calves 128 Foot 87
* Males >40 years old. †
Calf 11 microscopic thrombi in addition. Sample of 37 patients. DVT, deep venous thrombosis; n, number of patients with DVT; N, number of patients necropsied. ‡
Havig, 1977 [10]
BLUK077-Stein
March 12, 2007
CHAPTER 1
14:45
PE and DVT autopsy
7
Figure 1.3 Extensive antemortem thrombus located in popliteal and calf veins. Previously unpublished figure from Stein and Evans [35].
prevalent near the thrombus–vessel wall junction. Mononuclear cells were prominent [27]. 5 and 7 days. Beginnings of vascularization were apparent. Capillaries were within the thrombus mass and in cellular areas of attachment to the intima. The central area of the thrombus showed mainly debris [27]. 7 days. Occluding thrombi had retracted in places and were covered by flattened cells, and showed one or more firm cellular attachments to the intima. Macrophages were conspicuous and contained lipid, fibrin, and cellular debris together with fibroblastic cells [27]. 14 days. Thrombi consisted of cellular masses containing small clumps of fibrin and variable amounts of fat and fibrous tissue [27]. 20 days. Some thrombi appeared as polypoid masses protruding into the lumen and containing variable amounts of fat, fibrous, and elastic tissue, and on occasion calcium, while others showed lipid within foamy cells and a fibrous tissue cap containing fibroblasts, collagen, and elastic tissue [27]. 30 days. Thromboemboli were converted to eccentric fibrofatty thickenings of the intima [27].
Deep venous thrombosis at autopsy Data on patients who had complete dissection of the lower extremities at autopsy are from prior decades,
and before the general use of antithrombotic prophylaxis [10, 29–36]. Among patients at autopsy who had full limb dissection, 1213 of 2810 (43%) showed deep venous thrombosis (DVT) [10, 29–36] (Table 1.4, Figures 1.1 and 1.3). Among 161 patients with DVT at autopsy, 7 patients had thrombi in the common iliac vein and 22 had thrombi in the external iliac vein. Each of these patients also showed DVT in the femoral vein [10]. The external iliac vein showed thrombi in 12 of 161 patients (7%) without femoral vein involvement. In 4 of these patients, the calf veins showed DVT, but not the femoral veins [10]. Deep venous thrombosis affected the veins of the calves more frequently than the veins of the thighs, and both were more frequently affected than the veins of the pelvis. The distribution of 601 thrombi found in 311 patients who had dissection of the pelvic, thigh, and calf veins was 54% in the veins of the calves, 32% in the veins of the thighs, 12% in the pelvic veins, and 1% in the inferior vena cava [10, 30, 35, 36] (Figure 1.4). The distribution of 563 thrombi among 261 necropsied patients who had dissection of the veins of the foot as well as the veins of the calf, thigh, and pelvis was 28% in the veins of the foot, 38% in the calf, 27% in the thigh, 6% in the pelvic veins, and 1% in the inferior vena cava (IVC) [10, 30] (Figure 1.5). Among 282 necropsied patients who had complete dissection of the veins of the thighs and veins of the calves, the thrombi were located only in the veins of the calves in 54% of patients [31, 33–35] (Figure 1.6).
BLUK077-Stein
March 12, 2007
14:45
60
54
Thrombi (%)
50 40 32 30 20 12 10 0
1 IVC
Pelvis
Calf
Thigh
40
Figure 1.4 Distribution of deep venous thrombosis among patients at autopsy in whom pelvic, thigh, and calf veins were dissected.
38
35 Thrombi (%)
30
28
27
25 20 15 10 6 5 1 0 IVC
Pelvis
Thigh
Calf
Foot
Figure 1.5 Distribution of deep venous thrombosis among patients at autopsy in whom veins of the foot as well as pelvic, thigh, and calf veins were dissected.
Patients (%)
60
54
50 40 30
30 20
16
10 0 Thigh only
Thighs and calves
Calves only
Figure 1.6 Percentage of patients at autopsy with deep venous thrombosis who had involvement of veins of thigh only, veins of thighs and calf veins, and veins of calf only.
Figure 1.7 Normal postmortem venogram of calf (lateral projection) showing anterior tibial (AT), posterior tibial (PT), and peroneal (Pe) veins. The deep veins are paired. (Reproduced from Stein and Evans [35], with permission.)
8
BLUK077-Stein
March 12, 2007
CHAPTER 1
14:45
PE and DVT autopsy
9
Both the veins of the thighs and calves were affected in 30% of patients. Only the veins of the thighs showed DVT in 16% of patients. Bilateral DVT was observed in 81 of 96 patients (84%) with extensive DVT at autopsy and in 26 of 65 (40%) of patients with minor DVT at autopsy [10]. Postmortem venography illustrates the extent and location of DVT at autopsy in unselected patients [35]. For comparison, normal postmortem venograms of the calf and thighs are shown (Figures 1.7 and 1.8). Postmortem venograms of DVT involving the veins of the thighs are shown in Figures 1.9 and 1.10. Figure 1.8 Normal postmortem venogram of the thighs (anteroposterior projection) showing the femoral (F), deep femoral (DF), greater saphenous (GS), and popliteal (P) veins. Valve pockets are shown. (Reproduced from Stein and Evans [35], with permission.)
Figure 1.9 Postmortem venogram of the veins of both thighs. Extensive thrombosis of the femoral, deep femoral, and popliteal veins was found by dissection of the left thigh. The venogram of the left thigh shows absence of filling of the popliteal and deep femoral veins and only a faint outline of the femoral vein (F). The left greater saphenous vein is dilated and joined by numerous collateral vessels. The veins of the right thigh were normal. (Reproduced from Stein and Evans [35], with permission.)
Figure 1.10 Postmortem venogram of right thigh. The femoral vein has not filled with contrast material because of a completely occluding thrombus. The greater saphenous (GS) vein is distended. Collateral vessels formed at the site of an occluding thrombus in the greater saphenous vein (arrow). (Reproduced from Stein and Evans [35], with permission.)
BLUK077-Stein
March 12, 2007
14:45
10
PART I
Prevalence, risks, and prognosis of PE and DVT
Figure 1.11 Organized thrombus in anterior tibial vein (same patient as Figure 1.12). This thrombus is older than the thrombus in the femoral vein, and there is no phlebitis here. Hematoxylin and eosin ×40. (Previously unpublished figure from Stein and Evans [35].)
Forward thrombosis versus retrograde thrombosis
Thrombophlebitis and phlebothrombosis
In every case that we examined in which the veins of the thigh and the calf showed DVT in continuity, the thrombi in the calf were older than those in the thigh [35] (Figures 1.11 and 1.12). This supports the concept that forward thrombosis is more common than retrograde thrombosis.
The terms “thrombophlebitis” and “phlebothrombosis” in prior years were used to distinguish between DVT associated with inflammation (thrombophlebitis) and DVT not associated with inflammation (phlebopthrombosis). These are outdated terms. Histological investigations have not supported a distinction between the clinical diagnoses of thrombophlebitis and phlebothrombosis. Thrombosis of the veins of the lower extremities usually occurs without inflammation [35] (Figures 1.11 and 1.14–1.16). Inflammation of the walls of the veins, when it occurs (Figure 1.12), is usually secondary to the thrombosis [35]. No clear evidence indicates that inflammation
Collateral veins around occlusions Clinically unsuspected DVT at autopsy was often extensive, causing collateral circulation around occlusions and dilatation of collateral veins [35] (Figures 1.10 and 1.13).
Figure 1.12 Thrombus attached to femoral vein (same patient as Figure 1.11). Lymphocytic infiltrate is shown throughout the wall of the vein. The patient had signs and symptoms of deep venous thrombosis. Hematoxylin and eosin ×13. (Previously unpublished figure from Stein and Evans [35].)
BLUK077-Stein
March 12, 2007
CHAPTER 1
14:45
PE and DVT autopsy
Figure 1.13 Postmortem venogram of the thighs. There is definite radiographic evidence of occlusion of the femoral vein between points 1 and 2. There is no filling of the femoral vein (F) between these points. Dilated and tortuous collaterals pass around the site of occlusion. No thrombus was found on dissection of the veins of the thigh of this patient, presumably because dissection was carried out along the collateral vessels in this area rather than the femoral vein. (This apparent femoral vein occlusion was not included among the positive cases reported in Stein and Evans [35].)
Figure 1.14 Recent thrombus attached in vein of soleal plexus. Hematoxylin and eosin ×16. (Previously unpublished figure from Stein and Evans [35].)
11
of the veins prevents embolization, or that embolization is more frequent in those patients with thrombi not associated with venous inflammation. The distinction between “thrombophlebitis” and “phlebothrombosis” is of no clinical consequence [35]. A thrombus can induce inflammation in the underlying wall of the vein, and this inflammation in some patients is extensive enough to produce pain, tenderness, swelling, and fever compatible with the clinical diagnosis of thrombophlebitis [36]. However, the underlying pathogenic mechanism is primary thrombosis and not primary phlebitis [36]. The following historical background explains the evolution of these outdated diagnostic terms. John Hunter, after studying infected venesections in human beings and in horses, attributed the thrombosis to phlebitis [37]. Virchow, however, observed that the cellular reaction in the wall of the vein usually does not occur until after the thrombus has been laid down [38]. Welch [39], in studying DVT in patients with infectious diseases such as typhoid fever, found an inflammatory lesion beneath the endothelium in which he could not demonstrate any organisms. He termed this “toxic endophlebitis” and attributed some instances of DVT to inflammation of the veins. Subsequently, patients were described who had clinical evidence of thrombosed leg veins and also had clinical signs of inflammation (warmth, redness, tenderness). A diagnosis of thrombophlebitis was made. In view of Welch’s observations, it was concluded that the primary event was inflammation of the wall of the vein. In contrast, asymptomatic patients were later described who had thrombosis of
BLUK077-Stein
12
March 12, 2007
14:45
PART I
Prevalence, risks, and prognosis of PE and DVT
Figure 1.15 Fresh unattached thrombus in fomoral vein. Lines of Zahn distinguish this from postmortem clot. Hematoxylin and eosin ×4. (Previously unpublished figure from Stein and Evans [35].)
Figure 1.16 Photomicrograph showing thrombus originating in valve pocket of a posterior tibial vein. The well-organized fibrous point of attachment is capped by a fresh red cell, platelet, and fibrin clot. There is no inflammation of the vein. Hematoxylin and eosin ×4. (Previously unpublished figure from Stein and Evans [35].)
Figure 1.17 Thrombus attached to valve pocket in femoral vein and propagating along the vein. Venous valve is shown (arrow). Hematoxylin and eosin ×10. (Previously unpublished figure from Stein and Evans [35].)
BLUK077-Stein
March 12, 2007
CHAPTER 1
14:45
PE and DVT autopsy
13
Figure 1.18 Section of left posterior tibial vein. An antemortem thrombus, 0.2 cm in largest dimension, is located within a valve pocket. (Previously unpublished figure from Stein and Evans [35].)
the lower extremities that resulted in PE [40]. These patients, because of the lack of leg signs, were said to have phlebothrombosis. Although there are situations in which phlebitis is primary and thrombosis is secondary (such as mechanical and chemical injury) [36], these are rarely compared with the incidence of thrombosis without inflammation [31, 36]. In patients with DVT at autopsy, fresh components of the thrombus as well as older components were shown, indicating that the thrombosis was continuing [35] (Figure 1.16). None of the patients were diagnosed antemortem as having DVT. A patient with clinical signs and symptoms of DVT showed lymphocytic infiltration in the media of the veins (Figure 1.12). The inflammation occurred not only at the sites of attachment of the thrombus, but also where the thrombus was apposed to the endothelium without being attached, suggesting that the thrombus induced the inflammation.
Valve pockets as site of origin of DVT The valve pockets were a frequent site of origin of thrombi (Figures 1.16–1.18). Thrombi located in valve pockets consisted of organized fibrous points of attachment capped by fresh fibrin and red cell clot [35] (Figure 1.16). Dilated veins and enlarged valve pockets were frequently seen (Figure 1.19). There was no correlation of either of these abnormalities with the presence of thrombosis [35].
Figure 1.19 Postmortem venogram showing dilated valve pocket in femoral (F) vein of left thigh (arrow). The deep femoral vein (DF) is also shown. (Reproduced from Stein and Evans [35], with permission.)
BLUK077-Stein
March 12, 2007
14:45
14
PART I
References
18 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 19 Morgenthaler TI, Ryu JH. Clinical characteristics of fatal pulmonary embolism in a referral hospital. Mayo Clin Proc 1995; 70: 417–424. 20 Mandelli V, Schmid C, Zogno C, Morpurgo M. “False negatives” and “false positives” in acute pulmonary embolism: a clinical–postmortem comparison. Cardiologia 1997; 42: 205–210. 21 Pheby DF, Codling BW. Pulmonary embolism at autopsy in a normal population: implications for air travel fatalities. Aviat Space Environ Med 2002; 73: 1208–1214. 22 Simpson K. Shelter deaths from pulmonary embolism. Lancet 1940; 2: 744. 23 Gross JS, Neufeld RR, Libow LS, Gerber I, Rodstein M. Autopsy study of the elderly institutionalized patient. Review of 234 autopsies. Arch Intern Med 1988; 148: 173–176. 24 Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623. 25 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154; 159–162. 26 Poole JC, French JE, Cliff WJ. The early stages of thrombosis. J Clin Pathol 1963; 16: 523–528. 27 Still WJ. An electron microscopic study of the organization of experimental thromboemboli in the rabbit. Lab Invest 1966; 15: 1492–1507. 28 Thomas DP, Gurewich V, Ashford TP. Platelet adherence to thromboemboli in relation to the pathogenesis and treatment of pulmonary embolism. N Engl J Med 1966; 274: 953–956. 29 Rossle R. Uber die Bedeutung und die entstehung der wadenvenenthrombosen. Virchow Arch Path Anat 1937; 300: 180–189. 30 Neumann R. Ursprungszentren und entwicklungsformen der bein-thrombose. Virchow Arch Path Anat 1938; 301: 708–735. 31 Hunter WC, Krygier JJ, Kennedy JC, Sneeden VD. Etiology and prevention of thrombosis of the deep leg veins: a study of 400 cases. Surgery 1945; 17: 178–190. 32 Raeburn C. The natural history of venous thrombosis. BMJ 1951; 2: 517–520. 33 McLachlin J, Richards T, Paterson JC. An evaluation of clinical signs in the diagnosis of venous thrombosis. Arch Surg 1962; 85: 738–744. 34 Gibbs NM. Venous thrombosis of the lower limbs with particular reference to bed-rest. Br J Surg 1957; 45: 209– 236.
1 Dalen JE, Alpert JS. Natural history of pulmonary embolism. Prog Cardiovas Dis 1975; 17: 259–270. 2 Nordstrom M, Lindblad B. Autopsy-verified venous thromboembolism within a defined urban population— the city of Malmo, Sweden. Acta Path Microbiol Immunol Scand 1998; 106: 378–384. 3 Coon WW, Coller FA. Clinicopathologic correlation in thromboembolism. Surg Gynecol Obstet 1959; 109: 259– 269. 4 Smith GT, Dammin GJ, Dexter L. Postmortem arteriographic studies of the human lung in pulmonary embolization. JAMA 1964; 188: 143–151. 5 Uhland H, Goldberg LM. Pulmonary embolism: a commonly missed clinical entity. Dis Chest 1964; 45: 533–536. 6 Freiman DG, Suyemoto J, Wessler S. Frequency of pulmonary thromboembolism in man. N Engl J Med 1965; 272: 1278–1280. 7 Morrell MT, Dunnill MS. The post-mortem incidence of pulmonary embolism in a hospital population. Br J Surg 1968; 55: 347–352. 8 Coon WW. The spectrum of pulmonary embolism: twenty years later. Arch Surg 1976; 111: 398–402. 9 Schwarz N, Feigl W, Neuwirth E, Holzner JH. Venous thromboses and pulmonary emboli in autopsy material. Wien Klin Wochenschr 1976; 88: 423–428. 10 Havig O. Deep venous thrombosis and pulmonary embolism. Chapters 2, 4: Pulmonary thromboembolism. Acta Chir Scand 1977; 478(suppl): 4–11, 24–37. 11 Goldhaber SZ, Hennekens CH, Evans DA, Newton EC, Godleski JJ. Factors associated with correct antemortem diagnosis of major pulmonary embolism. Am J Med 1982; 73: 822–826. 12 Dismuke SE, VanderZwaag R. Accuracy and epidemiological implications of the death certificate diagnosis of pulmonary embolism. J Chronic Dis 1984; 37: 67–73. 13 Dismuke SE, Wagner EH . Pulmonary embolism as a cause of death. The changing mortality in hospitalized patients. JAMA 1986; 255: 2039–2042. 14 Rubenstein I, Murray D, Hoffstein V. Fatal pulmonary emboli in hospitalized patients—an autopsy study. Arch Int Med 1988; 148: 1425–1426. 15 Sandler DA, Martin JF. Autopsy proven pulmonary embolism in hospital patients: are we detecting enough deep vein thrombosis? J R Soc Med 1989; 82: 203–205. 16 Karwinski B, Svendsen E. Comparison of clinical and post-mortem diagnosis of pulmonary embolism. J Clin Pathol 1989; 42: 135–139. 17 Lindblad B, Sternby NH, Bergqvist D. Incidence of venous thromboembolism verified by necropsy over 30 years. BMJ 1991; 302: 709–711.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 12, 2007
CHAPTER 1
14:45
PE and DVT autopsy
35 Stein PD, Evans H. An autopsy study of leg vein thrombosis. Circulation 1967; 35: 671–681. 36 Allen EV, Barker NW, Hines EA, Jr. Peripheral Vascular Diseases. WB Saunders, Philadelphia, 1962: 559– 569. 37 Hunter J. Observation on the inflammation of the internal coats of veins. Trans Soc Imp Med Chir Knowl 1793; 1: 18. Quoted by Stein and Evans in Reference [35].
15
38 Virchow R. Cellular Pathology as Based Upon Physiological and Pathological Histology. J. & A. Churchill, Ltd., London, 1860: 197–203. Quoted from Gibbs NM in Reference [34]. 39 Welch WH. Thrombosis. In: Allbutt TC, ed. A System of Medicine, Vol. 6. Macmillan, New York, 1899: 180. 40 Homans J. Thrombosis of the deep veins of the lower leg causing pulmonary embolism. New Engl J Med 1934; 211: 993.
BLUK077-Stein
2
March 12, 2007
14:48
CHAPTER 2
Incidence of pulmonary embolism and deep venous thrombosis in hospitalized patients
Introduction In 1999, 140,000 patients were discharged from nonfederal short-stay hospitals in the United States with a diagnosis of pulmonary embolism (PE) and 370,000 patients were discharged with deep venous thrombosis (DVT) [1]. In 2001, the number of patients diagnosed with PE increased to 156,000, and the number discharged with DVT remained the same [2]. Either PE or DVT was diagnosed in 466,000 hospitalized patients [2]. Throughout the entire United States, from 1979 to 2001, the number of patients discharged from short-stay nonfederal hospitals with PE was 2,741,000, with DVT 6,475,000, and with either venous thromboembolism (VTE) (either PE or DVT) 8,575,000 [3]. During this 23-year period, the average populationbased rate of diagnosis of PE in hospitalized patients was 47/100,000 population, the population-based rate of diagnosis of DVT was 112/100,000 population, and for VTE it was 148/100,000 population [2].
Incidence of PE in hospitals The National Hospital Discharge Survey showed the prevalence of PE in patients ≥20 years of age averaged over a 21-year period of study from 1979 to 1999 of 0.40% [4] (Figure 2.1). During this period there were 612,000,000 short-stay nonfederal hospital admissions throughout the United States [4]. These data were based entirely on discharge codes. The results were comparable to the incidence of PE in hospitalized patients as was shown by much smaller but more rigidly defined retrospective evaluations (0.27–0.4%) [5–8]. (National Hospital Discharge Survey. Available at: http://www.cdc.gov/nchs/about/major/hdasd/nhds .htm.)
16
The incidence of PE in hospitalized patients did not change over 21 years [4] (Figure 2.2). The incidence of PE in hospitalized patients was nearly the same in men and women (relative risk of men to women 1.11) [4] (Figure 2.3). The incidence of PE in hospitalized patients was the same in white and black patients (relative risk of white patients to black patients 1.00) [4] (Figure 2.4). The prevalence of PE in a general hospital, based on clinical diagnoses, many of which were confirmed at autopsy, in an era prior to pulmonary angiography or ventilation–perfusion scans was 0.2% [9]. The prevalence of acute PE in patients in a clinic of digestive surgery, diagnosed by pulmonary angiography, high probability V-Q scans or autopsy was 0.3% [10]. Using comparable criteria, we found the same prevalence (0.3%) [8]. The inclusion of patients estimated to have PE based on non-high-probability interpretations of the ventilation–perfusion lung scans and the inclusion of patients with clinically undiagnosed PE at autopsy caused the estimated prevalence of PE to be higher, 1.0% [5]. There are, in addition, patients with silent PE, the frequency of which is undetermined. On average, PE occurs in 13% of patients at autopsy, among whom the diagnosis was unsuspected antemortem in 84% (Chapter 1).
Incidence of DVT in hospitals Based on results of the National Hospital Discharge Survey, the prevalence of DVT in patients ≥20 years of age averaged over the 21-year period of study was 0.93% [4] (Figure 2.1). Venous thromboembolic disease (VTE), defined as PE and/or DVT, occurred in 1.24% [4] (Figure 2.1). The incidence of DVT in
BLUK077-Stein
March 12, 2007
CHAPTER 2
14:48
17
PE and DVT in hospitalized patients
0.5 0.40
PE
VTE
2
VTE
1.5 DVT 1 0.5
PE
0 99
97
95
93
91
89
87
85
83
81
VTE, DVT, PE in hospitalized adults (%)
DVT
1.8 1.6
Male
1.4
DVT
1.2 Female
1 0.8
Male
0.6
PE
0.4
Female
0.2 99
97
95
93
91
89
87
85
83
81
PE and DVT in hospitalized adults according to sex (%)
Year
Year
2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2
White patients Black patients White patients Black patients
99
97
95
93
Year
91
89
87
85
83
81
79
Figure 2.4 Incidences of pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized adults from 1979 to 1999 according to race. The incidences in black and white patients were the same for PE and nearly the same for DVT. (Reprinted from Stein et al. [4], with permission from Elsevier.)
0.93
79
Figure 2.3 Incidences of pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized adults from 1979 to 1999 according to sex. The incidences in men and women were nearly the same. (Reprinted from Stein et al. [4], with permission from Elsevier.)
1.24
1
79
Figure 2.2 Incidences of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in hospitalized adults from 1979 to 1999. The incidence of DVT increased (slope = 0.028%/year, r = 0.92, P < 0.0005). The incidence of PE did not change. The incidence of VTE increased in parallel to the incidence of DVT. (Reprinted from Stein et al. [4], with permission from Elsevier.)
1.5
0
PE and DVT in hospitalized adults according to race (%)
Figure 2.1 Prevalence of pulmonary embolism (PE), deep venous thrombosis (DVT) and either PE or DVT, venous thromboembolism (VTE) in hospitalized adults (≥20 years). (Data from Stein et al. [4].)
PE and DVT in hospitalized adults (%)
2
BLUK077-Stein
March 12, 2007
14:48
18
PART I
hospitalized patients increased from 239,000 of 30,771,000 (0.8%) in 1979 to 363,000 of 28,504,000 (1.3%) in 1999 and the incidence of VTE increased in parallel (Figure 2.2). The incidence of DVT in hospitalized patients was nearly the same in men and women (relative risk of men to women 1.05) [4] (Figure 2.3). The incidence of DVT in hospitalized patients was nearly the same for white and black patients (relative risk 1.05) [4] (Figure 2.4). The increasing incidence of DVT in hospitalized patients from 1979 to 1999 may represent an increasing availability and use of venous ultrasound during much of that period [1]. Early diagnosis and treatment of DVT may have prevented a parallel increase in the incidence of PE in hospitalized patients. Whether the trend toward an increasing incidence of DVT in hospitalized patients will continue is uncertain, particularly in view of outpatient treatment of DVT, which was introduced in 1996 and 1997 [11–13]. The reported incidence of DVT in hospitals, 0.1– 3.17%, ranged more widely than the incidence of PE [14–17]. In Asian hospitals the prevalence of DVT was lower [18, 19], but VTE has been reported to be uncommon among Asians [20, 21] (see Chapter 15).
7 Proctor MC, Greenfield LJ. Pulmonary embolism: diagnosis, incidence and implications. Cardiovasc Surg 1997; 5: 77–81. 8 Stein PD, Patel KC, Kalra NJ et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. 9 Hermann RE, Davis JH, Holden WD. Pulmonary embolism. A clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961; 102: 19–28. 10 Huber O, Bounameaux H, Borst F et al. Postoperative pulmonary embolism after hospital discharge: an underestimated risk. Arch Surg 1992; 127: 310–313. 11 The Columbus Investigators. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med 1997; 337: 657–662. 12 Levine M, Gent M, Hirsh J et al. A comparison of low-molecular-weight heparin administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis. N Engl J Med 1996; 334: 677–681. 13 Koopman MMW, Prandoni P, Piovella F et al. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecular-weight heparin administered at home. N Engl J Med 1996; 334: 682–687. 14 Klatsky AL, Armstrong MA, Poggi J. Risk of pulmonary embolism and/or deep venous thrombosis in AsianAmericans. Am J Cardiol 2000; 85: 1334–1337. 15 Igbinovia A, Malik GM, Grillo IA et al. Deep venous thrombosis in Assir region of Saudi Arabia. Case–control study. Angiology 195; 46: 1107–1113. 16 Schuurman B, den Heijer M, Nijs AM. Thrombosis prophylaxis in hospitalized medical patients: does prophylaxis in all patients make sense? Neth J Med 2000; 56: 171–176. 17 Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. 18 Liam CK, Ng SC. A Review of patients with deep vein thrombosis diagnosed at university hospital, Kuala Lumpur. Ann Acad Med Singapore 1990; 19: 837–840. 19 Kueh YK, Wang TL, Teo CP et al. Acute deep vein thrombosis in hospital practice. Ann Acad Med Singapore 1992; 21: 345–348. 20 Stein PD, Kayali F, Olson RE, Milford CE. Pulmonary thromboembolism in Asian/Pacific Islanders in the United States: analysis of data from the National Hospital Discharge Survey and the United States Bureau of the Census. Am J Med 2004; 116: 435–442. 21 White RH, Zhou H, Romano PS. Incidence of idiopathic deep venous thrombosis and secondary thromboembolism among ethnic groups in California. Ann Intern Med 1998; 128: 737–740.
References 1 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diagnosing deep venous thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 2 Unpublished data from Stein PD, Kayali F, Olson RE. Regional differences in rates of diagnosis and mortality of pulmonary thomboembolism. Am J Cardiol 2004; 93: 1194–1197. 3 Stein PD, Kayali F, Olson RE. Regional differences in rates of diagnosis and mortality of pulmonary thomboembolism. Am J Cardiol 2004; 93: 1194–1197. 4 Stein PD, Beemath A, Olson RE. Trends in the incidence of pulmonary embolism and deep venous thrombosis in hospitalized patients. Am J Cardiol 2005; 95: 1525– 1526. 5 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 6 Stein PD, Huang H, Afzal A et al. Incidence of acute pulmonary embolism in a general hospital: relation to age, sex and race. Chest 1999; 116: 909–913.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 28, 2007
3
15:2
CHAPTER 3
Case fatality rate and population mortality rate from pulmonary embolism and deep venous thrombosis
Overview
Untreated patients
Among all patients with pulmonary embolism (PE) throughout the United States, irrespective of treatment or the severity of the PE, the estimated case fatality rate (death from PE/100 patients with PE) in 1998 was 7.7% [1]. The case fatality rate from untreated clinically apparent PE, obtained before anticoagulant therapy was universally used, was 26–37% [2, 3]. In recent years, however, the case fatality rate from mild untreated PE, based on a small number of patients was 5% [4]. The case fatality rate from acute PE in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) was 2.5% [5]. This rate applies to patients who were well enough to be recruited into an investigation of the diagnostic validity of ventilation– perfusion lung scans [5]. Most fatalities from PE occur within the first 2.5 hours after the diagnosis is made [6]. Such patients could not have been included in PIOPED. Case fatality rates from PE in trials of treatment with low-molecular-weight heparin were 0.6– 1.0% [7, 8]. In such trials patients with massive PE requiring thrombolytic therapy, and patients at risk of bleeding, among others, were excluded. The most important factor affecting mortality is shock due to right ventricular failure secondary to PE [9]. In patients with over 50% occlusion of the pulmonary circulation who were in shock, the case fatality rate was 32%, whereas in those with over 50% occlusion of the pulmonary circulation who were not in shock, the case fatality rate was 6% [9].
Regarding untreated patients with acute PE in decades before diagnostic imaging tests were available, Barritt and Jordan reported a 26% mortality from the initial PE [3]. Some of these patients perhaps died from recurrent PE [3]. The diagnosis was made on the basis of clinical features that included evidence of right ventricular failure, pulmonary infarction, or both. Clinical features of pulmonary infarction included pleuritic pain, hemoptysis, pleural friction rub, loss of resonance at the lung base, rales, and the chest radiograph. Features that they relied upon for the detection of acute right ventricular failure were faintness, chest pain, fall of blood pressure, rise of jugular venous pressure, and the electrocardiogram. In 1961, Hermann and associates calculated a 37% case fatality rate from the initial PE [2] (Figure 3.1). The diagnosis was based on clinical features, and autopsy among those who died. The treatment of these patients was not reported, although data were collected between 1943 and 1957 and anticoagulant therapy was not in general used before 1947 [2]. Hermann and associates also reported a 36% frequency of fatal recurrent PE. The total estimated frequency of death that included the original PE and recurrent PE was 73% (Figure 3.1). There was, in addition, a 21% frequency of nonfatal recurrent PE among untreated patients with clinically diagnosed PE (Figure 3.2). Presumably, PE was severe among these patients with apparent clinical features.
19
BLUK077-Stein
March 28, 2007
15:2
20
Untreated fatal PE Initial and recurrent 73%
80 60
Initial 37%
Treated patients in the modern era Case fatality rates (PE deaths/100 PE) have been reported in regional surveys [10], regional estimates [11], registries [12], and prospective investigations [5]
Non-fatal recurrent PE (%)
Untreated PE 40 30
21% 20
5%
10 0
Clinically apparent initial PE
Mild initial PE
Figure 3.2 Nonfatal recurrent pulmonary embolism (PE) among untreated patients diagnosed on the basis of clinical findings (clinically apparent PE) and among patients with mild PE diagnosed by objective tests. (Data from Hermann et al. [2] and Stein and Henry [4].)
1997
We evaluated the 20 patients who received no treatment for PE during the first 3 months of follow-up of PIOPED [4]. Only 1 of these patients (5%) died of PE [4] (Figure 3.1). These 20 patients from PIOPED are described in Chapter 5.
1995
Figure 3.1 Fatal initial pulmonary embolism (PE) and fatal initial and recurrent PE among untreated patients diagnosed on the basis of clinical findings (clinically apparent PE) and among patients with mild PE diagnosed by objective tests. (Data from Hermann et al. [2] and Stein and Henry [4].)
1993
Mild PE
1991
Clinically apparent PE
1989
Clinically apparent PE
0 1987
0
4
1985
Initial & recurrent 5%
20
8
1983
40
12
1981
Mortality (%)
100
Prevalence, risks, and prognosis of PE and DVT
1979
Fatal PE/all PE (%)
PART I
Year Figure 3.3 Estimated case fatality rate for pulmonary embolism (PE) from 1979 to 1998. (Reprinted from Stein et al. [1], with permission from Elsevier.)
where the number of deaths was in the hundreds, and in elderly patients [13] where the number of deaths was several thousand. We calculated case fatality rates of PE from a database with 194,000 PE deaths, based on the entire population of the United States from 1979 to 1998 [1]. The estimated case fatality rate from PE increased from 6.7% in 1979 to 10.5% in 1989 (Figure 3.3). From 1989 to 1998 the estimated case fatality rate decreased to 7.7 PE deaths/100 PE. The estimated case fatality rate in the Minneapolis-St. Paul metropolitan area in 1995 ranged from approximately 2 to 6% depending on age [11]. As in our investigation, these are estimated rates [11]. The case fatality rate in short-stay hospitals in metropolitan Worcester in 1985–1986 (12%) was somewhat higher than what we calculated during those years [10]. The estimated case fatality rate from PE increased with age (Figure 3.4). The relation of case fatality to age was described by an exponential function. The higher case fatality rate with age is concordant with regional investigations [10, 11] and studies in individual hospitals [2, 14]. The estimated case fatality rate from PE over the 20year interval of observation was higher among African Americans than Caucasians (Figure 3.5). The rate ratio of African Americans to Caucasians was 1.43. Others observed a higher case fatality rate among elderly African Americans than elderly Caucasians [13]. The estimated case fatality rate was comparable in men and women (Figure 3.6) as had been shown in regional studies [10, 11]. Among the patients ≥65 years of age, we showed no difference in the case fatality rate between men and women. In an investigation of Medicare patients aged 65–74 years, the case fatality
BLUK077-Stein
March 28, 2007
CHAPTER 3
15:2
21
Case fatality rate and population mortality rate
20
Figure 3.4 Estimated case fatality rates for pulmonary embolism (PE) according to decades of age. The estimated case fatality rates are the average of yearly values over a 20-year period. (Reprinted from Stein et al. [1], with permission from Elsevier.)
Fatal PE/all PE (%)
17.4 15 10.9 10
8.2 3.6
5
6.9
5.5
0 25−34 35−44
45−54 55−64 65−74 75−84
>85
Age group (years)
rate in 1991 of 10.0% in white men and 9.4% in white women [13] was comparable to our estimated case fatality rates of 7.3 and 9.2% in white men and women aged 65 years or older the same year. The case fatality rate in black females aged 65 years or older in 1991 was virtually identical in both studies (11.4 and 11.1%), but we calculated a higher rate in black males (20.9% versus 13.5%).
Massive PE: hypotensive patients Among patients with massive PE, defined as a systolic arterial pressure <90 mm Hg, death from PE within 90 days occurred in 35 of 108 (32%) [15] (Figure 3.7). Among patients with systolic arterial pressure ≥90 mm Hg, death from PE within 90 days occurred in 119 of 2284 (5%).
Treated patients with PE compared with treated patients with deep venous thrombosis (DVT) Pooled data among treated patients with PE and treated patients with DVT showed higher death rates from recurrent PE among the PE patients [16]. Among 2429 patients with DVT treated 5 days to 3 months with anticoagulants, death from PE occurred in 0.3% (Figure 3.8). Among 949 patients with PE treated 5 days to 3 months, death from recurrent PE occurred in 1.4%. The death rates from PE excluded deaths within the first 5 days of diagnosis [16]. Patients treated with thrombolytic agents or inferior vena cava filters were also excluded.
Population mortality rate from PE The number of patients who died from PE in 1998 based on death certificates was 24,947 [17]. With a
Caucasians 1997
1995
1993
1991
1989
1987
1985
1983
1981
Male
8
Female
4 0
1997
1995
1993
1991
1989
1987
1985
1983
1981
Figure 3.5 Case fatality rate from pulmonary embolism (PE) among African Americans and Caucasians. The estimated case fatality rate from PE was higher among African Americans than Caucasians over the 20-year period of observation (P = 0.014). (Reprinted from Stein et al. [1], with permission from Elsevier.)
12
1979
Year
Fatal PE/all PE (%)
African American
20 16 12 8 4 0 1979
Fatal PE/all PE (%)
4.7
Year Figure 3.6 Case fatality rate according to sex. The estimated case fatality rate among males and females was comparable. (Reprinted from Stein et al. [1], with permission from Elsevier.)
BLUK077-Stein
March 28, 2007
15:2
Mortality from PE (%)
22
PART I
40 35 30 25 20 15 10 5 0
32%
5%
PE BP < 90 mm Hg
PE BP > 90 mm Hg
Figure 3.7 Death rates from pulmonary embolism (PE) within 90 days among patients with massive PE, defined as a systolic arterial pressure (BP) <90 mm Hg, and patients with nonmassive PE (systolic arterial pressure ≥90 mm Hg). (Data are from Kucher et al. [15].)
population of 270,299,000 in 1998, this would amount to 9 PE deaths/100,000 population. Assuming that death certificates are only 26.7% accurate for the diagnosis of fatal PE [18], the estimated number of deaths from PE in 1998 may be 93,000 and the death rate may be 34 PE deaths/100,000 population. Over the last two decades the population mortality rate from PE (deaths from PE/100,000 population) decreased [17, 19]. This could be a consequence of a declining incidence of PE (diagnoses of PE/100,000 population) or a declining case fatality rate from PE (deaths from PE/100 cases of PE) or the combination. For the interval 1979–1989, our data showed no decline in the case fatality rate (Figure 3.3), which suggests that
1.4%
Fatal PE (%)
1.5
1
0.5
n = 949 0.3%
n = 2429 0 DVT
PE
Figure 3.8 Pooled data among patients with pulmonary embolism (PE) and patients with deep venous thrombosis (DVT) treated 5 days to 3 months with anticoagulants. The death rate from recurrent PE was higher in patients with PE than the death rate from PE in patients with DVT. (Data are from Douketis et al. [16].)
Prevalence, risks, and prognosis of PE and DVT
the declining population mortality rate from PE was largely due to a declining incidence of PE. From 1979 to 1999, the incidence of diagnosis has, in fact, decreased [20]. Improved prevention of PE may have been a key factor. From 1989 to 1998, the declining population mortality rate appears to be largely due to a declining case fatality rate. The rate of diagnosis of PE no longer declined during this interval [20], but the population mortality rate continued to decline [17]. The estimated case fatality rate also declined during this period (Figure 3.3). The declining population mortality rate during this period, therefore, reflected a decreased case fatality rate. Earlier diagnosis and improved management rather than improved prevention would account for the decreased population mortality rate during this period. The trends in estimated case fatality that we showed were concordant with trends from the Minneapolis-St. Paul metropolitan area [21]. Both databases showed either an increasing or unchanging case fatality rate from 1979 to the mid- or late 1980s [21] followed by a declining case fatality rate.
References 1 Stein PD, Kayali F, Olson RE. Estimated case fatality rate of pulmonary embolism, 1979–1998. Am J Cardiol 2004; 93: 1197–1199. 2 Hermann RE, Davis JH, Holden WD. Pulmonary embolism: a clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961; 102: 19–28. 3 Barritt DW, Jordan SC. Anticoagulant drugs in the treatment of pulmonary embolism: a controlled trial. Lancet 1960; 1: 1309–1312. 4 Stein PD, Henry JW, Relyea B. Untreated patients with pulmonary embolism: outcome, clinical and laboratory assessment. Chest 1995; 107: 931–935. 5 Carson JL, Kelley MA, Duff A et al. The clinical course of pulmonary embolism. New Engl J Med 1992; 326: 1240– 1245. 6 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 7 The Columbus Investigators. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med 1997; 337: 657–662. 8 Simonneau G, Sors H, Charbonnier B et al. A comparison of low-molecular-weight heparin with unfractionated heparin for acute pulmonary embolism. The THESSE Study Group. N Engl J Med 1997; 337: 663–669.
BLUK077-Stein
March 28, 2007
CHAPTER 3
15:2
Case fatality rate and population mortality rate
9 Alpert JS, Smith R, Carlson J et al. Mortality in patients treated for pulmonary embolism. JAMA 1976; 236: 1477– 1480. 10 Anderson FA, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151: 933–938. 11 Janke RM, MCGovern PG, Folsom AR. Mortality, hospital discharges, and case fatality for pulmonary embolism in the twin cities: 1980–1995. J Clin Epidemiol 2000; 53: 103–109. 12 Goldhaber SZ, Visani L, De Rosa M, for ICOPER. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999; 353: 1386–1389. 13 Siddique RM, Siddique MI, Conners AF, Jr, Rimm AA. Thirty-day case-fatality rates for pulmonary embolism in the elderly. Arch Intern Med 1996; 156: 2343–2347. 14 Byrne JJ. Phlebitis: a study of 748 cases at the Boston City Hospital. New Engl J Med 1955; 253: 579–586. 15 Kucher N, Rossi E, De Rosa M, Goldhaber SZ. Massive pulmonary embolism. Circulation 2006; 113: 577–582.
23
16 Douketis JD, Kearon C, Bates S, Duku EK, Ginsberg JS. Risk of fatal pulmonary embolism in patients with treated venous thromboembolism. JAMA 1998; 279: 458– 462. 17 Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979–1998: an analysis using multiple-cause mortality data. Arch Intern Med 2003; 163: 1711–1717. 18 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154: 159–162. 19 Lilienfeld DE. Decreasing mortality from pulmonary embolism in the United States, 1979–1996. Int J Epidemiol 2000; 29: 465–469. 20 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694. 21 Lilienfeld DE, Godbold JH, Burke GL, Sprafka JM, Pham DL, Baxter J. Hospitalization and case fatality for pulmonary embolism in the twin cities: 1979–1984. Am Heart J 1990; 120: 392–395.
BLUK077-Stein
4
April 6, 2007
15:6
CHAPTER 4
Prognosis in acute pulmonary embolism based on right ventricular enlargement, prognostic models, and biochemical markers
Right ventricular enlargement and prognostic models The prognosis in acute pulmonary embolism (PE) depends on the severity of the PE and the prior cardiorespiratory condition of the patient. The size of the PE has been assessed by the pulmonary angiogram, CT angiogram (Chapter 73), and perfusion lung scans [1]. Indices of severity include shock, pulmonary hypertension, right ventricular dilatation, and Pa O2 . Elderly patients had a worse prognosis than young patients (Chapter 3). Literature review of the prognostic value of echocardiographically assessed right ventricular dysfunction showed that short-term PE-related deaths occurred in 4–14% more patients with right ventricular dysfunction than in those without right ventricular dysfunction [2]. Among normotensive patients, the short-term mortality was 4 and 5% higher among those with right ventricular dysfunction [2]. Others, in a review, found a short-term mortality of 5% in normotensive patients with ventricular dysfunction [3]. Among patients with PE in whom the right ventricular to left ventricular short-axis ratio (RV/LV) was measured by CT, the positive predictive value of death from PE in 3 months was 10% in 69 patients with RV/LV > 1 and it was zero among 51 patients with RV/LV ≤ 1 [4]. However, among patients with PE in PIOPED II who were treated only with anticoagulants and/or an inferior vena cava filter, who were not hypotensive, in shock, critically ill, on ventilatory support, did not have a myocardial infarction within the previous month, and did not have an episode of ven-
24
tricular tachycardia or ventricular fibrillation within the previous 24 hours, in-hospital outcome was the same in those with and those without an enlarged right ventricle (Stein PD, Beemath A, Matta F, et al., unpublished data from PIOPED II). The in-hospital mortality from PE on in these patients with an RV/LV short axis dimension ratio >1 measured on CT angiograms was 0 of 76 (0%) versus 1 of 79 (1.3%) in those with an RV/LV dimension ratio ≤1. The in-hospital all cause mortality in those with an RV/LV dimension ratio >1 was 2 of 76 (2.6%) versus 2 of 79 (2.5%) in those with an RV/LV dimension ratio ≤1. The case fatality rate in hypotensive patients and in patients with right ventricular enlargement or dysfunction is described in Chapter 89. Correlation of pulmonary artery mean pressure with angiographic severity was low (r = 0.38) as was the correlation of Pa O2 (r = −0.34) [5]. Prognostic models based on weighting of several indices of severity have been described [6–8]. The most recent consists of 11 routinely available predictors of 30-day all-cause mortality [6]. These were age, male sex, cancer heart failure, chronic lung disease, pulse ≥110/min, respiratory rate ≥30/min, temperature <36◦ C, altered mental status (disorientation, lethargy, stupor, or coma), and oxygen saturation <90 mm Hg [6]. These were assigned point scores as shown in Table 4.1. Validation in a subsequent investigation showed a 30-day all-cause mortality of 0, 1.0, 3.1, 10.4, and 24.4% in the respective five classes of risk, which did not differ from the original derivation values [9]. Several laboratory tests were also independently associated with 30-day all-cause mortality [9].
BLUK077-Stein
April 6, 2007
CHAPTER 4
15:6
Prognosis in acute pulmonary embolism
Table 4.1 Predictors of 30-day mortality. Predictor
Point Score
Age
1 point/yr
Male
10
Cancer
30
Heart failure
10
Chronic lung disease
10
Pulse >110/min
20
Systolic blood pressure <100 mm HG
30
Respiratory rate ≥30/min
20
Temperature <36◦ C
20
Disorientation, lethargy, stupor, or coma
60
Oxygen saturation <90 mm Hg
20
Risk
Point Score
very low
≤65
low intermed
66–85 86–105
high
106–125
very high
>125
Table base on Aujesky et al. [9].
These included hemoglobin <12 g/dL, white blood cell count <4000 or >12,000/mm3 , platelets <100,000/mm3 , sodium <130 or >150 meq/L, blood urea nitrogen ≥30 mg/dL, arterial pH 7.25, and Pa CO2 <25 or >55 mm Hg [9]. The more complex model that included laboratory findings showed 30-day mortalities that were similar to mortalities with the less complex model [9].
Cardiac troponins Cardiac troponins are regulatory proteins of the thin actin filaments of the cardiac muscle [10]. They control the calcium-mediated interaction of actin and myosin [11]. Troponin T and troponin I are highly sensitive and specific markers of myocardial injury [10]. The release of cardiac troponin from the myocyte to the blood can be due to reversible or irreversible cell damage [11]. Ischemia without coronary stenosis (demand ischemia), resulting from a mismatch between myocardial oxygen supply and demand, may occur [11]. This may cause increased membrane permeability and release of smaller troponin fragments into the systemic circulation [11]. It has been thought for many years that some of the electrocardiographic changes in acute PE reflect myocardial ischemia [12–14]. Myocardial
25
infarction has been shown at autopsy of patients who died of PE and had normal coronary arteries [12, 13, 15, 16]. In fact, investigations in dogs [17] and in pigs [18] with experimentally induced PE showed that blood flow increased in both the right and left coronary arteries. Coronary blood flow increased concordantly with increasing pulmonary artery pressure and decreasing PaO2 [17, 18]. The troponin complex consists of three subunits: troponin T, which binds to tropomyosin and facilitates contraction; troponin I, which binds to actin and inhibits actin–myosin interactions; and troponin C, which binds to calcium ions [11]. The amino acid sequences of the skeletal and cardiac isoforms of cardiac troponin T and troponin I are sufficiently dissimilar and therefore detectable by monoclonal antibodybased assays [11]. Troponin C is not used clinically because both cardiac and smooth muscle share troponin C isoforms. In-hospital, all-cause mortality among patients with acute PE who had an elevated troponin I ranged from 14 to 36% and in patients with a normal troponin I, in-hospital mortality ranged from 2 to 7% [19–26] (Table 4.1). Pooled data among patients with an elevated troponin I showed an in-hospital mortality 24 of 109 (22%) versus 17 of 291 (6%) among patients with a normal troponin I. Patients with a markedly elevated troponin I level above 1.5 ng/mL had a higher mortality from PE (22% mortality) than those with a modest elevation of 0.07–1.5 ng/mL (10% mortality) [20]. Among unselected PE patients with an elevated troponin T, in-hospital, all-cause mortality was 13 of 57 (23%) versus 3 of 105 (3%) [20, 24] among patients with a normal troponin T (Table 4.2).
Troponin levels in combination with right ventricular dysfunction or right ventricular dilatation Acute right ventricular dilatation or dysfunction has been shown to indicate a poor prognosis by several investigators. This is reviewed in Chapter 57. Patients with right ventricular dysfunction in combination with an elevated troponin I [23] or troponin T [27] showed high rates of in-hospital mortality compared with those in whom the right ventricle was normal and troponin level was not elevated (Table 4.3). In-hospital mortality, however, was the same in patients with
BLUK077-Stein
April 6, 2007
15:6
26
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 4.2 Mortality and troponin levels in patients with acute pulmonary embolism. Troponin I
Cutoff
Abnormal, mortality
Normal, mortality
value
Cause of
Follow-up
First author [Ref]
n/ N (%)
n/ N (%)
(ng/mL)
death
duration
Yalamanchili [19]
8/24 (33)
All cause
In-hospital
Unselected
7/43 (16)*
9/123 (7) 1/63 (2)*
≥2.0
Konstantinides [20]
≥0.07
All cause
In-hospital
Unselected
Kucher [21]
4/28 (14)
1/63 (2)
≥0.06
All cause
In-hospital
Unselected
La Vecchia [22]
5/14 (36)
1/42 (2)
>0.6
All cause
In-hospital
Unselected
Scridon [23]
23/73 (32)*
5/68 (7)*
>0.1
All cause
30 days
Unselected
Troponin T
Selection
Cutoff
Abnormal, mortality
Normal, mortality
value
Cause of
Follow-up
First author [Ref]
n/ N (%)
n/ N (%)
(ng/mL)
death
duration
Konstantinides [20]
5/39 (13)*
2/67 (3)*
≥0.04
All cause
In-hospital
Unselected
Giannitsis [24]
8/18 (44)
1/38 (3)
≥0.1
All cause
In-hospital
Unselected
Selection
Pruszczyk [25]
8/32 (25)
0/32 (0)
>0.01
All cause
In-hospital
Normotensive
Pruszczyk [26]
6/24 (25)
1/22 (5)
>0.01
All cause
In-hospital
RV Dilatation
* Estimated from authors’ graphs.
elevations of troponin, irrespective of right ventricular function.
Myoglobin Myoglobin is a low-molecular-heme protein that is found in both cardiac and skeletal muscle and, therefore, is not cardiac specific [28, 29]. It is among the earliest markers released into the circulation [30], and may be detected as early as 2 hours after the onset of myocardial necrosis [28]. Serum myoglobin after myocardial infarction increases even before a detectable rise of cardiac troponin levels occurs [29]. Myoglobin measurement for the diagnosis of the acute coronary syndrome is most efficient when measured within 6 hours after the onset of myocardial infarc-
tion [28]. Literature on myoglobin levels in pulmonary embolism is sparse. Among patients with PE who had right ventricular distention and an elevated myoglobin level, in-hospital, all-cause mortality was 7 of 21 (33%) versus 0 of 25 (0%) among such patients who had a normal myoglobin level [16] (Table 4.4).
Natriuretic peptides The natriuretic peptides are useful diagnostic and prognostic biomarkers for patients with congestive heart failure [31]. In contrast to atrial natriuretic peptide that originates mainly from atrial tissue, brain natriuiretic peptide (BNP) is produced to a large degree from ventricular myocytes [31]. The principal stimulus for BNP synthesis and secretion is cardiomyocyte
Table 4.3 Mortality and right ventricular dysfunction with elevated troponin levels in patients with acute pulmonary embolism. RV dysfunction
RV normal and
and elevated
troponin not
troponins
elevated
[mortality
[mortality
First author [Ref]
n/N (%)]
n/N (%)]
Scridon [23]
17/45 (38)
Binder [27] ∗
3/16 (19)*
Estimated from authors’ graphs.
Troponin type
Cutoff value
Cause of
Follow-up
(ng/mL)
death
duration
— (5)
Trop I
>0.1
All cause
30 days
0/53 (0)
Trop T
≥0.04
All cause
In-hospital
BLUK077-Stein
April 6, 2007
CHAPTER 4
15:6
27
Prognosis in acute pulmonary embolism
Table 4.4 Mortality and myoglobin levels in patients with acute pulmonary embolism. Myoglobin High, mortality
Normal, mortality
Cutoff value
Cause of
Follow-up
First author [Ref]
n/N (%)
n/N (%)
(ng/mL)
death
duration
Selection
Pruszczyk [20]
7/21 (33)
0/25 (0)
>0.01
All cause
In-hospital
RV dilatation
stretch [31]. Brain natriuretic peptide is a 32-amino acid peptide hormone first isolated from porcine brain tissue [31]. In plasma, the intact 108 amino acid prohormone (proBNP), the biologically active 32-amino acid BNP, and the remaining part of the prohormone, N-terminal (NT)-proBNP, which is 76 amino acids, can be measured by immunoassay [31]. Prohormones in normal ventricular myocytes are not stored to a significant amount. Therefore, it takes several hours for the plasma natruiretic levels to increase significantly after the onset of stretch [32]. This process includes BNP messenger ribonucleic acid synthesis, prohormone synthesis, and prohormone release into the circulation. Elevations in BNP [33] and NT-proBNP [34] are associated with right ventricular dysfunction in acute PE. Natriuretic peptide levels are also increased in patients with right ventricular pressure overload due to causes other than PE, including primary pulmonary hypertension, chronic thromboembolic hypertension, congenital heart disease, and chronic lung disease [35–38].
Brain natriuretic peptides in patients with PE, when low, predict a benign clinical outcome, with few inhospital deaths from PE [31]. There were no inhospital deaths among 70 patients with PE who had a low NT-proBNP level [28, 39] and 2 in-hospital deaths among 66 patients with PE who had a low BNP level [40, 41] (Table 4.5). Published cutoff values for NT-proBNP [27, 34, 39] and for BNP [40–42] vary (Table 4.4). Because BNP release into the circulation may take several hours after the onset of myocardial injury, a second measurement should be obtained 6– 12 hours after an initially negative test in a PE patient with a symptom duration <6 hours [31].
Serum uric acid Uric acid is the final product of purine nucleotide degradation. Tissue hypoxia depletes adenosine triphosphate (ATP) and activates the purine nucleotide degradation pathway to uric acid, resulting in urate overproduction [43, 44]. Serum uric acid level is therefore determined by the imbalance between
Table 4.5 Mortality, brain natriuretic peptide, and N-terminal pro-brain natriuretic peptide levels in patients with acute pulmonary embolism. Natriuretic peptide
Peptide and
First author
High, mortality
Normal, mortality
cutoff value
Cause of
Follow-up
[Ref]
n/N (%)
n/N (%)
(pg/mL)
death
duration
Selection
Binder [27]
7/67 (10)*
0/57 (0)
NT-proBNP
All cause
In-hospital
Unselected
All cause
In-hospital
Unselected
Pulmonary
In-hospital
Unselected
In-hospital
Unselected
3 months
Unselected
≥1000 Pruszczyk [39]
15/66 (23)
0/13 (0)
NT-proBNP 88–334†
Kruger [40]
1/17 (6)
1/25 (4)
BNP >90
embolism Kucher [41]
4/32 (13)
1/41 (2)
BNP ≥90
Pulmonary embolism
ten Wolde [42]
4/36 (11)
1/74 (1)
BNP >21.7 pmol/L
∗
Pulmonary embolism
Estimated from authors’ graphs. NT-proBNP cutoff level age and sex dependent. NT-proBNP, N-terminal pro-brain natriuretic peptide; BNP, brain natriuretic peptide. †
BLUK077-Stein
April 6, 2007
15:6
28
PART I
production and excretion and is increased in hypoxic states such as left ventricular failure, cyanotic heart disease, and obstructive pulmonary disease [45–50]. Among 71 patients with acute PE, serum uric acid was elevated (mean ± SD 6.2 ± 2.3 mg/dL) compared with 62 age- and sex-matched controls (4.5 ± 0.9 mg/dL) [51]. Serum uric acid was higher in the 27 patients who died from PE during hospitalization than in the remaining survivors (8.3 ± 2.2 versus 6.5 ± 2.2 mg/dL). After treatment including thrombolysis and pulmonary embolectomy, serum uric acid level significantly decreased in patients with PE from 6.7 ± 2.0 to 5.8 ± 1.9 mg/dL.
10 Ammann P, Pfisterer M, Fehr T, Rickli H. Raised cardiac troponins: causes extend beyond acute coronary syndromes. BMJ 2004; 328: 1028–1029. 11 Ammann P, Maggiorini M, Bertel O et al. Troponin as a risk factor for mortality in critically ill patients without acute coronary syndromes. J Am Coll Cardiol 2003; 41: 2004–2009. 12 Horn H, Dack S, Friedberg CK. Cardiac sequelae of embolism of the pulmonary artery. Arch Int Med 1939; 64: 296. 13 Dack S, Master AM, Horn H, Grishman A, Field LE. Acute coronary insufficiency due to pulmonary embolism. Am J Med 1949; 7: 464. 14 Weber DM, Phillips JH, Jr. A re-evaluation of electrocardiographic changes accompanying acute pulmonary embolism. Am J Med Sci 1966; 251: 381–398. 15 Currens J, Barnes AR. The heart in pulmonary embolism. Arch Int Med 1943; 71: 325. 16 Coma-Canella I, Gamallo C, Martinez Onsurbe P, LopezSendon J. Acute right ventricular infarction secondary to massive pulmonary embolism. Eur Heart J 1988; 9: 534– 540. 17 Stein PD, Alshabkhoun S, Hatem C et al. Coronary artery blood flow in acute pulmonary embolism. Am J Cardiol 1968; 21: 32–37. 18 Stein PD, Alshabkhoun S, Hawkins HF, Hyland JW, Jarrett CE. Right coronary blood flow in acute pulmonary embolism. Am Heart J 1969; 77: 356–362. 19 Yalamanchili K, Sukhija R, Aronow WS, Sinha N, Fleisher AG, Lehrman SG. Prevalence of increased cardiac troponin I levels in patients with and without acute pulmonary embolism and relation of increased cardiac troponin I levels with in-hospital mortality in patients with acute pulmonary embolism. Am J Cardiol 2004; 93: 263– 264. 20 Konstantinides S, Geibel A, Olschewski M et al. Importance of cardiac troponins I and T in risk stratification of patients with acute pulmonary embolism. Circulation 2002; 106: 1263–1268. 21 Kucher N, Wallmann D, Carone A, Windecker S, Meier B, Hess OM. Incremental prognostic value of troponin I and echocardiography in patients with acute pulmonary embolism. Eur Heart J 2003; 24: 1651– 1656. 22 La Vecchia L, Ottani F, Favero L et al. Increased cardiac troponin I on admission predicts in-hospital mortality in acute pulmonary embolism. Heart 2004; 90: 633– 637. 23 Scridon T, Scridon C, Skali H, Alvarez A, Goldhaber SZ, Solomon SD. Prognostic significance of troponin elevation and right ventricular enlargement in acute pulmonary embolism. Am J Cardiol 2005; 96: 303– 305.
References 1 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Perfusion lung scanning. Circulation 1973; 47(2, suppl): II46–II50. 2 ten Wolde M, Sohne M, Quak E, Mac Gillavry MR, Buller HR. Prognostic value of echocardiographically assessed right ventricular dysfunction in patients with pulmonary embolism. Arch Intern Med 2004; 164: 1685–1689. 3 Gibson NS, Sohne M, Buller HR. Prognostic value of echocardiography and spiral computed tomography in patients with pulmonary embolism. Curr Opin Pulmon Med 2005; 11: 380–384. 4 van der Meer RW, Pattynama PM, van Strijen MJ et al. Right ventricular dysfunction and pulmonary obstruction index at helical CT: prediction of clinical outcome during 3-month follow-up in patients with acute pulmonary embolism. Radiology 2005; 235: 798–803. 5 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Interrelationships of pulmonary angiograms, lung scans, hemodynamic measurements, and fibrinolytic findigs. Circulation 1973; 47(2, suppl): II73– II80. 6 Aujesky D, Obrosky DS, Stone RA et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med 2005; 172: 1041–1046. 7 Wicki J, Perrier A, Perneger TV, Bounameaux H, Junod AF, Predicting adverse outcome in patients with acute pulmonary embolism: a risk score. Thromb Haemost 2000; 84: 548–552. 8 Nendaz MR, Bandelier P, Aujesky D et al. Validation of a risk score identifying patients with acute pulmonary embolism, who are at low risk of clinical adverse outcome. Thromb Haemost 2004; 91: 1232–1236. 9 Aujesky D, Roy PM, Le Manach CP et al. Validation of a model to predict adverse outcomes in patients with pulmonary embolism. Eur Heart J 2006; 27: 476–481.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
April 6, 2007
CHAPTER 4
15:6
Prognosis in acute pulmonary embolism
24 Giannitsis E, Muller-Bardorff M, Kurowski V et al. Independent prognostic value of cardiac troponin T in patients with confirmed pulmonary embolism. Circulation 2000; 102: 211–217. 25 Pruszczyk P, Bochowicz A, Torbicki A et al. Cardiac troponin T monitoring identifies high-risk group of normotensive patients with acute pulmonary embolism. Chest 2003; 123: 1947–1952. 26 Pruszczyk P, Bochowicz A, Kostrubiec M et al. Myoglobin stratifies short-term risk in acute major pulmonary embolism. Clin Chim Acta 2003; 338: 53– 56. 27 Binder L, Pieske B, Olschewski M et al. N-terminal pro-brain natriuretic peptide or troponin testing followed by echocardiography for risk stratification of acute pulmonary embolism. Circulation 2005; 112: 1573– 1579. 28 Braunwald E, Antman EM, Beasley JW, et al. ACC/AHA guidelines for the management of patients with unstable angina and non-ST-segment elevation myocardial infarction: a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Unstable Angina). J Am Coll Cardiol, 2000: 36: 970– 1062. 29 de Lemos JA, Morrow DA, Gibson M et al. The prognostic value of serum myoglobin in patients with nonST-segment elevation acute coronary syndromes. Results from the TIMI 11B and TACTICS-TIMI 1B Studies. J Am Coll Cardiol 2002; 40: 238–244. 30 Mair J, Wagner I, Jakob G et al. Different time courses of cardiac contractile proteins after acute myocardial infarction. Clin Chim Acta 1994; 231: 47–60. 31 Kucher N, Goldhaber SZ. Cardiac biomarkers for risk stratification of patients with acute pulmonary embolism. Circulation 2003; 108: 2191–2194. 32 Hama N, Itoh H, Shirakami G et al. Rapid ventricular induction of brain natriuretic peptide gene expression in experimental acute myocardial infarction. Circulation 1995; 92: 1558–1564. 33 Tulevski II, Hirsch A, Sanson BJ et al. Increased brain natriuretic peptide as a marker for right ventricular dysfunction in acute pulmonary embolism. Thromb Haemost 2001; 86: 1193–1196. 34 Kucher N, Printzen G, Doernhoefer T, Windecker S, Meier B, Hess OM. Low pro-brain natriuretic peptide levels predict benign clinical outcome in acute pulmonary embolism. Circulation 2003; 107: 1576–1578. 35 Nagaya N, Nishikimi T, Okano Y et al. Plasma brain natriuretic peptide levels increase in proportion to the extent of right ventricular dysfunction in pulmonary hypertension. J Am Coll Cardiol 1998; 31: 202– 208.
29
36 Nagaya N, Nishikimi T, Uematsu M et al. Plasma brain natriuretic peptide as a prognostic indicator in patients with primary pulmonary hypertension. Circulation 2000; 102: 865–870. 37 Bando M, Ishii Y, Sugiyama Y, Kitamura S. Elevated plasma brain natriuretic peptide levels in chronic respiratory failure with cor pulmonale. Respir Med 1999; 93: 507–514. 38 Tulevski II, Groenink M, van Der Wall EE et al. Increased brain and atrial natriuretic peptides in patients with chronic right ventricular pressure overload: correlation between plasma neurohormones and right ventricular dysfunction. Heart 2001; 86: 27–30. 39 Pruszczyk P, Kostrubiec M, Bochowicz A et al. Nterminal pro-brain natriuretic peptide in patients with acute pulmonary embolism. Eur Respir J 2003; 22: 649– 653. 40 Kruger S, Graf J, Merx MW et al. Brain natriuretic peptide predicts right heart failure in patients with acute pulmonary embolism. Am Heart J 2004; 147: 60–65. 41 Kucher N, Printzen G, Goldhaber SZ. Prognostic role of brain natriuretic peptide in acute pulmonary embolism. Circulation 2003; 107: 2545–2547. 42 ten Wolde M, Tulevski II, Mulder JW et al. Brain natriuretic peptide as a predictor of adverse outcome in patients with pulmonary embolism. Circulation 2003; 107: 2082–4208. 43 Fox AC, Reed GE, Meilman H, Silk BB. Release of nucleosides from canine and human hearts as an index of prior ischemia. Am J Cardiol 1979; 43: 52–58. 44 Mentzer RM, Jr, Rubio R, Berne RM. Release of adenosine by hypoxic canine lung tissue and its possible role in pulmonary circulation. Am J Physiol 1975; 229: 1625– 1631. 45 Leyva F, Anker S, Swan JW, Godsland IF, Wingrove CS, Chua TP et al. Serum uric acid as an index of impaired oxidative metabolism in chronic heart failure. Eur Heart J 1997; 18: 858–865. 46 Anker SD, Leyva F, Poole-Wilson PA, Kox WJ, Stevenson JC, Coats AJ. Relation between serum uric acid and lower limb blood flow in patients with chronic heart failure. Heart 1997; 78: 39–43. 47 Hayabuchi Y, Matsuoka S, Akita H, Kuroda Y. Hyperuricaemia in cyanotic congenital heart disease. Eur J Pediatr 1993; 152: 873–876. 48 Hasday JD, Grum CM. Nocturnal increase of urinary uric acid:creatinine ratio: a biochemical correlate of sleepassociated hypoxemia. Am Rev Respir Dis 1987; 135: 534– 538. 49 Braghiroli A, Sacco C, Erbetta M, Ruga V, Donner CF. Overnight urinary uric acid:creatinine ratio for detection of sleep hypoxemia: validation study in chronic
BLUK077-Stein
April 6, 2007
15:6
30
PART I
obstructive pulmonary disease and obstructive sleep apnea before and after treatment with nasal continuous positive airway pressure. Am Rev Respir Dis 1993; 148: 173– 178. 50 Elsayed NM, Nakashima JM, Postlethwait EM. Measurement of uric acid as a marker of oxygen tension
in the lung. Arch Biochem Biophys 1993; 302: 228– 232. 51 Shimizu Y, Nagaya N, Satoh T et al. Serum uric acid level increases in proportion to the severity of pulmonary thromboembolism. Circulation 2002; 66: 571– 575.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
April 6, 2007
5
15:7
CHAPTER 5
Changing risks of untreated deep venous thrombosis and acute pulmonary embolism
Introduction The frequency of fatal pulmonary embolism (PE) in patients with untreated deep venous thrombosis (DVT) has diminished as diagnostic tests have made it possible to diagnose mild DVT [1]. Prior to the use of venography and of sensitive noninvasive tests for the early detection of DVT, the risk of fatal PE in untreated patients with clinically apparent DVT was 37% [2] (Figure 5.1). In pooled data of patients with un-
100
60 Clinical DVT (%)
Fatal PE (%)
80
treated DVT identified by radioactive fibrinogen scintiscans, most of which was distal and subclinical, fatal PE occurred in 5% [3]. It is apparent that the risk of fatal PE was greater among patients with more severe DVT. The percentage of patients with acute PE who have clinically detectable DVT has diminished as physicians have developed the ability to diagnose subtle PE [2]. Among patients who died from acute PE, 53% had clinically identified DVT [4] (Figure 5.2). In an investigation of patients with massive or submassive angiographically diagnosed acute PE (the Urokinase Pulmonary Embolism Trial), 34% of patients had clinically identifiable DVT [5]. Among patients with mild as well as severe acute PE (PIOPED), only 15% of patients had clinically apparent DVT [6] and 47% had signs of DVT in PIOPED II [7].
40
20
60 53 34 20 15 0
0 Clinical DVT
Subclinical DVT
Figure 5.1 Frequency of fatal pulmonary embolism (PE) in untreated patients with clinically apparent deep venous thrombosis (DVT), and patients, most of whom had subclinical DVT diagnosed by radioactive fibrinogen scintiscans. (Data are from Byrne [2] and Collins et al. [3]. Reprinted from Stein [1] with permission.)
47
40
Fatal PE
Massive Massive or Massive or or mild PE mild PE submassive (PIOPED) (PIOPED II) PE
Figure 5.2 Frequency of clinically apparent deep venous thrombosis (DVT) among patients with fatal acute pulmonary embolism (PE), massive or submassive pulmonary embolism, or massive or mild pulmonary embolism. (Data are based on Byrne and O’Neil [4], the Urokinase Pulmonary Embolism Trial [5], Stein [6] and Stein et al. [7].)
31
April 6, 2007
15:7
32
In the present era of early diagnosis of acute PE, the risk of fatal recurrent PE as well as the risk of death from the initial PE in untreated patients is lower than in patients with severe PE reported in prior years [1]. Among untreated patients with acute PE diagnosed on the basis of clinical features that included evidence of right ventricular failure, pulmonary infarction, or both, Barritt and Jordan reported a 26% mortality from the initial PE, although some of these patients, perhaps died from recurrent PE [8] (see Chapter 3, untreated patients). In 1961, Hermann and associates [9] calculated a 37% mortality from the initial PE and a 36% frequency of fatal recurrent PE (see Chapter 3, untreated patients).
Untreated patients in present era of early diagnosis In an investigation of the clinical course of acute PE, Carson and associates observed that in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED), 24 patients escaped treatment in the hospital [10]. However, 4 patients were begun on anticoagulant therapy during the first month after discharge. We evaluated the 20 patients who received no treatment for PE during the first 3 months of follow-up [11]. Only 1 of these patients died of PE [10, 11]. In the era of early diagnosis by ventilation–perfusion lungs scans and pulmonary angiography, mortality from the initial PE and from recurrent PE among patients with untreated mild PE was 1 of 20 (5%) (Figure 5.3) [11]. The circumstances involving no therapy in these 20 patients from PIOPED are as follows: 19 had pulmonary angiograms interpreted as showing no PE by the local radiologist, but the interpretation of no PE was reversed in 18 following reevaluation by the central panel of angiogram readers [11]. The diagnosis of no PE was reversed in 1 patient because PE was found at autopsy 6 days after the pulmonary angiogram. One patient had no pulmonary angiogram; the patient died of unrelated causes 4 days after a ventilation– perfusion lung scan and autopsy showed small peripheral PE. The untreated patient who died was a 33-year-old woman with underlying primary pulmonary hypertension with right ventricular failure [11]. Organized
PART I
Prevalence, risks, and prognosis of PE and DVT
100
Fatal initial and recurrent PE (%)
BLUK077-Stein
80
60
40
20
0 Clinically apparent PE
Mild PE
Figure 5.3 Fatal initial and fatal recurrent pulmonary embolism (PE) among untreated patients. Comparison is made between patients in whom the diagnosis was clinically apparent, and presumably pulmonary embolism was severe and patients in whom pulmonary embolism was mild. (Data are from Hermann et al. [9], and Stein and Henry [11]. Reprinted from Stein [1] with permission.
and fresh pulmonary emboli were shown at autopsy 6 days after a pulmonary angiogram that failed to show PE. Whether this death resulted from the original PE or recurrent PE is uncertain. In regard to the course of untreated mild PE over 1 year, there were no instances of fatal recurrent PE [11]. This assumes that the cause of death in the only patient who died was the original PE. Fatal recurrent PE, therefore, was 0 of 19 (0%) among untreated survivors of mild PE during months 4–12 of observation [11]. One patient died of aspiration pneumonia following hysterectomy for endometrial carcinoma 4 days after an intermediate probability ventilation–perfusion lung scan was obtained [11]. Multiple small thromboemboli in peripheral branches were observed at autopsy. These thromboemboli did not contribute to death. Based on clinical assessment, recurrent PE was thought to have occurred. The frequency of nonfatal recurrent PE among survivors of the first PE was 1 of 19 (5.3%).
BLUK077-Stein
April 6, 2007
CHAPTER 5
15:7
100 90 80 Fatal initial PE and fatal and nonfatal recurrent PE (%)
33
Changing risks of untreated DVT and acute PE
70 60 50 40 30 20 10 0 Clinical PE
Mild PE
Figure 5.4 Fatal initial pulmonary embolism (PE), fatal recurrent PE, and nonfatal recurrent PE among patients with clinically apparent severe PE and patients with mild PE. (Data are from Hermann et al. [8] and Stein and Henry [10]).
The frequency of fatal initial PE, fatal recurrent PE, and nonfatal recurrent PE over the course of 1 year among untreated patients with acute PE was 2 of 20 (10%) (Figure 5.4). Untreated patients with PE had mild PE as suggested by the following data. Among patients who had measurements of the PaO2 while breathing room air, the PaO2 was lower in untreated patients compared to treated patients (39 ± 16 versus 55 ± 31 mm Hg) (P < 0.001) [11]. The pulmonary artery mean pressure did not show a statistically significant difference between untreated and treated patients (23 ± 13 mm Hg versus 24 ± 10 mm Hg). Ventilation–perfusion (V–Q) lung scans were interpreted as high probability in a smaller percent of untreated patients with PE than treated patients, 0 of 20 (0%) versus 160 of 376 (43%) [11]. Low probability, nearly normal, or normal ventilation–perfusion scans were more frequent among untreated patients 10 of 20 (50%) versus 58 of 376 (15%). Ventilation– perfusion lung scans among untreated patients more
often showed no mismatched segmental equivalent perfusion defects than among treated patients, 14 of 20 (70%) versus 122 of 376 (32%) [11]. All untreated patients, 20 of 20 (100%) showed fewer than 3 mismatched segmental equivalent perfusion defects compared with 227 of 376 (60%) among treated patients. Pulmonary angiograms at the time of PIOPED entry were obtained in 19 of the untreated patients [11]. Thromboemboli involved only segmental pulmonary arteries or smaller branches in 16 of 19 (84%) of untreated patients compared with 132 of 362 (36%) treated patients. Thromboemboli were not observed on the angiogram of 1 untreated patient, but were shown at autopsy 6 days later. The frequency of fatal initial and fatal recurrent PE in untreated patients with mild PE (5.0%) is strikingly lower than the mortality from untreated PE reported in past decades among patients who presumably had severe PE [8, 9]. This lower mortality appears to relate to the milder severity of PE in these untreated patients. The mortality of untreated patients with mild PE is comparable to the mortality from fatal PE in untreated patients with subtle DVT, approximately 5% [3].
Rates of recurrent PE in untreated patients with PE based on calculations in patients with suspected PE and negative serial noninvasive leg tests The estimated frequency of PE during the 3-month follow-up of untreated patients with nonmassive PE and serial noninvasive leg tests was between 3 and 9%. The estimated frequency of fatal PE was 1% [12] (see Chapter 88).
References 1 Stein PD. Changing patterns of risk of untreated thromboembolic disease. Semin Respir Crit Care Med 1996; 17: 3–6. 2 Byrne JJ. Phlebitis: a study of 748 cases at the Boston City Hospital. New Engl J Med 1955; 253: 579–586. 3 Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. Overview of results of randomized trials in general, orthopedic, and urologic surgery. New Engl J Med 1988; 318: 1162–1173.
BLUK077-Stein
April 6, 2007
15:7
34
4 Byrne JJ, O’Neil EE. Fatal pulmonary emboli. A study of 130 autopsy-proven fatal emboli. Am J Surg 1952; 83: 47–49. 5 A National Cooperative Study. Clinical and electrocardiographic observations. The Urokinase Pulmonary Embolism Trial. Circulation 1973; 47/48(suppl II): II-60– II-65. 6 Stein PD. Unpublished data from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). 7 Stein PD, Beemath A, Matta F, et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med 2007 (In press). 8 Barritt DW, Jordan SC. Anticoagulant drugs in the treatment of pulmonary embolism: a controlled trial. Lancet 1960; 1: 1309–1312.
PART I
Prevalence, risks, and prognosis of PE and DVT
9 Hermann RE, Davis JH, Holden WD. Pulmonary embolism: a clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961; 102: 19–28. 10 Carson JL, Kelley MA, Duff A et al. The clinical course of pulmonary embolism. New Engl J Med 1992; 326: 1240– 1245. 11 Stein PD, Henry JW, Relyea B. Untreated patients with pulmonary embolism: outcome, clinical and laboratory assessment. Chest 1995; 107: 931–935. 12 Stein PD, Hull RD, Raskob GE. Withholding treatment in patients with acute pulmonary embolism who have a high risk of bleeding and negative serial noninvasive leg tests. Am J Med 2000; 109: 301– 306.
BLUK077-Stein
March 28, 2007
6
15:5
CHAPTER 6
Resolution of pulmonary embolism
Pulmonary emboli resolve because of natural thrombolytic processes [1, 2]. The rate of resolution of perfusion defects, calculated as a percent of the pretreatment defect among 70 patients treated with anticoagulants in the Urokinase Pulmonary Embolism Trial [1], is shown in Figure 6.1. After 24 hours, there was only a 7% mean resolution of the pretreatment perfusion scan defect [1] (Figure 6.1). By 2 days, there was 16% mean resolution. The mean resolution progressively increased to 75% by 3 months, and thereafter, increased only slightly. Among patients with no prior cardiopulmonary disease, ≥90% resolution was shown at 1 year in 29 of 32 (91%) [1] (Figure 6.2).
However, among patients who had prior cardiopulmonary disease, ≥90% resolution was shown at 1 year in only 13 of 18 (72%). Others showed complete clearing of the perfusion scan in 7 of 10 (70%) patients with no prior cardiopulmonary disease [3], and in 22 of 33 (67%), many of who had prior cardiopulmonary disease [4]. In the Urokinase Pulmonary Embolism Trial, the proportion of patients with ≥90% resolution of the perfusion defect was similar in those treated with anticoagulants and those treated with Urokinase in both patients with prior cardiopulmonary disease and patients with no prior cardiopulmonary disease [1] (Figure 6.3).
Figure 6.2 Proportion of patients with <10% residual perfusion defects after 1 year of treatment with anticoagulants according to whether they had cardiopulmonary disease (CPD) or no cardiopulmonary disease. (Data from the Urokinase Pulmonary Embolism Trial [1].)
Mean resolution (%)
60
77
53 42
40
32 16
20
21
7 0 24hrs day 2
Percent of patients with <10% residual perfusion defects
Figure 6.1 Mean resolution of perfusion defects shown as a percentage (%) of pretreatment defects in relation to time of treatment. Hrs = hours, mos = months. (Data from the Urokinase Pulmonary Embolism Trial [1].)
77
75
80
day 3
100 80
day 5
day 7 day 14 3 mos 6 mos 12 mos
91 72
60 40 20 0 CPD (n=18)
No CPD (n=32)
35
BLUK077-Stein
March 28, 2007
15:5
Percent of patients with <10% residual perfusion defects
36
PART I
100 80
AC 72
UK 77
AC 91
Prevalence, risks, and prognosis of PE and DVT
UK 88
60 40 20 0
CPD (n=18)
CPD (n=22)
No CPD No CPD (n=32) (n=37)
Using newer scintigraphic techniques than used in the Urokinase Pulmonary Embolism Trial, and somewhat different methods of calculating the improvement of perfusion, de Groot showed a mean rate of improvement of the perfusion scan of 4% in 2–4 days [5]. Importantly, a high probability interpretation of the lung scan remained high probability in 77 of 79 patients (97%) during the 2–4 day period. The V–Q scan became nondiagnostic in 1 of 79 (1%) and it became normal in 1 of 79 (1%) [5]. Resolution of acute PE based on pulmonary angiograms obtained 1 to 7 days after the initial angiogram in 6 patients showed no change in 1 and minimal resolution with persistent marked abnormalities in 5 [2]. Ten to 21 days after the initial pulmonary angiogram, among 10 patients, no resolution was shown in 1, minimal resolution in 2, moderate resolution with persistent minimal abnormalities in 5, and complete resolution (normal angiogram) in 2. After 34 days, among 2 patients, 1 showed minimal resolution and 1 showed complete resolution. The pulmonary diffusing capacity of carbon monoxide (DLCO) after 1 year was only 72% of predicted among 21 patients with PE and no prior cardiopulmonary disease, who were treated with anticoagulants [6]. Others also observed a continuing impairment of DLCO in 10 patients treated with anticoagulants, although the perfusion lung scans tended to return to normal [3]. Among 11 patients such pa-
Figure 6.3 Percent of patients with <10% residual perfusion defects after treatment with anticoagulants (AC) or Urokinase (UK) according to whether they had cardiopulmonary disease (CPD) or no cardiopulmonary disease. (Data from The Urokinase Pulmonary Embolism Trial [1].)
tients followed a mean of 7.4 years, mean pulmonary artery pressure (22 mm Hg) was somewhat elevated and pulmonary vascular resistance at rest was also elevated (171 dyne –sec/cm−5 ) [7].
References 1 The Urokinase Pulmonary Embolism Trial. A national cooperative study. Perfusion lung scanning. Circulation 1973; 47(2 suppl): II46–II50. 2 Dalen JE, Banas JS, Jr, Brooks HL, Evans GL, Paraskos JA, Dexter L. Resolution rate of acute pulmonary embolism in man. N Engl J Med 1969; 280: 1194–1199. 3 Wimalaratna HS, Farrell J, Lee HY. Measurement of diffusing capacity in pulmonary embolism. Respir Med 1989; 83: 481–485. 4 Paraskos JA, Adelstein SJ, Smith RE et al. Late prognosis of acute pulmonary embolism. N Engl J Med 1973; 289: 55–58. 5 de Groot MR, Oostdijk AH, Engelage AH, van Marwijk Kooy M, Buller HR. Changes in perfusion scintigraphy in the first days of heparin therapy in patients with acute pulmonary embolism. Eur J Nucl Med 2000; 27: 1481– 1486. 6 Sharma GV, Burleson VA, Sasahara AA. Effect of thrombolytic therapy on pulmonary–capillary blood volume in patients with pulmonary embolism. N Engl J Med 1980; 303: 842–845. 7 Sharma GV, Folland ED, McIntyre KM, Sasahara AA. Long-term benefit of thrombolytic therapy in patients with pulmonary embolism. Vasc Med 2000; 5: 91–95.
BLUK077-Stein
March 12, 2007
7
15:14
CHAPTER 7
Upper extremity deep venous thrombosis
Prior to 1967, thrombosis of the upper extremities constituted less than 2% of cases of deep venous thrombosis (DVT) [1]. Since the 1970s, there has been an increased recognition of upper extremity DVT [2–6]. The incidence of upper extremity DVT in a community teaching general hospital was reviewed during the 2-year period, July 1998 through June 2000 [7]. The incidence of upper extremity DVT in adults (≥20 years) was 64 of 34,567 hospital admissions (0.19%). The incidence of upper extremity DVT that we observed among hospitalized patients was the same as reported by Kroger and associates (0.2%) [3]. Upper extremity DVT in adults was accompanied by proximal DVT of the lower extremity in 2 of 64 patients [7] (see Figure 7.1). Upper extremity DVT involved the subclavian vein in 48 patients (75%) and the axillary vein in 25 patients (39%) [7]. The internal jugular vein was included among patients with upper extremity DVT, and it involved 29 patients (45%) [7]. Nine patients (14%) had involvement only of a deep distal vein (brachial
Figure 7.1 Patient with upper extremity deep venous thrombosis (DVT). (Courtesy of Syed Mustafa, MD, St. Joseph Mercy-Oakland Hospital, Pontiac Michigan.)
veins in 6 patients, ulnar vein alone in 1 patient, radial vein alone in 1 patient, and both radial and ulnar veins in 1 patient). In 7 patients, the upper extremity DVT was shown by venography to extend proximally to the brachiocephalic vein. Among these, 2 extended to the superior vena cava. In addition to these patients with DVT of the upper extremity, 16 patients had involvement only of the superficial veins of the upper extremity. These patients are not included in the various computations. All the patients with upper extremity DVT received therapy with anticoagulants [7]. None developed pulmonary embolism (PE). Cancer was diagnosed in 30 of 65 (46%). Upper extremity DVT in the past had been considered to be benign and self-limited [8]. More recent data suggest that this is not the case [8–11]. Systematic review of the literature in 1991 showed a 9% incidence of symptomatic PE in patients with upper extremity DVT, half of which was diagnosed by objective tests [12]. Review in 1991 by others showed a 7% incidence of symptomatic PE in patients with upper extremity DVT, none of which was confirmed by objective tests [13]. More recently, symptomatic PE diagnosed by ventilation–perfusion lung scans was observed in 7% of patients with upper extremity DVT [8]. Most PE (94%) in patients with upper extremity DVT occurred in untreated patients [9]. Fatal PE in patients with upper extremity DVT has been reported [10]. Routine ventilation/perfusion lung scans in patients with upper extremity DVT were high probability for PE in 13% [13]. The increased incidence of upper extremity DVT has been attributed to the use of central venous catheters and transvenous pacemakers [2]. As the use of central venous lines and pacemaker wires has increased, their role in the etiology of upper extremity DVT has become prominent [3, 14–17]. The incidence of central venous catheter-related DVT assessed by venography has been reported in 27–66% of patients [18]. Most of
37
BLUK077-Stein
March 12, 2007
15:14
38
PART I
the thrombi in these cases were asymptomatic. The reported incidence of symptomatic catheter-related DVT in adults ranged from 0.3 to 28.3% [18]. In the series we reported, central venous access lines on the side of the upper extremity DVT were inserted in 39 of 65 (60%) [7]. Nearly the same percentage (55%) of upper extremity DVT was associated with central venous access lines in the experience of others [19]. Others reported the use of indwelling catheters in 28–33% of patients with upper extremity DVT [3, 9, 14]. Six patients in our series had arteriovenous shunts on the side of the upper extremity DVT [7]. There were 3 additional patients who had thrombosis only of the arteriovenous shunt without involvement of contiguous veins. These patients were not included among the patients we reported. Nineteen patients (29%) had no apparent cause, although all patients developed upper extremity DVT in the hospital and had received intravenous infusions of medications. Three of 39 patients (8%) who had central venous access lines had been on antithrombotic prophylaxis with lowdose warfarin (1–2 mg/day) prior to developing upper extremity DVT. Among 30 patients with venous access lines and upper extremity DVT in whom there was data, the lines had been inserted 3–14 days in 90%. Swelling of the arm was the most frequent sign, and was present in all patients with upper extremity DVT. Pain was present in 26 of 65 (40%) patients. Some discomfort due to the swelling was present in all. Four of 65 (6%) had erythema over the affected site. One patient with internal jugular vein thrombosis had swelling of the neck. He also had thrombosis of the superior vena cava. Campbell and associates, among 25 patients with upper extremity DVT, observed swelling in 96%, pain in 76%, discoloration in 52%, prominent veins in 52%, and a palpable cord in 8% [2].
We observed an association of malignancy with upper extremity DVT in 29 of 65 (45%) [7]. Among these patients, 23 also had a central line. An association of upper extremity DVT with malignancy is well established [20] and was reported in 64% by Baarslag et al. [19]. A hypercoagulable state may also be associated with proximal upper vein thrombosis [21], but our patients generally were not evaluated for this. During the 2-year period of this investigation, 10 of 34,567 patients (0.03%) had thrombosis of the superior vena cava or brachiocephalic vein, unaccompanied by DVT of the subclavian or axillary veins or more distal veins of the upper extremity and unaccompanied by proximal DVT of the lower extremity [22] (Figure 7.2). During the same period, proximal DVT of the lower extremity was diagnosed in 271 of 34,567 patients (0.78%) [23] (Figure 7.2). Deep venous thrombosis of the upper extremity, sometimes accompanied by thrombosis of the superior vena cava or brachiocephalic vein, but unaccompanied by proximal DVT of the lower extremity, was observed in 62 of 34,567 adult patients (0.18%) [7]. Among all patients with symptomatic DVT, thrombosis of the superior vena cava or brachiocephalic vein alone constituted 10 of 343 (3%) [7, 22, 23] (Figure 7.3). Upper extremity DVT, unaccompanied by proximal lower extremity DVT but sometimes accompanied by thrombosis of the superior vena cava or brachiocephalic vein, constituted 62 of 343 (18%), and proximal lower extremity DVT constituted 271 of 343 (79%). More recently, a prospective registry of 324 patients with central venous catheter-associated upper extremity DVT and 268 patients with noncentral venous catheter-associated upper extremity DVT was published [24]. An indwelling central venous catheter was
Prevalence, risks, and prognosis of PE and DVT
DVT in hosp pts (%)
1 0.8 0.6
0.78%
0.4 0.18%
0.2
0.03%
0 Lower extremity
Upper extremity, SVC, or brachioceph
SVC or brachioceph
Figure 7.2 Prevalence in hospitalized (hosp) patients (pts) of proximal lower extremity deep venous thrombosis (DVT), upper extremity, superior vena cava (SVC) or brachiocephalic (brachioceph) DVT unaccompanied by lower extremity DVT, and isolated SVC or brachiocephalic DVT. (Data from Mustafa et al. [7], Otten et al. [22], and Stein et al. [23].)
BLUK077-Stein
March 12, 2007
CHAPTER 7
15:14
Upper extremity deep venous thrombosis
SVC or brachiocephalic 3%
Upper extremity 18%
Lower extremity 79%
Figure 7.3 Proportion of patients with proximal deep venous thrombosis (DVT) of lower extremities, upper extremity DVT, unaccompanied by proximal lower extremity DVT but sometimes accompanied by thrombosis of the superior vena cava or brachiocephalic vein. (Data from Mustafa et al. [7], Otten et al. [22], and Stein et al. [23].)
the strongest independent predictor of upper extremity DVT. Only 20% of patients with upper extremity DVT who did not have a contraindication to anticoagulation were receiving anticoagulant prophylaxis at the time of diagnosis of upper extremity DVT [24]. Frequent risk factors in patients with central venous catheter-associated upper extremity DVT and in patients with noncentral venous catheter-associated upper extremity DVT were cancer, personal history, or family history of venous thromboembolism (VTE). Surgery or immobilization within 30 days, and ongoing chemotherapy were frequent risk factors in patients with central venous catheter-associated upper extremity DVT. Ongoing chemotherapy was a frequent risk factor in patients with noncentral venous catheterassociated upper extremity DVT [24]. Among 324 patients with upper extremity DVT, 25% had upper extremity swelling, 11% had upper extremity discomfort, and 4% had erythema [24]. Pulmonary embolism at the time of diagnosis of upper extremity DVT was confirmed in 0.9%.
References 1 Coon WW, Willis PW, III. Thrombosis of axillary and subclavian veins. Arch Surg 1967; 94: 657–663. 2 Campbell CB, Chandler JG, Tegtmeyer CJ, Bernstein EF. Axillary, subclavian, and brachiocephalic vein obstruction. Surgery 1977; 82: 816–826.
39
3 Kroger K, Schelo C, Gocke C, Rudofsky G. Colour Doppler sonographic diagnosis of upper limb venous thromboses. Clin Sci 1998; 94: 657–661. 4 Huber P, Hauptli W, Schmitt HE, Widmer LK. [Axillary and subclavian vein thrombosis and its sequelae]. Internist (Berl) 1987; 28: 336–343. 5 Theis W, Zaus M, Klefhaber M et al. Primare und sekundare Schultergurtelvenenthrombose; eine Analyse von 227 Patienten [Abstract]. VASA 1994; l43(suppl): 167– 169. 6 Layher T, Heinrich F. Retrospektive Betrachtung von Arm- bzw-Schultervenenthrombosen am Krankenhaus Bruchsal im Zeitraum von 1973 und 1993 [Abstract]. VASA 1994; 43(suppl): 169. 7 Mustafa S, Stein PD, Patel KC, Otten TR, Holmes R, Silbergleit A. Upper extremity deep venous thrombosis. Chest 2003; 123: 1953–1956. 8 Hingorani A, Ascher E, Lorenson E et al. Upper extremity deep venous thrombosis and its impact on morbidity and mortality rates in a hospital-based population. J Vasc Surg 1997; 26: 853–860. 9 Horattas MC, Wright DJ, Fenton AH et al. Changing concepts of deep venous thrombosis of the upper extremityreport of a series and review of the literature. Surgery 1988; 104: 561–567. 10 Monreal M, Raventos A, Lerma R et al. Pulmonary embolism in patients with upper extremity DVT associated to venous central lines—a prospective study. Thromb Haemost 1994; 72: 548–550. 11 Prandoni P, Polistena P, Bernardi E et al. Upperextremity deep vein thrombosis. Risk factors, diagnosis, and complications. Arch Intern Med 1997; 157: 57–62. 12 Becker DM, Philbrick JT, Walker FB, IV. Axillary and subclavian venous thrombosis. Prognosis and treatment. Arch Intern Med 1991; 151: 1934–1943. 13 Monreal M, Lafoz E, Ruiz J, Valls R, Alastrue A. Upperextremity deep venous thrombosis and pulmonary embolism. A prospective study. Chest 1991; 99: 280– 283. 14 Timsit JF, Farkas JC, Boyer JM et al. Central vein catheterrelated thrombosis in intensive care patients: incidence, risks factors, and relationship with catheter-related sepsis. Chest 1998; 114: 207–213. 15 Haire WD, Lieberman RP, Edney J et al. Hickman catheter-induced thoracic vein thrombosis. Frequency and long-term sequelae in patients receiving high-dose chemotherapy and marrow transplantation. Cancer 1990; 66: 900–908. 16 Ryan JA, Jr, Abel RM, Abbott WM et al. Catheter complications in total parenteral nutrition. A prospective study of 200 consecutive patients. N Engl J Med 1974; 290: 757–761.
BLUK077-Stein
March 12, 2007
15:14
40
PART I
17 Dollery CM, Sullivan ID, Bauraind O et al. Thrombosis and embolism in long-term central venous access for parenteral nutrition. Lancet 1994; 344: 1043–1045. 18 Verso M, Agnelli G, Venous thromboembolism associated with long-term use of central venous catheters in cancer patients. J Clin Oncol 2003; 21: 3665–3675. 19 Baarslag HJ, van Beek EJ, Koopman MM, Reekers JA. Prospective study of color duplex ultrasonography compared with contrast venography in patients suspected of having deep venous thrombosis of the upper extremities. Ann Intern Med 2002; 136: 865–872. 20 Prandoni P, Bernardi E. Upper extremity deep vein thrombosis. Curr Opin Pulm Med 1999; 5: 222–226.
21 Martinelli I, Cattaneo M, Panzeri D, Taioli E, Mannucci PM. Risk factors for deep venous thrombosis of the upper extremities. Ann Intern Med 1997; 126: 707–711. 22 Otten TR, Stein PD, Patel KC, Mustafa S, Silbergleit A. Thromboembolic disease involving the superior vena cava and brachiocephalic veins. Chest 2003; 123: 809– 812. 23 Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. 24 Joffe HV, Kucher N, Tapson VF, Goldhaber SZ; for the deep vein thrombosis (DVT) FREE Steering Committee. Upper-extremity deep vein thrombosis: a prospective registry of 592 patients. Circulation 2004; 110: 1605–1611.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 12, 2007
8
15:15
CHAPTER 8
Thromboembolic disease involving the superior vena cava and brachiocephalic veins
Isolated venous thromboembolic disease of the brachiocephalic veins and superior vena cava (SVC) was observed in hospitalized adults ≥20 years old in 10 of 34,567 (0.03%) [1] (see Chapter 7). It was accompanied by deep venous thrombosis (DVT) at other sites in an additional 13 patients (0.04%). Previous data are limited to 2 case reports [2, 3]. Thrombosis of the SVC is an uncommon cause of the SVC syndrome [4–6]. Rarely, pulmonary embolism (PE) with the SVC syndrome has been observed at autopsy [7, 8]. It was suggested, however, that SVC thrombosis may pose a significant risk for PE [9]. Brachiocephalic vein thrombosis was observed in 22 patients and SVC thrombosis was observed in 6 patients [1]. The number of patients with brachiocephalic vein thrombosis alone or in combination with SVC thrombosis or with thrombosis of the subclavian or axillary veins is shown in Table 8.1 [1]. The diagnosis was made by contrast venography in 21 patients, and by contrast enhanced spiral CT in 2 patients. All 6 patients with SVC thrombosis showed incomplete obstruction of the SVC. Among the 22 patients with brachiocephalic vein involvement, 13 showed total occlusion or occlusion sufficient to produce collateral veins.
Table 8.2 Predisposing factors among patients with brachiocephalic or superior vena cava thrombosis (n = 23).
Table 8.1 Vessels showing thrombosis (n = 23). SVC only
Two of 23 (8.7%) patients had nonfatal PE, both of which were diagnosed by high-probability ventilation– perfusion lung scans [1]. One patient had brachiocephalic vein thrombosis and, in addition, had bilateral thrombosis of the axillary, subclavian, and internal jugular veins. The PE occurred before the venous thrombosis was diagnosed and treated. The other patient had brachiocephalic vein thrombosis with no other upper body DVT, but in addition had proximal lower extremity DVT. The PE occurred while on treatment for lower extremity DVT. This was the only patient with coincident DVT of the lower extremities. To our knowledge, only one case of PE from brachiocephalic thrombosis has been reported previously [10]. Predisposing factors for patients with thrombosis of the SVC or brachiocephalic vein are shown in Table 8.2 [1]. Among patients who had central venous access lines, 10 had Groshong catheters, 4 had Infusaports, and 1 had a PICC line (peripherally inserted central catheter) (all from Bard Access Systems, Salt Lake City, Utah). An association of use of central venous access lines with thrombosis has been described [11–15]. Three patients with brachiocephalic vein thrombosis had neither cancer nor central venous lines, but had
1
Number
Percent
SVC + brachiocephalic
1
Cancer and central venous access line
14
61
SVC + brachiocephalic + subclavian or axillary
4
Cancer alone
3
13
Brachiocephalic only
9*
Central venous access line alone
1
4
Brachiocephalic + subclavian or axillary
8
External mechanical compression;
1
4
4
17
not cancer * One of these patients also had DVT of a lower extremity.
Cause not apparent
SVC, superior vena cava. Reprinted with permission from Otten et al. [1].
Reprinted with permission from Otten et al. [1].
41
BLUK077-Stein
March 12, 2007
15:15
42
PART I
Table 8.3 Signs and symptoms, SVC and/or brachicephalic vein thrombosis.
2 Kwong T, Leonidas JC, Ilowite NT. Asymptomatic superior vena cava thrombosis and pulmonary embolism in an adolescent with SLE and antiphospholipid antibodies. Clin Exp Rheum 1994; 12: 215–217. 3 Goldstein MF, Nestico P, Olshan AR et al. Superior vena cava thrombosis and pulmonary embolus: association with right atrial mural thrombus. Arch Intern Med 1982; 142: 1726–1728. 4 Gucalp R, Dutcher J. Oncologic emergencies. In: Braunwald E, Fauci AS, Kasper DL et al., eds. Harrison’s Principles of Internal Medicine, 15th edn. McGraw-Hill, New York, 2001: 642–650. 5 Abner A. Approach to the patient who presents with superior vena cava obstruction. Chest 1993; 103: 394S–397S. 6 Salsali M, Cliffton EE. Superior vena caval obstruction with lung cancer. Ann Thorac Surg 1968; 6: 437–442. 7 Maddox A-M, Valdivieso M, Lukeman J et al. Superior vena cava obstruction in small cell bronchogenic carcinoma. Clinical parameters and survival. Cancer 1983; 52: 2165–2172. 8 Ryan JA, Abel RM, Abbott WM et al. Catheter complications in total parenteral nutrition. A prospective study of 200 consecutive patients. N Engl J Med 1974; 290: 757– 761. 9 Adelstein DJ, Hines JD, Carter SG et al. Thromboembolic events in patients with malignant superior vena cava syndrome and the role of anticoagulation. Cancer 1988; 62: 2258–2262. 10 Black MD, French GJ, Rasuli P et al. Upper extremity deep venous thrombosis. Underdiagnosed and potentially lethal. Chest 1993; 103: 1887–1890. 11 Prandoni P, Polistena P, Bernardi E et al. Upper-extremity deep vein thrombosis. Risk factors, diagnosis, and complications. Arch Intern Med 1997; 157: 57–62. 12 Torosian MH, Meranze S, McLean G et al. Central venous access with occlusive superior central venous thrombosis. Ann Surg 1986; 203: 30–33. 13 Haire WD, Lieberman RP, Edney J et al. Hickman catheter- induced thoracic vein thrombosis. Frequency and long-term sequelae in patients receiving high-dose chemotherapy and marrow transplantation. Cancer 1990; 66: 900–908. 14 Gore JM, Matsumoto AH, Layden JJ et al. Superior vena cava syndrome. Its association with indwelling ballooning-tipped pulmonary artery catheters. Arch Intern Med 1984; 144: 506–508. 15 Dollery CM, Sullivan ID, Bauraind O. Thrombosis and embolism in long-term central venous access for parenteral nutrition. Lancet 1994; 344: 1043–1045. 16 Knudson GJ, Wiedmeyer DA, Erickson SJ et al. Color Doppler sonographic imaging in the assessment of upperextremity deep venous thrombosis. Am J Roentgenol 1990; 154: 399–403.
Number (percent) All patients
Brachiocephalic vein
(n = 23)
only* (n = 9)
Arm edema
18 (78)
7 (78)
Pain/discomfort
15 (65)
5 (56)
Head/neck edema
10 (43)
3 (33)
Distended veins
7 (30)
2 (22)
Erythema
5 (22)
2 (22)
2 (9)
0 (0)
1 (4)
1 (11)
CNS symptoms None (incidental finding)
* Patients with accompanying thrombosis of axillary, subclavian, or internal jugular vein were excluded. CNS, central nervous system. Reprinted with permission from Otten et al. [1].
arteriovenous shunts for renal dialysis [1]. There was no evidence of thrombosis of the shunts. One of these patients had subclavian vein thrombosis in addition to thrombosis of the brachiocephalic vein. The other 2 patients had no evidence of thrombosis at other sites. The patient who had SVC thrombosis alone had no clear cause for the thrombosis. She was taking estrogen replacement therapy following menopause. Signs and symptoms observed in all patients with SVC thrombosis and/or brachiocephalic vein thrombosis and in 9 patients who showed involvement only of a brachiocephalic vein are shown in Table 8.3 [1]. In the latter group, the thrombosis caused partial occlusion in 6 of 9 (67%). Three of those with partial occlusion showed collateral vessels. Contrast venography is the most conclusive test [16–20]. Helical computed tomographic phlebography [21], magnetic resonance angiography [19, 20], and gadolinium-enhanced magnetic resonance venography [22] may be useful. Color duplex Doppler ultrasonography cannot image the SVC and proximal segment of the brachiocephalic veins [16, 23].
References 1 Otten TR, Stein PD, Patel KC, Mustafa S, Silbergleit A. Thromboembolic disease involving the superior vena cava and brachiocephalic veins. Chest 2003; 123: 809–812.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 12, 2007
CHAPTER 8
15:15
VTE involving SVC and brachiocephalic veins
17 Gooding GA, Hightower DR, Moore EH et al. Obstruction of the superior vena cava or subclavian veins: sonographic diagnosis. Radiology 1986; 159: 663–665. 18 Schwartz EE, Goodman LR, Haskin ME. Role of CT scanning in the superior vena cava syndrome. Am J Clin Oncol 1986; 9: 71–78. 19 Hartnell GG, Hughes LA, Finn JP et al. Magnetic resonance angiography of the central chest veins. A new gold standard? Chest 1995; 107: 1053–1057. 20 Finn JP, Zisk JHS, Edelman RR et al. Central venous occlusion: MR angiography. Radiology 1993; 187: 245–251.
43
21 Qanadli SD, EL Hajjam M, Bruckert F et al. Helical CT phlebography of the superior vena cava: diagnosis and evaluation of venous obstruction. Am J Roentgenol 1999; 172: 1327–1333. 22 Kroencke TJ, Taupitz M, Arnold R et al. Threedimensional gadolinium-enhanced magnetic resonance venography in suspected thrombo-occlusive disease of the central chest veins. Chest 2001; 120: 1570–1576. 23 Falk RL, Smith DF. Thrombosis of upper extremity thoracic inlet veins: diagnosis with duplex Doppler sonography. Am J Roentgenol 1987; 149: 677–682.
BLUK077-Stein
March 28, 2007
9
15:5
CHAPTER 9
Venous thromboembolic disease in the four seasons
Introduction
26
30
11.0
11.8
12.1
10
12.0
20
O ct −D ec
ep −S ly Ju
un e Ap r− J
Ja
n−
M ar
0
44
There is no meaningful seasonal variation of mortality from PE [13]. From 1980 to 1998, quarterly mortality rates throughout the entire United States, based on data from the United States National Center for Health Statistics, ranged from 0.93 to 1.05 PE deaths/quarter/100,000 population [13] (Table 9.1). Quarterly mortality rates from PE in four regions of the United States (Northeast, South, Midwest, and West) were evaluated separately [13] (Table 9.2). All were within 1 PE death/quarter/100,000 population [13]. Small differences were statistically significant due to the large number of patients evaluated (184,201 death certificates indicating PE as the cause of death). Recognizing that the death certificate diagnosis of fatal or large PE is accurate in only 32–35% of patients [14, 15] we assumed that whatever inaccuracy exists in the death certificates was constant throughout the seasons [13]. Differing observations had been reported previously on seasonal differences of mortality from acute PE. Several investigators reported peak mortality rates
.2
.7 27
28
.0 28
40 Rate per 100,000
.1
An absence of seasonal variation was shown in all regions of the United States, including the Southern region where winters are mild, and the Northeastern and Midwestern regions where seasons are sharply defined [1] (Figure 9.1). These observations were based on 21 years of data from the National Hospital Discharge Survey [1]. The results were based on data in 2,457,000 hospitalized patients with pulmonary embolism (PE) and 5,767,000 hospitalized patients with deep venous thrombosis (DVT). The data apply to patients with broad differences of ethnic, social, and racial backgrounds and to regions with wide differences in climate. This absence of seasonal variation was concordant with results of others who found no seasonal variations in PE [2] or DVT [3–5]. Seasonal variability of thromboembolic disease has been suggested to be present since 1939 [6]. Some showed a peak incidence in winter months for PE [2–8], fatal PE [7], DVT [8, 9], and thromboembolic disease [10]. Others showed a decreased incidence of PE in the winter, or a peak in spring and autumn [11] or peaks in both summer and winter [12].
Mortality from acute PE according to season
DVT PE
Figure 9.1 Rates of diagnosis in hospitalized patients/100,000 population for pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) according to quarter of year. Data are averaged from 1979 to 1999. (Reprinted from Stein et al. [1], with permission from Elsevier.)
BLUK077-Stein
March 28, 2007
CHAPTER 9
15:5
45
Venous thromboembolic disease in the four seasons
Table 9.1 Average quarterly rate of deaths from pulmonary embolism: 1980–1998 (PE deaths/100,000 population). Jan–Mar
Apr–Jun
Jul–Sep
Oct–Dec
All seasons
Northeast
1.11
0.98
0.97
1.03
1.02
Midwest
1.14
1.00
0.99
1.06
1.05
South
1.22
1.06
1.09
1.16
1.13
West
0.65
0.59
0.60
0.65
0.62
All regions
1.03
0.91
0.91
0.98
0.96
Differences between regions: South > Northeast, Midwest, West (P < 0.05); Northeast, Midwest > West (P < 0.05). Differences between seasons: Jan–Mar > Apr–Jun, Jul–Sep, Oct–Dec (P < 0.05); Oct–Dec > Apr–Jun, Jul–Sep (P < 0.05). PE, pulmonary embolism. Reprinted with permission from Stein et al. [13].
in the first quarter of the year [16–20], sometimes with overlap in the last quarter [21] and sometimes with second peaks in the third quarter [16, 17]. Others reported peak mortality rates in the second quarter [22–24], sometimes with second peaks in the third and fourth quarter [22], or fourth quarter alone [23]. Some reported peaks only in the third and fourth quarter [25]. Some reported more frequent fatal PE during “fine weather phases” of the year and “at the beginning of fine weather” [26]. Some reported no quarterly varia-
tion [27, 28]. Many of these investigations were based on observations in less than 200 patients [16–18, 20, 21]. The largest investigation included less than 1500 patients [26]. The absence of meaningful seasonal variation of the rate of diagnosis in hospitalized patients [1] and mortality from PE [13] based on data from several thousands of patients indicate that PE is not affected by the season, contrary to reports based on smaller investigations.
Table 9.2 Regions of United States defined according to states and district of Columbia. West
Midwest
South
Northeast
Alaska
Illinois
Delaware
Maine
Arizona
Indiana
Maryland
New Hampshire
California
Iowa
District of Columbia
Vermont
Colorado
Kansas
Virginia
Massachusetts
Hawaii
Michigan
West Virginia
Connecticut
Idaho
Minnesota
North Carolina
Rhode Island
Montana
Missouri
South Carolina
New York
Nevada
Nebraska
Georgia
New Jersey
New Mexico
North Dakota
Florida
Pennsylvania
Oregon
Ohio
Kentucky
Utah
South Dakota
Tennessee
Washington
Wisconsin
Wyoming
Alabama Mississippi Arkansas Louisiana Oklahoma Texas
Reprinted from Stein et al. [1], with permission from Elsevier.
BLUK077-Stein
March 28, 2007
15:5
46
PART I
References
15 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154: 159–162. 16 Chau KY, Yuen ST, Wong MP. Seasonal variation in the necropsy incidence of pulmonary thromboembolism in Hong Kong. J Clin Pathol 1995; 48: 578–579. 17 Colantonio D, Casale R, Natali G, Pisqualetti P. Seasonal periodicity in fatal pulmonary thromboembolism. Lancet 1990; 335: 56–57. 18 Gallerani M, Manfredini R, Ricci L et al. Sudden death from pulmonary thromboembolism: chronobiological aspects. Eur Heart J 1992; 13: 661–665. 19 Mobius C, Gunther U, Klinker L, Putzke HP. [Meteoropathologic effects on the development of fatal lung embolism]. Z Gesamte Hyg 1989; 35: 391–392. 20 Manfredini R, Gallerani M, Salmi R, Zamboni P, Fersini C. Fatal pulmonary embolism in hospitalized patients: evidence for a winter peak. J Int Med Res 1994; 22: 85– 89. 21 Wroblewski BM, Siney PD, White R. Fatal pulmonary embolism after total hip arthroplasty. Seasonal variation. Clin Orthop Relat Res 1992; 276: 222–224. 22 Green J, Edwards C. Seasonal variation in the necropsy incidence of massive pulmonary embolism. J Clin Pathol 1994; 47: 58–60. 23 Hackl H. [Environmental effects and pulmonary embolism]. Dtsch Med J 1968; 19: 475–477. 24 Montes Santiago J, Rey Garcia G, Mediero Dominguez A. [Seasonal changes in morbimortality caused by pulmonary thromboembolism in Galicia]. An Med Interna 2003; 20: 457–460. 25 Steiner I, Matejek T. [Pulmonary embolism–temporal aspects]. Cesk Patol 2003; 39: 185–188. 26 Putzke HP, Mobius C, Gunther U, Bargenda M, Dobberphul J. [The incidence of fatal lung emboli with special reference to the underlying disease and the effect of weather]. Z Gesamte Inn Med 1989; 44: 106–110. 27 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. 28 Golin V, Sprovieri SR, Bedrikow R, Salles MJ. Pulmonary thromboembolism: retrospective study of necropsies performed over 24 years in a university hospital in Brazil. Sao Paulo Med J 2002; 120: 105–108.
1 Stein PD, Kayali F, Olson RE. Analysis of occurrence of venous thromboembolic disease in the four seasons. Am J Cardiol 2004; 93: 511–513. 2 Galle C, Wautrecht JC, Motte S et al. The role of season in the incidence of deep vein thrombosis. J Mal Vasc 1998; 23: 99–101. 3 Bounameaux H, Hicklin L, Desmarais S. Seasonal variation in deep vein thrombosis. BMJ 1996; 312: 284–285. 4 Luthi H, Gruber UF. Is there a seasonal fluctuation in the appearance of deep venous thrombosis? Anasth Intensivther Notfallmed 1982; 17(3): 158–160. 5 National Hospital Discharge Survey Multi-year Data File 1979–1999. CD-ROM Series 13, No. 19A. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics, Hyattsville, MD, reissued March 2001. http://www.cdc.gov/nchs/about/major/hdasd/nhds.htm. 6 Oschner A, DeBakey M. Thrombophlebitis and phlebothrombosis. South Surg 1939; 8: 269–290. 7 Lawrence JC, Xabregas A, Gray L, Ham JM. Seasonal variation in the incidence of deep vein thrombosis. Br J Surg 1977; 64: 777–780. 8 Boulay F, Berthier F, Schoukroun G, Raybaut C, Gendreike Y, Blaive B. Seasonal variations in hospital admission for deep vein thrombosis and pulmonary embolism: analysis of discharge data. BMJ 2001; 323: 601–602. 9 Ferrari E, Baudouy M, Cerboni P et al. Clinical epidemiology of venous thromboembolic disease. Results of a French multicentre registry. Eur Heart J 1997; 18: 685– 691. 10 Green J, Edwards C. Seasonal variation in the necropsy incidence of massive pulmonary embolism. J Clin Pathol 1994; 47: 58–60. 11 Bilora F, Manfredini R, Petrobelli F, Vettore G, Boccioletti V, Pomerri F. Chronobiology of non fatal pulmonary thromboembolism. Panminerva Med 2001; 43: 7–10. 12 Coon WW. The spectrum of pulmonary embolism: twenty years later. Arch Surg 1976; 111: 398–402. 13 Stein PD, Kayali F, Beemath A et al. Mortality from acute pulmonary embolism according to season. Chest 2005; 128: 3156–3158. 14 Dismuke SE, VanderZwaag R. Accuracy and epidemiological implications of the death certificate diagnosis of pulmonary embolism. J Chronic Dis 1984; 37: 67–73.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
April 6, 2007
10
15:8
CHAPTER 10
Regional differences in the United States of rates of diagnosis of pulmonary embolism and deep venous thrombosis and mortality from pulmonary embolism
Regional rates of diagnosis of pulmonary embolism and deep venous thrombosis The Western region of the United States from 1979 to 2001 showed lower rates of diagnosis of deep venous thrombosis (DVT) and venous thromboembolism (VTE) in hospitalized patients than any other region [1] (Table 10.1, Figure 10.1). The rates of diagnosis of DVT and VTE were lower in the Western region than in other regions from 1979 to 1989 and remained lower from 1990 to 2001 (Table 10.1). Rate ratios of the rates of the diagnosis of DVT, PE, and VTE comparing the Western region to other regions ranged from 0.65 to 0.87 [1] (Table 10.2). Rates of diagnosis were based on data from The National Hospital Discharge Survey [2]. Population estimates were from the United States Bureau of the Census [3]. Regions of the United States were defined by the National Hospital Discharge Survey (see Table 10.2, Chapter 9). In Caucasians, from 1979 to 2001, the rates of diagnosis of DVT and VTE were lower in the Western region than all other regions and the rate of diagnosis of PE was lower in the West than other regions except the Midwest [1]. In African Americans, the rates of diagnosis of PE, DVT, and VTE were lower in the West than in the Midwest (Table 10.3). In both men and women, the rates of diagnosis of DVT and VTE were lower in the West than any other region [1] (Table 10.4). Within each region, the rates of diagnosis of DVT and VTE in men were lower than in women [1] (Table 10.4).
In patients ≥65 years, rates of diagnosis of DVT and VTE were lower in the Western region than other regions, but there was only a trend toward a lower rate of PE in the Western region [1]. In the Western region, rates of DVT in men and women ≥65 years were comparable, although in the Midwestern and Southern regions, rates of DVT were higher in women ≥65 years than in men ≥65 years. Caucasians ≥65 years of age and <65 years of age, from 1979 to 2001, had lower rates of diagnosis of DVT and VTE in the Western region than in other regions (P < 0.01 to P < 0.001). African Americans ≥65 years showed no regional differences in rates of diagnosis. Younger African Americans (aged <65 years) showed lower rates of DVT, PE, and VTE in the West than the Midwest [1]. Relatively low rates of diagnosis of PE and DVT were observed on the Pacific and Atlantic coasts of the United States from 1986 to 1989 in patients ≥65 years, based on a sample of Medicare enrollees [4].
Regional mortality rates from PE The PE mortality rate from 1979 to 1998 was lower in the Western region than any other region [1] (Figure 10.2). It was lower in the Western region in both men and women and in African Americans and Caucasians (all P < 0.001). Rate ratios for rates of mortality comparing the Western region to other regions ranged from 0.55 to 0.60. In patients ≥65 years and patients <65 years, the mortality rates were lower in the Western region in both Caucasians and African
47
BLUK077-Stein
April 6, 2007
15:8
48
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 10.1 Rate of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region. Year
West
Midwest
South
Northeast
Deep venous thrombosis/100,000/yr 1979–1989
88
116
101
114
1990–2001
78
131
120
138
1979–2001
83
124
112
127
Pulmonary embolism/100,000/yr 1979–1989
48
59
51
55
1990–2001
34
46
41
50
1979–2001
40
52
46
53
Venous thromboembolism/100,000/yr 1979–1989
128
164
145
160
1990–2001
102
163
150
173
1979–2001
113
163
147
167
16 7
VTE DVT
53
40
50
46
100
52
83
112
124
150
127
11 3
200
PE
or
th
ea
st
So ut h N
M
id
w
es
es t
t
0 W
DX/100,000/yr
14 7
16 3
Deep venous thrombosis: 1979–1989: West < Northeast, Midwest (P < 0.001); West < South (P < 0.01) 1990–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–1989: South < Northeast, Midwest (P < 0.01) 1990–2001: South < Northeast (P < 0.05) 1979–2001: South < Northeast, Midwest (P < 0.01). Pulmonary embolism: 1990–2001: West < Northeast, Midwest (P < 0.001); West < South (P < 0.05) 1979–2001: West < Northeast, Midwest (P < 0.001) 1990–2001: South < Northeast (P < 0.05). Venous thromboembolism: 1979–1989: West < Northeast, Midwest (P < 0.001); West < South (P < 0.05) 1990–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–1989: South < Northeast, Midwest (P < 0.05) 1990–2001: South < Northeast (P < 0.01) 1979–2001: South < Northeast, Midwest (P < 0.01). Comparisons with probabilities of P > 0.05 were excluded. Reprinted from Stein et al. [1], with permission from Elsevier.
Figure 10.1 Rates of diagnosis (Dx)/100,000 population/year of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) according to region from 1979 to 2001. (Reprinted from Stein et al. [1], with permission from Elsevier.)
BLUK077-Stein
April 6, 2007
CHAPTER 10
15:8
Americans (P < 0.01 to P < 0.001). The mortality rate in women was higher than in men within all regions (P = 0.04 to P = 0.004). The rate ratios of mortality of women to men within each region ranged from 1.09 to 1.16. Death from PE was obtained from the United States Bureau of the Census records. The Western region, from 1990 to 2001, had a higher percentage of Asian Americans-Pacific Islanders (8.2%) than the Northeastern (3.2%), Midwestern (1.6%), or Southern (1.7%) regions [1]. Racial categories for Asian-Pacific Islanders did not exist in census data before 1990 [5]. Data were insufficient to make regional comparisons with Asian-Pacific Islanders. Men ≥65 years constituted 9.5% of the population in the Western region and 10.6–11.5% in other regions. Women ≥65 years constituted 12.1% of the population in the Western region and 13.8–15.4% in other regions. The rate ratio of women to men aged ≥65 years (1.3) was the same in all regions.
Table 10.2 Rate ratios of rates of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region. Year
49
Regional differences of VTE in the United States
West/Northeast
West/Midwest
West/South
Deep venous thrombosis/100,000/yr 1979–1989
0.77
0.76
0.87
1990–2001
0.57
0.60
0.65
1979–2001
0.65
0.66
0.74
Pulmonary embolism/100,000/yr 1979–1989
0.87
0.82
0.94
1990–2001
0.67
0.74
0.82
1979–2001
0.76
0.77
0.87
Venous thromboembolism/100,000/yr 1979–1989
0.80
0.78
0.88
1990–2001
0.59
0.63
0.68
1979–2001
0.68
0.69
0.77
Reprinted from Stein et al. [1], with permission from Elsevier.
Table 10.3 Rates of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region and race (1979–2001). Race
West
Midwest
South
Northeast
Deep venous thrombosis/100,000/yr White patients
68
95
109
122
Black patients
75
113
96
96
Pulmonary embolism/100,000/yr White patients
32
39
44
52
Black patients
38
50
39
40
Venous thromboembolism/100,000/yr White patients
92
125
143
162
Black patients
106
150
128
126
Deep venous thrombosis: Black patients: West < Midwest (P < 0.001) White patients: West < Midwest, South, Northeast (P < 0.001); Midwest < Northeast (P < 0.001); Midwest < South (P < 0.01); South < Northeast (P < 0.01). Pulmonary embolism: Black patients: West < Midwest (P < 0.05) White patients: West < Northeast (P < 0.001); West < South (P < 0.01); Midwest < Northeast (P < 0.01); South < Northeast (P < 0.05). Venous thromboembolism: Black patients: West < Midwest (P < 0.001); South, Northeast < Midwest (P < 0.05) White patients: West < Midwest, South, Northeast (P < 0.001); Midwest < Northeast (P < 0.001); Midwest < South (P < 0.01); South < Northeast (P < 0.01). Comparisons with probabilities of P > 0.05 were excluded. Reprinted from Stein et al. [1], with permission from Elsevier.
BLUK077-Stein
April 6, 2007
15:8
50
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 10.4 Rates of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region and sex (1979–2001). Sex
West
Midwest
South
Northeast
Deep venous thrombosis/100,000/yr Men
76
107
90
114
Women
90
143
134
141
Pulmonary embolism/100,000/yr Men
37
48
41
48
Women
43
56
50
57
Venous thromboembolism/100,000/yr Men
104
143
121
148
Women
123
185
174
185
Deep venous thrombosis: Men: West < Midwest, Northeast (P < 0.001); West < South (P < 0.05); South < Northeast (P < 0.001); South < Midwest (P < 0.01) Women: West < Midwest, South, Northeast (P < 0.001). Pulmonary embolism: Men: West < Midwest, Northeast (P < 0.05) Women: West < Midwest, Northeast (P < 0.001); West < South (P < 0.05). Venous thromboembolism: Men: West < Midwest, Northeast (P < 0.001); West < South (P < 0.01); South < Northeast, Midwest (P < 0.001) Women: West < Midwest, South, Northeast (P < 0.001). Comparisons with probabilities of P > 0.05 were excluded. Reprinted from Stein et al. [1], with permission from Elsevier.
PE deaths/100,000/yr
Kniffin and associates, in a population of patients ≥65 years of age, showed lower age adjusted rates of diagnosis of PE in women than men and a tendency toward higher rates of DVT in women than men [4]. We showed comparable rates of diagnosis of PE in men and women ≥65 years of age in each of the regions and comparable rates of DVT in the Western and
10 8 6 4
4.2
4.5
4.1
Midwest
South
Northeast
2.5
2 0 West
Figure 10.2 Mortality rates from PE (deaths from PE/100,000 population/year) from 1979 to 1998. The mortality rate in the Western region was lower than the mortality rate in all other regions (P < 0.001). (Reprinted from Stein et al. [1], with permission from Elsevier.)
Northeastern regions, but higher rates of DVT in elderly women than elderly men in the Midwestern and Southern regions [1]. Throughout the United States for patients of all ages, the rate of diagnosis of DVT was higher in women than men [6], but in elderly patients the rates of DVT and PE were comparable in men and women [7]. A somewhat younger population would have contributed to the lower rates of DVT and VTE and the lower mortality rate in the Western region [7]. However, such lower rates were observed in patients ≥65 years as well. A higher percentage of Asian Americans and/or Pacific Islanders in the Western region than in other regions would also have contributed to the lower rates of diagnosis and lower mortality rate in the Western region, because the incidences of PE and of DVT are lower in Asian Americans than in African Americans or Caucasians [8–10]. However, lower rates of diagnosis of DVT, PE, and VTE were shown in Caucasians in the Western region and lower mortality rates from PE were shown in Caucasians and African Americans in the Western region. The observed difference
BLUK077-Stein
April 6, 2007
CHAPTER 10
15:8
Regional differences of VTE in the United States
in regional rates of diagnosis of DVT and VTE are unlikely to be related to differences in climate. We observed no seasonal variation in the rate of diagnosis of DVT, PE, or VTE in any of the regions, including the Southern region, where winters are mild, and the Northeastern and Midwestern regions, where seasons are sharply defined [5] (see Chapter 9). Lilienfeld and Godbold, based on data from 1980 to 1984, showed lower mortality rates from PE in the Pacific and Mountain regions than other parts of the country [11].
References 1 Stein PD, Kayali F, Olson RE. Regional differences in rates of diagnosis and mortality of pulmonary thomboembolism. Am J Cardiol 2004; 93: 1194–1197. 2 National Hospital Discharge Survey 1979–2001 Multiyear Public-use data file documentation. US Department of Health and Human Services, Public Health Service, National Center for Health Statistics. http://www.cdc.gov/ nchs/about/major/hdasd/nhds.htm. 3 Bureau of the Census, Department Of Commerce, United States Department of Health and Human Services (US DHHS) Centers for Disease Control and Prevention (CDC), CDC WONDER On-line Database. http:// wonder.cdc.gov/census.shtml.
51
4 Kniffin WD, Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA, Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994; 154: 861–866. 5 Stein PD, Kayali F, Olson RE. Analysis of venous thromboembolic disease in the four seasons. Am J Cardiol 2004; 93: 511–513. 6 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694. 7 Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260–2265. 8 Klatsky AL, Armstrong MA, Poggi J. Risk of pulmonary embolism and/or deep venous thrombosis in AsianAmericans. Am J Cardiol 2000; 85: 1334–1337. 9 White RH, Zhou H, Romano PS. Incidence of idiopathic deep venous thrombosis and secondary thromboembolism among ethnic groups in California. Ann Intern Med 1998; 128: 737–740. 10 Stein PD, Kayali F, Olson RE, Milford, CE. Pulmonary thromboembolism in Asian-Pacific Islanders in the United States: analysis of data from the National Hospital Discharge Survey and the United States Bureau of the Census. Am J Med 2004; 116: 435–442. 11 Lilienfeld DE, Godbold JH. Geographic distribution of pulmonary embolism mortality rates in the United States, 1980 to 1984. Am Heart J 1992; 124: 1068–1072.
BLUK077-Stein
March 28, 2007
11
15:8
CHAPTER 11
Venous thromboembolism in the elderly patients than in younger patients (20–69 years). The 21-year trends for the diagnosis of DVT according to age are shown in Figure 11.1b [1]. The elderly population showed the greatest increase in the 1990s. The diagnosis of pulmonary embolism (PE) in patients 70 years or older was 6.2 than the rate in younger patients (Figure 11.2a) [1]. Contrary to DVT, the rate of diagnosis of PE decreased from 370 PE/100,000 population in 1979 to 254 PE/100,000 population in 1990 (Figure 11.2a) and then remained constant.
Rates of diagnosis and trends in the diagnosis of deep venous thrombosis and pulmonary embolism in the elderly Deep venous thrombosis (DVT) in elderly patients (70 years or older) increased 44% from 454 DVT/100,000 population in 1990 to 655 DVT/100,000 population in 1999 (Figure 11.1a) [1]. Deep venous thrombosis was diagnosed 4.7 times more frequently in elderly (a) DVT DVT/100,000 population
800
Age >70
600 400 Age 20−69
200 0
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
(b) DVT: age distribution
DVT/100,000 population
1000 800
70−89
600 400
60−69 50−59 40−49 20−39
200 0 1999
1997
1995
1993
52
1991
1989
1987
1985
1983
1981
1979
Year
Figure 11.1 (a) Among elderly patients (70 years or older), trends over 21 years in the rate of diagnosis of deep venous thrombosis (DVT) was constant from 1979 to 1990 and increased from 1990 to 1999. In younger patients (20–69 years), there was a slight but significant decline in the rate of diagnosis of DVT between 1979 and 1990. The rate then increased somewhat between 1990 and 1999. (b) From 1979 to 1990, the rate of diagnosis of DVT was constant in patients aged 60–69 and 70–89 years. During this time interval, the rate decreased in younger age groups. From 1990 to 1999, the rate of diagnosis of DVT increased in all age groups. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
BLUK077-Stein
March 28, 2007
CHAPTER 11
15:8
53
Venous thromboembolism in the elderly (a) PE
200 100
Age 20−69
0 1999
1997
1995
1993
1991
1989
1987
1985
1983
PE/100,000 population
500
PE: age distribution
400 300
70−89
200 60−69 50−59 40−49 20−39
100 0 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
Year
DVT/100,000 population/yr
Smaller changes were observed in patients aged 20– 69 years from 67 PE/100,000 population in 1979 to 30 PE/100,000 population in 1990 (Figure 11.2b). The rate increased somewhat between 1990 and 1999. The rates of diagnosis of DVT or PE in elderly men and women, and elderly black and white patients were comparable [1].
Figure 11.3 Deep venous thrombosis (DVT)/100,000 population/year, diagnosed at hospital discharge, shown according to age for the year 1999. (Data from Stein et al. [1, 2].)
1981
Year
(b)
1979
Figure 11.2 (a) Among elderly patients (70 years or older), trends over 21 years in the rate of diagnosis of pulmonary embolism (PE) decreased from 1979 to 1990 and then remained constant from 1990 to 1999. In younger patients there was a slight but significant decline in the rate of diagnosis of PE between 1979 and 1990. The rate then increased somewhat between 1990 and 1999. (b) Trends over 21 years in the rate of diagnosis of PE in patients as shown by age group. In all age groups, the rate of diagnosis of PE significantly decreased from 1979 to 1990. From 1990 to 1999, the rate of diagnosis of PE remained constant in all age groups except age 20–39, which showed a slight increase. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
Age >70
300
1979
PE/100,000 population
400
Deep venous thrombosis, based on hospital discharges, was diagnosed in 700 patients/100,000 population/year aged 70–89 years, 300/100,000 population/year aged 60–69 years, and lower proportions of the population of younger people [1, 2] (Figure 11.3). Comparing the rate of DVT at each decade of age with the rate at age 20–29, the rate ratio increased 700
700 600 500 400
300
300 200
200 100 0
<5
10
30
60
100
0−14 15−19 20−29 30−39 40−49 50−59 60−69 70−89
Age groups (years)
BLUK077-Stein
March 28, 2007
15:8
54
PART I
40
DVT 1980−1999
Prevalence, risks, and prognosis of PE and DVT
PE 1980−1999
40
.7
11
1
Rate ratio
Rate ratio
.7
20
1
6.
2.
6
10
7 7
90−99
80−89
70−79
60−69
250
50−59
300
300
40−49
exponentially up to age 89 (Figure 11.4) [1]. In patients aged 70–79 years and 80–89 years, the rate ratios for DVT were 12.7 and 17.7, respectively. Pulmonary embolism, based on hospital discharges, was diagnosed in 300 patients/100,000 population/year aged 70–89 years, 100/100,000 population/year aged 60–69 years, and lower proportions of the population of younger people [1, 2] (Figure 11.5). Comparing the rate of PE at each decade of age with the rate at age 20–29, the rate ratio increased exponentially up to age 89 (Figure 11.6) [1]. In patients aged 70–79 years and 80–89 years, the rate ratios for
30−39
Figure 11.4 Rate ratios for the rate of diagnosis of deep venous thrombosis (DVT), comparing each decade of age with the rate at age 20–29 years. Rate ratios were averaged over 21 years. Between ages 20–29 and 80–89 years, the rate ratios for the diagnosis of DVT increased exponentially. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
0 20−29
90−99
80−89
70−79
60−69
50−59
40−49
30−39
20−29
Age groups (years)
PE/100,000 population/yr
1
1.
7
3.
1. 0
Age groups (years) Figure 11.6 Rate ratios for the rate of diagnosis of pulmonary embolism (PE), comparing each decade of age with the rate at age 20–29 years. Rate ratios were averaged over 21 years. Between ages 20–29 and 80–89 years, the rate ratios for the diagnosis of PE increased exponentially. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
PE were 20.6 and 27.9, respectively. There was no step change of rate of diagnosis of PE or DVT at any age. The recommended approach to the diagnosis of PE is applicable to elderly patients [3]. An abundance of literature documents that the risk of venous thromboembolism increases with age [1, 4–19].
Case fatality rate From 1989 to 1998, the estimated case fatality rate from PE in the United States was 7.7 PE deaths/100 cases of PE [20]. The estimated case fatality rate from PE is strongly age-dependent, and increased exponentially with age from 3.6% in patients aged 25–34 years to 17.4% in patients aged >85 years [20] (see Chapter 3).
Antithrombotic prophylaxis and age
200 150
100
100
80
50 0
.3
.9 .6
12 8. 4.
10
20
.7
20
27
.9
17
17
30
27
30
<1 0−14
4
10
20
20
15−19 20−29 30−39 40−49 50−59 60−69 70−89
Age groups (years)
Figure 11.5 Pulmonary embolism (PE)/100,000 population/ year, diagnosed at hospital discharge, shown according to age for the year 1999. (Data from Stein et al. [1, 2].)
Recommendations for antithrombotic prophylaxis in patients undergoing surgical procedures are partially based on the age of the patient [21]. Patients less than 40 years have been considered at low-risk for venous thromboembolism for specific in-hospital surgical groups, such as general surgical patients, and age as a risk factor has become more important at 40 years or older with regard to thromboprophylaxis [21]. This led to the general belief that age as a risk factor has a
BLUK077-Stein
March 28, 2007
15:8
Venous thromboembolism in the elderly
55
break point at the age of 40 years. For patients aged 30–39 years, there is almost a 2-fold increase in the risk of DVT or PE compared with younger patients [1]. The concept that the elderly are at the greatest need for thromboprophylaxis is emphasized by these data which show an 18–28-fold increase in the risk of DVT or PE for patients ≥70 years compared with those 20–29 years of age [1].
the PE in 44% [25]. Comparable percentages of patients in the younger age groups were immobilized or underwent surgery before the PE. Malignancy was more frequent among patients ≥70 years (26%) than among the patients <40 years old (2%), but patients 40–69 years had malignancy nearly as frequently as did patients ≥70 years old (24%). Estrogen use was infrequent among female patients ≥70 years old. Its use among female patients <40 years of age preceded PE in 35% and childbirth in women <40 years of age preceded PE in 25% [25].
CHAPTER 11
Diagnosis of acute PE The diagnosis of PE among elderly patients has been thought to be particularly difficult because the expected signs and symptoms may be absent or ignored [22–24]. This did not seem to be the case in the experience of the Prospective Investigation of the Pulmonary Embolism Diagnosis (PIOPED I) [25] and, in general, not in the experience of PIOPED II [26]. In PIOPED II, however, dyspnea and tachypnea were less frequent in patients ≥70 years old than in patients <40 years old [26]. The typical signs and symptoms known to occur among younger patients were common among elderly patients [25, 26]. In the absence of these signs and symptoms, unexplained radiographic abnormalities were important diagnostic clues [25]. When the diagnosis of PE is uncertain, computed tomographic (CT) angiography can be performed safely in elderly patients, providing renal function is adequate [3]. Renal failure was a problem among elderly patients who underwent conventional angiography [25].
Predisposing factors according to age Among 72 patients ≥70 years old in PIOPED I, 67% were immobilized before the PE, and surgery preceded
Syndromes of PE according to age The usual syndromes of PE, among all patients, irrespective of prior cardiopulmonary disease, characterized by (1) hemoptysis or pleuritic pain, (2) isolated dyspnea, or (3) circulatory collapse were observed in PIOPED I among elderly patients [25]. However, 11% of patients ≥70 years of age in PIOPED I and 15% in PIOPED II did not show these syndromes [25, 26]. In PIOPED I they were identified on the basis of unexpected radiographic abnormalities, which may have been accompanied by tachypnea or a history of thrombophlebitis [25]. Unexplained radiographic abnormalities may be an important clue to the diagnosis of PE, particularly among elderly patients in whom the expected signs and symptoms are absent, as has been previously observed [22]. Among all patients with PE in PIOPED II, the syndrome of hemoptysis or pleuritic pain was less frequent in patients ≥70 years than in patients <40 years (P < 0.025) and isolated dyspnea was more frequent (P < 0.05) (Table 11.1). In PIOPED II, among patients with no prior cardiopulmonary disease, the prevalence of the various syndromes was similar among all age groups (Table 11.2).
Table 11.1 Syndromes of acute PE according to age: all patients with pulmonary embolism: PIOPED II. Syndromes Hypotension, LOC
≥70 yr (n = 55) [n (%)]
40–69 yr (n = 106) [n (%)]
<40 yr (n = 31) [n (%)]
2 (4)
10 (9)
3 (10)
Hemoptysis or pleuritic pain
15 (27)
43 (41)
16 (52)
Isolated dyspnea
30 (55)
39 (37)
10 (32)
8 (15)
14 (13)
2 (6)
No syndrome
The syndrome of hemoptysis or pleuritic pain was less frequent in patients ≥70 years than in patients <40 years (P <0.025) and isolated dyspnea was more frequent (P <0.05). PE, pulmonary embolism; LOC, loss of consciousness. Data from Stein et al. [26].
BLUK077-Stein
March 28, 2007
15:8
56
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 11.2 Syndromes of acute PE according to age: patients with no prior cardiopulmonary disease: PIOPED II. ≥70 yr (n = 35) [n (%)]
Syndromes Hypotension, LOC
40–69 yr (n = 76) [n (%)]
<40 yr (n = 22) [n (%)]
2 (6)
7 (9)
2 (9)
Hemoptysis or pleuritic pain
11 (31)
30 (39)
10 (45)
Isolated dyspnea
17 (49)
29 (38)
8 (36)
6 (17)
11 (14)
2 (9)
No syndrome
Differences comparing age groups not significant. PE, pulmonary embolism; LOC, loss of consciousness. Data from Stein et al. [26].
Symptoms according to age Dyspnea was the most frequent symptom in all patients with PE, occurring in 78 and 75% of ≥70 years old in PIOPED I and PIOPED II [25, 26] (Tables 11.3) and in 66% of patients ≥70 years old with no prior cardiopulmonary disease (Table 11.4). Pleuritic pain occurred in 51 and 33% of all patients with PE in PIOPED I and PIOPED II. Pleuritic pain occurred more frequently than hemoptysis in all age groups. In gen-
eral, all symptoms occurred with equal frequency in the different age groups, but there were occasional exceptions, as noted in Tables 11.3–11.4
Signs according to age Tachypnea (respiratory rate ≥20/min) was the most frequent sign in all age groups with PE [25, 26] (Table 11.5) and in those with no prior cardiopulmonary disease (Table 11.6). All signs, occurred with
Table 11.3 Symptoms in all patients with acute pulmonary embolism according to age: PIOPED I and PIOPED II. ≥70 yr
Dyspnea
<40 yr
40–69 yr
PIOPED I
PIOPED II
PIOPED I
PIOPED II
PIOPED I
(n = 72)
(n = 53–55)
(n = 144)
(n = 100–106)
(n = 44)
PIOPED II (n = 30–31)
(%)
(%)
(%)
(%)
(%)
(%) 87
78
75
78
79
82
Dyspnea (rest or exertion)
—
75
—
79
—
87
Dyspnea (at rest)
—
60
—
59
—
71
Dyspnea (exertion only)
—
13
—
19
—
13
Orthopnea (>2-pillow)
—
31
—
40
—
32
51
33
58
53
70
53
Cough
35
44
42
46
45
35
Purulent
—
11
—
10
—
6
Clear
—
9
—
11
—
6
Nonproductive
—
20
—
21
—
23
Leg swelling
35
26
33
50
14
33
Leg pain
31
28
26
45
20
48
Palpitation
13
—
15
—
9
—
Wheezing
10
25
12
32
16
35
Angina-like pain
10
13
13
18
7
23
8
4
4
6
32*
10
Pleuritic pain
Hemoptysis
*P <0.01, ≥70 years vs. <40 years; P <0.001, 40–69 years vs. <40 years among patients in PIOPED I. Data from Stein et al. [25, 26].
BLUK077-Stein
March 28, 2007
15:8
Table 11.4 Symptoms in patients with PE and no preexisting cardiac or pulmonary disease according to age: PIOPED II.
Syndromes
≥70 yr
40–69 yr
<40 yr
(n = 30–35)
(n = 72–76)
(n = 21–22)
(%)
(%)
(%)
Dyspnea Dyspnea (rest or exertion)
66
75
82
Dyspnea (at rest)
49
55
68
Dyspnea (exertion only)
14
19
9
Orthopnea (≥2-pillow)
23
33
18
Pleuritic pain
35
46
45
Chest pain (not pleuritic)
18
20
23
Cough
29
36
36
Hemoptysis
3
5
9
Purulent
6
5
5
Clear Nonproductive Wheezing
6
5
5
11
21
27
9
24
27
Calf swelling
26
49*
24
Thigh swelling
11
5
9
0
0
5
Calf pain
26
47
41
Thigh pain
11
24
19
0
3
5
Thigh swelling, no calf swelling
Thigh pain, no calf pain
*P = 0.025, age ≥70 years vs. age 40–69 years, P = 0.048, age 40–69 years vs. age <40 years. All other differences between age groups not significant. Data from Stein et al. [26].
Table 11.5 Signs in all patients with acute pulmonary embolism according to age: PIOPED I and PIOPED II. ≥70 yr
<40 yr
40–69 yr
PIOPED I
PIOPED II
PIOPED I
PIOPED II
PIOPED I
PIOPED II
(n = 72)
(n = 52–55)
(n = 144)
(n = 101–106)
(n = 44)
(n = 29–31)
(%)
(%)
(%)
(%)
(%)
(%)
Tachypnea (>20/min)
74
51
69
58
82
60
Rales
65
26
61
17
41
26
Tachycardia (>100/min)
29
21
26
24
32
43
Increased P2
15
7
20
19
34
18
Deep venous thrombosis
15
47
17
49
9
42
Diaphoresis
8
2
10
7
18
0
Wheezes
8
4
10
2
5
6
Temperature >38.5◦ C
7
0
5
2
14
3
Third heart sound
7
—
6
—
5
—
Right ventricular lift
—
4
—
7
—
0
Jugular venous distention
—
19
—
12
—
10
Rhonchi
—
6
—
6
—
0
Decreased breath sounds
—
29
—
19
—
13
Pleural friction rub
6
2
5
1
0
0
Homans’ sign
4
—
2
—
2
—
Cyanosis
3
0
3
1
2
0
Differences between age groups in both PIOPED I and PIOPED II were not significant. Data from Stein et al. [25, 26].
57
BLUK077-Stein
March 28, 2007
15:8
58
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 11.6 Signs in patients with PE and no preexisting cardiac or pulmonary disease according to age: PIOPED II. ≥70 yr
40–69 yr
<40 yr
(n = 30–35)
(n = 71–76)
(n = 20–22)
(%)
(%)
(%)
Tachypnea (≥20/min)
47
59
52
Tachycardia (>100/min)
23
21
38
Diaphoresis
3
3
0
Cyanosis
0
0
0
Signs General
Temperature >38.5◦ C (>101.3◦ F)
0
3
5
Cardiac examination (any)
23
22
14
Increased P2*
12
15
19
0
6
0
Jugular venous distension
20
14
5
Lung examination (any)
43
22
27
Rales (crackles)
Right ventricular lift†
26
14
19
Wheezes
3
1
0
Rhonchi
0
3
0
23
14
18
0
0
0
Decreased breath sounds Pleural friction rub DVT Calf or thigh
51
48
36
Calf only
34
31
32
Calf and thigh
14
16
5
3
1
0
Thigh only
All differences comparing age groups not significant. * Data in 26 patients ≥70 years, 61 patients 40–69 years, 16 patients ≤40 years. † Data in 30 patients ≥70 years, 64 patients 40–69 years, 16 patients <40 years. P2, pulmonary component of second heart sound; DVT, deep venous thrombosis. Data from Stein et al. [26].
similar frequency among all age groups (Tables 11.5 and 11.6).
Chest radiograph The chest radiograph among all patients with PE in PIOPED I was normal in 4% ≥70 years old [25]. Atelectasis or pulmonary parenchymal abnormalities were the most frequent radiographic abnormalities among all age groups. All radiographic abnormalities occurred with a comparable frequency among all age groups (Table 11. 7).
Combinations of symptoms and signs and radiographic abnormalities Even among patients ≥70 years old, a combination of nonspecific symptoms and signs that typically occur
with PE was present in the great majority of patients with PE [25, 26]. Dyspnea or tachypnea among all patients with PE occurred in 92% of patients ≥70 years of age in PIOPED I, but in only 77% in PIOPED II (Table 11.8) and in 70% with no prior cardiopulmonary disease in PIOPED II (Table 11.9). Among all patients with PE, dyspnea or tachypnea or pleuritic pain occurred in 94% of patients ≥70 years old in PIOPED I and in 87% in PIOPED II [25, 26]. If signs of DVT were added, 94% of patients ≥70 years in PIOPED I and 96% in PIOPED II had 1 or more of these findings (Tables 11.8). Combinations of signs and symptoms occurred with similar frequency in all age groups, except dyspnea or tachypnea among patients with PE and no prior cardiopulmonary disease were less frequent in patients 70 years than in patients <40 years [26] (Tables 11.8 and 11.9).
BLUK077-Stein
March 28, 2007
CHAPTER 11
15:8
59
Venous thromboembolism in the elderly
Table 11.7 Chest radiograph in all patients with acute pulmonary embolism according to age: PIOPED I. ≥70 yr
40–69 yr
<40 yr
(n = 72) (%)
(n = 144) (%)
(n = 44) (%)
4
8
Atelectasis or pulmonary parenchymal abnormality
Normal
71
69
14 64
Pleural effusion
57
46
45
Pleural-based opacity
42
34
43
Prominent central pulmonary artery
29
20
11
Elevated diaphragm
28
27
18
Cardiomegaly
22
17
14
Decreased pulmonary vascularity
19
22
20
Pulmonary edema
13
12
7
Westermark’s sign
7
8
0
Differences among age groups were not significant. Westermark’s sign = prominent central pulmonary artery and decreased pulmonary vascularity. Reprinted from Stein et al. [25], with permission from the American College of Cardiology Foundation.
The electrocardiogram according to age Nonspecific ST segment or T wave changes were the most frequent electrocardiographic abnormalities in all patients with PE, either or both occurring in 56% of patients ≥70 years old and with nearly the same
frequency in younger patients [25]. With the exception of left anterior hemiblock (left axis deviation) among patients ≥70 years of age, other electrocardiographic abnormalities occurred in 12% or fewer patients in all age groups in PIOPED I. No differences in the frequency of occurrence of any ECG abnormalities
Table 11.8 Combinations of signs and symptoms in all patients with acute pulmonary embolism according to age: PIOPED I and PIOPED II. ≥70 yr
<40 yr
40–69 yr
PIOPED I
PIOPED II
PIOPED I
PIOPED II
PIOPED I
PIOPED II
(n = 72)
(n = 52)
(n = 144)
(n = 104)
(n = 44)
(n = 30)
(%)
(%)
(%)
(%)
(%)
(%)
Dyspnea or tachypnea
92
77
90
88
95
Dyspnea or tachypnea or hemoptysis
92
—
91
—
98
93 —
Dyspnea or tachypnea or pleuritic pain*
94
87
98
93
100
Dyspnea or tachypnea or signs of deep
92
—
91
—
98
97
94
96
99
96
100
100
—
97
—
98
—
100
—
99
—
100
—
—
venous thrombosis Dyspnea or tachypnea or pleuritic pain or
100
signs of deep venous thrombosis* Dyspnea or tachypnea or radiographic atelectasis or parenchymal abnormality Dyspnea or tachypnea or pleuritic pain or radiographic atelectasis or parenchymal abnormality* *The addition of hemoptysis did not improve the sensitivity of the combination for the detection of pulmonary embolism. Tachypnea = respiratory rate ≥20/min. Differences among age groups in PIOPED I and PIOPED II were not significant. Data from Stein et al. [25, 26].
BLUK077-Stein
March 28, 2007
15:8
60
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 11.9 Combinations of signs and symptoms in patients with acute pulmonary embolism and no prior cardiopulmonary disease according to age: PIOPED II. ≥70 yr (n = 33) (%)
40–69 yr (n = 75) (%)
<40 yr (n = 21) (%)
Dyspnea or tachypnea (≥20/min)
70*
88
90
Dyspnea or tachypnea (≥20/min) or
85
93
95
97
97
100
pleuritic pain Dyspnea or tachypnea (≥20/min) or pleuritic pain or signs of DVT Tachypnea = respiratory rate ≥20/min. P = 0.025, age ≥70 years vs. age <40 years. Other differences among age groups were not significant. Data from Stein et al. [26].
were apparent between patients ≥70 years of age and younger patients, although incomplete right bundle branch block was less frequent among patients 40–69 years than among patients less than 40 years of age (Table 11.10).
Blood gases according to age The partial pressure of oxygen in arterial blood (PaO2 ) was lower among patients ≥70 years than among those <40 years of age based on data in PIOPED I among all patients with PE [25]. Among patients 40–69 years, the PaO2 was lower than in patients <40 years, but not significantly lower than in patients ≥70 years of age [25]. The PaO2 among patients with PE ≥70 years of age, 40–69 years of age, and <40 years of age was 61 ± 12,
67 ± 15, and 75 ± 18 mm Hg, respectively (mean ± standard deviation). In PIOPED II, the PaO2 in all patients with PE who were 70 years old was 91 ± 72 mm Hg and did not differ significantly from values in patients 40–69 years (83 ± 31 mm Hg) (unpublished data from [26]). In those with PE and no prior cardiopulmonary disease, the PaO2 in patients 70 years old was 88 ± 45 mm Hg and it also did not differ significantly from values in patients 40–69 years (84 ± 27 mm Hg) (unpublished data from [26]). The alveolar–arterial oxygen difference (gradient) among patients with PE ≥70 years of age was 47 ± 14 mm Hg, which was higher than among patients 40–69 years old (40 ± 17 mm Hg) and it was higher than in patients <40 years old (31 ± 17 mm Hg). The alveolar–arterial oxygen difference in normal adults increases with age [27–30].
Table 11.10 Electrocardiographic findings in all patients with acute pulmonary embolism according to age: PIOPED I. ≥70 yr (n = 72) (%)
40–69 yr (n = 113) (%)
<40 yr (n = 36) (%)
Normal
21
27
22
ST segment or T wave changes
56
51
56
Left axis deviation
18
11
8
Left ventricular hypertrophy
12
7
11
Acute myocardial infarction pattern
12
4
6
Low voltage QRS
9
5
0
Complete right bundle branch block
7
4
3
Right ventricular hypertrophy
4
3
3
Right axis deviation
2
3
8
P-pulmonale
2
2
0
Incomplete right bundle branch block
2
0
11*
*P < 0.01, age 40–69 years vs. age <40 years. Reprinted from Stein et al. [25], with permission from the American College of Cardiology Foundation.
BLUK077-Stein
March 28, 2007
CHAPTER 11
15:8
61
Venous thromboembolism in the elderly
Clinical assessment according to age The accuracy of clinical assessment was comparable among patients in all age groups [25]. When physicians were 80–100% confident that PE was present on the basis of clinical judgment and laboratory tests with the exception of ventilation–perfusion scans, they were correct in 90% of 10 patients ≥70 years old. When they believed that there was less than 20% likelihood of PE, it was present in 19% of 69 patients. In most patients physicians were uncertain of the diagnosis, believing that there was a 20–79% chance of PE.
Ventilation–perfusion lung scan according to age Among patients ≥70 years of age the positive predictive value of a high-probability interpretation of the ventilation–perfusion lung scan (94%) was higher than in patients <40 years of age (69%) [25] (Table 11.11). The positive predictive value of other interpretations was comparable in all age groups. The sensitivity of ventilation–perfusion lung scans indicating a high probability of PE among patients ≥70 years of age (47%) did not differ significantly from the sensitivity of such scans among younger age groups [25].
CT angiography according to age Among 773 patients with an adequate CT pulmonary angiography and 737 patients with an adequate CT pulmonary angiography/CT venography, the sensitivity and specificity for PE for age groups 18–59, 60–79, and 80–99 years did not differ to a statistically significant extent [31]. Multidetector CT pulmonary an-
giography and CT pulmonary angiography/CT venography may be used with various diagnostic strategies in adults of all ages.
Acute hemorrhage/infarction syndrome in the elderly The syndrome of hemoptysis or pleuritic pain, in addition to signs and symptoms, was investigated in detail in elderly patients [32]. The electrocardiogram was normal in 62% of such patients [32]. If abnormal, the most frequent abnormalities were nonspecific ST segment or T wave changes (38%). The chest radiograph showed atelectasis or a pulmonary parenchymal abnormality in 82% of elderly patients with the hemoptysis/pleuritic pain syndrome. The central pulmonary artery dilated in 29% of such elderly patients. A normal chest radiograph was uncommon, occurring in only 6% of elderly patients. The ventilation–perfusion lung scan was interpreted as high probability for PE in 41% of elderly patients with the hemoptysis/pleuritic pain syndrome. Elderly patients with the hemoptysis/pleuritic pain syndrome had a higher pulmonary artery mean pressure (25 ± 9 versus 17 ± 7 mm Hg) and lower PaO2 (64 ± 10 versus 81 ± 14 mm Hg) than patients <40 years of age, and elderly patients tended to have more mismatched segmental perfusion defects on the ventilation–perfusion lung scan than patients <40 years of age [32].
Use of diagnostic tests in the elderly Diagnostic approaches to DVT and PE have changed markedly over the past two decades in temporal
Table 11.11 Results of ventilation–perfusion lung scans in all patients with acute pulmonary embolism according to age: PIOPED I. ≥70 yr V–Q scan probability High Intermediate Low Near normal/normal
PE/n
≤40 yr
40–69 yr (%)
PE/n
(%)
PE/n
(%)
34/36
(94)*
60/68
(88)
11/16
(69)
27/100
(27)
59/199
(30)
22/52
(42)
10/71
(14)
24/172
(14)
8/57
(14)
1/8
(13)
1/55
(2)
3/68
(4)
*P ≤ 0.025 ≤70 years vs ≤40, all other differences among age groups were not significant. n, number of patients with the scan result shown in column; PE, pulmonary embolism. Reprinted from Stein et al. [25], with permission from the American College of Cardiology Foundation.
BLUK077-Stein
March 28, 2007
15:8
62
Venograms/100,000 population
PART I
Prevalence, risks, and prognosis of PE and DVT
200 150 Age >70
100 Age 20−69
50 0
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
harmony with the evolving literature [3, 33]. Changes in the use of diagnostic tests in the elderly parallel the changes in the general population. The use of diagnostic tests in elderly black and white patients was comparable [1]. During the 1980s and early 1990s, the use of contrast venography of the lower extremities was strikingly higher in elderly patients than in younger patients
Figure 11.7 Trends over 21 years in the use of contrast venography of the lower extremities in elderly patients compared with younger patients. The use of contrast venography peaked in 1986 and began to decline sharply in 1989 and thereafter in patients aged 70 or older. The use of contrast venography was higher in elderly patients than in younger patients. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
(Figure 11.7) [1]. The use of contrast venography sharply declined as the use of ultrasonography increased. Doppler ultrasonography has supplanted ascending contrast venography as the preferred diagnostic approach for DVT [33]. Between 1989 and 1999, the elderly population utilized 5.7 times far more venous ultrasound tests of the lower extremities than the younger population (20–69 years) (Figure 11.8a) [1].
(a) Ultrasound/100,000 population
Ultrasound 300
Age >70 200 100
Age 20−69 0 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
(b) Ultrasound 1989−1999
30
25.3
Rate ratio
20.4
20
14.6 8.4
10 1.4
2.5
30−39
40−49
4.8
0 90−99
80−89
70−79
60−69
50−59
20−29
Age groups (years)
Figure 11.8 (a) The use of Doppler ultrasonography in the elderly markedly increased after 1982 and stabilized in the 1990s. (b) Rate ratios comparing the rate of use of venous ultrasound at age 20–29 to older decades for the interval of 1989 to 1999. Between ages 20–29 and age 80–89, the rate ratios increased exponentially with age. The 95% confidence intervals were too narrow to illustrate. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
March 28, 2007
63
Venous thromboembolism in the elderly
80−89
400
70−79
300 200
60−69 50−59
100
40−49 20−39
0 1999
1997
1995
1993
1991
1989
1987
1985
Year
427 patients 40–69 years, and in 0.7% of 135 patients <40 years of age. Renal failure, either major or minor, was the most frequent complication of conventional pulmonary angiography among elderly patients [25]. It occurred in 3% of 200 patients ≥70 years of age, compared with 0.7% of 562 patients ≤69 years of age. Among the 10 patients who developed renal failure, 3 required dialysis. “Minor” complications of renal failure were important complications, although dialysis was not required. Patients with these complications showed either an elevation of the serum creatinine from previously normal levels to ≥2.1 mg/100 mL (range 2.1–3.5 mg/100 mL) or an increase in a previously abnormal serum creatinine level ≥2 mg/100 mL.
40 30 Age >70
20
AGE 20−69
10 0
9 −9 97 19 6 −9 94 19 3 −9 91 19 0 −9 88 19 7 −8 85 19 4 −8 82 19 1 −8 79
19
The rate ratio of use of ultrasound, comparing each decade of age with the rate at age 20–29, based on average values from 1989 to 1999, was the highest in the elderly (Figure 11.8b). The use of lung scans over 21 years of observation was highest in elderly patients (Figure 11.9a) [1]. In all age groups, the use of lung scans has decreased since mid-1980s (Figure 11.9a). The utility of ventilation– perfusion lung scans among patients ≥70 years old was comparable with that in younger patients [25]. The positive predictive value of all probabilities of ventilation–perfusion lung scans using original PIOPED criteria [34] were comparable in all age groups (Table 11.11) [25]. However, a higher proportion of patients ≥70 years of age had nondiagnostic (intermediate or low probability) V–Q scans, (80%), than patients ≤40 years of age (56%) (Table 11.11). Among patients ≥70 years of age with ventilation–perfusion lung scans indicating a high probability of PE, 94% had PE (Table 11.11). The rate of use of pulmonary angiograms over a 21-year period of observation was higher in elderly patients than in younger patients (Figure 11.10) [1]. Both for elderly patients and for younger patients the use of pulmonary angiograms increased from 1979 to 1999. Pulmonary angiography was not more hazardous among the elderly, although renal failure was a more frequent sequela among patients ≥70 years of age than among younger patients [25]. Major complications occurred in 1.0% of 200 patients ≥70 years, in 1.2% of
1983
1981
1979
Figure 11.9 Trends over 21 years in the use of radioisotopic lung scans in elderly patients (70 years or older) showed a higher utilization compared with younger patients. The use of lung scans in elderly patients began to increase sharply in 1983, peaked in 1986, and showed a progressive decline in 1987. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
500
Angios/100,000 population
CHAPTER 11
15:8
Lung scans/100,000 population
BLUK077-Stein
Year Figure 11.10 Trends over 21 years in the use of pulmonary angiograms in elderly patients compared with younger patients. The rate of use of pulmonary angiograms was higher in elderly patients than in younger patients. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
BLUK077-Stein
March 28, 2007
15:8
64
PART I
References 1 Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260–2265. 2 Stein PD, Kayali F, Olson RE. Incidence of venous thromboembolism in infants and children: data from the National Hospital Discharge Survey. J Pediatr 2004; 145: 563–565. 3 Stein PD, Woodard PK, Weg JG et al. Diagnostic pathways in acute pulmonary embolism: recommendations of the PIOPED II Investigators. Am J Med 2006; 119: 1048–1055. 4 Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT study. Arch Intern Med 1991; 151: 933–938. 5 Silverstein MD, Heit JA, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ, III. Trends in the incidence of deep vein thrombosis and pulmonary embolism. A 25year population based study. Arch Intern Med 1998; 158: 585–593. 6 Gillum RF. Pulmonary embolism and thrombophlebitis in the United States, 1970–1985. Am Heart J 1987; 114: 1262–1264. 7 Giuntini C, Ricco GD, Marini C, Mellilo E, Palla A. Pulmonary embolism: epidemiology. Chest 1995; 107 (suppl): 3S–9S. 8 Nordstrom M, Linblad B, Bergqvist D, Kjellstrom T. A prospective study of the incidence of deep-vein thrombosis within a defined urban population. J Int Med Res 1992; 232: 155–160. 9 Coon WW, Willis PW, III, Keller JB. Venous thromboembolism and other venous disease in the Tecumseh Community Health Study. Circulation 1973; 48: 839–846. 10 Kniffin WD, Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA, Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994; 154: 861–866. 11 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. 12 Goldhaber SZ, Visani L, De Rosa M. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999; 353: 1386–1389. 13 Ferrari E, Baudouy M, Cerboni P et al. Clinical epidemiology of venous thromboembolic disease. Results of a French Multicentre Registry. Eur Heart J 1997; 18: 685– 691. 14 Hansson P-O, Welin L, Tibblin G, Eriksson H. Deep vein thrombosis and pulmonary embolism in the general pop-
15
16
17 18
19
20
21 22
23 24 25
26
27
28
29
30
31
Prevalence, risks, and prognosis of PE and DVT
ulation. The study of men born in 1913. Arch Intern Med 1997; 157: 1665–1670. Hume M, Sevitt S, Thomas DP. Venous Thrombosis and Pulmonary Embolism. A Commonwealth Fund Book. Harvard University Press, Cambridge, MA, 1970. Nicolaides AN, Irving D. Clinical factors and the risk of deep venous thrombosis. Thromboembolism. In: Nicolaides AN, ed. Etiology, Advances in Prevention and Management. University Park Press, Baltimore, 1975: 199–204. Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. Stein PD, Huang H-L, Afzal A, Noor H. Incidence of acute pulmonary embolism in a general hospital: relation to age, sex, and race. Chest 1999; 116: 909–913. Stein PD, Patel KC, Kalra NK et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. Stein PD, Kayali F, Olson RE. Estimated case fatality rate from pulmonary embolism, 1979–1998. Am J Cardiol 2004; 93: 1197–1199. Geerts WH, Heit JA, Clagett GP et al. Prevention of venous thromboembolism. Chest 2001; 119(suppl): 132S–175S. Taubman LB, Silverstone FA. Autopsy proven pulmonary embolism among the institutionalized elderly. J Am Geriatr Soc 1986: 34: 752–756. Morrell MT. The incidence of pulmonary embolism in the elderly. Geriatrics 1970; 25: 138–153. Busby W, Bayer A, Pathy J. Pulmonary embolism in the elderly. Age Ageing 1988; 17: 205–209. Stein PD, Gottschalk A, Saltzman HA, Terrin ML. Diagnosis of acute pulmonary embolism in the elderly. J Am Coll Cardiol 1991; 18: 1452–1457. Stein PD, Beemath A, Matta F et al. Clinical Characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med 2007 (In press). Mellemgaard K. The alveolar–arterial oxygen difference: its size and components in normal man. Acta Physiol Scand 1966; 67: 10–20. Harris EA, Kenyon AM, Nisbet HD, Seelye ER, Whitlock RML. The normal alveolar–arterial oxygen-tension gradient in man. Clin Sci (Colch) 1974; 46: 89–104. Filley GF, Gregoire F, Wright GW. Alveolar and arterial oxygen tensions and the significance of the alveolar– arterial oxygen tension difference in normal men. J Clin Invest 1954; 33: 517–529. Kanber GJ, King FW, Eshchar YR, Sharp JT. The alveolar– arterial oxygen gradient in young and elderly men during air and oxygen breathing. Am Rev Respir Dis 1968; 97: 376–381. Stein PD, Beemath A, Quinn DA et al. Usefulness of multidetector spiral computed tomography according to age and sex for diagnosis of acute pulmonary embolism. Am J Cardiol (In press).
BLUK077-Stein
March 28, 2007
15:8
Venous thromboembolism in the elderly
65
32 Stein PD, Henry JW. Acute pulmonary embolism presenting as pulmonary hemorrhage/infarction syndrome in the elderly. Am J Geriatr Cardiol 1998; 7: 36– 42. 33 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diag-
nosing deep vein thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 34 The PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.
CHAPTER 11
BLUK077-Stein
March 28, 2007
12
15:10
CHAPTER 12
Pulmonary thromboembolism in infants and children
Twenty-three years of data were analyzed in children discharged from hospitals in the United States were based on the database of the National Hospital Discharge Survey (NHDS) [1]. Venous thromboembolic disease was coded at discharge from short-stay hospitals with sufficient frequency to indicate that it should be considered in the diagnosis of children with appropriate clinical findings [1]. Infants, teenage girls, and black children had the highest rates of diagnosis [1]. From 1979 to 2001, pulmonary embolism (PE) was diagnosed at discharge from short-stay non-Federal hospitals throughout the United States in 13,000 infants and children ≤17 years of age, deep venous thrombosis (DVT) in 64,000, and venous thromboembolism (VTE) in 75,000 [1]. Rates of diagnosis were 0.9 PE/100,000 children/year, 4.2 DVT/100,000 children/year, and 4.9 VTE/100,000 children/year (Figure 12.1) [1]. The rates of diagnosis of DVT and of VTE did not change from the triennial periods 1979–1982 to 1999–2001. Rates of diagnosis of PE, DVT, and VTE were higher in infants than children aged 2–14 years (Figures 12.2– 12.4) [1]. The rates were also higher in teenagers aged 15–17 years than in children aged 2–14 years, but the rates in teenagers were comparable to the rates in infants.
In teenagers aged 15–17 years, the rate of diagnosis of DVT was 2.1 times higher in girls than boys. Teenage girls with DVT had an associated pregnancy in 27%. The rate of DVT in nonpregnant teenage girls was 10 DVT/100,000 teenage girls/year, and the rate of pregnancy-associated DVT was 109 DVT/100,000 teenage girls/year. The rate of DVT in nonpregnant teenage girls did not differ significantly from the rate for teenage boys. The rate of diagnosis of PE in black children, 1.6/100,000/year, was 2.4 times higher than in white children, 0.7/100,000/year. The rate of DVT in black children, 5.7/100,000/year, was 1.7 times the rate in white children, 3.3/100,000/year. The rate of VTE in black children was 1.8 times the rate in white children. A double-peaked curve was shown for DVT [1, 2] and PE [1] with the highest rates of diagnosis in infants less than 1-year-old and a second peak in teenagers. Teenage girls had twice the rates of DVT and VTE as teenage boys, although in younger children the frequencies were comparable. Bernstein and associates, among adolescents, observed PE in twice as many girls as boys [3]. Pregnancy-related DVT accounted for the difference in rates between teenage boys and teenage girls.
6 5 4 3 2 1 0
4.9 4.2
0.9
3 2.2
1
0.4 0−1
PE
2.0
2
0
DVT
VTE
Figure 12.1 Thromboembolic disease in children ≤17 years of age. (Data from Stein et al. [1].)
66
PE/100,000/yr
PE, DVT, VTE/100,000 children/yr
4
2−14
15−17
Age (years) Figure 12.2 Pulmonary embolism (PE) in children. (Data from Stein et al. [1].)
March 28, 2007
CHAPTER 12
DVT/100,000/yr
10
15:10
67
Pulmonary thromboembolism in infants and children 9.9 8.7
8 6 4 2.1
2
12 VTE/100,000/yr
BLUK077-Stein
11.4
10.5
10 8 6 4
2.4
2 0
0 0−1
2−14
15−17
0−1
Age (years) Figure 12.3 Deep venous thrombosis (DVT) in children. (Data from Stein et al [1].)
Abortion and/or contraceptives were shown to be risk factors in 75% of female adolescents who had PE [3]. Teenage users of oral contraceptives did not appear to be at an increased risk of VTE compared to older users [4]. Indwelling catheter use was the most common predisposing factor for PE or DVT in children and adolescents, followed by surgery and trauma [5]. Neonatal thrombosis, with the exception of spontaneous renal vein thrombosis was almost always associated with indwelling catheters (86 of 97 cases, 89%) [6]. Lower extremity DVT in children, when unrelated to venous catheterization or surgery, appeared to be related to local infection of the involved extremity, trauma or immobilization [7]. One or more coagulopathies were reported in 26 of 40 (65%) children with venous thrombosis who had an evaluation for a deficiency of protein C, protein S, or antithrombin III and assessment for a lupus anticoagulant [8]. In the Canadian registry, 9% of children with VTE had a coagulopathy, but the majority did not receive coagulation testing [2, 9]. Among patients with heart disease at autopsy, the prevalence of PE as a cause of death was particularly high in children <10 years old [10]. Among medical patients with heart disease <10 years old, PE as a cause of death was found at autopsy in 14.5% and in postoperative children with heart disease at autopsy PE caused death in 8.0% [10]. Prior to 1962, under 50 cases of PE in children had been reported in the world literature [11]. Only 36 children with DVT and 10 with PE were identified among all Scottish hospital inpatients from 1968 to 1971 [12]. From 1975 to 1991, 308 children with DVT or PE were reported in the English and French literature [5]. A
2−14 Age (years)
15−17
Figure 12.4 Venous thromboembolism (VTE) in children. (Data from Stein et al. [1].)
Canadian registry of 15 hospitals from July 1990 to December 1992 identified 137 children aged 1 month to 17 years old who had DVT [2].
References 1 Stein PD, Kayali F, Olson RE. Incidence of venous thromboembolism in infants and children: data from the National Hospital Discharge Survey. J Pediatr 2004; 145: 563–565. 2 Andrew M, David M, Adams M et al. Venous thromboembolic complications (VTE) in children: first analyses of the Canadian Registry of VTE. Blood 1994; 83: 1251–1257. 3 Bernstein D, Coupey S, Schonberg SK. Pulmonary embolism in adolescents. Am J Dis Child 1986; 140: 667–671. 4 Royal College of General Practitioners’ Oral Contraception Study. Oral contraceptives, venous thrombosis, and varicose veins. J R Coll Gen Pract 1978; 28: 393–399. 5 David M, Andrew M. Venous thromboembolic complications in children. J Pediatr 1993; 123: 337–346. 6 Schmidt B, Andrew M. Neonatal thrombosis: report of a prospective Canadian and international registry. Pediatrics 1995; 96: 939–943. 7 Wise RC, Todd JK. Spontaneous, lower-extremity venous thrombosis in children. Am J Dis Child 1973; 126: 766– 769. 8 Nuss R, Hays T, Manco-Johnson M. Childhood thrombosis. Pediatrics 1995; 96: 291–294. 9 Manco-Johnson MJ. Disorders of hemostasis in childhood: risk factors for venous thromboembolism. Thromb Haemost 1997; 78: 710–714. 10 Pulido T, Aranda A, Zevallos MA et al. Pulmonary embolism as a cause of death in patients with heart disease. An autopsy study. Chest 2006; 129: 1282–1287. 11 Emery JL. Pulmonary embolism in children. Arch Dis Child 1962; 37: 591–595. 12 Jones DR, Macintyre IM. Venous thromboembolism in infancy and childhood. Arch Dis Child 1975; 50: 153–155.
BLUK077-Stein
March 12, 2007
13
15:31
CHAPTER 13
Venous thromboembolism in men and women The Tecumseh Community Health Study showed 4.5 PE/10,000 women/year compared with 1.75 PE/10,000 men/year [4]. The largest investigation was based on data from the National Hospital Discharge Survey [7]. From 1979 to 1999, 2,448,000 patients were discharged
(a)
(b)
80 Women
60 40 Men
20
80
Age-adjusted PE/100,000
PE/100,000 population
Pulmonary embolism (PE) has been reported to occur more frequently in women than in men due to estrogen use, childbearing, and a higher frequency of deep venous thrombosis (DVT) [1–5]. A postmortem study showed PE in 11% of women and 7% of men [6].
Women
60
40
Men
20
(d) Women
160
120 Men
80
40
Men
100
50 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
Figure 13.1 (a) Unadjusted rates of diagnosis of pulmonary embolism (PE) per 100,000 population over a 21-year period. The unadjusted rate of diagnosis was higher in women than men. (b) Age-adjusted rates of diagnosis of PE per 100,000 population over a 21-year period. The age-adjusted rates of diagnosis were comparable in men and women. (c) Unadjusted rates of diagnosis of deep venous thrombosis (DVT) per 100,000
Women
150
1979
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
Age-adjusted DVT/100,000
DVT/100,000 population
1999
1997
1995
1993
1991
Year
(c)
68
1989
1987
1985
1983
1981
1979
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
Year
population over a 21-year period. The unadjusted rate of diagnosis of DVT was higher in women than in men. (d) Age-adjusted rates of diagnosis of DVT per 100,000 population over a 21-year period. The age-adjusted rate of diagnosis was higher in women. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)
March 12, 2007
1999
1997
1995
1993
1991
1989
(b) 60 45
Women
30
Men
15 0 1999
1997
1995
1993
1991
1989
1987
1985
Figure 13.2 Rates of use per 100,000 population of ventilation–perfusion (V–Q) lung scans. The rates were higher among women than in men. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)
1987
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
Year
1983
Men
0
Men
0 1985
20
Women
1981
Women
15
1983
40
30
1981
60
45
1979
80
(a)
1979
from short-stay non-Federal hospitals with PE and 5,691,000 were discharged with DVT. Either PE or DVT was listed as the discharge diagnosis in 7,589,000 patients [7]. Given that PE and DVT are conditions that increase in incidence with age, age-adjusted rates of diagnosis as well as crude (unadjusted) rates were calculated [7]. In 1999, 139,000 patients were discharged from nonFederal short-stay hospitals with a diagnosis of PE. The rate of diagnosis not adjusted for age was 42 PE/100,000 men/year and 60 PE/100,000 women/year (Figure 13.1a) [7]. The age-adjusted rate of diagnosis of PE/100,000 population/year in men and women was comparable (Figure 13.1b) [7]. This was concordant with some prior investigations of PE [8–11], but PE has also been reported more frequently in men [12– 15] and older men [16]. Both the unadjusted rate of diagnosis of DVT/ 100,000 population/year and the age-adjusted rates of diagnosis of DVT/100,000 population/year were higher in women (Figures 13.1c and 13.1d) [7]. This is concordant with most prior literature [8, 16, 17], but the prior literature is not uniform. Objectively diagnosed DVT had also been reported to be more frequent in men [9]. In 1999, the unadjusted rate of diagnosis was 115 DVT/100,000 men/year and 154 DVT/100,000 women/year [7]. In 1999, 369,000 patients were discharged from non-Federal short-stay hospitals with a diagnosis of DVT. The rates of use/100,000 population/year of ventilation–perfusion lung scans (Figure 13.2) and
V−Q scans/100,000 population
69
Venous thromboembolism in men and women
Venograms/100,000 population
CHAPTER 13
15:31
Ultrasonography/100,000 population
BLUK077-Stein
Year Figure 13.3 (a) Rates of venography per 100,000 population over a 21-year period. The rates of use among men and women were comparable. (b) Rates of venous ultrasonography of the lower extremities per 100,000 population over a 21-year period. The rate of use was higher among women. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)
venous ultrasound examinations of the lower extremities were higher among women (Figure 13.3b) [7]. In 1999, the rate of use of ventilation–perfusion lung scans was 18/100,000 men/year and 24/100,000 women/year. During the same year the rate of use of venous ultrasound examinations of the lower extremities was 31/100,000 men/year and 38/100,000 women/year. A more frequent use of venous ultrasonography for the diagnosis of DVT had been observed previously in women [18]. The rates of use/100,000 population/year of contrast venography among men and women were comparable (Figure 13.3a). The durations of hospitalization among patients with a primary discharge diagnosis of PE and of DVT were comparable among men and women (Figure 13.4a) [7].
BLUK077-Stein
March 12, 2007
15:31
70
PART I
(a)
Table 13.1 Signs and symptoms of acute pulmonary embolism in women and men.
Prevalence, risks, and prognosis of PE and DVT
PE hospitalization days
16 12 Women
8 Men
4 0 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
DVT hospitalization days
8
Women
Men
4
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
0
(n = 132)
percent
percent
80
78
Pleuritic pain
60
57
Cough
41
40
Leg swelling
24
36†
Leg pain
23
30
Hemoptysis
10
21‡
Rales (crackles)
60
57
2
7*
* P < 0.05. † P < 0.04. ‡ P < 0.02, women vs. men. Data are from Quinn et al. [19].
16 12
Men
(n = 119)
Dyspnea
Pleural friction rub
(b)
Women
Year Figure 13.4 (a) Duration of hospitalization for men and women with a primary discharge diagnosis of pulmonary embolism (PE). Duration was comparable among genders. (b) Duration of hospitalization for men and women with a primary discharge diagnosis of deep venous thrombosis (DVT). The duration of hospitalization was comparable among genders. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)
Regarding the signs and symptoms of acute PE, women recruited in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) less frequently had hemoptysis (10% versus 21%), leg swelling (24% versus 36%), and pleural friction rub (2% versus 7%) (Table 13.1). Except for minor differences of heart rate, the prevalence of other signs and symptoms of acute PE did not differ significantly between women and men [19]. The sensitivity and specificity of the ventilation– perfusion scan, using original PIOPED criteria [20], were the same in women and men. The positive predic-
tive values of a high-probability ventilation–perfusion scan were similar between women (86%) and men (90%) and the negative predictive values of normal or nearly normal ventilation–perfusion scans were also similar, 93% in women and 88% in men [19]. There was no difference between women and men during 1-year follow-up in the recurrence of PE [19].
References 1 Palevsky HI. Pulmonary hypertension and thromboembolic disease in women. Cardiovasc Clin 1989; 19: 267– 283. 2 Bernstein D, Goupey S, Schonberg SK. Pulmonary embolism in adolescents. AJDC 1986; 140: 667–671. 3 Coon W. Epidemiology of venous thromboembolism. Ann Surg 1977; 186: 149–164. 4 Coon WW, Willis PW, III, Keller JB. Venous thromboembolism and other venous disease in the Tecumseh Community Health Study. Circulation 1973; 48: 839–846. 5 Breckenridge RT, Ratnoff OD. Pulmonary embolism and unexpected death in supposedly normal persons. N Engl J Med 1964; 270: 298–299. 6 Karwinski B, Svendsen E. Comparison of clinical and postmortem diagnosis of pulmonary embolism. J Clin Pathol 1989; 42: 135–139. 7 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694. 8 Ferrari E, Baudouy M, Cerboni P et al. Clinical epidemiology of venous thromboembolic disease. Results of a
BLUK077-Stein
March 12, 2007
CHAPTER 13
9
10
11
12
13
14
15:31
Venous thromboembolism in men and women
French multicentre registry. Eur Heart J 1997; 18: 685– 691. Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151: 933–938. Stein PD, Huang H-L, Afzal A, Noor H. Incidence of acute pulmonary embolism in a general hospital: relation to age, sex, and race. Chest 1999; 116: 909–913. Stein PD, Patel KC, Kalra NK et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. Lilienfeld DE, Godbold JH, Burke GL, Sprafka JM, Pham DL, Baxter J. Hospitalization and case fatality for pulmonary embolism in the twin cities: 1979–1984. Am Heart J 1990; 120: 392–395. Janke RM, McGovern PG, Folsom AR. Mortality, hospital discharges, and case fatality for pulmonary embolism in the twin cities 1980–1995. J Clin Epidemiol 2000; 53: 103– 109. Silverstein MD, Heit JA, Mohr DN et al. Trends in the incidence of deep vein thrombosis and pulmonary
15
16
17 18
19
20
71
embolism. A 25-year population-based study. Arch Intern Med 1998; 158: 585–593. Giuntini C, Di Ricco G, Marini C, Melillo E, Palla A. Pulmonary embolism: epidemiology. Chest 1995; 107(suppl): 3S–9S. Kniffin WD, Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA, Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994; 154: 861–866. Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. Beebe HG, Scissons RP, Salles-Cunha SX et al. Gender bias in use of venous ultrasonography for diagnosis of deep vein thrombosis. J Vasc Surg 1995; 22: 538– 542. Quinn DA, Thompson BT, Terrin ML et al. A prospective investigation of pulmonary embolism in women and men. JAMA 1992; 268: 1689–1696. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.
BLUK077-Stein
April 14, 2007
14
14:48
CHAPTER 14
Comparison of the diagnostic process in black and white patients
Introduction The death rate among patients with pulmonary embolism (PE) in the United States is higher among black patients than among white patients [1]. Recognizing that there is a racial disparity in the use of some health services [2–10], one could speculate that such a disparity could contribute to the higher death rate among black patients with PE. However, nothing was found to suggest that diagnostic tests were being withheld in black patients [11]. Also, there was no evidence of a failure to reach a diagnosis in black patients with thromboembolic disease [11]. Caution has been recommended in the use of race as a variable [12]. It is necessary to account for distinctions between race and socioeconomic status [12]. An emphasis on ethnic groups rather than on race implies an appreciation of cultural and behavioral attitudes, beliefs, lifestyle patterns, diet, environmental living conditions, and other factors [13]. Disparities in health care may be due to a lack of appropriate health care messages, limited access to care and services, lack of trust of health care providers, and racial bias among medical care providers [12]. Data on trends over 21 years and relative differences among races in the prevalence of PE and DVT (deep venous thrombosis) and the use of diagnostic tests were obtained from the National Hospital Discharge Survey from 1979 to 1999 [11]. During these 21 years, 2,448,000 patients were discharged from short-stay non-Federal hospitals with PE, 5,691,000 patients were discharged with DVT and 7,589,000 patients were discharged with either PE or DVT. The rate of diagnosis of PE/100,000 population/year not adjusted for age, was comparable among black and white patients (Figure 14.1a) [11]. In 1999, the rate of diagnosis among black patients was 41 PE/100,000
72
population/year and among white patients it was 42 PE/100,000 population/year. Despite some year-toyear fluctuation, the rate of diagnosis of PE among black patients did not change appreciably over the 21year period studied. White patients, however, showed a decreasing rate of PE in the population over the period of survey [11]. Adjustment for age caused the rate of diagnosis of PE among black patients to separate from white patients after the mid-1980s (Figure 14.1b) [11]. The age-adjusted rates indicate the hospitalization rates for black and white patients assuming identical age distributions for both populations. The age-adjusted rate of diagnosis of PE/100,000 population/year was higher in black patients than in white patients (Figure 14.1b). From 1979 to 1992, there was a decline in the rate of diagnosis both among black patients and among white patients. From 1992 to 1999, the rate of diagnosis was constant in black patients, but the rate of diagnosis increased somewhat in white patients. The rate of diagnosis of DVT/100,000 population/ year, not adjusted for age, was comparable among black and white patients (Figure 14.1c) [11]. In 1999, the rate of diagnosis among black patients was 110 DVT/100,000 population/year and among white patients it was 115 DVT/100,000 population/year. The rate of diagnosis increased over the 21-year period of survey in black patients. Among white patients, the rate of diagnosis remained unchanged during the period of survey. The age-adjusted rate of diagnosis of DVT/100,000 population/year was higher in black patients than in white patients (Figure 14.1d). From 1979 to 1992, the age-adjusted rate of diagnosis was constant in black patients, whereas white patients showed a gradual decrease in the rate. From 1992 to 1999, the rate of diagnosis increased sharply in black patients; the rate increased more gradually in white patients.
BLUK077-Stein
April 14, 2007
CHAPTER 14
14:48
73
VTE in black and white patients
60
Black patients
40 White patients
20
Black patients
60 40
White patients
20 0
Year
150
Black patients
100 White patients
50 0
200
Black patients
150 100 White patients
50 0
1999
1997
1995
60 45
Black patients
30 15
White patients
0 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
V−Q scans/100,000 population
diagnosis over a 21-year period of deep venous thrombosis (DVT) per 100,000 population according to race. The rate of diagnosis was comparable among black and white patients. (d) Age-adjusted rates of diagnosis of DVT per 100,000 population according to race. The rate was higher in black patients than in white patients. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)
1979
The use of pulmonary angiograms/100,000 population/year was low among both black and white patients. The frequency of use of angiograms among black patients was too low to calculate rates. However, inspection of the data showed no suggestion of a disparity of use between black and white patients. Rates of use of ventilation–perfusion lung scans among black and white patients from 1979 to 1999 were comparable (Figure 14.2). The use of ventilation– perfusion lung scans in both groups increased sharply between 1979 and 1986. The use then declined in black as well as white patients.
1993
Use of diagnostic tests for PE and deep vein thrombosis
1991
Year
Year Figure 14.1 (a) Unadjusted (crude) rates of diagnosis over a 21-year period of pulmonary embolism (PE) per 100,000 population according to race. The rates of diagnosis were comparable among black and white patients. (b) Ageadjusted rates of diagnosis of PE per 100,000 population according to race. There was no change in the rate of diagnosis among black patients over the 21-year period. White patients, however, showed a decreasing rate of PE over the period of survey. (c) Unadjusted (crude) rates of
1989
1987
1985
1983
1981
1979
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Age-adjusted DVT/100,000
(d)
(c) DVT/100,000 population
1999
1997
1995
1993
1991
Year
1989
1987
1985
1983
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1981
0
80
1979
PE/100,000 population
80
Age-adjusted PE/100,000
(b)
(a)
Year Figure 14.2 Rates of use per 100,000 population of ventilation–perfusion (V–Q) lung scans among black and white patients. The rates were comparable. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)
BLUK077-Stein
April 14, 2007
14:48
74
PART I
Prevalence, risks, and prognosis of PE and DVT
Venography/100,000 population
(a) 50
25
White patients Black patients
0 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
60 Black patients 40 White patients
20
0 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Ultrasonography/100,000 population
(b)
Year
Figure 14.3 (a) Rates of use per 100,000 population of contrast venography among black and white patients. The rates were comparable. (b) Rates of use per 100,000 population of venous ultrasound of the lower extremities among black and white patients. The rates were comparable. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)
(a)
PE, days
18
12 Black patients White patients 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
6
Year
(b)
DVT, days
15 10 Black patients 5
White patients
0 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year
Figure 14.4 (a) Duration of hospitalization for black and white patients with a primary discharge diagnosis of pulmonary embolism (PE). The durations were comparable among races. (b) Duration of hospitalization for black and white patients with a primary discharge diagnosis of deep venous thrombosis (DVT). The durations were comparable among races. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)
BLUK077-Stein
April 14, 2007
CHAPTER 14
14:48
VTE in black and white patients
The rates of use of contrast venography among black and white patients from 1979 to 1999 were comparable (Figure 14.3a) [11]. The rates of use of venograms in black and white patients peaked in the late 1980s. The rates of use of venous ultrasound of the lower extremities among black and white patients from 1979 to 1999 were comparable (Figure 14.3b). Between 1988 and 1999, the ratio of use of ultrasounds in black to white patients ranged from 1.2 to 1.8. The rate of use of venous ultrasounds increased in both until 1991.
Use of medical facilities for treatment of PE and deep vein thrombosis The duration of hospitalization for black and white patients with a primary discharge diagnosis of PE was comparable (Figure 14.4a) [11]. The duration of hospitalization decreased over time in both groups. The duration of hospitalization for black and white patients with a primary discharge diagnosis of DVT was also comparable (Figure 14.4b). As with PE, the duration of hospitalization decreased over time in both. This reflects the use of protocols for a rapid attainment of therapeutic levels of heparin, a shorter duration of therapy with heparin, and use of low-molecular heparin [14]. These data were obtained entirely from hospitalized patients [11]. Information related to care after hospitalization was not available. Some have reported a shorter duration of anticoagulant therapy in black patients [15].
References 1 Lilienfeld DE. Decreasing mortality from pulmonary embolism in the United States, 1979–1996. Int J Epidemiol 2000; 29: 465–469. 2 Freeman HP, Payne R. Racial injustice in health care. N Engl J Med 2000; 342: 1045–1047.
75
3 Schneider EC, Zaslavsky AM, Epstein AM. Racial disparities in the quality of care for enrollees in Medicare managed care. JAMA 2002; 287: 1288–1294. 4 Ayanian JZ, Udvarhelyi S, Gatsonis CA et al. Racial differences in the use of revascularization procedures after coronary angiography. JAMA 1993; 269: 2642–2646. 5 Gornick ME, Eggers PW, Reilly TW et al. Effects of race and income on mortality and use of services among Medicare beneficiaries. N Engl J Med 1996; 335: 791–799. 6 Roetzheim RG, Pal N, Tennant C, et al. Effect of health insurance and race on early detection of cancer. J Natl Cancer Inst 1999; 91: 1409–1415. 7 Kasiske BL, Neylan JF, III, Riogio RR et al. The effect of race on access and outcome in transplantation. N Engl J Med 1991; 324: 302–307. 8 Bach PB, Cramer LD, Warren JL, Begg CB. Racial differences in the treatment of early stage of lung cancer. N Engl J Med 1999; 341: 1198–1205. 9 Brawley OW, Freeman HP. Race and outcomes: is this the end of the beginning for minority health research? J Natl Cancer Inst 1999; 91: 1908–1909. 10 Roach M, III, Cirrincione C, Budman D et al. Race and survival from breast cancer based on Cancer and Leukemia Group B Trial 8541. Cancer J Sci Am 1997; 3: 107–112. 11 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in blacks and whites. Arch Intern Med 2003; 163: 1843–1848. 12 Thomas SB. The color line: race matters in the elimination of health disparities. Am J Public Health 2001; 91: 1046– 1048. 13 Haynes MA, Smedley BD (eds.) The Unequal Burden of Cancer: An Assessment of NIH Research and Programs for Ethnic Minorities and the Medically Underserved. National Academy Press, Washington, DC, 1999: 19. 14 Hyers TM, Agnelli G, Hull RD et al. Antithrombotic therapy for venous thromboembolic disease. Chest 2001; 119(suppl): 176S–193S. 15 Ganz DA, Glynn RJ, Mogun H et al. Adherence to guidelines for oral anticoagulation after venous thrombosis and pulmonary embolism. J Gen Intern Med 2000; 15: 776– 781.
BLUK077-Stein
March 12, 2007
15
15:39
CHAPTER 15
Pulmonary thromboembolism in Asians/Pacific Islanders
Introduction Most investigations of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in countries in Asia suggest that the rate of diagnosis is lower in Asians [1–13] than in persons from North America or Europe [14–18], although some did not describe differences [19–22]. Data from the National Hospital Discharge Survey combined with data from the Bureau of the Census show that in the United States the rate of diagnosis of DVT, PE, and VTE; the incidence in hospitalized patients; and mortality rate from PE were all lower in Asians/Pacific Islanders than in whites and African Americans [23].
Unadjusted population-based rates of diagnosis From 1990 to 1999, rates of diagnosis of PE in Asian/Pacific/Islanders were 5/100,000 population/ year, compared with 34/100,000/year in whites and 39/100,000/year African Americans [23] (Table 15.1, Figure 15.1). Regarding DVT, rates of diagnosis in Asian/Pacific Islanders were 20/100,000 population/year compared with 98/100,000/year in whites and 105/100,000/ year in African Americans. Rates for VTE were 23/ 100,000 population/year in Asian/Pacific Islanders, 122/100,000/year in whites and 134/100,000/year in African Americans [23]. Others too, reported VTE among Asian Americans to be lower than in whites and African Americans in two investigations [12, 13]. Klatsky et al. [12] studied the rate of a primary diagnosis of VTE among hospitalized patients in California from 1978 to 1985, and reported rates of 2/100,000 population/year in Asian Americans, 21/100,000 population/year in whites, and 22/100,000 population/year in African Americans.
76
Klatsky et al. [12] reported even lower rates of diagnosis of venous thromboembolic disease among Asian Americans, whites, and African Americans in California than we observed throughout the United States [23]. The likely reason is that they required a primary diagnosis of PE or DVT, whereas we included all patients with PE or DVT [23]. White et al. [13] investigated idiopathic DVT in hospitalized patients in California between 1991 and 1994, and reported rates of 6/100,000 population/year in Asian Americans, 23/100,000 population/year in whites, and 29/100,000 population/year in African Americans. White et al. [13] also reported lower rates of DVT among Asian Americans, whites, and African Americans in California, which would seem to relate to the study requirement of idiopathic DVT and the exclusion of patients with cancer and those with temporary risk factors such as surgery and trauma. We included all patients with DVT [23].
Age-adjusted population-based rates of diagnosis Age-adjusted rates of diagnosis of PE were also lower in Asians/Pacific Islanders than in African Americans and whites (Table 15.1, Figure 15.1). Considering VTE (DVT and/or PE) [23], age-adjusted rates for the 10year period were also lower among Asians/Pacific Islanders than in African Americans and whites [23] (Table 15.1, Figure 15.1). Among Asians/Pacific Islanders, the age-adjusted rates of DVT were similar in men and women (19/100,000/year versus 24/100,000/year), as were the rates for PE (5/100,000/year versus 8/100,000/year), and VTE (22/100,000/year versus 30/100,000/year) [23].
Incidence in hospitals There was a disproportionately low rate of hospitalizations among Asians/Pacific Islanders (5000
March 12, 2007
CHAPTER 15
15:39
77
Pulmonary thromboembolism in Asians
Figure 15.1 Age-adjusted rates (diagnoses/100,000 population/year) of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in hospitalized patients from 1990 to 1999. (Reprinted from Stein et al. [23], with permission from Elsevier.)
Age-adjusted rates/100,000/year
BLUK077-Stein
200 138
130 150 107
104
100
26 50
VTE
22 36
40
7
0 Whites
hospitalizations/100,000 population/year) compared with whites (9000 hospitalizations/100,000 population/year) and African Americans (11,000 hospitalizations/100,000 population/year) [23]. The incidence of DVT among hospitalized Asians/Pacific Islanders aged ≥20 years (0.4/100 hospitalizations) was lower than among African Americans (1.1/100 hospitalizations) and whites (1.1/100 hospitalizations) (Table 15.2) [23]. In patients ≥20 years, the incidence of PE in hospitalized patients was also lower among Asians/Pacific Islanders (0.1/100 hospitalizations) than among whites (0.4/100 hospitalizations) and African Americans (0.4/100 hospitalizations) [23]. Similarly, in patients aged ≥20 years, the incidence of VTE in hospitalized patients was lower in Asians/Pacific Islanders (0.5/100 hospitalizations) than in whites (1.4/100 hospitalizations) and African Americans (1.4/100 hospitalizations) [23].
Population mortality rates Age-adjusted mortality from 1990 to 1998 for PE was lower in “others” (1.0 deaths/100,000 population/year) than in whites (3.4 deaths/100,000 population/year) and African Americans (6.9 deaths/100,000 population/year) (Table 15.3) [23]. Mortality from PE among Asians/Pacific Islanders was included in mortality rates described in “others.” Lower rates of fatal PE in Asians were reported at autopsy in Singapore [1] but not in Chinese in Hong Kong [21]. The low rate of diagnosis was related partly to the low rate of all hospitalizations in Asians/Pacific Islanders. The lower incidence of thromboembolic disease in hospitalized patients was independent of effects of a disproportionately low rate of all-cause hospitalization
African Americans
DVT PE
Asian/Pacific Islanders
among Asians/Pacific Islanders. The lower mortality among Asians/Pacific Islanders was also independent of the lower rate of all-cause hospitalizations. A lower rate of VTE in Asians than whites has been observed in several Asian ethnicities, as well as in many different clinical settings, including autopsy, where possible ethnic customs related to outpatient therapy would hardly have affected the results (Table 15.4). In the study by Klatsky et al. on Asians in northern California, 42.2% were Chinese, 11.9% were Japanese, 29.5% were Filipinos, 5.0% were South Asians, and 11.2% were other Asians [13]. Lower rates of thromboembolic disease have been shown in investigations in China [1, 5, 8–11], Japan [2], Thailand [3, 4], and Malaysia [6, 7]. Some of these investigations included patients from other Asian regions, including India [1, 5, 6]. A lower rate of PE was found at autopsy among Indians from Vellore, South India, than among patients in Boston and Los Angeles [2]. A lower prevalence of genetically induced abnormalities predisposing to VTE, such as factor V Leiden, has been speculated to contribute to the lower incidence of VTE in Asians [24]. Factor V Leiden, an abnormal factor V protein that is relatively resistant to degradation by protein C [25], is the most common genetic mutation predisposing to VTE [25, 26]. Factor V Leiden has been found in 4–5% of whites in North America and Europe [27, 28], 0.9–1.2% of African Americans, and in only 0–0.5% of Asians [27–29]. Differences in coagulation factors may also contribute to the low rate of VTE in Asians. Plasma fibrinogen levels are lower in Japanese than in whites, irrespective of whether they live in a rural or urban area of Japan or whether they are Japanese Americans [30]. Blood levels of factor VIIc and factor VIIIc were
78 18,000
5,000
127,000
736,000
embolism
21,000
440,000
2,641,000
embolism
P < 0.0005 Asian-Pacific Islanders vs. Caucasians and African American. Reprinted from Stein et al. [23], with permission from Elsevier.
†
* 10 year sum of census estimates.
Asian-Pacific Islander
344,000
2,128,000
Caucasian
African American
thrombosis
thrombo-
Race
Venous
venous
Pulmonary
Deep
Number of patients (1990–1999)
92,179,599
328,136,068
2,170,606,118
Population*
39 5†
20†
34
embolism
Pulmonary
105
98
thrombosis
venous
Deep
23†
134
122
embolism
thrombo-
Venous
(Diagnoses/100,000 population/year)
Crude rate
22†
107
104
thrombosis
venous
Deep
7†
40
36
embolism
Pulmonary
26†
138
130
embolism
thrombo-
Venous
(Diagnoses/100,000 population/year)
Age-adjusted rate
March 12, 2007
Table 15.1 Deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to race (1990–1999).
BLUK077-Stein
15:39
BLUK077-Stein
March 12, 2007
CHAPTER 15
15:39
79
Pulmonary thromboembolism in Asians
Table 15.2 Deep venous thrombosis and venous thromboembolic disease in hospitalized patients according to race and age groups (1990–1999). Number of patients
Diagnoses/100
(1990–1999)
hospitalizations
Deep
Venous
Age
venous
thrombo-
All discharges
Deep venous
Venous thrombo-
(years)
thrombosis
embolism
(1990–1999)
thrombosis
embolism
Caucasians 0–19
25,000
29,000
20,222,000
0.1
0.1
20–49
419,000
503,000
67,389,000
0.6
0.7
50–69
705,000
869,000
48,630,000
1.5
1.8
≥70
979,000
1,240,000
67,112,000
1.5
1.8
2,128,000
2,641,000
203,352,000
1.0
1.3
All ages African Americans 0–19 20–49
9,000
12,000
6,620,000
0.1
0.2
130,000
166,000
16,553,000
0.8
1.0
50–69
104,000
133,000
7,683,000
1.4
1.7
≥70
101,000
129,000
6,429,000
1.6
2.0
All ages
344,000
440,000
37,285,000
0.9
1.2
Asian-Pacific Islanders 0–19
—*
—*
20–49
6,000
8,000
592,000
—*
—*
2,110,000
0.3
0.4
50–69
7,000
8,000
892,000
0.8
0.9
≥70
4,000
5,000
936,000
0.4
0.5
18,000†
21,000
4,530,000
0.4
0.5
All ages * Insufficient data. †
Represents the sum of all age groups before rounding. Reprinted from Stein et al. [23], with permission from Elsevier.
lower in rural and urban Japanese than in whites and Japanese Americans [30]. Some differences of coagulation factors between Asians and whites are attributable to environmental factors, especially diet and smoking, as well as genetic differences [26, 30]. In some venographic studies of patients recovering from total hip replacement, the incidence of asymp-
tomatic DVT in Asians was reported to be comparable with rates observed in the general population in North America [19, 20]. This may suggest that Asians have a more efficient inactivation of coagulation by activated protein C or more fibrinolytic activity [24], thereby lowering the rate of PE and the rate of symptomatic DVT.
Table 15.3 Mortality from pulmonary embolism (1990–1998). Crude mortality rate Death count Race
(1990–1998)
Population*
Age-adjusted mortality rate
(deaths/100,000
deaths/100,000
population/year)
(population/year)
Caucasian
66,879
1,946,786,446
3.4
3.4
African American
14,033
293,437,621
4.8
6.9
Other
584
100,803,416
0.6
1.0
* Sum of census estimates 1990–1998. Reprinted from Stein et al. [23], with permission from Elsevier.
Clinical/laboratory Clinical/laboratory
Major gynecological surgery Hospitalized patients
Deep venous thrombosis
Deep venous thrombosis
Venogram
80 Hospital Hospital idiopathic
Primary venous thromboembolism
Deep venous thrombosis
Discharge codes
Discharge codes and chart review
Clinical/laboratory autopsy
Autopsy
Venogram
Venogram
Venogram
Venogram
Fibrinogen uptake test
Fibrinogen uptake test/venogram
Clinical/laboratory
Reprinted from Stein et al. [23], with permission from Elsevier.
* Diagnosis of deep venous thrombosis was made by venous ultrasound or venography only [16].
Autopsy Hospital
Femoral fracture
Deep venous thrombosis
Fatal pulmonary embolism
Femoral fracture
Deep venous thrombosis
Venous thromboembolism
Total hip replacement
Gynecological surgery
Deep venous thrombosis Total knee replacement
Stroke
Deep venous thrombosis
Fibrinogen uptake test
General surgery
Deep venous thrombosis
Deep venous thrombosis
Deep venous thrombosis
Fibrinogen uptake test
General surgery General surgery
Deep venous thrombosis
Deep venous thrombosis
Laboratory
Pregnancy Pregnancy
Deep venous thrombosis
Venous thromboembolism
Fibrinogen uptake test
Fibrinogen uptake test
Total hip replacement
Autopsy
Autopsy
Hysterectomy
Autopsy
Pulmonary embolism
Deep venous thrombosis
Autopsy
Pulmonary embolism
Method
6/100,000/ year
2/100,000/year
0.013
3.6
53
50
77
64
2.6
17
2.6
12
0.02
0.2
0.04
0.08
3.8
1.7
4
1.4
0.16
Asia (%)
13–19/100,000/year
21–22/100,000/year
0.27
4.2
48
48
64
54
16
55
25
25
25
0.1
0.1
0.8
35
16
54
14.0, 14.7
14
European (%)
North American/
California
California
China
China
China
Malaysia
Malaysia
Malaysia
China
China
China
Malaysia
Malaysia
China
Malaysia
Singapore
Thailand
Thailand
Thailand
Japan
Singapore
location
Asian
13
12
11, 18
21
15, 20
15, 19
15, 19
15, 19
10, 15
9, 15
8, 15
7, 15
15
17, 22
6, 17
5, 16*
4
4
3, 15
2
1, 14
References
March 12, 2007
Deep venous thrombosis
Group
Venous thromboembolism
Rate of diagnosis
Table 15.4 Comparison of rates of venous thromboembolism, pulmonary embolism, and deep venous thrombosis in Asians versus North Americans/Europeans.
BLUK077-Stein
15:39
BLUK077-Stein
March 12, 2007
15:39
Pulmonary thromboembolism in Asians
81
Asians appear to be more sensitive to warfarin than whites [31]. The target range of the International Normalized Ratio (INR) for patients with nonvalvular atrial fibrillation, 1.5–2.1 in Japanese [32], and for mechanical prosthetic heart valves, 1.5–2.5 in Japanese [33], is lower than the target range of 2.0–3.0 for most indications in patients from North America and Europe [34, 35]. Clinical problems with toxicity and dosage adjustment of warfarin have been found in patients with a genetic polymorphism to CYP2C9 [36, 37], but this polymorphism does not appear to be more frequent in Asians than whites, nor does it explain the observed greater sensitivity to warfarin [38]. Differences in the activity of CYP2C9 between Asians and whites may relate to differences in the pharmacokenetics of warfarin [38].
bolism in the Chinese in Hong Kong. Int J Cardiol 1988; 20: 373–380. Klatsky AL, Armstrong MA, Poggi J. Risk of pulmonary embolism and/or deep venous thrombosis in AsianAmericans. Am J Cardiol 2000; 85: 1334–1337. White RH, Zhou H, Romano PS. Incidence of idiopathic deep venous thrombosis and secondary thromboembolism among ethnic groups in California. Ann Intern Med 1998; 128: 737–740. Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. Geerts WW, Heit JA, Clagett GP et al. Prevention of venous thromboembolism. Chest 2001; 119(suppl): 132S– 175S. Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. Andersen BS, Steffensen FH, Sorensen HT, Nielsen GL, Olsen J. The cumulative incidence of venous thromboembolism during pregnancy and puerperium—an 11 year Danish population-based study of 63,300 pregnancies. Acta Obstet Gynecol Scand 1998; 77: 170–173. Stein PD, Patel KC, Kalra NK et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. Dhillon KS, Askander A, Doraisamy S. Postoperative deep-vein thrombosis in Asian patients is not a rarity: a prospective study of 88 patients with no prophylaxis. J Bone Joint Surg 1996; 78B: 427–430. Mok CK, Hoaglund FT, Rogoff SM, Chow SP, Ma A, Yau AC. The incidence of deep vein thrombosis in Hong Kong Chinese after hip surgery for fracture of the proximal femur. Br J Surg 1979; 66: 640–642. Dickens P, Knight BH, Ip P, Fung WS. Fatal pulmonary embolism: a comparative study of autopsy incidence in Hong Kong and Cardiff, Wales. Forensic Sci Int 1997; 90: 171–174. Chan LY, Tam WH, Lau TK. Venous thromboembolism in pregnant Chinese women. Obstet Gynecol 2001; 98: 471–475. Stein PD, Kayali F, Olson RE, Milford, CE. Pulmonary thromboembolism in Asian-Pacific Islanders in the United States: analysis of data from the National Hospital Discharge Survey and the United States Bureau of the Census. Am J Med 2004; 116: 435–442. White RH. The epidemiology of venous thromboembolism. Circulation 2003; 107: I4–I8. Svensson PJ, Dahlback B. Resistance to activated protein C as a basis for venous thrombosis. N Engl J Med 1994; 330: 517–522. Franco RF, Reitsma PH. Genetic risk factors of venous thrombosis. Hum Genet 2001; 109: 369–384.
CHAPTER 15
12
13
14
15
16 17
References 1 Hwang WS. The rarity of pulmonary thromboembolism in Asians. Singapore Med J 1968; 9: 276–279. 2 Hirst AE, Gore I, Tanaka K, Samuel I, Krishtmukti I. Myocardial infarction and pulmonary embolism. Arch Pathol 1965; 80: 365–370. 3 Atichartakarn V, Pathepchotiwong K, Keorochana S, Eurvilaichit C. Deep vein thrombosis after hip surgery among Thai. Arch Intern Med 1988; 148: 1349–1353. 4 Chumnijarakij T, Poshyachinda V. Postoperative thrombosis in Thai women. Lancet 1975; 1: 1357–1358. 5 Kueh YK, Wang TL, Teo CP, Tan YO. Acute deep vein thrombosis in hospital practice. Ann Acad Med Singapore 1992; 21: 345–348. 6 Liam CK, Ng SC. A review of patients with deep vein thrombosis diagnosed at university hospital, Kuala Lumpur. Ann Acad Med 1990; 19: 837–840. 7 Cunningham IGE, Yong NK. The incidence of postoperative deep vein thrombosis in Malaysia. Br J Surg 1974; 61: 482–483. 8 Nandi P, Wong KP, Wei WI, Ngan H, Ong GB. Incidence of postoperative deep vein thrombosis in Hong Kong Chinese. Br J Surg 1980; 67: 251–253. 9 Tso SC. Deep vein thrombosis after strokes in Chinese. Aust N Z J Med 1980; 10: 513–514. 10 Tso SC, Wong V, Chan V, Chan TK, Ma HK, Todd D. Deep vein thrombosis and changes in coagulation and fibrinolysis after gynaecological operations in Chinese: the effect of oral contraceptives and malignant disease. Br J Haematol 1980; 46: 603–612. 11 Woo KS, Tse LK, Tse CY, Metreweli C, Vallance-Owen J. The prevalence and pattern of pulmonary thromboem-
18
19
20
21
22
23
24 25
26
BLUK077-Stein
March 12, 2007
15:39
82
PART I
27 Ridker PM, Miletich JP, Hennekens CH, Buring JE. Ethnic distribution of factor V Leiden in 4047 men and women. Implications for venous thromboembolism screening. JAMA 1997; 277: 1305–1307. 28 Rees DC, Cox M, Clegg JB. World distribution of factor V Leiden. Lancet 1995; 346: 1133–1134. 29 Gregg JP, Yamane AJ, Grody WW. Prevalence of the factor V-Leiden mutation in four distinct American ethnic populations. Am J Med Genet 1997; 73: 334–336. 30 Iso H, Folsom AR, Wu KK et al. Hemostatic variables in Japanese and Caucasian men. Plasma fibrinogen, factor VIIc, factor VIIIc, and von Willebrand factor and their relations to cardiovascular disease risk factors. Am J Epidemiol 1989; 130: 925–934. 31 Takahashi H, Echizen H. Pharmacogenetics of CYP2C9 and interindividual variability in anticoagulant response to warfarin. Pharmacogenomics J 2003; 3: 202–214. 32 Yamaguchi T. Optimal intensity of warfarin therapy for secondary prevention of stroke in patients with nonvalvular atrial fibrillation: a multicenter, prospective, randomized trial. Stroke 2000; 31: 817–821.
33 Matsuyama K, Matsumoto M, Sugita T et al. Anticoagulant therapy in Japanese patients with mechanical mitral valves. Circulation 2002; 66: 668–670. 34 Hirsh J, Dalen JE, Anderson DR et al. Managing oral anticoagulant therapy. Chest 2001; 119(suppl): 3S–7S. 35 Stein PD, Alpert JS, Bussey HI, Dalen JE, Turpie AG. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 2001; 119(suppl): 220S–227S. 36 Tabrizi AR, Zehnbauer BA, Borecki IB, McGrath SD, Buchman TG, Freeman BD. The frequency and effects of cytochrome P450 (CYP) 2C9 polymorphisms in patients receiving warfarin. J Am Coll Surg 2002; 194: 267–273. 37 Aithal GP, Day CP, Kesteven PJL, Daly AK. Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications. Lancet 1999; 353: 717–719. 38 Takahashi H, Wilkinson GR, Caraco Y et al. Population differences in S-warfarin metabolism between CYP2C9 genotype-matched Caucasian and Japanese. Clin Pharmacol Ther 2003; 73: 253–263.
Prevalence, risks, and prognosis of PE and DVT
January 5, 2007
CHAPTER 16
Pulmonary thromboembolism in American Indians and Alaskan Natives 155 131 150
71
100 126
107
50
55
VTE DVT
A Afr m ic er a ic n an A s A me la ri sk ca an n N Ind at ia iv n es s
ca
si
an
s
0
au
Rates of diagnosis of deep venous thrombosis (DVT), pulmonary embolism (PE), and venous thromboembolism (VTE) are lower in American Indians and Alaskan Natives than rates in African Americans or Caucasians [1, 2]. From 1996 to 2001, the rate of diagnosis of VTE (PE and/or DVT) in American Indians/Alaskan Natives, based on combined data from the National Hospital Discharge Survey and the Indian Health Service, was 71/100,000/year, compared with 155/100,000/year in African Americans and 131/100,000 in Caucasians (Figure 16.1) [1]. Rates of diagnosis of DVT according to race are shown in Figure 16.1 [1]. The number of PEs in American Indians/Alaskan Natives was too low to give an accurate estimate of the rate of diagnosis. Only 1 patient with PE was hospitalized in Indian Health Service hospitals between 1996 and 2001. During this interval, an estimated 420,000 patients were hospitalized [1]. The rate of diagnosis of VTE among patients discharged from Indian Health Service hospital care from 1980 to 1996 was reported as 33/100,000/year in American Indians/Alaskan Natives [2]. The relatively low incidence of VTE in American Indians/Alaskan Natives would seem to be due to as yet undetermined genetic factors. A lower prevalence of Factor V Leiden in American Indians/Alaskan Natives populations (1.25%) compared with Caucasians (5.3%) perhaps contributes to the lower incidence of VTE in American Indians/Alaskan Natives [3]. The concept that racial groups can differ genetically and the differences can have medical importance has recently been discussed [4]. The possibility that American Indians/Alaskan Natives have different diets or lifestyles
DX/100,000/yr
16
18:14
C
BLUK077-Stein
Figure 16.1 Rates of diagnosis of deep venous thrombosis (DVT) and venous thromboembolic disease (VTE) in American Indians/Alaskan Natives, Caucasians, and African Americans from 1996 to 2001 based on combined data from the National Hospital Discharge Survey and the Indian Health Service. Rates of diagnosis of DVT and of VTE were lower in Indians/Alaskan Natives than Caucasians or African Americans (all differences P < 0.001). The rates of diagnosis in Caucasians and African Americans were comparable. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)
that affect the rate of diagnosis of VTE cannot be excluded [5].
References 1 Stein PD, Kayali F, Olson RE, Milford, CE. Pulmonary thromboembolism in American Indians and Alaskan Natives. Arch Intern Med 2004; 164: 1804–1806. 2 Hooper WC, Holman RC, Heit, JA, Cobb N. Venous thromboembolism hospitalizations among American Indians and Alaska Natives. Thrombosis Res 2003; 108: 273– 278.
83
BLUK077-Stein
January 5, 2007
18:14
84
3 Ridker PM, Miletich JP, Hennekens CH, Buring JE. Ethnic distribution of Factor V Leiden in 4047 men and women. Implications for venous thromboembolism screening. JAMA 1997; 277: 1305– 1307.
PART I
Prevalence, risks, and prognosis of PE and DVT
4 Burchard EG, Ziv E, Coyle N et al. The importance of race and ethnic background in biomedical research and clinical practice. N Engl J Med 2003; 348: 1170–1175. 5 Cooper RS, Kaufman JS, Ward R. Race and genomics. N Engl J Med 2003; 348: 1166–1170.
BLUK077-Stein
April 14, 2007
17
14:54
CHAPTER 17
Venous thromboembolism in patients with cancer
Introduction Venous thromboembolism (VTE) is one of the most common complications in patients with cancer [1, 2]. The association of venous thrombosis with gastric carcinoma was described in 1865 [3]. A fourfold
increased risk of VTE has been found among patients with malignant neoplasm [4, 5]. Among patients insured through the Medicare program (patients aged 65 years or older), 0.6% of patients admitted with malignancy also had VTE [5]. A higher rate was reported by others [6]. Among patients hospitalized with solid
Table 17.1 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis in patients with cancer as well as no cancer. All ages 1979–1999 Diagnoses/ 100 hospitalizations
Number of patients Cancers Pancreas Brain Myeloproliferative, other lymphatic/ hematopoeitic Stomach Lymphoma, lymphosarcoma, reticulosarcoma Uterus Trachea, bronchus, and lung Esophagus Prostate Rectum, rectosigmoid junction, anus Kidney Colon Ovary Liver, gallbladder, intra- and extrahepatic ducts Leukemia Breast (female) Cervix Bladder Lip, oral cavity, pharynx Average incidences No cancer
VTE
PE
DVT
All discharges
VTE
PE
DVT
51,000 27,000 15,000
14,000 8,000 *
41,000 22,000 13,000
1,176,000 772,000 521,000
4.3 3.5 2.9
1.2 1.0 *
3.5 2.8 2.5
24,000 79,000
7,000 20,000
20,000 63,000
887,000 3,182,000
2.7 2.5
0.7 0.6
2.3 2.0
26,000 170,000 12,000 95,000 30,000
6,000 56,000 *
1,180,000 8,120,000 603,000 4,643,000 1,457,000
2.2 2.1 2.0 2.0 2.1
0.5 0.6 *
30,000 11,000
21,000 129,000 8,000 72,000 21,000
0.6 0.7
1.8 1.6 1.3 1.6 1.4
19,000 69,000 31,000 13,000
5,000 24,000 8,000 6,000
15,000 49,000 27,000 8,000
939,000 3,614,000 1,669,000 703,000
2.0 1.9 1.9 1.8
0.5 0.6 0.5 0.9
1.6 1.4 1.6 1.1
45,000 82,000 14,000 21,000 <5,000
10,000 22,000 *
2,655,000 4,932,000 875,000 2,011,000 849,000
1.7 1.7 1.6 1.0 <0.6
0.4 0.4 *
5,000 *
38,000 64,000 12,000 17,000 *
0.3 *
1.4 1.3 1.4 0.8 *
823,000 6,854,000
232,000 2,212,000
640,000 5,124,000
40,788,000 662,309,000
2.0 1.0
0.6 0.3
1.6 0.8
* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.
85
BLUK077-Stein
April 14, 2007
14:54
PART I
tumors, 7.8% also had VTE and 1.1% had PE (pulmonary embolism) [7]. The risk of VTE varies according to the type of cancer [5–8]. Carcinoma of the pancreas has been associated with the greatest risk of VTE [9]. High rates of VTE have been reported among patients with solid neoplasms originating at several sites including the lung, ovary, brain, pancreas, stomach, kidney, colon [5–7]. Patients with lymphoma, leukemia, and myeloproliferative syndromes also often have associated VTE [5, 6]. However, cancers originating at some sites, including head and neck and bladder, did not have a high incidence of associated VTE [5, 6]. A review of cancer patients who underwent surgery showed that they had twice the risk of postoperative deep venous thrombosis (DVT) as noncancer patients who underwent similar procedures [10].
Several mechanisms may be involved in the pathogenesis of thromboembolic events in patients with cancer [11]. These include (1) tumor cell procoagulants and/or cytokines, (2) tumor associated inflammatory cell procoagulants and/or cytokines, and (3) mediators of platelet adhesion or aggregation generated by tumor cells and/or tumor-associated inflammatory cells [11]. Stasis and endothelial damage may also be involved in the pathogenesis of thromboembolic events in patients with cancer [11]. The extent of cancer influences the risk of VTE [12]. Chemotherapy and radiation increase the risk of VTE [10, 12]. Risk factors may interact [10]. The risk of VTE increases with age [13] and age is also associated with malignancy. Among cancer patients undergoing surgery, advanced age, debility, prolonged and difficult surgery, and a lengthy and
(a)
4 3.5 3 2.5 2 1.5 1 0.5 0
Prevalence, risks, and prognosis of PE and DVT
Cancer patients
Noncancer patients 99
97
95
93
91
89
87
85
83
81
79
VTE in hospitalized cancer and noncancer patients (%)
86
(b)
2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
Cancer patients
Noncancer patients 99
97
95
93
91
89
87
85
83
81
79
PE in hospitalized cancer and noncancer patients (%)
Year
(c)
4 Cancer patients
3.5 3 2.5 2 1.5 1
Noncancer patients
0.5 0
99
97
95
93
Year
91
89
87
85
83
81
79
DVT in hospitalized cancer and noncancer patients (%)
Year
Figure 17.1 (a) Incidence of venous thromboembolism (VTE) in patients hospitalized with cancer and those without cancer. From 1989 to 1999, there was a prominent increase in the incidence of VTE in patients discharged with cancer. The incidence of VTE in patients without cancer also increased but the slope was lower. (b) Incidence of pulmonary embolism (PE) in patients hospitalized with cancer and those without cancer. From 1989 to 1999, the incidence of PE in patients discharged with cancer increased. The incidence of PE in patients without cancer increased a little. (c) Incidence of deep venous thrombosis (DVT) in patients hospitalized with cancer and those without cancer. From 1989 to 1999, there was a prominent increase in the incidence of DVT in patients discharged with cancer. The incidence of DVT in patients without cancer also increased but the slope was lower. (Reprinted from Stein et al. [15], with permission from Elsevier.)
BLUK077-Stein
April 14, 2007
CHAPTER 17
14:54
87
Venous thromboembolism in cancer patients
twice the incidence of VTE in hospitalized patients who did not have any of these malignancies, 6,854,000 of 662,309,000 (1.0%) [15]. The highest incidence of VTE was in patients with carcinoma of the pancreas (4.3%) (Table 17.1). The lowest incidence of VTE was
complicated postoperative course add to the risk of DVT [14]. Among patients hospitalized from 1979 to 1999, with any of the malignancies listed in Table 17.1, 827,000 of 40,787,000 (2.0%) had VTE [15]. This was
(a)
4 3.5 3 2.5 2 1.5 1 0.5
Pancreas Brain Myeloprol Stomach Lymphoma Uterus Lung Esophagus Prostate Rectal Kidney Colon Ovary Liver Leukemia Breast Cervix Bladder
Relative risk of VTE in cancer patients
4.5
4.5 4 3.5 3 2.5 2 1.5 1 0.5 Pancreas Brain Myeloprol Stomach Lymphoma Uterus Lung Esophagus Prostate Rectal Kidney Colon Ovary Liver Leukemia Breast Cervix Bladder
Relative risk of PE in cancer patients
(b)
(c) 5 4 3 2 1 Cervix
Bladder
Breast
Liver
Leukemia
Colon
Ovary
Rectal
Kidney
Prostate
Lung
Esophagus
Uterus
Stomach
Lymphoma
Myeloprol
Brain
0 Pancreas
Relative risk of DVT in cancer patients
Figure 17.2 (a) Relative risks of venous thromboembolism (VTE) in patients hospitalized with cancer compared to those without cancer. The relative risk of VTE ranged from 1.02 to 4.34. (b) Relative risks of pulmonary embolism (PE) in patients hospitalized with cancer compared to those without cancer. The relative risk of PE ranged from 0.77 to 3.66. (c) Relative risks of deep venous thrombosis (DVT) in patients hospitalized with cancer compared to those without cancer. The relative risk of DVT ranged from 1.07 to 4.65. (Reprinted from Stein et al. [15], with permission from Elsevier.)
BLUK077-Stein
April 14, 2007
14:54
88
PART I
in patients with carcinoma of the bladder (1.0%) (Table 17.1). The incidence of VTE in patients with carcinoma of the lip, oral cavity or pharynx was lower, but data were insufficient to calculate the incidence. Pulmonary embolism among patients hospitalized with carcinoma of the pancreas occurred in 1.2% and in patients with carcinoma of the bladder in 0.3% [15] (Table 17.1). Patients hospitalized with malignancies, on average, had twice the incidence of PE as in those who did not have malignancies (0.6% versus 0.3%) (Table 17.1).
The incidence of DVT associated with patients hospitalized with various neoplasms are shown in Table 17.1. As in patients with VTE and PE, patients hospitalized with cancer had twice the incidence of DVT as those who did not have cancer (1.6% versus 0.8%) (Table 17.1). The incidence of VTE in hospitalized patients with malignant neoplasms began to sharply increase in 1989 [15] (Figure 17.1a). The incidence of VTE in hospitalized patients who did not have cancer also showed an increasing incidence beginning in 1989, but
Prevalence, risks, and prognosis of PE and DVT
Table 17.2 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis in patients with cancer as well as no cancer. Age group 40–59, 1979–1999 Diagnoses/ Number of patients Cancers
VTE
PE
100 hospitalizations DVT
All discharges
VTE
PE
DVT
Pancreas
10,000
*
10,000
217,000
4.6
*
4.6
Brain
11,000 *
*
9,000 *
206,000
5.3 *
*
4.4 *
Myeloproliferative, other
*
49,000
*
lymphatic/ hematopoeitic Stomach Lymphoma, lymphosarcoma,
5,000
*
5,000
171,000
2.9
*
2.9
18,000
5,000
16,000
738,000
2.4
0.6
2.2
reticulosarcoma 6,000
*
5,000
318,000
1.9
*
1.6
52,000 *
13,000 *
43,000 *
2,057,000 157,000
2.5 *
0.6 *
2.1 *
Prostate
5,000
*
*
332,000
1.5
*
*
Rectum, rectosigmoid
5,000
*
*
291,000
1.7
*
*
Uterus Trachea, bronchus, and lung Esophagus
junction, anus *
*
*
232,000
*
*
*
Colon
11,000
*
8,000
612,000
1.8
*
1.3
Ovary
9,000 *
*
8,000 *
572,000
1.6 *
*
1.4 *
8,000
5,000
7,000
358,000
2.2
1.4
2.0
26,000
6,000 *
21,000
1,734,000
1.5
1.2
347,000
*
5,000 *
247,000
1.4 *
0.3 *
*
*
293,000
Kidney
Liver, gallbladder, intra- and
*
149,000
*
extrahepatic ducts Leukemia Breast (female) Cervix Bladder Lip, oral cavity, pharynx Average incidences No cancer
5,000 * *
*
1.4 *
*
*
*
171,000
29,000
137,000
9,080,000
1.9
0.3
1.5
1,693,000
504,000
1,295,000
131,893,000
1.3
0.4
1.0
* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.
BLUK077-Stein
April 14, 2007
CHAPTER 17
14:54
89
Venous thromboembolism in cancer patients
the increasing incidence was not as high. The incidence of PE among patients with cancer increased in the 1990s (Figure 17.1b), but the increasing incidence was not as high as with DVT (Figure 17.1c) or VTE (Figure 17.1a). The relative risk of VTE, PE, and DVT among hospitalized patients with various malignancies, compared to patients who did not have any of these malignancies, is shown in Figures 17.2a–17.2c [15]. The highest relative risk of VTE, 4.3, was among hospitalized patients with carcinoma of the pancreas. The lowest relative risk, 1.0, was among patients with carcinoma
of the bladder. The relative risks of PE and DVT paralleled the relative risks of VTE (Figures 17.2b and 17.2c). Among patients aged 40–59 years who were hospitalized with cancer, average incidence of VTE and DVT were 46 to 50% higher than in patients who did not have these types of cancer but the incidence of PE was not higher [15] (Table 17.2). Among patients aged 60–79 years who were hospitalized with cancer, average incidence of VTE and DVT were only 29 to 42% higher than in patients who did not have these cancers (Table 17.3).
Table 17.3 Venous thromboembolism, pulmonary embolism and deep venous thrombosis in patients with cancer as well as no cancer. Age group 60–79, 1979–1999 Diagnoses/ Number of patients Cancers
VTE
PE
100 hospitalizations DVT
All discharges
VTE
PE
DVT
Pancreas
37,000
724,000
5.1
9,000
245,000
4.9
1.7 *
3.9
12,000
12,000 *
28,000
Brain
9,000
*
8,000
263,000
3.4
*
3.0
Myeloproliferative, other
3.7
lymphatic/ hematopoeitic Stomach
13,000
*
11,000
494,000
2.6
*
2.2
Lymphoma, lymphosarcoma,
43,000
12,000
33,000
1,432,000
3.0
0.8
2.3
16,000
5,000
13,000
697,000
2.3
0.7
1.9
105,000
76,000 *
5,159,000
2.0
7,000
38,000 *
359,000
1.9
0.7 *
1.5 *
Prostate
66,000
21,000
51,000
2,992,000
2.2
0.7
1.7
Rectum, rectosigmoid
19,000
7,000
13,000
846,000
2.2
0.8
1.5
reticulosarcoma Uterus Trachea, bronchus, and lung Esophagus
junction, anus Kidney
12,000
*
9,000
507,000
2.4
*
1.8
Colon
44,000
32,000
2,087,000
2.1
17,000
15,000
842,000
2.0
0.8 *
1.5
Ovary
16,000 *
6,000
*
4,000
383,000
1.6
*
1.0
Liver, gallbladder, intra- and
1.8
extrahepatic ducts Leukemia
25,000
5,000
21,000
1,060,000
2.4
0.5
2.0
Breast (female)
42,000
13,000 *
32,000
2,300,000
1.8
1.4
4,000
263,000
1.9
0.6 *
*
11,000 *
1,217,000
*
425,000
1.2 *
*
0.9 *
Cervix Bladder Lip, oral cavity, pharynx Average incidences No cancer
5,000 14,000 *
*
1.5
492,000
129,000
370,000
22,033,704
2.2
0.6
1.7
3,018,000
1,038,000
2,151,000
174,352,000
1.7
0.6
1.2
* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.
BLUK077-Stein
April 14, 2007
14:54
90
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 17.4 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis according to sex in patients with cancer as well as no cancer. Incidence (%) VTE
PE
DVT
Cancers
Male
Female
Male
Female
Male
Female
Pancreas
4.4
4.3 3.0
1.3 *
3.5
3.9
1.2 *
3.5
Brain
3.4
2.2
Myeloproliferative, other lymphatic/hematopoeitic
2.8
3.1
*
*
2.3
2.6
Stomach
3.0
2.1
*
*
2.4
1.8
Lymphoma, lymphosarcoma, reticulosarcoma
2.4
2.6
0.8
0.5
1.8
2.2
Trachea, bronchus, and lung
2.1 2.2
2.1 *
0.7 *
0.7 *
1.5
Esophagus
1.6
1.7 *
Rectum, rectosigmoid junction, anus
1.9
2.2 2.2
0.7 *
1.5
1.9
0.7 *
1.4
Renal
1.5
1.7
Colon
2.0 2.6
0.7 *
0.7
Liver/gallbladder/extrahepatic duct
1.7 *
1.2 *
1.5
Leukemia
1.3
2.2
*
Bladder
1.1 *
*
0.5 *
Lip, oral cavity, pharynx
1.1 *
*
*
0.8 *
Average incidences
2.1
2.1
0.3
0.4
1.6
1.6
No cancer
1.1
1.0
0.4
0.3
0.8
0.8
1.4
1.0
1.5 1.9 * *
* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.
On average, the incidences of VTE, PE, and DVT in patients with cancer were similar in men and women, but differences were shown with various malignancies [15] (Table 17.4). The incidence of VTE, PE, and DVT according to race, on average, were similar in African Americans and whites (Table 17.5). Differences were observed with various malignancies. The rates we reported [15] were the same order of magnitude as reported by Levitan et al. [6]. In patients with carcinoma of pancreas, the relative risk of VTE was 4.3 times that of noncancer patients. The rates of VTE, PE, and DVT showed an increase starting in the late 1980s. The incidences of DVT and PE increase exponentially with age [13]. However, the relative risk of VTE among patients aged 40–59 years was paradoxically higher than in older patients. This reflected a disproportionately higher incidence of VTE in the older noncancer patients. Cancer age did not greatly increase the incidence of VTE in older patients.
The cancers that we evaluated were shown by others to have a higher incidence of VTE than in patients without cancer [5–7]. The extent of malignancy, invasion, metastasis, chemotherapy, radiotherapy, and surgery, which may increase the incidence of associated VTE, was not available on discharge codes.
Pulmonary embolism as a cause of death in patients who died with cancer Among 506 autopsies of patients in whom 96% had “some sort of neoplasm,” 35 (7%) died from PE [16]. Numerous other investigations describe the association of PE with cancer, but do not describe the frequency of fatal PE in patients who died with cancer. Among patients who died from PE, 22.9% had cancer [17]. Among patients with cancer who died from 1980 to 1998, the estimated average frequency of PE as the cause of death, after adjustment for the
BLUK077-Stein
April 14, 2007
CHAPTER 17
14:54
91
Venous thromboembolism in cancer patients
Table 17.5 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis according to race in patients with cancer as well as no cancer. Incidence (%) VTE
PE
DVT
Cancers
White
Black
White
Black
White
Black 2.9
Pancreas
4.4
4.2
1.2
1.7
3.5
Brain
3.8
5.1
1.0
3.2
3.2
2.0
Myeloproliferative, other lymphatic/hematopoeitic
3.0
3.7
0.7
0.4
2.6
3.3
Stomach
2.9
2.6
0.9
0.6
2.4
2.1
Lymphoma, lymphosarcoma, reticulosarcoma
2.5 2.2
0.6 *
0.5 *
2.1
Uterus
2.3 *
2.1 *
Trachea, bronchus, and lung
2.1
Esophagus
2.2
1.8 *
0.7 *
0.7 *
1.9 1.6
1.3 *
1.5
Prostate
2.2
1.6
0.6
0.6
1.7
1.3
Rectum, rectosigmoid junction, anus
1.9
3.3
0.7
1.2
1.3
2.1
Kidney
1.9
1.7
0.5
0.5
1.5
1.3
Colon
1.8
2.3
0.6
0.7
1.3
1.7
Ovary
1.9 1.8
0.5 *
0.5 *
1.6
Liver, gallbladder, intra- and extrahepatic ducts
2.0 *
1.6 *
1.2
Leukemia
1.8
1.8
0.4
0.6
1.6
1.2
Breast (female)
1.7 1.8
0.4 *
0.4 *
1.3
Cervix
1.7 *
1.2 *
Bladder
*
*
*
*
*
*
0.8 *
*
Lip, oral cavity, pharynx
1.0 *
Average incidences
2.2
1.8
0.5
0.6
1.7
1.3
No cancer
1.1
0.9
0.4
0.3
0.8
0.7
1.6
*
* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.
myeloproliferative disease and lowest in patients who died with cancer of the liver, gallbladder, intra- and extrahepatic ducts. The rate of fatal PE among patients who died with cancer decreased from 1980 to 1998 (Figure 17.3). This
0.6 0.5 0.4 0.3 0.2 0.1 0 98
96
94
92
Year
90
88
86
84
82
80
Figure 17.3 Frequency of fatal pulmonary embolism (PE) among patients who died with cancer from 1980 to 1998. The frequency declined over the 19-year period of study. (Reprinted from Stein et al. [18], with permission from Elsevier.)
Deaths from PE in patients who died with cancer (%)
inaccuracy of death certificates, was between 0.60 and 1.05% [18]. However, these values appear low compared to the prevalence of large or fatal PE in all patients at autopsy, about 4% (Chapter 1). The incidence of fatal PE was highest among those who died with
BLUK077-Stein
April 14, 2007
14:54
92
PART I
presumably reflects an improvement in achieving an early diagnosis and more vigorous prophylaxis over these 19 years.
8 Thodiyil PA, Kakkar AK. Variation in relative risk of venous thromboembolism in different cancers. Thomb Haemost 2002; 87: 1076–1077. 9 Rickles FR, Edwards RL. Activation of blood coagulation in cancer: Trousseau’s syndrome revisited. Blood 1983; 62: 14–31. 10 Bergqvist D. Venous thromboembolism and cancer: prevention of VTE. Thromb Res 2001; 102: V209–V213. 11 Rickles FR, Levine M, Edwards RL. Hemostatic alterations in cancer patients. Cancer Metast Rev 1992; 11: 237–248. 12 Lee AYY, Levine MN. Venous thromboembolism and cancer: risks and outcomes. Circulation 2003; 107: I-17–I-21. 13 Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260–2265. 14 Gallus AS. Prevention of post-operative deep leg vein thrombosis in patients with cancer. Thromb Haemost 1997; 78: 126–132. 15 Stein PD, Beemath A, Meyers FA, Skaf E, Sanchez J, Olson RE. Incidence of venous thromboembolism in patients hospitalized with cancer. Am J Med 2006; 119: 60–68. 16 Ambrus JL, Ambrus CM, Mink IB, Pickren JW. Causes of death in cancer patients. J Med 1975; 6: 61–64. 17 Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979–1998: an analysis using multiple-cause mortality data. Arch Intern Med 2003; 163: 1711–1717. 18 Stein PD, Beemath A, Meyers FA, Kayali F, Skaf E, Olson RE. Pulmonary embolism as a cause of death in patients who died with cancer. Am J Med 2006; 119: 163–165.
References 1 Arklel YS. Thrombosis and cancer. Semin Oncol 2000; 27: 362–374. 2 Donati MB. Cancer and thrombosis. Haemostatis 1994; 24: 128–131. 3 Trousseau A. Phlegmassia alba dolens. In: Clinique Medicale de l’Hotel-Dieu de Paris, Vol. 3. New Sydenham Society, London, 1865: 94. Quoted by Rickles and Edwards in Reference [9]. 4 Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ, III. Risk factors for deep vein thrombosis and pulmonary embolism. A populationbased case–control study. Arch Intern Med 2000; 160: 809– 815. 5 Blom JW, Doggen CJM, Osanto S, Rosendaal FR. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005; 293: 715–722. 6 Levitan N, Dowlati A, Remick SC et al. Rates of initial and recurrent thromboembolic diseases among patients with malignancy versus those without malignancy. Risk analysis using Medicare claims data. Medicine 1999; 78: 285–291. 7 Sallah S, Wan JY, Nguyen NP. Venous thrombosis in patients with solid tumors: determination of frequency characteristics. Thromb Haemost 2002; 87: 575–579.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 28, 2007
18
15:12
CHAPTER 18
Venous thromboembolism in patients with heart disease
Patients with heart failure Heart failure (HF) is considered a major risk factor for venous thromboembolism (VTE), defined as pulmonary embolism (PE) and/or deep venous thrombosis (DVT) [1–6]. In patients with advanced HF, thrombophlebitis of the pelvic or leg veins is a common complication and this may also be a source of PE [7]. Based on data from the National Hospital Discharge Survey (NHDS) [8], among hospitalized patients with HF, PE was diagnosed in 0.73%, DVT in 1.03%, and
VTE in 1.63% [9] (Table 18.1). Among hospitalized patients who were not diagnosed with HF, PE was diagnosed in 0.34%, DVT was diagnosed in 0.85%, and VTE in 1.11%. The prevalence of PE in HF was similar to findings reported in smaller investigations that used defined criteria for HF [10, 11]. Others showed that HF patients with lower ejection fractions had a higher risk of thromboembolic events [10, 12]. The reported frequency of PE in patients with HF has ranged widely from 0.9 to 39% of patients [3, 7, 10, 12–15]. At autopsy the incidence of PE in patients with HF ranged
Table 18.1 Prevalence and relative risk of pulmonary embolism, deep venous thrombosis, and venous thromboembolism in hospitalized patients according to age groups. n/N (%)* Age groups (yrs)
HF
No HF
HF vs. No HF(95% CI)
<40
13,000/1,129,000 (1.15)
333,000/339,063,000 (0.10)
11.72 (11.52–11.93)
40–59
65,000/6,410,000 (1.01)
655,000/164,010,000 (0.40)
2.54 (2.52–2.56)
60–79
236,000/28,745,000 (0.82)
1,229,000/205,140,000 (0.60)
1.37 (1.36–1.38)
>80
157,000/22,590,000 (0.69)
404,000/74,491,000 (0.54)
1.28 (1.27–1.29)
All ages
431,000/58,873,000 (0.73)
2,662,000/782,704,000 (0.34)
2.15 (2.15–2.16)
<40
19,000/1,129,000 (1.68)
1,045,000/339,063,000 (0.31)
5.46 (5.38–5.54)
40–59
85,000/6,410,000 (1.33)
1,762,000/164,010,000 (1.07)
1.23 (1.23–1.24)
60–79
317,000/28,745,000 (1.10)
2,833,000/205,140,000 (1.38)
0.80 (0.80–0.80)
>80
281,000/22,590,000 (1.24)
948,000/74,491,000 (1.27)
0.98 (0.97–0.98)
All ages
607,000/58,873,000 (1.03)
6,683,000/782,704,000 (0.85)
1.21 (1.20–1.21)
<40
30,000/1,129,000 (2.66)
1,304,000/339,063,000 (0.38)
6.91 (9.83–6.99)
40–59
139,000/6,410,000 (2.17)
2,227,000/164,010,000 (1.36)
1.60 (1.59–1.61)
60–79
509,000/28,745,000 (1.77)
3,749,000/205,140,000 (1.83)
0.97 (0.97–0.97)
>80
405,000/22,590,000 (1.79)
1,256,000/74,491,000 (1.69)
1.06 (1.06–1.07)
All ages
960,000/58,873,000 (1.63)
8,660,000/782,704,000 (1.11)
1.47 (1.47–1.48)
Pulmonary embolism
Deep venous thrombosis
Venous thromboembolism
* 95% confidence intervals are all ≤0.04%. HF, heart failure; yrs, years; CI, confidence interval. Reproduced from Beemath et al. [9], with permission from Elsevier.
93
BLUK077-Stein
March 28, 2007
15:12
94
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 18.2 Prevalence and relative risks of pulmonary embolism and deep venous thrombosis in hospitalized patients according to sex and race. Pulmonary embolism Prevalence (%)* HF
No HF
Deep venous thrombosis Prevalence (%)*
Relative risk HF vs. No HF
HF
No HF
Relative risk HF vs. No HF
Sex Male
0.69
0.38
1.80 (1.79–1.80)
0.89
0.90
0.98 (0.98–0.99)
Female
0.77
0.31
2.45 (2.44–2.46)
1.14
0.82
1.39 (1.39–1.40)
Race White
0.75
0.36
2.10 (2.09–2.11)
1.03
0.89
1.15 (1.15–1.15)
African American
0.85
0.30
2.82 (2.79–2.84)
1.13
0.79
1.44 (1.43–1.45)
* 95% confidence intervals are all ≤0.02%. HF, heart failure. Reproduced from Beemath et al. [9], with permission from Elsevier.
DVT and PE in patients hospitalized with heart failure (%)
from 28 to 48% [14, 16, 17]. The reported frequency of DVT in patients with HF also ranged widely from 10 to 59% [2, 3, 5]. The relative risk of PE in patients with HF was highest in patients <40 years of age (relative risk 11.72) and the relative risk for DVT was 5.46 [9] (Table 18.1). In patients aged 40–59 years, the relative risk for PE was 2.54 and for DVT 1.23. With older age groups the relative risks for both PE and DVT decreased (Table 18.1). With increasing age and its accompanying risk factors for PE and DVT, other risk factors balance or outweigh the risk of HF alone. In older patients therefore, the higher relative risk for PE or DVT is thereby reduced or eliminated. The high relative risk of PE and DVT in younger adults is readily explained, but the somewhat higher rates of PE and DVT in younger adults are not explained. This observation is not in accordance with
previous findings of an increased risk of PE and DVT with age [18]. The relative risk of PE was higher in patients with HF compared to those patients with no HF among both women and men and African Americans and whites [9] (Table 18.2). The relative risk of DVT was higher in patients with HF compared with no HF among women but not in men. Both African Americans and whites with HF had higher relative risks for DVT compared with those who did not have HF (Table 18.2). The rate of PE among patients hospitalized with HF decreased from 1.38% in 1979–1981 to 0.61% in 2000– 2003 [9] (Figure 18.1). The rate of diagnosis of DVT in patients hospitalized with HF decreased from 0.89% in 1979–1981 to 0.71% in 1988–1990 and then increased to 1.35% in 2000–2003 (Figure 18.1).
1.6 1.4
DVT
1.2 1 0.8 0.6
PE
0.4 '79−81 '82−84 '85−87 '88−90 '91−93 '94−96 '97−99 '00−03 Years
Figure 18.1 Rates of pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized patients with heart failure (HF). Data were averaged over 3-year periods from 1979 to 1999 and over the 4-year period 2000–2003. The rate of PE decreased from 1979–1981 to 2000–2003. The rate of DVT decreased from 1979–1981 to 1988–1990. The rate of DVT then increased from 1988–1990 to 2000–2003. (Reproduced from Beemath et al. [9], with permission from Elsevier.)
March 28, 2007
CHAPTER 18
15:12
95
VTE in patients with heart failure
Figure 18.2 Frequency of fatal pulmonary embolism (PE) among adults who died with heart failure (HF) from 1980 to 1998. The frequency declined over the 19-year period of study. (Reproduced from Beemath et al. [27], with permission from Elsevier.)
Deaths from PE in adults who died with heart failure (%)
BLUK077-Stein
6 5 4 3 2 1 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 Year
Figure 18.3 Frequency of fatal pulmonary embolism (PE) among adults who died with heart failure (HF) among the different age groups. (Reproduced from Beemath et al. [23], with permission from Elsevier.)
Deaths from PE in adults who died with heart failure (%)
Heart failure appears to be a stronger risk factor in woman. Dries et al. [10] reported a significant difference between the sexes in the distribution of thromboembolic events in patients with HF. Among women with HF, 24% had PE whereas among men with HF 14% had PE [10]. African Americans with HF also had a somewhat higher risk for PE and DVT than whites. In 2004, it was recommended that patients hospitalized with HF should receive antithrombotic prophylaxis [19]. Roberts et al. reported PE as the cause of death in 12 of 139 (9%) autopsied patients who died with HF [15]. Goldhaber et al. reported PE as the cause of death at autopsy in 13 of 41 (32%) patients with HF [16]. In other studies, PE was found in the wide range of 0.4–50% of autopsied patients who died with HF, but whether PE caused or contributed to these deaths was not stated [11, 14, 20–22]. We investigated the rate of death from PE in patients who died with HF from 1980 to 1998 based on data from death certificates, as listed by the United States Bureau of the Census [23].
Among adults with HF who died over a 19-year period of study, PE was the listed cause of death in 20,387 of 755,807 (2.7%) [23]. Assuming a sensitivity of 26.7% for the death certificate diagnosis of fatal PE [24], the frequency of fatal PE among patients who died with HF over the 19-year period of study would be 10.1%. The frequency of death from PE in patients who died with HF decreased from 5.0% in 1980 to 1.6% in 1998 (Figure 18.2). These rates were not adjusted for the 26.7% sensitivity of PE diagnoses on death certificates. We assume that whatever inaccuracy exists in the death certificates was constant throughout the 19-year period of study. The trend of a decreasing rate of death from PE in adults who died with HF, therefore, would appear to be accurate. The decreasing rate of fatal PE among patients who died with HF during the 19-year period of study presumably reflects an improvement in achieving an early diagnosis and more vigorous prophylaxis. The risk of PE in patients with HF is twice the risk of PE observed in patients who do not have
12 10 8 6 4 2 0
20−34
35−44 45−54
55−64
65−74
Age group (years)
75−84
85−99
BLUK077-Stein
March 28, 2007
15:12
96
HF [9]. Antithrombotic prophylaxis is recommended in such patients [19]. The rate of fatal PE in adults who died with HF decreased with age with the highest rate in the age group 20–34 years (10.6%) (Figure 18.3). The reason for this is not clear.
PART I
9
10
Patients with any heart disease Among 1032 autopsies of patients with heart disease, 231 patients (24.4%) had PE and 100 (9.7%) had massive PE, considered to be the cause of death [25]. Strikingly, the prevalence of PE as a cause of death was particularly high in young patients (<10 years old) with heart disease [25]. Among 59 patients who died from PE and in whom the cardiac pathology was described, 34 (58%) had right ventricular dilatation, and 8 (14%) biventricular dilatation [25]. An antemortem suspicion of PE was raised in only 18% of patients with heart disease who died from PE [25].
References 1 Shively BK. Deep venous thrombosis prophylaxis in patients with heart disease. Curr Cardiol Rep 2001; 3(1): 56–62. 2 Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160(22): 3415–3420. 3 Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151(5): 933–938. 4 Isnard R, Komajda M. Thromboembolism in heart failure, old ideas and new challenges. Eur J Heart Fail 2001; 3(3): 265–269. 5 Cogo A, Bernardi E, Prandoni P et al. Acquired risk factors for deep-vein thrombosis in symptomatic outpatients. Arch Intern Med 1994; 154(2): 164–168. 6 Jafri SM, Ozawa T, Mammen E, Levine TB, Johnson C, Goldstein S. Platelet function, thrombin and fibrinolytic activity in patients with heart failure. Eur Heart J 1993; 14(2): 205–212. 7 Segal JP, Harvey WP, Gurel T. Diagnosis and treatment of primary myocardial disease. Circulation 1965; 32: 837– 844. 8 US Department of Health and Human Services, Public Health Service, National Center for Health Statistics
11
12
13
14 15
16
17 18
19
20
21
22
Prevalence, risks, and prognosis of PE and DVT
National Hospital Discharge Survey 1979–1999 Multiyear Public-Use Data File Documentation. Available at: http://www.cdc.gov/nchs/about/major/hdasd/nhds.htm. Last accessed Jan/ 20/, 2006. Beemath A, Stein PD, Skaf E, Al Sibae MR et al. Risk of venous thromboembolism in patients hospitalized with heart failure. Am J Cardiol. 2006; 98(6): 793–795. Dries DL, Rosenberg YD, Waclawiw MA, Domanski MJ. Ejection fraction and risk of thromboembolic events in patients with systolic dysfunction and sinus rhythm: evidence for gender differences in the studies of left ventricular dysfunction trials. J Am Coll Cardiol 1997; 29(5): 1074–1080. Al-Khadra AS, Salem DN, Rand WM, Udelson JE, Smith JJ, Konstam MA. Warfarin anticoagulation and survival: a cohort analysis from the Studies of Left Ventricular Dysfunction. J Am Coll Cardiol 1998; 31(4): 749–753. Kyrle PA, Korninger C, Gossinger H et al. Prevention of arterial and pulmonary embolism by oral anticoagulants in patients with dilated cardiomyopathy. Thromb Haemost 1985; 54(2): 521–523. Dunkman WB, Johnson GR, Carson PE, Bhat G, Farrell L, Cohn JN, for The V-HeFT VA Cooperative Studies Group. Incidence of thromboembolic events in congestive heart failure. Circulation 1993; 87(6 suppl): VI94– VI101. Kinsey D, White P. Fever in congestive heart failure. Arch Intern Med 1940; 65: 163–170. Roberts WC, Siegel RJ, McManus BM. Idiopathic dilated cardiomyopathy: analysis of 152 necropsy patients. Am J Cardiol 1987; 60: 1340–1355. Goldhaber SZ, Savage DD, Garrison RJ et al. Risk factors for pulmonary embolism. The Framingham Study. Am J Med 1983; 74(6): 1023–1028. Greenstein J. Thrombosis and pulmonary embolism. South African Med J 1945; 19: 350–377. Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164(20): 2260–2265. Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126(3 suppl): 338S–400S. Harvey WP, Finch CA. Dicumarol prophylaxis of thromboembolic disease in congestive heart failure. N Engl J Med 1950; 242: 208–211. Anderson GM, Hull E. The effect of dicumarol upon the mortality and incidence of thromboembolic complications in congestive heart failure. Am Heart J 1950; 39: 697–702. Spodick DH, Littmann D. Idiopathic myocardial hypertrophy. Am J Cardiol 1958; 1: 610–623.
BLUK077-Stein
March 28, 2007
CHAPTER 18
15:12
VTE in patients with heart failure
23 Beemath A, Skaf E, Stein PD. Pulmonary embolism as a cause of death in patients who died with heart failure. Am J Cardiol 2006; 98: 1073–1075. 24 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an
97
autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154: 159–162. 25 Pulido T, Aranda A, Zevallos MA et al. Pulmonary embolism as a cause of death in patients with heart disease. An autopsy study. Chest 2006; 129: 1282– 1287.
BLUK077-Stein
March 10, 2007
19
14:23
CHAPTER 19
Venous thromboembolism in patients with ischemic and hemorrhagic stroke
Patients with stroke are at particular risk of developing deep venous thrombosis (DVT) and pulmonary embolism (PE) because of limb paralysis, prolonged bed rest, and increased prothrombotic activity [1]. Deep venous thrombosis in the paralyzed legs of patients with stroke was reported as early as 1810 by Ferriar and again by Lobstein in 1833 [2]. Among 14,109,000 patients hospitalized with ischemic stroke, PE occurred in 72,000 (0.51%), DVT occurred in 104,000 (0.74%), and VTE (venous thromboembolism) occurred in 165,000 (1.17%) [3] (Figure 19.1). Among 1,606,000 patients hospitalized with hemorrhagic stroke, rates were higher. Pulmonary embolism occurred in 11,000 (0.68%), DVT occurred in 22,000 (1.37%), and VTE occurred in 31,000 (1.93%)
[3] (Figure 19.1). The rate of VTE in hospitalized patients with ischemic stroke and with hemorrhagic stroke did not change significantly over the 25-year period of observation [3] (Figure 19.2). The higher rate of PE, DVT, and VTE among patients with hemorrhagic stroke compared with patients with ischemic stroke may represent more frequent use of antithrombotic prophylaxis in patients with ischemic stroke although treatment was not known [3]. Since 1980, with the use of modern diagnostic techniques and general awareness of the importance of antithrombotic prophylaxis, only a few previous case series have been reported, the largest of which had 607 patients [4–10]. In most, antithrombotic prophylaxis was either not given or not described [4–7, 9].
2 1.93
1.8 1.6 1.4 PE, DVT or VTE 1.2 (%) 1 0.8 0.6 0.4 0.2 0
1.37
0.68 0.51
1.17 0.74 Hemorrhagic stroke Ischemic stroke
PE DVT VTE Figure 19.1 Rates of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in hospitalized patients with ischemic and
98
hemorrhagic stroke. Data were averaged from 1979 to 2003. (Reprinted from Skaf et al. [3], with permission from Elsevier.)
March 10, 2007
99
VTE in patients with stroke
Figure 19.2 Trends in the rates of venous thromboembolism (VTE) in hospitalized patients with ischemic and hemorrhagic stroke. Data were pooled in 5-year intervals. The rate of VTE in patients with ischemic stroke and in patients with hemorrhagic stroke did not change over the 25-year period of observation. (Reprinted from Skaf et al. [3], with permission from Elsevier.)
2.8 2.4
Hemorrhagic stroke
2 1.6 1.2 Ischemic stroke 0.8 1979−84
The prevalence of PE in these patients with stroke ranged from 0 to 4%. The prevalence of PE was 0.8% among 245 patients who received antithrombotic prophylaxis [8]. The International Stroke Trial Collaboration Group showed that among patients with ischemic stroke who received low-dose heparin (5000 IU twice daily), the rate of PE within 14 days was 33 of 4860 (0.7%) and among those who received medium-dose heparin (12,500 IU twice daily), the rate of PE within 14 days was 20 of 4856 (0.4%) [11]. Some received aspirin 300 mg/day in addition. Among all patients treated with aspirin, heparin, both, or neither, the rate of PE ranged from 0.5 to 0.8% [11]. Most investigations of antithrombotic therapy for stroke [11] or to prevent DVT among stroke patients [10, 12] excluded patients with hemorrhagic stroke. Clinically apparent DVT was reported in 1.7–5.0% of patients with stroke [4–8]. Subclinical DVT occurred in 28–73% of patients with stroke, usually in the paralyzed limb [10, 12, 13]. A high proportion of patients with DVT also have subclinical PE [14]. Although deaths within a few days of stroke are usually due to the direct consequence of brain damage, those occurring over the following weeks are mainly due to potentially preventable problems including PE [4]. In the experience of some, PE is the leading cause of death during the 2–4 weeks after onset of stroke [15, 16], yet PE is one of the preventable causes of death after stroke [16–18]. Prior to the general use of antithrombotic prophylaxis (1941–1952), 26% of immediate survivors of stroke who subsequently died, died of PE [19]. More recently, among patients with stroke who died and had autopsies, PE was the cause of death
1985−89
1990−94
1995−99
2000−03
Year
in 8–16% [15, 17, 18, 20]. Others, who described the rate of PE as the cause of death in patients who died with stroke, based on death certificates and autopsy between 1961 and 1984, reported rates of 1.3–5.9% [16, 21–25]. In 1997, the International Stroke Trial Collaborative Group showed PE as a cause of death in 75 of 1781 (4.2%) of patients with ischemic stroke who died [11]. Among patients with ischemic stroke who died from 1980 to 1998, PE was the listed cause of death on death certificates in 11,101 of 2,000,963 (0.55%) [26]. Adjusted rates of fatal PE in stroke, based on an assumed sensitivity for fatal PE on death certificates of 26.7– 37.2%, were 1.5–2.1% [26]. Death rates from PE among patients who died with ischemic stroke decreased from 1980 to 1998 (Figure 19.3) [26]. The uncorrected death rate from PE
Fatal PE among stroke deaths (%)
CHAPTER 19
14:23
VTE in hospitalized patients with ischemic or hemorrhagic stroke (%)
BLUK077-Stein
0.7 0.6 0.5 0.4 '80 '82 '84 '86 '88 '90 '92 '94 '96 '98 Year
Figure 19.3 Proportion of deaths from pulmonary embolism (PE) in patients who died with ischemic stroke. The proportion of deaths from PE with ischemic stroke decreased over the 19-year period of observation. (Reprinted from Skaf et al. [26], with permission from Elsevier.)
BLUK077-Stein
March 10, 2007
14:23
100
PART I
among patients with ischemic stroke, 0.55%, was similar to the average uncorrected death certificate rate from PE in the general population (0.45%) [27]. The decreasing proportion of deaths from PE among patients with ischemic stroke who died suggests an increased and effective use of antithrombotic prophylaxis in these patients.
13 Warlow C, Ogston D, Douglas AS. Deep venous thrombosis of the legs after strokes. BMJ 1976; 1: 1178– 1181. 14 Moser KM, Fedullo PF, LittleJohn JK, Crawford R. Frequent asymptomatic pulmonary embolism in patients with deep venous thrombosis. JAMA 1994; 271: 223– 225. 15 Viitanen M, Winblad B, Asplund K. Autopsy-verified causes of death after stroke. Acta Med Scand 1987; 222: 401–408. 16 Silver FL, Norris JW, Lewis AJ, Hachinski VC. Early mortality following stroke: a prospective review. Stroke 1984; 15: 492–496. 17 Brown M, Glassenberg M. Mortality factors in patients with acute stroke. JAMA 1973; 224: 1493–1495. 18 Bounds JV, Wiebers DO, Whisnant JP, Okazaki H. Mechanisms and timing of deaths from cerebral infarction. Stroke 1981; 12: 474–477. 19 Marquardsen J. The natural history of acute cerebrovascular disease: a retrospective study of 769 patients. Acta Neurol Scand 1969; 45: 9–88. 20 Ulbrich J. Woran sterben die apoplektiker. Ther Umsch 1981; 38: 703–708. 21 Marshall J, Kaeser AC. Survival after non-haemorrhagic cerebrovascular accidents. A prospective study. BMJ 1961; 2: 73–77. 22 Baker RN, Schwartz WS, Ramseyer JC. Prognosis among survivors of ischemic stroke. Neurology 1968; 18: 933– 941. 23 Marquardsen J. The natural history of acute cerebrovascular disease. Acta Neurol Scand 1969; 45: 131–137. 24 Matsumoto N, Whisnant JP, Kurland LT, Okazaki H. Natural history of stroke in Rochester, Minnesota, 1955 through 1969: an extension of a previous study, 1945 through 1954. Stroke 1973; 4: 20–29. 25 Miah K, von Arbin M, Britton M, de Faire U, Helmers C, Maasing R. Prognosis in acute stroke with special reference to some cardiac factors. J Chronic Dis 1983; 36(3): 279–288. 26 Skaf E, Stein PD, Beemath A, Sanchez J, Olson RE. Fatal pulmonary embolism and stroke. Am J Cardiol 2006; 97: 1776–1777. 27 Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979–1998. An analysis using multiple-cause mortality data. Arch Intern Med 2003; 163: 1711–1717.
References 1 Harvey RL. Prevention of venous thromboembolism after stroke. Topics Stroke Rehab 2003; 10: 61–69. 2 Lobstein JF. Traite d’ Anatomie Pathologique, Vol 2. Levranle FG, Paris, 1833: 610. Quoted by Warlow et al. in Reference [13]. 3 Skaf E, Stein PD, Beemath A, Sanchez J, Bustamante MA, Olson RE. Venous thromboembolism in patients with ischemic and hemorrhagic stroke. Am J Cardiol 2005; 96: 1731–1733. 4 Davenport RJ, Dennis MS, Wellwood I, Warlow CP. Complications after acute stroke. Stroke 1996; 27: 415–420. 5 Dromerick A, Reding M. Medical and neurological complications during inpatient stroke rehabilitation. Stroke 1994; 25: 358–361. 6 Dobkin BH. Neuromedical complications in stroke patients transferred for rehabilitation before and after diagnostic related groups. J Neurol Rehab 1987; 1: 3–7. 7 McClatchie G. Survey of the rehabilitation outcomes of stroke. Med J Aust 1980; 1: 649–651. 8 Kalra L, Yu G, Wilson K, Roots P. Medical complications during stroke rehabilitation. Stroke 1995; 26: 990–994. 9 Subbararao BJ, Smith J. Pulmonary embolism during stroke rehabilitation. Illinois Med J 1984; 165: 328–332. 10 Turpie AGG, Levine MN, Hirsh J et al. Double-blind randomized trial of org 10172 low-molecular-weight heparinoid in prevention of deep-vein thrombosis in thrombotic stroke. Lancet 1987; 1: 523–527. 11 International Stroke Trial Collaborative Group. The International Stroke Trial (IST): a randomized trial of aspirin, subcutaneous heparin, both, or neither among 19 435 patients with acute ischaemic stroke. Lancet 1997; 349: 1569–1581. 12 McCarthy ST, Turner JJ, Robertson D, Hawkey CJ. Lowdose heparin as a prophylaxis against deep-vein thrombosis after acute stroke. Lancet 1977; 2: 800–801.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 28, 2007
20
15:14
CHAPTER 20
Pulmonary embolism and deep venous thrombosis in hospitalized adults with chronic obstructive pulmonary disease
When patients with chronic obstructive pulmonary disease (COPD) are routinely evaluated for pulmonary embolism (PE), the rate of diagnosis has been reported as 9 of 31 (29%) and 49 of 197 (25%) [1, 2]. Rates of diagnosis of PE among hospitalized patients with COPD who were not routinely evaluated for PE were much lower than found among patients with COPD who underwent routine testing for PE [3]. Over the 25-year period of evaluation (1979–2003), 58,392,000 adults ≥20 years of age were hospitalized with COPD [3]. Among these patients, PE was diagnosed in 0.65%, DVT (deep venous thrombosis) in 1.08%, and VTE (venous thromboembolism) (PE and/or DVT) in 1.58% [3] (Tables 20.1 and 20.2). Among 789,403,000 adults hospitalized with illnesses other than COPD, PE was diagnosed in 0.34%, DVT in 0.83%, and VTE in 1.09% [3] (Tables 20.1 and 20.2). The relative risk for PE in adults hospitalized with COPD compared with those who did not have COPD was 1.92 [3] (Table 20.1, Figure 20.1) [3]. Among hos-
pitalized patients aged 20–39 years with COPD, the relative risk for PE compared with hospitalized adults the same age who did not have COPD was 5.34 [3]. Among patients aged 40–59 years, the relative risk for PE decreased to 2.02. Among patients aged 60–79 years and 80–99 years, the relative risk for PE was 1.23 and 1.41, respectively. The relative risk for DVT among hospitalized adults with COPD compared with those who did not have COPD was 1.30 and as with patients with PE, the relative risk was high in younger patients (Table 20.2) [3]. In young adults other risk factors in combination with COPD are uncommon, so the contribution of COPD to the risk of PE and DVT becomes more apparent than in older patients. The relative risk for PE and DVT in all patients increases exponentially with age [4]. Congestive heart failure, cancer, and stroke have been shown to be risk factors for VTE [5–7]. The prevalence of these risks, as we showed in this study and as others have shown previously [6, 7], is age dependent. The prevalence of congestive heart failure, cancer, and
Table 20.1 Rates of pulmonary embolism and relative risk in hospitalized patients with COPD. n/N(%) Age group (yrs)
PE, COPD
PE, No COPD
Relative risk
20–39
5,500/846,000 (0.65)
313,000/257,457,000 (0.12)
5.34
40–59
67,000/9,700,000 (0.69)
653,000/190,519,000 (0.34)
2.02
60–79
226,000/35,057,000 (0.64)
1,238,000/236,328,000 (0.52)
1.23
80–99
82,000/12,789,000 (0.64)
479,000/105,099,000 (0.46)
1.41
All adults ≥20
381,000/58,392,000 (0.65)
2,684,000/789,403,000 (0.34)
1.92
PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease; yrs, years. Reprinted from Stein et al. [3], with permission.
101
BLUK077-Stein
March 28, 2007
15:14
102
PART I
Prevalence, risks, and prognosis of PE and DVT
Table 20.2 Rates of deep vein thrombosis and relative risk in hospitalized patients with COPD. n/N(%) Age group (yrs)
DVT, COPD
DVT, No COPD
Relative risk
20–39
8,000/846,000 (0.95)
952,000/257,457,000 (0.37)
2.56
40–59
98,000/9,700,000 (1.01)
1,749,000/190,519,000 (0.92)
1.10
60–79
389,000/35,057,000 (1.11)
2,761,000/236,328,000 (1.17)
0.95
80–99
138,000/12,789,000 (1.08)
1,092,000/105,099,000 (1.04)
1.04
All adults ≥20
632,000/58,392,000 (1.08)
6,554,000/789,403,000 (0.83)
1.30
DVT, deep venous thrombosis; COPD, chronic obstructive pulmonary disease; yrs, years. Reprinted from Stein et al. [3], with permission.
stroke increased with age among patients who did not have COPD [3]. The relative risk of PE in patients with COPD, therefore, decreased with age because of such competing risk factors. In younger patients with COPD, however, comorbid conditions were uncommon in the control population, so the contribution of COPD to the risk of VTE was evident. Symptomatic DVT was found in 3 of 196 (1.5%) of patients hospitalized in a respiratory intensive care unit with an exacerbation of COPD [8]. Subclinical DVT was diagnosed in an additional 18 of 196 (9.2%) [8]. The prevalence of subclinical DVT in other studies of patients with an exacerbation of COPD, each with fewer than 60 patients, ranged from 0 to 45% [9–12]. The prevalence of PE in hospitalized patients throughout the United States with COPD (0.65%),
Table 20.3 Predisposing factors in patients with COPD and suspected acute PE.
6 Relative risk for PE with COPD
compared with the rate of diagnosis of PE in hospitalized patients with COPD, all of whom underwent diagnostic tests for PE, 25–29% [1, 2] indicates that PE in patients with COPD is generally underdiagnosed. The clinical features in 108 patients in PIOPED I who had COPD and were suspected of having PE were evaluated [13]. In patients with COPD, wheezing was less frequent and crackles were more frequent in patients with PE than in those with COPD who did not have PE. The predisposing factors, symptoms, other signs, chest radiographic findings, blood gases, and alveolar–arterial oxygen differences (gradients) did not differ to a statistically significant extent in patients with COPD who had PE compared with patients with COPD who did not have PE [13] (Tables 20.3–20.6,
PE,* No. (%)
No PE, No. (%)
(n = 21)
(n = 87)
Immobilization
9 (43)
36 (41)
Surgery
7 (33)
21 (24)
2
Thrombophlebitis, ever
3 (14)
5 (6)
1 ----------------------------------------------------------
Malignancy
2 (10)
12 (14)
Trauma, lower
1 (5)
7 (8)
Stroke
1 (5)
4 (5)
Estrogen
0 (0)
2 (2)
5 4 3
0
20−39
40−59
60−79
80−99
Age (years) Figure 20.1 Relative risk for pulmonary embolism (PE) in patients with chronic obstructive pulmonary disease (COPD) compared to patients the same decade of age hospitalized without COPD. (Reprinted from Stein et al. [3], with permission.)
extremities
All differences between PE and No PE were not significant. * Some patients had more than one predisposing factor. PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].
BLUK077-Stein
March 28, 2007
CHAPTER 20
15:14
103
PE and DVT in patients with COPD
Table 20.4 Symptoms in patients with COPD and suspected acute PE.
Table 20.6 Chest radiographs: patients with COPD and suspected acute PE.
PE, No. (%)
No PE, No. (%)
PE, No. (%)
No PE, No. (%)
(n = 21)
(n = 87)
(n = 21)
(n = 87)
Dyspnea
19 (90)
80 (92)
Atelectasis
12 (57)
36 (41)
Cough
13 (62)
48 (55)
Effusion
10 (48)
31 (34)
Pleuritic pain
9 (43)
32 (37)
Volume loss
10 (48)
31 (34)
Leg swelling
9 (43)
23 (26)
Infiltrate
9 (43)
30 (33)
Wheezing
8 (38)
35 (40)
Elevated hemidiaphragm
6 (29)
18 (20)
Leg pain
4 (19)
13 (15)
Oligemia
4 (19)
16 (18)
Hemoptysis
4 (19)
7 (8)
Palpations
2 (10)
16 (18)
Angina-like pain
0 (0)
8 (9)
Figures 20.2 and 20.3). Changes of the alveolar–arterial oxygen gradients from prior values to values at the time of the suspected PE were no greater in those with PE than in those in whom PE was excluded [13] (Figure 20.4). Physicians, when confident of a low-probability clinical assessment, were usually correct in excluding
Table 20.5 Signs of acute PE in patients with COPD and suspected acute PE. PE, No. (%)
No PE, No. (%)
(n = 21)
(n = 87)
Crackles
17 (81)
46 (53)
Tachypnea (≥20/min)
15 (71)
71 (82)
Tachycardia (>100/min)
7 (33)
31 (36)
Wheezes
2 (10)
34 (39)
Deep vein thrombosis
2 (10)
9 (10)
Third heart sound
2 (10)
7 (8)
Diaphoresis
1 (5)
10 (11)
Temperature >38.5◦ C
1 (5)
7 (8)
Cyanosis
1 (5)
3 (3)
Increased pulmonary
0 (0)
11 (13)
component of second
Table 20.7 V–Q findings in 108 patients with COPD and suspected acute PE. V–Q scan probability
PE/total (%)
High
5/5 (100)
Intermediate
14/65 (22)
Low
2/33 (6)
Near normal/normal
0/5 (0)
Total
21/108 (19)
V–Q, ventilation perfusion; PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].
75
PaCO2 (mm Hg)
All differences between PE and No PE were not significant. PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].
All differences between PE and No PE were not significant. PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].
NS
60 45 30 15 0 PE (n = 14)
heart sound Pleural friction rub
0 (0)
4 (5)
Homan’s sign
0 (0)
1 (1)
PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].
NO PE (n = 57)
Figure 20.2 Partial pressure of carbon dioxide in arterial blood (PaCO2 ) while breathing room air in patients with pulmonary embolism (PE) and in patients in whom PE was excluded (No PE). The difference was not statistically significant (NS). (Reprinted with permission from Lesser et al. [13].)
BLUK077-Stein
March 28, 2007
15:14
A-a gradient (mm Hg)
104
PART I
225 200 175 150 125 100 75 50 25 0
Prevalence, risks, and prognosis of PE and DVT
NS
PE (n = 21)
NO PE (n = 83)
Figure 20.3 Alveolar–arterial (A-a) oxygen gradient among patients with pulmonary embolism (PE) and patients in whom PE was excluded (No PE). The difference was not statistically significant (NS). (Reprinted with permission from Lesser et al. [13].)
PE [correct in 29 of 30 (97%)] but physicians most often were uncertain of the diagnosis (67 patients) and rarely were confident of a high probability of PE (correct in 3 of 3) [13]. In patients with COPD, V–Q scans sometimes give diagnostic information, but less often than in those with no cardiopulmonary disease or cardiopulmonary disease exclusive of COPD [13, 14] (Table 20.7). Presumably, contrast enhanced CT is more likely to make or exclude the diagnosis of PE. The majority of ventilation–perfusion scans in patients with COPD, 60%, were interpreted as intermediate (indeterminate) probability for PE [13]. The cause of the perfusion defect, at least in patients with emphysema due to alpha-1
250
PE (n = 10)
250
A-a gradient (mm Hg)
A-a gradient (mm Hg)
150 100 50 0
antitrypsin deficiency, is destruction of the distal pulmonary arterial branches and capillary bed [15]. This is apparent on pulmonary angiograms and wedge angiograms of such patients, compared with the perfused capillary network of normal patients [15, 16] (Figures 20.5 and 20.6) (see normally perfused capillary network illustrated by wedge angiogram in Chapter 71).
NO PE (n = 41)
P < .01
NS 200
Figure 20.5 Pulmonary wedge arteriogram in a patient with emphysema associated with alpha-1 antitrypsin deficiency. There is diminished arborization of small pulmonary artery branches. (Reproduced with permission from Stein et al. [15].)
200 150 100 50 0
Pre Current
Pre
Current
Figure 20.4 Left: Alveolar–arterial (A-a) oxygen gradient among patients with pulmonary embolism (PE) who had both prior assessments and assessments at the time of the PE (current). Right: Prior and current values of A-a gradient among patients in whom PE was excluded. The A-a gradient increased in both, but the difference was not significant comparing those with PE and those who did not have PE. (Reprinted with permission from Lesser et al. [13].)
BLUK077-Stein
March 28, 2007
CHAPTER 20
15:14
105
PE and DVT in patients with COPD
(a)
(b)
Figure 20.6 (a) Pulmonary arteriogram in a patient with emphysema associated with alpha-1 antitrypsin deficiency. A diminished number of branches is shown in arteries of the lower lung zones. Capillary hypoperfusion in the lower zones shown on this film was confirmed in subsequent films. (Reproduced with permission from Stein et al. [15].) (b) Pulmonary wedge arteriogram in the same patient with emphysema associated with alpha-1 antitrypsin deficiency as shown in Figure 20.6a. There is a prominent reduction in the number of vessels branching from this artery. (Reproduced with permission from Stein et al. [15].)
References 1 Mispeleare D, Glerant JC, Audebert M et al. Pulmonary embolism and sibilant types of chronic obstructive pulmonary disease decompensations. Rev Mal Respir 2002; 19: 415–423. 2 Tillie-Leblond I, Marquette CH, Perez T et al. Pulmonary embolism in patients with unexplained exacerbation of chronic obstructive pulmonary disease: prevalence and risk factors. Ann Intern Med 2006; 144: 390– 396. 3 Stein PD, Beemath A, Meyers FA, Olson RE. Pulmonary embolism and deep venous thrombosis in hospitalized adults with chronic obstructive pulmonary disease. J Cardiovasc Med 2007 (in press). 4 Stein PD, Hull RD, Kayali F et al. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260– 2265. 5 Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism—the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 338s–400s. 6 Stein PD, Beemath A, Meyers FA, Skaf E, Sanchez J, Olson RE. Incidence of venous thromboembolism in patients hospitalized with cancer. Am J Med 2006; 119(1): 60– 68. 7 Skaf E, Stein PD, Beemath A, Sanchez J, Bustamante MA, Olson RE. Venous thromboembolism in patients with ischemic and hemorrhagic stroke. Am J Cardiol 2005; 96(12): 1731–1733. 8 Schonhofer B, Kohler D. Prevalence of deep-vein thrombosis of the leg in patients with acute exacerbation of chronic obstructive pulmonary disease. Respiration 1998; 65: 173–177. 9 Erelel M, Cuhadaroglu C, Ece T, Arseven O. The frequency of deep venous thrombosis and pulmonary embolus in acute exacerbation of chronic obstructive pulmonary disease. Respir Med 2002; 96: 515–518. 10 Pek WY, Johan A, Stan S et al. Deep vein thrombosis in patients admitted for exacerbation of chronic obstructive pulmonary disease. Singapore Med J 2001; 42: 308–311. 11 Prescott SM, Richards KL, Tikoff G et al. Venous thromboembolism in decompensated chronic obstructive pulmonary disease. A prospective study. Am Rev Respir Dis 1981; 123: 32–36. 12 Winter JH, Buckler PW, Bautista AP et al. Frequency of venous thrombosis in patients with an exacerbation of chronic obstructive lung disease. Thorax 1983; 38: 605– 608. 13 Lesser BA, Leeper KV, Stein PD et al. The diagnosis of acute pulmonary embolism in patients with chronic
BLUK077-Stein
March 28, 2007
15:14
106
PART I
obstructive pulmonary disease. Chest 1992; 102: 17– 22. 14 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603.
15 Stein PD, Leu JD, Welch MH, Guenter CA. Pathophysiology of the pulmonary circulation in emphysema associated with alpha antitrypsin deficiency. Circulation 1971; 43: 227–239. 16 Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 10, 2007
21
14:27
CHAPTER 21
Pulmonary embolism and deep venous thrombosis in hospitalized patients with asthma
Asthma is not included among the recognized risk factors for pulmonary embolism (PE) or deep venous thrombosis (DVT) [1–3]. The association of PE with asthma had been described only in case reports [4, 5]. However, among 18,041,000 patients ≥20 years of age who were hospitalized with asthma from 1979 to 2003, based on data from the National Hospital Discharge Survey (NHDS), PE was diagnosed in 93,000 (0.51%) and DVT was diagnosed in 169,000 (0.93%). Venous thromboembolism (VTE) (PE and/or DVT) occurred in 242,000 (0.97%) (Stein et al., unpublished data from the National Hospital Discharge Survey). On average, the risk for PE in patients hospitalized with asthma was higher than the risk for PE in patients who did not have asthma (relative risk 1.21) (Stein et al., unpublished data from the National Hospital Discharge Survey). The relative risk was age-dependent (Figure 21.1). Among patients aged 20–39 years, the relative risk for PE was 2.58. Among patients aged
40–59 years, the relative risk for PE decreased to 1.31. The relative risk in older patients ranged from 0.86 to 1.21 (Stein et al., unpublished data from the National Hospital Discharge Survey) (Figure 21.1). The risk for DVT in patients hospitalized with asthma, on average, was not higher than the risk for DVT in patients who did not have asthma (relative risk 0.94) (Stein et al., unpublished data from the National Hospital Discharge Survey). However, the relative risk was higher in younger patients aged 20–39 years (relative risk 1.36) (Figure 21.2). These observations indicate that young adults hospitalized with asthma have a high risk for PE and DVT in comparison to hospitalized patients the same age who do not have asthma. With increasing age and its accompanying risk factors for PE and DVT, other risk factors balance or outweigh the risk of asthma alone. In older patients, therefore, a higher relative risk for VTE in patients with asthma is thereby eliminated.
2
2
1 ----------------------------------------------------------
0
20−39
40−59
60−79
80−99
Age (years) Figure 21.1 Relative risk for pulmonary embolism (PE) in hospitalized patients with asthma compared to patients who did not have asthma shown in relation to age. The relative risk was age-dependent.
Relative risk for DVT with asthma
Relative risk for PE with asthma
3
1 ----------------------------------------------------------
0
20−39
40−59
60−79
80−99
Age (years) Figure 21.2 Relative risk for deep venous thrombosis (DVT) in hospitalized patients with asthma compared to patients who did not have asthma shown in relation to age. The relative risk was higher in younger patients.
107
BLUK077-Stein
March 10, 2007
14:27
108
PART I
References
3 Cohen AT. Venous thromboembolic disease management of the nonsurgical moderate- and high-risk patient. Semin Hematol 2000; 37: 19–22. 4 Labay Matias MV, Hervas Palazon J, Puges Bassols E, Perez Ferron J. [Status asthmaticus, adult respiratory distress syndrome and pulmonary vascular thrombosis]. An Esp Pediatr 1985; 22: 169–170. 5 Divac A, Djordjevic V, Jovanovic D et al. Recurrent pulmonary embolism in a patient with asthma. Respiration 2004; 71: 428.
1 Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 338S–400S. 2 Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160: 3415– 3420.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 10, 2007
22
14:33
CHAPTER 22
Deep venous thrombosis and pulmonary embolism in hospitalized patients with sickle cell disease
Nearly every component of hemostasis, including platelet function and the procoagulant, anticoagulant, and fibrinolytic states is altered in sickle cell disease (SCD) [1]. This has led to the view that SCD is a hypercoagulable state [1, 2]. Interactions among sickle erythrocytes, other blood cells, and the various matrix proteins that line the endothelium contribute to the multifactorial vaso-occlusive process [1]. Prior to our evaluation of the prevalence of deep venous thrombosis (DVT) in patients with SCD [3], DVT had been described only in case reports [4, 5]. Deep venous thrombosis is not included among the usual complications of SCD [6]. In past years, there may have been a reluctance to fully evaluate patients with SCD who had suspected DVT, because highosmolarity contrast material used for venograms had a potential for inducing sickling [7]. With the ready availability of venous ultrasound after 1991 [8], a definitive noninvasive diagnosis of DVT became possible in these patients. Among 1,804,000 patients hospitalized with SCD from 1979 to 2003, a discharge diagnosis of DVT was made in 11,000 (0.61%) [3]. From 1991 to 2003, when venous ultrasound was in general use, a somewhat higher percentage of patients hospitalized with SCD had DVT, 7800 of 1,107,000 (0.70%). Acute pulmonary embolism (PE) in patients with SCD is a more contentious issue than acute DVT. As would be expected with a hypercoagulable state, pulmonary thromboembolism occurs in SCD [9–11]. Its frequency, however, is undetermined, largely because of difficulties in distinguishing it from thrombosis in situ [10–13]. Pulmonary embolism has been documented in case reports of patients with SCD [14, 15],
and in 25% [16] and 60% [11] of autopsied patients with SCD. The distinction between PE and thrombosis in situ is particularly important in patients with SCD because both perhaps could be the cause of the acute chest syndrome (pulmonary infiltrates, fever, chest pain, and respiratory symptoms) in these patients. The acute chest syndrome is a leading cause of death in patients with SCD [17–19]. Even though pulmonary infarction occurs in 16% of patients with the acute chest syndrome [20], PE generally is not considered a cause [7]. The likely cause of pulmonary infarction in patients with SCD has been suggested to be thrombosis in situ [10, 12, 21]. Others suggested that engorge ment of the small pulmonary vessels, perhaps due to vasospasm and sickling, can produce pulmonary infarction in the absence of thrombosis of these vessels [22–24]. Whether acute PE is an initiating cause of the acute chest syndrome or whether recurrent PE leads to pulmonary hypertension in SCD has been an unsolved issue [9]. Imaging tests for PE were not obtained in a large study of causes of the acute chest syndrome in SCD [20]. Generally considered causes of the acute chest syndrome include infection, fat embolism, bone marrow embolism, fluid overload, hypoxemia, and vascular obstruction due to sickling and endothelial adherence of erythrocytes [7, 25]. Treatment of the acute chest syndrome includes antibiotics, oxygen, fluid management, respiratory therapy, pain management, bronchodilator therapy, and exchange transfusion [20]. Anticoagulant therapy is not included [20] and its role in the treatment of the acute chest syndrome is uncertain [7] or unconvincing [1, 26, 27]. Low doses of oral anticoagulants (average INR = 1.6) have been speculated to be a relevant form of therapy
109
BLUK077-Stein
March 10, 2007
14:33
PART I
because low doses can normalize the hypercoaguable state in patients with SCD [13]. This presumes that PE is not an etiologic factor. However, once pulmonary hypertension occurs following the acute chest syndrome, PE contributes substantially to further morbidity and mortality [12]. Over the 25-year period from 1979 to 2003, a discharge diagnosis of PE was made in 9000 of 1,804,000 patients hospitalized with SCD. (0.50%) [3]. During this same period, among hospitalized African Americans who did not have SCD, 322,000 of 97,463,000 (0.33%) had a discharge diagnosis of PE. Approximately, 7000 of 9000 (78%) patients with SCD who had a discharge diagnosis of PE were <40 years of age. In patients <40 years old, 7000 of 1,581,000 (0.44%) with SCD had a discharge diagnosis of PE, whereas only 59,000 of 48,611,000 (0.12%) of African Americans without SCD had a discharge diagnosis of PE (Figure 22.1). The prevalence of DVT was similar in patients <40 years old with SCD, 7000 of 1,581,000 (0.44%) and in African Americans who did not have SCD, 193,000 of 48,611,000 (0.40%). The relatively high prevalence of apparent PE in patients with SCD, and the comparable prevalences of DVT in both groups are consistent with the possibility of thrombosis in situ. Thrombosis in situ, however, may not be the exclusive cause of pulmonary vascular occlusion in patients with SCD. If the prevalence of PE in SCD patients aged <40 years were the same as in non-SCD African Americans the same age, then 2000 of 1,581,000 (0.13%) of patients with SCD would be
expected to have PE. This accounts for 29% of the 7000 patients with SCD who had a discharge diagnosis of PE. Thrombosis in situ is limited to muscular pulmonary artery branches (<1 mm diameter) and arterioles [12] and cannot be identified by ordinary pulmonary angiography or CT pulmonary angiography. The smallest PE that have been identified in living patients were in 1–2-mm-diameter pulmonary artery branches [28]. These were shown with wedge pulmonary arteriography [28]. Based on pulmonary angiography, PE occurs in main, lobar, or segmental pulmonary arteries in 96% of patients [29]. In view of the fact that thrombosis in situ occurs only in muscular pulmonary arteries or arterioles, the demonstration of thrombi in elastic vessels (>1 mm diameter) is diagnostic of PE. To diagnose PE, an imaging study is required to show an intraluminal filling defect in an elastic artery (>1 mm). Standard pulmonary angiography and CT pulmonary angiography are the commonly used methods for imaging intraluminal filling defects. However, the risk of injection of radio-opaque contrast material requires assessment in view of evidence that ionic contrast material may induce sickling [30]. Whether nonionic contrast material carries the same risk is uncertain. The package insert of nonionic contrast material cautions against its use in patients with SCD [31]. Magnetic resonance angiography has been used in a few patients with PE and shows promise [32–34]. The package insert for gadopentetate dimeglumine, a magnetic resonance contrast agent, warns that deoxygenated sickle erythrocytes align perpendicular to a magnetic field in vitro [35]. This may result in vasoocclusive complications in vivo. Gadopentetate dimeglumine has not been studied in patients with SCD. A ventilation–perfusion lung scan would not contribute to the differentiation between thrombosis in situ and pulmonary thromboembolism. In patients with the acute chest syndrome, if DVT were present it would be reasonable to assume that venous thromboembolism occurred. Clinically apparent DVT, however, is present in only 11% of patients with acute PE [36]. A venous compression ultrasound would be useful if it were positive, but it lacks sensitivity for DVT in patients with suspected PE who do not have clinical evidence of DVT, being positive in only 29% [37]. Pulmonary thromboembolism is not rare in patients with SCD [3]. This raises the issue of whether
Hospitalized patients
110
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
0.44
0.44 0.40
SCD
0.12 Af-Amer No SCD
PE
SCD AfAmer No SCD
DVT
Figure 22.1 Hospitalized patients younger than 40 years of age with sickle cell disease (SCD) who had a discharge diagnosis of pulmonary embolism (PE) or deep venous thrombosis (DVT) (light bars) compared with African Americans (Af-Amer) the same age who did not have SCD (dark bars). (Reprinted from Stein et al. [3], with permission from Elsevier.)
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 10, 2007
CHAPTER 22
14:33
Patients with sickle cell disease
PE might be an etiologic factor in patients with SCD who have the acute chest syndrome or pulmonary hypertension. In such patients, the risks of an imaging procedure need to be balanced with the advantage of proper diagnosis and treatment.
References 1 Ataga KI, Orringer EP. Hypercoagulability in sickle cell disease: a curious paradox. Am J Med 2003; 115: 721–728. 2 Francis RB, Jr. Platelets, coagulation, and fibrinolysis in sickle cell disease: their possible role in vascular occlusion. Blood Coagul Fibrinolysis 1991 2: 341–353. 3 Stein PD, Beemath A, Meyers FA, Skaf E, Olson RE. Deep venous thrombosis and pulmonary embolism in hospitalized patients with sickle cell disease. Am J Med 2006; 119: 897 e7–11. 4 Brion L, Dupont M, Fondu P, Rutsaert J. Sickle cell anemia and venous thrombosis. Acta Paediatr Belg 1978; 31: 241– 244. 5 Koren A, Zalman L, Levin C et al. Venous thromboembolism, Factor V Leiden, and methylenetetrahydrofolate reductase in a sickle cell anemia patient. Pediatr Hematol Oncol 1999; 16: 469–472. 6 Wang WC, Lukens JN. Sickle cell anemia and other sickling syndromes, In: Lee GR, Paraskevas S, Foerster J, Greer JP et al., eds. Wintrobe’s Clinical Hematology, 10th edn. Williams & Wilkins, Baltimore, 1999: 1346–1357. 7 Kirkpalrick MB, Haynes J. Sickle cell disease and the pulmonary circulation. Semin Respir Crit Care Med 1994; 15: 473–481. 8 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diagnosing deep vein thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 9 Serjeant GR. Sickle Cell Disease, 2nd edn. Oxford University Press, Oxford, UK, 1992: 184–187. 10 Moser KM., Shea JG. The relationship between pulmonary infarction, cor pulmonale and the sickle states. Am J Med 1957; 22: 561–579. 11 Oppenheimer EH, Esterly JR. Pulmonary changes in sickle cell disease. Am Rev Respir Dis 1971; 103: 858–859. 12 Adedeji MO, Cespedes J, Allen K, Subramony C, Hughson MD. Pulmonary thrombotic arteriopathy in patients with sickle cell disease. Arch Pathol Lab Med 2001; 125: 1436– 1441. 13 Walker B K, Ballas SK, Burka ER. The diagnosis of pulmonary thromboembolism in sickle cell disease. Am J Hematol 1979; 7: 219–232. 14 Maggi JC, Nussbaum E. Massive pulmonary infarction in sickle cell anemia. Pediatr Emerg Care 1987; 3: 30–32.
111
15 Rubler S, Fleischer RA. Sickle cell states and cardiomyopathy. Sudden death due to pulmonary thrombosis and infarction. Am J Cardiol 1967; 19: 867–873. 16 Haupt HM, Moore GW, Bauer TW, Hutchins GM. The lung in sickle cell disease. Chest 1982; 81: 332–337. 17 Vichinsky E. Comprehensive care in sickle cell disease: its impact on morbidity and mortality. Semin Hematol 1991; 28: 220–226. 18 Castro O, Brambilla DJ, Thorington B et al. The acute chest syndrome in sickle cell disease: incidence and risk factors: the Cooperative Study of Sickle Cell Disease. Blood 1994; 84: 643–649. 19 Platt OS, Brambilla DJ, Rosse WF et al. Mortality in sickle cell disease: life expectancy and risk factors for early death. N Engl J Med 1994; 330: 1639–1644. 20 Vichinsky EP, Neumayr LD, Earles AN et al., for National Acute Chest Syndrome Study Group. Causes and outcomes of the acute chest syndrome in sickle cell disease. N Engl J Med 2000; 342: 1855–1865. 21 Durant JR, Cortes FM. Occlusive pulmonary vascular disease associated with hemoglobin SC disease. Am Heart J 1966; 71: 100–106. 22 Athanasou NA, Hatton C, McGee JO, Weatherall DJ. Vascular occlusion and infarction in sickle cell crisis and the sickle chest syndrome. J Clin Pathol 1985; 38: 659– 664. 23 Francis RB, Johnson CS. Vascular occlusive in sickle cell disease: current concepts and unanswered questions. Blood 1991; 77: 1405–1414. 24 Gladwin MT, Rodgers GP. Pathogenesis and treatment of acute chest syndrome of sickle-cell anaemia. Lancet 2000; 355: 1476–1478. 25 Minter KR, Gladwin MT. Pulmonary complications of sickle cell anemia. A need for increased recognition, treatment, and research. Am J Respir Crit Care Med 2001; 164: 2016–2019. 26 Collins FS, Orringer EP. Pulmonary hypertension and cor pulmonale in the sickle hemoglobinopathies. Am J Med 1982; 73: 814–821. 27 Barrett-Connor E. Pneumonia and pulmonary infarction in sickle cell anemia. JAMA 1973; 224: 997–1000. 28 Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623. 29 Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/perfusion lung scans. Chest 1997; 111: 1246– 1248. 30 Richards D, Nulsen FE. Angiographic media and the sickling phenomenon. Surg Forum 1971; 22: 403–404. 31 Optiray (Ioversol Injection). Mallinckrodt Inc., St. Louis, Missouri, September 2000.
BLUK077-Stein
March 10, 2007
14:33
112
PART I
32 Meaney JFM, Weg JG, Chenevert TL et al. Diagnosis of pulmonary embolism with magnetic resonance angiography. N Engl J Med 1997; 336: 1422–1427. 33 Oudkerk M, van Beek EJ, Wielopolski P et al. Comparison of contrast-enhanced magnetic resonance angiography and conventional pulmonary angiography for the diagnosis of pulmonary embolism: a prospective study. Lancet 2002; 359: 1643–1647. 34 Gupta A, Frazer CK, Ferguson JM et al. Acute pulmonary embolism: diagnosis with MR angiography. Radiology 1999; 210: 353–359.
35 Magnevist Injection (Brand of Gadopentetate Dimeglumine). Berlix Laboratories, Wayne, New Jersey, May 2000. 36 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 37 Turkstra F, Kuijer PM, van Beek EJ, Brandjes DP, ten Cate JW, Buller HR. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
April 6, 2007
23
16:40
CHAPTER 23
Venous thromboembolism in pregnancy
Introduction
Trends in DVT according to age
Pulmonary venous thromboembolism in developed countries is one of the leading causes of maternal death [1–6]. Limited survey data from the United States, Denmark, and the United Kingdom have yielded conflicting findings as to whether venous thromboembolism associated with pregnancy is increasing or declining [3, 7–10]. We used the database of the National Hospital Discharge Survey, available at: http:// www.cdc.gov/nchs/about/major/hdasd/nhds.htm, to assess trends in venous thromboembolism during pregnancy [11].
The rate of pregnancy-associated DVT was higher among women aged 35–44 years than in younger women (Figure 23.2). On average, 91% of maternal patients were 15– 34 years of age (Figure 23.3). The percentage of deliveries in women 35–44 years of age increased over 21 years from 4.5 to 13.1% and the percentage of deliveries in younger women decreased proportionately (Figure 23.3).
Pregnancy-associated DVT according to race
Twenty-one-year trends in rate of diagnosis of deep venous thrombosis
Figure 23.1 Triennial rates of deep venous thrombosis (DVT) in women aged 15–44 years. (Reprinted from Stein et al. [11], with permission from Elsevier.)
Deep venous thrombosis/ 100,000/yr
Pregnancy-associated deep venous thrombosis (DVT) was diagnosed in 93,000 of 80,798,000 women (0.12%) from 1979 to 1999 [11]. The rate of pregnancyassociated DVT (vaginal delivery and cesarean section) increased from 1982–1984 to 1997–1999 (Figure 23.1). The rate of nonpregnancy-associated DVT decreased from 1979–1981 to 1991–1993. Thereafter, the rate of nonpregnancy-associated DVT remained constant.
The rate of pregnancy-associated DVT among black women was higher than among white women during 1979–1988 and 1989–1999 (Figure 23.4). The rate of pregnancy-associated DVT in both black women and white women increased from 1979–1988 to 1989– 1999. Higher rates of pregnancy-associated DVT in black women compared to white women have previously been observed [6, 12]. Higher rates of fatal pulmonary embolism (PE) have also been reported among black maternity patients [9].
200 Pregnancy-associated 150 100 Nonpregnancy-associated
50
0 1979−1981 1982−1984 1985−1987 1988−1990 1991−1993 1994−1996 1997−1999 Years
113
BLUK077-Stein
April 6, 2007
16:40
114
Deep venous thrombosis/ 100,000/yr
PART I
Prevalence, risks, and prognosis of PE and DVT
350 300 250 200
35−44
150 100 50
25−34 15−24
Figure 23.2 Rates of pregnancy-associated deep venous thrombosis (DVT) according to decades of age, 1979–1988 to 1989–1999. (Data from Stein et al. [11].)
0 1979−1988
1989−1999 Years
The percentage of deliveries in black and white women remained constant over the 21-year period of observation (Figure 23.5). The rate of nonpregnancy-associated DVT was higher in black women than in white women (Figure 23.6). The rates of nonpregnancy-associated DVT decreased from 1979–1988 to 1989–1999 in both black and white women (Figure 23.6). The observation of a higher rate of DVT in black women in the nonmaternal population is compatible with our previous observation of an increased rate of DVT in age-matched blacks compared to whites [13, 14].
Pregnancy-associated DVT according to mode of delivery The rate of diagnosis of DVT following cesarean section and following vaginal delivery increased from 1979–1988 to 1989–1999 (Figure 23.7).
All deliveries (%)
100
The percentage of all cesarean section deliveries among women aged 15–34 years increased from 15.9% in 1979 to 22.6% in 1987. Thereafter, the rate decreased to 17.0% in 1999 (Figure 23.8). Among women aged 35–44 years, the percentage of all cesarean section deliveries increased linearly from 1.1% in 1979 to 3.9% in 1999 (Figure 23.8). Among all women who had cesarean sections, the proportion performed in women aged 35–44 years increased from 6.2% in 1979 to 18.7% in 1999 (Figure 23.9). Concordantly, the proportion of cesarean sections in women aged 15–34 years decreased.
Twenty-one-year trends in the rate of diagnosis of PE The rate of pregnancy-associated PE was lower than the rate of nonpregnancy-associated PE (Figure 23.10). However, the rate of pregnancy-associated DVT was higher than the rate of nonpregnancy-associated DVT.
15−34
50
35−44 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Years
Figure 23.3 Proportion of deliveries according to age. Over the 21-year period of observation, deliveries in women 35–44 years of age increased. (Data from Stein et al. [11].)
April 6, 2007
CHAPTER 23
16:40
115
Venous thromboembolism in pregnancy
Figure 23.4 Rates of deep venous thrombosis (DVT) between 1979–1988 and 1989–1999 among black and white pregnancy-associated women. (Data from Stein et al. [11].)
Deep venous thrombosis/ 100,000/yr
BLUK077-Stein
200 144
150
118 100
100 61 50 0 White
Black
1979−1988
White
Black
1989−1999
All deliveries (%)
100 White
50
Figure 23.5 The proportion of deliveries in black and white women. The proportions were constant over the 21-year period of observation. (Data from Stein et al. [11].)
Figure 23.7 Rates of DVT following cesarean section and vaginal delivery during 1979–1988 and 1989–1999. (Data from Stein et al. [11].)
0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999
Deep venous thrombosis/ 100,000/yr
Years
200 150 100
70
55 50
57 30
0
White
Black
1979−1988 Deep venous thrombosis/ 100,000/yr
Figure 23.6 Rates of nonpregnancy-associated deep venous thrombosis (DVT) in black and white women. (Data from Stein et al. [11].)
Black
White
Black
1989−1999
120
104
80
63 47
40
29
0 Vaginal delivery
C-section
1979−1988
Vaginal delivery
C-section
1989−1999
BLUK077-Stein
April 6, 2007
16:40
Cesarean section (%)
116
PART I
Prevalence, risks, and prognosis of PE and DVT
30 15 −34
20
10 35 −44 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Years
The ratio of pregnancy-associated DVT to pregnancyassociated PE, 12.8, was higher than the ratio of nonpregnancy-associated DVT to nonpregnancyassociated PE, 2.9 (Figure 23.10). The reason for this difference is unknown and could reflect a difference of the natural history of DVT in pregnancy. It also could reflect a reluctance to expose pregnant women to ionizing radiation associated with imaging for PE resulting in a decreased frequency of diagnosis of PE. Our findings of an increasing rate of pregnancyassociated DVT in the United States [11] is in harmony with that observed in Denmark [7] where a similar upward trend was noted. The upward trend is concordant with reported findings in the United Kingdom for venous thromboembolism [3]. However, some reported a relatively constant rate of antepartum DVT and PE and a decreasing rate of postpartum PE [10]. Our observations support the impression held for many years that pregnancy predisposes to thromboembolism. The rate of pregnancy-associated DVT was twice the rate of nonpregnancy-associated DVT in women the same age [11]. A sixfold increase in the rate of thromboembolism during pregnancy and the
Figure 23.8 Cesarean sections, shown as a percentage of all deliveries according to age groups from 1979 to 1999. (Reprinted from Stein et al. [11], with permission from Elsevier.)
puerperium in comparison to nonpregnant women has been reported by others [15]. The reason for the diverging trends for the rates of diagnosis of pregnancy-associated DVT and nonpregnancy-associated DVT is not apparent. Analysis of covariance showed that age, race, and the percentage of deliveries by cesarean section did not explain the increasing rates of pregnancy-associated DVT over time. Although the percentage of deliveries in women aged 35–44 years increased over the 21-year period of observation, the increased risk of pregnancyassociated DVT was not limited to this age group. In fact, the greatest increase in rates of pregnancyassociated DVT was in younger women. Even though the rate of pregnancy-associated DVT was higher in black women than in white women, the percentage of deliveries in black women remained constant over the 21-year period of observation. Therefore, possible changes of racial characteristics of the maternal population would not explain the increasing rate of pregnancy-associated DVT. The proportion of deliveries by cesarean section decreased during most of the years that the rate of pregnancy-associated DVT was increasing (from 1987 to 1999). Therefore, the
Cesarean section (%)
100
15 −34
50
35 −44 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999
Years
Figure 23.9 Percentage of cesarean sections according to age from 1979 to 1999. (Data from Stein et al. [11].)
BLUK077-Stein
April 6, 2007
CHAPTER 23
16:40
117
Venous thromboembolism in pregnancy
150
Rate per 100,000/yr
11 5 100
49
50
17
9 Figure 23.10 Pregnancy-associated and nonpregnancy-associated deep venous thrombosis (DVT) and pulmonary embolism (PE) from 1979 to 1999. (Data From Stein et al. [11].)
0
percentage of deliveries by cesarean section could not explain the increasing rate of pregnancy-associated DVT. Increasing pregnancy-associated comorbidity due to acute and chronic medical conditions and comorbidity, which previously precluded conception may be co-factors, but this could not be assessed. Regarding nonpregnancy-associated DVT, a decreasing rate from 1973–1975 to 1988–1990 in the general population of women (pregnancy-associated and nonpregnancy-associated) was reported in women aged 15–44 years [16]. We observed the same trend during this time period in women of all ages, but from 1991 to 1999 the rate increased [14]. Our observations support the need for continued vigilance in the prevention of pregnancy-associated DVT. Further understanding is needed of the factors that contribute to this trend for an increasing rate of a potentially lethal condition in young healthy women.
References 1 Atrash HK, Koonin LM, Lawson HW, Franks AL, Smith JC. Maternal mortality in the United States, 1979–1986. Obstet Gynecol 1990; 76: 1055–1060. 2 Koonin LM, MacKay AP, Berg CJ, Atrash HK, Smith JC. Pregnancy-related mortality surveillance-United States, 1987–1990. MMWR 1997; 46: 17–36. 3 Department of Health, Welsh Office, Scottish Office Department of Health, Department of Health and Social Services, Northern Ireland. Why Mothers Die. Report
DVT
PE
Pregnancyassociated
4
5
6
7
8
9
10
11
12
DVT
PE
Nonpregnancyassociated
on Confidential Enquiries into Maternal Deaths in the United Kingdom, 1994–1996. 1998; Chapt 2. Hogberg U, Innala E, Sandstrom A. Maternal mortality in Sweden, 1980–1988. Obstet Gynecol 1994; 84: 240– 244. Berg CJ, Atrash HK, Koonin LM, Tucker M. Pregnancyrelated mortality in the United States, 1987–1990. Obstet Gynecol 1996; 88: 161–167. Rochat RW, Koonin LM, Atrash HK, Jewett JF. Maternal mortality in the United States: report from the Maternal Mortality Collaborative. Obstet Gynecol 1988; 72: 91– 97. Andersen BS, Steffensen FH, Sorensen HT, Nielsen GL, Olsen J. The cumulative incidence of venous thromboembolism during pregnancy and puerperium—an 11 year Danish population-based study of 63,000 pregnancies. Acta Obstet Gynecol Scand 1998; 77: 170–173. Macklon NS, Greer IA. Venous thromboembolic disease in obstetrics and gynecology: the Scottish experience. Scot Med J 1996; 41: 83–86. Franks AL, Atrash HK, Lawson HW, Colberg KS. Obstetrical pulmonary embolism mortality, United States, 1970–85. Am J Pub Health 1990; 80: 720–722. Heit JA, Kobbervig CE, James AH, Petterson TM, Bailey KR, Melton LJ, 3rd. Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study. Ann Intern Med 2005; 143: 697–706. Stein PD, Hull RD, Kayali F et al. Venous thromboembolism in pregnancy: 21 year trends. Am J Med 2004; 117: 121–125. Buehler JW, Kaunitz AM, Hogue CJR, Hughes JM, Smith JC, Rochat RW. Maternal mortality in women aged 35 years or older: United States. JAMA 1986; 255: 53–57.
BLUK077-Stein
April 6, 2007
16:40
118
PART I
13 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in blacks and whites. Arch Intern Med 2003; 163: 1843–1848. 14 Stein PD, Hull RD, Patel KC, Olson RE, Ghali WA, Meyers FA. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694.
15 Anonymous. Oral contraception and thrombo-embolic disease. J R Coll Gen Pract 1967; 13: 267–279. 16 Silverstein MD, Heit JA, Mohr DN, Petterson TM, O’Fallon WM, Melton, LJ, III. Trends in the incidence of deep vein thrombosis and pulmonary embolism. A 25-year population-based study. Arch Intern Med 1998; 158: 585–593.
Prevalence, risks, and prognosis of PE and DVT
BLUK077-Stein
March 12, 2007
24
15:45
CHAPTER 24
Air travel as a risk for pulmonary embolism and deep venous thrombosis
Introduction Air travel was suggested to be a risk for deep venous thrombosis (DVT) and pulmonary embolism (PE) in 1954, presumably induced by stasis [1]. However, the entire gamut of risk factors for venous thromboembolism has been suggested to participate in the occurrence of venous thromboembolic disease (VTE) following air travel [2]. The possibility of VTE after travel is not unique to air travel, having been reported with various modes of transportation [1, 3, 4]. Prolonged periods in cramped quarters, irrespective of travel, can lead to PE [5]. The term “economy class syndrome” was introduced in 1988 [6], but has since been replaced with “flight-related DVT” in recognition of the fact that all travelers are at risk, irrespective of the class of travel [7]. Systematic review indicates that there is a strong and significant association between prolonged air travel and PE and DVT [8]. Those with preexisting risk factors for VTE were most vulnerable [8, 9]. Flight duration between 3 and 18 hours appears to be a risk for DVT [9–13] and fatal PE [14]. Among patients who died of PE during flight, 10 of 11 died during travel lasting 12–18 hours [14]. Among 6.58 million passengers who arrived at Sydney International Airport following travel of ≥9 hours duration, 17 passengers had acute PE upon arrival (2.6 PE/million travelers) [15] (Figure 24.1). Among passengers who arrived at Madrid-Barajas Airport, 15 of 9.07 million passengers (1.65/million passengers) who traveled ≥8 hours had acute PE on arrival [16]. Only 1 of 3.93 million passengers who traveled 6–8 hours (0.25 per million) had acute PE on arrival and 0 of 28.04 million passengers who traveled ≤6 hours had acute PE on arrival [16] (Figure 24.1).
In a prospective investigation of travelers who traveled ≥10 hours, 4 of 878 (0.5%) developed PE and 5 of 878 (0.6%) developed DVT [17]. In terms of distance traveled, among those who traveled >6200 miles, 4.8 passengers per million passenger arrivals required transfer to a hospital by a French emergency medical team for acute PE [18]. Among those who traveled 4650–6199 miles, 2.7 passengers per million passenger arrivals had acute PE. The proportion with acute PE decreased to 0.4 passengers per million passenger arrivals with 3100–4649 miles traveled, 0.1 acute PE passengers per million passenger arrivals with 1550–3099 miles traveled, and no acute PE with <1549 miles traveled [18]. Recent travel has shown a positive association with DVT and PE in some [19–21], but not all investigations [22, 23]. Among inpatients, 41 of 168 (24.4%) with DVT or PE reported recent travel >4 hours by plane compared with 12 of 160 (7.5%) with other illnesses [19]. Other patients with DVT reported long distance travel in 62 of 494 (12.6%), whereas only 31 of 494 (6.3%) patients who did not have DVT reported long distance travel [20]. Still others showed that 20 of 185 patients with DVT (10.8%) traveled >3 hours compared with 31 of 383 (8.1%) who did not have DVT [21]. Regarding studies that showed no association of VTE with travel, 9 of 186 (5%) patients with DVT traveled ≥3 hours by train, car, or boat within the past 4 weeks, whereas even a somewhat higher proportion with no DVT, 42 of 602 (7%) also traveled >3 hours [22]. Similarly, 14 of 198 (7.1%) patients with DVT traveled >3 hours during the past 4 weeks, whereas the same proportion, 44 of 615 (7.2%) with no DVT also traveled >3 hours during the past 4 weeks [23]. The positive reports suggest a contribution of travel to the risk of PE and DVT, but the apparently negative
119
BLUK077-Stein
March 12, 2007
15:45
PE/million passenger arrivals
120
PART I
Prevalence, risks, and prognosis of PE and DVT
3 2.6 2.5 2 1.7 1.5 1 0.5 0
0.3 0 <6
≥8 6−8 Flight duration (hours)
studies failed to exclude a moderate effect because controls may have been identified, in part, by a history of recent travel [24].
Trials of prophylactic agents Elastic stockings appear to be effective in preventing flight-related DVT. Pooled data of randomized trials showed asymptomatic DVT, identified by screening with ultrasound, in 1 of 868 passengers (0.1%) who received elastic stockings and 38 of 883 passengers (4.3%) who received no prophylaxis [25–29]. Low-molecular-weight heparin, in small numbers of patients, also appeared to prevent DVT. No DVT was found in 84 passengers who received enoxaparin 1 mg/kg 2–4 hours before travel, whereas 4 of 82 (4.9%) untreated passengers developed DVT [30]. Aspirin, in small numbers of patients, was ineffective in preventing flight-related DVT. Among patients who received aspirin 400 mg/day for 3 days, starting 12 hours before travel, DVT (usually asymptomatic) developed in 3 of 84 (3.6%) versus 4 of 82 (4.9%) in controls [30].
Recommendations Recommendations of the American College of Chest Physicians for travel ≥6 hours are [31]. General measures (Clear risk/benefit, strong recommendation):
1 Avoid constrictive clothing around the lower extremities or waist. 2 Avoid dehydration.
≥9
Figure 24.1 Prevalence of pulmonary embolism (PE) per million passenger arrivals according to duration of flight. (Data from Hertzberg et al. [15] and Perez-Rodriguez et al. [16].)
3 Stretch calf muscles frequently. Specific measures:
1 Below knee graduated compression stockings providing 15–30 mm Hg at the ankle (Unclear risk/benefit, weak recommendation), or 2 Single prophylactic dose of low-molecularweight heparin injected prior to departure (Unclear risk/benefit, weak recommendation), 3 It is recommended that aspirin not be used (Clear risk/benefit, strong recommendation).
References 1 Homans J. Thrombosis of the deep leg veins due to prolonged sitting. N Engl J Med 1954; 250: 148–149. 2 Giangrande PL. Air travel and thrombosis. Br J Haematol 2002; 117: 509–512. 3 Symington IS, Stack BH. Pulmonary thromboembolism after travel. Br J Dis Chest 1977; 71: 138–140. 4 Tardy B, Page Y, Zeni F et al. Phlebitis following travel. Presse Med 1993; 22: 811–814. 5 Simpson K. Shelter deaths from pulmonary embolism. Lancet 1940; 2: 744. 6 Cruickshank JM, Gorlin R, Jennett B. Air travel and thrombotic episodes: the economy class syndrome. Lancet 1988; 2: 497–498. 7 Collins J. Thromboembolic disease related to air travel: what you need to know. Semin Roentgen 2005; 40: 1–2. 8 Ansari MT, Cheung BM, Qing Huang J, Eklof B, Karlberg JP. Traveler’s thrombosis: a systematic review. J Travel Med 2005; 12: 142–154. 9 Arfvidsson B. Risk factors for venous thromboembolism following prolonged air travel: a “prospective” study. Cardiovasc Surg 2001; 9: 158–159.
BLUK077-Stein
March 12, 2007
CHAPTER 24
15:45
121
Air travel as a risk for PE and DVT
10 Paganin F, Laurent Y, Gauzere BA, Blanc P, Roblin X. Pulmonary embolism on non-stop flights between France and Reunion Island. Lancet 1996; 347: 1195–1196. 11 Ribier G, Zizka V, Cysique J, Donatien Y, Glaudon G, Ramialison C. Venous thromboembolic complications following air travel. Retrospective study of 40 cases recorded in Martinique. Rev Med Interne 1997; 18: 601– 604. 12 Eklof B, Kistner RL, Masuda EM, Sonntag BV, Wong HP. Venous thromboembolism in association with prolonged air travel. Dermatol Surg 1996; 22: 637–641. 13 Mercer A, Brown JD. Venous thromboembolism associated with air travel: a report of 33 patients. Aviat Space Environ Med 1998; 69: 154–157. 14 Sarvesvaran R. Sudden natural deaths associated with commercial air travel. Med Sci Law 1986; 26: 35–38. 15 Hertzberg SR, Roy S, Hollis G, Brieger D, Chan A, Walsh W. Acute symptomatic pulmonary embolism associated with long haul air travel to Sydney. Vasc Med 2003; 8: 21–23. 16 Perez-Rodriguez E, Jimenez D, Diaz G et al. Incidence of air travel-related pulmonary embolism at the Madrid-Barajas airport. Arch Intern Med 2003; 163: 2766– 2770. 17 Hughes RJ, Hopkins RJ, Hill S et al. Frequency of venous thromboembolism in low to moderate risk long distance air travellers: the New Zealand Air Traveller’s Thrombosis (NZATT) study. Lancet 2003; 362: 2039–2044. 18 Lapostolle F, Surget V, Borron SW et al. Severe pulmonary embolism associated with air travel. N Engl J Med 2001; 345: 779–783. 19 Ferrari E, Chevallier T, Chapelier A, Baudouy M. Travel as a risk factor for venous thromboembolic disease: a casecontrol study. Chest 1999; 115: 440–444. 20 Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160: 3415–3420. 21 Arya R, Barnes JA, Hossain U, Patel RK, Cohen AT. Longhaul flights and deep vein thrombosis: a significant risk
22
23
24 25
26
27
28
29
30
31
only when additional factors are also present. Br J Haematol 2002; 116: 653–654. Kraaijenhagen RA, Haverkamp D, Koopman MM, Prandoni P, Piovella F, Buller HR. Travel and risk of venous thrombosis. Lancet 2000; 356: 1492–1493. Ten Wolde M, Kraaijenhagen RA, Schiereck J et al. Travel and the risk of symptomatic venous thromboembolism. Thromb Haemost 2003; 89: 499–505. Gallus AS, Goghlan DC. Travel and venous thrombosis. Curr Opin Pulm Med 2002; 8: 372–388. Belcaro G, Geroulakos G, Nicolaides AN, Myers KA, Winford M. Venous thromboembolism from air travel: the LONFLIT study. Angiology 2001; 52: 369–374. Scurr JH, Machin SJ, Bailey-King S, Mackie IJ, McDonald S, Smith PD. Frequency and prevention of symptomless deep-vein thrombosis in long-haul flights: a randomised trial. Lancet 2001; 357: 1485–1489. Belcaro G, Cesarone MR, Shah SS et al. Prevention of edema, flight microangiopathy and venous thrombosis in long flights with elastic stockings. A randomized trial: The LONFLIT 4 Concorde Edema-SSL Study. Angiology 2002; 53: 635–645. Cesarone MR, Belcaro G, Nicolaides AN et al. The LONFLIT4-Concorde–Sigvaris Traveno stockings in long flights (EcoTraS) study: a randomized trial. Angiology 2003; 54: 1–9. Cesarone MR, Belcaro G, Errichi BM et al. The LONFLIT4–Concorde deep venous thrombosis and edema study: prevention with travel stockings. Angiology 2003; 54: 143–154. Cesarone MR, Belcaro G, Nicolaides AN et al. Venous thrombosis from air travel: the LONFLIT3 study– prevention with aspirin vs low-molecular-weight heparin (LMWH) in high-risk subjects: a randomized trial. Angiology 2002; 53: 1–6. Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest 2004; 126: 338S–400S.
BLUK077-Stein
April 14, 2007
25
14:57
CHAPTER 25
Estrogen-containing oral contraceptives and venous thromboembolism
Risks of venous thromboembolic disease in users of estrogen-containing oral contraceptives compared with nonusers Most preparations of estrogen-containing oral contraceptives consist of a mixture of estrogen and progesterone [1]. The US Food and Drug Administration in 1980 recommended the use of the lowest possible dose of estrogen for birth control [2]. In the past it had been thought that different types of progestins did not affect the risk of deep venous thrombosis (DVT) [3], but more recent data indicate that some third generation progestins have double the risk of second generation progestins [4]. In most countries, medications containing 50 μg of ethinyl estradiol or its equivalent are no longer available except as a shortterm postcoital medication [1]. Modern-day oral contraceptives contain 20–35 μg of ethinyl estradiol [5, 6]. Although the relative risk of VTE is higher among users of oral estrogen-containing contraceptives than nonusers [5, 7], the absolute risk is low [8]. Review by Vandenbroucke et al. [8] showed relative risks for VTE that, in general, ranged from 2.5 to 6.1 [9–12] although one study showed a relative risk of 11 [13]. A review by Lewis [5] of second and third generation estrogen-containing oral contraceptives, showed odds ratios that ranged from 0.8 to 2.3 [5, 9, 14–19]. Importantly, an absolute baseline risk <1/10,000 patients/ year increased to only 3 to 4/10,000 patients/year during the time oral contraceptives are being used [8]. Among 234,218 users of estrogen-containing oral contraceptives from 1980 to 1986, at doses that ranged from <50 μg/day to >50 μg/day, inpatient diagnoses of PE (pulmonary embolism) or DVT were
122
made in only 142 women (5.8/10,000 oral contraceptive users/year) [20]. Others, in women taking <50 μg/day of estrogen-containing contraceptives, reported VTE in 4.2/10,000 users/year [20], 4.7/10,000 [6] and 2.5/10,000 users/year [21]. Pooled data showed VTE in 127/417,915 patient-years (3.0/10,000 contraceptive users/year) [6, 20, 21]. Users of a patch designed to deliver 20 μg of ethinyl estradiol/day and 150 μg of norelgestromin/day over a period of 1 week [22] showed VTE in 4.1/10,000 users/year [21] and 5.3/10,000 users/year [6]. With earlier generation estrogen-containing oral contraceptives evaluated in 1961–1969, the incidence of DVT among users was higher than reported in recent years, 9 of 4965 (18/10,000), but higher rates were also reported among nonusers, 8 of 4933 (16/10,000) [23].
Relative risk in relation to dose The relative risk for VTE in women using oral contraceptives containing 50 μg of estrogen, compared with users of oral contraceptives that contained <50 μg was 1.5 [20]. The relative risk for VTE in women using oral contraceptives containing >50 μg of estrogen, compared with users of oral contraceptives that contained <50 μg was 1.7 [20]. No difference in the risk of VTE was found with various levels of low doses of 20, 30, 40, and 50 μg/day [3]. With doses of estrogen of 50 μg/day, the rate of VTE was 7.0/10,000 contraceptive users/year and with >50 μg/day, the rate of VTE was 10.0/10,000 oral contraceptive users/year [20] (Figure 25.1). Some, however, found no appreciable difference in the relative risk of VTE in relation to low or higher estrogen doses [11].
April 14, 2007
CHAPTER 25
14:57
VTE/10,000 estrogen users/yr
10.0 10 9 8 7 6 5 4 3 2 1 0
123
Estrogen-containing oral contraceptives and VTE
7.0
3.0
Relative risk of VTE (estrogen users vs. nonusers)
BLUK077-Stein
4 3.5 3 2.5 2 1.5 1 0.5 0
3.7
1.9 1.0
<20
1.2
21−25
26−30
>30
Body mass index (kg/m2) <50
50 Estrogen (μg/day)
>50
Figure 25.1 Venous thromboembolism (VTE) per 10,000 estrogen using patients per year according to daily dose. (Data from Gerstman et al. [20], Jick et al. [6] and Drug Safety Final Report [21]. Data with estrogen <50 μg/day were pooled.)
Relative risk according to duration of use Reports of risk of VTE in relation to the duration of use of oral contraceptives are inconsistent. Some showed relative risks increased as the duration of use of estrogen-containing oral contraceptives increased [13]. The relative risks were 0.7 in women who used oral contraceptives <1 year, 1.4 for those who used oral contraceptives 1–4 years and 1.8 in those who used it ≥5 years [13]. Others showed quite the opposite effect with a decreasing relative risk with duration of use [3]. The relative risk for DVT or PE was 5.1 with use <1 year, 2.5 with use 1–5 years, and 2.1 with use >5 years [3]. Finally, some showed the risk to be unaffected by the duration of use [11].
Estrogen-containing oral contraceptives in combination with smoking Users of oral contraceptives in Europe who smoked ≥10 cigarettes/day had odds ratios of 2.59 compared with users of oral contraceptives who did not smoke [11]. Users of oral contraceptives in developing countries who smoked ≥10 cigarettes/day had odds ratios of 1.22 compared with users of oral contraceptives who did not smoke [11]. Users of oral contraceptives who smoked <10 cigarettes/day did not have an increased risk of VTE compared to users of oral contraceptives who did not smoke [11].
Figure 25.2 Relative risk of venous thromboembolism (VTE) comparing estrogen users to nonusers shown in relation to body mass index (BMI). (Data from Lidegaard et al. [3].)
Estrogen-containing oral contraceptives and obesity The odds ratio for VTE comparing users of oral estrogen-containing oral contraceptives with matched nonusers of various body mass indexes (BMIs), showed that the combination of obesity with oral contraceptives carried a higher relative risk [3] (Figure 25.2). The World Health Organization reported higher odds ratios with oral contraceptives among women with BMIs >25 kg/m2 [11].
Estrogen-containing oral contraceptives and postoperative VTE The possibility of an increased risk of postoperative thromboembolism with oral contraceptive use was raised by Vessey et al. in 1970 [24]. The risk of postoperative PE appears to have increased in women who use oral contraceptives, even when the oral contraceptives have a low estrogen content [2]. In PIOPED, if PE was suspected, and if the women who underwent surgery used oral contraceptives, 50% had PE. If PE was suspected, and if the women who underwent surgery did not use oral contraceptives, 12% had PE [2].
References 1 Carter CJ. Epidemiology of venous thromboembolism. In: Hull RD & Pineo GF, eds. Disorders of Thrombosis. W. B. Saunders Co., Philadelphia, PA, 1996: 159–174. 2 Quinn DA, Thompson BT, Terrin ML et al. A prospective investigation of pulmonary embolism in women and men. JAMA 1992; 268: 1689–1696.
BLUK077-Stein
April 14, 2007
14:57
124
3 Lidegaard O, Edstrom B, Kreiner S. Oral contraceptives and venous thromboembolism. A case–control study. Contraception 1998; 57: 291–301. 4 Tanis BC, Rosendaal FR. Venous and arterial thrombosis during oral contraceptive use: risks and risk factors. Semin Cardiovasc Med 2003; 3: 69–83. 5 Lewis MA. The epidemiology of oral contraceptive use: a critical review of the studies on oral contraceptives and the health of young women. Am J Obstet Gynecol 1998; 179: 1086–1097. 6 Jick SS, Kaye JA, Russmann S, Jick H. Risk of nonfatal venous thromboembolism in women using a contraceptive transdermal patch and oral contraceptives containing norgestimate and 35 microg of ethinyl estradiol. Contraception 2006; 73: 223–228. 7 Realini JP, Goldzieher JW. Oral contraceptives and cardiovascular disease: a critique of the epidemiologic studies. Am J Obstet Gynecol 1985; 152: 729–798. 8 Vandenbroucke JP, Rosing J, Bloemenkamp KW et al. Oral contraceptives and the risk of venous thrombosis. N Engl J Med 2001; 344: 1527–1535. 9 Jick H, Jick SS, Gurewich V, Myers MW, Vasilakis C. Risk of idiopathic cardiovascular death and nonfatal venous thromboembolism in women using oral contraceptives with differing progestagen components. Lancet 1995; 346: 1589–1593. 10 Vessey M, Mant D, Smith A, Yeates D. Oral contraceptives and venous thromboembolism: findings in a large prospective study. BMJ (Clin Res Ed) 1986; 292: 526. 11 World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Venous thromboembolic disease and combined oral contraceptives: results of international multicentre case– control study. Lancet 1995; 346: 1575–1582. 12 Lewis MA, Heinemann LA, MacRae KD, Bruppacher R, Spitzer WO, The increased risk of venous thromboembolism and the use of third generation progestagens: role of bias in observational research. Transitional Research Group on Oral Contraceptives and the Health of Young Women. Contraception, 1996; 54: 5–13. [Erratum, Contraception 1996; 54: 121.] 13 Helmrich SP, Rosenberg L, Kaufman DW, Strom B, Shapiro S. Venous thromboembolism in relation to oral contraceptive use. Obstet Gynecol 1987; 69: 91–95. 14 World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contracep-
PART I
15
16
17
18
19
20
21
22
23
24
Prevalence, risks, and prognosis of PE and DVT
tion. Effect of different progestagens in low oestrogen oral contraceptives on venous thromboembolic disease. Lancet 1995; 346: 1582–1588. Spitzer WO, Lewis MA, Heinemann LA, Thorogood M, MacRae KD, for Transnational Research Group on Oral Contraceptives and the Health of Young Women. Third generation oral contraceptives and risk of venous thromboembolic disorders: an international case–control study. BMJ 1996; 312: 83–88. Bloemenkamp KW, Rosendaal FR, Helmerhorst FM, Buller HR, Vandenbroucke JP. Enhancement by factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a thirdgeneration progestagen. Lancet 1995; 346: 1593–1596. Farmer RDT, Preston TD. The risk of venous thromboembolism associated with low oestrogen oral contraceptives. J Obstet Gynaecol 1995; 1: 13–20. Farmer RD, Lawrenson RA, Thompson CR, Kennedy JG, Hambleton IR. Population-based study of risk of venous thromboembolism associated with various oral contraceptives. Lancet 1997; 349: 83–88. Herings RMC, de Boer A, Urquhart J, Leufkens HGM. Non-causal explanations for the increased risk of venous thromboembolism among users of third generation oral contraceptives [Abstract]. Pharmacoepidemiol Drug Saf 1996; 5(suppl 1): S88. Gerstman BB, Piper JM, Tomita DK, Ferguson WJ, Stadel BV, Lundin FE. Oral contraceptive estrogen dose and the risk of deep venous thromboembolic disease. Am J Epidemiol 1991; 133: 32–37. i3 Drug Safety Final Report. The risk of venous thromboembolism, myocardial infarction, and ischemic stroke among women using the transdermal contraceptive system compared to women using norgestimate-containing oral contraceptives with 35 μg of ethinyl estradiol. Prepared for Johnson and Johnson PRD, August 3, 2006. Abrams LS, Skee D, Natarajan J, Wong FA. Pharmacokinetic overview of Ortho EvraTM /EvraTM . Fertility Sterility 2002; 77(Suppl 2): S3–S12. Fuertes-de la Haba A, Curet JO, Pelegrina I, Bangdiwala I. Thrombophlebitis among oral and nonoral contraceptive users. Obstet Gynecol 1971; 38: 259–263. Vessey M, Doll R, Fairbain A, Glober G. Postoperative thromboembolism and the use of oral contraceptives. BMJ 1970; 3: 123–126.
BLUK077-Stein
March 11, 2007
26
7:7
CHAPTER 26
Obesity as a risk factor in venous thromboembolism
Obesity was shown to be a risk factor for pulmonary embolism (PE) and deep venous thrombosis (DVT) [1]. Among hospitalized patients diagnosed with obesity, based on data from the National Hospital Discharge Survey (NHDS) [2], 91,000 of 12,015,000 (0.8%) had PE [1] (Figure 26.1). Among hospitalized patients who were not diagnosed with obesity, PE was diagnosed in 2,366,000 of 691,000,000 (0.3%). Deep venous thrombosis was diagnosed in 243,000 of 12,015,000 (2.0%) patients diagnosed with obesity, and in 5,524,000 of 691,000,000 (0.8%) who were not diagnosed with obesity. The relative risk of PE, comparing obese patients with non-obese patients, was 2.21 and for DVT it was 2.50 [1]. The relative risks for PE and DVT were agedependent (Table 26.1). Obesity had the greatest impact on patients <40 years of age, in whom the relative risk for PE in obese patients was 5.19 and the relative risk for DVT was 5.20 (Table 26.1) [1]. Obese females had a greater relative risk for DVT than obese males, 2.75 versus 2.02 [1]. The prevalence of both PE and DVT in hospitalized obese females was
PE, DVT (%)
2.5
DVT
2 2.0 1.5 PE
1
DVT PE
0.8
0.5 0
0.8
0.3 Non-obese
Obese
Figure 26.1 Pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized patients from 1979 to 1999 showing the prevalence in obese and non-obese patients. (Data based on Stein et al. [1].)
higher than in obese males (Figure 26.2). In females <40 years of age, the relative risk for DVT comparing obese to non-obese patients was 6.10. In males <40 years of age, the relative risk for DVT was 3.71. The proportion of hospitalized patients diagnosed with obesity was within a narrow range (1.4–2.4%) over the 21-year period of observation from 1979 to 1999, indicating consistency in the diagnostic process
Table 26.1 Relative risks of pulmonary embolism and deep venous thrombosis according to age among obese and non-obese patients. Obese vs. non-obese Pulmonary embolism Age groups
Relative risk
(95% CI)
Deep venous thrombosis Relative risk
(95% CI)
<40
5.19
(5.11–5.28)
5.20
(5.15–5.25)
40–49
1.94
(1.91–1.97)
2.13
(2.11–2.15)
50–59
1.25
(1.23–1.27)
1.67
(1.65–1.68)
60–69
1.42
(1.40–1.44)
1.88
(1.87–1.90) (1.87–1.91)
70–79
2.07
(2.04–2.10)
1.89
>80
3.15
(3.08–3.22)
2.16
(2.12–2.20)
All ages
2.21
(2.20–2.23)
2.50
(2.49–2.51)
Reprinted from Stein et al. [1], with permission from Elsevier.
125
BLUK077-Stein
March 11, 2007
7:7
126
PART I
DVT
PE, DVT (%)
2.5 DVT
2 1.5 1 0.5 0
PE
PE
2.2
1.7
0.6 Obese male
0.8 Obese females
Figure 26.2 Pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized patients from 1979 to 1999 showing the prevalence in obese men and obese women. (Data based on Stein et al. [1].)
[1]. Previous investigators used several indices of obesity including a body mass index (BMI) >35 kg/m2 as well as BMI 30–35 kg/m2 [3], >30 kg/m2 [4, 5], BMI ≥29 kg/m2 [6], weight >20% of median recommended weight for height [3], and waist circumference ≥100 cm [7]. In 28–33% of patients reported by Anderson and associates, the physicians’ assessment was accepted [8, 9]. It is likely that all patients diagnosed with obesity in the NHDS database were in fact obese, irrespective of the criteria used. However, some obese patients may not have had a listed discharge diagnosis of obesity, and they would have been included in the non-obese group. This would have tended to reduce the relative risk of obesity in venous thromboembolism (VTE). The relative risks for PE and DVT were similar to relative risks reported in smaller investigations that used defined criteria for obesity. In women with a BMI >30 kg/m2 , the relative risk for DVT, compared to non-obese women was 2.4 [10]. In women with a BMI ≥29 kg/m2 , the relative risk for PE was 2.9 [6]. In the Framingham Study, obesity was a risk factor only in women [4]. Coon and Coller also showed that obesity was a risk factor only in women. Oral contraceptives in obese women increased the relative risk of DVT to 9.8 [10]. Men with a waist circumference ≥100 cm had a relative risk of 3.9 for VTE compared to men with smaller waists [7]. Among men and women together, the risk ratio for DVT, comparing obese to non-obese patients, was 2.39 [11]. Some found that obesity was not an independent risk factor for VTE [12]. Obesity has been suggested to be a risk factor for fatal PE since 1927 [13]. Investigations that reported an increased risk due to obesity have been criticized because they failed to control for hospital confinement
Prevalence, risks, and prognosis of PE and DVT
or other risk factors [12]. High proportions of patients with venous thromboembolic disease have been found to be obese [8, 9], but the importance of the association is diminished because of the high proportion of obesity in the general population [14]. Some investigations showed an increased risk ratio for DVT or PE, in women [4, 6, 10, 15], but data in men were less compelling. One investigation showed obesity to be a risk factor in men [7] and two did not [4, 15]. Some found no evidence that obesity was an independent risk factor in men or women [12]. Case series of morbidly obese patients (>100 pounds overweight or twice ideal weight) who underwent gastric bypass surgery, showed only a small incidence of postoperative VTE [5, 16]. Enoxaparin was shown to be effective for thromboprophylaxis in morbidly obese patients following bariatric surgery [17]. With various dosing regimens among 544 patients, PE occurred in 0.7% and none developed DVT [17]. All PE occurred after the cessation of enoxaparin, 7 days to 1 month after operation.
References 1 Stein PD, Beemath A, Olson RE. Obesity as a risk factor in venous thromboembolism. Am J Med 2005; 118: 978–980. 2 US Department of Health and Human Services, Public Health Service, National Center for Health Statistics National Hospital Discharge Survey 1979–1999 Multiyear Public-Use Data File Documentation. Available at: http://www.cdc.gov/nchs/about/major/hdasd/nhds.htm. 3 Farmer RD, Lawrenson RA, Todd JC et al. A comparison of the risks of venous thromboembolic disease in association with different combined oral contraceptives. Br J Clin Pharmacol 2000; 49: 580–590. 4 Goldhaber SZ, Savage DD, Garrison RJ et al. Risk factors for pulmonary embolism. The Framingham Study. Am J Med 1983; 74: 1023–1028. 5 Printen KJ, Miller EV, Mason EE, Barnes RW. Venous thromboembolism in the morbidly obese. Surg Gynecol Obstet 1978; 147: 63–64. 6 Goldhaber SZ, Grodstein F, Stampfer MJ et al. A prospective study of risk factors for pulmonary embolism in women. JAMA 1997; 277: 642–645. 7 Hansson PO, Eriksson H, Welin L, Svardsudd K, Wilhelmsen L. Smoking and abdominal obesity: risk factors for venous thromboembolism among middle-aged men: “the study of men born in 1913.” Arch Intern Med 1999; 159: 1886–1890. 8 Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence
BLUK077-Stein
March 11, 2007
CHAPTER 26
9
10
11
12
7:7
Obesity as a risk factor in VTE
and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151: 933–938. Anderson FA, Jr, Wheeler HB, Goldberg RJ, Hosmer DW, Forcier A. The prevalence of risk factors for venous thromboembolism among hospital patients. Arch Intern Med 1992; 152: 1660–1664. Abdollahi M, Cushman M, Rosendaal FR. Obesity: risk of venous thrombosis and the interaction with coagulation factor levels and oral contraceptive use. Thromb Haemost 2003; 89: 493–498. Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160: 3415–3420. Heit JA, Silverstein MD, Mohr DN et al. The epidemiology of venous thromboembolism in the community. Thromb Haemost 2001; 86: 452–463.
127
13 Snell AM. The relation of obesity to fatal postoperative pulmonary embolism. Arch Surg 1927; 15: 237– 244. 14 Hedley AA, Ogden CL, Johnson CL et al. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA 2004; 291: 2847–2850. 15 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. 16 Kerstein MD, McSwain NE, Jr, O’Connell RC, Webb WR, Brennan LA. Obesity: is it really a risk factor in thrombophlebitis? South Med J 1987; 80: 1236–1238. 17 Hamad GG, Choban PS. Enoxaparin for thromboprophylaxis in morbidly obese patients undergoing bariatric surgery: findings of prophylaxis against VTE outcomes in bariatric surgery patients receiving enoxaparin (PROBE) study. Obesity Surg 2005; 15: 1368–1374.
BLUK077-Stein
April 6, 2007
27
15:23
CHAPTER 27
Hypercoagulable syndrome
Inherited thrombophilia Patients with inherited thrombophilia (Table 27.1) tend to have clinical episodes of venous thromboembolism (VTE) before 45 years of age, and episodes tend to be recurrent [1].
Antithrombin III deficiency Antithrombin III is a naturally occurring anticoagulant that inactivates a number of enzymes in the coagulation cascade (factors IIa, IXa, Xa, XIa, and XIIa) (Figure 27.1). A diagram of the coagulation cascade is shown in Chapter 85, Figure 85.1. Antithrombin III deficiency is inherited as an autosomal dominant trait with heterozygotes having an increased risk of VTE [2]. There are two types of antithrombin III deficiency [3]. In type I deficiency, there is a reduction of functional antithrombin and in type II deficiency, there is an abnormal molecule [3]. Heterozygosity for antithrombin III deficiency is found in about 4% of families with inherited thrombophilia, in 1% of patients with a first episode of deep venous thrombosis (DVT), and
in 0.02% of healthy individuals [4] (Table 27.2). The prevalence of antithrombin III deficiency among patients with thrombosis (predominantly venous thrombosis) ranges from 0.5 to 8% [5]. Deep venous thrombosis, with or without pulmonary embolism was present in 90% of patients with antithrombin III deficiency [1]. Patients with antithrombin III deficiency have a 8–10 times greater risk of developing thrombosis than individuals with normal coagulation [1, 6]. The antithrombin–heparin cofactor assay using a factor Xa and a thrombin inhibition assay are laboratory screening tests for this disorder. Since antithrombin III is a cofactor for heparin, heparin will not be effective in patients with antithrombin III deficiency [7]. In fact, heparin resistance may be an indication of antithrombin III deficiency [7].
Protein C deficiency Protein C deficiency is inherited as an autosomal dominant disorder and heterozygosity is a significant risk factor for VTE [8]. Two types of protein C deficiency have been reported [9]. Type I deficiency is a quantitative deficiency with decreased amounts protein C
Table 27.1 Inherited and acquired thrombophilic factors. Inherited thrombophilic factors Antithrombin III deficiency
Antiphospholipid syndrome
Protein C deficiency
Heparin induced thrombocytopenia
Protein S deficiency
Dysfibrinogenemia
Factor V Leiden
Myeloproliferative disorders
Prothrombin 20210A mutation
Malignancy
Elevated levels of factor VIII Elevated levels of factor XI Heparin cofactor II deficiency Dysfibrinogenemia Decreased levels of plasminogen Decreased levels of plasminogen activator Hyperhomocystenemia Sticky platelet syndrome
128
Acquired thrombophilic factors
BLUK077-Stein
April 6, 2007
CHAPTER 27
15:23
129
Hypercoagulable syndrome in VTE
XIIa
Tissue factor (TF) VIIa TF/VIIa complex
XIa
XI IX X
AT III IXa AT III
VIIIa Figure 27.1 Simplified coagulation cascade showing sites of action of antithrombin III (AT III). Deficiency of AT III results in failure to naturally inhibit the action of factors XIIa, XIa, IXa, Xa, and IIa.
activity due to reduced synthesis. Type II deficiency is a qualitative defect due to a defective protein C molecule [9]. Heterozygosity for protein C deficiency is found in about 6% of families with inherited thrombophilia, in 3% of patients with a first episode of unexplained DVT, and in 0.3% of healthy individuals [4] (Table 27.2). The prevalence of protein C deficiency among patients with thrombosis ranges from 1.5 to 11.5% [5]. Deep venous thrombosis, with or without pulmonary embolism was present in 88% of patients with protein C deficiency [1]. Patients with Protein C deficiency have a 4–10 times greater risk of thrombosis as compared with control groups with normal coagulation [1, 6, 10]. The best screening tests for deficiencies of protein C are functional assays that detect both quantitative and qualitative defects of protein C. Immunologic assays detect only quantitative deficiencies of protein C [7]. Coagulation assays for protein C can give falsely low values if the factor V Leiden mutation is present, and as a result, the presence of factor V Leiden mutation should be assessed prior to application of coagulation assays for protein C. Short-term management of thrombosis among patients with protein C deficiency should be with heparin or low-molecular-weight heparin. A vitamin K antagonist such as warfarin should be considered for long-term treatment [7].
Xa IIa II prothrombin Va thrombin I fibrinogen
Ia fibrin
inadequate amount of normally functional protein S present in both the free and bound forms. Type II deficiency is a defective protein S molecule. Type III protein S deficiency is characterized by a low amount of free protein S, but an overall normal amount of total protein S. The large majority of patients with protein S deficiency have a type I deficiency, the prevalence of which is 6% in families with inherited thrombophilia and 1–2% of patients with first time unexplained DVT [4] (Table 27.2). The prevalence of protein S deficiency among patients with thrombosis ranges from 1.5 to 13.2% [5]. Among patients with protein S deficiency, 74% develop DVT and 72% develop superficial thrombophlebitis [11]. Others reported DVT, with or without PE, to be present in 100% of patients with protein S deficiency [1]. Patients with protein S deficiency have an 8–10 times higher risk of thrombosis compared with individuals with normal coagulation [1, 6]. As with protein C deficiency, the screening tests for protein S deficiency are functional assays that are most reliable when factor V Leiden mutation is ruled out. It is important to measure free protein S, since some patients with hereditary protein S deficiency may have normal or borderline total protein S levels. Treatment for protein S deficiency is the same as in protein C deficiency.
Protein S deficiency Protein S deficiency is inherited as an autosomal dominant disorder, with heterozygotes having an increased risk of VTE when compared with their unaffected family members [11]. There are three classifications of protein S deficiency. Type I deficiency results from an
Activated protein C resistance associated with factor V mutation (factor V Leiden) The Leiden mutation of factor V is the most common genetic abnormality associated with VTE. It is found in
130 —
6 — — — — — — —
Protein S deficiency
Factor V Leiden
Prothrombin 20210A mutation
Elevated factor VIII levels
Elevated factor XI levels
Heparin cofactor II deficiency
Dysfibrinogenemia
—
Malignancy
—
—
—
—
25
—
—
1–2
Prevalence of
—
—
10
—
—
—
25
6
20
1.5–13.2
1.5–11.5
0.5–8
unexplained VTE (%)
disorders in Pts with
* Prevalence of disorder among Caucasians in general population. Pts, patients; Gen’l, general; DVT, deep venous thrombosis; VTE, venous thromboembolism.
—
Antiphospholipid syndrome
Acquired thrombophilic factors
Hyperhomocystenemia
—
6
3
4
1
DVT (%)
Protein C deficiency
thrombophilia (%)
Antithrombin III deficiency
Inherited thrombophilic factors
Thrombophilic disorders
Patients with first episode of
family members with
—
—
5–10
—
—
—
2* 11*
— 5*
0.3
0.02
population (%)
disorders in Gen’l
Prevalence of
15–50
29–55
—
10
36
—
—
6
57
74–100
88
90
disorder (%)
VTE with
Frequency of
2
11
2.5
—
—
2.2
5–6
2.8
2–8
8–10
4–10
8–10
VTE
risk For
Relative
[38, 39]
[33–35]
[7, 10, 13, 23]
[22]
[21]
[10, 19]
[10, 13, 16]
[3, 10, 13]
[1, 10, 13]
[1, 4, 6, 11]
[1, 4, 6, 10]
[1, 4, 6]
References
April 6, 2007
Incidence of disorder among
Table 27.2 Incidence of various thrombophilic disorders and associated venous thromboembolism.
BLUK077-Stein
15:23
BLUK077-Stein
April 6, 2007
CHAPTER 27
15:23
131
Hypercoagulable syndrome in VTE
Tissue factor (TF)
XIIa
VIIa TF/VIIa complex
XI
XIa IX
IXa
X VIIIa
Figure 27.2 Simplified coagulation cascade showing the involvement of factor V Leiden in the coagulation pathway. Activated factor V Leiden is resistant to inhibition by activated protein C.
about 5% of Caucasians, but is extremely rare in people of African and Asian descent [7, 12, 13]. Factor V Leiden is an autosomal dominant condition in which the mutated factor V is resistant to inactivation by activated protein C (Figure 27.2). About 4–7% of the general population is heterozygous for factor V Leiden and about 0.06–0.25% of the population is homozygous for factor V Leiden [1] (Table 27.2). The prevalence of factor V Leiden among patients with unexplained venous thrombosis is 20% [13]. Deep venous thrombosis, with or without pulmonary embolism, was present in 57% of patients with factor V Leiden mutations [1]. Patients with factor V Leiden mutation presented with a 2- to 8-fold increase risk of thrombosis compared with individuals with normal coagulation [1, 10]. The relative risk of thrombosis for carriers was shown to have increased 7-fold for heterozygotes and 80-fold for homozygotes among patients <70 years of age with no malignancy [14]. One way to test for the factor V Leiden mutation is to measure the plasma clotting time in the absence and presence of activated protein C [3]. Newer versions of this test can be performed when patients are receiving anticoagulants [15]. The most direct test for factor V Leiden is DNA testing (polymerase chain reactionbased test) [3].
Prothrombin 20210A mutation The prothrombin 20210A allele is due to a glycine-toarginine transition in position 20210 of the prothrombin gene [3]. The abnormal gene causes increased concentrations of prothrombin. It is thought that the increased amount of circulating prothrombin upreg-
Xa Va Leiden II prothrombin
V Leiden Activated protein C
IIa thrombin
I fibrinogen
Ia fibrin
ulates the coagulation cascade [3]. The prevalence of prothrombin 20210A mutation among Caucasians in the general population is 2% [13]. About 6.2% of patients with the prothrombin 20210A gene mutation had venous thrombosis [3, 10, 13]. The diagnosis of prothrombin 20210A is based entirely on DNA analysis of the genes since it is the only genetic coagulation defect that cannot be reliably diagnosed by functional or immunological tests [3].
Elevated factor VIII levels The prevalence of elevated factor VIII levels among Caucasians in the general population is 11% [13]. Patients with high levels of factor VIII (>1500 IU/L) had a 5-fold increase risk for thrombosis when compared to patients with lower levels of factor VIII (<1000 IU/L) [16]. Among patients presenting with an initial episode of unexplained DVT, 25% had factor VIII levels above 1500 IU/L [10, 13]. For every 100 IU/L rise in levels of factor VIII, the risk for a single episode of DVT increased 10% and the risk for recurrent DVT increased 24% [17]. Factor VIII may be increased during an acute illness [18]. A C-reactive protein level should therefore be measured to determine if the increase of factor VIII is likely to be an acute phase reactant [18]. Pregnancy and use of oral contraceptives may also raise levels of factor VIII [7].
Elevated factor XI levels Increased levels of factor XI, a component of the intrinsic pathway, has been implicated as a risk factor for VTE [19]. A 120.8% increase in levels of factor XI
BLUK077-Stein
April 6, 2007
15:23
132
PART I
was associated with a 2.2-fold increased risk of venous thrombosis [10, 19].
Disorders of plasmin generation
Heparin cofactor II deficiency Heparin cofactor II, also known as heparin cofactor A or dermatan sulfate cofactor, inhibits thrombin by forming a heparin cofactor II-thrombin complex [20]. The physiologic function of the molecule is unclear, but it may serve as a natural antithrombotic agent in the presence of dermatan sulfate. Inherited deficiency of heparin cofactor II is rare. It is inherited as an autosomal dominant trait. Limited studies have shown that heterozygosity for heparin cofactor II deficiency is not a likely risk for thrombosis without other concomitant risk factors [21]. Some, however, have reported thrombotic episodes in 36% of individuals with the deficiency [21].
Inherited dysfibrinogenemias A number of abnormal fibrinogens are associated with the hypercoagulable syndrome of dysfibrinogenemias. Ten percent of patients with dysfibrinogenemias develop venous thrombosis with arterial thrombosis being rare [22]. Hereditary dysfibrinogenemias is an autosomal dominant trait except for a few cases that appear to be recessive traits [22]. The thrombin time (TT) is the most sensitive screening test for dysfibrinogenemias. The thrombin time may be prolonged or shortened and reptilase time is also prolonged. Patients with recurrent thrombotic events may require long-term anticoagulation with Coumadin or subcutaneous heparin [20].
Prevalence, risks, and prognosis of PE and DVT
Dysplasminogenemia, decreased levels of plasminogen, decreased synthesis or release of tissue plasminogen activator, and increased levels of plasminogen activator inhibitor are associated with impaired fibrinolysis (Figure 27.3). Routine screening is thought not to be cost-effective and is not indicated [20]. Long-term treatment is with warfarin or low-molecular-weight heparin [7].
Hyperhomocystenemia Elevated plasma homocysteine levels constitute a risk factor for venous as well as arterial thrombosis [23]. Some suggest that hyperhomocystenemia contributes to venous and arterial thrombosis by its toxic effect on the vascular endothelium and on the clotting cascade [23]. Hyperhomocystenemia occurs in 5–10% of the general population [7] and in 10% of patients with VTE [13], with a relative risk of venous thromboembolism of 2.5 [10, 23]. Hyperhomocystenemia can be diagnosed by measuring fasting homocysteine plasma levels [15]. Hyperhomocysteinemia can also be diagnosed by fluorescence polarization immunoassays [24]. Thrombosis is treated in a standard fashion in addition to vitamin B6, B12, and folic acid supplementation [7].
Sticky platelet syndrome Sticky platelet syndrome is inherited in an autosomal dominant trait that results in platelets that are hyperaggregable to epinephrine and/or adenosine
Thrombin Fibrinogen
Fibrin
Plasmin
Plasminogen
PA inhibitor
tPA
Fibrin degradation products Figure 27.3 Fibrinolysis pathway. Dysplasminogenemia, decreased levels of plasminogen, decreased synthesis or release of tissue plasminogen activator (tPA), and increased levels of plasminogen activator inhibitor (PAI) contribute to the maintenance of fibrin, thereby promoting a hypercoagulable state.
BLUK077-Stein
April 6, 2007
CHAPTER 27
15:23
133
Hypercoagulable syndrome in VTE
phosphate [7]. Both arterial, and to a lesser-degree venous thrombi may occur in patients with sticky platelet syndrome [25]. Treatment is with either 81 mg or 325 mg aspirin daily [25]. If aspirin fails, an adenosine diphosphate receptor agonist such as clopidrogel, may be used [25].
Acquired thrombophilia The antiphospholipid syndrome Antiphospholipid antibodies are associated with both arterial and venous thrombosis [26]. The most commonly detected subgroups of antiphospholipid antibodies are lupus anticoagulant antibodies, anticardiolipin antibodies, and anti-B2 -glycoprotein I antibodies [27]. Increased expression and activity of tissue factor is mediated by antibodies against phospholipidbinding proteins present in endothelial cells or other cells. The antigen–antibody reaction releases tissue factor which initiates the coagulation cascade leading to the prothrombotic state [27–29]. The exact mechanism by which these antibodies induce the transduction signal to produce tissue factor are not yet clarified [28]. Primary antiphospholipid syndrome occurs in patients without clinical evidence of an autoimmune disease, whereas secondary antiphospholipid syndrome occurs in association with autoimmune diseases. Systemic lupus erythematosus (SLE) is the most common autoimmune disorder associated with the antiphospholipd syndrome [27]. Among patients with SLE who had the antiphospholipid antibodies, anticardiolipin antibodies were present in 12–30% [30, 31] and lupus anticoagulant antibodies were present in 15–34% [31, 32]. Deep venous thrombosis, the most common manifestation of the antiphospolipid syndrome, occurs in 29–55% of patients with the syndrome, and about half of these patients have pulmonary emboli [33–35]. The estimated relative risk for a first episode of venous thromboembolism in patients with a thrombophilic defect as compared to healthy individuals is 11 [6]. Arterial thrombosis is less common than venous thrombosis in patients with antiphospholipid syndrome [33– 35]. Since there is no definitive association between specific clinical manifestations and particular subgroups of antiphospholipid antibodies, and patients may be negative according to one test and positive for another,
multiple tests for antiphospholipid antibodies are recommended [27].
Heparin-induced thrombocytopenia In patients with heparin-induced thrombocytopenia, platelet activation occurs due to the binding of heparin to platelet factor 4. This forms a platelet– heparin complex that causes endothelial cell injury and activates tissue factor and the coagulation cascade [20, 36]. Thrombosis, either venous or arterial, occurs in about 20% of patients with heparininduced thrombocytopenia [20]. Platelet transfusion in patients with heparin-induced thrombocytopenia is contraindicated because in such patients intravascular platelet aggregation also occurs, thus further contributing to the thrombosis [20]. Management involves cessation of heparin and switching to alternative anticoagulant options.
Acquired dysfibrinogenemia Acquired dysfibrinogenemia occurs most often in patients with severe liver disease [22]. The impairment of the fibrinogen is due to a structural defect caused by an increased carbohydrate content impairing the polymerization of the fibrin, depending on the degree of abnormality of the fibrinogen molecule [22]. Screening tests and treatment are the same as with acquired dysfibrinogenemia.
Myeloproliferative disorders Polycythemia vera, essential thrombocytopenia, and agnogenic myeloid metaplasia are associated with thrombocytosis and an increase in whole blood viscosity [20], thereby contributing to the hypercoagulable state. These complications may result because of altered interactions between platelets, white blood cells, or endothelial cells, due to either altered receptor expression, receptor–ligand interactions, or signaling events. Thrombosis may occur in arteries or veins [37].
Malignancy Cancer is the second most common cause of hypercoagulability, accounting for 10–20% of spontaneous DVTS [7]. Approximately 15% of patients with cancer
BLUK077-Stein
April 6, 2007
15:23
134
have clinical thombosis and up to 50% have thrombosis on autopsy [38]. The mechanism by which tumor cells contribute to thrombosis is not clear. It has been shown that tumor cells interact with thrombin and plasmin and directly influence thrombus formation [20]. Patients with malignancy have a 2-fold increase of developing venous thromboembolism [39].
Diagnostic approach to patients suspected of thrombophilia “Strongly” thrombophilic patients are those patients who sustained their first VTE event before 50 years of age, have a history of recurrent thrombosis, or have a first-degree relative with documented VTE events occurring before 50 years of age [40]. If one or more of these features are present, a complete evaluation for hereditary thrombophilia is appropriate [40]. Testing should be extended to their first-degree family members as well. Because most of the tests are not reliable during anticoagulation, it is preferable to postpone laboratory testing until after discontinuation of treatment [40].
References 1 Martinelli I, Mannucci PM, De Stefano V et al. Different risks of thrombosis in four coagulation defects associated with inherited thrombophilia: a study of 150 families. Blood 1998; 92: 2353–2358. 2 Thaler E, Lechner K. Antithrombin III deficiency and thromboembolism. Clin Haematol 1981; 10: 369–390. 3 Bertina RM. Factor V Leiden and other coagulation factor mutations affecting thrombotic risk. Clin Chem 1997; 43: 1678–1683. 4 Lane DA, Mannucci PM, Bauer KA et al. Inherited thrombophilia: Part 1. Thromb Haemost 1996; 76: 651–662. 5 Mateo J, Oliver A, Borrell M, Sala N, Fontcuberta J. Laboratory evaluation and clinical characteristics of 2,132 consecutive unselected patients with venous thromboembolism—results of the Spanish Multicentric Study on Thrombophilia (EMET-Study). Thromb Haemost 1997; 77: 444–451. 6 Weitz JI, Middeldorp S, Geerts W, Heit JA. Thrombophilia and new anticoagulant drugs. Hematology (Am Soc Hematol Educ Program) 2004: 424–438. 7 Thomas RH. Hypercoagulability syndromes. Arch Intern Med 2001; 161: 2433–2439. 8 Allaart CF, Poort SR, Rosendaal FR, Reitsma PH, Bertina RM, Briet E. Increased risk of venous thrombosis in carri-
PART I
9
10
11
12
13 14
15 16
17
18
19
20 21
22 23
24
Prevalence, risks, and prognosis of PE and DVT
ers of hereditary protein C deficiency defect. Lancet 1993; 341: 134–138. Reitsma PH, Bernardi F, Doig RG et al. Protein C deficiency: a database of mutations, 1995 update. On behalf of the Subcommittee on Plasma Coagulation Inhibitors of the Scientific and Standardization Committee of the ISTH. Thromb Haemost 1995; 73: 876–889. Kamphuisen PW, Rosendaal FR. Thrombophilia screening: a matter of debate. Neth J Med 2004; 62: 180– 187. Engesser L, Broekmans AW, Briet E, Brommer EJ, Bertina RM. Hereditary protein S deficiency: clinical manifestations. Ann Intern Med 1987; 106: 677–682. Bounameaux H. Factor V Leiden paradox: risk of deepvein thrombosis but not of pulmonary embolism. Lancet 2000; 356: 182–183. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999; 353: 1167–1173. Rosendaal FR, Koster T, Vandenbroucke JP, Reitsma PH. High risk of thrombosis in patients homozygous for factor V Leiden (activated protein C resistance). Blood 1995; 85: 1504–1508. Federman DG, Kirsner RS. An update on hypercoagulable disorders. Arch Intern Med 2001; 161: 1051–1056. Koster T, Blann AD, Briet E, Vandenbroucke JP, Rosendaal FR. Role of clotting factor VIII in effect of von Willebrand factor on occurrence of deep-vein thrombosis. Lancet 1995; 345: 152–155. Kraaijenhagen RA, in’t Anker PS, Koopman MM et al. High plasma concentration of factor VIIIc is a major risk factor for venous thromboembolism. Thromb Haemost 2000; 83: 5–9. Cumming AM, Shiach CR. The investigation and management of inherited thrombophilia. Clin Lab Haematol 1999; 21: 77–92. Meijers JC, Tekelenburg WL, Bouma BN, Bertina RM, Rosendaal FR. High levels of coagulation factor XI as a risk factor for venous thrombosis. N Engl J Med 2000; 342: 696–701. Nachman RL, Silverstein R. Hypercoagulable states. Ann Intern Med 1993; 119: 819–827. Adcock DM, Jensen R, Johns CS, Macy PA. Coagulation Handbook. Esoterix Coagulation, Aurora, Colorado, 2002: 25–27. Brick W, Burgess R, Faguet GB. Dysfibrinogenemia. WebMD. www.webmd.com. Last accessed June 26, 2006. den Heijer M, Koster T, Blom HJ et al. Hyperhomocysteinemia as a risk factor for deep-vein thrombosis. N Engl J Med 1996; 334: 759–762. Tripodi A, Negri B, Bertina RM, Mannucci PM. Screening for the FV:Q506 mutation–evaluation of thirteen plasmabased methods for their diagnostic efficacy in comparison with DNA analysis. Thromb Haemost 1997; 77: 436–439.
BLUK077-Stein
April 6, 2007
CHAPTER 27
15:23
Hypercoagulable syndrome in VTE
25 Mammen EF. Sticky platelet syndrome. Semin Thromb Hemost 1999; 25: 361–365. 26 Greaves M. Antiphospholipid antibodies and thrombosis. Lancet 1999; 353: 1348–1353. 27 Levine JS, Branch DW, Rauch J. The antiphospholipid syndrome. N Engl J Med 2002; 346: 752–763. 28 Amengual O, Atsumi T, Khamashta MA. Tissue factor in antiphospholipid syndrome: shifting the focus from coagulation to endothelium. Rheumatology (Oxford) 2003; 42: 1029–1031. 29 Ames PR. Antiphospholipid antibodies, thrombosis and atherosclerosis in systemic lupus erythematosus: a unifying ‘membrane stress syndrome’ hypothesis. Lupus 1994; 3: 371–377. 30 Merkel PA, Chang Y, Pierangeli SS, Convery K, Harris EN, Polisson RP. The prevalence and clinical associations of anticardiolipin antibodies in a large inception cohort of patients with connective tissue diseases. Am J Med 1996; 101: 576–583. 31 Cervera R, Khamashta MA, Font J et al. Systemic lupus erythematosus: clinical and immunologic patterns of disease expression in a cohort of 1,000 patients. The European Working Party on Systemic Lupus Erythematosus. Medicine (Baltimore) 1993; 72: 113–124. 32 Love PE, Santoro SA. Antiphospholipid antibodies: anticardiolipin and the lupus anticoagulant in systemic lupus erythematosus (SLE) and in non-SLE disorders. Prevalence and clinical significance. Ann Intern Med 1990; 112: 682–698.
135
33 Asherson RA, Khamashta MA, Ordi-Ros J et al. The “primary” antiphospholipid syndrome: major clinical and serological features. Medicine (Baltimore) 1989; 68: 366– 374. 34 Alarcon-Segovia D, Perez-Vazquez ME, Villa AR, Drenkard C, Cabiedes J. Preliminary classification criteria for the antiphospholipid syndrome within systemic lupus erythematosus. Semin Arthritis Rheum 1992; 21: 275– 286. 35 Vianna JL, Khamashta MA, Ordi-Ros J et al. Comparison of the primary and secondary antiphospholipid syndrome: a European Multicenter Study of 114 patients. Am J Med 1994; 96: 3–9. 36 Arepally GM, Mayer IMM. Antibodies from patients with heparin-induced thrombocytopenia stimulate monocytic cells to express tissue factor and secrete interleukin8. Blood 2001; 98: 1252–1254. 37 Kessler CM. Propensity for hemorrhage and thrombosis in chronic myeloproliferative disorders. Semin Hematol 2004; 41: 10–14. 38 Luzzatto G, Schafer AI. The prethrombotic state in cancer. Semin Oncol 1990; 17: 147–159. 39 Stein PD, Beemath A, Meyers FA, Skaf E, Sanchez J, Olson RE. Incidence of venous thromboembolism in patients hospitalized with cancer. Am J Med 2006; 119: 60– 68. 40 Bauer KA. The thrombophilias: well-defined risk factors with uncertain therapeutic implications. Ann Intern Med 2001; 135: 367–373.
BLUK077-Stein
March 12, 2007
II
15:49
PART II
Diagnosis of deep venous thrombosis
137
BLUK077-Stein
March 12, 2007
28
15:49
CHAPTER 28
Deep venous thrombosis of the lower extremities: clinical evaluation
Physical examination remains the initial diagnostic modality that calls attention to the potential diagnosis and upon which physicians must rely. Only 13 of 32 patients (41%) with autopsy proven deep venous thrombosis (DVT) who died of trauma or burns were diagnosed antemortem, and 63 of 118 patients (53%) with autopsy proven DVT who died of pulmonary embolism (PE) had antemortem clinical signs of DVT [1, 2]. Among 37 legs of patients screened by 125 I fibrinogen scans, the clinical signs showed a similar sensitivity (49%) [3]. The specificity of clinical signs among 16 legs evaluated by venography was also low (56%), but these data are biased because patients were selected due to clinical findings [4]. In 12 extremities of patients with DVT shown by dissection at autopsy, the sensitivity of ankle asymmetry ≥1.27 cm was 83%, Homans’ sign was 8%, and local tenderness was 41% [5]. The specificity among 18 extremities of patients in whom DVT was excluded by dissection at autopsy for both ankle asymmetry and Homans’ sign was 94% and the specificity of local tenderness was 89% [5]. Homans’ sign is active and/or passive dorsiflexion of the foot associated with any of the following: (1) pain, (2) incomplete dorsiflexion (with equal pressure applied) to prevent pain, or (3) flexion of the knee to release tension in the posterior muscles with dorsiflexion [6]. A Homans’ sign was present in 44% of patients with DVT of the lower leg, and in 60% of patients with femoral venous thrombosis [7]. The elicitation of pain with inflation of a blood pressure cuff around the calf to 60–150 mm Hg has been recommended as a test for DVT [8]. This test, however, was not shown to be more helpful than the assessment of direct tenderness or leg circumference [7].
Calf asymmetry indicates a need for noninvasive diagnostic tests of the lower extremities to determine whether DVT is present [9]. Asymmetry of the circumference of the ankle, calf, or thigh ≥1 cm has been shown in 90% of patients with DVT, but such asymmetry was also shown in 92% of patients with suspected DVT in whom the diagnosis was excluded [10]. A difference ≥3 cm of calf circumference, however, was associated with a high likelihood of having DVT [11, 12]. A difference of circumference of the calves ≥1 cm, measured 10 cm below the tibial tuberosity, is abnormal [9] and was defined as a quantitative sign. As the difference of calf circumference increased from ≥1 cm to ≥4 cm, the sensitivity decreased from 43% or 53% to 4% or 10% and the specificity increased from 57–98% (Table 28.1, Figure 28.1) [13]. The sensitivity and specificity of signs of DVT in two different diagnostic categories of patients was assessed, the results of which gave an envelope of values [13]. The first diagnostic category was 350 patients from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I) [14] with PE proven by pulmonary angiography. Deep venous thrombosis was assumed to be present based on the 83% and 91% incidence of DVT at autopsy of patients with PE [2, 15]. This category is advantageous for the investigation of the sensitivity of signs of DVT because more than 99% of these patients were identified on the basis of respiratory complaints associated with PE [16]. Although DVT was presumably present, the patients were not selected for evaluation because of signs of DVT in this diagnostic category. The second diagnostic category was 30 patients with suspected PE in PIOPED I in whom DVT was
139
BLUK077-Stein
March 12, 2007
15:49
140
PART II
Diagnosis of DVT
Table 28.1 Calf asymmetry with deep venous thrombosis. Positive leg test
PE +, Assumed DVT
Negative leg test
Circum diff (cm)
sensitivity [N/n (%)]
sensitivity [N/n (%)]
specificity [N1 /n1 (%)]
≥1
16/30 (53)
149/350 (43)
27/47 (57)
≥2
8/30 (27)
59/350 (17)
41/47 (87)
≥3
4/30 (13)
23/350 (7)
45/47 (96)
≥4
3/30 (10)
13/350 (4)
46/47 (98)
N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis; Circum diff, circumference difference of calves. Stein et al. [13], and based on unpublished data from PIOPED I.
diagnosed by objective leg tests of the lower extremities. The advantage of this diagnostic category is that the diagnosis of DVT was made with confidence. The disadvantage is that many patients presumably had leg tests obtained because of clinical manifestations suggestive of DVT. This bias would increase the apparent sensitivity of clinical findings. However, among the entire group of patients in whom leg tests were performed (30 with DVT and 47 with no DVT) 48 of 77 (62%) had no qualitative signs of DVT, and 29 of 77 (38%) had no qualitative signs or measured asymmetry. Therefore, a significant number of patients were referred for objective leg tests only because they had a suspicion of PE. The individual qualitative signs of DVT (edema, erythema, calf tenderness, palpable cord, Homans’ sign) showed a sensitivity of 47% or less, irrespective of whether DVT was diagnosed by objective tests of the
lower extremities or whether it was assumed to be present in patients with PE (Table 28.2) [13]. Edema was the most sensitive sign. Unilateral edema was not statistically significantly more sensitive or specific than bilateral edema. All signs showed a specificity of 83% or higher (Table 28.2). The specificities of signs did not differ to a statistically significant extent. The presence of any sign (edema, erythema, calf tenderness, palpable cord, or Homans’ sign) or measured asymmetry of the calves increased the sensitivity for detection of DVT above the sensitivity of a physical sign alone or asymmetry alone (Table 28.3). The specificity varied inversely with the sensitivity [13]. The combination of a sign on physical examination (edema, erythema, calf tenderness, palpable cord, or Homans’ sign) plus ipsilateral asymmetry was associated with a sensitivity of 33% or lower, but the specificity was 87% or higher (Table 28.4). Among the
100
Sensitivity (%)
80
60
Leg test +
40
PE +
20 0 0
20
40
60
False positives (%)
80
100
Figure 28.1 Relation of sensitivity of measured calf asymmetry to the frequency of false-positive findings. The upper curve (leg test +) shows data of patients in whom objective tests of the lower extremities showed deep venous thrombosis (DVT) to be present or absent. The lower curve (PE +) shows data of patients in whom DVT was estimated to be present because of pulmonary embolism (PE). The numbers indicate the measured difference of calf circumference. (Data from Stein et al. [13], based on unpublished data from PIOPED I.)
BLUK077-Stein
March 12, 2007
CHAPTER 28
15:49
141
Clinical evaluation of DVT
Table 28.2 Signs of deep venous thrombosis. Positive leg test
PE +, Assumed DVT
Negative leg test
Sign
sensitivity [N/n (%)]
sensitivity [N/n (%)]
specificity [ N1 /n1 (%)]
Edema, any
14/30 (47)
79/350 (23)
40/47 (85)
Edema, unilat
30/350 (9)
43/47 (91)
Edema, bilat
49/350 (14)
44/47 (94)
21/350 (6)
43/47 (91)
Erythema
5/30 (17)
Calf tender
5/30 (17)
47/350 (13)
39/47 (83)
Palp cord
0/30 (0)
8/350 (2)
46/47 (98)
Homan’s sig
1/30 (3)
9/350 (3)
44/47 (94)
Unilat, unilateral; bilat, bilateral; N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis. Stein et al. [13], and based on unpublished data from PIOPED I.
Table 28.3 Signs of deep venous thrombosis and/or calf asymmetry. Sign +/or circum
Positive leg test
PE +, Assumed DVT
Negative leg test
diff (cm)
sensitivity [N/n (%)]
sensitivity [N/n (%)]
specificity [N1 /n1 (%)]
Any sign +/or ≥1 cm
24/30 (80)
192/350 (55)
23/47 (49)
Any sign +/or ≥2 cm
20/30 (67)
131/350 (37)
33/47 (70)
Any sign +/or ≥3 cm
19/30 (63)
111/350 (32)
35/47 (74)
Any sign +/or ≥4 cm
18/30 (60)
104/350 (30)
35/47 (74)
N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis; circum diff, circumference difference of calves. Stein et al. [13], and based on unpublished data from PIOPED I.
Table 28.4 Ipsilateral signs of deep venous thrombosis and calf asymmetry. Sign and circum
Positive leg test
PE +, Assumed DVT
Negative leg test
diff (cm)
sensitivity [N/n (%)]
sensitivity [N/n (%)]
specificity [N1 /n1 (%)]
Any sign and ≥1
10/30 (33)
52/350 (15)
41/47 (87)
Any sign and ≥2
5/30 (17)
27/350 (8)
43/47 (91)
Any sign and ≥3
3/30 (10)
11/350 (3)
45/47 (96)
Any sign and ≥4
2/30 (7)
8/350 (2)
46/47 (98)
N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis; circum diff, circumference difference of calves. Stein et al. [13], and based on unpublished data from PIOPED I.
BLUK077-Stein
March 12, 2007
15:49
142
PART II
Diagnosis of DVT
Figure 28.2 Relation of sensitivity of combinations of qualitative signs and measured calf asymmetry to the frequency of false-positive findings. The upper curve (leg test +) shows data of patients in whom objective tests of the lower extremities showed deep venous thrombosis (DVT). The lower curve (PE +) shows data of patients in whom DVT was estimated to be present because of pulmonary embolism (PE). The numbers indicate the measured difference of calf circumference. Sign +/or asym indicates patients who had a qualitative sign and/or asymmetry. Sign + ipsi asym indicates patients who had a qualitative sign plus ipsilateral asymmetry. (Data from Stein et al. [13], based on unpublished data from PIOPED I.)
3–10% of patients who had one or more qualitative signs plus ≥3 cm calf asymmetry, the specificity for DVT was 96% [13]. Among patients in whom DVT was diagnosed by objective leg tests, signs of DVT or measured asymmetry showed a higher sensitivity than in patients in whom DVT was estimated to be present because of PE (Figure 28.2) [13]. If either a qualitative sign or measured asymmetry was present, the presence of one or both was more sensitive than a qualitative sign plus measured asymmetry. The former, however, showed more false-positive values than patients with qualitative signs and ipsilateral measured calf swelling. As the measured difference of calve circumferences increased from ≥1 cm to ≥4 cm in combination with qualitative signs, the sensitivity diminished, and the presence of false-positive values also diminished. Among patients with DVT diagnosed by objective leg tests, 2 of 30 (7%) had leg pain, but no qualitative signs or asymmetry ≥1 cm. An additional 1 of 30 (3%) reported swelling in the leg or foot, but had no leg pain, qualitative signs, or asymmetry ≥1 cm. The sensitivity for the detection of DVT of one or more symptoms, qualitative signs, or asymmetry ≥1 cm was 27 of 30 (90%). Among patients with DVT estimated to be present on the basis of PE, the addition of symptoms of tenderness or swelling to the qualitative or quantitative signs increased the sensitivity for the detection of DVT to 229 of 350 (65%) [13]. The addition of symptoms of tenderness or swelling to qualitative signs or measured asymmetry ≥1 cm
decreased the specificity of one or more of any of these findings to 18 of 47 (38%) [13]. These data show, therefore, that combinations of qualitative signs and measured asymmetry of the calves do not reliably identify patients with DVT, nor does the absence of such signs exclude DVT. However, in patients who have ≥3 cm asymmetry of the calves, the finding is 96% specific [13]. Meta-analysis identified covariates that provided diagnostic accuracy for DVT. Only malignancy and previous DVT were useful for ruling in DVT and recent immobilization, difference in calf diameter, and recent surgery were of borderline value [17]. The positive likelihood ratios [18] ranged from 1.76 to 2.71 indicating that none provided high certainty for a diagnosis [17]. Only absence of calf swelling and absence of a difference in calf diameter were somewhat (borderline) useful for ruling out DVT [17]. The negative likelihood ratios were 0.67 and 0.57 [17].
References 1 Sevitt S, Gallagher N. Venous thrombosis and pulmonary embolism: a clinico-pathological study in injured and burned patients. Br J Surg 1961; 48: 475–489. 2 Byrne JJ, O’Neil EE. Fatal pulmonary emboli. A study of 130 autopsy-proven fatal emboli. Am J Surg 1952; 83: 47–49. 3 Milne RM, Gunn AA, Griffiths JMT, Ruckley CX. Postoperative deep venous thrombosis: a comparison of diagnostic techniques. Lancet 1971; 2: 445–447. 4 Johnson WC. Evaluation of newer techniques for the diagnosis of venous thrombosis. J Surg Res 1974; 16: 473–481.
BLUK077-Stein
March 12, 2007
CHAPTER 28
15:49
Clinical evaluation of DVT
5 McLachlin J, Richards T, Paterson JC. An evaluation of clinical signs in the diagnosis of venous thrombosis. Arch Surg 1962; 85: 738–744. 6 Homans J. Disease of the veins. N Engl J Med 1944; 231: 51–60. 7 DeWeese JA, Rogoff SM. Phlebographic patterns of acute deep venous thrombosis of the leg. Surgery 1963; 53: 99– 108. 8 Lowenberg RI. Early diagnosis of phlebothrombosis with aid of a new clinical test. JAMA 1954; 155: 1566–1570. 9 Stein PD, Henry JW, Gopalakrishnan D, Relyea B. Asymmetry of the calves in the assessment of patients with suspected acute pulmonary embolism. Chest 1995; 107: 936–939. 10 Cranley JJ, Canos AJ, Sull WJ. The diagnosis of deep venous thrombosis: fallibility of clinical symptoms and signs. Arch Surg 1976; 111: 34–36. 11 Lambie JM, Mahaffy RG, Barber DC, Karmody AM, Scott MM, Matheson NA. Diagnostic accuracy in venous thrombosis. BMJ 1970; 2: 142–143. 12 Nypaver TJ, Shepard AD, Kiell CS, McPharlin M, Fenn N, Ernst CB. Outpatient duplex scanning for deep vein thrombosis: parameters predictive of a negative study result. J Vasc Surg 1993; 18: 821–826.
143
13 Stein PD, Henry JW, Relyea B. Sensitivity and specificity of physical examination of the lower extremity in evaluation of deep venous thrombosis. Unpublished data from PIOPED I. 14 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 15 Short DS. A survey of pulmonary embolism in a general hospital. BMJ 1952; 1: 790–796. 16 Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. 17 Goodacre SG, Sutton AJ, Sampson FC. Meta-analysis: the value of clinical assessment in the diagnosis of deep venous thrombosis. Ann Intern Med 2005; 143: 129– 139. 18 Jaeschke R, Guyatt GH, Sackett DL, for The EvidenceBased Medicine Working Group. Users’ guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? JAMA 1994; 271: 703– 707.
BLUK077-Stein
March 11, 2007
29
7:5
CHAPTER 29
Clinical model for assessment of deep venous thrombosis
Scoring systems for deep venous thrombosis Recognizing the difficulty in making a clinical diagnosis of acute deep venous thrombosis (DVT), a clinical model was developed which permits an assessment of the clinical probability [1] (Table 29.1). An earlier model with more features had been developed by the same group [2]. Although it gave good results, it was more complex and seems to have been abandoned. A clinical model for assessment of probability of DVT may be particularly useful when combined with another test, such as D-dimer. This will be discussed in Chapter 32.
Table 29.1 Clinical model for assessment of probability of deep venous thrombosis. Feature Active cancer (treatment ongoing or within
Score 1
previous 6 months or palliative) Paralysis, paresis, or recent immobilization of lower
1
extremities Recently bedridden >3 days or major surgery within
1
4 weeks Localized tenderness along the deep venous system
1
Entire leg swollen
1
Calf swelling >3 cm compared with the
1
asymptomatic leg (measured 10 cm below tibial tuberosity) Pitting edema (greater in the symptomatic leg) Collateral superficial veins (non-varicose) Alternative diagnosis as likely or greater than that
1 1 −2
of deep vein thrombosis High probability ≥3; moderate probability 1–2; low probability ≤0. In patients with symptoms in both legs, the more symptomatic leg is used. Based on data from Wells et al. [1].
144
Positive predictive values of the Wells test in patients with suspected DVT are shown in Table 29.2. Among patients with a low probability Wells test, DVT was shown in 3–13% [1–7] (Table 29.2). Among patients with a high probability assessment by the Wells test, DVT was present in 38–75%. Review of the Wells test in a meta-analysis of 22 studies showed that 11% of patients with a low probability Wells score had DVT [8], A low Wells score markedly reduced the probability of DVT (negative likelihood ratio 0.25 [8]. In this meta-analysis, 56% of patients with a high probability Wells score had DVT. A high Wells score markedly increased the probability of DVT (positive likelihood ratio 5.2) [8]. Risk stratification was more accurate for proximal DVT than for distal DVT [8]. Empirical assessment showed that in patients with suspected DVT who had a low probability clinical assessment, DVT was present in 1–13% [3, 5, 9] (Table 29.2, Figure 29.1). With a high probability clinical assessment, DVT was present in 63–100% [3, 5, 9]. In articles reviewed for a meta-analysis, 52% of patients with a high probability empirical assessment had DVT [8]. Meta-analysis showed that empirical assessment gave similar likelihood ratios as the Wells score, but there were a limited number of studies, and the confidence intervals were wide [8]. In patients with suspected DVT who had a low probability clinical assessment, DVT was present in 8% [8]. The Wells scoring system has been criticized because DVT is not entirely excluded in patients with a low score [6]. Resident physicians calculated a probability score for DVT that disagreed with senior staff in 30 of 165 patients (18%) [10]. Awareness of the subjective aspect of some parts of the Wells’ score was recommended before its implementation in clinical practice [10]. Putting this in perspective, the real value of the Wells score (and any clinical decision aid) is its ability to complement, rather than displace, physicians’
BLUK077-Stein
March 11, 2007
CHAPTER 29
7:5
145
Clinical model for assessment of DVT
Table 29.2 Positive predictive value of empirical assessment and clinical model for probability of deep venous thrombosis.
Method
First author [Ref]
Low probability
Intermediate probability
High probability
[DVT/ N (%)]
[DVT/ N (%)]
[DVT/ N (%)]
Empirical
Cornuz [5]
11/86 (13)
30/127 (24)
41/65 (63)
Empirical
Perrier [9]
3/29 (1)
56/291 (19)
52/54 (96)
Empirical
Miron [3]
1/78 (1)
30/166 (18)
26/26 (100)
15/193 (8)
116/584 (20)
119/145 (82)
Wells [2]
16/301 (5)
47/143 (33)
72/85 (85) 34/46 (74)
Average Wells expanded Wells
Miron [3]
4/126 (3)
19/98 (19)
Wells
Wells [4]
5/50 (10)
14/71 (20)
22/29 (76)
Wells
Cornuz [5]
14/121 (12)
36/121 (30)
32/48 (67)
Wells
Odega [6]
61/507 (12)
53/321 (17)
175/467 (38)
Wells
Kraaijn’n [7]
71/896 (8)
133/508 (26)
208/322 (65)
Wells
Wells [1]
10/39 (3)
32/193 (17)
53/71 (75)
165/1739 (9)
287/1312 (22)
524/983 (53)
Average
Figure 29.1 Pooled data showing prevalence of deep venous thrombosis (DVT) among patients with low, intermediate, and high probability (prob) assessments based upon empirical (empir) evaluation and upon the Wells scoring system. Data based on values shown in Table 29.2.
Deep vein thrombosis (%)
100
Empir 82%
75 Wells 53%
50
25
Empir Wells 8% 9%
Empir Wells 20% 22%
0 Low prob
empirical assessment [11]. The Wells score will classify the patient into a probability range that is probably correct, but it is up to the physician to apply clinical judgment to fine-tune this estimate of disease probability [11].
4
5
References 1 Wells PS, Anderson DR, Bormanis J et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 1997; 350: 1795–1798. 2 Wells PS, Hirsh J, Anderson DR et al. Accuracy of clinical assessment of deep-vein thrombosis. Lancet 1995; 345: 1326–1330. 3 Miron M-J, Perrier A, Bounameaux H. Clinical assessment of suspected deep vein thrombosis: comparison be-
6
7
Intermediate prob
High prob
tween a score and empirical assessment. J Intern Med 2000; 247: 249–254. Wells PS, Anderson DR, Bormanis J et al. Application of a diagnostic clinical model for the management of hospitalized patients with suspected deep-vein thrombosis. Thromb Haemost 1999; 81: 493–497. Cornuz J, Ghali WA, Hayoz D et al. Clinical prediction of deep venous thrombosis using two risk assessment methods in combination with rapid quantitative D-dimer testing. Am J Med 2002; 112: 198–203. Oudega R, Hoes AW, Moons KGM. The Wells rule does not adequately rule out deep venous thrombosis in primary care patients. Ann Intern Med 2005; 143: 100– 107. Kraaijenhagen RA, Piovella F, Bernardi E et al. Simplification of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002; 162: 907–911.
BLUK077-Stein
March 11, 2007
7:5
146
8 Goodacre SG, Sutton AJ, Sampson FC. Meta-analysis: the value of clinical assessment in the diagnosis of deep venous thrombosis. Ann Intern Med 2005; 143: 129–139. 9 Perrier A, Desmarais S, Miron M-J et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet 1999; 353: 190–195.
PART II
Diagnosis of DVT
10 Bigaroni A, Perrier A, Bounameaux H. Is clinical probability assessment of deep vein thrombosis by a score really standardized? Thromb Haemost 2000; 83: 788–789. 11 Douketis JD. Use of clinical prediction score in patients with suspected deep venous thrombosis: two steps forward, one step back? Ann Intern Med 2005; 143: 140–142.
BLUK077-Stein
March 11, 2007
30
7:6
CHAPTER 30
Clinical probability score plus single negative ultrasound for exclusion of deep venous thrombosis
Ultrasonography A clinical model that combines a pretest probability by a scoring system with compression ultrasonography is safe and feasible, and reduces the need for serial ultrasound testing [1]. The Wells point score system for assessment of probability of deep venous thrombosis (DVT) is shown in Chapter 29 [1]. A low probability score for DVT (score ≤0) in combination with a negative compression ultrasound of the lower extremities had a negative predictive value of 97.8–99.7% [1–4]
and an “unlikely” probability (score <2) had a negative predictive value of 98.3–98.9% [5, 6] (Table 30.1). A moderate clinical probability of DVT in combination with a negative ultrasound had a negative predictive value of 96.6–97.0% [1, 2] (Table 30.1). However, if the clinical probability by the point score system was high probability for DVT, a negative venous ultrasound could not be relied upon to exclude DVT. The negative predictive value of a high-probability clinical score and negative ultrasound, in small numbers of patients ranged from 71 to 73% [1, 2] (Table 30.1).
Table 30.1 Negative predictive value in patients with suspected deep vein thrombosis and negative ultrasound according to clinical assessment scores. First author [Ref]
Population
Negative predictive value (%)
Ultrasound negative and low probability or “unlikely” clinical score (Wells score)* Wells score ≤0 (low probability) Wells [2]
Hospitalized
45/46 (97.8)
Wells [1]
Outpatients
317/318 (99.7)
Kraaijenhagen [4]
Outpatients
821/834 (98.4)
Tick [3]
Outpatients
245/250 (98.0)
Wells score <2 (unlikely) Wells [5]*
Outpatients
264/267 (98.9)
Perrier [6]†
Outpatients
234/238 (98.3)
Wells [2]
Hospitalized
58/60 (96.7)
Wells [1]
Outpatients
161/166 (97.0)
Wells [2]
Hospitalized
5/7 (71.4)
Wells [1]
Outpatients
11/15 (73.3)
Wells score 1–2 (moderate probability)
Wells score ≥3 (high probability)
*Unlikely = Wells score <2. † Low or moderate probability. Low probability = Wells score ≤0; moderate probability = Wells score 1–2; high probability = Wells score ≥3.
147
BLUK077-Stein
March 11, 2007
7:6
148
PART II
Diagnosis of DVT
References 1 Wells PS, Anderson DR, Bormanis J et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 1997; 350: 1795–1798. 2 Wells P, Anderson D, Bormanis J et al. Application of a diagnostic clinical model for the management of hospitalized patients with suspected deep-vein thrombosis. Thromb Haemost 1999; 81: 493–497. 3 Tick LW, Ton E, van Voorthuizen T et al. Practical diagnostic management of patients with clinically suspected deep vein thrombosis by clinical probability test, compression
ultrasonography and D-dimer test. Am J Med 2002; 113: 630–635. 4 Kraaijenhagen RA, Piovella F, Bernardi E et al. Simplification of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002; 162: 907–911. 5 Wells PS, Anderson DR, Rodger M et al. Evaluation of Ddimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003; 349: 1227–1235. 6 Perrier A, Desmarais S, Miron MJ et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet 1999; 353: 190–195.
BLUK077-Stein
March 12, 2007
31
15:51
CHAPTER 31
D-dimer for the exclusion of acute deep venous thrombosis
The D-dimer assays in patients with suspected deep venous thrombosis (DVT) differ in sensitivity, specificity, and likelihood ratios from values in patients with suspected pulmonary embolism (PE) [1]. Values of sensitivity of the enzyme-linked immunosorbent assay (ELISA) and quantitative rapid ELISA assays are significantly superior to those for the quantitative latex, semiquantitative latex, and whole blood agglutination assays in patients with suspected DVT [1] (Figure 31.1). The quantitative rapid ELISA assay is more convenient than the conventional ELISA and it provides a high certainty for a negative diagnosis of DVT in patients with a low or intermediate clinical probability (but not high clinical probability) (Chapter 32).
Negative likelihood ratios of the quantitative rapid ELISA assay were 0.09, 0.10, and 0.08 in Tier 1, Tier 2, and Tier 3 studies (Tables 31.1 and 31.2). Tier 1 studies (20 investigations) compared an ELISA assay and at least one other D-dimer assay for exclusion of DVT [2–21] (Table 31.3). Tier 2 included the Tier 1 studies and 29 additional studies [22–50] that met all inclusion criteria (Table 31.4). Tier 3 combined 21 methodologically weaker studies [51–71] with the 49 Tier-2 studies. Negative likelihood ratios <0.1 generate large and often conclusive changes from pretest to posttest probability [72] providing high certainty for excluding DVT. As will be shown, combining a negative rapid ELISA with a low or moderate clinical probability for
Table 31.1 Deep vein thrombosis. Sensitivity
Specificity
Positive likelihood
Negative likelihood
(95% CI)
(95% CI)
ratio (95% CI)
ratio (95% CI)
ELISA
0.96 (0.91–1.00)
0.38 (0.28–0.48)
1.55 (1.32–1.81)
0.12 (0.04–0.33)
Quant rapid ELISA
0.96 (0.90–1.00)
0.44 (0.32–0.55)
1.70 (1.39–2.09)
0.09 (0.02–0.41)
Semiquant rapid ELISA
0.89 (0.81–0.98)
0.39 (0.28–0.50)
1.47 (1.21–1.78)
0.27 (0.12–0.60)
Qual rapid ELISA
0.93 (0.84–1.00)
0.47 (0.30–0.63)
1.75 (1.28–2.39)
0.15 (0.04–0.56)
Quant latex
0.85 (0.74–0.95)
0.66 (0.55–0.78)
2.49 (1.77–3.51)
0.24 (0.12–0.45)
Semiquant latex
0.78 (0.67–0.89)
0.66 (0.56–0.76)
2.30 (1.69–3.13)
0.33 (0.21–0.54)
Whole blood
0.87 (0.68–1.00)
0.83 (0.65–1.00)
4.97 (1.84–13.42)
0.16 (0.04–0.65)
ELISA
0.95 (0.91–0.99)
0.40 (0.32–0.49)
1.60 (1.39–1.83)
0.12 (0.05–0.29)
Quant rapid ELISA
0.96 (0.90–1.00)
0.44 (0.34–0.54)
1.71 (1.43–2.05)
0.10 (0.03–0.36)
Semiquant rapid ELISA
0.90 (0.83–0.98)
0.39 (0.29–0.50)
1.48 (1.24–1.78)
0.25 (0.12–0.55)
Qual rapid ELISA
0.93 (0.87–0.99)
0.46 (0.35–0.57)
1.73 (1.40–2.13)
0.15 (0.07–0.37)
Quant latex
0.86 (0.78–0.94)
0.61 (0.51–0.71)
2.20 (1.70–2.84)
0.23 (0.13–0.41)
Semiquant latex
0.79 (0.69–0.88)
0.66 (0.57–0.75)
2.33 (1.75–3.11)
0.32 (0.20–0.51)
Whole blood
0.86 (0.80–0.93)
0.67 (0.61–0.73)
2.62 (2.17–3.16)
0.20 (0.13–0.32)
Test Tier 1 analysis
Tier 2 analysis
Quant, quantitative; semiquant, semiquantitative; qual, qualitative. Modified and reprinted with permission from Stein et al. [1].
149
BLUK077-Stein
March 12, 2007
15:51
150
Diagnosis of DVT
Proportion
Proportion
PART II
Sensitivity
Specificity
Figure 31.1 Boxplots of sensitivity and specificity among the D-dimer assays for patients with suspected deep venous thrombosis (DVT). (Reprinted with permission from Stein et al. [1].)
DVT essentially rules out DVT (Chapter 32). However, a negative D-dimer does not reliably exclude DVT if the clinical probability is high (Chapter 32). Sensitivity and specificity for DVT according to various assays using a cutoff level of 500 ng/mL are shown
in Figure 31.1. In patients with DVT, the least variability for sensitivity was seen with the ELISA, qualitative rapid ELISA, and quantitative rapid ELISA assays. Limited data for sensitivity, specificity, and likelihood ratios were available for cutoff values of 250 ng/mL and
BLUK077-Stein
March 12, 2007
CHAPTER 31
15:51
151
D-dimer for exclusion of acute DVT
Table 31.2 Deep vein thrombosis.
Test
Sensitivity
Specificity
Positive likelihood
Negative likelihood
(95% CI)
(95% CI)
ratio (95% CI)
ratio (95% CI)
Cutoff 500 ng/mL – Tier 3 analysis (all data) ELISA
0.94 (0.89–0.98)
0.43 (0.36–0.50)
1.65 (1.46–1.87)
0.15 (0.07–0.30)
Quant rapid ELISA
0.97 (0.92–1.00)
0.42 (0.32–0.52)
1.67 (1.42–1.97)
0.08 (0.02–0.38)
Semiquant rapid ELISA
0.91 (0.85–0.98)
0.43 (0.34–0.52)
1.60 (1.37–1.88)
0.21 (0.10–0.42)
Qual rapid ELISA
0.93 (0.87–0.99)
0.53 (0.43–0.64)
1.99 (1.60–2.48)
0.13 (0.06–0.32)
Quant latex
0.88 (0.80–0.95)
0.59 (0.49–0.69)
2.14 (1.68–2.73)
0.21 (0.12–0.38)
Semiquant latex
0.78 (0.69–0.87)
0.70 (0.62–0.78)
2.60 (1.95–3.46)
0.31 (0.21–0.47)
Whole blood
0.82 (0.76–0.89)
0.70 (0.64–0.76)
2.77 (2.27–3.38)
0.25 (0.18–0.36)
Cutoff 250 ng/mL – All studies meeting inclusion criteria ELISA
NA
NA
NA
NA
Quant rapid ELISA
0.98
0.39
1.58
0.07
Semiquant rapid ELISA
0.92
0.56
2.10
0.14
Qual rapid ELISA
NA
NA
NA
NA
Quant Latex
0.91
0.53
1.96
0.16
Semiquant latex
0.91
0.47
1.73
0.19
Whole blood
0.88
0.66
2.57
0.18
Cutoff 1000 ng/ml – All studies meeting inclusion criteria ELISA
0.90
0.72
3.21
0.14
Quant rapid ELISA
0.93
0.58
2.23
0.13
Semiquant rapid ELISA
0.94
0.57
2.19
0.11
Qual rapid ELISA
NA
NA
NA
NA
Quant latex
0.81
0.70
2.73
0.27
Semiquant latex
0.73
0.78
3.35
0.34
Whole blood
0.88
0.66
2.57
0.18
Quant, quantitative; semiquant, semiquantitative; qual, qualitative. Modified and reprinted with permission from Stein et al. [1].
1000 ng/mL for some but not all of the D-dimer assays (Table 31.2). A negative D-dimer has highest negative predictive values in populations with a low prevalence of DVT. The negative predictive value of the quantitative rapid ELISA for exclusion of acute DVT was estimated based on a sensitivity of 96% and specificity of 44% (Tier 1 and Tier 2 studies), knowing the prevalence of DVT in the population tested. The prevalence of DVT in the included investigations of D-dimer ranged from 20 to 78% (average 36%) [1]. If the prevalence of DVT in the population tested were 36%, the negative predictive value of a negative quantitative rapid ELISA would be 95%. If the prevalence of DVT were 20%, the neg-
ative predictive value would be 98%. If the prevalence of DVT were 78%, the negative predictive value would be only 75% (Figure 31.2). A negative D-dimer may be useful for the exclusion of recurrent DVT as well as the exclusion of a suspected first episode of DVT [73]. The reliability of a negative D-dimer for the exclusion of DVT is highest in populations with a low prevalence of DVT and limited when the prevalence is high. Non-ELISA assays should not be used as stand-alone tests [74] given their inferior sensitivity and negative likelihood ratio values. However, in patients with a low clinical probability, non-ELISA assays provide a reasonable certainty for ruling out DVT [1].
BLUK077-Stein
March 12, 2007
15:51
Table 31.3 Patients with suspected DVT.
Study
Year
D-dimer type
Cutoff analyzed (ng/mL)
No. Pts enrolled
Patients with DVT (%)
1989
ELISA,
<500
56
37.5
<500
100
45.0
<500
97
40.2
<500
116
29.3
<500
32
78.1
Quant rapid ELISA,
<500
100
40.0
semiquant rapid ELISA,
<250 <500
114
42.1
ELISA,
<250
96
37.5
semiquant latex
<500
ELISA,
<500
171
43.9
Quant rapid ELISA,
<500
81
51.9
semiquant rapid ELISA,
<250
Tier 1 analysis Bounameaux [3]
semiquant latex Elias [6]
1990
ELISA, semiquant latex
Speiser [18]
1990
Quant rapid ELISA, qual latex
Boneu [2]
1991
ELISA, semiquant latex
Chang-Liem [4]
1991
Quant rapid ELISA, quant latex
Dale [5]
1994
quant latex Hansson [10]
1994
ELISA, quant latex, semiquant latex
Tengborn [19] Elias [7]
1994 1996
Quant rapid ELISA, semiquant rapid ELISA, qual rapid ELISA, semiquant latex Legnani [12]
1997
qual rapid ELISA, ELISA, quant latex, semiquant latex Mayer [15]
1997
Semiquant rapid ELISA, whole blood
<500
108
30.6
Escoffre-Barbe [8]
1998
ELISA,
<500
464
59.5
<250
180
33.3
99
39.4
97
39.2
177
20.3
99
50.5
quant latex Lindahl [14]
1998
Semiquant rapid ELISA, semiquant latex
Legnani [13]
˚ Wahlander [21]
1999
1999
Quant rapid ELISA,
<700
ELISA,
<155
quant latex
<70
Quant rapid ELISA,
≤ 500 ≤ 1000
semiquant rapid ELISA,
<500
Qual rapid ELISA, ELISA, semiquant latex Sadouk [16] van der Graaf [20]
2000 2000
Quant rapid ELISA,
<500
quant latex
<250
Quant rapid ELISA,
<500
semiquant rapid ELISA,
<1000
qual rapid ELISA,
<250
ELISA, quant latex, semiquant latex, whole blood (Continued )
152
BLUK077-Stein
March 12, 2007
15:51
Table 31.3 Patients with suspected DVT (Continued ).
Study
Year
D-dimer type
Cutoff analyzed (ng/mL)
No. Pts enrolled
Patients with DVT (%)
Funfsinn [9]
2001
Quant rapid ELISA,
<500
106
44.3
108
25.0, 20.4, 26.9
113
43.4
ELISA, quant latex Shitrit [17]
Larsen [11]
2001
2002
Quant rapid ELISA,
<1000
quant latex,
<500
semiquant latex
<250
Quant rapid ELISA,
<500
semiquant rapid ELISA, quant latex
DVT, deep venous thrombosis; semiquant, semiquantitative; quant, quantitative; qual, qualitative. Modified and reprinted with permission from Stein et al. [1]. Table 31.4 Tier 2 analysis (Tier 2 includes Tier 1 and the following investigations).
Study
Year
D-dimer type
Cutoff analyzed (ng/mL)
No. Pts enrolled
Patients with DVT (%)
Van Bergen [47]
1989
ELISA
<250
239
24.7
de Boer [28]
1991
Qual latex
<200
33
63.6
Kroneman [39]
1991
Quant rapid ELISA
<268
239
24.7
Ibrahim [34]
1992
Semiquant latex
<250
85
45.9
Jossang [37]
1992
Semiquant latex
<500
69
53.6
Heijboer [33]
1992
ELISA
<300
474
12.2
Pini [45]
1993
Semiquant latex
<200
425
45.9
Wells [49]
1995
Whole blood
—
214
24.8
D’Angelo [27]
1996
Quant rapid ELISA
<500
103
21.4
<1000 Gavaud [30]
1996
ELISA
<370
80
32.5
Borg [23]
1997
ELISA
<500
76
42.1
Gauzzaloca [31]
1997
Qual rapid ELISA
<500
68
52.9
Jacq [35]
1997
Whole blood
—
50
48.0
Knecht [38]
1997
Quant latex
<500
154
38.3
Leroyer [42]
1997
Qual rapid ELISA,
<500
448
59.2
ELISA Khaira [37]
1998
Semiquant rapid ELISA
<500
79
36.7
Wijns [50]
1998
Quant rapid ELISA,
<1000
74
43.2
qual rapid ELISA,
<500
ELISA Aschwanden [22]
1999
Whole blood
—
343
24.2
Caliezi [26]
1999
ELISA
<500
106
44.3
Legnani [40]
1999
Quant latex
<230
92
38.0
Lennox [41]
1999
Whole blood
—
200
23.0
Lindahl [43]
1999
Quant latex
<700
236
43.2
Wells [48]
1999
Whole blood
—
150
26.0
Bradley [24]
2000
Semiquant rapid ELISA
<300
138
31.9
Farrell [29]
2000
Whole blood
—
198
8.1
Permpikul [44]
2000
Whole blood
—
65
66.2
Bucek [25]
2001
Quant latex,
<500
100
35.0
Harper [32]
2001
<250
235
21.7, 21.3
—
1739
24.4
whole blood Quant latex, whole blood ten Wolde [42]
2002
Whole blood
DVT, deep venous thrombosis; qual, qualitative; quant, quantitative; semiquant, semiquantitative. Modified and reprinted with permission from Stein et al. [1].
153
BLUK077-Stein
March 12, 2007
15:51
154
PART II
Diagnosis of DVT
Negative predictive value (%)
100 98
95
80 75
60 40 20 0 20
36 Prevalence of DVT (%)
References 1 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 2 Boneu B, Bes G, Pelzer H, Si´e P, Boccalon H. D-dimers, thrombin antithrombin III complexes and prothrombin fragments 1 + 2: diagnostic value in clinically suspected deep vein thrombosis. Thromb Haemost 1991; 65: 28–31. 3 Bounameaux H, Schneider PA, Reber G, de Moerloose P, Krahenbuhl B. Measurement of plasma D-dimer for diagnosis of deep venous thrombosis. Am J Clin Pathol 1989; 91: 82–85. 4 Chang-Liem GS, Lustermans FA, van Wersch JW. Comparison of the appropriateness of the latex and Elisa plasma D-dimer determination for the diagnosis of deep venous thrombosis. Haemostasis 1991; 21: 106–110. 5 Dale S, Gogstad GO, Brosstad F et al. Comparison of three D-dimer assays for the diagnosis of DVT: ELISA, latex and immunofiltration assay (NycoCard D-dimer). Thromb Haemost 1994; 71: 270–274. 6 Elias A, Aillaud F, Roul C et al. Assessment of D-dimer measurement by ELISA or latex methods in deep vein thrombosis diagnosed by ultrasonic duplex scanning. Fibrinolysis 1990; 4: 237–240. 7 Elias A, Aptel I, Huc B et al. D-dimer test and diagnosis of deep vein thrombosis: a comparative study of 7 assays. Thromb Haemost 1996; 76: 518–522. 8 Escoffre-Barbe M, Oger E, Leroyer C et al. Evaluation of a new rapid D-dimer assay for clinically suspected deep venous thrombosis (Liatest D-dimer). Am J Clin Pathol 1998; 109: 748–753. 9 Funfsinn N, Caliezi C, Biasiutti FD et al. Rapid D-dimer testing and pre-test clinical probability in the exclusion of deep venous thrombosis in symptomatic outpatients. Blood Coagul Fibrinolysis 2001; 12: 165–170.
78
Figure 31.2 Negative predictive value for deep venous thrombosis (DVT) of quantitative rapid ELISA shown in relation to prevalence of DVT in population studied. (Data from Stein et al. [1].)
10 Hansson PO, Eriksson H, Eriksson E, Jagenburg R, Lukes P, Risberg B. Can laboratory testing improve screening strategies for deep vein thrombosis at an emergency unit? J Intern Med 1994; 235: 143–151. 11 Larsen TB, Stoffersen E, Christensen CS, Laursen B. Validity of D-dimer tests in the diagnosis of deep vein thrombosis: a prospective comparative study of three quantitative assays. J Intern Med 2002; 252: 36–40. 12 Legnani C, Pancani C, Palareti G et al. Comparison of new rapid methods for D-dimer measurement to exclude deep vein thrombosis in symptomatic outpatients. Blood Coagul Fibrinolysis 1997; 8: 296–302. 13 Legnani C, Pancani C, Palareti G, Guazzaloca G, Coccheri S. Contribution of a new, rapid, quantitative and automated method for D-dimer measurement to exclude deep vein thrombosis in symptomatic outpatients. Blood Coagul Fibrinolysis 1999; 10: 69–74. 14 Lindahl TL, Lundahl TH, Ranby M, Fransson SG. Clinical evaluation of a diagnostic strategy for deep venous thrombosis with exclusion by low plasma levels of fibrin degradation product D-dimer. Scand J Clin Lab Invest 1998; 58: 307–316. 15 Mayer W, Hirschwehr R, Hippmann G, Odpadlik H, Bayer P, Partsch H. Whole-blood immunoassay (SimpliRED) versus plasma immunoassay (Nyco-Card) for the diagnosis of clinically suspected deep vein thrombosis. VASA 1997; 26: 97–101. 16 Sadouk M, Desmarais S, Patenaude JV, Lepage R. Comparison of diagnostic performance of three new fast Ddimer assays in the exclusion of deep vein thrombosis. Clin Chem 2000; 46: 286–287. 17 Shitrit D, Heyd J, Raveh D, Rudensky B. Diagnostic value of the D-dimer test in deep vein thrombosis: improved results by a new assay method and by using discriminate levels. Thromb Res 2001; 102: 125–131. 18 Speiser W, Mallek R, Koppensteiner R et al. D-dimer and TAT measurement in patients with deep venous thrombosis: utility in diagnosis and judgement of anticoagulant
BLUK077-Stein
March 12, 2007
CHAPTER 31
19
20
21
22
23
24
25
26
27
28
29
30
15:51
155
D-dimer for exclusion of acute DVT
treatment effectiveness. Thromb Haemost 1990; 64: 196– 201. Tengborn L, Palmblad S, Wojciechowski J, Peterson LE, Stigendal L. D-dimer and thrombin/antithrombin III complex—diagnostic tools in deep venous thrombosis? Haemostasis 1994; 24: 344–350. van der Graaf F, van den Borne H, van der Kolk M, de Wild PJ, Janssen GW, van Uum SH. Exclusion of deep venous thrombosis with D-dimer testing—comparison of 13 D-dimer methods in 99 outpatients suspected of deep venous thrombosis using venography as reference standard. Thromb Haemost 2000; 83: 191–198. W˚ahlander K, Tengborn L, Hellstr¨om M et al. Comparison of various D-dimer tests for the diagnosis of deep venous thrombosis. Blood Coagul Fibrinolysis 1999; 10: 121–126. Aschwanden M, Labs KH, Jeanneret C, Gehrig A, Jaeger KA. The value of rapid D-dimer testing combined with structured clinical evaluation for the diagnosis of deep vein thrombosis. J Vasc Surg 1999; 30: 929–935. Borg JY, Levesque H, Cailleux N, Franc C, Hellot MF, Courtois H. Rapid quantitative D-dimer assay and clinical evaluation for the diagnosis of clinically suspected deep vein thrombosis [Letter]. Thromb Haemost 1997; 77: 602– 603. Bradley M, Bladon J, Barker H. D-dimer assay for deep vein thrombosis: its role with colour Doppler sonography. Clin Radiol 2000; 55: 525–527. Bucek RA, Quehenberger P, Feliks I, Handler S, Reiter M, Minar E. Results of a new rapid D-dimer assay (cardiac D-dimer) in the diagnosis of deep vein thrombosis. Thromb Res 2001; 103: 17–23. Caliezi C, Funfsinn N, Mauron T et al. Performance of a new fibrin monomer assay to exclude deep vein thrombosis in symptomatic outpatients. Thromb Haemost 1999; 81: 50–53. D’Angelo A, D’Alessandro G, Tomassini L, Pittet JL, Dupuy G, Crippa L. Evaluation of a new rapid quantitative D-dimer assay in patients with clinically suspected deep vein thrombosis. Thromb Haemost 1996; 75: 412– 416. de Boer WA, de Haan MA, Huisman JW, Klaassen CH. Ddimer latex assay as screening method in suspected deep venous thrombosis of the leg. A clinical study and review of the literature. Neth J Med 1991; 38: 65–69. Farrell S, Hayes T, Shaw M. A negative SimpliRED Ddimer assay result does not exclude the diagnosis of deep vein thrombosis or pulmonary embolus in emergency department patients. Ann Emerg Med 2000; 35: 121– 125. Gavaud C, Ninet J, Ville D et al. [Diagnosis of venous thrombosis and/or pulmonary embolism by determination of D-dimers using ELISA. Review based on a study
31
32
33
34
35
36
37
38
39
40
41
42
of 80 consecutive patients hospitalized in an emergency unit]. J Mal Vasc 1996; 21: 22–30. Guazzaloca G, Palareti G, Legnani C et al. [Deep venous thrombosis. Evaluation of a non-invasive diagnostic procedure based on compression ultrasonography and measurement of D-dimer plasma levels]. Minerva Cardioangiol 1997; 45: 259–266. Harper P, Marson C, Grimmer A, Monahan K, Humm G, Baker B. The rapid whole blood agglutination D-dimer assay has poor sensitivity for use as an exclusion test in suspected deep vein thrombosis. N Z Med J 2001; 114: 61–64. Heijboer H, Ginsberg JS, Buller HR, Lensing AW, Colly LP, Wouter ten Cate J. The use of the D-dimer test in combination with non-invasive testing versus serial non-invasive testing alone for the diagnosis of deep-vein thrombosis. Thromb Haemost 1992; 67: 510–513. Ibrahim KM, O’Neill AI, Parkin DJ. D-dimer (latex) assay in the diagnosis of deep venous thrombosis [Letter]. Med J Aust 1992; 157: 574–575. Jacq F, Heron E, Rance A, Cesarini ML, Emmerich J, Fiessinger JN. [Evaluation of a test for rapid detection of D-dimers for the exclusion of the diagnosis of venous thrombosis]. Presse Med 1997; 26: 1132–1134. Jossang B, Runde I. [Diagnostic value of C-reactive protein and D-dimer in deep venous thrombosis]. Tidsskr Nor Laegeforen 1992; 112: 1153–1155. Khaira HS, Mann J. Plasma D-dimer measurement in patients with suspected DVT—a means of avoiding unnecessary venography. Eur J Vasc Endovasc Surg 1998; 15: 235–238. Knecht MF, Heinrich F. Clinical evaluation of an immunoturbidimetric D-dimer assay in the diagnostic procedure of deep vein thrombosis and pulmonary embolism. Thromb Res 1997; 88: 413–417. Kroneman H, Van Bergen PF, Knot EA, Jonker JJ, de Maat MP. Diagnostic value of D-dimer for deep venous thrombosis in outpatients. Haemostasis 1991; 21: 286–292. Legnani C, Pancani C, Palareti G, Guazzaloca G, Coccheri S. Performance of a new, fast D-dimer test (IL Test Ddimer) for the management of outpatients with suspected deep vein thrombosis in emergency situations. Fibrinolysis Proteolysis 1999; 13: 139–141. Lennox AF, Delis KT, Serunkuma S, Zarka ZA, Daskalopoulou SE, Nicolaides AN. Combination of a clinical risk assessment score and rapid whole blood D-dimer testing in the diagnosis of deep vein thrombosis in symptomatic patients. J Vasc Surg 1999; 30: 794–803. Leroyer C, Escoffre M, Le Moigne E et al. Diagnostic value of a new sensitive membrane based technique for instantaneous D-dimer evaluation in patients with clinically suspected deep venous thrombosis. Thromb Haemost 1997; 77: 637–640.
BLUK077-Stein
March 12, 2007
15:51
156
43 Lindahl TL, Lundahl TH, Fransson SG. Evaluation of an automated microlatex D-dimer assay (Tina-quant on Hitachi 911 analyser) in symptomatic outpatients with suspected DVT [Letter]. Thromb Haemost 1999; 82: 1772– 1773. 44 Permpikul C, Chantavatcharakorn K, Bouranasompop C, Chaiyasoot W. Whole blood agglutination D-dimer test for the diagnosis of deep vein thrombosis. J Med Assoc Thai 2000; 83: 732–736. 45 Pini M, Quintavalla R, Pattacini C et al. Combined use of strain-gauge plethysmography and latex D-dimer test in clinically suspected deep venous thrombosis. Fibrinolysis 1993; 7: 391–396. 46 ten Wolde M, Kraaijenhagen RA, Prins MH, Buller HR. The clinical usefulness of D-dimer testing in cancer patients with suspected deep venous thrombosis. Arch Intern Med 2002; 162: 1880–1884. 47 van Bergen PF, Knot EA, Jonker JJ, de Boer AC, de Maat MP. Is quantitative determination of fibrin(ogen) degradation products and thrombin–antithrombin III complexes useful to diagnose deep venous thrombosis in outpatients? Thromb Haemost 1989; 62: 1043–1045. 48 Wells PS, Anderson DR, Bormanis J et al. Application of a diagnostic clinical model for the management of hospitalized patients with suspected deep-vein thrombosis. Thromb Haemost 1999; 81: 493–497. 49 Wells PS, Brill-Edwards P, Stevens P et al. A novel and rapid whole-blood assay for D-dimer in patients with clinically suspected deep vein thrombosis. Circulation 1995; 91: 2184–2187. 50 Wijns W, Daoud N, Droeshout I et al. Evaluation of two D-dimer assays in the diagnosis of venous thromboembolism. Acta Clin Belg 1998; 53: 270–274. 51 Bouman CS, Ypma ST, Sybesma JP. Comparison of the efficacy of Ddimer, fibrin degradation products and prothrombin fragment 1–2 in clinically suspected deep venous thrombosis. Thromb Res 1995; 77: 225–234. 52 Brenner B, Pery M, Lanir N et al. Application of a bedside whole blood D-dimer assay in the diagnosis of deep vein thrombosis. Blood Coagul Fibrinolysis 1995; 6: 219–222. 53 Carter CJ, Doyle DL, Dawson N, Fowler S, Devine DV. Investigations into the clinical utility of latex D-dimer in the diagnosis of deep venous thrombosis. Thromb Haemost 1993; 69: 8–11, 24. 54 Carter CJ, Serrano K, Breen DJ, Buckley AR, Devine DV. Rapid fibrin D-dimer tests for deep venous thrombosis: factors affecting diagnostic utility. J Emerg Med 1999; 17: 605–610. 55 Chapman CS, Akhtar N, Campbell S, Miles K, O’Connor J, Mitchell VE. The use of D-dimer assay by enzyme immunoassay and latex agglutination techniques in the diagnosis of deep vein thrombosis. Clin Lab Haematol 1990; 12: 37–42.
PART II
Diagnosis of DVT
56 Heaton DC, Billings JD, Hickton CM. Assessment of Ddimer assays for the diagnosis of deep vein thrombosis. J Lab Clin Med 1987; 110: 588–591. 57 Janssen MC, Verbruggen BW, ter Hark PJ, Novakova IR. Factor VIIA determination compared to D-dimer in diagnosis of deep venous thrombosis. Thromb Res 1997; 86: 423–426. 58 Janssen MC, Heebels AE, de Metz M et al. Reliability of five rapid D-dimer assays compared to ELISA in the exclusion of deep venous thrombosis. Thromb Haemost 1997; 77: 262–266. 59 Killick SB, Pentek PG, Mercieca JE, Clarke MF, Bevan DH. Comparison of immunofiltration assay of plasma Ddimer with diagnostic imaging in deep vein thrombosis. Br J Haematol 1997; 96: 846–849. 60 Kozman H, Flemmer MC, Rahnama M. Deep venous thrombosis: prediction by D-dimer? South Med J 1997; 90: 907–910. 61 LaCapra S, Arkel YS, Ku DH, Gibson D, Lake C, Lam X. The use of thrombus precursor protein, D-dimer, prothrombin fragment 1.2, and thrombin antithrombin in the exclusion of proximal deep vein thrombosis and pulmonary embolism. Blood Coagul Fibrinolysis 2000; 11: 371–377. 62 Mauron T, Baumgartner I, Z’Brun A et al. SimpliRED D-dimer assay: comparability of capillary and citrated venous whole blood, between-assay variability, and performance of the test for exclusion of deep vein thrombosis in symptomatic outpatients [Letter]. Thromb Haemost 1998; 79: 1217–1219. 63 Ott P, Astrup L, Jensen RH, Nyeland B, Pedersen B. Assessment of D-dimer in plasma: diagnostic value in suspected deep venous thrombosis of the leg. Acta Med Scand 1988; 224: 263–267. 64 Pannocchia A, Chiappino I, Valpreda S et al. DVT exclusion in symptomatic outpatients: use of a rapid D-dimer immunofiltration assay and comparison with three enzyme immuno assays. Fibrinolysis 1995; 10: 117–119. 65 Rowbotham BJ, Carroll P, Whitaker AN et al. Measurement of crosslinked fibrin derivatives—use in the diagnosis of venous thrombosis. Thromb Haemost 1987; 57: 59–61. 66 Sassa H, Sone T, Tsuboi H, Kondo J, Yabashi T. Diagnostic significance of thrombin-antithrombin III complex (TAT) and D-dimer in patients with deep venous thrombosis. Jpn Circ J 1996; 60: 201–206. 67 Scarano L, Bernardi E, Prandoni P et al. Accuracy of two newly described D-dimer tests in patients with suspected deep venous thrombosis. Thromb Res 1997; 86: 93–99. 68 Siragusa S, Terulla V, Pirrelli S et al. A rapid D-dimer assay in patients presenting at the emergency room with suspected acute venous thrombosis: accuracy and relation to clinical variables. Haematologica 2001; 86: 856–861.
BLUK077-Stein
March 12, 2007
CHAPTER 31
15:51
D-dimer for exclusion of acute DVT
69 Trujillo-Santos AJ, Garcia de Lucas MD, Rios-Tamayo R, Jimenez-Puete A, Garcia-Sanchez JE. [Clinical and analytic diagnostic evaluation of deep venous thrombosis of the lower limbs.] Med Clin (Barc) 2000; 114: 46–49. 70 Wildberger JE, Vorwerk D, Kilbinger M, Lentner A, Wienert V, G¨unther RW. [The diagnosis of deep venous thromboses of the leg using a new rapid test (SimpliRED)]. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 1997; 167: 79–82. 71 Wildberger JE, Vorwerk D, Kilbinger M et al. Bedside testing (SimpliRED) in the diagnosis of deep vein thrombosis. Evaluation of 250 patients. Invest Radiol. 1998; 33: 232–235.
157
72 Jaeschke R, Guyatt GH, Sackett DL, for The EvidenceBased Medicine Working Group. Users’ guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? JAMA 1994; 271: 703–707. 73 Rathbun SW, Whitsett TL, Raskob GE. Negative Ddimer result to exclude recurrent deep venous thrombosis: a management trial. Ann Intern Med 2004; 141: 839– 845. 74 Lee AY, Julian JA, Levine MN et al. Clinical utility of a rapid whole-blood D-dimer assay in patients with cancer who present with suspected acute deep venous thrombosis. Ann Intern Med 1999; 131: 417–423.
BLUK077-Stein
March 11, 2007
32
7:13
CHAPTER 32
D-dimer combined with clinical probability assessment for exclusion of acute deep venous thrombosis
Introduction If a D-dimer assay is negative, deep venous thrombosis (DVT) will be ruled out in a high proportion of patients, providing clinical probability assessment also indicates that the likelihood of DVT is low [1, 2]. How-
ever, calculations based on negative likelihood ratios, according to Bayes theorem, show that DVT would often be present in patients with a discordantly high clinical assessment in combination with a negative Ddimer [2]. This holds true for all D-dimer assays [1] (Figure 32.1a). A high clinical probability, therefore,
DVT 100 90 80 70 60 50 40 30 20 10 0
Whole blood
Quant rapid ELISA
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 10
D-Dimer posttest probability (%)
(a)
Pretest probability (%)
DVT
12 10 8
Whole blood 6 4 Quant rapid ELISA
2
Pretest probability (%)
158
40
35
30
25
20
15
10
0 5
D-Dimer posttest probability (%)
(b)
Figure 32.1 (a) Relation of posttest probability of deep venous thrombosis (DVT) to pretest probability of DVT in patients with a negative D-dimer assay. (Data from Stein et al. [1].) (b) Enlargement of Figure 32.1a.
BLUK077-Stein
March 11, 2007
CHAPTER 32
7:13
D-dimer and clinical probability to exclude DVT
Table 32.1 Negative D-dimer and clinical assessment for DVT by Wells score. Proportion with First author [Ref]
D-dimer
DVT [n/N (%)]
Clinical assessment low probability (score ≤0) Likhanipour [6]
Rapid ELISA
0/60 (0)
Bates [7]
Quantitative latex
0/193 (0)
Kraaijenhagen [3]
Whole blood
10/561 (1.8)
Kearon [4]
Whole blood
1/177 (0.6)
Janes [5]
Whole blood
1/98 (1.0)
Clinical assessment unlikely (score <2) Wells [8]
Whole blood
Rapid ELISA
blood agglutination [3–5] and in none tested by rapid ELISA or quantitative latex [6, 7] (Table 32.1). With an unlikely clinical probability and negative whole blood agglutination D-dimer, DVT was present in 0.9% [8] (Table 32.1). With a moderate probability Wells score, DVT was found in 1.1–3.0% with a negative rapid ELISA or quantitative latex assay [6, 7] and DVT occurred in 5.8% with a negative whole blood agglutination assay [4] (Table 32.1). Results were highly variable with a high probability score and negative D-dimer assay [4, 7] (Table 32.1), but sparse data showed that DVT was not reliably excluded [4] (Table 32.1).
2/218 (0.9)
Clinical assessment moderate probability (score 1–2) Likhanipour [6]
159
2/67 (3.0)
Bates [7]
Quantitative latex
1/90 (1.1)
Kearon [4]
Whole blood
7/120 (5.8)
Clinical assessment high probability (score ≥3) Bates [7]
Quantitative latex
0/20 (0)
Kearon [4]
Whole blood
33/41 (80.5)
Likhanipour [6]
Rapid ELISA
0/19 (0)
DVT, deep venous thrombosis.
indicates a need for further testing even though the Ddimer assay is negative. Differences according to assay are apparent when clinical assessment indicates approximately a 20–30% likelihood of DVT, as may occur with an intermediate probability Wells score. This is illustrated in Figure 32.1b, which is an enlargement of Figure 32.1a. For example, if the pretest probability assessment is 25% and a rapid ELISA is negative, approximately 3% of patients would have DVT, whereas with a negative whole blood assay, approximately 5% would have DVT [1, 2] (Figure 32.1b). Among patients with a negative D-dimer assay and a low-probability clinical assessment, DVT on average was shown in 1.4% of patients tested by whole
References 1 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 2 Sox HC. Commentary. Ann Intern Med 2004; 140: 602. 3 Kraaijenhagen RA, Piovella F, Bernardi E et al. Simplification of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002; 162: 907–911. 4 Kearon C, Ginsberg JS, Douketis J et al. Management of suspected deep venous thrombosis in outpatients by using clinical assessment and D-dimer testing. Ann Intern Med 2001; 135: 108–111. 5 Janes S, Ashford N. Use of a simplified clinical scoring system and D-dimer testing can reduce the requirement for radiology in the exclusion of deep vein thrombosis by over 20%. Br J Haematol 2001; 112: 1079–1082. 6 Likhanipour K, Wolfson AB, Walker H et al. Combining clinical risk with D-dimer testing to rule out deep vein thrombosis. J Emerg Med 2004; 27: 233–239. 7 Bates SM, Kearon C, Crowther M et al. A diagnostic strategy involving a quantitative latex D-dimer assay reliably excludes deep venous thrombosis. Ann Intern Med 2003; 138: 787–794. 8 Wells PS, Anderson DR, Rodger M et al. Evaluation of Ddimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003; 349: 1227–1235.
BLUK077-Stein
March 11, 2007
33
7:14
CHAPTER 33
D-dimer and single negative compression ultrasound for exclusion of deep venous thrombosis
D-dimer alone (Chapter 31), combined with clinical assessment (Chapter 32), or combined with a single compression ultrasound examination of the lower extremities (sometimes with clinical assessment) have been used in various management algorithms. Bernardi and associates, among all patients with suspected deep venous thrombosis (DVT), showed DVT in 0.2% on 3-month follow-up after a normal rapid ELISA D-dimer and single negative venous ultrasound [1] (Table 33.1). Tick and associates, in patients with a high or moderate clinical assessment, negative whole blood agglutination test, and negative ultrasound showed no DVT on follow-up [2]. Oudega and associates, among patients with a lowprobability Wells clinical assessment for DVT, negative rapid ELISA, and single negative compression ultrasound showed DVT in 2.3% on 3-month follow-up [3]. The average rate of DVT on 3-month follow-up after a single negative compression ultrasound and negative
D-dimer was 6 of 1268 (0.5%) [1–3]. Clearly, however, the studies were heterogeneous with differences in the pretest probability of DVT and differences in the D-dimer assay used.
References 1 Bernardi E, Prandoni P, Lensing A et al. D-dimer testing as an adjunct to ultrasonography in patients with clinically suspected deep vein thrombosis: prospective cohort study. BMJ 1998; 17: 1037–1040. 2 Tick LW, Ton E, van Voorthuizen T et al. Practical diagnostic management of patients with clinically suspected deep vein thrombosis by clinical probability test, compression ultrasonography and D-dimer test. Am J Med 2002; 113: 630–635. 3 Oudega R, Hoes AW, Moons KGM. The Wells rule does not adequately rule out deep venous thrombosis in primary care patients. Ann Intern Med 2005; 143: 100–107.
Table 33.1 Negative D-dimer, single compression ultrasound, and clinical evaluation for DVT. Proportion with DVT: First author [Ref]
Clinical
D-dimer
3-month outcome (%)
Bernardi [1]
All suspected DVT
Rapid ELISA
1/598 (0.2)
Tick [2]
Moderate/high (≥1) Wells
Whole blood agglutination
0/148 (0)
Oudega [3]
Low (≤0) Wells
Rapid ELISA
5/222 (2.3)
DVT, deep venous thrombosis.
160
BLUK077-Stein
April 6, 2007
34
15:24
CHAPTER 34
Contrast venography
Introduction Contrast venography continues to be the diagnostic reference standard for deep venous thrombosis (DVT) [1], but by 1999 it was rarely performed [2] and by 2003 fewer than 5000 hospitalized patients in the United States had a contrast venogram (Stein unpublished) (Figure 34.1). For the evaluation of patients with suspected DVT, contrast venography has been replaced by compression ultrasound [2] and perhaps venous phase computed tomography (CT) when used in combination with contrast enhanced CT pulmonary angiography [3].
Method of contrast venography
125 100
US
75 50 25
Venograms
0 2003
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Number of patients (×103)
The procedure for contrast venography recommended in PIOPED II [4] was modified from Rabinov and Paulen [5]. The patient is placed in a semi-upright position (30–45◦ ). The patient stands on a box or rotating
Year Figure 34.1 Number of hospitalized patients in United States from 1979 to 2003 who had venous ultrasonography or contrast venography. By 2003, fewer than 5000 hospitalized patients in the United States had contrast venography. (Modified and reproduced from Stein et al. [2], with permission from American Medical Association. All rights reserved.)
platform with the other foot so that the injected leg is non-weight-bearing and completely relaxed. A 19–21gauge scalp vein needle or a plastic cannula is placed in a superficial vein of the foot, preferably in the superficial dorsal metatarsal vein of the great toe. A tourniquet is placed at the ankle. Low- or iso-osmolar nonionic contrast material is recommended at a concentration of 40–50% (200–250 mg/mL). The injection is done by hand using a 60 cc syringe. Power injectors are not recommended. The injection is performed under fluoroscopic monitoring at a rate of 0.5–1.0 cc per second. The amount of contrast material necessary for complete opacification of the veins is determined by direct fluoroscopic observation. Usually, 125–150 mL is required for each leg. The injection site should be observed fluoroscopically and/or visually in order that any extravasation can be recognized at once. Each segment of the venous system from the foot to the groin is recorded on spot films or digitally and/or on 14 × 17 films with an overhead tube at 40 inches target-film distance and Bucky film tray in various projections (at least two orthogonal views) for the calf veins and popliteal veins, and a single AP film for the thigh veins, pelvic veins, and IVC, so that optimal display of the venous pathways is accomplished without overlapping. Gentle palpation can be helpful to achieve complete opacification in local areas of imperfect filling. For the best radiographic detail, a small focal-spot tube (0.6 mm) and an automatic phototimer are recommended. After an initial bolus of approximately 50 cc of contrast material, 15–25 mL boluses of contrast material are injected prior to each overhead film. The examiner manually compresses the ipsilateral common femoral vein and the fluoroscopic table is turned into the horizontal or slight Trendelenburg position. The imaging field is positioned to cover the ipsilateral femoral vein, iliac vein, and the lower inferior vena. The ipsilateral leg is elevated as much as possible by the examiner and manual pressure on the femoral vein is released. A single AP image is obtained.
161
BLUK077-Stein
April 6, 2007
15:24
162
Diagnostic criteria Criteria for diagnosis of acute DVT by contrast venography that were agreed upon by PIOPED II investigators [4] were (1) a complete filling defect, i.e., failure to opacify the entire lumen due to a central filling defect (the vessel may enlarge compared to the opposite vein) and (2) a partial filling defect surrounded by contrast material. Criteria for chronic DVT are large collateral veins, webs, and synechiae (signs of recanalization).
Side-effects with low-osmolar, nonionic contrast material Severe discomfort among 282 patients who received high concentration standard contrast material, sodium and meglumine diatrizoates (Renografin 60) occurred in 9% and moderate discomfort occurred in 14% [6]. Postvenography thrombosis occurred in 9% [6]. New subclinical DVT was shown in 20 of 61 (33%) initially normal patients following venograms, based on substantially increased uptake of I-125 labeled fibrinogen after the procedure [7]. In 4 of 61 patients (7%), there was evidence of PE or DVT by ventilation–perfusion lung scan or repeat venography [7]. With Renografin 45, among 212 patients, severe pain was reduced to 2%, moderate pain reduced to 12%, and postvenography thrombosis reduced to 4% [6]. With nonionic lowosmolar contrast material (iopamidol), among 200 patients, severe and moderate pain was eliminated and postvenography thrombosis decreased to 2% or less [6]. All patients had mild pain with venography [6]. In another investigation with nonionic low-osmolar contrast material (iohexol), among 463 patients, local pain occurred on 12% and persisted for 1 week in 4% of patients [8]. Postvenography superficial phlebitis occurred in 0.6% and serious bronchospasm occurred in 0.4%. Altogether, minor side-effects including local pain, nausea, dizziness, skin reactions, vomiting, delayed edema, and superficial phlebitis occurred in 22% [8].
Adequacy of visualization with clinical venograms Clinical venograms showed adequate visualization of the veins of the thighs in 108 of 111 (97%) and adequate visualization of the veins of the calves in 109
PART II
Diagnosis of DVT
of 111 (98%) [6]. Some indicate, however, that filling is inadequate in as many as 5% of patients, and there is disagreement about the presence of thrombi in approximately 10% of patients [9].
Correlation of contrast venography with I-125-fibrinogen uptake Among patients with normal contrast venography of the calves, I-125-fibrinogen uptake showed DVT in 8 of 71 (11%) in one investigation [6] and 10 of 41 (24%) in another [8].
Correlation of postmortem contrast venography with dissection Postmortem contrast venography in 47 limbs compared with careful extensive dissection showed 87– 100% correlation with thrombi in the femoral vein, deep femoral vein, popliteal vein, and greater saphenous vein [10]. However, postmortem venography of the veins of the calves was an unreliable guide to the presence of thrombosis because of apparent filling defects and absence of vessels [10]. Postmortem venography had the advantage over clinical contrast venography of being able to inject an unlimited volume of contrast material. The potential problem of false positive filling defects caused by postmortem clots was rare. In the veins of the thighs, intraluminal filling defects, absence of vessels, dilated veins, enlarged valve pockets, and collateral circulation were readily shown.
References 1 Redman HC. Deep venous thrombosis: is contrast venography still the diagnostic “gold standard”? Radiology 1988; 168: 277–278. 2 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diagnosing deep venous thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 3 Cham MD, Yankelevitz DF, Shaham D et al., for The Pulmonary Angiography-Indirect CT Venography Cooperative Group. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216: 744–751.
BLUK077-Stein
April 6, 2007
CHAPTER 34
15:24
Contrast venography
4 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182. 5 Rabinov. K, Paulen S. Venography of the lower extremities. In: Abrams HL, ed. Abrams Angiography, 3rd edn. Little Brown & Co., Boston, Massachusetts, 1983: 1885– 1887. 6 Bettmann MA, Robbins A, Braun SD, Wetzner S, Dunnick NR, Finkelstein J. Contrast venography of the leg: diagnostic efficacy, tolerance, and complication rates with ionic and nonionic contrast media. Radiology 1987; 165: 113–116.
163
7 Albrechtsson U, Olsson CG. Thrombotic side-effects of lower-limb phlebography. Lancet 1976; 1: 723–724. 8 Lensing AW, Prandoni P, Buller HR, Casara D, Cogo A, ten Cate JW. Lower extremity venography with iohexol: results and complications. Radiology 1990; 177: 503– 505. 9 Baldt MM, Zontsich T, Stumpflen A et al. Deep venous thrombosis of the lower extremity: efficacy of spiral CT venography compared with conventional venography in diagnosis. Radiology 1996; 200: 423–428. 10 Stein PD, Evans H. An autopsy study of leg vein thrombosis. Circulation 1967; 35: 671–681.
BLUK077-Stein
April 6, 2007
35
15:28
CHAPTER 35
Compression ultrasound for the diagnosis of deep venous thrombosis
Definitions Conventional gray-scale real-time ultrasonography, as the name indicates, gives a gray-scale real-time (Bmode) two-dimensional image of the vein. The image is obtained by real-time computation of the reflected signals from an array of ultrasound sources. The reflections are due to boundaries between adjacent structures with different acoustic properties [1]. Duplex ultrasonography is the combined use of conventional gray-scale real-time ultrasonography with pulsed Doppler technology. The combination permits simultaneous imaging of anatomic structures and audible or visual characterization of venous and arterial blood flow [1]. Color duplex ultrasonography gives Doppler shift information from moving erythrocytes. Color coding shows both velocity and direction of flow. This is superimposed in the real-time gray-scale image [1].
Equipment preference As long as compression is used, a clear advantage of one system over another has not been shown in prospective clinical trials [1]. Even so, nonocclusive thrombi may be more easily documented with color flow imaging, and studies of obese patients are generally more easily achieved with this technique [1].
Diagnostic criteria for acute DVT Noncompressibility of the vein is the most reliable sign of acute deep venous thrombosis (DVT) [1]. In Prospective Investigation of Pulmonary Embolism
164
Diagnosis II (PIOPED II), color duplex imaging of the common femoral, superficial femoral, popliteal, and proximal greater saphenous veins was used [2]. Acute DVT was diagnosed if there was noncompressibility of the vein in combination with at least one of the following: an enlarged vein, a hypoechoic lumen, or absence of collaterals [2]. Usually, an enlarged vein is found in combination with noncompressibility (T. Wakefield, personal communication, 2006). Some use, as a secondary diagnostic criterion, an echogenic thrombus within the vein lumen [1]. This may be false positive, however, and acute thrombi may be anechoic [1]. Loss of flow phasicity, response to Valsalva, and augmentation have also been cited as secondary diagnostic criteria [1]. An abbreviated imaging study of only the common femoral veins and popliteal veins but not the superficial femoral veins has been used by some to conserve time and resources. However, among 269 patients with DVT identified by compression ultrasound, the DVT was isolated to the superficial femoral vein in 60 patients (22%) [3]. The remaining patients showed DVT in the common femoral or popliteal veins or both [3]. Abbreviated imaging studies, therefore, may fail to identify DVT in a significant number of patients.
Accuracy of compression ultrasound Symptomatic DVT In patients with clinically suspected symptomatic acute DVT, compression ultrasound of the proximal deep veins is highly sensitive and specific. In prospective investigations of symptomatic patients with suspected DVT, using contrast venography as a reference
BLUK077-Stein
April 6, 2007
CHAPTER 35
15:28
165
Compression ultrasound for diagnosis of DVT
Symptomatic High risk asymptomatic
100 97 97
94
96 96
97
Figure 35.1 Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of compression ultrasound in symptomatic patients with deep venous thrombosis (DVT) and in high-risk asymptomatic patients with DVT. Data are pooled data described in text.
Percent
80 60
68
63
40 20 0 Sensitivity
standard with blinded readers, sensitivity with realtime B-mode ultrasound [4–6], duplex ultrasound [7, 8], and color duplex ultrasound [9, 10], the respective sensitivities were 192 of 207 (93%), 120 of 126 (95%), and 45 of 47 (96%). Specificities were 293 of 299 (98%), 111 of 120 (93%), and 124 of 125 (99%). All these patients had ultrasonic examination of the common femoral vein, superficial femoral, and popliteal vein. On average, using any of these three methods with compression ultrasound (recognizing heterogeneity of the methods, so these averages are estimates), sensitivity was 94%, specificity 97%, positive predictive value 97%, and negative predictive value 96% (Figure 35.1).
Asymptomatic DVT In asymptomatic patients with a high risk for DVT, sensitivity 62 of 99 (63%) and positive predictive value 62 of 91 (68%) of compression ultrasound were lower than in symptomatic patients (Figure 35.1) [11–17]. Similar values to those in symptomatic patients, however, were shown with specificity 1024 of 1060 (97%) and negative predictive value 1001 of 1038 (96%).
Calf veins The sensitivity of compression ultrasound for isolated DVT of the calf veins, in a limited number of symptomatic patients, was 92 of 112 (82%) [6, 10, 18]. For this reason, serial studies to examine for extension to proximal veins are recommended [1]. In the event that serial studies cannot be obtained, contrast venography has been recommended [1].
Specificity
PPV
NPV
The sensitivity of compression ultrasound for isolated DVT of the calf veins, in a small number of asymptomatic patients, was low, 36 of 58 (62%) [14, 17, 19].
Compression ultrasound in patients with acute pulmonary embolism Among 149 patients with acute pulmonary embolism (PE) diagnosed by a high-probability lung scan or positive pulmonary angiogram, bilateral B-mode grayscale compression ultrasonography of the proximal leg veins showed DVT in 43 of 149 (29%) [20]. Others showed DVT by B-mode compression ultrasound in 70 of 162 patients with PE (43%) diagnosed by a high-probability ventilation–perfusion lung scan and in 35 of 60 (58%) diagnosed by a high-probability ventilation–perfusion lung scan in combination with a high clinical probability [21]. Autopsies have shown that 80% or more of patients with PE have thrombi that originate in the lower extremities [22–25]. Even so, only 11% of patients with acute PE had clinically apparent DVT in PIOPED [26] and 47% in PIOPED II [27]. Therefore, the limited sensitivity of compression ultrasound for the detection of DVT in patients with acute PE may partially reflect the fact that the patients are asymptomatic from DVT. Among 178 patients in whom PE was excluded, compression ultrasound showed DVT in 5 of 178 (3%) [20]. Others showed a lower prevalence of coincidental DVT, 2 of 334 (0.6%) [21].
BLUK077-Stein
April 6, 2007
15:28
166
Outcome studies of untreated patients with suspected DVT and negative compression ultrasound: conversion to positive after 1 week Among outpatients with suspected DVT who had negative initial real-time B-mode ultrasonography and were untreated, 2% showed DVT on days 2–8 [28– 31]. Detailed results were 7 of 342 (2%) [28], 12 of 1302 (0.9%) [29], 0 of 118 (0%) [30], and 6 of 413 (1.5%) [31]. One of the 12 patients evaluated by Cogo and associates died of massive PE during the first week after a negative ultrasound [29].
Outcome of patients with suspected DVT and negative ultrasound after repeat testing Patients with suspected DVT were followed off treatment for 3 months after repeat testing with ultrasound in ≤8 days of the first negative test [28–31]. Venous thromboembolic events occurred in 2 of 335 (0.6%) [28], 8 of 1290 (0.6%) [29], and 6 of 378 (1.6%) [30]. Among 118 outpatients with suspected DVT, 1.3% of the negative ultrasound patients developed PE or DVT in 3 months [31]. Results in inpatients were quite different, however. Among 56 inpatients with suspected DVT, 10% of the ultrasound negative patients developed thromboembolic complications within 3 months [31]. It has been shown that compression ultrasonography repeated only once at 5–7 days, and if negative, permits a safe withholding of anticoagulants in patients with suspected DVT [28].
Outcome studies of compression ultrasound and clinical score See Chapter 30.
Outcome studies of compression ultrasound and D-dimer See Chapter 33.
Comparison of compression ultrasound and CT venous phase imaging for DVT There was 95.5% concordance between CT venous phase imaging and compression ultrasound [32].
PART II
Diagnosis of DVT
Kappa statistic was 0.809 (“almost perfect”). Among patients with PE by the reference test the proportion with a positive CT venogram and positive venous ultrasound was nearly the same and showed no statistically significant difference (32).
References 1 Tapson VF, Carroll BA, Davidson BL et al., for the American Thoracic Society. The diagnostic approach to acute venous thromboembolism. Clinical practice guideline. Am J Respir Crit Care Med 1999; 160: 1043–1066. 2 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173– 182. 3 Maki DD, Kumar N, Nguyen B, Langer JE, Miller WT, Jr, Gefter WB. Distribution of thrombi in acute lower extremity deep venous thrombosis: implications for sonography and CT and MR venography. Am J Roentgenol 2000; 175: 1299–1301. 4 Cronan JJ, Dorfman GS, Scola FH, Schepps B, Alexander J. Deep venous thrombosis: US assessment using vein compression. Radiology 1987; 162: 191–194. 5 Pedersen OM, Aslaksen A, Vik-Mo H, Bassoe AM. Compression ultrasonography in hospitalized patients with suspected deep venous thrombosis. Arch Intern Med 1991; 151: 2217–2220. 6 Lensing AW, Prandoni P, Brandjes D et al. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med 1989; 320: 342–345. 7 O’Leary DH, Kane RA, Chase BM. A prospective study of the efficacy of B-scan sonography in the detection of deep venous thrombosis in the lower extremities. J Clin Ultrasound 1988; 16: 1–8. 8 Mitchell DC, Grasty MS, Stebbings WS et al. Comparison of duplex ultrasonography and venography in the diagnosis of deep venous thrombosis. Br J Surg 1991; 78: 611–613. 9 Lewis BD, James EM, Welch TJ, Joyce JW, Hallett JW, Weaver AL. Diagnosis of acute deep venous thrombosis of the lower extremities: prospective evaluation of color Doppler flow imaging versus venography. Radiology 1994; 192: 651–655. 10 Rose SC, Zwiebel WJ, Nelson BD et al. Symptomatic lower extremity deep venous thrombosis: accuracy, limitations, and role of color duplex flow imaging in diagnosis. Radiology 1990; 175: 639–644. 11 Froehlich JA, Dorfman GS, Cronan JJ, Urbanek PJ, Herndon JH, Aaron RK. Compression ultrasonography for the detection of deep venous thrombosis in patients who have a fracture of the hip. A prospective study. J Bone Joint Surg Am 1989; 71: 249–256.
BLUK077-Stein
April 6, 2007
CHAPTER 35
15:28
Compression ultrasound for diagnosis of DVT
12 Barnes CL, Nelson CL, Nix ML, McCowan TC, Lavender RC, Barnes RW. Duplex scanning versus venography as a screening examination in total hip arthroplasty patients. Clin Orthop Relat Res 1991; 271: 180–189. 13 Woolson ST, Pottorff G. Venous ultrasonography in the detection of proximal vein thrombosis after total knee arthroplasty. Clin Orthop Relat Res 1991; 273: 131–135. 14 Elliott CG, Suchyta M, Rose SC et al. Duplex ultrasonography for the detection of deep vein thrombi after total hip or knee arthroplasty. Angiology 1993; 44: 26–33. 15 Davidson BL, Elliott CG, Lensing AW, for the RD Heparin Arthroplasty Group. Low accuracy of color Doppler ultrasound in the detection of proximal leg vein thrombosis in asymptomatic high-risk patients. Ann Intern Med 1992; 117: 735–738. 16 Mattos MA, Londrey GL, Leutz DW et al. Color-flow duplex scanning for the surveillance and diagnosis of acute deep venous thrombosis. J Vasc Surg 1992; 15: 366–375; discussion 375–376. 17 Lensing AW, Doris CI, McGrath FP et al. A comparison of compression ultrasound with color Doppler ultrasound for the diagnosis of symptomless postoperative deep vein thrombosis. Arch Intern Med 1997; 157: 765–768. 18 Habscheid W, Hohmann M, Wilhelm T, Epping J. Realtime ultrasound in the diagnosis of acute deep venous thrombosis of the lower extremity. Angiology 1990; 41: 599–608. 19 Rose SC, Zwiebel WJ, Murdock LE et al. Insensitivity of color Doppler flow imaging for detection of acute calf deep venous thrombosis in asymptomatic postoperative patients. J Vasc Interv Radiol 1993; 4: 111–117. 20 Turkstra F, Kuijer PM, van Beek EJ, Brandjes DP, ten Cate JW, Buller HR. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781. 21 Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. 22 Sevitt S, Gallagher N. Venous thrombosis and pulmonary embolism: a clinico-pathological study in injured and burned patients. Br J Surg 1961; 48: 475–489.
167
23 Cohn R, Walsh J. The incidence and anatomical site of origin of pulmonary emboli. Stanford Med Bull 1946; 4: 97–99. 24 Short DS. A survey of pulmonary embolism in a general hospital. BMJ 1952; 1: 790–796. 25 Byrne JJ, O’Neil EE. Fatal pulmonary emboli. A study of 130 autopsy-proven fatal emboli. Am J Surg 1952; 83: 47–49. 26 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 27 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med 2007 (In press). 28 Birdwell BG, Raskob GE, Whitsett TL et al. The clinical validity of normal compression ultrasonography in outpatients suspected of having deep venous thrombosis. Ann Intern Med 1998; 128: 1–7. 29 Cogo A, Lensing AW, Koopman MM et al. Compression ultrasonography for diagnostic management of patients with clinically suspected deep vein thrombosis: prospective cohort study. BMJ 1998; 316: 17–20. 30 Heijboer H, Buller HR, Lensing AW, Turpie AG, Colly LP, ten Cate JW. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N Engl J Med 1993; 329: 1365– 1369. 31 Sluzewski M, Koopman MM, Schuur KH, van Vroonhoven TJ, Ruijs JH. Influence of negative ultrasound findings on the management of in- and outpatients with suspected deep-vein thrombosis. Eur J Radiol 1991; 13: 174–177. 32 Goodman LG, Stein PD, Beemath A, Sostman HD, Wakefield TW, Woodard PK, Yankelevitz DF. Equivalence of venous phase CT venograms and compression ultrasound in the evaluation of patients with suspected acute pulmonary embolism: data from the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II). (Submitted for publication).
BLUK077-Stein
March 12, 2007
36
18:9
CHAPTER 36
Impedance plethysmography and fibrinogen uptake tests for diagnosis of deep venous thrombosis
Impedance plethysmography Impedance plethysmography (IPG) has been largely replaced by compression ultrasonography for the diagnosis of deep venous thrombosis (DVT). Data from the National Hospital Discharge Survey show that from 1990 through 2003 fewer than 5000 impedance plethysmograms were coded at hospital discharge of all patients in the United States (Stein PD, unpublished). Impedance plethysmography has the advantage of being less expensive than other diagnostic tests for DVT; it is portable and it requires less technician training [1]. Even so, it is operator dependent, and requires trained technicians and adherence to a validated protocol [2]. Impedance plethysmography is sensitive for proximal DVT, but not distal DVT. Only 17–25% of limbs with isolated calf DVT showed an abnormality with IPG [3, 4]. Among patients with DVT in the proximal veins shown by venography, 86% (30 of 35) had a positive impedance plethysmogram [5]. Others showed that impedance plethysmography was indicative of DVT in 15 of 16 (94%) of legs that showed positive venograms of both the proximal and distal leg veins [4]. Hull and associates showed that 124 of 133 limbs with DVT (93%) were identified by IPG and 386 of 397 limbs without DVT (97%) showed a negative IPG [3]. One study of IPG, however, showed a sensitivity of only 26 of 40 (65%) and specificity of 79 of 85 (93%) for proximal DVT [6]. This may have reflected referral of patients with less severe symptoms and smaller, less occlusive thrombi [6]. On average, based on 10 investigations, IPG for proximal DVT showed a sensitivity
168
of 650 of 738 (88%) and specificity of 330 of 354 (93%) [7]. Outcome study has shown that only 9 of 365 (2.5%) of patients with suspected DVT and negative serial impedance plethysmograms over a period of 8 days had symptoms of PE (pulmonary embolism) or DVT in 6 months [8]. Although IPG detects proximal DVT in approximately 88% of patients, noninvasive testing for DVT in patients with documented PE is positive in only 43– 57% of the patients [5, 9]. This may reflect the following possibilities: (1) The residual DVT may be nonobstructive or confined to the calf and remain undetected. (2) The deep venous thrombi causing the pulmonary emboli may embolize entirely from the deep veins of the thighs, leaving no residual thrombosis to detect. (3) The source of the pulmonary emboli may be from sites other than the lower extremities (for example, deep pelvic veins, renal veins, and subclavian veins). Comparison of IPG with compression ultrasound has shown the latter to be more accurate [10]. Among symptomatic patients with clinically suspected DVT, the sensitivity of IPG was 100 of 130 (77%) versus 117 of 130 (90%) with compression ultrasound. The specificities were 339 of 365 (93%) with IPG and 358 of 365 (98%) with compression ultrasound.
Outcome of patients with suspected DVT who had a single negative impedance plethysmogram Only 1 of 284 patients with suspected DVT (0.4%) and a negative impedance plethysmogram developed DVT on follow-up [11]. None developed PE.
BLUK077-Stein
March 12, 2007
CHAPTER 36
18:9
Impedance plethysmography and fibrinogen uptake tests
Outcome of patients with suspected DVT who had negative serial impedance plethysmograms Among 139 pregnant women with suspected DVT and negative serial impedance plethysmograms obtained over 14 days, none developed DVT or PE within a 3month follow-up period [12]. Among 289 patients with suspected DVT who had negative serial impedance plethysmograms obtained over 10 days, none developed PE within a 6-month follow-up period after all serial tests were completed [13]. One patient, however, may have had a minor PE during the 10-day period of serial testing [13]. One patient (0.3%) had DVT 9 weeks after the initial evaluation.
7
8
9
10
Fibrinogen uptake test Fibrinogen leg scanning, also called the fibrinogen uptake test, had been used extensively to detect subclinical DVT [14]. The test is no longer available because of concerns about the potential for viral transmission with this human blood product [15]. Geerts et al. [15] indicated that the fibrinogen uptake test lacks both specificity and sensitivity for the detection of DVT [16–21] and is poorly correlated with major thromboembolic events [22].
References 1 Tapson VF, Carroll BA, Davidson BL et al., for the American Thoracic Society. The diagnostic approach to acute venous thromboembolism. Clinical practice guideline. Am J Respir Crit Care Med 1999; 160: 1043–1066. 2 Taylor DW, Hull R, Sackett DL, Hirsh J. Simplification of the sequential impedance plethysmograph technique without loss of accuracy. Thromb Res 1980; 17: 561–565. 3 Hull R, van Aken WG, Hirsh J et al. Impedance plethysmography using the occlusive cuff technique in the diagnosis of venous thrombosis. Circulation 1976; 53: 696– 700. 4 Moser KM, LeMoine JR. Is embolic risk conditioned by location of deep venous thrombosis? Ann Intern Med 1981; 94: 439–444. 5 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. 6 Ginsberg JS, Wells PS, Hirsh J et al. Reevaluation of the sensitivity of impedance plethysmography for the
11
12
13
14 15
16
17
18
19
169
detection of proximal deep vein thrombosis. Arch Intern Med 1994; 154: 1930–1933. Kraaijenhagen RA, Lensing AW, Wallis JW, van Beek EJ, ten Cate JW, Buller HR. Diagnostic management of venous thromboembolism. Baillieres Clin Haematol 1998; 11: 541–586. Heijboer H, Buller HR, Lensing AW, Turpie AG, Colly LP, ten Cate JW. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N Engl J Med 1993; 329: 1365–1369. Hull RD, Hirsh J, Carter CJ et al. Diagnostic value of ventilation–perfusion lung scanning in patients with suspected pulmonary embolism. Chest 1985; 88: 819–828. Wells PS, Hirsh J, Anderson DR et al. Comparison of the accuracy of impedance plethysmography and compression ultrasonography in outpatients with clinically suspected deep vein thrombosis. A two centre paireddesign prospective trial. Thromb Haemost 1995; 74: 1423– 1427. Wheeler HB, Anderson FA, Jr, Cardullo PA, Patwardhan NA, Jian-Ming L, Cutler BS. Suspected deep vein thrombosis. Management by impedance plethysmography. Arch Surg 1982; 117: 1206–1209. Hull RD, Rascob GE, Carter CJ. Serial impedance plethysmography in pregnant patients with clinically suspected deep-vein thrombosis. Clinical validity of negative findings. Ann Intern Med 1990; 112: 663–667. Huisman MV, Buller HR, ten Cate JW, Vreeken J. Serial impedance plethysmography for suspected deep venous thrombosis in outpatients: The Amsterdam General Practitioner Study. N Engl J Med 1986; 314: 823–828. Kakkar VV. The diagnosis of deep vein thrombosis using the 125 I fibrinogen test. Arch Surg 1972; 104: 152–159. Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 338S–400S. Moskovitz PA, Ellenberg SS, Feffer HL et al. Low-dose heparin for prevention of venous thromboembolism in total hip arthroplasty and surgical repair of hip fractures. J Bone Joint Surg Am 1978; 60: 1065–1070. Fauno P, Suomalainen O, Bergqvist D et al. The use of fibrinogen uptake test in screening for deep vein thrombosis in patients with hip fracture. Thromb Res 1990; 60: 185–190. Lensing AW, Hirsh J. 125 I-fibrinogen leg scanning: reassessment of its role for the diagnosis of venous thrombosis in post-operative patients. Thromb Haemost 1993; 69: 2–7. Agnelli G, Radicchia S, Nenci GG. Diagnosis of deep vein thrombosis in asymptomatic high-risk patients. Haemostasis 1995; 25: 40–48.
BLUK077-Stein
March 12, 2007
18:9
170
20 Bergqvist D, Burmark US, Flordal PA et al. Low molecular weight heparin started before surgery as prophylaxis against deep vein thrombosis: 2500 versus 5000 XaI units in 2070 patients. Br J Surg 1995; 82: 496–501. 21 Nurmohamed MT, Verhaeghe R, Haas S, et al. A comparative trial of a low molecular weight heparin (enoxaparin) versus standard heparin for the; prophylaxis of postope-
PART II
Diagnosis of DVT
rative deep vein thrombosis in general surgery. Am J Surg 1995 169: 567–571. 22 Flordal PA, Bergqvist D, Ljungstrom KG et al. Clinical relevance of the fibrinogen uptake test in patients undergoing elective general abdominal surgery: relation to major thromboembolism and mortality. Fragmin Multicentre Study Group. Thromb Res 1995; 80: 491–497.
BLUK077-Stein
April 14, 2007
37
14:54
CHAPTER 37
Computed tomography for diagnosis of deep venous thrombosis
Ascending venography Spiral CT ascending venography compared with conventional contrast venography In an evaluation of spiral CT venography (CTV) compared with conventional contrast venography combined with duplex ultrasonography and follow-up, 52 patients with suspected deep venous thrombosis (DVT) were evaluated [1]. Sensitivity was 100%. Regarding results according to limb and site, sensitivity in calf veins was 36 of 36, thigh veins 32 of 32, and pelvic veins 13 of 13. Specificities in limbs were calf veins 64 of 68 (94%), thigh veins 68 of 70 (97%), and thigh veins 88 of 90 (98%). False-positive results with CTV may have been caused by false-negative results of conventional venography [1]. Spiral CTV more clearly demonstrated extension of DVT into the pelvic veins and inferior vena cava than conventional contrast venography [1]. In another investigation of 102 lower extremities compared with ascending venography and color duplex ultrasonography, sensitivity was 100% and specificity was 96% [2]. The quality of CTV was superior to ascending venography in all veins [2]. Methods of ascending CTV The method of ascending CTV used by Baldt and associates is as follows: A 22-gauge cannula was inserted in a dorsal vein of both feet and a tourniquet was placed around both ankles [1]. Forty milliliters of iohexol diluted with 200 mL of saline was injected via a Y adapter simultaneously into both legs at 4 mL/s using a power injector. This achieved a flow of 2 mL/s for both legs. After a 35-second delay, a 100-cm section was imaged from the ankle to the inferior vena cava. Immediately
after the study, 100 mL of saline was injected into both cannulas to flush contrast material from the veins. A single slice scanner was used. From the data set, 200 images with a nominal thickness of 5 mm were reconstructed.
Complications Complications of ascending contrast venography are described in Chapter 34. Complications relate to local and systemic effects of contrast material, and presumably the complications rate are the same with CTV and ascending contrast venography.
Venous phase CTV Validity of CT venous phase imaging in patients with suspected acute pulmonary embolism On average, CTV of the veins of the calf, thigh, and pelvis in patients with suspected PE (pulmonary embolism) showed a sensitivity of 34 of 37 (92%) and specificity of 71 of 82 (87%) using ascending venography as a reference standard (Table 37.1) [3, 4]. Venous phase imaging limited to the veins of the thighs showed an average sensitivity of 83 of 85 (98%) and average specificity of 399 of 405 (99%) using compression ultrasonography as a reference standard (Table 37.1) [5–7]. With an average prevalence of DVT of 17% in these patients, positive predictive value was 83 of 89 (93%) and negative predictive value was 399 of 401 (99.5%) (Figure 37.1) [5–7]. Venous phase imaging of the veins of the thighs also was sensitive and specific based on Doppler sonography as the reference standard [8] (Table 37.1). Technical quality of CTV, among 541 patients, was excellent in 47%, good in 29%, fair in 17%, and poor in 6% [7]. 171
BLUK077-Stein
April 14, 2007
14:54
172
PART II
Diagnosis of DVT
Table 37.1 Sensitivity and specificity of spiral CT venous phase imaging in patients with suspected pulmonary embolism. n/N (%) First author (Ref)
Number slice CT
Sites evaluated
Reference test
Sensitivity
Coche [3]
2
Mid-calf, thigh, pelvis
Venogram
16/17 (94)
Ghaye [4]
1
Calf, thigh, pelvis
Venogram
18/20 (90)
17/26 (65)
Ghaye [4]
1
Calf, thigh, pelvis
Compression ultrasound
82/84 (98)
81/93 (87)
Garg [5]
1
Thigh
Compression ultrasound
Loud [6]
1
Thigh
Compression ultrasound
63/65 (97)
243/243 (100)
Cham [7]
1, 2, 4
Thigh
Compression ultrasound
15/15 (100)
93/97 (96)
Begemann [8]
4
Thigh
Doppler sonography
11/11 (100)
28/29 (97)
Methods The methods for CT venous phase imaging of the veins of the pelvis and lower extremities used in the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II) were as follows [9]. An injection of 135 mL of low osmolar nonionic contrast material (300– 320 mg iodine/mL) was made into an arm vein at 4 mL/s. Scans of the iliac, femoral, and popliteal veins were obtained 3 minutes after the beginning of the injection of contrast material for chest CT. Scans commenced at the iliac crest and continued in a helical mode to the level of the knees. Veins of the calves were not studied because CT venous phase imaging usually fails to opacify the calf veins. Diagnostic criteria for DVT The diagnostic criteria for acute DVT based on venous phase imaging in the PIOPED II [9] were: Occlusive = complete filling defect, i.e., failure to opacify the entire lumen due to a central filling defect (the vessel may enlarge compared to the opposite vein); Nonocclusive = partial filling defect surrounded by contrast material.
5/5 (100)
Specificity 54/56 (96)
63/65 (97)
The criteria for chronic DVT are not well worked out in the CT literature, but are well established in the venous compression ultrasound literature. The following criteria were designated as DVT of unknown age: a Complete filling defect, but the vein is smaller than its peers; b Contrast flowing through a thickened, often smaller vein; c Calcification in vein; d Secondary sign, i.e., increased collateral vessels.
Increased sensitivity for venous thromboembolism using venous phase imaging in patients with suspected acute PE In patients with suspected acute PE, the number of patients with only a positive contrast-enhanced chest CT, positive chest CT and positive CTV, and only a positive CTV is shown in Table 37.2 [3, 5–8, 10–15]. Images were obtained with 1-slice, 2-slice, 4-slice, 8-slice, and 16-slice CT (Table 37.2). Pooled data showed that 647 patients with suspected acute PE had a positive chest CT and an additional 121 patients (19%) had only a
100 99.5
95 Percent
90
93
92
85
87
80 75 70 SENS
SPEC
PPV
NPV
Figure 37.1 Sensitivity (sens), specificity (spec), positive predictive value (PPV), and negative predictive value (NPV) of venous phase imaging (CTV) of veins of thighs using conventional contrast venography as a reference standard. Data are pooled data described in text.
BLUK077-Stein
April 14, 2007
CHAPTER 37
14:54
173
Computed tomography for diagnosis of DVT
Table 37.2 Distribution of positive (POS) findings on contrast enhanced chest CT and venous phase imaging (CTV) in patients with suspected pulmonary embolism. POS CT
POS CT chest
POS CT chest
First author (Ref)
Number slice CT
chest only
and CTV
± POS CTV
POS CTV only
POS CTV only/VTE (%)*
Richman [10]
4
31
42
88
15
15/88 (17)
Coche [3]
2
9
13
25
3
3/25 (12)
Loud [6]
1
27
58
116
31
31/116(27)
Garg [5]
1
7
5
12
0
0/12 (0)
Cham [7]
1, 2, 4
62
29
107
16
16/107(15)
Revel [11]
4
33
21
59
5
5/59 (8)
Nicolas [12]
1
18
29
52
5
5/52 (10)
Begemann [8]
4
7
13
31
11
11/31 (35)
Stein [13]
4, 8, 16
71
79
164
14
14/183 (8)
Ghaye [14]
16
15
38
73
20
20/73 (27)
Johnson [15]
4, 8
30
10
41
1
1/41 (2)
CT = computed tomography, POS = positive, Neg = negative, CTV = CT venogram, Anglo = angiogram, VTE = venous thromboembolic disease.
positive CTV [3, 5–8, 10–15] (Figure 37.2). Among patients in whom single slice CT was used for the diagnosis of suspected PE, 36 of 180 (20%) patients with VTE were diagnosed on the basis of a positive CTV [5, 6, 12]. Among patients with suspected acute PE who were shown to have venous thromboembolism by multidetector contrast-enhanced spiral CT, the diagnosis was made by a positive CTV in 66 of 456 (14%) [8, 10, 11, 13–15]. In PIOPED II, CTV’s were obtained using continuous helical imaging with 7.5 mm collimation and 7.5 mm reconstruction (14). Subsequently, a random sample of the original data set was examined after every two of three 7.5 mm images were deleted. The refor-
matted cases, therefore, consisted of 7.5 mm images with a 15 mm gap (14). Per patient agreement with continuous images for DVT was 89%. Use of 5 mm axial images, skipping 15 mm, should yield similar results (14). With 5 mm, skip 15 mm images, plus automated dose modulation, the calculated radiation dose dropped from 9.1 to 3 mSv. In PIOPED II, thrombi isolated to the inferior vena cava and iliac vessels occurred in only 3% of patients (13). Therefore, scanning from the acetabulum to the tibial plateau, rather than from the iliac crest, would further reduce the radiation to 2.0 mSv (16), Continuous helical CTVs and compression ultrasound showed similar results (17) (See Chapter 35).
100 84
VTE (%)
80 60 40 16
20 Figure 37.2 Patients with suspected pulmonary embolism in whom venous thromboembolism (VTE) was diagnosed by multidetector CT.
0 Positive Chest CT
Positive CTV Only
BLUK077-Stein
April 14, 2007
14:54
174
References 1 Baldt MM, Zontsich T, Stumpflen A et al. Deep venous thrombosis of the lower extremity: efficacy of spiral CT venography compared with conventional venography in diagnosis. Radiology 1996; 200: 423–428. 2 Zontsich T, Turetschek K, Baldt M. CT-phlebography. A new method for the diagnosis of venous thrombosis of the upper and lower extremities. Radiologe 1998; 38: 586–590. 3 Coche EE, Hamoir XL, Hammer FD, Hainaut P, Goffette PP. Using dual-detector helical CT angiography to detect deep venous thrombosis in patients with suspicion of pulmonary embolism: diagnostic value and additional findings. AJR Am J Roentgenol 2001; 176: 1035–1039. 4 Ghaye B, Szapiro D, Willems V, Dondelinger RF. Combined CT venography of the lower limbs and spiral CT angiography of pulmonary arteries in acute pulmonary embolism: preliminary results of a prospective study. JBRBTR 2000; 83: 271–278. 5 Garg K, Kemp JL, Wojcik D et al. Thromboembolic disease: comparison of combined CT pulmonary angiography and venography with bilateral leg sonography in 70 patients. Am J Roentgenol 2000; 175: 997–1001. 6 Loud PA, Katz DS, Bruce DA, Klippenstein DL, Grossman ZD. Deep venous thrombosis with suspected pulmonary embolism: detection with combined CT venography and pulmonary angiography. Radiology 2001; 219: 498–502. 7 Cham MD, Yankelevitz DF, Shaham D et al., for the Pulmonary Angiography-Indirect CT Venography Cooperative Group. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216: 744–751. 8 Begemann PG, Bonacker M, Kemper J et al. Evaluation of the deep venous system in patients with suspected pulmonary embolism with multi-detector CT: a prospective study in comparison to Doppler sonography. J Comput Assist Tomogr 2003; 27: 399–409. 9 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182.
PART II
Diagnosis of DVT
10 Richman PB, Wood J, Kasper DM et al. Contribution of indirect computed tomography venography to computed tomography angiography of the chest for the diagnosis of thromboembolic disease in two United States emergency departments. J Thromb Haemost 2003; 1: 652– 657. 11 Revel MP, Petrover D, Hernigou A, Lefort C, Meyer G, Frija G. Diagnosing pulmonary embolism with fourdetector row helical CT: prospective evaluation of 216 outpatients and inpatients. Radiology 2005; 234: 265– 273. 12 Nicolas M, Debelle L, Laurent V et al. Incremental lower extremity CT venography, a simplified approach for the diagnosis of phlebitis in patients with pulmonary embolism. J Radiol 2001; 82: 251–256. 13 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 14 Ghaye B, Nchimi A, Noukoua C, Dondelinger RF. Does multi-detector row CT pulmonary angiography reduce the incremental value of indirect CT venography compared with single-detector row CT pulmonary angiography? Radiology 2006; 240: 256–262. 15 Johnson JC, Brown MD, McCullough N, Smith S. CT lower extremity venography in suspected pulmonary embolism in the ED. Emerg Radiol 2006; 12: 160–163. 16 Goodman LR, Stein PD, Beemath A, Sostman HD, Wakefield TW, Woodard PK, Yankelevitz D. CT venography: continuous helical images versus reformatted discontinuous images using PIOPED II data. Am J Roentgen 2007 (In press). 17 Goodman LG, Stein PD, Beemath A, Sostman HD, Wakefield TW, Woodard PK, Yankelevitz DF. Equivalence of venous phase CT venograms and compression ultrasound in the evaluation of patients with suspected acute pulmonary embolism: data from the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II). (Submitted for publication).
BLUK077-Stein
March 28, 2007
38
19:12
CHAPTER 38
Magnetic resonance angiography for diagnosis of deep venous thrombosis
Introduction Magnetic resonance angiography (MRA) has advantages compared to contrast-enhanced spiral CT, which include no ionizing radiation, no iodinated contrast material, and ability to perform repeated examinations safely [1]. Disadvantages are that MRA is expensive and cumbersome [1]. At present, MRA of the veins of the lower extremity is a second-line diagnostic tool because of higher cost, technical limitations, limited availability, and logistical constraints [2]. Magnetic resonance imaging was available, however, in 59% of non-Federal short-stay general and special hospitals in the United States in 2003 [3].
MR venography without contrast material Deep venous thrombosis (DVT) can be readily detected on MRI sequences without requiring a contrast medium. Results with spin-echo-black blood imaging [4], two-dimensional gradient-echo, time-of-flight MR venograms [5–9], or direct thrombus imaging on spin echo with fat saturation [10] have reported high sensitivity and specificity (87–100%) for detection of thrombosis of the pelvic and lower extremity veins. A disadvantage is the time required to obtain multiple axial images, and small superficial and perforating veins are often not delineated [6].
Gd-MR venography of the lower extremities Compared to two-dimensional time-of-flight techniques (which do not employ contrast material) three-
dimensional gadolinium-enhanced peripheral venography has a much shorter acquisition time, images look sharper, and more veins can be seen [11]. The use of low-dose contrast permits repeated measurements [11]. The pedal injection of diluted paramagnetic contrast material in combination with a dedicated vascular coil showed images comparable to conventional contrast venograms [7]. Sensitivity for venous segments, compared with contrast venography, was 13 of 13 (100%) and specificity was 88 of 90 (98%).
Delayed-enhanced MR venography of the veins of the thighs Fraser and associates [8] following an antecubital injection of gadopentetate dimeglumine showed that in the femoral vein, Gd-MRV (Gd-MR venography)was sensitive in 20 of 20 (100%) and specific in 31 of 32 (97%). In the iliac vein, Gd-MRV was sensitive in 7 of 7 (100%) and specific in 39 of 39 (100%). A normal Gd-MRA venogram following an intravenous injection of Gd is shown in Figure 38.1. Good visualization of the veins of the lower extremities was obtained with this technique immediately after imaging the pulmonary arteries (Figures 38.2 and 38.3) using a single intravenous injection of a gadoliniumcontaining contrast agent. No additional gadolinium was necessary following Gd-MRA pulmonary angiography. The gadolinium injected for the pulmonary MR angiogram was used for venous phase imaging of the thigh veins. Veins of the calf in general are not clearly shown.
175
BLUK077-Stein
March 28, 2007
19:12
176
(a)
PART II
(b)
Diagnosis of DVT
Femoral vein
(c)
Saphenous vein
Superficial femoral vein (d)
Popliteal vein Figure 38.1 (a) Subtracted maximum intensity projection images of the lower extremity venous system. (b) Contiguous reconstructed transverse images through the data set without subtraction. These transverse images show both veins (arrows) and arteries. (c) Contiguous reconstructed transverse images through the data set without subtraction. These transverse images show both
(a)
veins (arrows) and arteries. (d) Contiguous reconstructed transverse images through the data set without subtraction. These transverse images show both veins (arrows) and arteries. (Courtesy of Thomas Chenevert, PhD, Department of Radiology, University of Michigan, Ann Arbor, MI.)
Thrombus
Figure 38.2 Deep vein thrombosis shown with antecubital injection of gadolinium: comparison of techniques. (a) 2D interleaved fast spoiled gradient recalled echo with fat suppression 512 × 192 with one average; 28 locations 5 mm thick in 0.5 minutes. (b) 2D sequential time of flight
(b)
256 × 129 with two averages; 31 locations 5 mm thick in 6.3 minutes. (Courtesy of Pamela K. Woodard, MD, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO.)
BLUK077-Stein
March 28, 2007
CHAPTER 38
19:12
177
Magnetic resonance angiography for DVT
Vein Thrombus
Figure 38.3 Reformatted 2D fast spoiled gradient recalled echo showing deep venous thrombosis following an antecubital injection with gadolinium. (Courtesy of Pamela K. Woodard, MD, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO.)
Gd-MRV: diagnostic criteria for acute DVT The diagnostic criteria for acute DVT on Gd-MRV are: Occlusive = complete filling defect, i.e., failure to opacify the entire lumen due to a central filling defect (the vessel may enlarge compared to the opposite vein); Nonocclusive = partial filling defect surrounded by opacification.
Comprehensive examination The use of MRV of the veins of the lower extremities in combination with Gd-MRA of the pulmonary arteries creates a comprehensive study for thromboembolism comparable to the combination of contrast-enhanced pulmonary spiral CT of the pulmonary arteries in combination with venous phase CT of the veins of the lower extremities (CTV) [12, 13]. A pilot study of Gd-MRA in 20 patients indicated that such a comprehensive study was feasible and appeared to be a reliable and rapid diagnostic approach [14].
Gd-MRV of the upper extremities and central veins Patients with suspected thrombois, stenosis, occlusion, or compression of the axillary vein, subclavian vein, brachiocephalic vein, or superior or inferior vena cava were imaged with three-dimensional Gd-MRA [15]. The Gd-MRA was sensitive in 4 of 4 (100%) with
axillary or subclavian vein thrombosis [15]. Absence of disease was correctly identified in 13 of 13 (100%). A high number of patients, however, are not able to undergo MRA in this setting [16].
Complications Severe reactions to magneto-pharmaceuticals are rare, 1 in 350,000–450,000 [17]. In the 2 reported cases of anaphylactoid reactions, the patients had a history of reactive airway disease [17]. Any adverse reaction to gadolinium contrast material has been reported in 0.06% [18] but perhaps as many as 1% may have mild allergic reactions (hives) according to the package insert [19]. In June 2006, the US Food and Drug Administration (FDA) issued an alert indicating that nephrogenic systemic fibrosis or nephrogenic fibrosing dermopathy (NSF/NFD) had been reported in patients with advanced renal failure who received high doses of gadolinium-containing contrast agent [20]. In December 2006, the FDA reported NSF/NFD following gadolilnium-containing contrast agents in patients with only moderately severe renal disease [21]. This is discussed in Chapter 81.
References 1 Sostman HD. MRA for diagnosis of venous thromboembolism. Q J Nucl Med 2001; 45: 311–323. 2 Kanne JP, Lalani TA. Role of computed tomography and magnetic resonance imaging for deep venous thrombosis
BLUK077-Stein
March 28, 2007
19:12
178
3
4
5
6
7
8
9
10
11
12
and pulmonary embolism. Circulation 2004; 109: 115– 121. AHA Hospital Statistics, 2005 Edition. Health Forum LLC (an affiliate of American Hospital Association), Chicago, 2005: 159. Catalano C, Pavone P, Laghi A et al. Role of MR venography in the evaluation of deep venous thrombosis. Acta Radiol 1997; 38: 907–912. Spritzer CE, Norconk JJ, Jr, Sostman HD, Coleman RE. Detection of deep venous thrombosis by magnetic resonance imaging. Chest 1993; 104: 54–60. Ruehm SG, Zimny K, Debatin JF. Direct contrastenhanced 3D MR venography. Eur Radiol 2001; 11: 102– 112. Ruehm SG, Wiesner W, Debatin JF. Pelvic and lower extremity veins: contrast-enhanced three-dimensional MR venography with a dedicated vascular coil-initial experience. Radiology 2000; 215: 421–427. Fraser DG, Moody AR, Davidson IR, Martel AL, Morgan PS. Deep venous thrombosis: diagnosis by using venous enhanced subtracted peak arterial MR venography versus conventional venography. Radiology 2003; 226: 812–820. Gottschalk A, Stein PD, Goodman LR et al. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182. Botner RM, Buecker A, Wiethoff AJ et al. In vivo magnetic resonance imaging of coronary thrombosis using a fibrin binding molecular magnetic resonance contrast agent. Circulation 2004; 110: 1463–1466. Li W, David V, Kaplan R, Edelman RR. Three-dimensional low dose gadolinium-enhanced peripheral MR venography. J Magn Reson Imaging 1998; 8: 630–633. Loud PA, Katz DS, Klippenstein DL, Shah RD, Grossman ZD. Combined CT venography and pulmonary angiography in suspected thromboembolic disease: diagnostic accuracy for deep venous evaluation. AJR 2001; 174: 61– 65.
PART II
Diagnosis of DVT
13 Cham MD, Yankelevitz DF, Shaham D et al., for the Pulmonary CT Angiography-Indirect CT Venography Cooperative Group. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216: 744– 751. 14 Obernosterer A, Aschauer M, Portugaller H, Koppel H, Lipp RW. Three-dimensional gadolinium-enhanced magnetic resonance angiography used as a “one-stop shop” imaging procedure for venous thromboembolism: a pilot study. Angiology 2005; 56: 423–430. 15 Thornton MJ, Ryan R, Varghese JC, Farrell MA, Lucey B, Lee MJ. A three-dimensional gadolinium-enhanced MR venography technique for imaging central veins. AJR Am J Roentgenol 1999; 173: 999–1003. 16 Baarslag HJ, Van Beek EJ, Reekers JA. Magnetic resonance venography in consecutive patients with suspected deep vein thrombosis of the upper extremity: initial experience. Acta Radiol 2004; 45: 38–43. 17 Carr JJ. Magnetic resonance contrast agents for neuroimaging. Safety issues. Neuroimaging Clin N Am 1994; 4: 43–54. 18 Cochran ST, Bornyea K, Sayre JW. Trends in adverse events after IV administration of contrast media. AJR 2001; 176: 1385–1388. 19 Package Insert: Magnevist. Berlex Laboratories, Wayne, NJ. Revised May 2000. 20 U.S. Food and Drug Administration, Healthcare professional sheet. Gadolinium-containing contrast agents for magnetic resonance imaging (MRI) (marketed as Omniscan, OptiMARK, Magnevist, ProHance and MultiHance). http://www.fda.gov/cder/drug/InfoSheets/HCP/ gccaHCP.htm Accessed 23 June 2006. 21 US Food and Drug Administration Public Health Advisory. Update on magnetic resonance imaging (MRI) contrast agents containing gadolinium and nephrogenic fibrosing dermopathy. http://www.fda.gov/cder/drug/ advisory/gadolinium agents 20061222.htm.
BLUK077-Stein
March 11, 2007
39
8:17
CHAPTER 39
P-selectin and microparticles to predict deep venous thrombosis
A preliminary investigation of the predictive value of a combination of D-dimer, P-selectin, and microparticles showed that plasma markers have potential for diagnostic evaluation, but sensitivity and specificity need to be improved for clinical applicability [1]. Pselectin is present in the alpha granules of platelets and the Weibel-Palade bodies of endothelial cells [1]. P-selectin is the first upregulated glycoprotein on activated endothelial cells and platelets and has procoagulant properties [1]. Microparticles are fragments of cell membranes shed from platelets [2], leucocytes [3], and endothelial cells [4]. They are <1 μm in size and circulate within the vascular system [1]. Hematopoietic cells may also be a significant source of microparticles that participate in thrombosis via P-selectindependent mechanisms [5]. Microparticles are rich in tissue factor [6], phosphatidylserine [7], and the receptor for P-selectin, PSGL-1 [8]. Through interactions between P-selectin and its receptor, PSGL-1, microparticles are elaborated and then taken up into thrombi [1]. The presence of these microparticles has an important role in thrombus generation and its subsequent amplification [9, 10]. Combinations of P-selectin, microparticles, and Ddimer were evaluated among 22 patients with deep venous thrombosis (DVT) and 21 symptomatic patients in whom DVT was excluded [1]. Threshold values were soluble P-selectin 0.68 ng/mg of total protein, total microparticles 125% compared to controls, and D-dimer 3 mg/L by latex turbidimetric assay. Sensitivity of P-selectin alone was 68%, total microparticles alone 50%, and D-dimer 64%. Respective specificities were 81, 67, and 76%. Combined application of these variables gave a sensitivity of 73% and specificity of 81%.
References 1 Rectenwald JE, Myers DD, Jr, Hawley AE et al. D-dimer, P-selectin, and microparticles: novel markers to predict deep venous thrombosis. A pilot study. Thrombo Haemost 2005; 94: 1312–1317. 2 Gilbert GE, Sims PJ, Wiedmer T, Furie B, Furie BC, Shattil SJ. Platelet-derived microparticles express high affinity receptors for factor VIII. J Biol Chem 1991; 266: 17261– 17268. 3 Mesri M, Altieri DC. Endothelial cell activation by leukocyte microparticles. J Immunol 1998; 161: 4382–4387. 4 Sabatier F, Roux V, Anfosso F, Camoin L, Sampol J, Dignat-George F. Interaction of endothelial microparticles with monocytic cells in vitro induces tissue factordependent procoagulant activity. Blood 2002; 99: 3962– 3970. 5 Chou J, Mackman N, Merrill-Skoloff G, Pedersen B, Furie BC, Furie B. Hematopoietic cell-derived microparticle tissue factor contributes to fibrin formation during thrombus propagation. Blood 2004; 104: 3190–3197. 6 Falati S, Liu Q, Gross P et al. Accumulation of tissue factor into developing thrombi in vivo is dependent upon microparticle P-selectin glycoprotein ligand 1 and platelet P-selectin. J Exp Med 2003; 197: 1585–1598. 7 Martinez MC, Tesse A, Zobairi F, Andriantsitohaina R. Shed membrane microparticles from circulating and vascular cells in regulating vascular function. Am J Physiol Heart Circ Physiol 2005; 288: H1004–H1009. 8 Furie B, Furie BC. Role of platelet P-selectin and microparticle PSGL-1 in thrombus formation. Trends Mol Med 2004; 10: 171–178. 9 Myers DD, Hawley AE, Farris DM et al. P-selectin and leukocyte microparticles are associated with venous thrombogenesis. J Vasc Surg 2003; 38: 1075–1089. 10 Myers D, Jr, Farris D, Hawley A et al. Selectins influence thrombosis in a mouse model of experimental deep venous thrombosis. J Surg Res 2002; 108: 212–221.
179
BLUK077-Stein
April 6, 2007
16:0
III Diagnosis of acute PART III
pulmonary embolism
181
BLUK077-Stein
April 6, 2007
40
16:0
CHAPTER 40
Clinical characteristics of patients with no prior cardiopulmonary disease
Introduction The clinical manifestations of pulmonary embolism (PE) are recognized to be nonspecific. When considered together, frequent signs and symptoms, and combinations of signs and symptoms, particularly in patients with predisposing factors for PE, can strongly suggest the need for further diagnostic evaluation [1, 2]. The most extensive databases on the clinical characteristics of acute PE are from the Urokinase Pulmonary Embolism Trial (UPET) [3], PIOPED I (Prospective Investigation of Pulmonary Embolism Diagnosis I) [4], and more recently, PIOPED II [5]. It is clear, however, that the clinical characteristics reported among patients in PIOPED I, as in PIOPED II and other investigations of patients enrolled in clinical trials, are characteristics of patients whose findings were sufficient to alert physicians to the diagnosis and characteristics of patients who met the inclusion criteria. Patients who were too ill to participate, or who died suddenly, or were unidentified because the clinical findings were mild or atypical would not be included. Patients in whom PE was suspected but excluded would also be likely to have clinical findings similar to patients with PE, because such clinical findings are recognized by physicians as those associated with PE, and for this reason patients were evaluated. In spite of these constraints, recognition of the clinical characteristics is vital in the initial identification of patients who may have PE. Critical to identification of patients with PE is a high level of suspicion, particularly in patients with risk factors for PE, which would then trigger further evaluation. In order to characterize the diagnostic features of the history and physical examination in patients with acute PE, patients with no preexisting cardiac or pulmonary disease were evaluated. In such patients, the features
of acute PE become clear and manifestations related to coexistent disease are excluded. In PIOPED I [4] there were 117 patients with PE and no prior cardiopulmonary disease. All were diagnosed by pulmonary angiography or autopsy. PE was excluded in 248 such patients. It was excluded by a normal pulmonary angiogram in 176 patients and by the absence of adverse events over the course of 1 year in 72 patients who did not receive anticoagulant therapy [6]. In PIOPED II, there were 133 patients with PE and no prior cardiopulmonary disease and PE was excluded in 366 such patients. The diagnosis and exclusion of PE was on the basis of a composite set of reference tests [7] (Chapter 76). Patients in PIOPED I were considered to have no preexisting cardiac or pulmonary disease if they had no history or evidence of valvular heart disease, coronary artery disease, myocardial infarction, “other heart disease,” or if prior to this acute episode, they had not suffered left- or right-sided congestive heart failure. Such patients had no history of asthma, chronic obstructive pulmonary disease, interstitial lung disease, “other lung disease,” and no recognized acute pneumonia or acute respiratory distress syndrome at the time of the evaluation for suspected PE. These patients also had no previous history of PE. In PIOPED II, the definitions of “no prior cardiopulmonary disease” differed somewhat because the information collected in the two investigations was not exactly the same. In PIOPED II, patients were defined as having no prior cardiopulmonary disease if they had no current asthma or pneumonia, no history of chronic bronchitis, emphysema, chronic obstructive pulmonary disease, lung cancer, or prior PE, and no current or history of rightor left-sided heart failure. Patients in PIOPED I with PE and no prior cardiopulmonary disease were 54 ± 17 years of age 183
BLUK077-Stein
April 6, 2007
16:0
184
(mean ± standard deviation) and 49% were males [4]. Patients in PIOPED II with PE were 56 ± 16 years of age. In PIOPED II, 38% were males [5].
Predisposing factors Immobilization, irrespective of the cause, was shown to be the most prevalent predisposing factor. Data from PIOPED I [4] indicate, as others have shown [8], that immobilization of even 1 or 2 days may predispose to PE. Among 117 patients with PE who had no preexisting cardiac or pulmonary disease, immobilization usually due to surgery occurred in 56% of the patients (Table 40.1) [4]. Among the 66 patients in whom immobilization was a predisposing factor, 43 (65%) were immobilized for ≤2 weeks. Immobilization of 2 days or less preceded PE in 8 of 117 (7%). One or more predisposing factors were present in 96 of 117 (82%) patients with PE. In PIOPED II, immobilization was not as frequent a predisposing factor as in PIOPED I (Table 40.1) [4, 5]. Recent travel ≥4 hours was reported in 14% of the patients with PE. Almost half of the patients had smoked at one time. One or more risk
PART III
Diagnosis of acute PE
factors were reported in 92% of the patients with PE in PIOPED II. Subsequent to PIOPED II, we evaluated the incidence of venous thromboembolism in patients with the nephrotic syndrome using data from the National Hospital Discharge Survey (Kayali F, Najjar R, Stein PD. Unpublished data). The rate of diagnosis of deep venous thrombosis (DVT) in patients with the nephrotic syndrome was 1.5%. The relative risk, comparing with patients who did not have the nephrotic syndrome was 1.67. Among patients 18–39 years old, the rate of DVT with the nephrotic syndrome was 2.8% (relative risk 7.0). There were too few patients with renal vein thrombosis to calculate its rate of occurrence. The data show that the nephrotic syndrome is a risk factor for DVT. Pulmonary embolism, if it occurs, is likely to be due to DVT and not renal vein thrombosis, because of the infrequent occurrence of the latter.
Syndromes of acute PE Consideration of the diagnosis of PE in terms of its presenting syndrome may focus one’s thinking about the diverse manifestations of PE. The syndrome of
Table 40.1 Predisposing factors in patients with pulmonary embolism and no previous cardiac or pulmonary disease. n (%) Risk factors
PIOPED I ( N = 117)
PIOPED II ( N = 131–133)
Immobilization
66 (56)*
27 (20)†
Surgery (≤3 mo)
63 (54)
30 (23)‡
Malignancy (excluding lung cancer)*
27 (23)
29 (22)
Thrombophlebitis, ever
16 (14)
11 (8)
12 (10)
14 (11)
Travel ≥4 h in last month
19 (14)
Trauma (≤3 mo) Lower extremities Other
4 (3)
Smoke (ever)
55 (42)
<1 pack/day
29 (22)
1–2 packs/day
20 (15)
≥2 packs/day
0 (0)
Central venous instrumentation (≤3 mo)
12 (9)
Stroke, paresis, or paralysis
8 (7)
Postpartum (≤3 mo)
5 (4)
4 (3)
*Actively treated in last 3 months. † Within last month. ‡ Among 30 patients with surgery as a risk factor, the prevalence of heart, abdominal, pelvic, hip/knee – open, hip/knee – replacement, and neurosurgery ranged from 3 to 5. Includes pelvis. Data from Stein et al. [4, 5].
BLUK077-Stein
April 6, 2007
CHAPTER 40
16:0
185
Patients with no prior cardiopulmonary disease
Table 40.2 Syndromes of acute pulmonary embolism. n (%) Syndrome
PIOPED I ( N = 117)
PIOPED II ( N = 133)
Pleuritic pain or hemoptysis
76 (65)
55 (41)
Uncomplicated dyspnea
26 (22)
48 (36)
Circulatory collapse*
9 (8)
11 (8)
Different Presentation
6 (5)
19 (14)
*Systolic blood pressure <80 mm Hg or loss of consciousness. Data from Stein et al. [4, 5].
pleuritic pain or hemoptysis, in the absence of circulatory collapse, was the most frequent mode of presentation occurring in 65% of the patients with PE and no prior cardiopulmonary disease in PIOPED I and in 41% in PIOPED II (Table 40.2) [4, 5]. Circulatory collapse (systolic blood pressure <80 mm Hg or loss of consciousness) was an uncommon mode of presentation in these investigations occurring in 8% of the patients with PE and no prior cardiopulmonary disease in PIOPED I and PIOPED II. A syndrome of dyspnea in the absence of hemoptysis or pleuritic pain or circulatory collapse occurred in 22% in PIOPED I and in 36% in PIOPED II. Many of the remaining patients in PIOPED II were identified on the basis of testing in patients with deep venous thrombosis (DVT), or other findings in the absence of any of these syndromes. The syndrome of pleuritic pain or hemoptysis may be more frequent in PIOPED I and PIOPED II than the other syndromes of PE because such patients usually have less extensive PE than patients with other syndromes of PE [1]. Such patients, therefore, may have been more likely to be recruited.
Symptoms of acute PE Among patients with PE and no prior cardiac or pulmonary disease, dyspnea was the most common symptom, occurring in 73% in PIOPED I and 73% in PIOPED II (Table 40.3) [4, 5]. Among patients with PE and dyspnea, 21 of 97 (22%) had dyspnea only on exertion, whereas 73 of 97 (75%) had dyspnea at rest [5]. In a few patients, the information was not known (Table 40.3). Orthopnea was shown to be a symptom of PE. Orthopnea occurred in 37 of 73 (51%) of the patients with PE who had dyspnea at rest and it occurred in 7 of 21 (33%) with dyspnea only on exertion (information not known in 2 patients with orthopnea) [5]. The rate of onset of dyspnea was within seconds
in 46% of patients, within minutes in 26%, and within hours in 11% (5). In some patients (17%), the rate of onset was within days (5). In both PIOPED I and PIOPED II, pleuritic chest pain was more frequent in patients with PE and no prior cardiopulmonary disease than was hemoptysis (Table 40.3). Cough was common among patients with PE and no prior cardiopulmonary disease in PIOPED I and PIOPED II, occurring in 37 and 34% [4, 5]. It was usually nonproductive (Table 40.3). In those in whom the cough was productive, similar proportions of patients had clear, purulent, or bloody sputum (Table 40.3). In small number of patients, blood-streaked sputum (4 patients) was more frequent than all blood (1 patient), and the amount of blood in all instances was <1 teaspoonful (Table 40.3). Calf or thigh pain occurred in 44% of the patients with PE and no prior cardiopulmonary disease in PIOPED II (Table 40.3) Thigh pain alone was uncommon, 3% [5]. The relation of the sensitivity of individual symptoms (percent of patients with PE who had the particular symptoms) to the percent of false positives among patients in PIOPED I who did not have PE is shown in Figure 40.1.
Signs of acute PE Tachypnea (respiratory rate ≥ 20/min) was the most common sign of acute PE among patients with no prior cardiac or pulmonary disease, occurring in 70% in PIOPED I and 54% in PIOPED II (Table 40.4) [4, 5]. Tachycardia (heart rate > 100/min) occurred in 30 and 24% of the patients with PE. One of the signs of right atrial, right ventricular, or pulmonary artery pressure elevation occurred in 21% of the patients (Table 40.4).
BLUK077-Stein
April 6, 2007
16:0
Table 40.3 Symptoms in patients with pulmonary embolism and no preexisting cardiac or pulmonary disease. PE n (%) Symptoms
PIOPED I ( N = 117)
PIOPED II ( N = 127–133)
85 (73)
97 (73)
Dyspnea Dyspnea (rest or exertion) Dyspnea (at rest)
73 (55)
Dyspnea (exertion only)
21 (16)
Orthopnea (≥2-pillow)
37 (28)
Pleuritic pain Chest pain (not pleuritic) Cough Hemoptysis
77 (66)
58 (44)
5 (4)
25 (19)
43 (37)
45 (34)
15 (13)
7 (5)*
Purulent
7 (5)
Clear
7 (5)
Nonproductive
26 (20)
Wheezing
10 (9)
Palpitations
12(10)
27 (21)
Calf swelling
51 (40)
Thigh swelling
10 (8)
Thigh swelling, no calf swelling
1 (1) 30 (26)†
Calf or thigh pain
56 (44)
Calf pain only
30 (23)
Calf and thigh pain
22 (17)
Thigh pain only
4 (3)
*Hemoptysis; patients with PE: 2 = slightly pinkish, 4 = blood-streaked, 1 = all blood (<1 teaspoonful). † “Leg pain.” PE, pulmonary embolism. Data from Stein et al. [4, 5].
100
80
Sensitivity (%)
1 2 60
40
1. Dyspnea 2. Pleuritic pain 3. Cough 4. Leg swelling 5. Leg pain 6. Hemoptysis 7. Palpitations 8. Wheezing 9. Anginal pain
3 4 5
20 6 9
8
7
0 0
20
40 60 False positives (%)
Figure 40.1 Sensitivity and frequency of false-positive symptoms in patients with suspected acute pulmonary embolism (PE) who had no prior cardiopulmonary disease
80
(CPD) (n = 365). The numbers refer to the individual symptoms shown in the insert. (Data from Stein et al. [4].)
186
100
BLUK077-Stein
April 6, 2007
CHAPTER 40
16:0
187
Patients with no prior cardiopulmonary disease
Table 40.4 Signs in patients with pulmonary disease and no preexisting cardiac or pulmonary disease. PE n (%) Signs
PIOPED I ( N = 117)
PIOPED II ( N = 128–132)
General Tachypnea (≥20/min)
82 (70)
71 (54)
Tachycardia (>100/min)
35 (30)
32 (24)
Diaphoresis
13 (11)
3 (2)
Cyanosis
1 (1)
0 (0)
Temperature >38.5◦ C (>101.3◦ F)
8 (7)
1 (1)
Cardiac examination (any) Increased P2 Third heart sound
28 (21) 27 (23)
Fourth heart sound
28 (24)
Right ventricular lift
5 (4)
4 (4)∗∗
Jugular venous distension
18 (14)
Lung examination (any) Rales (crackles) Wheezes
38 (29) 60 (51)
23 (18)
6 (5)
2 (2)
Rhonchi
2 (2)
Decreased breath sounds Pleural friction rub
15 (15)∗
3 (3)
22 (17) 3 (3)
0 (0)
13 (11)
62 (47)*
DVT Calf or thigh Calf only
42 (32)
Calf and thigh
18(14)
Thigh only Homans’ sign
2 (2) 5 (4)
Number of patients with PE who had one or more signs of DVT: edema = 55, erythema = 5, tenderness = 32, palpable cord = 2. PE, pulmonary embolism; P2, pulmonary component of second sound; DVT, deep venous thrombosis. Data from Stein et al. [4, 5]. *Data in 103 pts. **Data in 110 pts.
Examination of the lungs showed an abnormality in 29% of the patients with PE and no prior cardiopulmonary disease (Table 40.4). Rales (crackles) and decreased breath sounds were the most frequently detected abnormalities. Wheezes and rhonchi were heard occasionally. A pleural friction rub was rare in PIOPED I and was not observed in PIOPED II. In PIOPED I, most patients with PE who had rales (88%) had pulmonary parenchymal abnormalities, atelectasis, or a pleural effusion on the chest radiograph. Signs of DVT were observed in the calf or thigh in 47% of the patients with PE and no prior cardiopulmonary disease (Table 40.4). Both the thigh and calf showed signs of DVT in 14%, but the thigh alone uncommonly showed signs of DVT in 2%.
The relation of the sensitivity of individual signs (percent of patients with PE who had the particular signs) to the percent of false positives among patients in PIOPED I who did not have PE is shown in Figure 40.2.
Combinations of signs and symptoms Combinations of clinical characteristics in patients with PE and no prior cardiopulmonary disease are shown in Table 40.5. Dyspnea or tachypnea was shown in 90 and 84% of such patients in PIOPED I and PIOPED II [4, 5]. Dyspnea or tachypnea or pleuritic pain was shown in 97 and 92%. One of these findings or signs of DVT was found in 98% of such patients in
BLUK077-Stein
April 6, 2007
16:0
188
PART III
Diagnosis of acute PE
100
80 1. RR ≥ 20/min 2. Rales 3. HR > 100/min 4. S4 5. Increased P2 6. DVT 7. Diaphoresis 8. TEMP > 38.5 9. Wheezes 10. Homans' 11. RV Lift 12. Pleural rub 13. S3 14. Cyanosis
Sensitivity (%)
1 60 2 40 3 4
5
20
6
7 9
0
8
0
20
40 60 False positives (%)
80
100
Figure 40.2 Sensitivity and frequency of false-positive signs in patients with suspected acute pulmonary embolism (PE) who had no prior cardiopulmonary disease
(CPD) (n = 365). The numbers refer to the individual signs shown in the insert. (Data from Stein et al. [4].)
PIOPED II. Conversely, in the absence of dyspnea or tachypnea, pleuritic pain, or signs of DVT, PE was infrequently diagnosed [4, 5]. Comparable observations among patients with no prior cardiopulmonary disease were also made in the
UPET [3]. Among 215 patients, either dyspnea or tachypnea (respiratory rate ≥20/min) occurred in 96% of the patients with no preexisting cardiac or pulmonary disease in UPET [3]. Either dyspnea, tachypnea, or signs of DVT occurred in 99% [3].
Table 40.5 Combinations of clinical characteristics in patients with pulmonary embolism and no prior cardiopulmonary disease. PE n (%) Signs
PIOPED I ( N = 117)
PIOPED II ( N = 131)
Dyspnea or tachypnea (≥20/min)
105 (90)
110 (84)
Dyspnea or tachypnea (≥20/min) or pleuritic pain
113 (97)
120 (92)
Dyspnea or tachypnea (≥20/min) or pleuritic pain
128 (98)
or signs of DVT Dyspnea or tachypnea (≥20/min) or signs of DVT
107 (91)
Dyspnea or tachypnea (≥20/min) or pleuritic pain
115 (98)
or radiographic evidence of atelectasis or a parenchymal abnormality PE, pulmonary embolism. Data from Stein et al. [4, 5].
BLUK077-Stein
April 6, 2007
CHAPTER 40
16:0
189
Patients with no prior cardiopulmonary disease
Table 40.6 Clinical assessment in patients with no preexisting cardiac or pulmonary disease. PE/n (%) Pulmonary embolism
Pulmonary embolism
Clinical assessment
(assessment by fellow)
(assessment by staff)
High probability
26/41 (63)
15/17 (88)
Intermediate probability
47/99 (47)
74/200 (37)
8/74 (11)
14/103 (14)
Low probability
Differences between assessments by fellows and staff were not significant. PE, pulmonary embolism. Data are from Stein et al. [4].
Clinical assessment Physicians, on the basis of clinical assessment, were more consistent in their ability to exclude PE than they were in their ability to make the diagnosis of PE (Table 40.6). In the majority of patients (46–63%), however, physicians were suspicious of the diagnosis of PE, but uncertain in their assessment. Senior staff, when they made a high-probability assessment, were correct in 88% and fellows were correct in 63% (Table 40.6) [4]. The clinical assessment of the likelihood of acute PE was based upon all available noninvasive data, not any specific predetermined criteria. In conclusion, symptoms may be mild and traditionally important symptoms may be absent. There may be dyspnea only on exertion. Orthopnea may be a symptom of PE.
References 1 Stein PD, Willis PW, III, DeMets DL. History and physical examination in acute pulmonary embolism in patients without pre-existing cardiac or pulmonary disease. Am J Cardiol 1981; 47: 218–223.
2 Wenger NK, Stein PD, Willis PW, III. Massive acute pulmonary embolism: the deceivingly nonspecific manifestations. JAMA 1972; 220: 843–844. 3 A National Cooperative Study. Clinical and electrocardiographic observations (Chapter X of The Urokinase Pulmonary Embolism Trial). Circulation 1973; 47(suppl II): 60–65. 4 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 5 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, 2007 (In press). 6 The PIOPED Investigators. Value of the ventilation/ perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 7 Stein PD. Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II). N Engl J Med 2006; 354: 2317–2327. 8 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487– 501.
BLUK077-Stein
March 12, 2007
41
16:18
CHAPTER 41
Relation of right-sided pressures to clinical characteristics of patients with no prior cardiopulmonary disease
The pulmonary artery mean pressure in patients with pulmonary embolism (PE) and no prior cardiopulmonary disease in Prospective Investigating Pulmonary Embolism Diagnosis I (PIOPED I) did not exceed the upper limits of normal (20 mm Hg) in 58% of the 111 randomized patients (Figure 41.1) (P.D. Stein and B. Relyea unpublished data from PIOPED I, 1990). Pulmonary artery mean pressure was 30 mm Hg or less in 85% of these patients. Right atrial mean pressure was less than the upper limits of normal (8 mm Hg) in 75% of the 109 patients with PE and no prior cardiopulmonary disease (Figure 41.2). Right atrial mean pressure was rarely elevated in patients in whom the pulmonary artery mean pressure was normal. Even among patients in whom pulmonary artery mean pressure exceeded 20 mm Hg, there was a poor correlation with right atrial mean pressure (r = 0.11). Right-sided pressures were not routinely obtained in PIOPED II. The relation of signs and symptoms to pulmonary artery mean pressure, right atrial mean pressure,
and the partial pressure of oxygen in arterial blood (Pa O2 ) while breathing room air is shown in Table 41.1. The pulmonary artery mean pressure was lowest in patients with pleuritic chest pain and highest in those with a third heart sound, right ventricular lift, diaphoresis, and palpitations. A few statistically significant differences were shown at the P < 0.01 level. The pulmonary artery mean pressure was higher in patients with a third heart sound than in patients with the following signs and symptoms who did not have a third heart sound: pleural friction rub, accentuated pulmonary component of the second sound, tachycardia, tachypnea, wheezing, hemoptysis, cough, pleuritic pain, and dyspnea. Patients with pleuritic pain had a lower pulmonary artery mean pressure than patients with palpitations and patients with tachypnea who did not have pleuritic pain. There was a poor correlation between pulmonary artery mean pressure and the Pa O2 (Figure 41.3). 100
25
80
n = 111 20
60 15
40 10
20
5
0
4
8
12
16
20
24
28
32
36
40
48
52
PA mean pressure (mm Hg)
190
0
Cumulative frequency (%)
Number of patients
30
Figure 41.1 Distribution and cumulative frequency of pulmonary artery mean pressure among randomized patients with pulmonary embolism and no prior cardiopulmonary disease (P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).
BLUK077-Stein
March 12, 2007
16:18
100
Figure 41.2 Distribution and cumulative frequency of right atrial mean pressure among randomized patients with pulmonary embolism and no prior cardiopulmonary disease (P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).
30
n = 109
25
80
60
20 15
40
10 20 5
Cumulative frequency (%)
Number of patients
35
0
0 2
4
6
8
10 12 14 16 18
20 22 24 26
RA mean pressure (mm Hg)
Table 41.1 Relation between signs, symptoms, right-sided pressures, and Pa O2 . Signs/symptoms
PA mean (mm Hg)
RA mean (mm Hg)
Pa O2 (mm Hg)
Dyspnea
23 ± 9 (n = 121)
6 ±5 (n = 114)
69 ± 14 (n = 98)
Pleuritic pain
21 ± 8 (n = 115)
5 ± 4 (n = 114)
73 ± 14 (n = 97)
Cough
24 ± 9 (n = 68)
6 ± 5 (n = 63)
71 ± 13 (n = 54)
Hemoptysis
22 ± 8 (n = 24)
5 ± 5 (n = 24)
74 ± 14 (n = 23)
Palpitations
27 ± 10 (n = 18)
6 ± 6 (n = 15)
64 ± 15 (n = 15)
Wheezing
21 ± 7 (n = 16)
4 ± 3 (n = 14)
70 ± 15 (n = 13)
Angina-like pain
25 ± 14 (n = 9)
3 ± 3 (n = 6)
62 ± 9 (n = 8)
Tachypnea (≥20/min)
22 ± 9 (n = 118)
6 ± 5 (n = 112)
69 ± 15 (n = 95)
Rales (crackles)
22 ± 8 (n = 81)
6 ± 4 (n = 78)
68 ± 14 (n = 68)
Tachycardia (>100/min)
24 ± 9 (n = 59)
7 ± 5 (n = 53)
69 ± 17 (n = 42)
Increased P2
23 ± 9 (n = 42)
6 ± 5 (n = 40)
67 ± 14 (n = 33)
Diaphoresis
26 ± 11 (n = 16)
8 ± 7 (n = 14)
61 ± 15 (n = 12)
Right ventricular lift
28 ± 12 (n = 3)
8 ± 9 (n = 3)
67 ± 23 (n = 4)
Friction rub
21 ± 8 (n = 6)
9 ± 6 (n = 6)
67 ± 4 (n = 5)
Third heart sound
39 ± 2 (n = 3)
12 ± 8 (n = 4)
71 ± 9 (n = 3)
Values are mean ± standard deviation. PA, pulmonary artery; RA, right atrium; RV, right ventricle; Pa O2 , partial pressure of oxygen in arterial blood in patients breathing room air. P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).
Figure 41.3 Relation of the partial pressure of oxygen in arterial blood (Pa O2 ) while breathing room air to the pulmonary artery (PA) mean pressure in patients with no prior cardiopulmonary disease. Both randomized and referred patients are included (P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).
191
BLUK077-Stein
March 28, 2007
42
15:25
CHAPTER 42
The history and physical examination in all patients irrespective of prior cardiopulmonary disease
Introduction In this chapter, the clinical characteristics of all patients with acute pulmonary embolism (PE) are described. Evaluation of the clinical characteristics of acute PE in patients with no prior cardiopulmonary disease eliminates patients with PE whose clinical findings may be due to underlying disease and was described in Chapter 40. Data from Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) include patients who were randomized for pulmonary angiography as described in the primary PIOPED report [1] and patients who were referred for angiography and not described in the original PIOPED report. The patients with acute PE in PIOPED I were 58 ± 17 years of age (mean ± standard deviation) and 54% were males [2]. In PIOPED II, patients with PE were 57 ± 17 years of age and 42% were males [3].
Predisposing factors Predisposing factors are shown in Table 42.1. In PIOPED II, 94% had one or more risk factors. Immobilization was the most frequent predisposing factor, and surgery was the usual cause of immobilization. In PIOPED I and PIOPED II, 19 and 10% of the patients, respectively, had a history of ever having had deep venous thrombosis (DVT). Six percent and 4% had prior PE in PIOPED I and PIOPED II (Table 42.1).
Syndromes of acute PE The syndrome of Hemoptysis or pleuritic pain, in the absence of circulatory collapse, and the syndrome of
192
uncomplicated dyspnea were frequent modes of presentation in patients with acute PE in PIOPED II (Table 42.2) [3]. Circulatory collapse (systolic blood pressure <80 mm Hg or loss of consciousness) was an uncommon mode of presentation, possibly because such patients were not recruited.
Symptoms of acute PE Dyspnea was the most frequent symptom, occurring in 78 and 79% of patients with PE in PIOPED I and PIOPED II (Table 42.3) [2, 3]. Dyspnea was usually present at rest (61% of the patients with PE), but in some (16% with PE) it was present only with exertion (Table 42.3) [3]. Orthopnea was reported in 36% of the patients with PE [3]. The onset of dyspnea was within seconds in 41%, minutes in 26%, hours in 14%, and over a period of days in 19% [3] Pleuritic chest pain occurred four to eight times more frequently than hemoptysis in patients with PE. Pure bloody sputum was observed in only 2 patients, and it was <1 teaspoonful. Cough was common and occurred in 43% of the patients with PE in both PIOPED I and PIOPED II (Table 42.3).
Signs of acute PE Tachypnea (respiratory rate ≥20/min) was the most common sign of acute PE and was observed in 73% of the patients with PE in PIOPED I and 57% in PIOPED II (Table 42.4) [2, 3]. Rales and decreased breath sounds were the most frequent findings on examination of the lungs, but rhonchi and wheezes were also observed. Rarely, signs of DVT were found only in the
BLUK077-Stein
March 28, 2007
CHAPTER 42
15:25
193
All patients irrespective of prior cardiopulmonary disease
Table 42.1 Predisposing factors in all patients with pulmonary embolism. n (%) PIOPED I (N = 383)
Risk factors Immobilization
206 (54)
†
Travel ≥4 h in last month Surgery (≤3 mo)
PIOPED II (N = 185 – 192)* 48 (25)‡ 23 (12)
160 (42)
Coronary heart disease
76 (20)
Myocardial infarction
48 (13)
Heart failure
45 (12)
Collagen vascular disease
15 (4)
41 (21)§
10 (5)
Malignancy†
69 (18)
Stroke, paresis, or paralysis
37 (10)
37 (19)¶ 7 (4)
Thrombophlebitis, ever
71 (19)
19 (10)
Prior pulmonary embolism
23 (6)
7 (4)
47 (12)
16 (8)||
27 (7)
22 (12)
Trauma (≤3 mo) Lower extremities Other
5 (3)
Asthma Pneumonia (current)
27 (7)
5 (3)
Chronic obstructive pulmonary disease
37 (10)
10 (5)
Emphysema
7 (4)
Interstitial lung disease
6 (2)
Lung cancer
5 (3)
Estrogen
22 (6)
Males, therapeutic
1 (1)
Smoke (ever)
90 (47)
<1 pack/day
43 (22)
1–2 packs/day
37 (19)
>2 packs/day
1 (1)
Central venous instrumentation (≤3 mo) Postpartum (≤3 mo)
22 (12) 9 (2)
Sepsis (current)
0 (0)
* Nominal value. If value was not reported, it was assumed to be absent. †
Actively treated in last 3 months. Within last month. § Among 41 patients with surgery as a risk factor, abdominal surgery was in 9, and heart, pelvic, hip/knee – open, hip/knee – replacement, and neurosurgery ranged from 3 to 5. ¶ Excluding lung cancer. || Includes pelvis. Data from PIOPED I (unpublished) and Stein et al.[3]. ‡
Table 42.2 Syndromes of acute pulmonary embolism. Syndrome
n (%) in PIOPED II (N = 192)
Hemoptysis or pleuritic pain
84 (44)
Uncomplicated dyspnea Circulatory collapse*
70 (36) 15 (8)
* Systolic blood pressure <80 mm Hg or loss of consciousness. Data from Stein et al. [3].
BLUK077-Stein
March 28, 2007
15:25
194
PART III
Diagnosis of acute PE
Table 42.3 Symptoms in all patients with pulmonary embolism. PE n (%) Symptoms
PIOPED I (N = 383)
PIOPED II (N = 184 – 192)
Dyspnea Dyspnea (rest or exertion)
299 (78)
Dyspnea (at rest) Dyspnea (exertion only)
31 (16)
Orthopnea (≥2 pillows) Pleuritic pain Chest pain (not pleuritic) Cough Hemoptysis
151 (79) 117 (61) 69 (36)
225 (59)
89 (47)
23 (6)
33 (17)
166 (43)
82 (43)
62 (16)
11 (6)*
Purulent
18 (10)
Clear
18 (10)
Nonproductive Wheezing
40 (21) 54 (14)
Palpitations
48 (13)
Calf swelling
118 (31)†
58 (31) 54 (29)‡
Calf or thigh swelling
72 (39)
Calf and thigh swelling
15 (8)
Thigh swelling, no calf swelling Calf pain
3 (2) 103 (27)
43 (23)‡
Calf or thigh pain
78 (42)
Calf and thigh pain
30 (16)
Thigh pain, no calf pain
5 (3)
* Hemoptysis; patients with PE: 3 = slightly pinkish; 6 = blood-streaked; 2 = all blood (<1 teaspoonful). † “Leg.” ‡ Calf only. PE, pulmonary embolism. Data from Stein et al. [2, 3].
thighs. Tachycardia (heart rate >100/min) was present in only 30 and 26% of the patients with PE in PIOPED I and PIOPED II (Table 42.4). In the Urokinase Pulmonary Embolism Trial, among all patients with PE irrespective of the presence or absence of prior cardiopulmonary disease, syncope, accentuated pulmonary component of the second, third, or fourth heart sound (S3 or S4 ), and cyanosis were more frequent among patients with massive PE than submassive PE [4]. Pleuritic pain was less frequent among patients with massive PE than submassive PE.
Combinations of signs and symptoms Among all patients with acute PE , dyspnea or tachypnea (respiratory rate ≥20/min) was present in 91 and
86% of the patients in PIOPED I and PIOPED II (Table 42.5) [2, 3]. Dyspnea or tachypnea or pleuritic pain was present in 97 and 92% [2, 3]. Dyspnea or tachypnea or pleuritic pain or signs of DVT were observed in 97 in both PIOPED I and PIOPED II [2, 3] In PIOPED I, all patients with acute PE had dyspnea or tachypnea or pleuritic pain or unexplained radiographic evidence of atelectasis or a parenchymal abnormality or an unexplained low Pa O2 [2].
Dyspnea, hypoxia, and normal chest radiograph Dyspnea and a low partial pressure of oxygen in arterial blood (Pa O2 ≤70 mm Hg) in a patient with a normal chest radiograph suggested the diagnosis of PE and was seen in 9 of 17 (53%) with PE in PIOPED I and
BLUK077-Stein
March 28, 2007
CHAPTER 42
15:25
195
All patients irrespective of prior cardiopulmonary disease
Table 42.4 Signs in all patients with pulmonary embolism. PE n (%) PIOPED I (N = 383)
PIOPED II (N = 184 – 191)
Tachypnea (≥20/min)
278 (73)
108 (57)
Tachycardia (>100/min)
115 (30)
49 (26)
Signs General
Diaphoresis
38 (10)
8 (4)
Cyanosis
13 (3)
1 (1)
Temperature >38.5◦ C (>101.3◦ F)
27 (7)
3 (2)
Cardiac examination (any)
42 (22)
Increased P2
87 (23)
Third heart sound
21 (5)
22 (15)
Right ventricular lift
8 (5)
Jugular venous distension
25 (13)
Lung examination (any) Rales (crackles) Wheezes
70 (37) 210 (55)
40 (21)
44 (11)
6 (3)
Rhonchi
9 (5)
Decreased breath sounds Pleural friction rub
40 (21) 17 (4)
2 (1)
56 (15)
90 (47)
DVT Calf or thigh Calf only
65 (34)
Calf and thigh
23 (12)
Thigh only Homans’ sign
2 (1) 10 (3)
Number of patients with PE who had one or more signs of DVT: edema = 79, erythema = 7, tenderness = 36, palpable cord = 2. PE, pulmonary embolism; P2, pulmonary component of second sound; DVT, deep venous thrombosis. Data from Stein et al. [2, 3].
Table 42.5 Combinations of clinical characteristics in patients with pulmonary embolism. n (%) Signs
PIOPED I (N = 383)
PIOPED II (N = 188)
Dyspnea or tachypnea (≥20/min)
347 (91)
162 (86)
Dyspnea or tachypnea (≥20/min) or pleuritic pain
371 (97)
173 (92)
Dyspnea or tachypnea (≥20/min) or pleuritic pain or
373 (97)
182 (97)
signs of DVT Dyspnea or tachypnea (≥20/min) or pleuritic pain or signs of DVT or radiographic evidence of atelectasis or a parenchymal abnormality Data from Stein et al. [2, 3].
381 (99)
BLUK077-Stein
March 28, 2007
15:25
196
in 22 of 93 (24%) with no PE [5]. This combination draws attention to the likelihood of PE and suggests the necessity for further diagnostic studies. Neither this combination nor other combinations, however, were diagnostic. The signs and symptoms in patients with PE were not specific. The frequency of individual signs and symptoms in patients with PE in most instances did not differ statistically from signs and symptoms in patients in whom PE was suspected, but excluded. This is expected, because patients with suspected PE were identified for further diagnostic studies on the basis of signs and symptoms suggestive of PE. The nonspecific signs and symptoms as well as the nonspecific but important findings on the chest radiograph and electrocardiogram, particularly in a patient with risk factors for PE, and a necessary high level of awareness of the possibility of PE, may lead to further diagnostic tests and a correct diagnosis.
PART III
Diagnosis of acute PE
References 1 The PIOPED Investigators. Value of the ventilation/ perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 2 Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. 3 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, 2007 (In press). 4 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Associated laboratory and clinical findings. Circulation 1973; 47/48(suppl II): II81– II85. 5 Stein PD, Alavi A, Gottschalk A et al. Usefulness of noninvasive diagnostic tools for diagnosis of acute pulmonary embolism in patients with a normal chest radiograph. Am J Cardiol 1991; 67: 1117–1120.
BLUK077-Stein
March 28, 2007
43
15:25
CHAPTER 43
Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes
Introduction Many of the findings in the various syndromes of pulmonary embolism (PE) can be understood in terms of the severity of PE as it increases from mild with the pulmonary infarction syndrome to moderate with the isolated dyspnea syndrome to severe with circulatory collapse. Patients with submassive PE more commonly have the pulmonary infarction syndrome than patients with massive PE [1]. Patients with pulmonary infarction were shown to have less severe PE than patients with isolated dyspnea based on an objective pulmonary angiography scoring system [2]. Patients with circulatory collapse had the most severe PE based on the angiographic score, but the score was not statistically significantly higher than in patients with isolated dyspnea [2]. The prevalence of various clinical and laboratory characteristics of patients with the syndrome of pulmonary infarction, isolated dyspnea, or circulatory collapse may give clues to the diagnosis or suggest characteristics which may reduce the likelihood of inadvertently discarding the diagnosis of PE. Among patients with PE and no prior cardiopulmonary disease who survived long enough to undergo diagnostic evaluation, 65% in PIOPED I (Prospective Investigation of Pulmonary Embolism Diagnosis I) and 41% in PIOPED II presented with the syndrome of pulmonary hemorrhage or infarction characterized by pleuritic pain or hemoptysis [3, 4]. The syndrome of isolated dyspnea in the absence of circulatory collapse, pleuritic pain, or hemoptysis occurred in 22 and 36% in PIOPED I and PIOPED II [3, 4]. Circulatory collapse among patients who survived long enough for diagnostic evaluation was observed in only 8% with
PE and no prior cardiopulmonary disease in both PIOPED I and PIOPED II. However, approximately one third of patients with acute PE die within 2.5 hours [5]. The diagnosis of PE is occasionally made in asymptomatic patients with deep venous thrombosis (DVT) [3, 4] or, particularly in elderly patients, the diagnosis is suggested on the basis of unexplained radiographic abnormalities in asymptomatic patients [6]. The mean ages of patients in PIOPED I with pulmonary infarction syndrome, isolated dyspnea, and circulatory collapse were 51 ± 17 (mean ± SD), 55 ± 17, and 66 ± 18 years, respectively, which were not statistically significantly different [7] (Table 43.1).
Predisposing factors There were no statistically significant differences of predisposing factors between the three groups of presenting syndromes in PIOPED I [7] (Table 43.2). Absence of predisposing factors was observed in 13 and 15% of patients with pulmonary infarction or isolated dyspnea syndrome, respectively. Two of 5 patients with circulatory collapse had no predisposing factors. By definition none of these patients had prior cardiac or pulmonary disease, which, in some instances, may predispose them to PE.
Symptoms There were no statistically significant differences of symptoms between the groups of patients with PE and no prior cardiopulmonary disease in PIOPED I, excluding differences due to the defining characteristics [7] (Table 43.3). In patients with PE and no prior
197
BLUK077-Stein
March 28, 2007
15:25
Table 43.1 Age of patients in PIOPED I with pulmonary embolism and no prior cardiopulmonary disease. n (%)
Age (years) <21
Pulmonary infarction syndrome
Isolated dyspnea syndrome
(N = 119)
(N = 31)
1 (1)
1 (3)
21–30
20 (17)
3 (10)
31–40
15 (13)
3 (10)
41–50
16 (13)
4 (13)
51–60
27 (23)
6 (19)
61–70
27 (23)
7 (23)
71–80
9 (8)
7 (23)*
81–90
4 (3)
0 (0)
Patients with circulatory collapse ranged between age 31 and 90 years. * P < 0.02, pulmonary embolism vs. isolated dyspnea. Reprinted with permission from Stein and Henry [7]. Table 43.2 Predisposing factors for patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction syndrome
Isolated dyspnea syndrome
Predisposing factor
(N = 119)
(N = 31)
Immobilization
63 (53)
19 (61)
Surgery
61 (51)
14 (45)
Malignancy
23 (19)
4 (13)
Thrombophlebitis, ever
15 (13)
8 (26)
Trauma, lower extremity
22 (18)
2 (6)
Estrogen
11 (9)
4 (13)
Stroke
8 (7)
3 (10)
Postpartum ≤3 mo
5 (4)
2 (6)
None of the above
18 (15)
4 (13)
Only 5 patients had circulatory collapse. Two had no apparent predisposing factor. By definition, none of these patients had prior cardiopulmonary disease, which in some instances may predispose to PE. Reprinted with permission from Stein and Henry [7]. Table 43.3 Symptoms in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%)
Symptoms Dyspnea Pleuritic pain Cough
Pulmonary infarction
Isolated dyspnea
Circulatory collapse
syndrome (N = 119)
syndrome (N = 31)
syndrome (N = 5)
31 (100)* 0 (0)*
2 (40)
11 (35)
2 (40)
86 (72) 115 (97)* 52 (44)
0 (0)
Leg swelling
33 (28)
12 (39)
2 (40)
Leg pain
30 (25)
10 (32) 0 (0)*
2 (40)
5 (16)
0 (0)
Hemoptysis
25 (21)*
Palpitations
11 (9)
Wheezing
10 (8)
5 (16)
1 (20)
5 (4)
2 (6)
0 (0)
Angina-like pain
* Presence or absence of symptom defines syndrome. Reprinted with permission from Stein and Henry [7].
198
0 (0)
BLUK077-Stein
March 28, 2007
CHAPTER 43
15:25
199
Clinical characteristics according to presenting syndromes
cardiopulmonary disease in PIOPED I who had circulatory collapse, dyspnea was present in 2 of 5 (40%) [7], and in PIOPED II dyspnea was present in 9 of 11 (82%) [4]. Among all patients with PE and circulatory collapse in PIOPED II, dyspnea was present in 13 of 15 (87%) [4].
Signs Among patients with no prior cardiopulmonary disease in PIOPED I, rales were more prevalent in the pulmonary infarction group, 67 of 119 (56%), than in the isolated dyspnea group, 6 of 31 (19%) (P < 0.001) (Table 43.4). This was not, however, observed in PIOPED II in which 20% of the patients with the hemoptysis/pleuritic pain syndrome had rales compared with 26% in the isolated dyspnea syndrome [4]. Tachycardia in PIOPED I was more prevalent in patients with PE who had isolated dyspnea, 14 of 31 (45%), than in the pulmonary infarction group, 28 of 119 (24%) (P < 0.02). The frequency of tachycardia, however, was similar in both the groups in PIOPED II, 32 and 20%, respectively.
Combinations of signs and symptoms Among patients with pulmonary infarction syndrome and no prior cardiopulmonary disease in PIOPED I,
102 of 119 (86%) had dyspnea or tachypnea [7]. In PIOPED II, 41 of 51 (80%) had dyspnea or tachypnea [4]. Among patients with PE and circulatory collapse in PIOPED I and PIOPED II, dyspnea or tachypnea was not always present [4, 7]. Even dyspnea or tachypnea or pleuritic pain was not always present, but patients with circulatory collapse had one of these signs or symptoms or signs of DVT [4, 7].
Clinical assessment Physicians were confident of the diagnosis of PE (80– 100% clinical likelihood) in only 14–23% of the patients irrespective of the presenting syndrome. Physicians thought PE was probably absent (0–19% clinical likelihood) in 10% or fewer patients irrespective of the group.
Electrocardiographic manifestations A normal electrocardiogram (ECG) was more prevalent among patients with pulmonary infarction syndrome, 45 of 97 (46%), than among patients with isolated dyspnea syndrome, 2 of 21 (10%) (P < 0.01) (Table 43.5). Right bundle branch block was more
Table 43.4 Signs in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%)
Signs
Pulmonary infarction
Isolated dyspnea
Circulatory collapse
syndrome (N = 119)
syndrome (N = 31)
syndrome (N = 5)
Tachypnea (≥20/min)
84 (71)
20 (65)
2 (40)
Rales (crackles)
67 (56)*
6 (19)
1 (20)
Tachycardia (>100/min)
28 (24)†
14 (45)
3 (60)
Increased pulmonary component of sec-
32 (27)
10 (32)
0 (0)
ond sound Deep venous thrombosis
11 (9)
3 (10)
1 (20)
Diaphoresis
8 (7)
5 (16)
1 (20)
Temperature >38.5◦ C (>101.3◦ F)
7 (6)
2 (6)
0 (0)
Wheezes
8 (7)
2 (6)
0 (0)
Homans’ sign
2 (2)
2 (6)
1 (20)
Pleural friction rub
6 (5)
0 (0)
0 (0)
Third heart sound
2 (2)
2 (6)
0 (0)
Cyanosis
2 (2)
1 (3)
0 (0)
* P < 0.001, pulmonary infarction vs. isolated dyspnea. † P < 0.02, pulmonary infarction vs. isolated dyspnea. Reprinted with permission from Stein and Henry [7].
BLUK077-Stein
March 28, 2007
15:25
200
PART III
Diagnosis of acute PE
Table 43.5 Electrocardiographic manifestations in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction
Isolated dyspnea
Circulatory collapse
Manifestations
syndrome (N = 97)
syndrome (N = 21)
syndrome (N = 5)
Normal electrocardiogram
45 (46)*
2 (10)
1 (20)
ST-segment or T-wave changes
39 (40)
12 (57)
3 (60)
Left-axis deviation
13 (13)
3 (14)
0 (0)
Complete right bundle branch block
3 (3)
2 (10)
2 (40)†
Left ventricular hypertrophy
9 (9)
0 (0)
0 (0)
Incomplete right bundle branch block
4 (4)
2 (10)
0 (0)
Acute myocardial infarction pattern
2 (2)
1 (5)
0 (0)
Low-voltage QRS
3 (3)
1 (5)
1 (20)
P pulmonale
1 (1)
1 (5)
1 (20)
Right-axis deviation
1 (1)
0 (0)
0 (0)
Right ventricular hypertrophy
1 (1)
1 (1)
0 (0)
Some patients had more than one abnormality. * P < 0.01, pulmonary infarction vs. isolated dyspnea. † P < 0.001, pulmonary infarction vs. circulatory collapse. Reprinted with permission from Stein and Henry [7].
prevalent among patients with circulatory collapse than among patients with pulmonary infarction syndrome, 2 of 5 (40%) versus 3 of 97 (3%) (P < 0.001).
Plain chest radiograph The pulmonary infarction group tended to have a higher prevalence of atelectasis or pulmonary parenchymal abnormalities, 89 of 119 (75%), than the isolated dyspnea group, 16 of 31 (52%) (P < 0.02), and it had a higher prevalence than the circulatory collapse group, 1 of 5 (20%) (P < 0.01) (Table 43.6). The pulmonary infarction group also had a higher prevalence of pleural effusion, 67 of 119 (56%), than the isolated dyspnea group, 8 of 31 (26%) (P < 0.01), and it tended to have a higher prevalence than the circulatory collapse group, 0 of 5 (0%) (P < 0.02) (Table 43.6). The circulatory collapse group tended to have a higher prevalence of cardiomegaly, 2 of 5 (40%), than the pulmonary infarction group, 10 of 119 (8%) (P < 0.02) (Table 43.6).
creased and the Pa O2 decreased. The pulmonary infarction group had a higher mean Pa O2 (72.6 ± 14.2 mm Hg) than the isolated dyspnea group (63.2 ± 12.9 mm Hg) (P < 0.01) (Table 43.7). A Pa O2 greater than 80 mm Hg was observed in 27 of 99 (27%) patients with pulmonary infarction and 2 of 19 (11%) with isolated dyspnea. The pulmonary infarction group had a lower pulmonary artery mean pressure than the isolated dyspnea group (20.1 ± 8.1 mm Hg versus 24.5 ± 9.3 mm Hg (P < 0.02).
Ventilation–perfusion lung scan A high-probability ventilation–perfusion (V–Q) lung scan was more prevalent among the isolated dyspnea group than the pulmonary infarction group, 20 of 31 (65%) versus 38 of 119 (32%) (P < 0.001) (Table 43.8). Intermediate probability V–Q lung scans were most prevalent among patients with pulmonary infarction syndrome.
Pulmonary artery pressure, Pa O2 , Pa CO2 , and pH
Pulmonary infarction syndrome compared with isolated dyspnea syndrome
With increasing severity of PE, from pulmonary infarction to isolated dyspnea to circulatory collapse, trends suggest that the pulmonary artery mean pressure in-
Rales are more prevalent among patients with pulmonary infarction and tend to be less prevalent among patients with isolated dyspnea or circulatory collapse
BLUK077-Stein
March 28, 2007
CHAPTER 43
15:25
201
Clinical characteristics according to presenting syndromes
Table 43.6 Plain chest radiograph findings in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction
Isolated dyspnea
Circulatory collapse
Findings
syndrome (N = 119)
syndrome (N = 31)
syndrome (N = 5)
Atelectasis or pulmonary
89 (75)*
16 (52)
1 (20)†
Pleural effusion
67 (56)‡
8 (26)
0 (0)§
Pleural-based opacity
43 (36)
7 (23)
0 (0)
Elevated diaphragm
31 (26)
6 (19)
1 (20)
Decreased pulmonary vasculature
24 (20)
8 (26)
0 (0)
Prominent central pulmonary artery
17 (14)
5 (16)
0 (0)
Cardiomegaly
10 (8)
5 (16)
2 (40)§
7 (6)
2 (6)
0 (0)
Pulmonary edema
3 (3)
1 (3)
0 (0)
Normal chest X-ray
17 (14)
8 (26)
2 (40)
parenchymal abnormality
Westermark’s sign¶
* P < 0.02, pulmonary infarction vs. isolated dyspnea. † P < 0.01, pulmonary infarction vs. circulatory collapse. ‡ P < 0.01, pulmonary infarction vs. isolated dyspnea. § P < 0.02, pulmonary infarction vs. circulatory collapse. ¶ Prominent central pulmonary artery and decreased pulmonary vascularity. Reprinted with permission from Stein and Henry [7].
(Table 43.4). We previously showed that rales are associated with patients who have radiographic evidence of a parenchymal abnormality [3]. A normal ECG is frequently seen in patients with the pulmonary infarction syndrome, but a normal ECG is uncommon in patients with isolated dyspnea (Table 43.5). A Pa O2 higher than 80 mm Hg is not uncommon in patients with the pulmonary infarction syndrome,
but such levels are uncommon in patients with isolated dyspnea (Table 43.7). Abnormalities on the chest radiograph, although more common among patients with pulmonary infarction, are often observed in patients with isolated dyspnea (Table 43.6). A high-probability interpretation of the V–Q scan occurs in a minority of patients with the pulmonary infarction syndrome, but it was found
Table 43.7 Pulmonary artery mean pressure, Pa O2 , Pa CO2 , and pH (on room air) in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. Pulmonary infarction
Isolated dyspnea
Circulatory collapse
syndrome
syndrome
syndrome
Mean PA pressure (mm Hg)
20.1 ± 8.1* (n = 115)
24.5 ± 9.3 (n = 30)
25.8 ± 5.9 (n = 5)
Mean Pa O2 (mm Hg)
72.6 ± 14.2† (n = 99)
63.2 ± 12.9 (n = 19)
56.5 ± 2.7‡ (n = 4)
Mean Pa CO2 (mm Hg)
34.6 ± 5.2 (n = 99)
34.8 ± 4.3 (n = 19)
34.5 ± 4.7 (n = 4)
Mean pH
7.45 ± 0.04* (n = 99)
7.42 ± 0.08 (n = 19)
7.41 ± 0.07‡ (n = 4)
Values are mean ± standard deviation. * P < 0.02, pulmonary infarction vs. isolated dyspnea. † P < 0.01, pulmonary infarction vs. isolated dyspnea. ‡ P < 0.05, pulmonary infarction vs. circulatory collapse. PA, pulmonary artery; Pa O2 , partial pressure of oxygen in arterial blood; Pa CO2 , partial pressure of carbon dioxide in arterial blood. Reprinted with permission from Stein and Henry [7].
BLUK077-Stein
March 28, 2007
15:25
202
PART III
Diagnosis of acute PE
Table 43.8 Ventilation–perfusion lung scan in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction
Isolated dyspnea
Circulatory collapse
syndrome (N = 119)
syndrome (N = 31)
syndrome (N = 5)
38 (32)* 63 (53)*
20 (65)
4 (80)†
Intermediate
4 (13)
0 (0)†
Low
15 (13)
4 (13)
1 (20)
3 (3)
3 (10)
0 (0)
High
Near normal/normal
* P < 0.001, pulmonary infarction vs. isolated dyspnea. † P < 0.05, pulmonary infarction vs. circulatory collapse. Reprinted with permission from Stein and Henry [7].
in the majority of patients with isolated dyspnea (Table 43.8).
reduce the likelihood of inadvertently discarding the diagnosis.
References Circulatory collapse syndrome Some patients with circulatory collapse from PE, even though they have the most severe PE, are not short of breath. Based on pooled data from PIOPED I and PIOPED II, dyspnea was absent in 5 of 16 (31%) of patients with circulatory collapse who had PE and no prior cardiopulmonary disease [4, 7]. Among patients in shock in the Urokinase Pulmonary Embolism Trial, 6 of 21 (29%) did not have dyspnea, and among those with syncope, 2 of 19 (11%) did not have dyspnea [2]. Even frequent combinations of signs and symptoms may be absent in these patients with severe PE. Combined data from patients with PE and no prior cardiopulmonary disease in PIOPED I and PIOPED II showed that dyspnea and tachypnea were absent in 3 of 16 ( 19%) and the same number had an absence of dyspnea, tachypnea, and pleuritic pain [4, 7]. Patients with PE and circulatory collapse may have a normal radiograph (Table 43.6) and they may have a low-probability interpretation of the V–Q scan (Table 43.8). Awareness that some typical characteristics of PE may be absent, even with severe PE, may
1 A National Cooperative Study: the Urokinase Pulmonary Embolism Trial. Associated clinical and laboratory findings. Circulation 1973; 47(suppl II): II-51–II-85. 2 Stein PD, Willis PW, III, DeMets DL. History and physical examination in acute pulmonary embolism in patients without preexisting cardiac or pulmonary disease. Am. J Cardiol 1981; 47: 218–223. 3 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 4 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, 2007 (In press). 5 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 6 Stein PD, Gottschalk A, Saltzman HA, Terrin ML. Diagnosis of acute pulmonary embolism in the elderly. J Am Coll Cardiol 1991; 18: 1452–1457. 7 Stein PD, Henry JW. Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes. Chest 1997; 112: 974–979.
March 11, 2007
44
7:53
CHAPTER 44
Clinical assessment in the critically ill
Introduction Autopsy studies have consistently shown that pulmonary embolism (PE) is frequently not considered antemortem in patients who died from PE (see Chapter 1). It is apparent, therefore, that the clinical diagnosis of PE in critically ill patients is difficult. The clinical characteristics in several categories of critically ill patients were evaluated in order to identify features that may alert physicians to the diagnosis [1].
Ventilatory support, intensive care units, hypoxemic not on ventilatory support The clinical diagnosis of acute PE in an intensive care unit is generally considered to be exceedingly difficult [2]. A clinical assessment of the likelihood of PE was based on bedside evaluation, the plain chest radiograph, electrocardiogram, and ordinary laboratory tests [3]. Physicians were uncertain of whether PE was present in 67–78% of critically ill patients [1]. In noncritically ill patients, physicians were uncertain whether PE was present in 67%. Among patients on ventilatory support, patients in intensive care units not on ventilatory support, and hypoxemic patients not on ventilatory support, when physicians assessed an 80–100% likelihood of PE, 75–88% of patients in fact had PE [1]. This was comparable to the positive predictive value in noncritically ill patients (77%). There were few critically ill patients in whom physicians were confident enough to make a high-likelihood clinical assessment. A high-likelihood clinical assessment was made in only 9% of the critically ill patients. However, the proportion of patients with a high-likelihood clinical assessment in noncritically ill patients was comparable, 8%. If physicians assessed a 0–19% clinical likelihood of PE in these critically ill patients, only 0– 5% had PE. The number of critically ill patients with a
low-likelihood clinical assessment was also small. The frequency of PE in noncritically ill patients with a lowlikelihood clinical assessment was 15%. After making a clinical assessment, physicians in Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) were uncertain whether PE was present in 63–78% of patients with no prior cardiopulmonary disease [4], any prior cardiopulmonary disease (P.D. Stein and J.W. Henry, unpublished data from PIOPED, 1990), chronic obstructive pulmonary disease (COPD) [5], patients in intensive care units not on ventilatory support [1], patients on ventilatory support [1], and hypoxemic patients not on ventilatory support [1] (Figure 44.1). However, physicians were more often uncertain of whether PE was present in making a clinical assessment among patients in an intensive care unit not on ventilatory support than among patients with no prior cardiopulmonary disease. The percent of patients with an uncertain clinical assessment in other categories did not differ from each other to a statistically significant extent.
80
Patients (%)
BLUK077-Stein
60 40 20 0 NO CPD (n = 320)
ANY CPD (n = 478)
COPD (n = 100)
ICU NO VENT (n = 85)
VENT (n = 46)
Figure 44.1 Percent of patients in whom clinical assessment was uncertain regarding the likelihood of pulmonary embolism. NO CPD, no prior cardiopulmonary disease; ANY CPD, any prior cardiopulmonary disease; COPD, chronic obstructive pulmonary disease; ICU NO VENT, intensive care unit not on ventilatory support; VENT, ventilatory support. NO CPD versus ICU NO VENT, P < 0.01. All other differences were not statistically significant.
203
BLUK077-Stein
March 11, 2007
7:53
Pulmonary embolism (%)
204
PART III
Table 44.1 Positive predictive value of lung scan interpretation and clinical assessment in patients on ventilatory support.
100 80 60 40 20 0 NO CPD (n = 17)
ANY CPD (n = 22)
COPD (n = 3)
ICU NO VENT (n = 8)
VENT (n = 4)
Figure 44.2 Frequency of pulmonary embolism among patients in whom a high-likelihood clinical assessment was made. Abbreviations are as in Figure 44.1. There were no statistically significant differences in the frequency of pulmonary embolism between the various groups.
40 Pulmonary embolism (%)
Diagnosis of acute PE
30 20 10 0 NO CPD (n = 103)
ANY CPD (n = 123)
COPD (n = 30)
ICU NO VENT (n = 11)
VENT (n = 11)
Figure 44.3 Frequency of pulmonary embolism among patients in whom the clinical assessment indicated a low likelihood of pulmonary embolism. Abbreviations are as in Figure 44.1. Pulmonary embolism was more frequent among patients with no cardiopulmonary disease (CPD) than with any CPD (P < 0.05). Other differences were not statistically significant.
V–Q scan
Total PE+/Test+ (%)
High
4/4 (100)
Intermediate
6/24 (25)
Low
2/17 (12)
Near normal/normal
0/1 (0)
Total
12/46 (26)
PE, pulmonary embolism. Reprinted with permission from Henry et al. [1].
Among patients with no prior cardiopulmonary disease, any prior cardiopulmonary disease, COPD, patients in intensive care units not on ventilatory support, patients on ventilatory support, and hypoxemic patients not on ventilatory support, if clinical assessment indicated a high likelihood of PE, it was present in 75% or more of the patients in each of these categories [1, 4–6] (Figure 44.2). Few patients, however, had a high-likelihood clinical assessment among those with COPD, in the intensive care unit not on ventilatory support, and patients on ventilatory support. Among patients in whom physicians made a clinical assessment of PE as being low likelihood, PE was present in 0–14% of the patients in the various categories (Figure 44.3).
Ventilation–perfusion lung scans Most patients on ventilatory support, 41 of 46 (89%), showed nondiagnostic (intermediate or low probability) ventilation–perfusion lung scans (Table 44.1, Figure 44.4).
Patients (%)
80 60 40 20
References
0 NO CPD (n = 365)
ANY CPD (n = 526)
COPD (n = 98)
ICU NO VENT (n = 89)
VENT (n = 46)
Figure 44.4 Percent of patients with intermediate probability ventilation–perfusion (V–Q) lung scans. Abbreviations are as in Figure 44.1. Fewer patients with no prior CPD had intermediate probability V–Q scans than any other category (P < 0.01–0.001). Patients with COPD had intermediate probability V–Q scans more frequently than any other group (P < 0.01–0.001) except in patients on ventilatory support.
1 Henry JW, Stein PD, Gottschalk A, Relyea B, Leeper KV, Jr. Scintigraphic lung scans and clinical assessment in critically ill patients with suspected acute pulmonary embolism. Chest 1996; 109: 462–466. 2 Cowen JC, Kelley MA. An organized approach to detecting pulmonary embolism in the critically ill. J Crit Illn 1994; 9: 551–567. 3 The PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary
BLUK077-Stein
March 11, 2007
CHAPTER 44
7:53
Clinical assessment in critically ill
Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753– 2759. 4 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing
205
cardiac or pulmonary disease. Chest 1991; 100: 598– 603. 5 Lesser BA, Leeper KV, Stein PD et al. The diagnosis of pulmonary embolism in patients with chronic obstructive pulmonary disease. Chest 1992; 102: 17–22.
BLUK077-Stein
March 12, 2007
45
16:20
CHAPTER 45
The electrocardiogram
In 1935, McGinn and White described the electrocardiograms of 7 patients with acute cor pulmonale secondary to pulmonary embolism (PE) [1]. They described the presence of a Q wave and late inversion of the T wave in lead III and a low origin of the T wave with a gradual staircase ascent of the ST segment in lead II. They also noted an S wave in lead I. Their description of the S1 Q3 T3 complex became a traditional electrocardiographic manifestation of acute cor pulmonale. This observation was expanded by Murnaghan and associates [2] in 1943, with the report of 10 additional patients with acute cor pulmonale and no prior heart disease. An S1 Q3 T3 pattern was again observed, sometimes associated with right-axis deviation. Some of the patients showed a depressed ST segment; others showed ST-segment elevation [2]. However, even in the presence of preexistent cardiac disease, these electrocardiographic indications of acute cor pulmonale were present in only about one-third of the patients. Based upon studies of patients and of dogs, Love and associates, in 1938, concluded that the typical electrocardiogram of PE was ST-segment depression and T-wave inversion [3]. Transient right bundle branch block in PE was reported in 1939 [4]. Wood [5], and subsequently others [6–9], reported sharp T-wave inversion in the precordial leads facing the right ventricle. Clockwise rotation was described by Wilson and associates in 1947 [10] and was confirmed subsequently [9]. A shift of the transitional zone to the left and evidence of incomplete right bundle branch block was found helpful in distinguishing the prominent Q waves of PE from those of myocardial infarction [7–11]. P pulmonale in acute PE was described by Katz in 1946 [12]. By 1940, Sokolow and associates stated that no single electrocardiographic abnormality is consistently present in PE [13]. In a study of 50 patients, they found that a large percentage had either normal electrocardiograms or nonspecific changes [13]. The distinctive pattern described by McGinn and White occurred only in 10% of the patients [1].
206
The prevalence of various electrocardiographic abnormalities in 90 patients with massive or submassive PE and in 89 patients with mild to massive PE, all of whom had no prior cardiopulmonary disease, is shown in Table 45.1 [14, 15]. Definitions of various electrocardiographic abnormalities are shown in Table 45.2 [14]. The electrocardiographic abnormalities observed in patients with no prior cardiopulmonary disease who had submassive or massive PE were comparable to the electrocardiographic changes observed in patients with no prior cardiopulmonary disease in whom the severity of the PE ranged from mild to severe (Table 45.1) [15]. Definitions of electrocardiographic changes are shown in Table 45.2 [16–21]. A normal electrocardiogram was shown in only 6% of the patients with massive acute PE , whereas in those with submassive acute PE , it was normal in 23%. Although this would appear to suggest a trend, the difference was not statistically significant. In patients with mild to massive PE, a normal electrocardiogram was shown in 30% [15]. Cutforth and Oram reported 50 patients with PE, 49 of whom had patent coronary arteries and 24% of whom had normal electrocardiograms [19]. None of the patients who had massive or submassive PE with no prior cardiopulmonary disease had atrial flutter or atrial fibrillation [14], and only 5% of the patients with no prior cardiopulmonary disease in whom the severity of the PE ranged from mild to severe had atrial fibrillation or atrial flutter [15]. None had second- or third-degree atrioventricular block [14, 15]. Only 1 patient had first-degree atrioventricular block [14, 15]. Among 23 patients with acute PE and no heart disease reported by Szucs and associates, only 4% (1 of 23) had atrial flutter or atrial fibrillation [22]. This observation led Dalen to conclude that atrial flutter or atrial fibrillation in patients with PE is nearly always limited to individuals with prior heart disease [23]. Right-axis deviation in PE has been observed for many years [1, 11]. Electrocardiographic
BLUK077-Stein
March 12, 2007
CHAPTER 45
16:20
207
Electrocardiogram
Table 45.1 Electrocardiographic manifestations: patients without prior cardiac or pulmonary disease. Pulmonary embolism (%) Electrocardiogram
Mild to massive (n = 89)*
Submassive to massive (n = 90)†
Normal electrocardiogram
30
13
Atrial flutter
1
0
Atrial fibrillation
4
0
Atrial premature contractions
4
2
Ventricular premature contractions
4
3
2
6
Rhythm disturbances
P wave P pulmonale QRS abnormalities 2
7
Left-axis deviation
Right-axis deviation
13
7
Clockwise rotation (V5 )
—
7
S1 S2 S3
—
7
S1 Q3 T3
—
12
Incomplete right bundle branch block
4
6
Complete right bundle branch block
6
9
Right ventricular hypertrophy
2
6
Pseudoinfarction
3
11
Low voltage (frontal plane)
3
6
ST segment and T wave Nonspecific T wave
—
42
ST-segment depression
—
26
ST segment elevation
—
16
Nonspecific ST segment or T wave
49
—
Some patients had more than one abnormality. ∗ Data from Stein et al. [15]. † Data from Stein et al. [14].
manifestations of acute cor pulmonale (S1 Q3 T3 , complete right bundle branch block, P pulmonale, or right-axis deviation) were less common than ST-segment or T-wave changes (Table 45.1) [14]. One or more of these abnormalities occurred in 26% of the patients with submassive or massive acute PE not associated with cardiac or pulmonary disease and occurred in 32% of such patients who suffered massive PE. Cutforth and Oram reported that 14% of the patients with PE had right bundle branch block and 28% had an S1 Q3 T3 [19]. A pseudoinfarction pattern was seen in 11% of the patients with submassive or massive PE who had no prior cardiopulmonary disease [14]. A pseudoinfarction pattern, however, was seen in only 3% of the patients with no prior cardiopulmonary disease who had PE that ranged in severity from mild to massive [15]. The problem of distinguishing the prominent Q waves
known to occur with PE from those of myocardial infarction has been recognized for many years [7, 9]. The electrocardiogram may simulate an inferior infarction with Q waves and T-wave inversion in leads II, III, and aVF [7, 24] or anteroseptal infarction characterized by QS or QR waves in V1 and T-wave inversion in the right precordial leads [19, 20]. The development of Q waves and extensive T-wave inversion in the anterior and lateral leads was also observed in 1 patient following PE [25]. Leftward shifts of the frontal plane axis in PE are frequent [26, 27]. Karlen and Wolff noted a shift of the frontal plane axis toward the horizontal position in 4 patients with PE [24]. An example of a patient with a previously normal frontal plane QRS axis that showed a shift to −25◦ following an acute PE and subsequently reverted to normal is shown in Figure 45.1 [26]. In a review of 115 patients with acute PE, a leftward shift
BLUK077-Stein
March 12, 2007
16:20
208
PART III
Diagnosis of acute PE
Table 45.2 Definitions of the various electrocardiographic abnormalities. Abnormalities of the P wave Right atrial enlargement. The P wave was considered to be suggestive of right atrial enlargement if it was 0.25 mV (2.5 mm) in the extremity leads or over 0.15 mV in lead V1 [16, 17]. Abnormalities of the QRS complex Right-axis deviation. Right-axis deviation was defined as a mean frontal plane QRS electrical axis greater than 90◦ [17]. Left-axis deviation. This was defined as a mean frontal plane QRS axis equal to or leftward of −30◦ . The A QRS was measured by the simplified method of Grant [18]. Many patients with marked left-axis deviation undoubtedly had left anterior fascicular block, but this was not indicated as a separate interpretation. S1 S2 S3 pattern. This was defined as the occurrence of S waves in leads I, II, and III of at least 0.15 mV (1.5 mm) amplitude in each lead. Clockwise rotation. This was defined as a shift in the transition zone (R = S) in the precordial leads. Usually a shift to the V4 position or further leftward is considered as clockwise rotation [19, 20]. For the purpose of this study, a shift in the transition zone to V5 was analyzed to avoid problems of interpretation that may occur with minor errors of precordial electrode position. Right bundle branch block [19, 20]. Incomplete right bundle branch block was diagnosed when the QRS duration was 0.10–0.11 s and when the terminal QRS forces were directed rightward and anteriorly causing the production of S waves in leads I and R waves in lead V1 . Complete right bundle branch block. This was considered to be present in those patients with a QRS duration of 0.12 s or greater and with terminal QRS forces directed as previously described. S1 Q3 T3 pattern [1, 19, 20]. The presence of S waves in lead I and Q waves in lead III, each of amplitude exceeding 0.15 mV (1.5 mm), associated with inversion of the T wave in lead III constituted the S1 Q3 T3 pattern. Right ventricular hypertrophy [17, 21]. Right ventricular hypertrophy was considered to be present if the R wave in lead V1 exceeded 0.5 mV (5 mm) or the R/S ratio in lead V1 was greater than 1. In this study, right-axis deviation was not required to make this diagnosis nor was the presence of right-axis deviation alone considered adequate for the diagnosis of right ventricular hypertrophy. Low-voltage QRS complexes. The QRS complexes were considered to be low voltage if the greatest overall deflection of the QRS was 0.5 mV (5 mm) or less in all of the limb leads [17]. Pseudoinfarction pattern. This was defined as ST-segment or T-wave abnormalities associated with prominent Q waves, which simulated a recent myocardial infarction. ST-segment depression. This was considered present if the ST segment was depressed 0.05 mV (0.5 mm) or greater in any lead except aVR. No distinction was made between junctional or ischemic ST-segment depression. Changes were considered primary in the absence of complete bundle branch block or ventricular hypertrophy. ST-segment elevation. This was recorded as present when found in any lead except aVR if the ST segment was elevated at least 0.1 mV (1 mm) in the absence of ST depression in any other lead. T-wave inversion. This was considered to be present if it occurred in any lead except aVL, III, aVR, or V1 . Changes were considered primary in the absence of complete bundle branch block or ventricular hypertrophy. Reprinted from Stein et al. [14], with permission from Elsevier.
of the frontal plane QRS axis was observed more frequently than a rightward shift, and left-axis deviation was present more often than right-axis deviation [27]. In patients with submassive or massive PE and no prior cardiopulmonary disease, left-axis deviation was recorded as frequently as right-axis deviation (Table 45.1) [14]. Among patients with PE who had no prior cardiopulmonary disease in PIOPED, left-axis deviation was more frequent than right-axis deviation (Table 45.1) [15]. Emphysema is also a cause of left-axis deviation [28]. Since the electrocardiogram of emphy-
sema is usually easily recognized, the term pseudo leftaxis deviation has been termed for such individuals to distinguish the electrocardiogram from that of left ventricular conduction abnormalities. Abnormal conduction through the emphysematous lungs has been postulated as a possible cause of the left-axis deviation in these patients [28]. Low-voltage frontal plane QRS complexes in PE was first observed by Stein and associates [14]. It was observed in 6% of the patients with massive or submassive PE who did not have prior cardiopulmonary
BLUK077-Stein
March 12, 2007
CHAPTER 45
16:20
209
Electrocardiogram
(a)
(b)
Figure 45.1 (a) Electrocardiogram 1 week prior to acute pulmonary embolism. Maximal frontal plane QRS axis, +70◦ , is within normal limits. (b) Electrocardiogram on the day of hospitalization for acute pulmonary embolism. The frontal plane QRS axis has shifted leftward to −25◦ . Nonspecific ST-segment and T-wave changes are also present. (c) Electrocardiogram 5 days after transvenous embolectomy shows return of frontal plane QRS axis toward control value. Frontal plane QRS axis is now +15◦ . (Reprinted from Stein and Bruce [26], with permission from Elsevier.)
(c)
BLUK077-Stein
March 12, 2007
16:20
210
PART III
Diagnosis of acute PE
Figure 45.2 (a) Electrocardiogram on the day of diagnostic pulmonary angiogram showing low-voltage QRS complexes in frontal plane. (b) Electrocardiogram 13 days later. Amplitude of QRS complexes is normal. (Reprinted from Stein et al. Bruce [14], with permission from Elsevier.)
disease (Table 45.1) [14]. Among patients in Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) with mild to massive PE who had no prior cardiopulmonary disease, 3% had low-voltage QRS complexes [15]. An electrocardiogram showing lowvoltage frontal plane QRS complexes that subsequently resolved is shown in Figure 45.2. Right ventricular endocardial potential is reduced in massive PE, possibly reflecting acute myocardial stress [29]. In 1938, Love and associates considered changes of the ST segment and T wave to be “the significant electrocardiographic changes noted” in acute PE [3]. Abnormalities of the ST segment and T wave are by far the most frequent electrocardiographic manifestation of PE [14, 15]. Nonspecific T-wave changes occurred in 42% of the patients with massive or submassive embolization [14]. A similar number of patients (41%) showed either depression or elevation of the ST seg-
ment [14]. Nonspecific ST-segment or T-wave changes were observed in 49% of the patients who had no prior cardiopulmonary disease and in whom the severity of PE ranged from mild to severe [15]. Inversion of the T wave was also the most common electrocardiographic abnormality reported by Cutforth and Oram (46%) [19]. They observed depression of the ST segment in 18% of the patients with PE [19]. As early as 1935, the electrocardiographic abnormalities in acute PE were known to be transient and in some instances lasted only 48 hours [1]. Among patients with submassive or massive PE and no prior cardiopulmonary disease, inversion of the T waves was the most persistent electrocardiographic abnormality [14]. Abnormalities of the QRS complex appeared to be more transient [15]. The duration of electrocardiographic abnormalities following acute PE is shown in Table 45.3 [14]. (The
BLUK077-Stein
March 12, 2007
CHAPTER 45
16:20
211
Electrocardiogram
Table 45.3 Disappearance of electrocardiographic abnormalities following acute massive or submassive pulmonary embolism. Days after diagnostic pulmonary angiography Electrocardiographic abnormality*
5 or 6
14
P-wave abnormality* P pulmonale
80% (4 of 5)
80% (4 of 5)
Right-axis deviation
60% (3 of 5)
60% (3 of 5)
Left-axis deviation
50% (3 of 6)
50% (3 of 6)
Clockwise rotation (V5 )
67% (4 of 6)
67% (4 of 6)
Incomplete right bundle branch block
80% (4 of 5)
100% (5 of 5)
QRS abnormalities
Complete right bundle branch block
75% (6 of 8)
75% (6 of 8)
Right ventricular hypertrophy
60% (3 of 5)
80% (4 of 5)
S 1 S2 S3
67% (4 of 6)
80% (4 of 5)
S 1 Q3 T3
100% (11 of 11)
100%(11 of 11)
67% (6 of 9)
67% (6 of 9)
0% (0 of 4)
67% (2 of 3)
Pseudoinfarction Low voltage (frontal plane) Primary RST-segment and T-wave abnormalities RST-segment depression (not reciprocal)
48% (11 of 23)
74% (17 of 23)
RST-segment elevation (not reciprocal)
31% (4 of 13)
14% (2 of 14)†
T-wave inversion
22% (8 of 37)
49% (18 of 37)
*Some patients had more than one abnormality. † Recurrence noted in 2 patients. Reprinted from Stein et al. [14], with permission from Elsevier.
day of diagnostic arteriography and baseline electrocardiogram is considered as day 0.) The acute PE, by history, may have occurred as many as 5 days previously. Abnormalities of repolarization tended to persist. Inversion of the T wave disappeared in only 22% of the patients 5 or 6 days after the PE was diagnosed by angiography, although it resolved in 49% by 2 weeks [14]. Depression of the ST segment tended to resolve somewhat faster. In contradistinction to repolarization, many of the abnormalities of depolarization disappeared faster. In well over half of the patients with pseudoinfarction, S1 S2 S3 , S1 Q3 T3 , right ventricular hypertrophy, or right bundle branch block, these abnormalities were no longer apparent on the electrocardiograms 5 or 6 days after the angiographic diagnosis. Examples of the occurrence and disappearance of abnormal left-axis deviation (left anterior fascicular block), low-voltage QRS complexes in the frontal plane, pseudoinfarction, and apparent acute ischemic event are shown in Figures 45.1–45.4, respectively. Electrocardiographic abnormalities of acute massive or submassive pulmonary embolization were frequently related to the severity of the embolization as
indicated by lung scans or pulmonary arteriography [14]. Patients with ST-segment abnormalities, T-wave inversion, pseudoinfarction patterns, S1 Q3 T3 patterns, incomplete right bundle branch block, right-axis deviation, right ventricular hypertrophy, or ventricular premature beats had statistically significantly larger perfusion defects on the lung scan or larger defects on the pulmonary arteriogram than those with normal electrocardiograms [14]. Elevation of the pulmonary arterial mean pressure or right ventricular end-diastolic pressure was often associated with electrocardiographic abnormalities. In patients with right-axis deviation, incomplete right bundle branch block, S1 S2 S3 pattern, pseudoinfarction pattern, ST-segment depression, or T-wave inversion, either the pulmonary arterial mean pressure or right ventricular end-diastolic pressure or both were significantly higher than in patients who had normal electrocardiograms [14]. The partial pressure of oxygen in arterial blood (Pa O2 ) in patients with various electrocardiographic abnormalities, with the exception of patients with right-axis deviation, did not differ significantly from
BLUK077-Stein
March 12, 2007
16:20
212
PART III
Diagnosis of acute PE
Figure 45.3 (a) Electrocardiogram on the day of pulmonary angiogram showing Q waves in II, II, aVF, and V1 simulating recent inferior and anteroseptal myocardial infarction. Pulmonary artery pressure was 63/24 mm Hg and Pa O2 was 63 mm Hg. (b) Electrocardiogram next day shows no significant Q waves. Inversion of T waves persists. Pulmonary artery pressure decreased to 32/12 mm Hg and Pa O2 increased to 89 mm Hg. (Reprinted from Stein et al. [14], with permission from Elsevier.)
the depressed levels found in patients who had normal electrocardiograms [14]. Even so, a contribution of hypoxemia to the electrocardiographic changes cannot be excluded. Neither right ventricular work nor the total pulmonary resistance of patients with abnormal electrocardiograms differed significantly from values measured in patients with normal electrocardiograms. The product of heart rate and pulmonary arterial peak pressure also showed no difference between groups. Peak brachial arterial pressure was 90 mm Hg or higher in all but 3 patients (96%). The magnitude of the T-wave inversion was not related to any of the measured hemodynamic abnormalities or to the severity of PE [14]. Patients in whom the
T wave was inverted, 0.05–0.1 mV, had hemodynamic, angiographic, and perfusion abnormalities comparable to patients in whom the T waves were inverted more than 0.3 mV in the frontal plane. The electrocardiographic changes in PE have been postulated to be related to several etiologic mechanisms. Since some of the electrocardiographic abnormalities simulate myocardial ischemia, and even infarction, it has been suggested that myocardial underoxygenation is the cause [30]. The right ventricle would be especially embarrassed by myocardial ischemia because of the load placed upon it by obstruction of the pulmonary circulation. The main cardiac effects of PE are (1) strain upon the right ventricle and
BLUK077-Stein
March 12, 2007
CHAPTER 45
16:20
Electrocardiogram
213
Figure 45.4 (a) 56-year-old man with acute pulmonary embolism shown by CT angiography. Coronary angiogram on same day was normal. Electrocardiogram simulates acute ischemic event showing prominent ST-segment elevation in leads V1 and V2 . (b) Electrocardiogram next day no longer shows ST-segment elevation. (Courtesy of Elias Skaf, MD.)
(2) myocardial ischemia [30]. The latter has been postulated to be due to shock, hypoxemia, or vagal reflexes from the lung to the coronary arteries [30]. Myocardial necrosis, usually involving the left ventricle, has been observed at autopsies of patients who died of PE and had normal coronary arteries [30, 31]. Because studies of left coronary flow [32] and right coronary blood flow [33] following experimentally induced acute PE in dogs and pigs showed an increased coronary flow (Figure 45.5), rather than a reduction, reflex spasm of the coronary arteries appears unlikely. Shock also appears to be an unlikely cause, because electrocardiographic changes frequently occur in the absence of shock [14]. An effect of hypoxemia on the myocardium cannot be excluded. Such an effect, however, would be secondary, since the partial pressure of oxygen in arterial blood (Pa O2 ) in patients with abnormal electrocardiograms, in general, did not differ from that of patients with normal electrocardiograms [14]. The possibility also exists that some patients may have had asymptomatic coronary heart disease. In the
presence of acute PE and hypoxemia, such individuals would not be able to increase coronary flow adequately to compensate for the underoxygenation of arterial blood. Love and associates concluded that dilatation of the right ventricle produced the electrocardiographic changes in PE [3]. Mechanical obstruction of the pulmonary artery produced electrocardiographic changes similar to those produced by PE. The electrocardiographic abnormalities were preceded by visible and marked dilatation of the right ventricle [3]. Section of the vagi or cervical sympathetic cords had no effect on the production of the electrocardiographic changes [3]. Wood also concluded that the essential factor in the production of electrocardiographic abnormalities in PE was right ventricular stress [5]. Electrocardiographic changes appeared to depend on the size of the PE [5]. Occlusion of the pulmonary artery of human beings during thoracic surgery produced some of the electrocardiographic changes observed in PE, thereby supporting this contention [34].
BLUK077-Stein
March 12, 2007
16:20
214
PART III
Diagnosis of acute PE
Figure 45.5 Simultaneous mean left circumflex coronary artery blood flow (LCBF), femoral artery mean pressure (FAM), pulmonary artery mean pressure (PAM), right atrial mean pressure (RAM), and electrocardiogram (ECG) in a dog during the administration of clot pulmonary embolism (PE). The LCBF was measured with an electromagnetic flow transducer placed around the artery. The LCBF increased from a control value of 88 to 210 cm3 /min 1 minute after PE. The PAM increased to 72 mm Hg 25 seconds after PE and decreased to 53 mm Hg 1 minute after PE. There was a transient reduction of FAM and a transient elevation of RAM. (Reprinted from Stein et al. [32], with permission from Elsevier.)
Occasional patients with a Pa O2 as low as 51 mm Hg and mean pulmonary arterial pressures as high as 31 mm Hg had normal electrocardiograms [14]. Even so, most of the patients with electrocardiographic abnormalities had more severe evidence of PE as indicated by the lung scan or pulmonary angiogram than patients with normal electrocardiograms. The frequent association of electrocardiographic abnormalities with an elevated pulmonary arterial mean pressure or elevated right ventricular end-diastolic pressure supports the hypothesis that ventricular dilatation represents the major cause of the electrocardiographic changes. Arterial hypoxemia did not correlate with the electrocardiographic changes, although the Pa O2 was low in many patients and may have contributed to the observed abnormalities [14].
References 1 McGinn S, White PD. Acute cor pulmonale resulting from pulmonary embolism. JAMA 1935; 104: 1473–1480. 2 Murnaghan D, McGinn S, White PD. Pulmonary embolism with and without acute cor pulmonale, with special reference to the electrocardiogram. Am Heart J 1943; 25: 573–597. 3 Love WS, Jr, Brugler GW, Winslow N. Electrocardiographic studies in clinical and experimental pulmonary embolization. Ann Intern Med 1938; 11: 2109–2123.
4 Durant TM, Ginsburg IW, Roesler H. Transient bundle branch block and other electrocardiographic changes in pulmonary embolism. Am Heart J 1939; 17: 423–430. 5 Wood P. Pulmonary embolism: diagnosis by chest lead electrocardiography. Br Heart J 1941; 3: 21–29. 6 Myers GB, Klein HA, Stoffer BE. The electrocardiographic diagnosis of right ventricular hypertrophy. Am Heart J 1948; 35: 1–40. 7 Phillips E, Levine HD. A critical evaluation of extremity and precordial electrocardiography in acute cor pulmonale. Am Heart J 1950; 39: 205–216. 8 Wood P. Electrocardiographic appearances in acute and chronic pulmonary heart disease. Br Heart J 1948; 10: 87. 9 Eliaser M, Jr, Giansiracusa F. The electrocardiographic diagnosis of acute cor pulmonale. Am Heart J 1952; 43: 533–545. 10 Wilson FN, Rosenbaum FF, Johnston FD. Interpretation of the ventricular complex of the electrocardiogram. Adv Intern Med 1947; 2: 1–63. 11 Kuo PT, Vander Veer JB. Electrocardiographic changes in pulmonary embolism with special reference to an early and transient shift of the electrical axis of the heart. Am Heart J 1950; 40: 825–838. 12 Katz LN. Electrocardiography, 2nd edn. Lea & Febiger, Philadelphia, 1946: 401–406. 13 Sokolow M, Katz LN, Muscovitz AN. The electrocardiogram in pulmonary embolism. Am Heart J 1940; 19: 166– 184. 14 Stein PD, Dalen JE, McIntyre KM, Sasahara AA, Wenger NK, Willis PW, III. The electrocardiogram in acute
BLUK077-Stein
March 12, 2007
CHAPTER 45
15
16
17 18 19 20
21
22
23
24
16:20
Electrocardiogram
pulmonary embolism. Prog Cardiovasc Dis 1975; 17: 247–257. Stein PD, Terrin ML, Hales CA et al.. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Chest 1991; 100: 598–603. Thomas P, Dejong D. The P wave in the electrocardiogram in the diagnosis of heart disease. Br Heart J 1954; 16: 241– 254. Friedberg CK. Diseases of the Heart, 3rd edn. Saunders, Philadelphia, 1966: 38, 171, 188. Grant RP. Clinical Electrocardiography. McGraw-Hill, New York, 1957: 13. Cutforth RH, Oram S. The electrocardiogram in pulmonary embolism. Br Heart J 1958; 20: 41–60. Weber DM, Phillips JH, Jr. A re-evaluation of electrocardiographic changes accompanying acute pulmonary embolism. Am J Med Sci 1966; 251: 381–398. Milnor WR. Electrocardiogram and vectorcardiogram in right ventricular hypertrophy and right bundle branch block. Circulation 1957; 16: 348–367. Szucs MM, Jr, Brooks HL, Grossman W et al. Diagnostic sensitivity of laboratory findings in acute pulmonary embolism. Ann Intern Med 1971; 74: 161–166. Dalen JE. Diagnosis of acute pulmonary embolism. In: Dalen JE, ed. Pulmonary Embolism., Medcom, New York, 1972: 28–39. Karlen WS, Wolff L. The vectorcardiogram in pulmonary embolism. II. Am Heart J 1956; 51: 839–860.
215
25 Romhilt D, Susilavorn B, Chou T. Unusual electrocardiographic manifestation of pulmonary embolism. Am Heart J 1970; 80: 237–241. 26 Stein PD, Bruce TA. Left axis deviation as an electrocardiographic manifestation of acute pulmonary embolism. J Electrocardiol 1971; 4: 67–69. 27 Lynch RE, Stein PD, Bruce TA. Leftward shift of frontal plane QRS axis as a frequent manifestation of acute pulmonary embolism. Chest 1972; 61: 443–446. 28 Pryor R, Blount SG, Jr. The clinical significance of true left axis deviation. Left intraventricular blocks. Am Heart J 1966; 72: 391–413. 29 Chatterjee K, Sutton GC, Miller GAH. Right ventricular endocardial potential in acute massive pulmonary embolism. Br Heart J 1972; 34: 271–273. 30 Horn H, Dack S, Friedberg CK. Cardiac sequelae of embolism of the pulmonary artery. Arch Intern Med 1939; 64: 296–321. 31 Dack S, Master AM, Horn H et al. Acute coronary insufficiency due to pulmonary embolism. Am J Med 1949; 7: 464–477. 32 Stein PD, Alshabkhoun S, Hatem C et al. Coronary artery blood flow in acute pulmonary embolism. Am J Cardiol 1968; 21: 32–37. 33 Stein, PD, Alshabkhoun S, Hawkins HF et al. Right coronary blood flow in acute pulmonary embolism. Am Heart J 1969; 77: 356–362. 34 Semisch CW, III, Merves L. Electrocardiographic studies on artificially produced pulmonary artery occlusion in human beings. Arch Intern Med 1942; 69: 417–428.
BLUK077-Stein
March 12, 2007
46
16:22
CHAPTER 46
The plain chest radiograph
Introduction Abnormalities on the plain chest radiograph of patients with pulmonary embolism (PE) were described in the late 1930s and early 1940s [1–3]. The radiographic signs of acute PE include oligemia, pleonemia of the obstructed lung, dilated hilar arteries, elevated hemidiaphragm, dilated pulmonary artery trunk, pleural effusion, atelectasis, and infiltrate [4, 5]. Occasionally, cavity formation may occur with pulmonary infarction [5]. Dilatation of the azygos vein and superior vena cava has been described [4]. Pulmonary edema has also been described [4]. It was thought by some to reflect underlying heart disease [5], although pulmonary edema has been observed in patients with no prior cardiopulmonary disease [6]. Clearly, the plain chest radiograph cannot be considered a definitive examination [7]. Importantly,
however, the chest radiograph may provide an additional data point in the evaluation of patients with suspected acute PE and may assist in the exclusion of disease processes, such as pneumothorax or rib fracture, which may simulate PE. The frequency of abnormalities on the plain chest radiograph in 383 patients in the Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) with acute PE, the severity of which ranged from mild to massive, is shown in Table 46.1 [8]. Many of these patients had prior cardiopulmonary disease. The frequency of abnormalities on the plain chest radiograph in 128 patients in the Urokinase Pulmonary Embolism Trial who had submassive to massive PE is also shown in Table 46.1 [9]. Approximately one-third of these patients had prior cardiopulmonary disease [10]. Radiographic findings among patients with acute PE and no prior cardiopulmonary disease are shown
Table 46.1 Plain chest radiograph in patients with acute pulmonary embolism, irrespective of prior cardiopulmonary disease. Mild to massive
Submassive to massive
PE (%) (n = 383)*
PE (%) (n = 128)†
Atelectasis or pulmonary parenchymal abnormality
69
—
Atelectasis
—
20
Consolidation
41
Pleural effusion
47
28
Pleural-based opacity
34
—
Elevated hemidiaphragm
28
41
Decreased pulmonary vascularity
20
15
Prominent central pulmonary artery
20
23
Cardiomegaly
16
—
Westermark’s sign‡
6
—
Pulmonary edema
10
—
Pulmonary venous hypertension pattern
—
3
Pulmonary arterial hypertension pattern
—
3
*Data are from Stein et al. [8] based on data from PIOPED. † Data are from the Urokinase Pulmonary Embolism Trial [9]. ‡ Prominent central pulmonary artery and decreased pulmonary vascularity.
216
BLUK077-Stein
March 12, 2007
CHAPTER 46
16:22
217
Plain chest radiograph
Table 46.2 Plain chest radiograph in patients with acute pulmonary embolism and no prior cardiopulmonary disease. Mild to massive
Submassive to massive
PE (%) (n = 117)*
PE (%) (n = 169)†
Atelectasis or pulmonary parenchymal abnormality
68
—
Atelectasis
—
28
Consolidation
39
Pleural effusion
48‡
30
Pleural-based opacity
35
— 46
Elevated hemidiaphragm
24
Decreased pulmonary vascularity
21
22
Prominent central pulmonary artery
15
21
Cardiomegaly
12
—
7
—
Westermark’s sign§ Pulmonary edema
4
—
Pulmonary venous hypertension pattern
—
4
Pulmonary arterial hypertension pattern
—
2
*Data are from Stein et al. [6]. † Data are from Stein et al. [11]. ‡ Among patients with a pleural effusion, 86% had only blunting of the costophrenic angle. None had a pleural effusion that occupied more than one-third of a hemithorax. § Prominent central pulmonary artery and decreased pulmonary vascularity. Reproduced from Stein et al. [11], with permission from S. Karger AG, Basel.
no prior cardiopulmonary disease, the plain chest radiograph was normal in 16% [6]. Among 169 patients with submassive or massive acute PE and no prior cardiopulmonary disease, one or more of the following signs (elevated hemidiaphragm, consolidation, pleural effusion, or atelectasis) occurred in 67% [11]. Parenchymal signs occurred with comparable frequency among patients with submassive PE (71%) and massive PE (65%). Either consolidation, atelectasis, or pleural effusion occurred in 62% with submassive PE and 50% with massive PE [11]. Atelectasis or a pulmonary parenchymal
in Table 46.2. The sensitivity of these findings in relation to the frequency of false-positive radiographic signs among 364 patients with no prior cardiopulmonary disease in whom PE was suspected is shown in Figure 46.1 [6]. Among 169 patients with submassive or massive acute PE and no prior cardiopulmonary disease, a normal chest radiograph was observed in 24% [11]. The frequency of a normal plain chest radiograph was similar among patients with submassive PE (25%) and massive PE (24%). Among 117 patients with PE that ranged in severity from mild to massive and who had
Figure 46.1 Sensitivity and frequency of false-positive values of various abnormalities on the plain chest radiograph. Data are from Stein et al. [6] in 364 patients with suspected acute pulmonary embolism and no prior cardiopulmonary disease. The numbers refer to the individual radiographic abnormalities listed in the insert.
Sensitivity (%)
100 80 1
60
1. Atelectasis/parenchymal abnormality 2. Pleural effusion 3. Pleural-based opacity 4. Elevated diaphragm 5. Decreased pulmonary vasculature 6. Prominent central PA 7. Cardiomegaly , 8. Westermark s sign 9. Pulmonary edema
2
40
3 4
5 6
20
7 9
0 0
8
20
40
60
False positives (%)
80
100
BLUK077-Stein
March 12, 2007
16:22
218
abnormality were the most frequent abnormalities on the plain chest radiographs in patients with no prior cardiopulmonary disease who had mild to massive PE (Table 46.2) [6]. Among these patients, a pleural effusion occurred in 48% [6]. Only blunting of the costophrenic angle was shown in 86% of the patients with PE who had a pleural effusion. No patient with PE and no prior cardiopulmonary disease had a pleural effusion that occupied more than one-third of a hemithorax [6]. Among patients with no prior cardiopulmonary disease, one or more vascular signs (focal oligemia, distension of the proximal pulmonary artery, pulmonary artery hypertension pattern, or pulmonary venous hypertension pattern) occurred in 37%. Vascular signs on the plain chest radiograph were more frequent among patients with massive PE than with submassive PE (43% versus 28%) [11]. Either focal oligemia or distension of the proximal portion of the pulmonary artery or both occurred more frequently among patients with massive PE than with submassive PE (42% versus 26%) [11]. Pulmonary edema occurred in 4% of the patients with PE and no prior cardiopulmonary disease [6]. Alveolar pulmonary edema occurred in only 1% and interstitial edema occurred in 3%. Among 25 patients with massive acute PE that involved 50% or more of the pulmonary arterial tree, 96% of whom had no prior cardiopulmonary disease, Kerr and associates identified zones of focal oligemia in all patients [12]. Other findings were hyperemic zones (40%), pulmonary infarction (56%), hilar enlargement (56%), elevated hemidiaphragm (32%), and enlarged pulmonary trunk (12%). Among 41 patients (the number with prior cardiopulmonary disease was unstated), Moses and associates observed that the most common signs on the plain chest radiograph were infiltrate (54%), pleural effusion (51%), and atelectasis (27%) [13]. An elevated hemidiaphragm was observed only in 17%. Focal oligemia was uncommon (2%) as was a pattern of pulmonary hypertension (5%). A normal plain chest radiograph was uncommon (7%). Among 126 patients who had PE (the number with prior cardiopulmonary disease was unstated), Laur [14] reported an elevated hemidiaphragm in 74% of severe cases and 50% in mild cases of PE. Dilatation of the hilar vessels was shown in 77% of the severe cases and 48% of the mild cases of PE. Dilatation of pul-
PART III
Diagnosis of acute PE
monary artery was noted in 57% of severe cases and 22% of mild cases of PE. Among 38 patients with PE, 5 of whom had heart failure, Wiener and associates [15] observed an elevated hemidiaphragm in 32% and changes suggestive of pleuritis in 45%. A negative plain chest radiograph was observed in 21%. Among 29 patients with PE, Follath and associates [16] reported dilatation of the right pulmonary artery in 59% and dilatation of the left pulmonary artery in 45%. Dilatation of the main pulmonary artery was observed in fewer, 27%. Infiltrates occurred in 45% and oligemia in 17%.
Radiographic abnormalities related to pulmonary artery mean pressure Patients with no prior cardiopulmonary disease and PE, who had normal plain chest radiographs, had the lowest pulmonary artery mean pressures (Table 46.3) [17]. Patients with normal plain chest radiographs had pulmonary artery mean pressures lower than in patients with an elevated diaphragm, pleural-based opacity, decreased pulmonary vascularity, prominent central pulmonary artery, Westermark’s sign, or cardiomegaly. Patients with normal chest radiographs also had pulmonary artery mean pressures lower than in patients with atelectasis or patients with a pleural effusion. The highest pulmonary artery mean pressures were in patients with no prior cardiopulmonary disease who had PE that ranged in severity from mild to massive with a prominent central pulmonary artery or cardiomegaly (Table 46.3) [17]. Patients with a prominent central pulmonary artery had pulmonary artery mean pressures higher than in patients with atelectasis or a pulmonary parenchymal abnormality or with a pleural effusion. Patients with cardiomegaly also had pulmonary artery mean pressures higher than in patients with atelectasis or a pulmonary parenchymal abnormality or with a pleural effusion. Patients with cardiomegaly also had pulmonary artery mean pressures higher than in patients with a pleural-based opacity, elevated diaphragm, or decreased pulmonary vascularity. Patients with cardiomegaly and a dilated central pulmonary artery had pulmonary artery mean pressures higher than in
BLUK077-Stein
March 12, 2007
CHAPTER 46
16:22
219
Plain chest radiograph
Table 46.3 Pulmonary artery mean pressure and chest radiographic abnormalities among 123 patients with acute pulmonary embolism and no prior cardiopulmonary disease. Number (%)
PA mean pressure (mm Hg)*
Normal
10 (8)
16 ± 5
Atelectasis or pulmonary parenchymal abnormality
94 (76)
21 ± 9
Pleural effusion
71 (58)
21 ± 10
Pleural-based opacity
45 (37)
23 ± 10
Elevated diaphragm
33 (27)
23 ± 8
Decreased pulmonary vascularity
26 (21)
24 ± 11
Prominent central pulmonary artery
20 (16)
30 ± 14
Cardiomegaly
15 (12)
34 ± 13
Westermark’s sign†
7 (6)
31 ± 11
Pulmonary edema
5 (4)
25 ± 10
*Values are mean ± standard deviation. † Prominent central pulmonary artery and decreased pulmonary vascularity. PA, pulmonary artery. Reprinted from Stein et al. [17], with permission from Elsevier.
patients with a prominent central pulmonary artery not accompanied by cardiomegaly. Pulmonary infarction or pulmonary hemorrhage (defined as the abrupt onset of pleuritic pain with or without hemoptysis and infiltrate on the chest radiograph) was uncommon when emboli obstructed the central pulmonary artery [18]. However, pulmonary infarction was frequent when distal arteries were occluded [18]. Presumably,
those with central pulmonary artery occlusion had higher pulmonary artery pressures. Radiographic abnormalities, related to the partial pressure of oxygen in arterial blood (Pa O2 ) and the alveolar–arterial oxygen gradient, of patients with PE that ranged in severity from mild to massive and who had no prior cardiopulmonary disease are shown in Table 46.4 [17].
Table 46.4 Partial pressure of oxygen in arterial blood alveolar–arterial oxygen gradient and chest radiographic abnormalities among 93 patients with acute pulmonary embolism and no prior cardiopulmonary disease. Number (%)
Pa O2 (mm Hg)*
A–a gradient (mm Hg)*
Normal
10 (11)
75 ± 14
30 ± 16
Atelectasis or pulmonary parenchymal abnormality
71 (76)
70 ± 13
36 ± 15
Pleural effusion
53 (57)
72 ± 15
35 ± 17
Pleural-based opacity
31 (33)
70 ± 15
37 ± 14
Elevated diaphragm
25 (27)
67 ± 12
39 ± 14
Decreased pulmonary vascularity
16 (17)
69 ± 15
39 ± 17
Prominent central pulmonary artery
11 (12)
61 ± 11
44 ± 16
Cardiomegaly
9 (10)
61 ± 16
48 ± 22
Westermark’s sign†
4 (4)
60 ± 14
41 ± 18
Pulmonary edema
2 (2)
55 ± 4
58 ± 2
*Values are mean ± standard deviation. † Prominent central pulmonary artery and decreased pulmonary vascularity. Pa O2 , partial pressure of oxygen in arterial blood; A–a = alveolar–arterial. Reprinted from Stein et al. [17], with permission from Elsevier.
BLUK077-Stein
March 12, 2007
16:22
220
References 1 Westermark N. On the Roentgen diagnosis of lung embolism. Acta Radiol 1938; 19: 357–372. 2 Hampton AO, Castleman B. Correlation of postmortem chest teleroentgenograms with autopsy findings: with special reference to pulmonary embolism and infarction. Am J Roentgenol 1940; 43: 305–326. 3 Fleischner F, Hampton AO, Castleman B. Linear shadows in the lung (interlobar pleuritis, atelectasis, and healed infarction). Am J Roentgenol 1941; 46: 610– 618. 4 Fleischner FG. Pulmonary embolism. Clin Radiol 1962; 13: 169–182. 5 Kelly MJ, Elliot LP. The radiologic evaluation of the patient with suspected pulmonary thromboembolic disease. Med Clin North Am 1974; 59: 3–36. 6 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 7 Greenspan RH, Ravin CE, Polansky SM, McLoud TC. Accuracy of the chest radiograph in diagnosis of pulmonary embolism. Invest Radiol 1982; 17: 539– 543. 8 Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. 9 National Cooperative Study: the Urokinase Pulmonary Embolism Trial: associated clinical and laboratory findings. Circulation 1973; 47/48(suppl II): 81–85.
PART III
Diagnosis of acute PE
10 National Cooperative Study. The Urokinase Pulmonary Embolism Trial: clinical and electrocardiographic observations. Circulation 1973; 47/48(suppl II): 60–65. 11 Stein PD, Willis PW, III, DeMets DL, Greenspan RH. Plain chest roentgenogram in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Am J Noninvas Cardiol 1987; 1: 171–176. 12 Kerr IH, Simon G, Sutton GC. The value of the plain radiograph in acute massive pulmonary embolism. Br J Radiol 1971; 44: 751–757. 13 Moses DC, Silver TM, Bookstein JJ. The complementary roles of chest radiography, lung scanning, and selective pulmonary angiography in the diagnosis of pulmonary embolism. Circulation 1974; 49: 179–188. 14 Laur A. Roentgen diagnosis of pulmonary embolism and its differential from myocardial infarction. Am J Roentgenol 1963; 90: 632–637. 15 Wiener SN, Edelstein J, Charms BL. Observations on pulmonary embolism and the pulmonary angiogram. Am J Roentgenol 1966; 98: 859–873. 16 Follath F, Burkart F, Fridrich R. On evaluation of the thoracic picture in acute pulmonary embolism. Schweiz Med Wochenschr 1968; 98: 1589–1592. 17 Stein PD, Athanasoulis C, Greenspan RH, Henry JW. Relation of plain chest radiographic findings to pulmonary arterial pressure and arterial blood oxygen levels in patients with acute pulmonary embolism. Am J Cardiol 1992; 69: 394–396. 18 Dalen JE, Haffajee CI, Alpert JS, Howe JP, III, Ockene IS, Paraskos JA. Pulmonary embolism, pulmonary hemorrhage, and pulmonary infarction. N Engl J Med 1977; 296: 1431–1435.
BLUK077-Stein
March 11, 2007
47
8:32
CHAPTER 47
Arterial blood gases and the alveolar–arterial oxygen difference in acute pulmonary embolism
The partial pressure of oxygen in arterial blood (Pa O2 ) when low in patients with suspected acute pulmonary embolism (PE) has been shown to be a helpful adjunct in the diagnostic assessment [1]. It is recognized, however, that patients with acute PE may have a normal Pa O2 [2–4]. Among 132 patients with acute PE and
no prior cardiopulmonary disease who had measurements of the Pa O2 while breathing room air, 24% had a Pa O2 of 80 mm Hg or higher (Figure 47.1) (P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990). These data were obtained from patients in the randomized arm of the Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) and in the referred arm [5]. In PIOPED II, the Pa O2 while breathing room
Figure 47.1 Distribution of partial pressure of oxygen in arterial blood (Pa O2 ) and cumulative frequency while breathing room air among 132 patients with angiographically proven pulmonary embolism and no preexisting cardiac or pulmonary disease. The number of
patients with indicated values of the Pa O2 is shown on the left and the cumulative percent of patients with the indicated values of the Pa O2 is shown of the right. (Data from P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990.)
Arterial blood gases
221
BLUK077-Stein
March 11, 2007
8:32
222
PART III
Diagnosis of acute PE
Table 47.1 Arterial blood gases and alveolar–arterial oxygen difference while breathing room air (PIOPED II). n (%) Probability PE
Pa O2 (mm Hg) ≤49
Probability PE
PE no Prior
No PE no prior
vs. No PE (No
PE all patients
No PE all patients
vs. No PE
CPD ( N = 48)
CPD ( N = 88)
Prior CPD)
( N = 74)
( N = 186)
(all patients)
1 (2)
2 (2)
NS
4 (5)
17 (9)
50–59
6 (13)
12 (14)
NS
12 (16)
32 (17)
NS
60-69
15 (31)
14 (16)
≤0.05
20 (27)
35 (19)
NS
8 (17)
13 (15)
NS
14 (19)
32 (17)
NS
18 (38)
47 (53)
NS
24 (32)
70 (38)
NS
70–79 ≥80 Pa CO2 (mm Hg) ≤35 36–39 ≥40
NS
30 (63)
39 (44)
≤0.05
42 (57)
65 (35)
≤0.01
12 (25)
17 (19)
NS
18 (24)
39 (21)
NS
6 (13)
32 (36)
≤0.01
14 (19)
82 (44)
≤0.001
≤0.05
pH (units) <7.35
0 (0)
7 (8)
0 (0)
13 (7)
≤0.025
7.35–7.45
29 (60)
60 (68)
NS
41 (55)
131 (70)
≤0.025
>7.45
19 (40)
21 (24)
NS
33 (45)
42 (23)
≤0.001
A–a O2 difference (mm Hg) ≤20
17 (35)
44 (50)
NS
24 (32)
70 (38)
NS
21–30
4 (8)
10 (11)
NS
5 (7)
32 (17)
≤0.05
31–40
11 (23)
13 (15)
NS
18 (24)
30 (16)
NS
41–50
9 (19)
13 (15)
NS
14 (19)
32 (17)
NS
51–60
5 (10)
6 (7)
NS
10 (14)
17 (9)
NS
≥61
2 (4)
2 (2)
NS
3 (4)
5 (3)
NS
CPD, cardiopulmonary disease; PE, pulmonary embolism; NS, not significant. Reprinted from Stein et al. [6], with permission.
air was measured in only 48 patients with PE and no prior cardiopulmonary disease [6] (Table 47.1). The Pa O2 was 80 mm Hg or higher in 38% of these patients [6]. Even among patients with submassive or massive acute PE, 12% had a Pa O2 of 80 mm Hg or higher [2]. The distribution of the Pa O2 among 88 randomized patients with PE who had no prior cardiopulmonary disease was comparable to the distribution of the Pa O2 among 202 such patients in whom PE was suspected, but excluded [3]. Among 277 patients with acute PE, 53% of whom had prior cardiopulmonary disease, the Pa O2 was 80 mm Hg or higher in 19% [7]. Clearly, acute PE cannot be excluded on the basis of a normal Pa O2 . Typically, however, among patients with acute PE, the Pa O2 is low. The distribution of the partial pressure of carbon dioxide in arterial blood (Pa CO2 ) among 132 patients with acute PE and no prior cardiopulmonary disease who had measurements while breathing room air is shown in Figure 47.2. The Pa CO2 was less than
35 mm Hg in 49% of these patients, and it was greater than 45 mm Hg in 2% of the patients (P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990). Among 48 patients in PIOPED II with PE and no prior cardiopulmonary disease, the Pa CO2 was ≤35 mm Hg in 65% and ≥40 mm Hg in 13% (6). The pH of arterial blood was greater than 7.44 in 53% of such patients in PIOPED I and it was less than 7.38 in 4% of patients (Figure 47.3) (P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990). In PIOPED II, more of the patients with PE and no prior cardiopulmonary disease had an arterial blood pH less than 7.35 and 40% had a pH greater than 7.45 (6).
Alveolar–arterial oxygen difference A normal alveolar–arterial oxygen difference (alveolar–arterial oxygen gradient) does not exclude acute PE [3, 6, 7, 8]. The distribution of values of
BLUK077-Stein
March 11, 2007
8:32
100
30
90 25
80
60 50
15
Cumulative frequency (%)
Number of patients
70 20
40 10
30 20
5 10 0
0 25
28
31
34
37
40
43
47
PaCO2 (mm Hg) Figure 47.2 Distribution of partial pressure of carbon dioxide in arterial blood (Pa CO2 ) and cumulative frequency while breathing room air among 132 patients with
angiographically proven pulmonary embolism and no prior cardiopulmonary disease. (Data from P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990.)
100
60
50 80
Number of patients
60 30 40 20
20 10
0
0 7.17
7.21
7.25
7.29
7.33
7.37
7.41
7.45
7.49
7.53
7.59
pH Figure 47.3 Distribution of pH of arterial blood and cumulative frequency while breathing room air among 132 patients with angiographically proven PE and no prior
cardiopulmonary disease. (Data from P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990.)
223
Cumulative frequency (%)
40
BLUK077-Stein
March 11, 2007
8:32
224
PART III
35
PE (n = 88)
30 Percent of patients
Diagnosis of acute PE
No PE (n = 202)
25 20 15 10 5 0 <10
10−20 21−30 31−40 41−50 51−60 >61 Alveolar-arterial oxygen gradient (mm Hg)
the alveolar–arterial oxygen differences in patients with acute PE and in patients in whom acute PE was excluded is shown in Figure 47.4 [3]. No value of the alveolar–arterial oxygen difference was diagnostic of PE and no value excluded PE. Subsequent to the publication of our observations on the alveolar–arterial oxygen difference in patients with no prior cardiopulmonary disease, there has been rekindling of interest in the ability of a normal alveolar– arterial oxygen difference to exclude PE. McFarlane and Imperiale indicated that only 1.9% of the patients with a normal alveolar–arterial oxygen difference and no history of prior PE or prior DVT (deep venous thrombosis) had PE [9]. They concluded, therefore, that a normal alveolar–arterial oxygen difference among patients with no prior PE or prior DVT makes the diagnosis of PE so unlikely that further diagnostic evaluation may be unnecessary in this subgroup. Because the alveolar–arterial oxygen difference seemed to be more useful in the experience of these investigators than what our data showed, we evaluated further the value of the alveolar–arterial oxygen difference among patients in PIOPED [7]. We used more than one definition of a normal alveolar–arterial oxygen difference. We stratified according to categories of patients in whom a normal alveolar–arterial oxygen difference is thought by some to assist in excluding the diagnosis of PE. We used an expanded database from PIOPED that included patients randomized for obligatory pulmonary angiography and patients referred for pulmonary angiography by their physicians.
Figure 47.4 Distribution of alveolar–arterial oxygen gradient among 88 patients with angiographically proven pulmonary embolism (PE) and no preexisting cardiac or pulmonary disease (solid bars) and 202 such patients in whom PE was excluded (striped bars). (Reprinted with permission from Stein et al. [3].)
The alveolar–arterial oxygen difference was calculated according to the following equation [10]: Alveolar–arterial oxygen difference = 150 − 1.25Pa CO2 − Pa O2 where Pa CO2 is the partial pressure of carbon dioxide in arterial blood (mm Hg) and Pa O2 is the partial pressure of oxygen in arterial blood (mm Hg). Measurements of arterial blood gases were obtained with the patient breathing room air. All measurements were obtained within 24 hours prior to the diagnostic pulmonary angiogram. This may have been as long as 48 hours after the onset of symptoms, although usually it was less than 36 hours after the onset of symptoms [5]. The normal alveolar–arterial oxygen difference increases with age. Most patients have an alveolar– arterial oxygen difference ≤20 mm Hg [11–14]. Normal age-related values obtained from data reported by Mellemgaard [11] are as follows: Patients 15–19 years of age had an alveolar–arterial oxygen difference ≤14 mm Hg; patients 20–29 years of age had an alveolar–arterial oxygen difference ≤20 mm Hg; and patients ≥30 years of age (with one exception) had an alveolar–arterial oxygen difference ≤27 mm Hg [11]. These values correspond closely with normal agerelated values of the alveolar–arterial oxygen difference reported by Harris and associates [12] and Kanber and associates [14]. Others reported somewhat lower normal values of the alveolar–arterial oxygen difference,
BLUK077-Stein
March 11, 2007
CHAPTER 47
8:32
225
Blood gases and alveolar–arterial oxygen difference
which did not exceed 20 mm Hg in subjects 23–60 years of age [13]. Normal values of the alveolar–arterial oxygen difference were also approximated by the equation age/4 + 4 [9, 15]. This equation is restrictive, however. It would have excluded 28% of the normal patients reported by Mellemgaard [11], 20% reported by Kanber and associates [14], 27% reported by Harris and associates [12], and 16% reported by Filley and associates [13]. Averaging these data, the age-corrected equation would incorrectly define as abnormal 19% of the patients [7]. Irrespective of how a normal alveolar–arterial oxygen difference was defined, comparable percentages of patients with normal alveolar–arterial oxygen differences had PE or no PE [7]. This applied to all patients with PE including patients with prior cardiopulmonary disease. It applied to patients stratified according to the absence of prior cardiopulmonary disease and to patients stratified according to the absence of prior PE or prior DVT.
All patients Among 280 patients with acute PE in PIOPED I, including patients with prior cardiopulmonary disease as well as patients with prior PE and/or prior DVT, 12% had an alveolar–arterial oxygen difference ≤20 mm Hg versus 16% among patients who did not have PE (Table 47.2) [7]. Using the same method in PIOPED II, among 74 patients with PE the alveolar-arterial oxygen difference was ≤20 mmHg in 32% (6). Among
patients with a normal alveolar–arterial oxygen difference based on reported age-related values, 20% with PE had a normal alveolar–arterial oxygen difference versus 25% among patients who did not have PE. If a normal alveolar–arterial oxygen difference was defined on the basis of the equation age/4 + 4, then 8% had a normal alveolar–arterial oxygen difference versus 12% who did not have PE. None of these comparisons of patients with PE to those without PE showed statistically significant differences.
Patients with no prior cardiopulmonary disease Among 130 patients with no prior cardiopulmonary disease in PIOPED I, 14% with an alveolar–arterial oxygen difference ≤ 20 mm Hg had acute PE versus 20% among patients who did not have PE (Table 47.2) [7]. Using the same method in PIOPED II, among 48 patients with PE and no prior cardiopulmonary disease, the alveolar-arterial oxygen difference was ≤ 20 mmHg in 35% (6). Among patients with a normal alveolar–arterial oxygen based on reported age-related values, 23% with PE had a normal alveolar–arterial oxygen difference versus 27% with no PE. Using the age/4 + 4 equation to define normal values of the alveolar–arterial oxygen difference, 10% with PE had normal values of the alveolar–arterial oxygen difference versus 12% among patients who did not have PE. None of these comparisons of patients with PE to those without PE showed statistically significant differences.
Table 47.2 Frequency of normal A–a gradients among patients with pulmonary embolism and patients with no pulmonary embolism. (PIOPED I). n/N (%) A–a gradient
A–a gradient
A–a gradient normal =
normal ≤20 mm Hg
normal ≤ Age/4 +
Age-related data
(A–a ≤ N1)
4 (A–a ≤ N1)
(A–a ≤ N1)
PE
No PE
PE
No PE
PE
No PE
All patients
33/280 (12)
82/499 (16)
23/280 (8)
59/499 (12)
57/280 (20)
123/499 (25)
No prior CPD
18/130 (14)
40/200 (20)
13/130 (10)
24/200 (12)
30/130 (23)
54/200 (27)
No prior PE or DVT
21/190 (11)
55/365 (15)
15/190 (8)
40/365 (11)
38/190 (20)
86/365 (24)
All differences (PE vs. no PE) not significant. CPD, cardiopulmonary disease; DVT, deep vein thrombosis; Nl, normal; Pts, patients. Reprinted with permission from Stein et al. [7].
BLUK077-Stein
March 11, 2007
8:32
226
PART III
Among 190 patients with no prior PE or prior DVT, 11% with acute PE had an alveolar–arterial oxygen difference ≤ 20 mm Hg versus 15% who did not have PE (Table 47.2) [7]. Among patients with a normal alveolar–arterial oxygen difference based on reported age-related values, 20% had PE and 24% did not have PE. Based on the equation for age-corrected values of the alveolar–arterial oxygen difference, 8% with PE had a normal alveolar–arterial oxygen difference versus 11% with no PE. None of these comparisons of patients with PE to those without PE showed statistically significant differences.
Relation of alveolar–arterial oxygen difference to the Pa O2
A–a gradient (mm Hg)
The alveolar–arterial oxygen difference showed a strong inverse linear correlation with the Pa O2 among patients with PE and no prior cardiopulmonary disease (Figure 47.5) [7]. The alveolar–arterial oxygen difference also showed a strong inverse linear correlation with the Pa O2 and patients with PE who had prior cardiopulmonary disease (Figure 47.6). This indicates that the value of both tests for screening for acute PE is similar. Neither values of the alveolar–arterial oxygen difference nor values of the Pa O2 can be used to exclude PE.
100 90 80 70 60 50 40 30 20 10 0 −10
No CPD
A−a gradient (mm Hg)
Patients with no prior PE or prior DVT
Diagnosis of acute PE
PaO2 (mm Hg) Figure 47.5 Alveolar–arterial (A–a) oxygen difference shown as a function of the partial pressure of oxygen in arterial blood (Pa O2 ) among patients with no prior cardiopulmonary disease (No CPD). Correlation coefficient r = −0.917 (P < 0.005). (Reprinted with permission from Stein et al. [7].)
Relation of alveolar–arterial oxygen difference to pulmonary pressure and to the ventilation–perfusion lung scan Patients with less severe PE were more likely to have a normal alveolar–arterial oxygen difference than those with more severe PE. The alveolar–arterial oxygen difference showed a loose linear correlation with the pulmonary artery mean pressure among patients with PE and no prior cardiopulmonary disease (r = 0.32, P < .005) and
CPD
20
40
60 80 PaO2 (mm Hg)
100
120
Figure 47.6 Alveolar–arterial (A–a) oxygen difference shown as a function of the partial pressure of oxygen in arterial blood (Pa O2 ) among patients with any prior cardiopulmonary disease (CPD). Correlation coefficient r = −0.842 (P < 0.005). (Reprinted with permission from Stein et al. [7].)
BLUK077-Stein
March 11, 2007
CHAPTER 47
8:32
227
Blood gases and alveolar–arterial oxygen difference
Table 47.3 Negative predictive value of blood gases and alveolar–arterial oxygen gradient, alone or in combination: PIOPED I. No PE/Test Neg (%) No prior CPD
Prior CPD
All patients
Pa CO2 ≥35
105/171 (61)
142/211 (67)
247/382 (65)
Pa O2 ≥80
55/87 (63)
62/84 (74)
117/171 (68)
A–a ≤20
40/58 (69)
38/50 (76)
78/108 (72)
Pa CO2 ≥35 and Pa O2 ≥80
30/49 (61)
28/36 (78)
58/85 (68)
Pa CO2 ≥35 and A–a ≤20
30/46 (65)
28/33 (85)
58/79 (73)
Pa O2 ≥80 and A–a ≤20
36/54 (67)
34/45 (76)
70/99 (71)
Pa CO2 ≥35, Pa O2 ≥80, and A–a ≤20
26/42 (62)
24/28 (86)
50/70 (71)
Blood gases and alveolar–arterial (A–a) oxygen differences are in mm Hg. PE, pulmonary embolism; CPD, cardiopulmonary disease. NEG, negative. Reprinted with permission from Stein et al. [17].
among patients with PE who had prior cardiopulmonary disease (r = 0.34, P < .005) [7]. The alveolar–arterial oxygen difference showed a loose linear correlation with the number of mismatched perfusion defects on the ventilation– perfusion lung scan among patients with PE and no prior cardiopulmonary disease (r = 0.40, P < .005) [7]. The alveolar–arterial oxygen difference also showed a loose linear correlation with the number of mismatched perfusion defects on the ventilation– perfusion lung scan among patients with PE who had prior cardiopulmonary disease (r = 0.39, P < .005).
Blood gases in combination with the alveolar-arterial oxygen difference It has been suggested that if both the Pa CO2 and the alveolar–arterial oxygen difference are normal, this can be used as evidence against the presence of PE [16]. This is based on the observation among patients with essentially no prior cardiac or pulmonary disease who had PE that 93% had hypoxemia or hypocapnia and 98% had an increased alveolar–arterial oxygen difference or hypocapnia [16]. We tested this among patients in PIOPED to determine if normal values of one or more blood gases, alone or in combination, could be used to exclude the diagnosis of acute PE [17]. Whether PE is absent if blood gases are normal is shown by the negative predictive value. Based on data
from PIOPED, we were not able to exclude PE based on an absence of abnormal values of the Pa O2 , Pa CO2 , or alveolar–arterial oxygen difference [17]. Among patients with no prior cardiopulmonary disease, PE was absent in 69% or fewer with a Pa O2 ≥80 mm Hg, Pa CO2 ≥35 mm Hg, and/or alveolar–arterial oxygen difference ≤20 mm Hg (Table 47.3). Even with values of the Pa O2 ≥80 mm Hg, Pa CO2 ≥35 mm Hg, and alveolar–arterial oxygen difference ≤20 mm Hg, 62% did not have PE. Conversely, in patients who did not have prior cardiopulmonary disease, PE was present in 38% when the Pa O2 and Pa CO2 were not low and the alveolar–arterial oxygen difference was normal. Among patients with prior cardiopulmonary disease, the highest negative predictive value was shown with the combination Pa O2 ≥80 mm Hg, Pa CO2 ≥35 mm Hg, and alveolar–arterial oxygen difference ≤20 mm Hg (Table 47.3) [17]. With this combination, PE was absent in 86% of patients who had prior cardiopulmonary disease. Conversely, in patients with prior cardiopulmonary disease, PE was present in 14% when the Pa O2 and Pa CO2 were not low and the alveolar–arterial oxygen difference was normal. We observed that 88% of the patients with PE and no prior cardiopulmonary disease had either a low Pa CO2 or a high alveolar–arterial oxygen difference [17]. We also observed that 97% of the patients with prior cardiopulmonary disease who had PE had a low Pa CO2 or a high alveolar–arterial oxygen difference. Pulmonary embolism could not be excluded on the basis of such
BLUK077-Stein
March 11, 2007
8:32
228
blood gases [17]. No combinations of blood gases were identified that reliably excluded pulmonary embolism. Our data indicate that the Pa O2 and the alveolar– arterial oxygen difference may be included among laboratory and clinical findings that typically are abnormal in patients with PE, but may be normal in some patients with PE. Normal values of an alveolar–arterial oxygen difference do not exclude the diagnosis of acute PE.
References 1 Szucs MM, Brooks HL, Grossman W et al. Diagnostic sensitivity of laboratory findings in acute pulmonary embolism. Ann Intern Med 1971; 74: 161–166. 2 The National Cooperative Study: The Urokinase Pulmonary Embolism Trial. Circulation 1973; 47/48(suppl II): II-18–II-21 and II-81–II-85. 3 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 4 Menzoian JO, Williams LF. Is pulmonary angiography essential for the diagnosis of acute pulmonary embolism? Am J Surg 1979; 137: 543–548. 5 A collaborative study by the PIOPED investigators. Value of the ventilation–perfusion scan in acute pulmonary embolism: Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 6 Stein PD, Beemath A, Weg JG et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, submitted.
PART III
Diagnosis of acute PE
7 Stein PD, Goldhaber SZ, Henry JW. Alveolar–arterial oxygen gradient in the assessment of acute pulmonary embolism. Chest 1995; 107: 139–143. 8 Overton DT, Bocka JJ. The alveolar–arterial oxygen gradient in patients with documented pulmonary embolism. Arch Int Med 1988; 148: 1617–1619. 9 McFarlane MJ, Imperiale TF. Use of the alveolar–arterial oxygen gradient in the diagnosis of pulmonary embolism. Am J Med 1994; 96: 57–62. 10 Guenter CA. Respiratory function of the lungs and blood. In: Guenter CA, Welch MH, eds. Pulmonary Medicine, 2nd edn. Lippincott, Philadelphia, 1982: 168. 11 Mellemgaard K. The alveolar–arterial oxygen difference: its size and components in normal man. Acta Physiol Scand 1966; 67: 10–20. 12 Harris EA, Kenyon AM, Nisbet HD, Seelye ER, Whitlock RML. The normal alveolar–arterial oxygen-tension gradient in man. Clin Sci Mol Med 1974; 46: 89– 104. 13 Filley GF, Gregoire F, Wright GW. Alveolar and arterial oxygen tensions and the significance of the alveolar– arterial oxygen tension difference in normal men. J Clin Invest 1954; 33: 517–529. 14 Kanber GJ, King FW, Eshchar YR, Sharp JT. The alveolar– arterial oxygen gradient in young and elderly men during air and oxygen breathing. Am Rev Respir Dis 1968; 97: 376–381. 15 Skorodin MS. Respiratory diseases and A–a gradient measurement. JAMA 1984; 252: 1344. 16 Cvitanic O, Marino PL. Improved use of arterial blood gas analysis in suspected pulmonary embolism. Chest 1989; 95: 48–51. 17 Stein PD, Goldhaber SZ, Henry JW, Miller AC. Arterial blood gas analysis in the assessment of suspected acute pulmonary embolism. Chest 1996; 109: 78–81.
BLUK077-Stein
March 28, 2007
48
15:26
CHAPTER 48
Fever in acute pulmonary embolism
Patients (%)
Low-grade fever based on data from Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I), is not uncommon in pulmonary embolism (PE) and high fever, although rare, may occur [1]. Low-grade fever need not be accompanied by pulmonary hemorrhage/infarction [1]. Temperature was measured on the day of recruitment into PIOPED I [1]. Symptoms of PE began within 24 hours of study entry [2]. Fever was defined as temperature of 100.0◦ F (37.8◦ C) or higher [3, 4]. Although the normal temperature in human beings has been said to be 98.6◦ F (37◦ C) based on Wunderlich’s original observations over 120 years ago, the maximum oral temperature in normal individuals at 4 p.m. is 99.9◦ F (37.7◦ C) [3, 4]. Among all patients, fever [temperature ≥ 100.0◦ F (37.8◦ C)] was present in 95 of 363 (26%) patients with PE. Other definite or possible causes of fever were present in 46 of these patients [1]. Among those with pulmonary hemorrhage/infarction, fever was present in 80 of 308 (26%). Among those who did not have pulmonary hemorrhage/infarction, fever was present in a similar percentage, 15 of 55 (27%). Among patients with PE and no other definite or possible cause of fever, temperature was normal in
100 90 80 70 60 50 40 30 20 10 0
268 of 311 (86%) (Figure 48.1) [1]. Among those with fever, the fever was usually low grade. Temperature of 101◦ F (38.3◦ C) or higher occurred only in 19 of 311 (6%) patients and only 5 of 311 (1.6%) had a temperature of 102◦ F (38.9◦ C) or higher. Some conditions causing fever may not have been listed in the discharge summaries or computer listing of discharge diagnoses. The frequency of patients that we reported with fever caused by PE, therefore, may have been higher than the actual value. The distribution of temperatures according to whether patients had pulmonary hemorrhage/infarction syndrome versus uncomplicated PE or circulatory collapse is shown in Figure 48.2 [1]. Signs or symptoms of deep venous thrombosis (DVT) were present in 24 of 43 patients (56%) with PE who had fever. Among patients with a temperature of 102◦ F (38.9◦ C) or higher, 3 of 5 (60%) had clinical findings compatible with DVT. Each of 4 patients with PE and otherwise unexplained fever who did not have pulmonary hemorrhage or infarction had clinical findings compatible with DVT. In PIOPED I, among patients with PE and no prior cardiopulmonary disease, a temperature higher than 101.3◦ F (38.5◦ C) was reported in 7% [5], but no attempt was made to eliminate patients with other causes
86
8 <100.0
100.0− 100.9
5 101.0− 101.9
1 102.0− 102.9
Temperature (°F)
<1
<1
103.0− 103.9
>103.9
Figure 48.1 Temperature distributions among patients with pulmonary embolism and no other definite or possible cause of fever. (Data from Stein et al. [1].)
229
BLUK077-Stein
March 28, 2007
15:26
230
PART III
90
85
91
80 Patients (%)
Diagnosis of acute PE
Hemorrhage/infarction syndrome
70
Uncomplicated PE or circulatory collapse syndrome
60 50 40 30 20
8 7
10
5
2
0 <100.0
100.0− 100.9
101.0− 101.9
1
0
102.0− 102.9
<1 0
<1 0
103.0− 103.9
>103.9
Temperature (°F)
of fever. Among patients with no prior cardiopulmonary disease who had the pulmonary infarction syndrome, temperature higher than 101.3◦ F (38.5◦ C) was reported in 6% [6] and with the syndrome of isolated dyspnea, 16% were reported with such a temperature elevation [6]. Again, no attempt was made to eliminate patients with other causes of fever. In the Urokinase Pulmonary Embolism Trial and the Urokinase-Streptokinase Embolism Trial, among patients with no prior cardiopulmonary disease, a temperature higher than 99.9◦ F (37.5◦ C) was observed in 50% [7]. The prevalence of fever was the same among patients with massive and with submassive PE. At a tertiary care hospital, among all patients with PE diagnosed by pulmonary angiography between 1980 and 1984, 14% had a temperature higher than 100◦ F (37.8◦ C) [8]. Other causes of fever were not excluded in any of these case series. Murray and associates attempted to clarify whether fever in PE was due to the PE or accompanying disease [9]. Among patients with angiographically diagnosed PE and no other apparent cause of fever, a temperature of 100.4◦ F (38◦ C) or higher was observed in 20 of 31 (64%) and 2 of 31 (6%) had a temperature of 103.1◦ F (39.5◦ C) or higher [9]. Pulmonary infarction was not more frequent among patients with fever [9]. Some reported that unless concomitant pulmonary infection is present, the temperature generally is 101◦ F (38.3◦ C) or less [10, 11]. Higher temperatures were generally observed in patients with extensive pulmonary infarction or in whom secondary pneumonitis developed distal to the embolus [11].
Figure 48.2 The distribution of temperatures according to whether patients had pulmonary hemorrhage/infarction syndrome versus uncomplicated pulmonary embolism (PE) or circulatory collapse. (Data from Stein et al. [1].)
Some have suggested that low-grade fever, especially after an operation, is occasionally the only symptom of PE [12]. Occasional case reports describe PE as a cause of fever of undetermined origin [13, 14].
References 1 Stein PD, Afzal A, Henry JW, Villarreal CG. Fever in acute pulmonary embolism. Chest 2000; 117: 39–42. 2 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 3 Mackowiak PA, Wasserman SS, Levine MM. A critical appraisal of 98.6◦ F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich. JAMA 1992; 268: 1578–1580. 4 Gelfand JA, Dinarello CA. Alterations in body temperature. In: Fauci AS, Braunwald E, Isselbacher KJ et al., eds. Harrison’s Principles of Internal Medicine, 14th edn. McGraw-Hill Inc., New York, 1998: 84–90. 5 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 6 Stein PD, Henry JW. Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes. Chest 1997; 112: 974– 979. 7 Stein PD, Willis PW, III, DeMets DL. History and physical examination in acute pulmonary embolism in patients
BLUK077-Stein
March 28, 2007
CHAPTER 48
15:26
Fever in acute PE
without preexisting cardiac or pulmonary disease. Am J Cardiol 1981; 47: 218–223. 8 Leeper KV, Jr, Popovich JP, Jr, Adams DA, Stein PD. The clinical manifestations of acute pulmonary embolism: HFH experience. A five year review. Henry Ford Hosp Med J 1988; 36: 29–34. 9 Murray HW, Ellis GC, Blumenthal DS, Sos TA. Fever and pulmonary thromboembolism. Am J Med 1979; 67: 232– 235. 10 Sasahara AA, Cannilla JE, Morse RL et al. Clinical and physiologic studies in pulmonary thromboembolism. Am J Cardiol 1967; 20: 10–20.
231
11 Sharma GV, Sasahara AA, McIntyre KM. Pulmonary embolism: the great imitator. Dis Mon 1976; 22: 4–38. 12 Hodgson CH. Pulmonary embolism and infarction. Dis Chest 1965; 47: 577–588. 13 Aburahma AF, Saiedy S. Deep vein thrombosis as a probable cause of fever of unknown origin. W V Med J 1997; 93: 368–370. 14 Stallman JS, Aisen PS, Aisen ML. Pulmonary embolism presenting as fever in spinal cord injury patients: report of two cases and review of the literature. J Am Paraplegia Soc 1993; 16: 157–159.
BLUK077-Stein
April 6, 2007
49
16:2
CHAPTER 49
Leukocytosis in acute pulmonary embolism syndrome) had a WBC count of >10,000/mm3 in 32 of 183 (17%) [1]. Among patients who did not have the pulmonary hemorrhage/infarction syndrome, 20 of 83 (24%) had a WBC count of >10,000/mm3 . Among patients with the pulmonary hemorrhage/infarction syndrome, the frequency of an elevated WBC count (16–17%) was the same irrespective of whether the pulmonary hemorrhage/infarction syndrome was characterized by atelectasis/infiltrate on the chest radiograph, hemoptysis, or pleuritic chest pain. The percent of neutrophils in patients with PE and a WBC count of >10,000/mm3 and in those with a WBC count of ≤10,000/mm3 is shown in Table 49.1 [1]. Older clinicians found the WBC count useful in distinguishing between pneumonia and pulmonary infarction [2]. Pulmonary infarction was considered to be more likely than pneumonia if the WBC count was <15,000/mm3 [2]. Sasahara showed that among patients with PE, not necessarily with the pulmonary hemorrhage/infarction syndrome, the WBC count was in the range of 15,000–25,000/mm3 in only 1 of 20 (5%) [2]. Others evaluated leukocytosis with pulmonary infarction. Among 200 patients with
Leukocytosis is uncommon in acute pulmonary embolism (PE) [1]. The frequency of an elevated white blood cell (WBC) count was not greater in patients with the pulmonary hemorrhage/infarction syndrome than in patients with acute PE who did not have pulmonary hemorrhage/infarction [1]. One reason that the prevalence of leukocytosis in acute PE has been unclear, in spite of an extensive number of investigations of PE over a period of several years, may relate to the fact that acute PE is usually associated with other conditions, which themselves may cause leukocytosis. Based on a detailed review of records from a tertiary care hospital, among all patients with PE, the WBC count was ≤10,000/mm3 in 214 of 386 (55%) and >10,000/mm3 in 172 of 386 (45%) [1]. Among patients in whom other definite causes for leukocytosis were eliminated, 214 of 324 (66%) had a WBC count of ≤10,000/mm3 and 110 of 324 (34%) had a WBC count of >10,000/mm3 [1]. The distribution of the WBC count in patients with PE and no other cause for WBC count elevation is shown in Figure 49.1. Among patients with no other cause for a WBC count elevation, those with pulmonary hemorrhage/infarction syndrome (hemoptysis/pleuritic pain 80
80
PE (%)
60 40 20
7
8
3
1
1
0
0 <10.1
10.1− 12.0
12.1− 14.0
14.1− 16.0
16.1− 18.0
WBC × 103/mm3
232
18.1− 20.0
>20
Figure 49.1 The distribution of white blood cell (WBC) count in patients with pulmonary embolism (PE) and no other cause for WBC count elevation. (Reproduced with minor modifications from Afzal et al. [1].)
BLUK077-Stein
April 6, 2007
CHAPTER 49
16:2
233
Leukocytosis in acute PE
Table 49.1 Distribution of percent neutrophils in patients with PE according to the presence or absence of leukocytosis. No. of patients (%) Neutrophils (%)
WBC >10,000/mm3 ( N = 27)
40–50
—
WBC ≤10,000/mm3 ( N = 81) 10 (12.3)
61–70
5 (18.5)
24 (29.6)
71–80
14 (51.8)
20 (24.6)
81–90
6 (22.2)
13 (16.0)
>90
1 (3.7)
2 (2.4)
WBC = white blood cells. Reproduced with minor modifications from Afzal et al. [1].
pulmonary hemorrhage/infarction shown at autopsy, the WBC count was usually between 10,000 and 15,000/mm3 [3]. In patients with pulmonary infarction accompanied by pneumonia or patients with large pulmonary infarctions, the WBC count was reported to be “around” 20,000/mm3 . In two patients a WBC count of 28,000/mm3 and a count of 48,000/mm3 were observed. Among patients with pulmonary infarction at autopsy, Miller and Berry observed a WBC count of >10,000/mm3 in 79 of 104 (76%) [4]. The WBC count was 15,000–20,000/mm3 in 34 of 104 (33%). It was >20,000/mm3 in 17 of 104 (16%).
References 1 Afzal A, Noor HA, Gill SA, Brawner C, Stein, PD. Leukocytosis in acute pulmonary embolism. Chest 1999; 115: 1329–1332. 2 Sasahara AA, Stein M. PulmonaryEmbolic Disease. Grune & Straton, New York, 1965: 258–259. 3 Gsell O. Der hamorrhagische lungeninfarct und seine komplikationen (Infarktpleuritis, infarktpnoumonie, infarktkaverne usw). Deut Med Wchschr 1935; 61:1317– 1360. 4 Miller R, Berry JB. Pulmonary infarction: a frequently missed diagnosis. Am J Med Sci 1951; 222: 197–206.
BLUK077-Stein
March 28, 2007
50
15:30
CHAPTER 50
Alveolar dead-space in the diagnosis of pulmonary embolism
Alveolar dead-space volume occurs in areas of the lung that are ventilated, but not perfused and that contain a very low partial pressure of carbon dioxide (PCO2 ) [1]. Exhaled dead-space volume dilutes the total amount of CO2 in exhaled breaths relative to the arterial partial pressure of CO2 (Pa CO2 ). Therefore, the alveolar dead-space volume can be estimated by simultaneously measuring CO2 in exhaled breaths and Pa CO2 [1]. Physiological dead-space fraction = VDphys /VT = (Pa CO2 − PE CO2 )/Pa CO2 where PE CO2 is the partial pressure of mixed expired CO2 . VDphys = physiological dead-space and VT = tidal volume. Because the physiological dead-space fraction (VDphys /VT ) measures the fraction of each tidal volume that is wasted on alveolar and airway dead-space, the airway dead-space fraction was subtracted from the physiological dead-space fraction. This yielded the alveolar dead-space fraction (VDalv ). Multiplication by 100 gives the fraction as a percentage. The airway deadspace volume was calculated from geometric analysis of the CO2 volume curve [2, 3]. Normal alveolar deadspace was defined as ≤20% [1]. Some calculate “late dead-space fraction” [3]. This represents the dead-space fraction of CO2 at the end of a hypothetical expiration with a volume equal to 15% of predicted total lung capacity [3]. Volumetric capnograms were obtained at the bedside using a commercially available volumetric-based capnography machine, which can simultaneously record time-based capnograms, expired CO2 content, breath volume, and flow data. Measurement of the alveolar dead-space volume with volumetric capnography and blood gas analysis can be completed in less than 5 minutes [2]. The most timeconsuming portion of the measurement was the col-
234
lection and processing of arterial blood for blood gas analysis [2]. The end-tidal to arterial PCO2 gradient as well as physiological dead-space is often increased in obstructive lung disease and interstitial lung disease [3]. These techniques, therefore, suffer an inherent limited ability to differentiate between pulmonary embolism (PE) and other pulmonary disease [3]. Efforts to increase the specificity of measurement include measurements after a deep expiration, which reduces the arterialend expiratory gradient in obstructive lung disease, but not in PE [3, 4]. Others suggested that measurements could be made in combination with spirometry [3, 5]. Physiological dead-space fraction, alveolar deadspace fraction, and late dead-space fraction are shown in Table 50.1. The sensitivities and specificities varied widely [1, 3, 5–7]. On the other hand, alveolar dead-space fraction [1, 6, 7] or a change in alveolar dead-space fraction [1] when used in combination with D-dimer had a high sensitivity, but it was not specific (Table 50.2). The combination has a high negative predictive value (Table 50.2). Alveolar dead-space fraction, therefore, in the opinion of the authors, may be useful for exclusion of PE when used in combination with a negative D-dimer [1]. The alveolar dead-space fraction showed a correlation with systolic pulmonary artery pressure and an approximate correlation with the percent perfusion defect [2]. Among patients in whom bedside evaluation showed any two of the following—alveolar deadspace fraction ≤15%, D-dimer normal, or Wells score “unlikely”, PE was safely excluded in a higher proportion of patients than by only an “unlikely” Wells score plus a normal D-dimer [8]. If any two of these assessments was normal or “unlikely,” PE was excluded in 34% of the patients in whom the diagnosis was suspected, and venous thromboembolism (VTE) on a 3month follow-up occurred in only 2.4% [8]. If clinical
BLUK077-Stein
March 28, 2007
CHAPTER 50
15:30
235
Alveolar dead-space in diagnosis of PE
Table 50.1 Physiological dead-space fraction, alveolar dead-space fraction, and late dead-space fraction in pulmonary embolism. n/N (%) First author (Ref.)
Test
Cutoff value
Sensitivity
Specificity
Burki [5]
VDphys /VT
>0.4
16/16 (100)
16/29 (55)
Eriksson [3]
VDphys /VT
>0.4
8/9 (89)
22/29 (76)
Eriksson [3]
fDlate
>0.2
6/9 (67)
Kline [6]
VDalv /VT
0.2
23/26 (88)
Johanning [7]
VDalv /VT
>0.45
Kline [1]
VDalv /VT *
>0.2
4/7 (57)* 43/64 (67%)
29/29 (100) 135/144 (94) 14/14 (100) 241/316 (76)
*VDalv /VT calculated as VDphys /VT – airway dead-space fraction. VDphys , physiological dead-space; VT , tidal volume; fDlate , late dead-space fraction; VDalv /VT , alveolar dead-space fraction, = (Pa CO2 – end-tidal CO2 )/Pa CO2 ). Table 50.2 Dead-space in combination with D-dimer in pulmonary embolism. Negative First author (Ref.)
Test
Index value
D-dimer
Sensitivity
Kline [6]
VDalv /VT
≥0.2
Semiquant latex agglut
26/26 (100)*
Johanning [7]
VDalv /VT
>0.45
Semiquant latex agglut
Kline [1]
VDalv /VT‡
>0.2
Whole blood agglut
4/5 (80%) 63/64 (98)§
Specificity
‡
95/144 (66)† 16/16 (100%) 163/316 (52)†
predictive value 95/95 (100) 16/17 (94) 163/164 (99)
*Both tests positive. † Both tests negative. ‡ PE positive required a dead-space level of >0.45 or a change in dead-space of >0.10 from control values before symptoms in combination with a positive D-dimer. PE negative required a dead-space level of <0.45 and a change in dead-space of <0.10 in combination with a negative D-dimer. § Either test positive. Semiquant, semiquantitative; agglut, agglutination.
assessment plus D-dimer alone were used, PE was excluded in 18% [8].
References 1 Kline JA, Israel EG, Michelson EA, O’Neil BJ, Plewa MC, Portelli DC. Diagnostic accuracy of a bedside D-dimer assay and alveolar dead-space measurement for rapid exclusion of pulmonary embolism: a multicenter study. JAMA 2001; 285: 761–768. 2 Kline JA, Kubin AK, Patel MM et al. Alveolar dead space as a predictor of severity of pulmonary embolism. Acad Emer Med 2000; 7: 611–617. 3 Eriksson L, Wollmer P, Olsson CG et al. Diagnosis of pulmonary embolism based upon alveolar dead space analysis. Chest 1989; 96: 357–362. 4 Hatle L, Rokseth R. The arterial to end-expiratory carbon dioxide tension gradient in acute pulmonary embolism and other cardiopulmonary diseases. Chest 1974; 66: 352– 357.
5 Burki NK. The dead space to tidal volume ratio in the diagnosis of pulmonary embolism. Am Rev Respir Dis 1986; 133: 679–685. 6 Kline JA, Meek S, Boudrow D, Warner D, Colucciello S. Use of the alveolar dead space fraction (Vd/Vt) and plasma D-dimers to exclude acute pulmonary embolism in ambulatory patients. Acad Emerg Med 1997; 4: 856– 863. 7 Johanning JM, Veverka TJ, Bays RA, Tong GK, Schmiege SK. Evaluation of suspected pulmonary embolism utilizing end-tidal CO2 and D-dimer. Am J Surg 1999; 178: 98–102. 8 Rodger MA, Bredeson CN, Jones G et al. The bedside investigation of pulmonary embolism diagnosis study: a double-blind randomized controlled trial comparing combinations of 3 bedside tests vs ventilation– perfusion scan for the initial investigation of suspected pulmonary embolism. Arch Intern Med 2006; 166: 181– 187.
BLUK077-Stein
March 12, 2007
51
18:30
CHAPTER 51
Neural network computer-assisted diagnosis
We attempted to determine if objective clinical characteristics that can be readily identified by physicians without special knowledge of pulmonary embolism (PE) could be evaluated by computer to achieve a correct estimate of the likelihood of PE [1]. To accomplish this, we used a neural network [1]. A neural network is an artificial intelligence paradigm that integrates multiple clinical variables to arrive at a correct diagnosis. It has been used for clinical problems such as myocardial infarction and low back disorders [2, 3]. Neural network paradigms also have
been applied to the evaluation of ventilation–perfusion lung scans, with good results [4, 5]. Neural network paradigms have been reported that perform computerized interpretations of scintillation data on thallium scans of the heart [6]. Artificial neural network models were developed by presenting clinical data to the neural network in the form of numeric input variables [1]. Fifty characteristics of the history, physical examination, chest radiograph, electrocardiogram, and arterial blood gases were presented to the computer algorithm (Table 51.1).
Table 51.1 Variables presented to the neural network. Alveolar–arterial gradient Pa O2 ∗
Complaints of dyspnea
Atrial flutter
Hemoptysis (≤1 mo) Stroke, ever
Hypoxia (Pa O2 < 60 mm Hg) QRS axis >120 and <−30 degrees
Postpartum (≤3 mo) Cough (≤3 days)
ST-segment and/or T-wave changes
Pa CO2 ∗∗
Recent surgery (≤3 mo) Lower extremity trauma (≤3 mo)
Cyanosis Prominent pulmonary artery (radiograph)
Angina-like pain (≤3 day) Palpitations (≤3 day)
Pulmonary edema (radiograph) Right ventricular hypertrophy (ECG)
History of chronic obstructive pulmonary disease A–a gradient > twice age-corrected History of immobilization (≤3 mo) T-wave changes alone Incomplete right bundle branch block
Asthma, ever Estrogen use (at onset of symptoms) Pleuritic chest pain (≤3 days) Wheezing Pleural friction rub
Loss of consciousness (≤1 wk) Calf tenderness Thrombophlebitis (ever) Leg or foot swelling, not joints
Cardiomegaly (radiograph) ST-segment depression alone P pulmonale Respiratory alkalosis (Pa CO2 ≤ 30 mm Hg)
Any leg pain, not arthritic
Palpable venous cord
Leg erythema Homans’ sign Low-voltage QRS complexes
Left ventricular hypertrophy (ECG) Atrial fibrillation Diaphoresis
Premature ventricular contractions
Premature atrial contractions Malignancy (≤3 mo) Complete right bundle branch block
∗
Pa O2 = partial pressure of oxygen in arterial blood. Pa CO2 = partial pressure of carbon dioxide in arterial blood.
∗∗
236
BLUK077-Stein
March 12, 2007
CHAPTER 51
18:30
237
Neural network computer-assisted diagnosis
Data were obtained from patients recruited in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). The interpretations of the ventilation–perfusion scans were not included in the data presented to the neural network. Variables were coded as either binary variables (present or absent) or continuous variables. The percentage of patients with PE was determined for deciles of risk estimated on the basis of clinical assessment by physicians in PIOPED [1]. This was compared with deciles of risk estimated by neural network clinical assessment (Figure 51.1). In general, there was good agreement between the likelihood of PE estimated by physicians in PIOPED and estimated by the neural network. The areas under the receiver operating curves (ROC) [7] relating sensitivity to the frequency of false-positive values were calculated for clinical assessment estimated by the neural network and by intuitive clinical assessment The areas under the ROC did not differ significantly, indicating that the neural networks performed as well in achieving a clinical diagnosis as physicians
knowledgeable and experienced in the diagnosis of acute PE. Data that were of particular value to the neural network in making or rejecting the diagnosis of PE were identified [1]. Among these were a high alveolar– arterial gradient, a low partial pressure of oxygen in arterial blood (Pa O2 ) on room air, atrial flutter, marked right-axis or left-axis deviation, ST-segment depression, T-wave inversion, and surgery within the last 3 months. It is important to keep in mind that all of the patients who were recruited in PIOPED were identified on the basis of well-known clinical characteristics suggestive of acute PE. Consequently, the well-known features of acute PE were present in patients in whom the diagnosis was made and also in patients in whom the diagnosis was excluded [8, 9]. Typical characteristics of PE were not useful to the neural network in distinguishing between patients with PE and those without PE. Even though similar signs and symptoms were present in patients with and without PE, experienced physicians in PIOPED were often able to correctly
100 90
Neural (n =1213)
80
PIOPED (n =1213)
70
PE (%)
60 50 40 30 20 10
91 –1 00
81 –9 0
71 –8 0
61 –7 0
51 –6 0
41 –5 0
31 –4 0
21 –3 0
11 –2 0
0– 10
0
Predicted PE (%) Figure 51.1 Percentage of patients with pulmonary embolism (PE) for deciles of the predicted risk based on PIOPED clinical assessment and on neural network clinical assessment. (Reprinted with permission from Patil et al. [1].)
BLUK077-Stein
March 12, 2007
18:30
238
estimate the likelihood of PE, based on undefined clinical judgment [10]. This suggests that there were details of the clinical characteristics that were not recorded in the PIOPED database. It also suggests that there may be recorded combinations of observations that were weighted by physicians in reaching a correct clinical assessment.
PART III
6
7
References 1 Patil S, Henry JW, Rubenfire M, Stein PD. Neural network in the clinical diagnosis of acute pulmonary embolism. Chest 1993; 104: 1685–1689. 2 Baxt WG. Use of an artificial neural network for the diagnosis of myocardial infarction. Ann Intern Med 1991; 115: 843–848. 3 Bounds DG, Lloyd PJ. A comparison of neural network and other pattern recognition approaches to the diagnosis of low back disorders. Neural Netw 1990; 3: 583–591. 4 Scott JA, Palmer EL. Neural network analysis of ventilation–perfusion lung scans. Radiology 1993; 186: 661–664. 5 Tourassi GD, Floyd CE, Sostman HD, Coleman RE. Ar-
8
9
10
Diagnosis of acute PE
tificial neural network for diagnosis of acute pulmonary embolism: effect of case and observer selection. Radiology 1995; 194: 889–893. Fujita H, Katafuchi T, Uehara T, Nishimura T. Application of artificial neural network to computer-aided diagnosis of coronary artery disease in myocardial SPECT bull’s-eye images. J Nucl Med 1992; 33(2): 272–276. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143: 29–36. Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. A Collaborative Study by the PIOPED Investigators. Value of the ventilation–perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.
BLUK077-Stein
April 6, 2007
52
16:5
CHAPTER 52
Empirical assessment and clinical models for diagnosis of acute pulmonary embolism
The most prominent standardized clinical scores based on objective or mostly objective clinical findings are the Wells (or “Canadian”) extended clinical model [1–3], the Wells simplified model [3–5], and the original “Geneva” clinical model [5, 6]. More recently, a revised Geneva clinical model was developed [7]. A Pisa model has also been developed [8]. In addition, many use an empirical assessment [3, 6, 9]. Differences in the Wells simplified model [4] and the Wells extended model [2] relate to whether leg swelling is measured (extended model) or appears swollen (simplified model) and whether pulmonary embolism (PE) is as likely or more likely than an alternative diagnosis (extended model) or just more likely (simplified model). The Wells scoring system is shown in Table 52.1. A Wells score of <2 points indicates a
low probability of PE, 2–6 points indicates a moderate probability, and >6 points indicates a high probability of PE [1, 4]. A Wells score of ≤4 points also has been used to indicate that PE is “unlikely” and >4 points indicates that PE is likely [4, 10]. The Wells extended and simplified models are not strictly objective. Judgment is required to evaluate the final variable “if pulmonary embolism is as likely or more likely than an alternative diagnosis.” To make this judgment, physicians may use the history, physical examination, chest radiograph, electrocardiogram, and whatever blood tests are considered necessary to diagnose PE [1]. Scoring for the original Geneva system is shown in Table 52.2. The original Geneva scoring system, although objective, requires an arterial blood gas analysis. A revised Geneva score has been developed based
Table 52.1 Clinical scoring system for suspected pulmonary embolism (Wells). Clinical feature*
Wells extended model*
Wells simplified model†
Score
Signs/symptoms of DVT
Measured leg swelling plus pain
Leg swelling plus pain with
3.0
with palpation in deep vein region
palpation in deep vein region
Heart rate >100/min
Yes
Yes
1.5
Immobilization: bedrest
Yes
Yes
1.5
Prior PE or DVT
Yes
Yes
1.5
Hemoptysis
Yes
Yes
1.0
Malignancy: treated within
Yes
Yes
1.0
PE as likely or more likely
PE more likely
3.0
(± bathroom privileges) ≥3 consecutive days or surgery in last 4 wk
6 mo or on palliative care Alternative diagnosis
Low probability of PE: score <2; moderate probability: score 2–6; high probability: score >6; unlikely probability: score ≤4† ; likely probability: score >4. ∗Wells et al. [1]. † Wells et al. [4]. PE, pulmonary embolsim; DVT, deep venous thrombosis.
239
BLUK077-Stein
April 6, 2007
16:5
240
PART III
Diagnosis of acute PE
Table 52.2 Original Geneva scoring system for suspected pulmonary embolism.
Table 52.3 Revised Geneva scoring system for suspected pulmonary embolism.
Feature
Feature
Score
Score
Age >65 yr
1
60–79
1
Previous PE or DVT
3
≥80
2
Surgery (under general anesthesia) or fracture of
2
Age (yr)
Previous PE or DVT
2
Recent surgery
3
Heart rate > 100/min
1
lower limbs (≤1 mo) Solid or hematologic malignant condition,
2
currently active or considered cured <1 yr
Pa CO2
Unilateral lower limb pain
3
<4.8 kPa (<36 mm Hg)
2
Hemoptysis
2
4.8–5.19 kPa (36–39 mm Hg)
1
Heart rate
Pa O2 <6.5 kPa (<49 mm Hg)
4
6.5–7.99 kPa (49–60 mm Hg)
3
8–9.49 kPa (61–71 mm Hg)
2
9.5–10.99 kPa (71–82 mm Hg)
1 1
Elevation of hemidiaphragm
1
3
≥95 beats/min
5
Pain on lower limb deep venous palpation and
4
unilateral edema Low probability of PE: score 0–3; intermediate probability: score 4–10; high probability: score ≥11. PE, pulmonary embolsim; DVT, deep venous thrombosis. From Le Gal et al. [7], with permission.
Chest X-ray Plate-like atelectasis
75–94 beats/min
Low probability of PE: score 0–4; moderate probability: score 5–8; high probability: score ≥9. Wicki et al. [6]. PE, pulmonary embolsim; DVT, deep venous thrombosis. Pa O2 = partial pressure of oxygen in arterial blood. Pa CO2 = partial pressure of carbon dioxide in arterial blood.
entirely on clinical variables and independent of arterial blood gases (Table 52.3) [7]. A “Pisa” clinical model showed excellent results (Table 52.4), but is more difficult to apply [8]. The Wells extended clinical model and simplified clinical model are applicable to inpatients as well as
outpatients [1, 2]. The original Geneva model and the revised Geneva model are applicable to patients seen on an emergency service [6, 7]. The Wells extended and simplified scores [2, 4], and the original and revised Geneva scores give points for recent surgery [6, 7]. The simplified Wells prediction rule and the original Geneva prediction rule performed poorly in patients referred from surgical wards [11]. Results with these prediction rules are shown in Table 52.4. Comparisons between the simplified Wells system and the original Geneva system in the same
80
0 Clin Low
Clin Moderate
Geneva revised
Geneva original
Empirical Geneva revised
Wells simplified Geneva original
10
Empirical Wells extended
20
Geneva revised
30
Wells simplified
40
Geneva original
50
Wells extended Wells simplified
60
Empirical Wells extended
Positive predictive value (%)
70
Clin High
Figure 52.1 Positive predictive value for pulmonary embolism of low, moderate, and high probability clinical assessments using empirical judgment, Wells extended model, Wells simplified model, original Geneva model, and revised Geneva model. Data based on values shown in Table 52.4.
BLUK077-Stein
April 6, 2007
CHAPTER 52
16:5
241
Empirical assessment and clinical models
Table 52.4 Positive predictive value of empirical clinical assessment and clinical prediction scores for pulmonary embolism. PE / N (%) Method/first author (Ref.)
Low probability
Moderate probability
High probability
Unlikely
Likely
9/178 (5%)
27/69 (39%)
Empirical Sanson [3]
80/278 (29)
35/77 (46)
PIOPED [9]
21/228 (9)
11/58 (19)
170/569 (30)
61/90 (68)
Wicki [6]
33/368 (9)
174/523 (33)
62/94 (66)
Average
65/654 (10)
424/1390 (31)
158/261 (61)
25/734 (3)
Wells extended Wells [1]
112/403 (28)
80/102 (78)
Sanson [3]
17/60 (28)
54/138 (39)
18/39 (46)
Wells [2]
7/357 (2)
45/168 (27)
28/43 (65)
Average
49/1151 (4)
211/709 (30)
126/184 (68)
2/99 (2)
24/128 (19)
10/20 (50)
41/147 (28)
78/259 (30)
3/8 (38)
Wells simplified Wells [4] Sanson [3] Chagnon [5]
19/162 (12)
42/104 (40)
10/11 (91)
Average
62/408 (15)
144/491 (29)
23/39 (59)
Wicki [6]
49/486 (10)
166/437 (38)
51/63 (81)
Chagnon [5]
20/152 (13)
43/113 (38)
8/12 (67)
Average
69/638 (11)
209/550 (38)
59/75 (79)
18/229 (8)
132/463 (29)
42/57 (74)
19/432 (4)
62/283 (22)
306/313 (98)
Geneva original
Geneva revised Le Gal [7] Pisa Miniati [8] PE, pulmonary embolsim.
patients were made by Chagnon and associates [5] and are also shown in Table 52.4. Receiver operating characteristic curve analysis showed no difference between results with these prediction rules [5]. The positive predictive value of a low-probability clinical assessment, whether empirical or by any of the prediction rules (Wells extended, Wells simplified, Geneva original, Geneva revised, or Pisa), on average, was 4–15% (Table 52.4, Figure 52.1). The positive predictive value of a moderate- or intermediateprobability clinical assessment was 22–38%. Highprobability clinical assessment showed PE in 59–98%.
2
3
4
References 1 Wells PS, Anderson DR, Rodger M et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pul-
5
monary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001; 135: 98–107. Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. Sanson BJ, Lijmer JG, Mac Gillavry MR, Turkstra F, Prins MH, Buller HR, for the ANTELOPE-Study Group. Comparison of a clinical probability estimate and two clinical models in patients with suspected pulmonary embolism. Thromb Haemost 2000; 83: 199–203. Wells PS, Anderson DR, Rodger M et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000; 83: 416– 420. Chagnon I, Bounameaux H, Aujesky D et al. Comparison of two clinical prediction rules and implicit assessment
BLUK077-Stein
April 6, 2007
16:5
242
among patients with suspected pulmonary embolism. Am J Med 2002; 113: 269–275. 6 Wicki J, Pernerger TV, Junod AF et al. Assessing clinical probability of pulmonary embolism in the emergency ward: a simple score. Arch Intern Med 2001; 161: 92–97. 7 Le Gal G, Righini M, Roy P-M, et al. Prediction of pulmonary embolism in the emergency department: the revised Geneva score. Ann Intern Med 2006; 144: 165– 171. 8 Miniati M, Monti S, Bottai M. A structured clinical model for predicting the probability of pulmonary embolism. Am J Med 2003; 114: 173–179.
PART III
Diagnosis of acute PE
9 A Collaborative Study by the PIOPED Investigators. Value of the ventilation–perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 10 van Belle A, Buller HR, Huisman MV et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA 2006; 295: 172–179. 11 Ollenberger GP, Worsley DF. Effect of patient location on the performance of clinical models to predict pulmonary embolism. Thromb Res 2006; 118: 685–690.
BLUK077-Stein
March 28, 2007
53
15:31
CHAPTER 53
D-dimer for the exclusion of acute pulmonary embolism
Tests for D-dimer have been available since the 1980s for use in the exclusion of venous thromboembolic disease [1]. In spite of this extensive literature, the role of D-dimer in the diagnosis of deep vein thrombosis (DVT) or pulmonary embolism (PE) was unclear, in part due to the multiple D-dimer assays that are available, the role of central laboratory testing versus bedside testing, and concerns about variability of the assays [1]. A systematic review of the literature showed that in ranking the D-dimer assays according to the sensitivity values and likelihood of increasing certainty for ruling out PE, the values for sensitivity for the ELISA (enzyme-linked immunosorbent assay) and quantitative rapid ELISA assays were significantly superior to those for the semiquantitative latex and whole blood agglutination assays [2]. The quantitative rapid ELISA assay is more convenient than the conventional ELISA, and provides a high level of certainty for a negative diagnosis of PE as well as DVT [2]. In this chapter, D-dimer as a stand-alone test for the exclusion of acute PE will be discussed. D-dimer combined with clinical probability assessment for the exclusion of PE will be discussed in Chapter 54. Chapter 31 discussed D-dimer as a stand-alone test for the exclusion of DVT. D-dimer for the exclusion of DVT in combination with clinical probability was reviewed in Chapter 32. A brief description of various D-dimer tests is shown in Table 53.1. Tier 1 (11 studies) compared an ELISA assay and at least one other D-dimer assay [3–13]. Tier 2 studies included the Tier 1 studies and 20 other studies that met all inclusion criteria [6, 14–33]. There were, therefore, 31 investigations that met all inclusion criteria, and these comprised Tier 2 (Table 53.2). Tier 3 combined 10 methodologically weaker studies [34–43] with the 31 Tier 2 studies. Sensitivity, specificity, and likelihood ratios differed according to whether D-dimer was being used to exclude PE or DVT. Non-ELISA assays should not be used as stand-alone tests for PE [44] given their in-
ferior sensitivity and negative likelihood ratio values. However, in patients with a low clinical probability of PE, non-ELISA assays provide a reasonable certainty for ruling it out [2]. In patients with PE, the highest sensitivity was seen among the ELISA assays using a cutoff level of 500 ng/mL (Figure 53.1) [2]. The variability of the sensitivity values is generally less for the ELISA assays as a group (Figure 53.1), probably reflecting assay reproducibility given the very large database in this systematic review [2]. The ELISA and quantitative rapid ELISA had a negative likelihood ratio of 0.13 in Tier 1 studies (Table 53.3), 0.07 in Tier 2 studies (Table 53.3), and 0.05 in Tier 3 studies (Table 53.4) [2]. Likelihood ratios of <0.1 generate large and often conclusive changes from pretest to posttest probability [45] providing high certainty for excluding PE. Combining a negative rapid ELISA with a low or moderate clinical probability for PE essentially rules out PE (Chapter 31). However, a negative D-dimer does not reliably exclude PE if the clinical probability is high (Chapter 31). Furthermore, a negative D-dimer has the highest negative predictive values in populations with a low prevalence of PE. It is possible to estimate the negative predictive value of the quantitative rapid ELISA for exclusion of acute PE from the sensitivity (97%) and specificity (41%) (based on Tier 2 studies) knowing the prevalence of PE in the population tested. The prevalence of PE in the included investigations of D-dimer ranged from 8 to 62% (average 25%) [2]. If the prevalence of PE in the population to be tested were 25%, the negative predictive value of a negative quantitative rapid ELISA would be 98% [2] (Figure 53.2). If the prevalence of PE were 8%, the negative predictive value would be 99%. If the prevalence of PE were 62%, the negative predictive value would be only 89%. Limited data for sensitivity, specificity, and likelihood ratios were available for cutoff values of 250 and
243
BLUK077-Stein
March 28, 2007
15:31
244
PART III
Diagnosis of acute PE
Table 53.1 Brief description of D-dimer tests. Type of test
Description
Time to results
Enzyme-linked
The conventional ELISA is considered a reference standard
Approximately 8 h; may
immunosorbent assay
for determination of D-dimer concentration. The
not be available the
(ELISA)
conventional ELISA is not practical for use in the
same day
diagnosis of venous thromboembolism in individual patients, because it is laboratory intensive. Availability in most centers is limited Quantitative rapid ELISA
The quantitative rapid ELISA test uses an antibody to
Approximately 35 min
D-dimer that is fluorescent labeled. An automated immunoanalyzer provides a numerical result Semiquantitative rapid ELISA
The semiquantitative rapid ELISA employs an antibody to
Less than 10 min
D-dimer that is tagged to a color-producing agent. Plasma D-dimer concentration is quantified from the degree of color intensity produced
Qualitative rapid ELISA
The qualitative rapid ELISA method involves the presence
10 min
of D-dimer detected by an anti-D-dimer monoclonal antibody coupled to alkaline phosphatase. Activity is revealed by the addition of a substrate that causes a color change. The intensity of the color is read visually by comparison with a positive and negative control Quantitative latex
Quantitative latex agglutination assays use monoclonal
agglutination
antibodies to D-dimer that are coated onto latex
7–15 min
particles. In the presence of D-dimer, the particles aggregate to form larger aggregates and light scattering decreases. Quantification of the D-dimer concentration is done with an analyzer that detects agglutination and precipitation turbidimetrically Semiquantitative latex agglutination
Semiquantitative latex agglutination assays rely on the
3–4 min
use of monoclonal antibodies to D-dimer that are coated onto latex particles. Macroscopic agglutinations are seen when elevated D-dimer levels are present in the plasma
Whole blood agglutination
The most frequently studied whole blood assay uses a
2 min
freshly collected drop of capillary or venous whole blood, which is mixed with a conjugate of monoclonal antibody to D-dimer. The latter is linked to a monoclonal antibody to human red blood cells. Visible agglutination of red cells takes place in the presence of elevated D-dimer levels. Interpreter experience is important due to difficulty in discriminating between weak positive and normal results
1000 ng/mL for some but not all of the D-dimer assays (Table 53.4) [2]. The clinical utility of the D-dimer assays is limited by the nonspecificity of a positive result due to factors such as inflammation, trauma, and surgery to name some. The clinical utility differs among patient populations. The proportion of inpatients in whom PE
can be excluded by a normal D-dimer is lower than in outpatients, because inpatients more often have unrelated disorders that cause a positive D-dimer test [46]. Similarly, the proportion of normal D-dimer tests in patients with a history of prior DVT or PE is lower than in patients without previous venous thromboembolic disease [47].
BLUK077-Stein
March 28, 2007
CHAPTER 53
15:31
245
D-dimer for exclusion of acute PE
Table 53.2 Patients with clinically suspected pulmonary embolism.
Study
Cutoff
No. of
Patients
analyzed
patients
with
enrolled
PE (%)
Year
D-dimer type
(ng/mL)
1991
Quant rapid ELISA, qual
<500
100
25.0
Tier 1 analysis Leitha [8]
latex Ginsberg [6]
1993
ELISA, semiquant latex
<500
221
19.5
Lenzhofer [9]
1993
ELISA, semiquant latex
<250
118
14.4
Flores [5]
1995
Quant rapid ELISA, quant
<500
85
36.5
<500, <1000
183
10.4, 10.4,
latex Veitl [12]
1996
ELISA, semiquant rapid ELISA, quant latex,
9.8, 10.4,
semiquant latex,
9.3
whole blood Duet [4]
1998
ELISA, quant latex
<500
85
18.8
Meyer [10]
1998
Qual rapid ELISA, quant
<500
142
35.9, 42.3
latex Oger [11]
1998
ELISA, quant latex
<500
386
37.8
Heit [7]
1999
Semiquant rapid ELISA,
≤ 500, < 250,
105
31.4, 31.4
<500
131
22.9, 25.2
<500
287
31.4
≤ 250
ELISA, quant latex, semiquant latex Wallis [13]
2000
Quant rapid ELISA, whole blood
de Monye [3]
2002
Quant rapid ELISA, quant latex
Tier 2 analysis (Tier 2 includes Tier 1 and the following patients) Bounameaux [15]
1989
ELISA
<500
99
9.1
Bounameaux [16]
1990
ELISA
<500
67
29.9
Bounameaux [17]
1992
ELISA
<500
170
32.4
Demers [21]
1992
ELISA
<500
156
15.4
Pappas [28]
1993
Semiquant latex
<250
20
35.0
Ginsberg [24]
1995
Whole blood
—
86
18.6
Reber [31]
1995
ELISA, semiquant rapid
<500
301
30.9
de Moerloose [20]
1996
Quant rapid ELISA
<500
195
23.6
Perrier [29]
1996
ELISA
<500
308
35.4
Bonnin [14]
1997
Quant rapid ELISA
<500
83
22.9
Knecht [26]
1997
Quant latex
<500
154
10.4
Perrier [30]
1997
ELISA
<500
671
29.2
Egermayer [22]
1998
Whole blood
—
499
8.0
Ginsberg [25]
1998
Whole blood
—
1177
16.7
de Groot [19]
1999
Whole blood
—
245
24.9
Kruip [32]
1999
Quant rapid ELISA
<500
137
24.1
ELISA
Kutinsky [33]
1999
Semiquant latex
<250, <500
98
30.6
Bova [18]
2000
Semiquant rapid ELISA
<300
140
32.1
Farrell [23]
2000
Whole blood
—
198
20.7
MacGillavry [27]
2001
Whole blood
—
403
31.0
PE, pulmonary embolism; Quant, quantitative; qual, qualitative; semiquant, semiquantitative. Reprinted with minor modifications with permission from Stein et al. [2].
Negative predictive value (%) 1.0
99
8 E Q L ra ua ISA p i lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit Se x as ativ m s e la iqu ay te an x as tita sa tiv y e W ho l e as b sa loo y d
E Q L ra ua ISA p i lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit Se x as ativ m s e la iqu ay te an x as tita sa tiv y e W ho l e as b sa loo y d
Proportion
March 28, 2007
E Q L ra ua ISA pi lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit x Se as ativ m s e la iqu ay te an x t as ita sa tiv y e W ho l as e b sa loo y d
E Q L ra ua ISA pi lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit x Se as ativ m s e la iqu ay te an x t as ita sa tiv y e W ho l as e b sa loo y d
Proportion
BLUK077-Stein
15:31
Tier 1
0.95 0.7
0.6
0.85 0.5
0.4
0.75 0.3
Tier 2
0.6
80
25
Prevalence of PE (%)
246 0.8
0.9 0.7
0.8 0.6
0.7 0.5
0.4
0.5 0.3
Sensitivity Specificity
Figure 53.1 Boxplots of the finding for sensitivity and specificity among the D-dimer assays for patients with suspected pulmonary embolism. (Reprinted with permission from Stein PD et al. [2].)
100 98
89
60
40
20
0
62
Figure 53.2 Negative predictive value of D-dimer quantitative rapid ELISA (enzyme-linked immunosorbent assay) for pulmonary embolism (PE) in relation to prevalence of PE in population tested. (Data from Stein et al. [2].)
BLUK077-Stein
March 28, 2007
15:31
Table 53.3 Pulmonary embolism. Sensitivity
Specificity
Positive likelihood
Negative likelihood
Test
(95% CI)
(95% CI)
ratio (95% CI)
ratio (95% CI)
ELISA
0.95 (0.85–1.00)
0.44 (0.34–0.54)
1.68 (1.44–1.95)
0.13 (0.03–0.58)
Quant rapid ELISA
0.95 (0.83–1.00)
0.39 (0.28–0.51)
1.56 (1.32–1.83)
0.13 (0.02–0.84)
Tier 1 analysis
Semiquant rapid ELISA
0.93 (0.79–1.00)
0.36 (0.23–0.50)
1.45 (1.20–1.76)
0.20 (0.04–0.96)
Qual rapid ELISA
0.93 (0.74–1.00)
0.68 (0.50–0.87)
2.92 (1.77–4.79)
0.11 (0.01–0.93)
Quant latex
0.89 (0.81–0.98)
0.45 (0.36–0.53)
1.62 (1.43–1.84)
0.24 (0.13–0.45)
Semiquant latex
0.92 (0.79–1.00)
0.45 (0.31–0.59)
1.68 (1.35–2.09)
0.17 (0.04–0.78)
Whole blood
0.78 (0.64–0.92)
0.74 (0.60–0.88)
2.93 (1.89–4.52)
0.31 (0.18–0.51)
Tier 2 analysis ELISA
0.96 (0.88–1.00)
0.51 (0.44–0.59)
1.97 (1.72–2.26)
0.08 (0.01–0.43)
Quant rapid ELISA
0.97 (0.87–1.00)
0.41 (0.30–0.51)
1.64 (1.40–1.91)
0.07 (0.00–1.55)
Semiquant rapid ELISA
0.93 (0.79–1.00)
0.40 (0.27–0.54)
1.55 (1.25–1.93)
0.18 (0.04–0.94)
Qual rapid ELISA
0.91 (0.68–1.00)
0.70 (0.47–0.93)
3.01 (1.52–5.96)
0.13 (0.01–1.28)
Quant latex
0.89 (0.80–0.99)
0.47 (0.38–0.57)
1.69 (1.44–1.99)
0.23 (0.11–0.48)
Semiquant latex
0.80 (0.65–0.94)
0.56 (0.42–0.70)
1.81 (1.35–2.42)
0.36 (0.20–0.67)
Whole blood
0.83 (0.74–0.92)
0.64 (0.55–0.73)
2.32 (1.87–2.88)
0.27 (0.17–0.42)
Quant, quantitative; qual, qualitative; semiquant, semiquantitative. Reprinted with minor modifications with permission from Stein et al. [2]. Table 53.4 Pulmonary embolism.
Test
Sensitivity
Specificity
Positive likelihood
Negative likelihood
(95% CI)
(95% CI)
ratio (95% CI)
ratio (95% CI)
Tier 3 analysis: Cutoff 500 ng/mL (all data) ELISA
0.95 (0.88–1.00)
0.45 (0.38–0.53)
1.74 (1.55–1.96)
0.11 (0.03–0.39)
Quant rapid ELISA
0.98 (0.88–1.00)
0.40 (0.29–0.50)
1.62 (1.38–1.91)
0.05 (0.00–4.15)
Semiquant rapid ELISA
0.94 (0.81–1.00)
0.39 (0.26–0.52)
1.55 (1.24–1.92)
0.15 (0.02–1.13)
Qual rapid ELISA
0.92 (0.71–1.00)
0.68 (0.46–0.90)
2.92 (1.52–5.61)
0.11 (0.01–1.45)
Quant latex
0.90 (0.81–1.00)
0.46 (0.37–0.56)
1.68 (1.42–1.98)
0.21 (0.09–0.49)
Semiquant latex
0.86 (0.74–0.97)
0.51 (0.39–0.62)
1.73 (1.39–2.16)
0.29 (0.14–0.58)
Whole blood
0.82 (0.74–0.91)
0.63 (0.54–0.71)
2.21 (1.81–2.70)
0.28(0.18–0.43)
0.07
Tier 3 analysis: Cutoff 250 ng/mL (all studies meeting inclusion criteria). ELISA
0.96
0.55
2.13
Quant rapid ELISA
NA
NA
NA
NA
Semiquant rapid ELISA
NA
NA
NA
NA
Qual rapid ELISA
NA
NA
NA
NA
Quant latex
0.94
0.44
1.69
0.14
Semiquant latex
0.90
0.63
2.39
0.17
Whole blood
0.84
0.62
2.22
0.26
NA
Tier 3 analysis: Cutoff 1000 ng/mL (all studies meeting inclusion criteria) ELISA
NA
NA
NA
Quant rapid ELISA
NA
NA
NA
NA
Semiquant rapid ELISA
0.74
0.75
2.96
0.35
Qual rapid ELISA
NA
NA
NA
NA
Quant latex
NA
NA
NA
NA
Semiquant latex
NA
NA
NA
NA
Whole blood
0.84
0.62
2.22
0.26
Quant, quantitative; qual, qualitative; semiquant, semiquantitative. Reprinted with minor modifications with permission from Stein et al. [2].
247
BLUK077-Stein
March 28, 2007
15:31
248
References 1 Becker DM, Philbrick JT, Bachhuber TL, Humphries JE. D-dimer testing and acute venous thromboembolism. A shortcut to accurate diagnosis? Arch Intern Med 1996; 156: 939–946. 2 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 3 de Monye W, Sanson BJ, Buller HR, Pattynama PMT, Huisman MV. The performance of two rapid quantitative D-dimer assays in 287 patients with clinically suspected pulmonary embolism. Thromb Res 2002; 107: 283–286. 4 Duet M, Benelhadj S, Kedra W et al. A new quantitative D-dimer assay appropriate in emergency: reliability of the assay for pulmonary embolism exclusion diagnosis. Thromb Res 1998; 91: 1–5. 5 Flores J, Lancha C, Perez Rodriguez E, Garcia Avello A, Bollo E, Garvia Frade LJ. Efficacy of D-dimer and total fibrin degradation products evaluation in suspected pulmonary embolism. Respiration 1995; 62: 258–262. 6 Ginsberg JS, Brill-Edwards PA, Demers C, Donovan D, Panju A. D-dimer in patients with clinically suspected pulmonary embolism. Chest 1993; 104: 1679–1684. 7 Heit JA, Minor TA, Andrews JC, Larson DR, Li H, Nichols WL. Determinants of plasma fibrin D-dimer sensitivity for acute pulmonary embolism as defined by pulmonary angiography. Arch Pathol Lab Med 1999; 123: 235–240. 8 Leitha T, Speiser W, Dudczak R. Pulmonary embolism: efficacy of D-dimer and thrombin–antithrombin III complex determinations as screening tests before lung scanning. Chest 1991; 100: 1536–1541. 9 Lenzhofer R, Wimmer F, Haydl H et al. Prospective study of determining the value of D-dimer in diagnosing pulmonary embolism. Wien Klin Wochenschr 1993; 105: 492– 496. 10 Meyer G, Fischer AM, Collignon MA et al. Diagnostic value of two rapid and individual D-dimer assays in patients with clinically suspected pulmonary embolism: comparison with microplate enzyme linked immunosorbent assay. Blood Coagul Fibrinolysis 1998; 9: 603–608. 11 Oger E, Leroyer C, Bressollette L, et al. Evaluation of a new, rapid, and quantitative D-dimer test in patients with suspected pulmonary embolism. Am J Respir Crit Care Med 1998; 158: 65–70. 12 Veitl M, Hamwi A, Kurtaran A, Virgolini I, Vukovich T. Comparison of four rapid D-dimer tests for diagnosis of pulmonary embolism. Thromb Res 1996; 82: 399–407. 13 Wallis JW, Kruip M, de Jongh-Leuvenink J, B¨uller HR. A comparative study of two rapid D-dimer tests for the exclusion of pulmonary embolism in symptomatic patients [letter]. Thromb Haemost 2000; 84: 925.
PART III
Diagnosis of acute PE
14 Bonnin F, Hadjikostova H, Jebrak G et al. Complementarity of lung scintigraphy and D-dimer test in pulmonary embolism. Eur J Nucl Med 1997; 24: 444–447. 15 Bounameaux H, Slosman D, de Moerloose P, Reber G. Laboratory diagnosis of pulmonary embolism: value of increased levels of plasma D-dimer and thrombin– antithrombin III complexes. Biomed Pharmacother 1989; 43: 385–388. 16 Bounameaux H, Schneider PA, Slosman D, de Moerloose P, Reber G. Plasma D-dimer in suspected pulmonary embolism: a comparison with pulmonary angiography and ventilation–perfusion scintigraphy. Blood Coagul Fibrinolysis 1990; 1: 577–579. 17 Bounameaux H. Significance of the determination of Ddimers in venous thromboembolic disease. J Mal Vasc 1992; 17(suppl B): 91–92. 18 Bova C, Greco F, Ferrari A, Serafini O et al. The usefulness of the association of clinical probability, rapid plasma measurement of D-dimer, compression echography of the lower limbs and echocardiography in the diagnosis of acute pulmonary embolism. Ital Heart J 2000; 1: 116– 121. 19 de Groot MR, van Marwijk Kooy M, Pouwels JG, Engelage AH, Kuipers BF, Buller HR. The use of a rapid D-dimer blood test in the diagnostic work-up for pulmonary embolism: a management study. Thromb Haemost 1999; 82: 1588–1592. 20 de Moerloose P, Desmarais S, Bounameaux H et al. Contribution of a new, rapid, individual and quantitative automated D-dimer ELISA to exclude pulmonary embolism. Thromb Haemost 1996; 75: 11–13. 21 Demers C, Ginsberg JS, Johnston M, Brill-Edwards P, Panju A. D-dimer and thrombin–antithrombin III complexes in patients with clinically suspected pulmonary embolism. Thromb Haemost 1992; 67: 408– 412. 22 Egermayer P, Town GI, Turner JG, Heaton DC, Mee AL, Beard ME. Usefulness of D-dimer, blood gas, and respiratory rate measurements for excluding pulmonary embolism. Thorax 1998; 53: 830–834. 23 Farrell S, Hayes T, Shaw M. A negative SimpliRED Ddimer assay result does not exclude the diagnosis of deep vein thrombosis or pulmonary embolus in emergency department patients. Ann Emerg Med 2000; 35: 121– 125. 24 Ginsberg JS, Wells PS, Brill-Edwards P et al. Application of a novel and rapid whole blood assay for D-dimer in patients with clinically suspected pulmonary embolism. Thromb Haemost 1995; 73: 35–38. 25 Ginsberg JS, Wells PS, Kearon C et al. Sensitivity and specificity of a rapid whole-blood assay for D-dimer in the diagnosis of pulmonary embolism. Ann Intern Med 1998; 129: 1006–1011.
BLUK077-Stein
March 28, 2007
CHAPTER 53
15:31
D-dimer for exclusion of acute PE
26 Knecht MF, Heinrich F. Clinical evaluation of an immunoturbidimetric D-dimer assay in the diagnostic procedure of deep vein thrombosis and pulmonary embolism. Thromb Res 1997; 88: 413–417. 27 MacGillavry MR, Lijmer JG, Sanson BJ, Buller HR, Brandjes DP. Diagnostic accuracy of triage tests to exclude pulmonary embolism. Thromb Haemost 2001; 85: 995– 998. 28 Pappas AA, Dalrymple G, Harrison K et al. The application of a rapid D-dimer test in suspected pulmonary embolus. Arch Pathol Lab Med 1993; 117: 977– 980. 29 Perrier A, Bounameaux H, Morabia A et al. Diagnosis of pulmonary embolism by a decision analysis-based strategy including clinical probability, D-dimer levels, and ultrasonography: a management study. Arch Intern Med 1996; 156: 531–536. 30 Perrier A, Desmarais S, Goehring C et al. D-dimer testing for suspected pulmonary embolism in outpatients. Am J Respir Crit Care Med 1997; 156: 492–496. 31 Reber G, Vissac AM, de Moerloose P, Bounameaux H, Amiral J. A new, semi-quantitative and individual ELISA for rapid measurement of plasma D-dimer in patients suspected of pulmonary embolism. Blood Coagul Fibrinolysis 1995; 6: 460–463. 32 Kruip MJHA, Slob MJ, Buller HR, van der Heul C. A prospective clinical study on the feasibility and safety of a combined noninvasive/pulmonary angiography strategy in patients with suspected pulmonary embolism: an interim analysis. Clin Appl Thromb Haemost 1999; 5(4): 222–225. 33 Kutinsky I, Blakley S, Roche V. Normal D-dimer levels in patients with pulmonary embolism. Arch Intern Med 1999; 159: 1569–1572. 34 Goldhaber SZ, Vaughan DE, Tumeh SS, Loscalzo J. Utility of cross-linked fibrin degradation products in the diagnosis of pulmonary embolism. Am Heart J 1988; 116: 505–508. 35 Goldhaber SZ, Simons GR, Elliott CG et al. Quantitative plasma D-dimer levels among patients undergoing pulmonary angiography for suspected pulmonary embolism. JAMA 1993; 270: 2819–2822. 36 Gosselin RC, Owings JT, Utter GH, Jacoby RC, Larkin EC. A new method for measuring D-dimer using immunoturbidometry: a study of 255 patients with suspected pulmonary embolism and deep vein thrombosis. Blood Coagul Fibrinolysis 2000; 11: 715–721.
249
37 Harrison KA, Haire WD, Pappas AA et al. Plasma Ddimer: a useful tool for evaluating suspected pulmonary embolus. J Nucl Med 1993; 34: 896–898. 38 Johanning JM, Veverka TJ, Bays RA, Tong GK, Schmiege SK. Evaluation of suspected pulmonary embolism utilizing end-tidal CO2 and D-dimer. Am J Surg 1999; 178: 98–102. 39 Kline JA, Israel EG, Michelson EA, O’Neil BJ, Plewa MC, Portelli DC. Diagnostic accuracy of a bedside D-dimer assay and alveolar dead-space measurement for rapid exclusion of pulmonary embolism: a multicenter study. JAMA 2001; 285: 761–768. 40 LaCapra S, Arkel YS, Ku DH, Gibson D, Lake C, Lam X. The use of thrombus precursor protein, D-dimer, prothrombin fragment 1.2, and thrombin antithrombin in the exclusion of proximal deep vein thrombosis and pulmonary embolism. Blood Coagul Fibrinolysis 2000; 11: 371–377. 41 Lichey J, Reschofski I, Dissmann T, Priesnitz M, Hoffmann M, Lode H. Fibrin degradation product D-dimer in the diagnosis of pulmonary embolism. Klin Wochenschr 1991; 69: 522–526. 42 Quinn RJ, Nour R, Butler SP et al. Pulmonary embolism in patients with intermediate probability lung scans: diagnosis with Doppler venous US and D-dimer measurement. Radiology 1994; 190: 509–511. 43 Quinn DA, Fogel RB, Smith CD et al. D-dimers in the diagnosis of pulmonary embolism. Am J Respir Crit Care Med 1999; 159: 1445–1449. 44 Jaeschke R, Guyatt GH, Sackett DL, for the EvidenceBased Medicine Working Group. Users’ guides to the medical literature. III: How to use an article about a diagnostic test. B: What are the results and will they help me in caring for my patients? JAMA 1994; 271: 703– 707. 45 Lee AY, Julian JA, Levine MN et al. Clinical utility of a rapid whole-blood D-dimer assay in patients with cancer who present with suspected acute deep venous thrombosis. Ann Intern Med 1999; 131: 417–423. 46 Schrecengost JE, LeGallo RD, Boyd JC et al. Comparison of diagnostic accuracies in outpatients and hospitalized patients of D-dimer testing for the evaluation of suspected pulmonary embolism. Clin Chem 2003; 49: 1483–1490. 47 Le Gal G, Righini M, Roy PM et al. Value of D-dimer testing for the exclusion of pulmonary embolism in patients with previous venous thromboembolism. Arch Intern Med 2006; 166: 176–180.
BLUK077-Stein
April 6, 2007
54
16:8
CHAPTER 54
D-dimer combined with clinical probability for exclusion of acute pulmonary embolism
Combining a negative rapid ELISA (enzyme-linked immunosorbent assay) with a low or moderate clinical probability for pulmonary embolism (PE) essentially rules out the diagnosis of PE [1]. The non-ELISA assays when combined with a low clinical probability
for PE (but not a moderate probability) also provide a reasonable certainty of ruling out PE [1]. Figure 54.1 shows the relation of posttest probability to pretest probability using negative likelihood ratios according to Bayesian analysis [2]. Posttest probabilities with a
(a)
100 80 Whole blood
60 40 20
Quantitative rapid ELISA 95
85
75
65
55
45
35
25
15
0 5
D-dimer posttest probability (%)
120
Pretest probability (%)
20 18 16 14 12
Whole blood
10 8 6 4
Quantitative rapid ELISA
2
Pretest probability (%)
250
40
35
30
25
20
15
10
0 5
D-dimer posttest probability (%)
(b)
Figure 54.1 (a) Relation of posttest probability to pretest probability. (Data from Stein et al. [1].) (b) Enlargement of part (a).
BLUK077-Stein
April 6, 2007
CHAPTER 54
16:8
251
D-dimer and clinical probability for exclusion of PE
Table 54.1 Negative D-dimer with clinical evaluation to exclude PE: 3-month or 6-month outcome. PE/N (%)
First author
Wells score
Wells score
Wells score
Wells score
Wells score
(Ref.)
D-dimer
<2 (low)
≤ 4 (unlikely)
2–6 (moderate)
>4 (likely)
>6 (high)
Wells [3]
Whole blood
2/73 (2.7)
2/118 (1.7)
2/69 (2.9)
3/29(10.3)
1/5 (20)
Kearon [4]
Whole blood
1 or 2 /373
Christopher
Rapid ELISA
4/1026(0.4)†
Rapid ELISA
Geneva score
(0.3–0.5)* Group [5] Perrier [6]
0–8 (low or moderate) 0/220 (0) *6 mo outcome (1 PE and 1 “VTE,” 2 additional DVT on testing). † In addition, 1/1026 (0.1%) had DVT on follow-up. PE, pulmonary embolism; DVT, deep venous thrombosis; VTE, venous thromboembolsim; ELISA, enzyme-linked immunosorbent assay.
negative D-dimer assessed with the whole blood agglutination assay are shown in comparison with the quantitative rapid ELISA (Figure 54.1). If the pretest clinical probability for PE is high, for example, 70% or higher, the probability of PE, even with a negative D-dimer, irrespective of whether measured with whole blood agglutination or quantitative ELISA, is greater than 20%. Clearly, a high-clinical probability, irrespective of a negative D-dimer, indicates a need for further testing. The advantage of a negative quantitative rapid ELISA compared to a negative whole blood agglutination assay becomes apparent with a low or moderate clinical probability [1]. An enlargement (Figure 54.1b) shows that when the clinical probability is 4–15%, which is typical of a low-probability clinical assessment, the posttest probability with D-dimer testing by whole blood agglutination may be as high as 6%, but with the rapid ELISA it is 2%. With a 29–38% clinical probability, which is typical of a moderate-probability assessment, the posttest probability with whole blood agglutination may be as high as 16%, but with the rapid ELISA it is approximately 7% (Figure 54.1b). The prevalence of PE with levels of clinical assessment is shown in Chapter 52 “Empirical assessment and clinical models for diagnosis of acute pulmonary embolism.” The quantitative rapid ELISA method per-
mits a reasonably safe exclusion of PE with a lowor moderate-probability clinical assessment, but the whole blood agglutination may give uncertain results with a moderate-probability clinical assessment. Tests of whole blood D-dimer in combination with clinical assessment in small numbers of patients showed that PE occurred in 2.7% of the patients with a low-probability Wells score (score <2) on 3-month follow-up and PE occurred in 0.3–1.7% with an “unlikely” probability (score ≤ 4) [3, 4] (Table 54.1). Results were better in larger investigations using the rapid ELISA [4, 5]. With the D-dimer rapid ELISA in combination with an unlikely probability by Wells criteria, PE on 3-month follow-up occurred in 0.4% and an additional 0.1% had deep venous thrombosis (DVT) [5]. With the D-dimer rapid ELISA in combination with a low or moderate probability by original Geneva criteria, none had PE on 3-month follow-up [6]. The criteria for the Wells clinical assessment and the Geneva clinical assessment are described in Chapter 52 “Empirical assessment and clinical models for diagnosis of acute pulmonary embolism.” Among patients with suspected acute PE, a negative rapid ELISA in combination with an unlikely probability clinical assessment by Wells criteria or either a low or moderate clinical assessment by Geneva criteria occurred in 31–32% of the patients [5, 6] (Figure 54.2).
BLUK077-Stein
April 6, 2007
16:8
252
PART III
Diagnosis of acute PE
References
PE excluded Further tests necessary
31−32%
68−69%
Figure 54.2 Proportion of patients with suspected pulmonary embolism (PE) in whom the diagnosis of PE was excluded by a negative rapid ELISA D-dimer plus low or moderate /original Geneva score or unlikely clinical probability by the simplified Wells score. (Data from Christopher Study Investigators [5] and Perrier et al. [6].)
1 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of deep venous thrombosis and acute pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 2 Sox HC. Commentary. Ann Intern Med 2004; 140: 602. 3 Wells PS, Anderson DR, Rodger M et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000; 83: 416– 420. 4 Kearon C, Ginsberg JS, Douketis J et al. An evaluation of D-dimer in the diagnosis of pulmonary embolism: a randomized trial. Ann Intern Med 2006; 144: 812–821. 5 van Belle A, Buller HR, Huisman MV et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing and computed tomography. JAMA 2006; 295: 172–179. 6 Perrier A, Roy P-M, Sanchez O et al. Multidetectorrow computed tomography in suspected pulmonary embolism. N Engl J Med 2005; 352: 1760–1768.
BLUK077-Stein
April 14, 2007
55
14:58
CHAPTER 55
D-dimer in combination with amino-terminal pro-B-type natriuretic peptide for exclusion of acute pulmonary embolism
Fragments of pro-B-type natriuretic peptide (pro BNP), B-type natriuretic peptide (BNP), and aminoterminal B-type natriuretic peptide (NT-proBNP) are released in response to myocardial stretch and volume overload [1]. Elevations of natriuretic peptide have been described in patients with pulmonary embolism (PE) [2–5], particularly in those with right ventricular strain [2]. On the basis of indicating right ventricular overload, both BNP and NT-proBNP have been used to predict adverse outcomes in patients with PE [3– 5]. It was speculated that a low value of NT-proBNP in combination with a low value of D-dimer would strengthen the validity of an excluded diagnosis of PE over exclusion of PE on the basis of a low value of D-dimer alone [6]. A preliminary investigation calls attention to a potential advantage of using these two tests in combination [6]. Results, however, are not yet clinically applicable. D-dimer, measured by a quantitative latex agglutination test, and NT-proBNP were assessed in only 19 patients with dyspnea who were found to have PE and in 199 patients with dyspnea in whom PE was not necessarily suspected [6]. The patients in the control group most commonly had a pulmonary infection, exacerbation of chronic obstructive pulmonary disease, or asthma [6]. Among 199 patients with no PE, specificity for a low D-dimer alone was 72% and for a low D-dimer or low NT-proBNP specificity was 85% [6].
However, the negative predictive values were similar. The negative predictive value of D-dimer alone was 143 of 144 (99%) and of NT-proBNP was 115 of 119 (97%) [6]. The negative predictive value of the combination of a low D-dimer and low NT-proBNP was 169 of 173 (98%).
References 1 Cowie MR, Struthers AD, Wood DA et al. Value of natriuretic peptides in assessment of patients with possible new heart failure in primary care. Lancet 1997; 350: 1349–1353. 2 Pruszczyk P, Kostrubiec M, Bochowicz A et al. N-terminal pro-brain natriuretic peptide in patients with acute pulmonary embolism. Eur Respir J 2003; 22: 649–653. 3 Kucher N, Printzen G, Doernhoefer T, Windecker S, Meier B, Hess OM. Low pro-brain natriuretic peptide levels predict benign clinical outcome in acute pulmonary embolism. Circulation 2003; 107: 1576–1578. 4 Kucher N, Printzen G, Goldhaber SZ. Prognostic role of brain natriuretic peptide in acute pulmonary embolism. Circulation 2003; 107: 2545–2547. 5 ten Wolde M, Tulevski, II, Mulder JW et al. Brain natriuretic peptide as a predictor of adverse outcome in patients with pulmonary embolism. Circulation 2003; 107: 2082– 2084. 6 Melanson SE, Laposata M, Camargo CA, Jr, et al. Combination of D-dimer and amino-terminal pro-B-type natriuretic peptide testing for the evaluation of dyspneic patients with and without acute pulmonary embolism. Arch Pathol Lab Med 2006; 130: 1326–1329.
253
BLUK077-Stein
April 6, 2007
56
16:10
CHAPTER 56
Low tissue plasminogen activator plasma levels and low plasminogen activator inhibitor-1 levels as an aid in exclusion of acute pulmonary embolism
D-dimer is one of the last products derived from the thrombotic/fibrinolytic process. Its levels rise when thrombin activates fibrinogen, fibrin is formed, and factor XIIIa stabilizes it [1]. Then tissue plasminogen activator (tPA) activates plasminogen to plasmin and plasmin splits fibrin. One of the fibrin degradation products is D-dimer [1, 2]. This fibrinolytic process is regulated by plasminogen activator inhibitor type-1 (PAI-1) inhibiting tPA in the circulation and alpha2antiplasmin inhibiting free plasmin [3]. Fibrinolytic system activation means that tPA must be released into the circulation [4]. This conclusion led to the investigation of tPA levels and PAI-1 levels in patients with suspected acute pulmonary embolism (PE) [4]. Plasma levels were measured using an enzyme-linked immunosorbent assay (ELISA). Flores and associates evaluated tPA levels and PAI-1 levels in 27 patients with acute PE and 39 patients with suspected PE in whom the diagnosis was excluded [4]. A cutoff level of 8.5 ng/mL tPA was used. In 13 of 39 (33%) in whom PE was absent, the tPA level was lower than 8.5 ng/mL. Sensitivity was 27 of 27 (100%) and negative predictive value was 13 of 13 (100%).
254
Levels of PAI-1 were investigated using a cutoff level of 15 ng/mL [4]. In 9 of 39 (23%) in whom PE was absent, the PAI-1 level was lower than 15 ng/mL [4]. Sensitivity was 27 of 27 (100%) and negative predictive value was 9 of 9 (100%). This preliminary investigation suggests that tPA levels and PAI-1 levels below a specified cutoff value may potentially be useful for the exclusion of the diagnosis of acute PE.
References 1 Garcia Frade LJ, Sureda A. Degradation products of fibrinogen/fibrin: methods of determination and clinical utility. Rev Iberoam Thromb Hemost 1991; 4: 89–98. 2 Lijnen HR, Collen D. Mechanisms of physiological fibrinolysis. Baillieres Clin Haematol 1995; 8: 277–290. 3 Booth NA, Bennett B. Fibrinolysis and thrombosis. Baillieres Clin Haematol 1994; 7: 559–572. 4 Flores J, Garcia-Avello A, Flores VM, Navarro JL, Canseco F, Perez-Rodriguez E. Tissue plasminogen activator plasma levels as a potential diagnostic aid in acute pulmonary embolism. Arch Pathol Lab Med 2003; 127: 310– 315.
BLUK077-Stein
March 28, 2007
57
15:45
CHAPTER 57
Echocardiogram in the diagnosis and prognosis of acute pulmonary embolism
Visualization of emboli within the right atrium, right ventricle, or pulmonary artery
Right ventricular enlargement or dysfunction: all patients or normotensive patients
Echocardiography has a low sensitivity for visualization of emboli within the right atrium, right ventricle, or pulmonary artery. Among any patient with pulmonary embolism (PE), pooled data showed emboli within the right ventricle or right atrium in 19 of 195 (10%) (Table 57.1) [1–3]. Thrombi in the right side of the heart or pulmonary artery were shown in 24 of 165 (15%) (Table 57.1, Figure 57.1) [1, 2]. Among patients with massive PE or unstable patients, the sensitivity of echocardiography for thrombi in the right atrium or ventricle ranged from 7 to 18% (Table 57.2) [4–6]. Patients with extremely mobile right heart thrombi that resembled a worm or snake had a high thrombusrelated mortality rate, 20 of 48 (42%) [7]. Among these, fatal PE occurred in 13 (27%), fatal paradoxical PE occurred in 1 (2%), and perioperative death from thrombectomy occurred in 6 (13%). Patients with more or less immobile right heart thrombi had a lower mortality. Among 57 patients, none had fatal PE and 2 (4%) died perioperatively at thrombectomy [7].
Echocardiographic evidence of right ventricular enlargement or dysfunction has been used for prognostic purposes and to guide therapy (see Chapter 89 on thrombolytic therapy). In-hospital mortality from PE among unselected patients and among normotensive patients with right ventricular dysfunction ranged from 5 to 13% compared with 0–1% among those with normal right ventricular function (Table 57.3) [8–10]. Pooled data showed an in-hospital mortality from PE in 23 of 222 (10%) with right ventricular dysfunction compared with 1 of 383 (0.3%) among those with normal right ventricular function [8–10]. Data based on right ventricular-to-left ventricular dimension ratios derived from CT angiography in normotensive, not critically ill patients, however, showed no increased in-hospital mortality in those with right ventricular enlargement (Stein PE, Beemath A, Matta F et al. Unpublished data from PIOPED II) (see Chapter 4). The cumulative 1-year mortality rate from acute PE was higher in patients with significant right ventricular hypokinesis [11].
Table 57.1 Intracardiac or pulmonary artery thrombi on echocardiography: any patient with pulmonary embolism. n/N (%) First author (Ref.)
RA or RV thrombi only sensitivit/
Cheriex [1]
10/60 (17)
10/60 (17)
Kasper [2]
4/105 (4)
14/105 (13)
Franzoni [3]
5/30 (17)
RA, RV, or PA thrombi sensitivity
5/30 (17)
RA, right atrium; RV, right ventricle; PA, pulmonary artery.
255
BLUK077-Stein
March 28, 2007
15:45
256
PART III
100
Diagnosis of acute PE
87
90 Sensitivity (%)
80 64
70 60 50 40 30
15
20 10
n = 393
n = 191
Echo RV Enlarged or dysfunction
Echo or US
n = 195
0
RA, RV, or PA thrombi
Figure 57.1 Sensitivity of echocardiographic findings for pulmonary embolism. Data based on information from Tables 57.1, 57.4, and 57.6. RA, right atrium; RV, right ventricle; PA, pulmonary artery; echo, echocardiogram; US, ultrasound.
Table 57.2 Thrombi within the right atrium or right ventricle on echocardiography: massive pulmonary embolism or unstable patient. First author (Ref.)
n/N (%): Sensitivity
Casazza [4]
23/130 (18)
Castillo-Fenoy [5]
4/34 (12)
Chapoutot [6]
12/170 (7)
Table 57.3 In-hospital mortality from pulmonary embolism according to the presence of right ventricular dysfunction. n/N (%) First Author (Ref.)
PE mortality: RV dysfunction
PE mortality: no RV dysfunction
Ribeiro [8]
9/70 (13)
Grifoni [9]
3/65 (5)
0/56 (0) 0/97 (0)
Kasper [10]
11/87 (13)
2/230 (1)
PE, pulmonary embolism; RV, right ventricle.
Table 57.4 All patients suspected with pulmonary embolism: echocardiographic right ventricular enlargement or dysfunction. n/N (%) First author (Ref.)
Sensitivity
Specificity
Positive predictive value
Negative predictive value
Grifoni [14]
32/63 (51)
Mansencal [15]
97/173 (56)
Perrier [16]
12/18 (67)
30/32 (94)
12/14 (86)
30/36 (83)
Miniati [17]
24/43 (56)
60/67 (90)
24/31 (77)
60/79 [76]
Cheriex [1]
46/46 (100)
Nazeyrollas [18]
29/31 (94)
32/39 (82)
29/36 (81)
32/34 (94)
Rudoni [19]
13/24 (54)
46/47 (98)
13/14 (93)
46/55 (84)
BLUK077-Stein
March 28, 2007
CHAPTER 57
15:45
257
Echocardiogram
Table 57.5 Echocardiograhic indices related to tricuspid regurgitation.
First author (Ref.)
TR velocity
TR velocity
TR velocity
TR velocity
TR color
> 2.5 m/s
≥ 2.7 m/s
≥ 3.0 m/s
moderate-severe
coded (1–4)
Grifoni [14] Mansencal [15] Perrier [16]
X
Miniati [17]
X
Cheriex [1]
X
Nazeyrollas [18]
X
X
Rudoni [19]
X
TR, tricuspid regurgitation.
Thirty-day all-cause mortality was 108 of 681 (16%) in PE patients with right ventricular dysfunction compared with 71 of 785 (9%) in PE patients with normal right ventricular function [12, 13]. Regarding the diagnosis of PE in all patients, the sensitivity of right ventricular enlargement or dysfunction ranged from 51 to 100% (Table 57.4) [1, 14–19]. On average, the sensitivity of right ventricular enlargement or dysfunction was 253 of 393 (64%), but it is recognized that the data are heterogeneous. The criteria used for right ventricular enlargement or dysfunction varied (Tables 57.5–57.7).
Echocardiogram plus venous ultrasound: all patients The sensitivity of echocardiographic findings of right ventricular enlargement or dysfunction combined with venous compression ultrasound in all patients
with suspected PE, on average, was 222 of 254 (87%) (Table 57.8, Figure 57.1) [15, 16]. If any two of clinical probability high, echocardiogram, and ultrasound were used, the sensitivity was 56 of 63 (89%) (Table 57.8, Figure 57.1) [14].
Right ventricular enlargement or dysfunction: massive PE or unstable patients The sensitivity of right ventricular enlargement or dysfunction for the diagnosis of PE in patients with massive PE or unstable patients, combining data from three small case series, was 33 of 33 (100%) (Table 57.9) [1, 15, 19]. Specificity and positive predictive value were not stated. Considering high-probability clinical assessment, echocardiogram, and ultrasound, if any 2 of 3 were positive, the sensitivity was 56 of 63 (89%) (Table 57.8) [14].
Table 57.6 Echocardiograhic indices related to RV dysfunction.
First author (Ref.)
RVEDD
RVEDD
>27 mm
>30 mm
Grifoni [14]
RVEDD high
RV/LV
RV/LV
RV/LV area ≥ 0.6 &
EDD > 0.5
EDD > 1.0
RV free wall < 7.0 mm
X
X
Mansencal [15]
X
Perrier [16] Miniati [17] Cheriex [1]
X X
X X
Nazeyrollas [18] Rudoni[19]
RV hypokinetic
X X
RV, right ventricle; LV, left ventricle; RVEDD, right ventricular end diastolic diameter.
BLUK077-Stein
March 28, 2007
15:45
Table 57.7 Echocardiograhic miscellaneous indices.
First author (Ref.)
IVS paradoxical
IVS systolic
Pulmonary acceleration
RV/RA gradient
IVC collapse
bulge
flattening
< 90 m/s
> 30 mm Hg
index < 40%
X
X
X
Grifoni [14] Mansencal [15]
X
Perrier [16] Miniati [17] Cheriex [1]
X
X
Nazeyrollas [18] Rudoni [19]
X
* Grifoni used any one criteria from Tables 57.5–57.7. Mansencal required both criteria from Tables 57.5 and 57.6. Perrier required both criteria from Tables 57.5 and 57.6. Miniati and Rudoni used any two criteria from Tables 57.5–57.7. IVS, intraventricular septum; IVC, inferior vena cava; RV, right ventricle; RA, right atrium. Table 57.8 All patients suspected with pulmonary embolism. n/N (%) Positive
Negative
First author (Ref.)
Basis for diagnosis
Sensitivity
Specificity
predictive value
predictive value
Grifoni [14]
Clin/echo/US*
56/63 (89)
40/54 (74)
56/70 (80)
40/47 (85)
Mansencal [15]
Echo or US
154/173 (89)
Perrier [16]
Echo or US
12/18 (67)
30/32 (94)
12/14 (86)
30/36 (83)
* Any 2 of 3. US, ultrasound.
Table 57.9 Massive pulmonary embolism or unstable patient: echocardiographic right ventricular enlargement or dysfunction. First author (Ref.)
Basis for Dx
n/N (%): sensitivity
Grifoni [14]
Clin/echo/US*
33/34 (97)
Mansencal [15]
Echo
17/17 (100)
Cheriex [1]
Echo
11/11 (100)
Rudoni [19]
Echo
5/5 (100)
* Any 2 of 3. US, ultrasound.
100 100
Sensitivity (%)
90 80 70 60 50 40 30 20 10 0
12
n = 334 RA or RV thrombi
n = 33 Echo RV dysfunction or enlargement
258
Figure 57.2 Sensitivity of echocardiogram in patients with massive pulmonary embolism (PE) or unstable from PE. Right atrial (RA) or right ventricular (RV) thrombi were shown in 12 patients and echocardiographic RV dysfunction or enlargement was shown in all patients. Data based on information from Tables 57.2 and 57.7.
BLUK077-Stein
March 28, 2007
CHAPTER 57
15:45
259
Echocardiogram
References 1 Cheriex EC, Sreeram N, Eussen YF, Pieters FA, Wellens HJ. Cross sectional Doppler echocardiography as the initial technique for the diagnosis of acute pulmonary embolism. Br Heart J 1994; 72: 52–57. 2 Kasper W, Meinertz T, Henkel B et al. Echocardiographic findings in patients with proved pulmonary embolism. Am Heart J 1986; 112: 1284–1290. 3 Franzoni P, Cuccia C, Zappa C, Volpini M, Gei P, Visioli O. Thromboembolus migrating into the right heart in pulmonary embolism. Echocardiographic and clinicotherapeutic aspects in 7 cases and review of the literature. G Ital Cardiol 1989; 19: 7–16. 4 Casazza F, Bongarzoni A, Centonze F, Morpurgo M. Prevalence and prognostic significance of right-sided cardiac mobile thrombi in acute massive pulmonary embolism. Am J Cardiol 1997; 79: 1433–1435. 5 Castillo-Fenoy A, Scheuble C, Benacerraf A, Brau J, Cereze P. Echographical symptomatology of acute pulmonary embolism. Arch Mal Coeur Vaiss 1980; 73: 971–978. 6 Chapoutot L, Metz D, Canivet E et al. Mobile thrombus of the right heart and pulmonary embolism: diagnostic and therapeutic problems. Apropos of 12 cases. Arch Mal Coeur Vaiss 1993; 86: 1039–1045. 7 European Working Group on Echocardiography. The European Cooperative Study on the clinical significance of right heart thrombi.. Eur Heart J 1989; 10: 1046– 1059. 8 Ribeiro A, Lindmarker P, Juhlin-Dannfelt A, Johnsson H, Jorfeldt L. Echocardiography Doppler in pulmonary embolism: right ventricular dysfunction as a predictor of mortality rate. Am Heart J 1997; 134: 479–487. 9 Grifoni S, Olivotto I, Cecchini P et al. Short-term clinical outcome of patients with acute pulmonary embolism, normal blood pressure, and echocardiographic right ventricular dysfunction. Circulation 2000; 101: 2817– 2822. 10 Kasper W, Konstantinides S, Geibel A, Tiede N, Krause T, Just H. Prognostic significance of right ventricular after-
11 12
13
14
15
16
17
18
19
load stress detected by echocardiography in patients with clinically suspected pulmonary embolism. Heart 1997; 77: 346–349. Ribeiro A. The role of echocardiography Doppler in pulmonary embolism. Echocardiography 1998; 15: 769–778. Kucher N, Rossi E, De Rosa M, Goldhaber SZ. Prognostic role of echocardiography among patients with acute pulmonary embolism and a sysytolic arterial pressure of 90 mm Hg or higher. Arch Intern Med 2005; 165: 1777–1781. Schoepf UJ, Kucher N, Kipfmueller F, Quiroz R, Costello P, Goldhaber SZ. Right ventricular enlargement on chest computed tomography: a predictor of early death in acute pulmonary embolism. Circulation 2004; 110(20): 3276– 3280. Grifoni S, Olivotto I, Cecchini P et al. Utility of an integrated clinical, echocardiographic, and venous ultrasonographic approach for triage of patients with suspected pulmonary embolism. Am J Cardiol 1998; 82: 1230– 1235. Mansencal N, Redheuil A, Joseph T et al. Use of transthoracic echocardiography combined with venous ultrasonography in patients with pulmonary embolism. Int J Cardiol 2004; 96: 59–63. Perrier A, Tamm C, Unger PF, Lerch R, Sztajzel J. Diagnostic accuracy of Doppler-echocardiography in unselected patients with suspected pulmonary embolism. Int J Cardiol 1998; 65: 101–109. Miniati M, Monti S, Pratali L et al. Value of transthoracic echocardiography in the diagnosis of pulmonary embolism: results of a prospective study in unselected patients. Am J Med 2001; 110: 528–535. Nazeyrollas P, Metz D, Jolly D et al. Use of transthoracic Doppler echocardiography combined with clinical and electrocardiographic data to predict acute pulmonary embolism. Eur Heart J 1996; 17: 779–786. Rudoni RR, Jackson RE, Godfrey GW, Bonfiglio AX, Hussey ME, Hauser AM. Use of two-dimensional echocardiography for the diagnosis of pulmonary embolus. J Emerg Med 1998; 16: 5–8.
BLUK077-Stein
March 11, 2007
58
10:5
CHAPTER 58
Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism
In 2004, CT (computed tomography) scanners, mostly single slice, were available in a larger proportion of nonfederal short-stay general and special hospitals in the United States than were radioisotope facilities (88% versus 61%) (Table 58.1) [1]. If one assumes that hospitals with <25 beds did not have CT, then 95% of the short-stay hospitals with ≥25 beds had CT. Multislice CT scanners were available in 39% of the short-stay hospitals [1]. Diagnostic ultrasound was available in 86% and magnetic resonance imaging was available in 59% of the short-stay hospitals. The distribution of sizes of short-stay nonfederal hospitals in the United States in 2004 was 7% with 6–24 beds, 20% with 25–49 beds, 21% with 50–99 beds, 23% with 100–199 beds, and 29% with ≥200 beds [1]. The proportion of hospitals with various imaging facilities differs in European countries. In the Netherlands in 1997, among 122 hospitals, facilities for ventilation–perfusion (V–Q) scans were available in about 57% of the hospitals, spiral CT in 61%, and pulmonary angiography in 89% [2]. In Austria in 1998,
Table 58.1 Proportion of short-stay nonfederal hospitals in the United States with computed tomography (CT), ventilation–perfusion (V–Q) lung scans, magnetic resonance imaging (MRI), and ultrasound. Percent of hospitals CT
88
Multislice CT
39
MRI
60
V–Q
61
Ultrasound
86
Data from American Hospital Association Hospital Statistics [1].
260
among 127 hospitals, ultrasound was available in 97%, V–Q scans in 19%, spiral CT in 54%, and pulmonary angiography in 59% [3]. In the United Kingdom, according to a survey published in 1999, 66% of 327 hospitals had facilities for V–Q scans, 44% had facilities for spiral CT, and 46% had facilities for pulmonary angiography [4]. In the United Kingdom, V–Q scans were frequently the only imaging test other than chest radiography [4]. It has been suggested that spiral CT replace V–Q scintigraphy as the primary modality for screening patients with suspected acute pulmonary embolism (PE) [5], particularly among hospitalized patients [6]. It has also been proposed that spiral CT be included in the workup of patients with nondiagnostic V–Q lung scans [6]. Among patients discharged from hospitals with a diagnosis of PE in 2003, 47% of the diagnostic tests for PE were with CT scans (Figure 58.1). In 2003, only 17% of the tests in patients discharged with PE were with V–Q scans, 24% were pulmonary angiograms, and venous ultrasound of the lower extremities was obtained in 13%. These data were assessed using the database of the National Hospital Discharge Survey from 1979 through 2001 [7] and updated through 2003 (unpublished). Whether venous ultrasound in patients discharged with PE served to identify deep venous thrombosis as a surrogate diagnosis for PE or whether there was an interest in diagnosing deep venous thrombosis in patients already diagnosed with PE is undetermined. We assumed that all imaging studies we reported were used for the diagnosis of PE in these patients [7]. This assumption is stronger for CT, V–Q scans, and pulmonary angiography. Tests in patients with suspected PE in whom the diagnosis was excluded were
BLUK077-Stein
March 11, 2007
CHAPTER 58
10:5
261
Trends in use of diagnostic imaging
Tests in patients with PE (%)
100
80 V−Q SCANS CT
60
40 ANGIOS 20 US 0 2003
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
Year Figure 58.1 Relative use of diagnostic imaging tests in patients hospitalized with pulmonary embolism (PE) from 1979 through 2003. V–Q, ventilation–perfusion, US,
ultrasound, ANGIOS, pulmonary angiograms. (Modified and reprinted from Stein et al. [7], with permission from Elsevier.)
not included in these trends because there would be less certainty about whether CT was used to exclude PE. It is possible that pulmonary angiography was coded more consistently than other diagnostic tests.
3 Schibany N, Fleischmann D, Thallinger C et al. Equipment availability and diagnostic strategies for suspected pulmonary embolism in Austria. Eur Radiol 2001; 11: 2287– 2294. 4 Burkill, GJC, Bell JRG, Padley SPG. Survey on the use of pulmonary scintigraphy, spiral CT and conventional pulmonary angiography for suspected pulmonary embolism in the British Isles. Clin Radiol 1999; 54: 807–810. 5 Goodman LR, Lipchik RJ. Diagnosis of acute pulmonary embolism: time for a new approach. Radiology 1996; 199: 25–27. 6 Goodman LR, Lipchik RJ, Kuzo RS. Acute pulmonary embolism: the role of computed tomographic imaging. J Thor Imaging 1997; 12: 83–86. 7 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316–1317.
References 1 AHA Hospital Statistics, 2006 Edition. Health Forum LLC (An Affiliate of American Hospital Association), Chicago, 2006. 2 Hagen PJ, van Strijen MJ, Kieft GJ, Prins MH, Postmus PE. Availability of diagnostic facilities in the Netherlands for patients with suspected pulmonary embolism. ANTELOPE Study Group, Advances in New Technologies Evaluating the Localisation of Pulmonary Embolism. Neth J Med 2000; 57: 142–149.
BLUK077-Stein
March 11, 2007
59
10:14
CHAPTER 59
Techniques of perfusion and ventilation imaging
Introduction With the advent and ready availability of contrastenhanced spiral CT, the use of ventilation–perfusion (V–Q) lung scans for the diagnosis of acute pulmonary embolism (PE) has decreased. By 2001, the use of CT surpassed the use of V–Q scans [1], and unpublished updated information show that divergent trends have continued through 2003 (Chapter 58). Even so, V–Q scans are recommended for the diagnosis of acute PE in patients with severe allergy to iodine containing contrast material, patients with renal failure, women of reproductive age in whom radiation is a particular concern, and sometimes in patients in whom the CT angiogram gives an uncertain reading [2].
Perfusion imaging In perfusion lung imaging, the amount of microembolization of radiolabeled particles is proportional to the pulmonary arterial blood flow to that region [3]. Macroaggregated human albumin labeled with 99m Tc is used. The aggregated albumin preparation typically contains 90% of particles within the 10–90 μm range. It is estimated that there are over 600 million pulmonary arterioles small enough to trap the administered particles [4]. Because approximately 200,000– 500,000 particles are routinely injected for an individual study, the effect is physiologically insignificant. The arterioles are only temporarily occluded because the material degrades into smaller particles, the biologic half-time being about 6–8 hours [5]. The fragments are then phagocytized by cells of the reticuloendothelial system. The minimum toxic dose for albumin is 20 mg/kg [6]. The usual imaging dose is about 0.14 μg/kg albumin. At least 60,000 particles should be injected for an adequate study. Otherwise, an uneven distribution of
262
radioactivity in the vascular bed is likely [7]. Because the usual 2–4 mCi dose from most kits is distributed among approximately 500,000 particles, low-particle injections are rarely a problem. Perfusion imaging is begun by an intravenous injection of macroaggregates with the patient in the supine position to permit an even distribution of flow from the base of the lung to the apex, although there is a ventral to posterior gradient [8]. After injection, if a single-head camera is used, imaging is best performed with the patient erect, which allows maximum expansion of the lungs, especially at the lung bases. With a dual-headed camera, the patient is usually imaged in the supine position, which allows simultaneous anterior and posterior views and other 180◦ combination simultaneous views to be made. Anterior, posterior, both laterals, and both posterior and anterior obliques have been advocated [9–11]. Segmental delineation of the basal segments of both lower lobes is best seen in posterior oblique projections [11]. Posterior oblique views separate the two lungs, thereby avoiding “shine through” artifacts sometimes present in the lateral images. Oblique images contributed to the definition and clarification of the abnormalities, and resulted in improved lesion localization in 73% of patients [12].
Ventilation imaging Current methods for performing a ventilation study utilize a xenon gas, either 133 Xe or 127 Xe, radioactive krypton gas, 81m Kr, or a radioactive aerosol, usually the radioaerosol of 99m Tc diethylenetriamine pentaacetic acid (DTPA) or 99m Tc pyrophosphate (PYP). In addition, an ultra-fine dry dispersion of carbon “soot” that can be labeled with 99m Tc has been used, particularly in Europe and Australia [13]. Some authorities believe the ventilation study with 133 Xe can be effectively performed following the
BLUK077-Stein
March 11, 2007
CHAPTER 59
10:14
Techniques of perfusion and ventilation imaging
perfusion portion of the ventilation–perfusion study [14]. Carrying out the ventilation scan study after the perfusion scan would be convenient, because the ventilation scan need be performed only on those patients whose perfusion scans are indeterminate [15]. The experience of others, however, indicates that scattered irradiation from the previously administered perfusion tracer (99m Tc) significantly decreases the accuracy of the washout phase of the ventilation scan [8]. 127 Xe has photon energies higher than that of 99m Tc. As a consequence, 127 Xe ventilation studies can be performed after the perfusion scan. The uncertain availability of this tracer and its considerable cost have limited its general application [8].
Xenon ventilation studies The preperfusion 133 Xe ventilation lung scan has three phases: the single-breath image, the equilibrium phase, and the washout phase. If a single-head camera is used, the single-breath image is obtained with the patient initially positioned with his back to a wide field of view gamma camera (ideally erect, but often supine). A bolus of approximately 15–20 mCi of 133 Xe is injected into the mouthpiece of the spirometer system at a time when the patient begins a maximal inspiration. The patient holds his breath for the next 15–20 seconds and a single-breath image is obtained. Following this, equilibrium phase imaging is obtained with the patient breathing in a closed spirometer for 4–5 minutes. During this interval, a posterior and both right and left posterior oblique images are obtained. If there are any areas of the lungs that xenon cannot reach, this will be demonstrated on these views. If a dual-headed camera is used, simultaneous anterior and posterior images are obtained, and the oblique views are not needed. It is important to employ a tight-sealing facemask or mouthpiece at this point to ensure that complete equilibrium is reached [8]. The washout phase is begun at the end of this time by readjusting the intake valves of the spirometer system to permit the patient to inhale ambient air and exhale into a shielded charcoal trap that absorbs the expired radioactive gas. In general, the washout phase should last at least 5 minutes. With a single-head camera, images are taken at 30–60 second intervals during the course of the washout. Posterior oblique views approximately halfway through the washout process enhance the ability to detect and lo-
263
cate xenon retention in the anterior–posterior plane. These are not necessary if a dual-headed camera is employed. The equilibrium image has an activity distribution that corresponds to the aerated lung. Images made during washout should show a rapid clearance of activity from the lungs, usually within 90 seconds. Often, there is slight retention in the upper zones compared to the lower zones. In contrast to the single-breath image, which displays an abnormality as a deficit in radioactivity, abnormal washout images show 133 Xe “hot spots” that characterize the uneven distribution of ventilation. In this situation, abnormally retained 133 Xe activity (“air trapping”) is superimposed on a background of decreasing activity from normally ventilated lung regions. A sufficiently long rebreathing equilibrium time (at least 4 minutes) is imperative to maximize the usefulness of the washout phase by permitting the radiotracer to enter abnormal lung zones by collateral airdrift [8]. The washout phase of the 133 Xe study is the most likely phase to show ventilation abnormalities [16]. The major disadvantage of the xenon technique is that images of the lung are required in a preselected view, typically in the posterior projection, with some ancillary anteroposterior information added by oblique views. The simultaneous anterior and posterior views obtained with a dualheaded camera largely avoid this problem. Additional information at times can be added by obtaining a “second-look” single-breath xenon study in a different view. The typical analogue display of a Xe ventilation study for pulmonary embolism relies on a compromise intensity setting. The intensity is usually determined by experience, such that the early phases of the washout examination may be slightly intense relative to the later phases. The intensity of the later phases, at the time when the “unevenness” of abnormal xenon washout is best detected, will have an intensity set to optimize the abnormal appearance. In practice, however, it is sometimes difficult to adjust the washout intensity correctly, and other variables such as chest wall thickness, lung thickness (i.e., volume), and the depth and frequency of respiration may all affect the washout phase of the ventilation study. To avoid these problems, serial computerized washout images (e.g., 30–45 second intervals) may be modified by normalizing each image in the series. In this way, regions with
BLUK077-Stein
March 11, 2007
10:14
264
delayed clearance become more prominent, particularly in the later phases of the washout. There is no absolute normal range of xenon clearance times because of the normal variability in the rate and depth of respiration between subjects. The normal lungs are homogeneous in horizontal appearance, although there may be changes from apex to base. Inhomogeneity of xenon lung clearance at any time during the washout phase indicates an abnormality.
Krypton ventilation studies 81m
Kr is a relatively insoluble inert gas with a 13second half-life. It decays to 81 Kr by isomeric transition, emitting 190 keV gamma rays (65%) and internal conversion electrons. The gas is obtained by eluting a generator containing its parent 81 Rb, which has a 4.7-hour half-life. The short half-life of the 81 Rb generator is a disadvantage, since the generator contains only enough activity to be used for clinical studies on the day it is delivered. This markedly increases cost, because a new generator is required for each day’s work. Inhalation of 81m Kr results in images that demonstrate the distribution of regional tidal ventilation. Because of the ultrashort half-life of 13 seconds, by the time collateral air drift occurs, the 81m Kr has decayed away. There is, therefore, no opportunity to obtain washout images. The inability to obtain washout images is balanced by the fact that ventilation images can be obtained in multiple projections. This facilitates direct comparison with perfusion scans for the evaluation of ventilation–perfusion mismatches. 81m Kr images provide a sensitive means for studying regional ventilation [17, 18]. If the minute ventilation becomes greater than normal, 81m Kr regional count rates become more dependent on lung volume than on ventilation. This rarely, if ever, occurs in adults. Thus, 81m Kr images tend to underestimate ventilation in regions with high airflow because of potential partial dependence on the regional lung volume. Furthermore, because 81m Kr has a higher photon energy than 99m Tc, a low energy collimator cannot be used without getting some septal penetration. This decreases the resolution of the ventilation scan. If a medium energy collimator is used to avoid septal penetration, the collimator resolution is inherently less than a low energy collimator, and resolution of both the perfusion and ventilation scans decreases. This is also true for 127 Xe.
PART III
Diagnosis of acute PE
Radioaerosol inhalation studies Radioaerosols also provide a means for investigating regional ventilation and are commercially available with the introduction of small, easily contained, and efficient aerosol nebulizers. Radioaerosols are small particles, rather than gases. The radioaerosols deposit in the lung by impaction in central airways, sedimentation in more distal airways, and random contact with alveolar walls during diffusion in the alveoli [19]. The commercially available nebulizers produce submicron size aerosols that penetrate relatively evenly to the lung periphery, whereas larger (3–5 μm) radioaerosols, which were used in the past, demonstrated substantial degrees of central airway deposition, resulting in poor images of peripheral ventilation patterns. Currently, the radioaerosol most often employed is 99m Tc-diethylenetriamine pentaacetate (DTPA). DTPA is a low-molecular-weight solute that is able to cross the respiratory epithelium and be removed from the lungs rapidly through the bloodstream. Aerosols are typically used as a preperfusion ventilation scan with counts about one-fifth the counts of the perfusion scan so they do not “bleed through” on the perfusions images. Typically, if an eight-view perfusion scan is to be obtained, the aerosol ventilation images will be obtained in the same projections allowing precise comparison of ventilation and perfusion. However, good results have been shown with obtaining perfusion lung scans before ventilation scans when the ventilation scans were obtained with DPTA [20]. Performing the perfusion scan prior to the ventilation scan permitted the ventilation study to be tailored for optimal positioning to determine if mismatched defects were present. In addition, ventilation scans were unnecessary in patients in whom the perfusion scans were normal. Aerosolized 99m Tc-pyrophosphate has also been used and has the advantage of a longer lung residence time, which is especially important in smokers [21]. To yield uniform apex-to-base deposition, the inhalation of radioaerosol can be performed with the patient in the supine position. However, with a singlehead camera, if postperfusion aerosol images are obtained, and the perfusion defects are basilar, the erect position is often used. A nebulizer that generates submicron size particles should be filled with about 30 mCi of 99m Tc DTPA in 2–3 mL of saline, and the nebulizer
BLUK077-Stein
March 11, 2007
CHAPTER 59
10:14
Techniques of perfusion and ventilation imaging
should be aerated at a flow rate of 8–10 L/minute [21]. After approximately 3 minutes of breathing, about 750 mCi of 99m Tc aerosol will have been deposited in the lungs, yielding 100 K count images in approximately 2 minutes with a standard wide-field gamma camera with an all-purpose low-energy collimator. If postperfusion aerosol scanning is done, the aerosol dose in the nebulizer is increased to as much as 45 mCi of 99m Tc-pyrophosphate or DTPA and the inhalation of aerosol continues until the count rate obtained for the perfusion scan is doubled [21]. There is no readily apparent difference between the various agents used for ventilation scans from a diagnostic standpoint [22]. Cost, patient logistics, and referral patterns determine which ventilation agent is best suited for a specific institution [8]. Regarding the distribution of ventilation, in upright normal subjects, the lung bases ventilate better than the apices. In the lateral decubitus position, the lower lung ventilates better per unit volume than the upper lung. When the subject is supine, there is little apex-tobase gradient, but there is still an anterior–posterior gradient, with the most dependent regions receiving more ventilation than the uppermost regions.
Single photon emission computed tomographic perfusion lung scan Single photon emission computed tomography (SPECT) is discussed in Chapter 70.
References 1 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316–1317. 2 Stein PD, Woodard PK, Weg JG et al. Diagnostic pathways in acute pulmonary embolism: recommendations of the PIOPED II Investigators. Am J Med 2006; 119: 1048– 1055. 3 Stein PD, Gottschalk A. Critical review of ventilation/ perfusion lung scans in acute pulmonary embolism. Prog Cardiovasc Dis 1994; 37: 13–24. 4 Miller WS. The structure of the lungs. J Morphol 1893; 8: 165. Quoted by Dalen JE, Haynes FW, Hoppin FG, Jr, Evans GL, Bhardwaj P, Dexter L. Cardiovascular responses to experimental pulmonary embolism. Am J Cardiol 1967; 20: 3–9.
265
5 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 6 Taplin GV, MacDonald NS. Radiochemistry of macroaggregated albumin and newer lung scanning agents. Semin Nucl Med 1971; 1: 132–152. 7 Heck LL, Duley JW. Statistical considerations in lung imaging with 99 m-Tc albumin particles. Radiology 1974; 113: 675–679. 8 Gottschalk A, Alderson PO, Sostman HD. Nuclear medicine techniques and applications. In: Murray JF & Nadel JA, eds. Textbook of Respiratory Medicine, 2nd edn. WB Saunders, Philadelphia, 1994: 682–710. 9 Mack JF, Wellman HN, Saenger EL, Friedman BI. Clinical experience with oblique views in pulmonary perfusion camera-imaging in normal and pathological anatomy. Radiology 1969; 92: 897–902. 10 Mack JF, Wellman HN, Saenger EL. Oblique pulmonary scintiphotography in the analysis of perfusion abnormalities due to embolism. J Nucl Med 1969; 10: 420. 11 Caride VJ, Puri S, Slavin JD, Lange RC, Gottschalk A. The usefulness of posterior oblique views in perfusion lung imaging. Radiology 1976; 121: 669–671. 12 Nielsen PE, Kirchner PT, Gerber FH. Oblique views in lung perfusion scanning: clinical utility and limitations. J Nucl Med 1977; 18: 967–972. 13 Fawdry RM, Gruenewald SM. Initial experience with technegas: a new ventilation agent. Australas Radiol 1988; 32: 232–238. 14 Kipper MS. Alazraki N. The feasibility of performing 133 Xe ventilation imaging following the perfusion study. Radiology 1982; 144: 581–586. 15 Stein PD, Terrin ML, Gottschalk A, Alavi A, Henry JW. Value of ventilation/perfusion scans versus perfusion scans alone in acute pulmonary embolism. Am J Cardiol 1992; 69: 1239–1241. 16 Alderson PO, Biello DR, Khan AR, Barth KH, McKnight RC, Siegel BA. Comparison of 133 Xe single-breath and washout imaging in the scintigraphic diagnosis of pulmonary embolism. Radiology 1980; 137: 481– 486. 17 Alderson PO, Line BR. Scintigraphic evaluation of regional pulmonary ventilation. Semin Nucl Med 1980; 10: 218–242. 18 Miller TR, Biello DR, Lee JI et al. Ventilation imaging with Kr-81 m: A comparison with Xe-133. Eur J Nucl Med 1981; 6: 11–16. 19 Stuart BO. Deposition of inhaled aerosols. Arch Intern Med 1973; 131: 60–73. 20 Freitas JE, Sarosi MG, Nagle CC, Yeomans ME, Freitas AE, Juni JE. Modified PIOPED criteria used in clinical practice. J Nucl Med 1995; 36: 1573–1578.
BLUK077-Stein
March 11, 2007
10:14
266
21 Krasnow AZ, Isitman AT, Collier BD et al. Diagnostic applications of radioaerosols in nuclear medicine. In: Freeman LM, ed. Nuclear Medicine Annual 1993. Raven Press, New York, 1993: 123–193.
PART III
Diagnosis of acute PE
22 Alderson PO, Biello DR, Gottschalk A et al. Tc-99 mDTPA aerosol and radioactive gases compared as adjuncts to perfusion scintigraphy in patients with suspected pulmonary embolism. Radiology 1984; 153: 515–521.
BLUK077-Stein
April 14, 2007
60
15:0
CHAPTER 60
Ventilation–perfusion lung scan criteria for interpretation prior to the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED)
Introduction Scintigraphic studies of normal lungs demonstrate homogeneous patterns of matched ventilation and perfusion [1]. Nearly all pulmonary diseases, including neoplasms, infections, and chronic obstructive pulmonary disease, can produce decreased pulmonary blood flow to affected lung zones [2]. Pulmonary consolidation, bronchial obstructive disease, and restrictive disease as well as vascular occlusion disease may cause abnormalities on the perfusion lung scan as well as the ventilation lung scan. Vascular occlusion may result from neoplastic vascular compression and pulmonary vasculitis as well as pulmonary embolism (PE). Other processes occurring in the pulmonary arterial lumen can cause a ventilation–perfusion mismatch, including embolism of material other than thrombus (fat embolism and foreign material injected by intravenous drug abusers), in situ thrombosis (sickle cell disease), and pulmonary artery tumors [3–7]. Ventilatory defects may accompany the perfusion defect due to bronchial constriction caused from substances released from emboli or by reductions in alveolar levels of carbon dioxide, a potent airway dilator. This response, however, is usually transient and uncommon in scintigraphic studies [8–10]. Pulmonary infection, hemorrhage, infarction, atelectasis, or other conditions associated with alveolar collapse and/or leakage of fluids into the alveolar space cause perfusion and ventilation abnormalities. The degree of impaired ventilation and perfusion can vary, but a matched ventilation and perfusion defect typically is present [11].
Emphysema due to alpha1-antitrypsin deficiency and presumably due to other etiologies results in destructive loss of alveoli and the capillary bed, thereby affecting perfusion as well as ventilation [12]. The pulmonary arterioles also constrict with hypoxia [12]. Destruction of alveoli and the capillary bed as well as arteriolar constriction tend to cause matched ventilation and perfusion abnormalities [1]. Restrictive disease, particularly if caused by chronic fibrosis, usually results in a greater impairment of perfusion than ventilation [1]. A generalized abnormality on the chest radiograph such as diffuse pulmonary edema or diffuse reticulonodular disease may not cause the perfusion lung scan to be abnormal. Among 55 patients with diffuse radiographic opacities, 73% had normal or near-normal perfusion images [13]. A ventilation–perfusion scan in such patients, therefore, may assist in excluding PE, often guiding management decisions in these patients.
Diagnostic criteria for PE The scintigraphic hallmarks of PE are large, wedgeshaped, pleural-based perfusion defects in areas that ventilate normally and are radiographically clear [14]. Wagner and associates [15] and DeNardo and associates [16] suggested combined ventilation–perfusion lung imaging as a way to improve the specificity of diagnosing pulmonary emboli by scintigraphy. McNeil and coworkers [17] highlighted the findings of numerous investigators. They pointed out that in regions where the chest radiograph is normal, abnormalities in the
267
BLUK077-Stein
April 14, 2007
15:0
268
PART III
Diagnosis of acute PE
Table 60.1 Criteria for high probability in the interpretation of ventilation–perfusion scans for pulmonary embolism. Sullivan and Gottschalk [18] Biello [19]
Biello et al. [20]
Based on McNeil [21]
Hull et al. [22]
≥2 large V–Q mismatches
≥1 large V–Q mismatches
Multiple V–Q mismatches
≥1 segmental V–Q
with Nl X-ray
with Nl X-ray
mismatch
largest being greater than or equal to segmental
or
or
≥2 moderate V–Q
≥2 moderate V–Q
mismatches with Nl X-ray
mismatches with Nl X-ray
or
or
≥1 Q > X-ray
≥1 Q > X-ray
Mod: 25–90% Seg
Mod: 25–75% Seg
Large: >90% Seg
Large: >75% Seg
V–Q mismatch = Nl V in
V–Q mismatch = Nl V in
region of Q defect
region of Q defect
V–Q match = abnormal V in
V–Q mismatch = Nl or mildly reduced V with Nl X-ray in
V–Q match = abnormal V in
region of Q defect
region of Q defect
region of Q defect
Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality; seg, segment. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.
Table 60.2 Criteria for intermediate probability in the interpretation of ventilation–perfusion scans for pulmonary embolism. Sullivan and Gottschalk [18] Biello [19]
Biello et al. [20]
Based on McNeil [21]
Hull et al. [22]
≥1 Q = X-ray
≥1 Q = X-ray
1 segmental V–Q mismatch
≥1 subsegmental V–Q
or
or
or
or
1 large V–Q mismatch
1 moderate V–Q mismatch
≥1 Q = X-ray*
≥1 segmental V–Q
mismatch
with Nl X-ray
match
or
or
or
Severe COPD with Q
Severe COPD with Q defects
Mixed V–Q mismatched and
defects
matched defects or 1 V–Q mismatch ≥lobar size or Multiple subsegmental V–Q mismatches
* Our interpretation of data. Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.
BLUK077-Stein
April 14, 2007
CHAPTER 60
15:0
269
V–Q scan diagnostic criteria before PIOPED
Table 60.3 Criteria for low probability in the interpretation of ventilation–perfusion scans for pulmonary embolism. Sullivan and Gottschalk [18] Biello [19]
Biello et al. [20]
Based on McNeil [21]
Hull et al. [22]
Small V–Q mismatches
Small V–Q mismatches
1 subsegmental V–Q
≥1 subsegmental V–Q
mismatch or
or
or
V–Q matches with Nl
V–Q matches with Nl
≥1 V–Q match
X-ray
match
X-ray
or
or
Q X-ray
Q X-ray
Small: <25% seg
Small: <25% seg
*Our interpretation of data. Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality; seg, segment. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.
perfusion scan that are matched by zones of abnormal ventilation are less likely to represent pulmonary emboli than the mismatched abnormalities (i.e., reduced perfusion with normal ventilation). Many studies have led to a gradual evolution of diagnostic criteria for scintigraphy. This was reviewed by Stein and Gottschalk [18]. Tables 60.1–60.3 list diagnostic criteria for the interpretation of ventilation–perfusion lung scans that were employed by various investigators or seem to be suggested by their observations [19–22]. The positive predictive values for PE with high, intermediate, low, and normal interpretations of ventilation–perfusion scans according to various proposed criteria were comparable (Table 60.4). When the chest radiograph is called “normal” this refers to the radiographic appearance in the region of the ventilation or perfusion defect and need not mean
Table 60.4 Probability of pulmonary embolism obtained with different criteria for the interpretation of ventilation–perfusion scans for pulmonary embolism. V–Q
Biello
Hull
probability
Biello [19]
et al. [20]
et al. [22]
High
90%
87–92%
86%
Intermediate
30%
20–33%
Low
10%
0–8%
Normal
0%
0%
21–40% 13% 0%
V–Q, ventilation–perfusion. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.
the entire radiograph. Similarly, a chest radiographic defect indicates a radiographic opacity in the region related to the ventilation or perfusion defect. A lung zone is one-third of a lung divided craniocaudally (i.e., upper, middle, and lower zones). Definitions of a mismatched defect, matched defect, large segmental defect, and moderate size segmental defect varied among investigators (Table 60.1). McNeil showed likelihood ratios for several ventilation–perfusion scan patterns [21]. Based upon her data, others developed criteria for the interpretation of ventilation–perfusion scans [23, 24]. These criteria differ somewhat from each other. Neither of these developed criteria showed areas under receiver operating characteristic curves that differed significantly from criteria suggested by Biello and associates [19, 20, 23, 24]. A common cause of ventilation–perfusion mismatched defects not due to acute PE is persistent mismatched defects from prior PE. Rates and extent of resolution of perfusion lung scans are described in Chapter 6.
References 1 Gottschalk A, Alderson PO, Sostman HD. Nuclear medicine techniques and applications. In: Murray JF and Nadel JA, eds. Textbook of Respiratory Medicine, 2nd edn., WB Saunders Co, Philadelphia, 1994: Chap. 25. 2 Secker-Walker RH, Siegel BA. The use of nuclear medicine in the diagnosis of lung disease. Radiol Clin North Am 1973; 11: 215–241.
BLUK077-Stein
April 14, 2007
15:0
270
3 McNeil BJ. A diagnostic strategy using ventilation– perfusion studies in patients suspect for pulmonary embolism. J Nucl Med 1976; 17: 613–616. 4 Smith RR, Hutchins GM. Pulmonary fecal embolization complicating the Budd–Chiari syndrome. N Engl J Med 1978; 298: 1069–1070. 5 Sostman HD, Brown M, Toole A, Bobrow S, Gottschalk A. Perfusion scan in pulmonary vascular/lymphangitic carcinomatosis: the segmental contour pattern. Am J Radiol 1981; 137: 1072–1074. 6 Myerson PJ, Myerson DA, Katz R, Lawson JP. Gallium imaging in pulmonary artery sarcoma mimicking pulmonary embolism. Case report. J Nucl Med 1976; 17: 893–895. 7 Velchik MG, Tobin M, McCarthy K. Nonthromboembolic causes of high-probability lung scans. Am J Physiol Imaging 1989; 4: 32–38. 8 Thomas D, Stein M, Tanabe G, Rege V, Wessler S. Mechanism of bronchoconstriction produced by thromboemboli in dogs. Am J Physiol 1964; 206: 1207–1212. 9 Kessler RM, McNeil BJ. Impaired ventilation in a patient with angiographically demonstrated pulmonary emboli. Radiology 1975; 114: 111–112. 10 Epstein J, Taylor A, Alazraki N, Coel M. Acute pulmonary embolus associated with transient ventilatory defect case report. J Nucl Med 1975; 16: 1017–1020. 11 Finley TN, Tooley WH, Swenson EW, Gardner RE, Clements JA. Pulmonary surface tension in experimental atelectasis. Am Rev Respir Dis 1964; 89: 372–378. 12 Stein PD, Leu JD, Welch MH, Guenter CA. Pathophysiology of the pulmonary circulation in emphysema associated with alpha1 antitrypsin deficiency. Circulation 1971; 43: 227–239. 13 Newman GE, Sullivan DC, Gottschalk A, Putman CE. Scintigraphic perfusion patterns in patients with diffuse lung disease. Radiology 1982; 143: 227–231.
PART III
Diagnosis of acute PE
14 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 15 Wagner HN, Jr, Lopez-Majano V, Langan JK, Joshi RC. Radioactive xenon in the differential diagnosis of pulmonary embolism. Radiology 1968; 91: 1168–1174. 16 DeNardo GL, Goodwin DA, Ravasini R, Dietrich PA. The ventilatory lung scan in the diagnosis of pulmonary embolism. N Engl J Med 1970; 282: 1334–1336. 17 McNeil BJ, Holman BL, Adelstein SJ. The scintigraphic definition of pulmonary embolism. JAMA 1974; 227: 753–756. 18 Stein PD, Gottschalk A. Critical review of ventilation– perfusion lung scans in acute pulmonary embolism. Prog Cardiovasc Dis 1994; 37: 13–24. 19 Biello DR. Radiological (scintigraphic) evaluation of patients with suspected pulmonary thromboembolism. JAMA 1987; 257: 3257–3259. 20 Biello DR, Mattar AG, McKnight RC, Siegel BA. Ventilation–perfusion studies in suspected pulmonary embolism. Am J Radiol 1979; 133: 1033–1037. 21 McNeil BJ. Ventilation–perfusion studies and the diagnosis of pulmonary embolism: concise communication. J Nucl Med 1980; 21: 319–323. 22 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. 23 Webber MM, Gomes AS, Roe D, LaFontaine RL, Hawkins RA. Comparison of Biello, McNeil, and PIOPED criteria for the diagnosis of pulmonary emboli on lung scans. Am J Roentgenol 1990; 154: 975–981. 24 Sullivan DC, Coleman RE, Mills SR, Ravin CE, Hedlund LW. Lung scan interpretation: effect of different observers and different criteria. Radiology 1983; 149: 803– 807.
BLUK077-Stein
March 11, 2007
61
10:38
CHAPTER 61
Observations from PIOPED: ventilation–perfusion lung scans alone and in combination with clinical assessment
Introduction The Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) was a national collaborative study designed to determine the sensitivity and specificity of ventilation–perfusion lung scans in patients with suspected acute pulmonary embolism (PE) [1]. All patients in whom attending physicians suspected acute PE were asked to consent to undergo obligatory angiography if their ventilation–perfusion lung scan was abnormal. Among those who consented for participation in PIOPED, a random sample was selected for investigation. PE was diagnosed or excluded by pulmonary angiography or at autopsy. Patients with a normal ventilation–perfusion scan were not required to undergo a pulmonary angiogram. PE was excluded in such patients by the absence of PE during 1 year of follow-up while not receiving anticoagulant therapy. In addition to the randomized patients described in the primary PIOPED report [1], an arm of PIOPED included patients who were referred for pulmonary angiography by attending physicians. These nonrandomized patients were not included in the analysis of sensitivity and specificity reported in PIOPED [1], and these patients are not included in the data described in this chapter unless specifically stated. Useful information, however, was obtained from data in these patients. A number of secondary investigations based on PIOPED data included both arms of PIOPED. Ventilation studies were performed, preferably in the upright position with 5.6 × 108 − 11.1 × 108 Bq of xenon-133. They started with a 100,000-count, posterior view, first-breath image, and then posterior equi-
librium (wash-in) images for two consecutive 120second periods. Then, perfusion scans were obtained with 1.5 × 108 Bq of 99m Tc macroaggregated albumin that contained 100,000–700,000 particles. Particles were injected into an antecubital vein. The perfusion images consisted of anterior, posterior, both posterior oblique, and both anterior oblique views, with 750,000 counts per image for each. For the lateral view with the best perfusion, 500,000 counts per image were collected; the other lateral view was obtained for the same length of time. Two nuclear medicine readers, not from the center that performed the scan and blinded to all clinical and laboratory data, independently interpreted the lung scans with chest roentgenograms. The criteria for the interpretation of ventilation– perfusion scans used in PIOPED were developed on the basis of 2 or more mismatched segmental equivalent defects being indicative of a high probability of PE (Table 61.1). A mismatched segmental equivalent defect is 1 large mismatched segmental defect or 2 moderate size segmental defects. A large segmental defect is >75% of a segment. A moderate size segmental defect is ≥25% of a segment, but ≤75% of a segment (Figures 61.1 and 61.2). These criteria assume that a mismatched moderate size segmental equivalent defect is of less diagnostic value than a mismatched large segmental defect [2]. One of the difficulties with this system of grading is the interpretation of the size of the mismatched defect. The distinguishability of large versus moderate size segmental defects requires skill and judgment. Experienced readers of radionuclide lung scans often underestimate the size of segmental defects [3].
271
BLUK077-Stein
March 11, 2007
10:38
272
PART III
Diagnosis of acute PE
Table 61.1 PIOPED central scan interpretation categories and criteria. High probability ≥2 Large (>75% of a segment) segmental perfusion defects without corresponding ventilation or roentgenographic abnormalities or substantially larger than either matching ventilation or chest roentgenogram abnormalities. ≥2 Moderate segmental (25 and 75% of a segment) perfusion defects without matching ventilation or chest roentgenogram abnormalities and 1 large mismatched segmental defect. ≥4 Moderate segmental perfusion defects without ventilation or chest roentgenogram abnormalities. Intermediate probability (indeterminate) Not falling into normal, very low, low-, or high-probability categories. Borderline high or borderline low. Difficult to categorize as low or high. Low probability Nonsegmental perfusion defects (e.g., very small effusion causing blunting of the costophrenic angle, cardiomegaly, enlarged aorta, hila, and mediastinum, and elevated diaphragm). Single moderate mismatched segmental perfusion defect with normal chest roentgenogram. Any perfusion defect with a substantially larger chest roentgenogram abnormality. Large or moderate segmental perfusion defects involving no more than four segments in one lung and no more than three segments in one lung region with matching ventilation defects either equal to or larger in size, and chest roentgenogram either normal or with abnormalities substantially smaller than perfusion defects. >3 Small segmental perfusion defects (<25% of a segment) with a normal chest roentgenogram. Very low probability ≤ 3 Small segmental perfusion defects with a normal chest roentgenogram. Normal No perfusion defects present. Perfusion outlines exactly the shape of the lungs as seen on the chest roentgenogram (hilar and aortic impressions may be seen, chest roentgenogram and/or ventilation study may be abnormal). PIOPED, Prospective Investigation of Pulmonary Embolism Diagnosis. From PIOPED [1], with permission.
The PIOPED criteria for interpretation of ventilation–perfusion lung scans defined a “very low probability” ventilation–perfusion scan, but the results of PIOPED reported “nearly normal/normal” interpretations. The latter interpretation includes nearly normal ventilation–perfusion lung scans and entirely normal ventilation–perfusion lung scans. A nearly normal ventilation–perfusion lung scan category included readings of very low probability by one central scan reader and low probability by the other, very low probability by both, very low probability by one, and normal by the other. An entirely normal ventilation–perfusion lung scan was one that was agreed upon by both central scan readers as being normal (Figure 61.3). Agreement among scan readers was excellent in PIOPED for high-probability (95%), very low probability (92%), and normal (94%) scan categories [1]. For intermediate-probability (indeterminate) and
low-probability scan categories, the readers agreed less frequently (75 and 70%, respectively). The sensitivity of a high-probability ventilation– perfusion lung scan interpretation among 251 patients in PIOPED with angiographically diagnosed PE was 41% [1]. Sensitivity was defined as the percentage of patients with PE who had a high-probability ventilation–perfusion lung scan. The specificity of a high-probability ventilation–perfusion lung scan among 480 patients with negative pulmonary angiograms was 97% [1]. Specificity was defined as the percentage of patients with pulmonary angiograms free of signs of PE who had non-high-probability interpretations of the ventilation–perfusion lung scans. The frequency of PE among patients in whom the diagnosis was made or excluded entirely by pulmonary angiography differed somewhat from the frequency in patients in whom outcome analysis as well as pulmonary angiography were employed
BLUK077-Stein
March 11, 2007
10:38
(a)
Figure 61.1 (a) (Left) Chest radiograph, right lower zone, showing a mild parenchymal abnormality; (right) perfusion scan, anterior view, showing a perfusion defect (arrow) that is much larger than the radiographic abnormality. (b) (Left) Perfusion scan, posterior view, showing the same large perfusion defect seen in the anterior view (arrow); (right) ventilation scan, posterior view, equilibrium phase, showing nearly normal ventilation. The washout phase was normal. The ventilation–perfusion lung scan was interpreted as high probability for pulmonary embolism. A pulmonary angiogram showed a large embolism in the right lower zone. The perfusion scan was obtained with 99m Tc macroaggregated albumin. The ventilation scan was obtained with xenon-133. (Courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, MI.)
(b)
Perfusion
Anterior Ventilation
Perfusion
Posterior (a)
Figure 61.2 (a) (Left) Chest radiograph showing mild linear atelectasis at left base; (right) perfusion scan, anterior view, showing a massive perfusion defect (arrow) that is much larger than the radiographic abnormality. (b) (Left) Perfusion scan, posterior view, showing the same large perfusion defect (arrow) seen in the anterior view in Figure 61.2(a); (right) ventilation scan, posterior view, equilibrium phase, showing nearly normal ventilation. The ventilation–perfusion lung scan was interpreted as high probability for pulmonary embolism. A pulmonary angiogram showed pulmonary embolism. Technique of ventilation–perfusion scans was as in Figure 61.1. (Courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, MI.)
L Perfusion
(b)
L
L Perfusion
273
Ventilation
BLUK077-Stein
March 11, 2007
10:38
274
PART III
(a)
Anterior
Diagnosis of acute PE
Posterior
R
R
(b)
RPO
(Table 61.2). A larger number of patients with low probability or nearly normal interpretations of the ventilation–perfusion scans were included in the database when outcome analysis was employed in addition to pulmonary angiography. The following results refer to patients in whom outcome analysis in combination with pulmonary angiography was employed, although both sets of data are shown in Table 61.2. Among the general population of patients suspected of PE, a high-probability ventilation–perfusion scan using original PIOPED criteria was indicative of PE in 87% (Table 61.2) [1]. Among patients in whom the ventilation–perfusion scan probability interpretation was intermediate, PE was present in 30%. Therefore, the intermediate or indeterminate interpretation was uninformative. Among patients in whom a low-probability interpretation of the ventilation–perfusion lung scan was made, PE was present in 14%. Some physicians believe that this percentage of patients with PE who have a low-probability ventilation–perfusion scan is too high to adequately exclude PE [4]. Therefore, patients
LPO
Figure 61.3 Normal perfusion lung scan in a patient with borderline cardiomegaly and a prominent aorta. (a) (Left) Anterior view; (right) posterior view. (b) (Left) Right posterior oblique (RPO) view; (right) left posterior oblique (LPO) view. A pulmonary angiogram was negative for pulmonary embolism. The perfusion scan was obtained with Tc-99m macroaggregated albumin. (Courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, MI.)
Table 61.2 Ventilation–perfusion lung scan results: all randomized patients using original pioped criteria. #PE/#PTS (%) V–Q
Pulmonary
scan probability
angiography and
Diagnostic quality pulmonary
category
outcome analysis*
angiography only
High probability
103/118 (87)
102/116 (88)
Intermediate
104/345 (30)
105/322 (33)
40/296 (14)
39/238 (16)
5/128 (4)
5/55 (9)
probability Low probability Near normal/ normal Total
252/887 (28)
251/731 (34)
*Pulmonary embolism was diagnosed by pulmonary angiography in 248 patients and by autopsy in 4 patients. Pulmonary embolism was excluded by pulmonary angiography in 465 patients. Pulmonary embolism was excluded by outcome assessment (the absence of adverse events during 1-year follow-up while not receiving antibiotics) in 170 patients. PE: pulmonary embolism; PTS, patients; V–Q, ventilation– perfusion. Data are from PIOPED [1].
BLUK077-Stein
March 11, 2007
CHAPTER 61
10:38
275
V–Q scan results from PIOPED
with low-probability ventilation–perfusion scans require further diagnostic studies [4–6]. A nearly normal or normal ventilation–perfusion lung scan by original PIOPED criteria showed PE in only 4% of the patients (Table 61.2). Conversely, such an interpretation excluded PE in 96% of the patients. Among patients with nearly normal ventilation– perfusion lung scans included in both arms of PIOPED (patients who volunteered for random assignment to obligatory angiography and patients referred for angiography), the frequency of PE was 8 of 165 (5%) [7]. PE was diagnosed more frequently among patients with nearly normal ventilation–perfusion lung scans who underwent pulmonary angiography, 8 of 75 (11%), than in those in whom PE was diagnosed on the basis of an adverse outcome while receiving no anticoagulant therapy, 0 of 90 (0%) [7]. At first glance, this may suggest that recurrent PE was not apparent in patients with PE of such mild severity that the ventilation–perfusion scan was nearly normal. There is no evidence that this was the case, however [7]. In patients with nearly normal ventilation–perfusion scans who were assessed by outcome events in comparison to the patients assessed by angiography, the ventilation–perfusion scan showed fewer mismatched segmental perfusion defects, very low probability, or normal ventilation–perfusion interpretations by one of the two ventilation–perfusion readers as well as a generally lower clinical assessment [7]. The observed lower frequency of PE in patients evaluated by outcome events in comparison to those who underwent angiography, therefore, can be attributed to a lower likelihood of PE. A normal ventilation–perfusion scan entirely excluded PE in the experience of PIOPED and others [1, 8–10]. There were 21 patients in PIOPED who had ventilation–perfusion scans read as normal by both the central readers. Three underwent angiography and none showed thromboemboli. None of the remaining 18 patients received anticoagulants and none had clinically evident PE on follow-up. Sporadic case reports suggest that PE may occur in the presence of a normal perfusion lung scan. If these reports are correct, such cases are extremely rare. The major theoretical reasons for PE associated with a normal perfusion scan are central, nonobstructing, nonlateralized PE or minimal defects on the perfusion scan that are not appreciated [11]. Patients with a single, small, or partially occluding embolus may not show
a perfusion defect because of limitations of perfusion scanning. A history of prior PE diminished the positive predictive value of a ventilation–perfusion lung scan for acute PE. Among 19 patients with histories of prior PE and a high-probability ventilation–perfusion scan, the positive predictive value of the high-probability ventilation–perfusion scan was only 74% compared with 91% for those without a history of PE. This difference of positive predictive values reflected a loss of specificity in the high-probability ventilation– perfusion scan diagnosis for patients with histories of PE. The specificity of the ventilation–perfusion scan was only 88% among patients with a history of PE whereas it was 98% among patients with no prior PE [1]. In PIOPED, only 13% of the patients had highprobability ventilation–perfusion scans. Intermediateprobability ventilation–perfusion scan readings occurred in 39% of the patients (Figure 61.4). Lowprobability ventilation–perfusion scans occurred in 34% and nearly normal ventilation–perfusion scans in 14%. Among patients with PE, 41% showed a highprobability ventilation–perfusion scan, but some had low probability or even nearly normal ventilation– perfusion scans (Figure 61.5).
Clinical assessment in combination with lung scans One of the important contributions of PIOPED was the demonstration of the value of clinical assessment in combination with ventilation–perfusion lung scan. Clinical assessment used in combination with the findings of ventilation–perfusion lung scans strengthened the diagnostic value of the ventilation–perfusion scan [1]. In PIOPED, if both the independent clinical assessment and findings by ventilation–perfusion lung scans were high probability for PE, the diagnosis was correct in 97% of the patients (Table 61.3) [1]. If both the independent clinical assessment and findings by ventilation–perfusion lung scans were low probability, PE was present in only 4% of the patients. Unfortunately, these concordant diagnostic combinations were uncommon, occurring in only 28% of the patients with clinically suspected PE. Either clinical uncertainty or uncertainty regarding the ventilation–perfusion lung scan findings (i.e., intermediate ventilation–perfusion scan pattern) was present in 72% of the patients.
BLUK077-Stein
March 11, 2007
Near NL/NL 14%
Low 34%
High 41%
10:38
High 13%
Intermed 39%
Figure 61.4 Distribution of interpretations of ventilation–perfusion lung scans in all patients with suspected pulmonary embolism randomized for investigation in PIOPED. Intermed, intermediate; Near NL /NL, near normal/normal. (Data from PIOPED [1].)
Near normal 16%
Low 16%
Intermed 41%
Figure 61.5 Distribution of interpretations of ventilation–perfusion lung scans in patients with proven pulmonary embolism randomized for investigation in PIOPED. Intermed, intermediate; Near NL /NL, near normal/normal. (Data from PIOPED [1].)
Table 61.3 Clinical assessment and ventilation–perfusion scan probability in PIOPED. #PE/#PTS (%) V–Q scan probability category
Highly likely
High probability
28/29 (97)
70/80 (88)
5/9 (56)
Intermediate probability
27/41 (66)
66/236 (28)
11/68 (16)
6/15 (40)
30/191 (16)
4/90 (4)
0/5 (0)
4/62 (6)
1/61 (2)
Low probability Near normal/normal Total
61/90 (68)
Uncertain
170/569 (30)
Unlikely
21/228 (9)
Highly likely = 80–100% likelihood of PE based on clinical assessment; uncertain = 20–79% likelihood of PE based on clinical assessment; unlikely = 0–19% likelihood of PE based on clinical assessment. Pulmonary embolism was diagnosed by pulmonary angiography in 248 patients and by autopsy in 4 patients. Pulmonary embolism was excluded by pulmonary angiography in 465 patients. Pulmonary embolism was excluded by outcome assessment (the absence of adverse events during 1-year follow-up while not receiving antibiotics) in 170 patients. PE, pulmonary embolism; PTS, patients; V–Q, ventilation–perfusion. Modified from PIOPED [1].
276
BLUK077-Stein
March 11, 2007
CHAPTER 61
10:38
V–Q scan results from PIOPED
The probability of PE with various ventilation– perfusion scan probabilities combined with various concordant and discordant clinical suspicions is shown in Table 61.3.
References 1 A Collaborative Study by the PIOPED Investigators: Value of the ventilation/perfusion scan in acute pulmonary embolism: Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 2 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 3 Morrell NW, Nijran KS, Jones BE, Biggs T, Seed WA. The underestimation of segmental defect size in radionuclide lung scanning. J Nucl Med 1993; 34: 370–374. 4 Hull RD, Raskob GE. Low-probability lung scan findings: a need for change. Ann Intern Med 1991; 114: 142–143. 5 Stein PD, Hull RD, Saltzman HA, Pineo G. Strategy for diagnosis of patients with suspected acute pulmonary embolism. Chest 1993; 103: 1553–1559.
277
6 Stein PD, Hull RD, Pineo G. Strategy that includes serial and noninvasive leg tests for diagnosis of thromboembolic disease in patients with suspected acute pulmonary embolism: estimated percentage of patients, based on data from PIOPED, in whom a noninvasive diagnosis or exclusion of thromboembolic disease might be safely made. Arch Intern Med 1995; 155: 2101–2104. 7 Henry JW, Stein PD, Gottschalk A, Raskob G. Pulmonary embolism among patients with a nearly normal ventilation/perfusion lung scan. Chest 1996; 110: 395– 398. 8 Stein PD. Low-dose heparin for prevention of pulmonary embolism and significance of normal lung scan. ACCP Bull 1982; 21: 12–14. 9 Hull RD, Raskob GE, Coates G, Panju AA. Clinical validity of a normal perfusion lung scan in patients with suspected pulmonary embolism. Chest 1990; 97: 23–26. 10 Kipper MS, Moser KM, Kortman KE, Ashburn WL. Long term follow-up of patients with suspected pulmonary embolism and a normal lung scan. Chest 1982; 82: 411– 415. 11 Stein PD, Gottschalk A. Critical review of ventilation/ perfusion lung scans in acute pulmonary embolism. Prog Cardiovasc Dis 1994; 37: 13–24.
BLUK077-Stein
April 6, 2007
62
16:14
CHAPTER 62
Ventilation–perfusion lung scans in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior cardiopulmonary disease, and patients with chronic obstructive pulmonary disease
Data from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) included 133 patients with a normal chest radiograph, among whom only 20 had pulmonary embolism (PE) [1] (Table 62.1, Figure 62.1). Although the ventilation–perfusion (V–Q) lung scan was diagnostic in a higher proportion of patients with a normal chest radiograph than with more complex chest findings, still only 52% were diagnostic (high probability or nearly normal/normal) [1]. Others, however, showed diagnostic V–Q scans (high probability or normal) in 201 of 220 patients (91%) with a normal chest radiograph [2]. Based on data from PIOPED and using the original criteria for the evaluation of ventilation–perfusion scans from PIOPED, as the complexity of associated
cardiopulmonary disease increased, the proportion of patients with diagnostic interpretations decreased [1, 3, 4]. In patients with no prior cardiopulmonary disease, 37% of patients showed diagnostic V–Q scans and in patients with any cardiopulmonary disease, 23% showed diagnostic interpretations [3] (Table 62.1, Figure 62.1). With even more complex disease (chronic obstructive pulmonary disease, COPD), only 10% showed diagnostic V–Q scans [4] (Table 62.1, Figure 62.1). Among each of these categories of patients, the positive predictive value of high, low, and nearly normal ventilation–perfusion lung scans was similar [1, 3, 4] (Table 62.1). The sensitivities of high probability ventilation– perfusion scans and specificities of a nearly normal/
Table 62.1 V–Q scan findings in patients with various degrees of complexity of associated cardiopulmonary disease. V–Q scan
Normal chest
No prior
Any prior
COPD‡
probability
radiograph* [PE/n (%)]
CPD† [PE/n (%)]
CPD† [PE/n (%)]
[PE/n (%)]
High
6/9 (67)
5/54 (93)
55/66 (83)
4/17 (24)
47/119 (39)
58/227 (26)
14/65 (22)
Low
8/47 (17)
17/113 (15)
25/182 (14)
2/33 (6)
Near normal/NL
2/60 (3)
3/79 (4)
2/51 (4)
0/5 (0)
Intermediate
5/5 (100)
*Stein et al. [1], † Stein et al. [3], ‡ Lesser et al. [4]. V–Q, ventilation–perfusion; PE, pulmonary embolism; CPD, cardiopulmonary disease; COPD, chronic obstructive pulmonary disease; NL, normal.
278
BLUK077-Stein
April 6, 2007
CHAPTER 62
16:14
Normal chest X-ray
Low 35%
279
V–Q scan and normal chest radiograph
Any CPD
No prior CPD
Near NL/NL Low 31% 45%
COPD
Near NL/NL 22%
Near NL/NL 22%
High Low 13% 31%
Low 35%
Near NL/NL High 10% 10%
High 15% Intermed High 13% 7%
Intermed 33%
Intermed 43%
Intermed 60%
Figure 62.1 Distribution of interpretations of ventilation–perfusion scans among patients with a normal chest radiograph, no prior cardiopulmonary disease (CPD), patients with any prior cardiopulmonary disease, and patients with chronic obstructive pulmonary disease (COPD). As the complexity of disease increased, the
proportion of patients with diagnostic ventilation– perfusion scans (high probability or nearly normal/normal) decreased from 52 to 10%. Some of the distributions do not add to exactly 100% because percentages were rounded off. NL, normal; Intermed, intermediate. (Data from Stein et al. [1, 3] and Lesser et al. [4].)
Table 62.2 Sensitivity and specificity of V–Q scans in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior
cardiopulmonary disease, and patients with chronic obstructive pulmonary disease.
Normal chest radiograph*
Sensitivity high prob
Specificity normal or near normal V–Q
V–Q high/PE + (%)
V–Q normal or near NL/PE − (%)
9/20 (45)
60/113 (53)
No prior cardiopulmonary disease†
54/72 (75)
79/293 (27)
Any prior cardiopulmonary disease†
66/140(47)
51/386 (13)
5/21 (24)
5/87 (6)
COPD‡
*Stein et al. [1], † Stein et al. [3], ‡ Lesser et al. [4]. V–Q, ventilation–perfusion; PE, pulmonary embolism; prob, probability; COPD, chronic obstructive pulmonary disease.
normal V–Q scan are shown in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior cardiopulmonary disease, and patients with COPD [1, 3, 4]. With increasing complexity of pulmonary disease, in general, the sensitivity of a high-probability V–Q scan decreased and the specificity of a nearly normal/normal V–Q scan also decreased (Table 62.2).
References 1 Stein PD, Alavi A, Gottschalk A et al. Usefulness of noninvasive diagnostic tools for diagnosis of acute pulmonary
embolism in patients with a normal chest radiograph. Am J Cardiol 1991; 67: 1117–1120. 2 Forbes KP, Reid JH, Murchison JT. Do preliminary chest X-ray findings define the optimum role of pulmonary scintigraphy in suspected pulmonary embolism? Clin Radiol 2001; 56: 397–400. 3 Stein PD, Coleman RE, Gottschalk A, Saltzman HA, Terrin ML, Weg JG. Diagnostic utility of ventilation/perfusion lung scans in acute pulmonary embolism is not diminished by pre-existing cardiac or pulmonary disease. Chest 1991; 100: 604–606. 4 Lesser BA, Leeper KV, Stein PD et al. The diagnosis of pulmonary embolism in patients with chronic obstructive pulmonary disease. Chest 1992; 102: 17–22.
BLUK077-Stein
April 14, 2007
63
15:1
CHAPTER 63
Perfusion lung scans alone in acute pulmonary embolism
Ventilation lung scans are obtained in combination with perfusion lung scans with the thought that ventilation would be abnormal in areas of pneumonia or local hypoventilation, but would be normal in pulmonary embolism (PE) [1]. However, among 98 patients in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I), the positive predictive values of high probability, intermediate probability, low probability, and near normal/normal interpretations of perfusion scans alone were comparable with the positive predictive values of ventilation– perfusion scans [2] (Table 63.1). A high probability perfusion scan had no less positive predictive value for acute PE than a high probability ventilation– perfusion scan [2]. Similarly, a low probability or near normal/normal perfusion scan excluded PE with no less validity than a low probability or near normal/normal ventilation–perfusion scan [2]. Original criteria for ventilation–perfusion scan interpretation from the PIOPED I were used [3]. Perfusion defects, in the absence of a ventilation scan, were interpreted on the basis of regional findings on the plain chest radiograph. Table 63.1 Positive predictive values of perfusion scans alone compared to ventilation–perfusion scans. Perfusion scan
V–Q scan
Scan interpretation
PE/n(%)
PE/n(%)
High
14/15 (93)
15/16 (94)
Intermediate
14/38 (37)
9/25 (36)
0/12 (0)
5/25 (20)
Low Near normal/normal
1/2 (50)
0/1 (0)
All differences between perfusions scan alone and ventilation–perfusion (V–Q) scans were not significant. PE, pulmonary embolism; n, number of patients; V–Q, ventilation–perfusion. Reprinted and modified from Stein et al. [2], with permission from Elsevier.
280
The positive predictive value of high probability interpretations of perfusion scans alone was 93%, which did not differ significantly from that of high probability interpretations of ventilation–perfusion scans 94% (Table 63.1) [2]. The positive predictive value for PE among patients with low probability interpretations of perfusion lung scans alone, 0%, did not differ significantly from that of low probability interpretations of ventilation– perfusion scans, 20% [2] (Table 63.1). Few patients had nearly normal/normal ventilation–perfusion or perfusion scans. Comparisons in this category, therefore, were not meaningful. Fewer patients who had perfusion scans alone had low probability interpretations than did those with ventilation–perfusion scans (23% versus 38%) and a trend suggested that more patients with perfusion scans alone had intermediate probability interpretations than did those with ventilation–perfusion scans [2] (Table 63.2). The sensitivity of high probability perfusion scans alone, 48%, was comparable with that of high probability ventilation–perfusion scans, 52%. The specificity of
Table 63.2 Distribution of lung scan interpretations in 98 patients with perfusion scans alone and ventilation– perfusion scans. Perfusion scan
V–Q scan
Scan interpretation
(%)
(%)
High
17
18
Intermediate
49
35
Low
23*
38
Near normal/normal
10
9
*P < 0.05. V–Q, ventilation–perfusion. Reprinted and modified from Stein et al. [2], with permission from Elsevier.
BLUK077-Stein
April 14, 2007
CHAPTER 63
15:1
281
Perfusion lung scans alone in acute PE
high probability perfusion scans alone, 97%, was identical to that of high probability ventilation–perfusion scans [2]. Investigators in the PISA-PED study used only perfusion lung scans; they did not use ventilation lung scans [4]. A perfusion scan compatible with PE was defined as a scan with a single or multiple wedge-shaped perfusion defects of any size. Diversion of blood away from an underperfused area usually resulted in associated overperfused single or multiple wedge-shaped areas. A perfusion scan was considered negative for PE if single or multiple perfusion defects did not exhibit a wedge shape. Positive perfusion scans had a positive predictive value of 95%. A negative perfusion scan had a negative predictive value of 81%. Combining clinical assessment with the perfusion scan showed good results when clinical assessment and the perfusion scan reading were concordant. Only the minority of patients (21%) had discordant clinical and perfusion scan assessments, and these patients required pulmonary angiography.These results at first glance appear better than the results of PIOPED. In the PISA-PED study, however, 34% of the patients had normal perfusion
scans and 61% had normal or nearly normal perfusion scans, whereas in PIOPED, only 2% had normal scintiscans and 14% had normal or nearly normal perfusion scans. The two study populations, therefore, were dissimilar.
References 1 Wagner HN, Jr, Lopez-Majano V, Langan JK, Joshi RC. Radioactive xenon in the differential diagnosis of pulmonary embolism. Radiology 1968; 91: 1168–1174. 2 Stein PD, Terrin ML, Gottschalk A, Alavi A, Henry JW. Value of ventilation/perfusion scans versus perfusion scans alone in acute pulmonary embolism. Am J Cardiol 1992; 69: 1239–1241. 3 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism—results of the prospective investigation of pulmonary embolism diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 4 Invasive and noninvasive diagnosis of pulmonary embolism. Preliminary results of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISAPED). Chest 1995; 107: 33S–38S.
BLUK077-Stein
March 28, 2007
64
15:49
CHAPTER 64
Probability interpretation of ventilation–perfusion lung scans in relation to largest pulmonary arterial branches in which pulmonary embolism is observed
In the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I) patients, with high probability interpretations of V–Q scans were likely to have PE (pulmonary embolism) in main or lobar pulmonary arteries, whereas patients with low probability V–Q scans had a higher prevalence of PE limited to subsegmental pulmonary arterial branches [1]. The prevalence of PE according to the largest pulmonary artery involved is shown for patients with high, intermediate, and low probability interpretations of V–Q scans (Table 64.1). Eighty-five percent of patients with a high probability interpretation of a V–Q scan had PE in main or lobar pulmonary arteries (Table 64.1) [1]. Conversely, those with low probability interpretations had PE in main or lobar pulmonary arteries in only 31%. Patients with a low probability interpretation of the V–Q lung scan had a higher prevalence of PE lim-
ited to the subsegmental branches, 17%, than patients with an intermediate probability interpretation, 6%, or patients with a high probability interpretation, 1%. These data are further analyzed in patients according to the presence or absence of prior cardiopulmonary disease (Table 64.2) [1]. Patients with low probability interpretations of the V–Q scan and no prior cardiopulmonary disease had PE limited to the subsegmental pulmonary arteries in 30% whereas patients with low probability V–Q scans who had prior cardiopulmonary disease had PE limited to subsegmental pulmonary arteries in 8%. Oser and associates showed PE limited to subsegmental pulmonary arteries or smaller in 23 of 76 (30%) patients with angiographically diagnosed PE [2]. Goodman and associates showed PE limited to subsegmental pulmonary arteries in 4 of 11
Table 64.1 Largest pulmonary artery with pulmonary embolism in relation to the ventilation–perfusion lung scan. Interpretation
Main or lobar [n/N(%)]
High prob V–Q
126/148 (85)
21/148 (14)
1/148 (1)
71/160 (44)
80/160 (50)
9/160 (6)
18/59 (31)
31/59 (53)
217/375 (58)
136/375 (36)
Intermediate prob V–Q Low prob V–Q All patients
Segmental [n/N(%)]
Subsegmental [n/N(%)]
10/59 (17)*† 22/375 (6)
*P < 0.01 low V–Q vs. intermediate V–Q. † P < 0.001 low V–Q vs. high V–Q. High prob V–Q, high probability interpretation of ventilation–perfusion (V–Q) lung scan; intermediate prob V–Q, intermediate probability interpretation of V–Q; low prob V–Q, low probability interpretation of V–Q. Reprinted with permission from Stein et al. [1].
282
BLUK077-Stein
March 28, 2007
CHAPTER 64
15:49
283
Probability interpretation according to largest pulmonary arter involved
Table 64.2 Largest pulmonary artery with pulmonary embolism according to ventilation–perfusion scan among patients stratified for prior cardiopulmonary disease. Interpretation
Main or lobar [n/N(%)]
Segmental [n/N(%)]
Subsegmental [n/N(%)]
No CPD
56/68 (82)
12/68 (18)
0/68 (0)
CPD
70/80 (88)
9/80 (11)
1/80 (1)
No CPD
29/73 (40)
40/73 (55)
4/73 (5)
CPD
40/83 (48)
39/83 (47)
4/83 (5)
7/23 (30)
9/23 (39)
7/23 (30)
11/36 (31)
22/36 (61)
3/36 (8)*
High prob V–Q
Intermediate prob V–Q
Low prob V–Q No CPD CPD
*P < 0.05 low probability V–Q with no CPD vs. low probability V–Q with CPD. High prob V–Q, high probability interpretation of ventilation–perfusion (V–Q) lung scan; intermediate prob V–Q, intermediate probability interpretation of V–Q; low prob V–Q, low probability interpretation of V–Q; CPD, cardiopulmonary disease. Reprinted with permission from Stein et al. [1].
(36%) patients with angiographically diagnosed PE [3]. Quinn and associates reported PE limited to subsegmental pulmonary arteries in 2 of 20 (10%) patients with PE [4]. The prevalence that we observed of PE limited to subsegmental pulmonary arteries (6%) was lower than reported by Oser and associates (30%) and by Goodman and associates (36%), but was comparable to the frequency reported by Quinn and associates (10%) [2–4]. Goodman and associates also reported a higher prevalence of low probability interpretations of the V–Q scan among patients with PE, 5 of 11 (45%) than observed in PIOPED, 59 of 375 (16%) [3]. The patients in PIOPED, therefore, may have had more severe PE than the patients reported by Goodman and associates.
References 1 Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/ perfusion lung scans. Chest 1997; 111: 1246–1248. 2 Oser RF, Zuckerman DA, Gutierrez FR, Brink JA. Anatomic distribution of pulmonary emboli at pulmonary angiography: implications for cross-sectional imaging. Radiology 1996; 199: 31–35. 3 Goodman LR, Curtin JJ, Mewissen MW et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. Am J Roentgenol 1995; 164: 1369–1374. 4 Quinn MF, Lundell CJ, Klotz TA et al. Reliability of selective pulmonary arteriography in the diagnosis of pulmonary embolism. Am J Roentgenol 1987; 149: 469–471.
BLUK077-Stein
March 11, 2007
65
11:1
CHAPTER 65
Revised criteria for evaluation of lung scans recommended by nuclear physicians in PIOPED
After the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) was concluded, nuclear physicians in PIOPED reevaluated the original criteria for interpretation of ventilation–perfusion lung scans [1]. They correlated probability estimates of pulmonary embolism (PE) based on the ventilation– perfusion lung scans with the actual frequency of PE as determined by pulmonary angiography. The nuclear physicians in PIOPED determined that most of the original PIOPED criteria appropriately categorized ventilation–perfusion scans [1]. However, it was recommended that three criteria should be reconsidered (Table 65.1):
1 Two large segmental mismatches may not be the optimum threshold for high probability, and in some cases should be considered for intermediate probability. However, due to the small number of cases with this finding, no definite, statistically founded recommendation could be made. 2 A single moderate mismatched perfusion defect is appropriately categorized as intermediate, rather than as low probability. 3 Extensive matched ventilation–perfusion abnormalities are appropriate for low probability, provided that the chest radiograph is clear. Single matched defects may be better categorized as intermediate probability. Due to the small number of cases with this finding, no definite, statistically founded recommendation could be made. It was suggested that the revised criteria resulting from these adjustments should now be used for the interpretation of ventilation–perfusion scans [1]. Analyses focused upon individual PIOPED criteria. In many instances, combined patterns were excluded. Combined patterns involving mismatched
284
defects were considered in the analysis of criteria for “high probability,” since mismatched defects and high probability diagnoses took priority over other patterns in the PIOPED criteria. For example, a patient with three segmental mismatches (which meets criteria for high probability) and matched ventilation–perfusion defects (which meet the criteria for low probability) should have been assigned to the high probability category. Therefore, all patients with mismatched perfusion defects, including those with combined patterns, were considered in the analysis of high probability [1]. When evaluating low probability criteria, since these did not take precedence in scan categorization, it was necessary to isolate analyses to those patients who did not have scan findings that would place them in a higher category [1]. Correlations of scintigraphic patterns that fulfilled individual PIOPED criteria were made with pulmonary angiograms read as definitely positive or definitely negative for PE [1]. There were 731 patients in PIOPED in the randomized obligatory pulmonary angiography group who had definitive angiographic results [2]. Correlations of ventilation–perfusion scans with pulmonary angiograms were made in 393 of those patients (54%) who satisfied pertinent individual PIOPED criteria on their ventilation–perfusion lung scans [1]. Analyses of the PIOPED criteria for the evaluation of ventilation–perfusion scans arbitrarily defined the following ranges:
1 High probability = 80–100% likelihood of PE. 2 Intermediate probability = 20–79% likelihood of PE. 3 Low probability = 0–19% likelihood of PE. Two large mismatched perfusion defects did not provide a reliable interpretation of high probability,
mismatch
V–Q are borderline
285
mismatches
mismatches
normal perfusion in lung
≥1 small Q with Nl X-ray
Nl X-ray
very low probability ≤3 small Q with
or
>3 small Q with Nl X-ray
or
elevated diaphragm
aorta, hilum, mediastinum, and
effusion, cardiomegaly, enlarged
Perfusion defects due to pleural
elevated diaphragm or
or
mediastinum, and
aorta, hilum,
probability
cardiomegaly, enlarged
region
borderline intermediate to high
pleural effusion,
lung and ≤3 segment in 1 lung
mismatches are
Perfusion defects due to
or
Q involving ≤4 segments in 1
V–Q matches with Nl X-ray or X-ray
or
Q X-ray
X-ray and some areas of
or ≥2 V–Q matches with Nl
or
≥1 Q X-ray defect
PIOPED revised (1)
X-ray (2 large V–Q
1 matched V–Q with Nl
or
mismatches
1–3 moderate V–Q
or
ray
1 moderate V–Q mismatch with Nl X-
1 large ±1 moderate V–Q mismatch
PIOPED (2)
PIOPED revised (1)
Low probability
Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality; mod, moderate; seg, segment. Moderate = 25–75% segment; large ≥75% segment; V–Q mismatch = Nl V and Nl X-ray or Q > V or X-ray in region of Q defect. Data from PIOPED [2] and Gottschalk et al. [1].
≥4 moderate V–Q
≥4 moderate V–Q
mismatches
mismatches or
moderate V–Q
moderate V–Q
or
2 or 3 moderate V–Q
1 large V–Q and ≥2
1 large V–Q and ≥2 mismatches
or
or
or
high probability)
±1 moderate V–Q
1 large V–Q mismatch
mismatches (2 large
≥2 large V–Q
≥2 large V–Q
PIOPED (2)
Intermediate probability
March 11, 2007
mismatches
PIOPED revised (1)
PIOPED (2)
High probability
Table 65.1 Original PIOPED (Prospective Investigation of Pulmonary Embolism Diagnosis) criteria for ventilation–perfusion lung scans compared with proposed revised criteria.
BLUK077-Stein
11:1
BLUK077-Stein
March 11, 2007
11:1
286
PART III
Diagnosis of acute PE
Figure 65.1 Left: Perfusion scan, posterior view, showing solitary perfusion defect (arrow), in the posterior basal segment of the right lower lobe. Right: Ventilation scan, posterior view, initial breath showing no ventilation defect. Equilibrium and washout phases also showed no ventilation defect. According to original PIOPED criteria
this single mismatched moderate size perfusion defect should have been interpreted as low probability for pulmonary embolism (PE), but according to the revised PIOPED criteria this ventilation–perfusion scan should be interpreted as intermediate probability. Pulmonary embolism was shown on the pulmonary angiogram.
whereas 2.5 mismatched segmental equivalent perfusion defects provided a more accurate categorization [1] (Table 65.2). A mismatched segmental equivalent perfusion defect is 1 large (>75% of a segment) mismatched perfusion defect or 2 moderate size (25–75% of a segment) perfusion defects. The nuclear physicians
in PIOPED could not definitely recommend changing the threshold for high probability, however, because the number of patients with this pattern was small. Ten of 28 patients (36%) with a single moderate mismatched perfusion defect had PE [1]. It was clear that this was not a valid criterion for low probability. Scans with this finding should be considered intermediate probability for PE (Figure 65.1). Patients with a single matched perfusion appeared to have a higher likelihood of PE than patients with 2 or more matched perfusion defects [1]. A single matched defect was associated with PE in 26% of patients (Table 65.3). Two or more matched perfusion defects were associated with PE in 14% of patients. A single matched perfusion defect, therefore, should be considered for intermediate probability.
Table 65.2 Frequency of pulmonary embolism in patients with various patterns of mismatched perfusion defects. Number of Type of mismatch
patients
PE (%)
1 segmental equivalent
33
52
1 large defect
24
46
2 moderate defects
9
67
1.5 segmental equivalents
18
72
1 large + 1 moderate defect
11
73
3 moderate defects
7
71
2 segmental equivalents
7
71
2 large defects
5
80
1 large + 2 moderate defects
1
0
4 moderate defects
1
100
10
100
2 large + 1 moderate defect
8
100
1 large + 3 moderate defects
1
100
5 moderate defects
1
100
2.5 segmental equivalents
Defect = mismatched perfusion defect. PE, pulmonary embolism. Reprinted by permission of the Society of Nuclear Medicine from Gottschalk et al. [1].
Table 65.3 Patients with only matched defects on the ventilation–perfusion scan. Number of points
PE (%)
Single matched defect (any size)
23
26
Multiple matched defects (any size)
66
14
All matched defects
89
17
PE, pulmonary embolism. Data from Gottschalk et al. [1].
BLUK077-Stein
March 11, 2007
CHAPTER 65
11:1
Revised criteria for evaluation of V–Q scans
References 1 Gottschalk A, Sostman HD, Coleman RE et al. Ventilation– perfusion scintigraphy in the PIOPED study. Part II. Evaluation of the scintigraphic criteria and interpretations. J Nucl Med 1993; 34: 1119–1126.
287
2 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism—results of the prospective investigation of pulmonary embolism diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.
BLUK077-Stein
March 12, 2007
66
16:44
CHAPTER 66
Criteria for very low probability interpretation of ventilation– perfusion lung scans
Introduction The criteria for the interpretation of low probability ventilation–perfusion (V–Q) lung scans in patients with suspected acute pulmonary embolism (PE) that were used in the collaborative study of Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I) [1] have been modified since its conclusion [2]. In PIOPED I, using the original PIOPED criteria, 14% of patients with ventilation–perfusion scans interpreted as low probability had PE [1]. The PIOPED I criteria for low probability included [1]:
1 Nonsegmental perfusion defects ≤ radiographic abnormality. 2 Perfusion defects ≤ ventilation defects with normal chest radiograph. 3 Perfusion defects < radiographic defect. 4 ≤3 small perfusion defects with a normal chest radiograph. 5 A single moderate size mismatched perfusion defect with a normal chest radiograph. The Nuclear Medicine Working Group of PIOPED recommended that the following modifications be made in regard to low probability interpretations [2]:
1 A single moderate mismatched perfusion defect should be categorized as intermediate rather than as low probability. 2 Multiple and relatively extensive matched ventilation–perfusion abnormalities are appropriate for low probability, provided that the chest radiograph is clear. 3 Single matched defects may be better categorized as intermediate probability, although this cannot be definitely validated statistically. The modifications of the PIOPED I criteria for low probability were made on the assumption that patients
288
with low probability interpretations of ventilation– perfusion scans should have a positive predictive value of PE <20% [2]. These revised PIOPED criteria were found to be more accurate than the original PIOPED criteria [3]. The modified PIOPED I criteria, however, were tested prospectively in a population of patients that differed significantly from the population of patients investigated in PIOPED I [4]. In the population of patients investigated, 36% had normal pulmonary scintiscans, in comparison to 2% in PIOPED and the prevalence of PE was also lower, 10% versus 28% in PIOPED. Using the modified criteria, patients with a low probability interpretation had a 6% frequency of PE [4], whereas in PIOPED such patients had a 14% frequency of PE A “very low probability” interpretation with a positive predictive value less than 10% would be more useful than a “low probability” interpretation [5]. To identify such “very low probability” criteria of this, we evaluated the individual characteristics and combinations of characteristics of the low probability ventilation–perfusion lung scan in PIOPED I [6]. In some instances, we maximized the useful PIOPED database by evaluation of individual lungs or lung zones rather than individual patients [6, 7].
Nonsegmental perfusion abnormalities Nonsegmental perfusion abnormalities that were associated with a low positive predictive value for PE were enlargement of the heart, hila, or mediastinum, or elevated hemidiaphragm (Figure 66.1). Lungs with such nonsegmental perfusion abnormalities had a positive predictive value for PE of 8% [6] (Table 66.1, Figure 66.2).
BLUK077-Stein
March 12, 2007
CHAPTER 66
16:44
289
Very low probability interpretation of V–Q scans
Perfusion
(a) Figure 66.1 (a) Plain chest radiograph showing mild cardiomegaly and prominence of hila. (b) Perfusion lung scan, anterior view, showing nonsegmental perfusion abnormalities. The scan shows prominence of the hila, more clearly shown on the right (arrow) and cardiomegaly.
(b) These types of nonsegmental perfusion abnormalities were associated with a very low positive predictive value for pulmonary embolism (PE). (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)
Table 66.1 Positive predictive value of individual criteria used for low probability assessment of ventilation–perfusion lung scans. PE/total (%) Lungs with only one type of perfusion defect Costophrenic angle effusion
4/14 (29)
Nonsegmental abnormality
8/103 (8)
Perfusion defect < X-ray 1 zone
2/24 (8)
2 or 3 zones
1/16 (6)
All zones
3/40 (8)
Matched V–Q (X-ray normal) 1 zone
4/34 (12)
2 or 3 zones
1/30 (3)
All zones
5/64 (8)
Lungs with two types of perfusion defects Costophrenic angle effusion and nonsegmental abnormality
1/10 (10)
Costophrenic angle effusion and matched V–Q (X-ray normal)
1/8 (13)
Nonsegmental abnormality and perfusion defect < X-ray
3/34 (9)
Nonsegmental abnormality and matched V–Q (X-ray normal)
4/29 (14)
Combinations of two perfusion defects were excluded from the table if the combination was observed in only 4 or fewer lungs. Costophrenic angle effusion = Pleural effusion with obliteration of the costophrenic angle with the perfusion defect ≤ X-ray abnormality. Nonsegmental abnormality = Nonsegmental perfusion abnormality including enlargement of the hilum, mediastinum, or heart or elevated diaphragm with the perfusion defect ≤ X-ray abnormality. Perfusion defect < X-ray = Parenchymal abnormality on the chest radiograph with the perfusion defect < radiographic abnormality. Matched V–Q (X-ray normal) = Matched ventilation–perfusion defect with normal chest radiograph and perfusion defect ≤ ventilation defect. PE, pulmonary embolism. Reprinted by permission of the Society of Nuclear Medicine from Stein et al. [6].
BLUK077-Stein
March 12, 2007
16:44
290
PART III
Perfusion
(a)
(b)
Perfusion defect smaller than radiographic defect A perfusion defect smaller than the radiographic defect, regardless of the findings on the ventilation scan, showed PE in 8% of patients [6] (Table 66.1).
Matched ventilation–perfusion abnormality in 2 or 3 zones of a single lung Matched ventilation–perfusion abnormalities, in the presence of a regionally normal chest radiograph, when occurring as the only type of perfusion defect in 2 or 3 zones of a single lung, showed a pos-
L (a)
Perfusion
L (b)
Diagnosis of acute PE
Figure 66.2 (a) Plain chest radiograph showing linear atelectasis in the left lower zone. (b) Perfusion lung scan, anterior view, showing small perfusion defect (arrow) in the region of the atelectasis. Perfusion defects smaller than associated parenchymal abnormalities on the chest radiograph, when occurring as the only type of perfusion defect, showed a positive predictive value for pulmonary embolism (PE) of 8%. The pulmonary angiogram in these patients showed no PE. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)
itive predictive value for PE of 3% [6] (Figure 66.3, Table 66.1). A matched ventilation–perfusion defect occurring as the only type of perfusion defect in 1 zone of a single lung showed a trend toward a higher positive predictive value for PE than matched ventilation–perfusion defects in 2 zones or 3 zones in a single lung. A matched ventilation–perfusion defect in only 1 zone was not considered a criterion for a very low probability interpretation [6]. The positive predictive value for PE with matched ventilation–perfusion defects in 1 zone was 4 of 34 (12%), in 2 zones it was 1 of 20 (5%), and in 3 zones it was 0 of 10 (0%) [6] (Table 66.1).
Ventilation
Figure 66.3 (a) Perfusion lung scan, posterior view, showing a perfusion defect (arrow) in the left lower zone. (b) Ventilation scan, posterior view, initial breath, showing impaired ventilation in the left lower zone. A matched ventilation–perfusion defect occurring as the only type of perfusion defect in 1 zone of a single lung showed a trend toward a higher positive predictive value for pulmonary embolism (PE) than matched perfusion defects in 2 zones or 3 zones in a single lung. This patient showed no PE on the pulmonary angiogram. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)
BLUK077-Stein
March 12, 2007
CHAPTER 66
16:44
291
Very low probability interpretation of V–Q scans
Figure 66.4 Posterior view of a perfusion lung scan. Both the round lesion in the left apex and the retrocardiac lesion in the base of the left lung were small segmental perfusion defects (<25% of a segment). The patient showed no PE on pulmonary angiography. (Reprinted by permission of the Society of Nuclear Medicine from Stein et al. [8].
One to three small perfusion defects The positive predictive value for PE of lung scans with 1–3 small (<25% of a segment) segmental perfusion defects with a regionally normal chest radiograph was 1–3% depending on the group analyzed [8] (Figure 66.4). Evaluation of this criterion was independent of findings on the ventilation scan; a regionally normal chest radiograph was required. This low positive predictive value for PE is comparable to the positive predictive value reported by Stein [9]. Among patients in whom PE was diagnosed or excluded by pulmonary angiography or follow-up, 1 of 68 (1%) showed PE [8]. Among patients with >3 small segmental perfusion defects, 3 of 27 (11%) showed PE [8].
Triple matched defects in the upper or middle lung zones A matched ventilation–perfusion defect with associated matching chest radiographic opacity is defined as a triple matched defect. The triple match can be caused by PE creating a pulmonary “infarction” (usually pulmonary hemorrhage). Worsley and associates, showed that PE was present more frequently with triple matches in the lower zones of the lung compared with the upper or middle zones [10] (Table 66.2). We showed PE with triple matched defects was infrequent in the upper or middle lung zones, occurring in 1 of 27 (4%) (Figure 66.5, Table 66.2) [11]. In the lower zone,
Table 66.2 Positive predictive value of triple matched defects according to lung zone. Gottschalk
Worsley
[PE/No. (%)]
[PE/No. (%)]
Upper zone
0/13 (0)
Middle zone
1/14 (7)
4/36 (11)
Lower zone
13/57 (23)
61/187 (33)
Total
14/84 (17)
71/275 (26)
6/52 (12)
PE, pulmonary embolism; No., number. Reprinted by permission of the Society of Nuclear Medicine from Gottschalk et al. [11].
however, with a triple match, PE occurred in 13 of 57 (23%) [11].
Stripe sign The stripe sign on perfusion images consists of a stripe of perfused lung tissue between a perfusion defect and the adjacent pleural surface [10]. In PIOPED, perfusion defects showing the stripe sign had a positive predictive value of PE of only 6 of 85 (7%) [12].
Pleural effusion A pleural effusion with blunting of the costophrenic angle in combination with a nonsegmental perfusion defect (enlargement of the heart, hila, or mediastinum, or elevated hemidiaphragm) showed a
BLUK077-Stein
March 12, 2007
16:44
292
PART III
Diagnosis of acute PE
Figure 66.5 (a) Chest radiograph, lateral view, showing a parenchymal abnormality in the posterior portion of the base of the right lung (arrow). (b) Perfusion scan, lateral view of right lung, showing perfusion defect in posterior inferior portion (arrow). (c) Ventilation scan, posterior view, equilibrium phase, showing absent ventilation in base of right lung. The finding of a matched ventilation–perfusion defect with associated matching
chest radiographic opacity (the triple match) in a lower lung zone has an intermediate/indeterminate probability with a positive predictive value for acute pulmonary embolism (PE) of 23%. This patient had PE. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)
positive predictive value for PE of 10% [2] (Table 66.1). A small pleural effusion alone was associated with a higher positive predictive value for PE (29%). We prospectively evaluated the database of PIOPED II to test these criteria for very low probability and the additional criterion of a large pleural effusion [13]. Some of the criteria, however, were defined somewhat differently in PIOPED II. The diagnostic reference standards for PE were CT angiogram, CT venogram, and standard pulmonary digital subtraction angiography (DSA) [13]. Criteria for a very low probability of PE tested prospectively from data in PIOPED II were:
6 Stripe sign (consists of a stripe of perfused lung tissue between a perfusion defect and the adjacent pleural surface, best seen on a tangential view). 7 Pleural effusion ≥1/3 of the pleural cavity with no other perfusion defect in either lung.
1 Nonsegmental perfusion abnormalities: these were enlargement of the heart or hilum, elevated hemidiaphragm, costophrenic angle effusion, or linear atelectasis with no other perfusion defect in either lung. 2 Perfusion defect smaller than corresponding radiographic lesion. 3 ≥2 matched V–Q defects with regionally normal chest radiograph and some areas of normal perfusion elsewhere in the lungs. 4 1–3 small segmental perfusion defects (<25% of a segment). 5 Solitary triple matched defect (defined as a matched V–Q defect with associated matching chest radiographic opacity) in the mid or upper lung zone confined to a single segment.
A very low probability consensus interpretation of the V–Q scan was made in 460 of 824 patients with suspected PE (56%). Among these patients, the CT angiogram was obtained and of diagnostic quality in 422 patients. In these patients, PE was shown in 38 of 422 (9.0%) [13]. Each of the component V–Q criteria for very low probability showed a positive predictive value <10% except 1–3 small segmental perfusion defects (Table 66.3). However, some categories had too few patients for a meaningful assessment. The positive predictive value of a very low probability interpretation was 15 of 185 (8.1%). If the criterion of 1–3 small perfusion defects were eliminated, the positive predictive value of a very low probability interpretation would be 12 of 161 (7.5%). The data from PIOPED II confirmed that the very low probability V–Q interpretation has a positive predictive value for PE of <10% [13]. This places the low probability interpretation into a useful category for the exclusion of acute PE, keeping in mind that a very low probability interpretation of the V–Q scan was made in 56% of patients with suspected PE [13]. All of the component V–Q features of a very low probability interpretation showed a low (<10%) positive
BLUK077-Stein
March 12, 2007
CHAPTER 66
16:44
293
Very low probability interpretation of V–Q scans
Table 66.3 Positive predictive value of criteria for very low probability interpretation. Criterion
PE/patients (%)
All very low probability*
15/185 (8.1)
Nonsegmental perfusion abnormalities
10/120 (8.3)
Perfusion defect smaller than corresponding radiographic lesion
0/9 (0)
≥2 matched V–Q defects
5/57 (8.8)
1–3 small segmental perfusion defects
8/64 (12.5)
A solitary triple matched defect
0/3 (0)
Stripe sign
1/25 (4.0)
Pleural effusion ≥1/3 of the pleural cavity with no other perfusion defect in either lung
0/5 (0)
*Includes only patients in whom central readers identified the criterion on which a very low probability interpretation was based, and in whom central reader 1 and central reader 2 agreed on the presence of the criterion. PE, pulmonary embolism. Reproduced with permission from Gottschalk et al. [13].
predictive value except 1–3 small segmental perfusion defects [13]. Many criteria, however, require further evaluation because of sparse data.
References 1 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 2 Gottschalk A, Sostman HD, Coleman RE et al. Ventilation–perfusion scintigraphy in the PIOPED study. Part II. Evaluation of the scintigraphic criteria and interpretations. J Nucl Med 1993; 34: 1119–1126. 3 Sostman HD, Coleman RE, DeLong DM, Newman GE, Paine S. Evaluation of revised criteria for ventilation– perfusion scintigraphy in patients with suspected pulmonary embolism. Radiology 1994; 193: 103–107. 4 Freitas JE, Sarosi MG, Nagle CC, Yeomans ME, Freitas AE, Juni JE. Modified PIOPED criteria used in clinical practice. J Nucl Med 1995; 36: 1573–1578. 5 Hull RD, Raskob GE. Low-probability lung scan findings: a need for change. Ann Intern Med 1991; 114: 142–143. 6 Stein PD, Relyea B, Gottschalk A. Evaluation of individual criteria for low probability interpretation of ventilationperfusion lung scans. J Nucl Med 1996; 37: 577–581.
7 Stein PD, Gottschalk A. Review of criteria appropriate for a very low probability of pulmonary embolism on ventilation–perfusion lung scans: a position paper. Radiographics 2000; 20: 99–105. 8 Stein PD, Henry JW, Gottschalk A. Small perfusion defects in suspected pulmonary embolism. J Nucl Med 1996; 37: 1313–1316. 9 Silberstein E, Worsley DF, Alavi A, Elgazzar A. The clinical significance of the very low probability (PIOPED) lung scan pattern [abstract]. J Nucl Med 1995; 36: 113. 10 Worsley DF, Kim CK, Alavi A, Palevsky HI. Detailed analysis of patients with matched ventilation–perfusion defects and chest radiographic opacities. J Nucl Med 1993; 34: 1851–1853. 11 Gottschalk A, Stein PD, Henry JW, Relyea B. Matched ventilation, perfusion and chest radiographic abnormalities in acute pulmonary embolism. J Nucl Med 1996; 37: 1636–1638. 12 Sostman HD, Gottschalk A. Prospective validation of the stripe sign in ventilation–perfusion scintigraphy. Radiology 1992: 184: 455–459. 13 Gottschalk A, Stein PD, Sostman HD et al. Very low probability interpretation of ventilation-perfusion lung scans in combination with low probability clinical assessment reliably excludes pulmonary embolism: Data from PIOPED II (Submitted for Publication).
BLUK077-Stein
March 11, 2007
67
11:27
CHAPTER 67
Probability assessment based on the number of mismatched segmental equivalent perfusion defects or number of mismatched vascular defects
Introduction The diagnostic value of ventilation–perfusion lung scans was strengthened by creating a table of positive predictive values and specificities based on the observed number of mismatched segmental equivalent perfusion defects [1] (Table 67.1, Figure 67.1). This table was more informative than “high,” “intermediate,” or “low” probability clinical assessments, and permitted the nuclear scan reader to assign specific risks of pulmonary embolism (PE) to individual patients. In addition, stratification of patients according to the presence or absence of prior cardiopulmonary disease enhanced the ventilation–perfusion scan assessment of the probability of PE among patients in both of these clinical categories [1]. One segmental equivalent was defined as 1 large segment or 2 moderate size segments [2]. For example, 1.5 mismatched segmental equivalents can be obtained with either 1 large and 1 moderate mismatched segmental perfusion defect or 3 moderate mismatched segmental perfusion defects. With increasing numbers of mismatched segmental equivalent perfusion defects, there was a continuum of gradually increasing specificities and positive predictive values, and decreasing sensitivities [1] (Table 67.1, Figure 67.1). No discrete criterion for “high probability” can be clearly identified from the data. Rather than define an arbitrary number of mismatched segmental equivalent defects as “high probability” for all patients, use of Table 67.1 enables readers of ventilation–perfusion scans to assign a positive predictive value and specificity
294
to their individual patients according to prior disease status.
Number of mismatched segmental equivalent defects and stratification according to prior cardiopulmonary disease Stratification of patients according to the presence or absence of prior cardiopulmonary disease enhanced the ventilation–perfusion scan assessment of PE among both clinical categories of patients [1]. Among patients with no prior cardiopulmonary disease ≥0.5 mismatched segmental equivalent perfusion defects was associated with a positive predictive value for PE of 80% (Figure 67.2, Table 67.1). With ≥1.5 mismatched segmental equivalents, the positive predictive value for PE was 89% in patients with no prior cardiopulmonary disease. Among patients with prior cardiopulmonary disease, more mismatched segmental equivalent perfusion defects were required to give a particular positive predictive value and specificity than among patients with no prior cardiopulmonary disease. Among patients with prior cardiopulmonary disease, ≥2 mismatched segmental equivalent perfusion were required for an 80% positive predictive value of PE, and ≥4.5 mismatched segmental equivalents were required for a 90% positive predictive value of PE [1]. Further stratification of patients with prior cardiopulmonary disease into those with prior cardiac disease (exclusive of prior pulmonary disease) and
BLUK077-Stein
March 11, 2007
CHAPTER 67
11:27
295
Mismatched segmental perfusion defects or mismatched vascular defects
Table 67.1 Positive predictive value of cumulative number of mismatched segmental equivalent perfusion defects among patients with no prior cardiopulmonary disease and patients with prior cardiopulmonary disease. #PE/#PTS (%) Segmental
Patients with no prior
Patients with prior
All patients
equivalents
cardiopulmonary disease (n = 421)
cardiopulmonary disease (n = 629)
(n = 1050)
≥0.0
173/421 (41)
205/629 (33)*
378/1050 (36)
≥0.5
123/154 (80)
130/192 (68)†
253/346 (73)
≥1.0
102/118 (86)
113/155 (73)*
215/273 (79)
≥1.5
91/102 (89)
99/128 (77)†
190/230 (83)
≥2.0
79/87 (91)
91/114 (80)‡
170/201 (85)
≥2.5
72/80 (90)
87/105 (83)
159/185 (86)
≥3.0
65/73 (89)
81/97 (84)
146/170 (86)
≥3.5
60/67 (90)
77/88 (88)
137/155 (88)
≥4.0
57/63 (90)
74/84 (88)
131/147 (89)
≥4.5
50/53 (94)
70/78 (90)
120/131 (92)
≥5.0
49/52 (94)
65/72 (90)
114/124 (92)
≥5.5
47/50 (94)
61/66 (92)
108/116 (93)
≥6.0
42/44 (95)
59/64 (92)
101/108 (94)
≥6.5
40/42 (95)
56/61 (92)
96/103 (93)
≥7.0
38/40 (95)
51/56 (91)
89/96 (93)
≥7.5
34/36 (94)
43/47 (91)
77/83 (93)
*P < 0.01. † P < 0.02. ‡ P < 0.05: No prior cardiopulmonary disease vs. Prior cardiopulmonary disease. These probabilities were calculated with chi-square and are higher than reported in Stein et al. [1] because Yates’ correction, which was unnecessarily conservative, was used in the previous analysis. Modified and reprinted with permission from Stein et al. [1].
Pulmonary embolism (%)
100
Figure 67.1 Positive predictive value for pulmonary embolism (PE) according to the number of mismatched segmental equivalent perfusion defects on the ventilation–perfusion lung scan. Bars represent standard error.
90
80
70
60
≥0
≥1
≥2 ≥3 ≥4 ≥5 Cumulative mismatched segmental equivalents
≥6
BLUK077-Stein
March 11, 2007
11:27
296
PART III
Diagnosis of acute PE
Figure 67.2 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects. Broken line indicates patients with no prior cardiopulmonary disease (NO CPD). Unbroken line indicates patients with any prior cardiopulmonary disease (CPD). Significant differences occurred with ≥0.5 (P < 0.02), ≥1.0 segmental equivalents (P < 0.01), ≥1.5 (P < 0.02), and ≥2.0 (P < 0.05) segmental equivalents. These probabilities were calculated by chi-square and are higher than reported by Stein et al. [1] because Yates’ correction was unnecessarily conservative, was used in the previous calculations. (Reprinted with permission from Stein et al. [1].)
those with prior pulmonary disease (exclusive of prior cardiac disease) did not strengthen the positive predictive value of either group. No significant differences were shown between patients with prior cardiac disease and patients with prior lung disease (Figure 67.3). As shown in Table 67.1, some patients with PE had no mismatched large or moderate size segmental defects. The patients who had PE and no mismatched segmental lesions had either small perfusion defects, mediastinal abnormalities, pleural effusions, a variety of parenchymal radiographic abnormalities, ventilation– perfusion matched defects or any combination of
these. None had an entirely normal ventilation– perfusion scan. Sensitivity and specificity according to the number of mismatched segmental equivalent perfusion defects among patients stratified according to the presence or absence of prior cardiopulmonary disease is shown in Table 67.2. With stratification, among patients with no prior cardiopulmonary disease the presence of 1.0 or more mismatched segmental equivalent perfusion defects was more sensitive for PE than the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) criteria for high probability (≥2
Figure 67.3 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects among patients with no prior cardiopulmonary disease (NO CPD), any prior cardiac disease (CARD DIS), and any prior pulmonary disease (PULM DIS). There were no significant differences between patients with prior cardiac disease and patients with prior pulmonary disease. (Reprinted with permission from Stein et al. [1].)
BLUK077-Stein
March 11, 2007
CHAPTER 67
11:27
297
Mismatched segmental perfusion defects or mismatched vascular defects
Table 67.2 Number of mismatched segmental equivalent defects, sensitivity, and specificity among patients with no prior cardiopulmonary disease and patients with prior cardiopulmonary disease. Patients with no prior
Patients with prior
cardiopulmonary
cardiopulmonary
All patients
disease (n = 421)
disease (n = 629)
(n = 1050)
Segmental equivalents
SENS (%)
SPEC (%)
SENS (%)
SPEC (%)
SENS (%)
SPEC (%)
≥ 0.0
100
0
100
0
100
0
≥ 0.5
71
88
63
85
67
86
≥ 1.0
59
94
55
90
57
91
≥ 1.5
53
96
48
93
50
94
≥ 2.0
46
97
44
95
45
95
≥ 2.5
42
97
42
96
42
96
≥ 3.0
38
97
40
96
39
96
≥ 3.5
35
97
38
97
36
97
≥ 4.0
33
98
36
98
35
98
≥ 4.5
29
99
34
98
32
98
≥ 5.0
28
99
32
98
30
99
≥ 5.5
27
99
30
99
29
99
≥ 6.0
24
99
29
99
27
99
≥ 6.5
23
99
27
99
25
99
≥ 7.0
22
99
25
99
24
99
≥ 7.5
20
99
21
99
20
99
SENS, Sensitivity; SPEC, Specificity. Reprinted with permission from Stein et al. [1].
segmental equivalent defects) applied to unstratified patients (59% versus 41%) (P < 0.001) [1, 3]. This increased sensitivity was associated with a comparable positive predictive value (86% versus 87%) and specificity (94% versus 97%) [1, 3].
References 1 Stein PD, Gottschalk A, Henry JW, Shivkumar K. Stratification of patients according to prior cardiopulmonary disease and probability assessment based on the num-
ber of mismatched segmental equivalent perfusion defects: approaches to strengthen the diagnostic value of ventilation/perfusion lung scans in acute pulmonary embolism. Chest 1993; 104: 1461–1467. 2 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 3 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.
BLUK077-Stein
March 11, 2007
68
11:28
CHAPTER 68
Probability assessment based on the number of mismatched vascular defects and stratification according to prior cardiopulmonary disease
The diagnosis of pulmonary embolism (PE) on the basis of the number of mismatched segmental equivalents assumes that a mismatched moderate size segmental perfusion defect is of less diagnostic value than a mismatched large size segmental perfusion defect [1]. This assumption has not been fully tested. There is considerable skill and judgment required in distinguishing a moderate size mismatched segmental perfusion defect (25–75% of a segment) from a large size mismatched segmental perfusion defect (>75% of a segment). Experienced readers of radionuclide lung scans often underestimate the size of segmental defects [2]. Ventilation–perfusion lung scans can be assessed without the necessity of distinguishing large mismatched segmental defects from moderate size mismatched seg-
mental defects [3]. This makes the interpretation of ventilation–perfusion lung scans easier and more objective. The positive predictive value of PE relative to the number of mismatched large segmental perfusion defects was similar to the positive predictive value of PE relative to the number of mismatched moderate size segmental defects [3]. The positive predictive value of PE based upon the cumulative number of mismatched large segmental defects, irrespective of the number of mismatched moderate size segmental defects, is shown in Figure 68.1 and Table 68.1. The positive predictive value of PE based upon the cumulative number of mismatched moderate size segmental perfusion defects, irrespective of the number
Figure 68.1 Positive predictive value of pulmonary embolism (PE) according to the cumulative number of large segmental (SEG) perfusion defects (irrespective of the number of moderate size segmental defects) (unbroken line) and according to the number of moderate size segmental perfusion defects (irrespective of the number of large segmental defects) (broken line). (Reprinted with permission from Stein et al. [3].)
298
BLUK077-Stein
March 11, 2007
CHAPTER 68
11:28
299
Interpretation of V–Q scans according to prior cardiopulmonary disease
Table 68.1 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched large segmental perfusion defects and moderate size segmental perfusion defects. Mismatched large segmental defects
Mismatched moderate size segmental
(n = 1064)
defects (n = 1064)
Cumulative number of defects
#PE/#PTS
%PE
#PE/#PTS
%PE
≥0
383/1064
36
383/1064
36
≥1
197/251
78
205/277
74
≥2
161/191
84
119/141
84
≥3
135/153
88
89/98
91
≥4
119/133
89
64/69
93
≥5
108/117
92
48/52
92
≥6
92/99
93
40/43
93
≥7
78/82
95
36/38
95
≥8
62/66
94
35/37
95
PE, pulmonary embolism; PTS, patients. Modified and reprinted with permission from Stein et al. [3].
of mismatched large segmental perfusion defects, is also shown in Figure 68.1 and Table 68.1. Based on the similarity of the positive predictive value of large mismatched perfusion defects and moderate size mismatched perfusion defects, we defined a mismatched vascular perfusion defect as a mismatched large or moderate segmental perfusion defect. The number of mismatched vascular perfusion defects, therefore, was the number of large and/or moderate size mismatched perfusion defects [3]. The size of the mismatched defect, providing it was ≥25% of a segment, was of no consequence. The positive predictive value of the cumulative number of mismatched vascular defects (mismatched large and/or
Figure 68.2 Positive predictive value of pulmonary embolism (PE) based on the cumulative number of mismatched vascular defects (the number of mismatched large and/or moderate size segmental defects). (Reprinted with permission from Stein et al. [3].).
moderate size segments) is shown in Table 68.2 and Figure 68.2. The sensitivity and specificity of various numbers of mismatched vascular defects are shown in Table 68.3. The cumulative number of mismatched segmental equivalents perfusion defects and the cumulative number of vascular defects were of comparable diagnostic value, as indicated by similar values of the maximum likelihood estimates of the areas under the receiver operating characteristics (ROC) curves (Table 68.4, Figure 68.3). Stratification of patients according to the presence or absence of prior cardiopulmonary disease enhanced the positive predictive value of the number of
BLUK077-Stein
March 11, 2007
11:28
300
PART III
Diagnosis of acute PE
Table 68.2 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched vascular perfusion defects (large and/or moderate size segmental perfusion defects). Cumulative number
All patients
of defects
(n = 1064)
Percent
(n = 421)
No prior CPD Percent
(n = 629)
Any prior CPD Percent
≥0
383/1064
36
173/421
41
205/629
33∗
≥1
255/350
73
123/154
80
130/190
68†
≥2
200/244
82
94/106
89
105/136
77‡
≥3
170/201
85
79/87
91
90/112
80‡
≥4
149/172
87
67/75
89
81/96
84
≥5
130/144
90
55/60
92
75/84
89
≥6
116/124
94
49/52
94
67/72
93
≥7
103/110
94
43/45
96
60/65
92
≥8
93/100
93
41/43
95
52/57
91
∗
P < 0.01. P < 0.02. ‡ P < 0.05: Prior cardiopulmonary disease vs. No prior cardiopulmonary disease. These probabilities are higher than reported by Stein et al. [3] because the chi-square probabilities previously calculated used the unnecessarily conservative Yates’ correction. CPD, cardiopulmonary disease. Modified and reprinted with permission from Stein et al. [3]. †
mismatched vascular defects in each category (Table 68.2, Figure 68.4). Among patients with no prior cardiopulmonary disease, ≥1 vascular defect indicated a positive predictive value for PE of 80% and ≥2 mismatched vascular defects indicated a positive predictive value of 89% (Figure 68.5). Among patients with
prior cardiopulmonary disease, ≥1 mismatched vascular perfusion defect indicated a positive predictive value of 68% and ≥2 mismatched vascular perfusion defects indicated a positive predictive value of 77%. Tables 68.2 and 68.3 can be used to assess a positive predictive value and specificity for individual patients
Table 68.3 Sensitivity and specificity in relation to the number of mismatched vascular perfusion defects large segmental and/or moderate size segmental defects). All patients
Cumulative number of defects
SENS (%)
SPEC (%)
No prior CPD SENS (%)
SPEC (%)
Prior CPD SENS (%)
SPEC (%)
≥0
100
0
100
0
100
0
≥1
67
86
71
88
63
86
≥2
52
94
54
95
51
93
≥3
44
95
46
97
44
95
≥4
39
97
39
97
40
96
≥5
34
98
32
98
37
98
≥6
30
99
28
99
33
99
≥7
27
99
25
99
29
99
≥8
24
99
24
99
25
99
All differences between NO prior cardiopulmonary disease and Prior cardiopulmonary disease were not significant. SENS, sensitivity; SPEC, specificity; CPD, cardiopulmonary disease. Reprinted with permission from Stein et al. [3].
BLUK077-Stein
March 11, 2007
CHAPTER 68
11:28
301
Interpretation of V–Q scans according to prior cardiopulmonary disease
Table 68.4 Maximum likelihood estimates of the areas under receiver operating characteristic curves based on the number of mismatched vascular perfusion defects and the number of mismatched segmental equivalent defects. Mismatched vascular
Mismatched segmental
defects (area)
equivalents (area)
All patients (n = 1064)
0.8512
0.8530
Patients with no prior
0.8875*
0.8905†
0.8280
0.8215
cardiopulmonary disease (n = 421) Patients with any prior cardiopulmonary disease (n = 629) ∗
P = 0.051 – NO prior cardiopulmonary disease vs. Any prior cardiopulmonary disease – mismatched vascular defects. P = 0.026 – NO prior cardiopulmonary disease vs. Any prior cardiopulmonary disease – mismatched segmental equivalents. Differences between mismatched vascular defects and mismatched segmental equivalent defects were not significant. Reprinted with permission from Stein et al. [3]. †
Figure 68.3 Receiver operating characteristic curves based on the cumulative number of mismatched segmental equivalent perfusion defects and on the cumulative number of mismatched vascular perfusion defects. The curves are superimposed.
Figure 68.4 Positive predictive values of pulmonary embolism (PE) based on the cumulative number of mismatched vascular defects among patients with no prior cardiopulmonary disease (NO CPD) and among patients with any prior cardiopulmonary disease (CPD). Differences between patients with CPD and NO CPD occurred with ≥1 mismatched vascular defect (P < 0.02), ≥2 defects (P < 0.05), and ≥3 defects (P < 0.05). Probabilities are based on chi-square without Yates’ correction. (Reprinted with permission from Stein et al. [3].)
BLUK077-Stein
March 11, 2007
11:28
302
PART III
Perfusion
Posterior
Ventilation
Posterior
L LAT
RPO
Perfusion
Diagnosis of acute PE
Perfusion
Figure 68.5 Top left: Perfusion scan, posterior view, showing suggestion of perfusion defects, seen better in left lateral view (below). Top right: Ventilation scan, posterior view, washout phase, showing no abnormality. Bottom left: Perfusion scan in left lateral view showing moderate size perfusion defects (arrows) in left anterior segment of left upper lobe and anterior basal segment of left lower lobe. Bottom right: Perfusion scan, right posterior oblique (RPO) view, showing a nonsegmental perfusion abnormality associated with atelectasis (arrow). Among patients with no prior cardiopulmonary disease, 2 or more mismatched vascular defects indicated a positive predictive value for pulmonary embolism (PE) of 89%. Pulmonary embolism was shown on the pulmonary angiogram in this patient. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)
Figure 68.6 Positive predictive value of pulmonary embolism (PE) according to the cumulative number of mismatched vascular defects among patients with no prior cardiopulmonary disease (NO CPD), prior pulmonary disease (PULM DIS), and prior cardiac disease (CARD DIS). The positive predictive value of PE among patients with prior cardiac disease and those with prior pulmonary disease was similar. (Reprinted with permission from Stein et al. [3].)
BLUK077-Stein
March 11, 2007
CHAPTER 68
11:28
Interpretation of V–Q scans according to prior cardiopulmonary disease
stratified according to prior cardiopulmonary disease and according to the number of mismatched vascular perfusion defects. Patients with prior cardiac disease and those with prior pulmonary disease showed a comparable positive predictive value (Figure 68.6). There was, therefore, no advantage in stratification according to the specific prior cardiac or pulmonary disease. The number of mismatched vascular defects is as powerful for the assessment of ventilation–perfusion scans as the number of mismatched segmental equivalents. The number of mismatched vascular defects is easier to interpret, and permits a more objective evaluation [3].
303
References 1 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 2 Morrell NW, Nijran KS, Jones BE, Biggs T, Seed WA. The underestimation of segmental defect size in radionuclide lung scanning. J Nucl Med 1993; 34: 370– 374. 3 Stein PD, Henry JW, Gottschalk A. Mismatched vascular defects: an easy alternative to mismatched segmental equivalent defects for the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest 1993; 104: 1468–1471.
BLUK077-Stein
March 12, 2007
69
16:48
CHAPTER 69
The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation–perfusion lung scans
Introduction The diagnostic value of ventilation–perfusion lung scans was enhanced by showing the positive predictive value for pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects [1] (Chapter 67) where one segmental equivalent is one large segmental defect (>75% of a segment) or two moderate size segmental defects (25–75% of a segment) [2]. The positive predictive value for PE, relative to the cumulative number of mismatched vascular perfusion defects (large and/or moderate size segmental defects) was also shown [3] (Chapter 68). Interpretation of ventilation–perfusion scans on the basis of mismatched vascular defects eliminated the necessity of estimating whether a segmental defect was large or moderate in size. This facilitated the interpretation of ventilation–perfusion lung scans. The diagnostic value of ventilation–perfusion lung scans was further enhanced by combining prior clinical assessment in patients stratified according to the presence of prior cardiopulmonary disease [4]. Clinical assessment was based on intuitive judgment. By combining prior clinical assessment with stratification of patients according to the presence or absence of prior cardiopulmonary disease, a family of curves was derived which allowed an accurate assessment of the positive predictive value of PE based upon the number of mismatched segmental equivalent defects or upon the number of mismatched vascular defects [4]. The families of curves were comparable, irrespec-
304
tive of whether assessment was based upon segmental equivalent defects or vascular defects [4]. The primary advantage of the use of vascular equivalent defects, as indicated previously [3] (Chapter 68), is the ability to assess ventilation–perfusion lung scans without differentiating large segmental defects from moderate size segmental defects.
Patients with no prior cardiopulmonary disease Patients with no prior cardiopulmonary disease, irrespective of whether they were evaluated on the basis of mismatched segmental equivalent perfusion defects (Figure 69.1) or mismatched vascular perfusion defects (Figure 69.2), showed comparable families of curves. A high likelihood clinical assessment, in combination with ≥0.5 mismatched segmental equivalent perfusion defects or ≥1 mismatched vascular perfusion defect resulted in a positive predictive value for PE of 100% (Tables 69.1 and 69.2) [4]. Among patients with no prior cardiopulmonary disease and an intermediate likelihood clinical assessment, a positive predictive value for PE ≥80% was shown with ≥1 mismatched segmental equivalent perfusion defect and ≥2 mismatched vascular perfusion defects. To achieve ≥90% positive predictive value for PE, ≥4 mismatched segmental equivalent perfusion defects or ≥5 mismatched vascular perfusion defects were required (Tables 69.1 and 69.2) [4]. Among patients with no prior cardiopulmonary disease who had a low likelihood clinical assessment, ≥1
BLUK077-Stein
March 12, 2007
CHAPTER 69
16:48
Clinical assessment and prior cardiopulmonary disease for V–Q interpretation
305
Figure 69.1 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects among patients with no prior cardiopulmonary disease (NO CPD). Patients were categorized according to a high likelihood clinical assessment (CLIN HIGH), intermediate likelihood clinical assessment (CLIN INTER), or low likelihood clinical assessment (CLIN LOW). Only few patients had a low likelihood clinical assessment and ≥1 mismatched segmental perfusion defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)
mismatched segmental equivalent perfusion defects or ≥2 mismatched vascular perfusion defects resulted in a positive predictive value for PE of ≥80%. With a larger number of mismatched defects there was inconsistency of the data due to small numbers of patients (Tables 69.1 and 69.2) [4].
Patients with prior cardiopulmonary disease Among patients with a high probability clinical assessment and prior cardiopulmonary disease, ≥0.5 mis-
Figure 69.2 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched vascular perfusion defects among patients with no prior cardiopulmonary disease (NO CPD). Abbreviations as in Figure 69.1. Only few patients had a low likelihood clinical assessment and ≥1 mismatched vascular defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)
matched segmental equivalent perfusion defects and ≥1 mismatched vascular perfusion defects resulted in a positive predictive value for PE ≥88% (Figures 69.3 and 69.4, Tables 69.3 and 69.4) [4]. Among patients with an intermediate probability clinical assessment, ≥1.5 mismatched segmental equivalent defects and ≥3 mismatched vascular defects indicated a positive predictive value ≥80%. To achieve ≥90% positive predictive value for PE, ≥3.5 mismatched segmental equivalent defects or ≥5 mismatched vascular defects were required (Figures 69.3 and 69.4, Tables 69.3 and 69.4) [4].
BLUK077-Stein
March 12, 2007
16:48
Table 69.1 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched segmental equivalent defects and clinical
assessment among patients with no prior cardiopulmonary disease (n = 324).
#PE/#PTS (%)
Mismatched segmental defects
High clinical likelihood
Intermed clin likelihood
Low clinical likelihood
≥0
27/31 (87)∗
107/251 (43)
16/42 (38)†
≥0.5
19/19 (100)‡
79/101 (78)
9/15 (60)§
≥1
18/18 (100)
64/78 (82)
7/8 (88)
≥1.5
16/16 (100)
58/67 (87)
6/7 (86)
≥2
15/15 (100)
51/57 (89)
4/5 (80)
≥2.5
14/14 (100)
46/52 (88)
3/4 (75)
≥3
14/14 (100)
41/47 (87)
2/3 (67)
≥3.5
13/13 (100)
38/43 (88)
1/2 (50)
≥4
13/13 (100)
35/39 (90)
1/2 (50)
≥4.5
11/11 (100)
31/32 (97)
1/2 (50)
≥5
11/11 (100)
30/31 (97)
1/2 (50)
≥5.5
10/10 (100)
29/30 (97)
1/2 (50)
≥6
8/8 (100)
26/27 (96)
1/1 (100)
≥6.5
8/8 (100)
26/27 (96)
1/1 (100)
≥7
8/8 (100)
24/25 (96)
1/1 (100)
≥7.5
7/7 (100)
21/22 (95)
1/1 (100)
Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001. † P < 0.001. ‡ P < 0.05: High vs. intermediate clinical likelihood. § P < 0.01: High vs. low clinical likelihood. PE, pulmonary embolism; PTS, patients; Intermed, intermediate; Clin, clinical. Modified and reprinted with permission from Stein et al. [4].
Table 69.2 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched vascular defects and clinical assessment
among patients with no prior cardiopulmonary disease (n = 324).
#PE/#PTS (%)
Mismatched vascular defects
High clinical likelihood
Intermediate clinical likelihood
Low clinical likelihood
≥0
27/31 (87)∗
107/251 (43)
16/42 (38)†
≥1
19/19 (100)‡
79/101 (78)
9/15 (60)§
≥2
16/16 (100)
60/70 (86)
7/8 (88)
≥3
16/16 (100)
50/56 (89)
4/5 (80)
≥4
15/15 (100)
42/48 (88)
2/3 (67)
≥5
11/11 (100)
36/39 (92)
1/2 (50)
≥6
10/10 (100)
31/32 (97)
1/2 (50)
≥7
8/8 (100)
28/29 (97)
1/2 (50)
≥8
8/8 (100)
26/27 (96)
1/1 (100)
Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001. † P < 0.001. ‡ P < 0.05: High vs. intermediate. § P < 0.01: High vs. low. PE, pulmonary embolism; PTS, patients; CI, 95% confidence interval. Modified and reprinted with permission from Stein et al. [4].
306
BLUK077-Stein
March 12, 2007
CHAPTER 69
16:48
Clinical assessment and prior cardiopulmonary disease for V–Q interpretation
307
Figure 69.3 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent defects among patients with prior cardiopulmonary disease (ANY CPD). Abbreviations as in Figure 69.1. Only few patients had a low likelihood clinical assessment and ≥1 mismatched segmental perfusion defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)
Among patients with a low probability clinical assessment, and prior cardiopulmonary disease, the positive predictive value for PE did not exceed 33% (Figures 69.2 and 69.4, Tables 69.2 and 69.4). Few patients, however, had ≥1 mismatched segmental equivalent or ≥2 mismatched vascular defects [4]. Presentation of the data as continuous numbers of mismatched defects provided a positive predictive value for PE based upon the ventilation–perfusion scan characteristics of any particular patient. This flexibility was not possible with the criteria used for ventilation– perfusion scan interpretation data as presented in the
Figure 69.4 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched vascular defects among patients with prior cardiopulmonary disease (ANY CPD). Abbreviations as in Figure 69.1. Only few patients had a low likelihood clinical assessment and ≥1 mismatched vascular defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)
Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED), which employed fixed numbers of defects for various probabilities of PE [5]. Clinical assessment was important [4]. Among patients with no prior cardiopulmonary disease, a high likelihood clinical assessment was indicative of PE in >85% of patients, irrespective of the findings on the ventilation–perfusion lung scan. Also, a sufficiently large numbers of mismatched segmental equivalent perfusion defects or mismatched vascular perfusion defects negated a low or intermediate likelihood clinical impression.
BLUK077-Stein
March 12, 2007
16:48
Table 69.3 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched segmental equivalent perfusion defects and
#PE/#PTS (%)
Mismatched segmental defects ≥0
clinical assessment among patients with prior cardiopulmonary disease (n = 569).
High clinical likelihood ∗
24/30 (80)
Intermed clin likelihood
Low clinical likelihood
144/396 (36)
† †
13/143 (9)‡ 3/15 (20)‡ 3/11 (27)‡
≥0.5
17/19 (89)
99/139 (71)
≥1.0
16/18 (89)
86/115 (75)†
≥1.5
15/17 (88)
78/98 (80)
1/6 (17)
≥2.0
13/14 (93)
72/88 (82)
1/6 (17)
≥2.5
13/14 (93)
68/80 (85)
1/5 (20)
≥3.0
13/14 (93)
63/73 (86)
1/5 (20)
≥3.5
12/13 (92)
61/67 (91)
1/4 (25)
≥4.0
11/12 (92)
59/65 (91)
1/3 (33)
≥4.5
11/12 (92)
55/60 (92)
1/3 (33)
≥5.0
10/11 (91)
51/54 (94)
1/3 (33)
≥5.5
10/10 (100)
47/50 (94)
1/3 (33)
≥6.0
10/10 (100)
45/48 (94)
1/3 (33)
≥6.5
9/9 (100)
44/47 (94)
1/3 (33)
≥7.0
8/8 (100)
41/44 (93)
0/2 (0)
≥7.5
7/7 (100)
39/41 (95)
0/2 (0)
Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001: High vs. intermediate. † P < 0.001: Intermediate vs. low. ‡ P < 0.001: High vs. low. PE, pulmonary embolism; PTS, patients; Intermed, intermediate; Clin, clinical. Modified and reprinted with permission from Stein et al. [4].
Table 69.4 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched vascular perfusion defects and clinical
#PE/#PTS (%)
Mismatched vascular defects ≥0
assessment among patients with prior cardiopulmonary disease (n = 569).
High clinical likelihood 24/30 (80)
∗
Intermed clin likelihood 144/396 (36)
† †
Low clinical likelihood 13/143 (9)‡ 3/15 (20)‡
≥1
17/19 (89)
99/139 (71)
≥2
16/18 (89)
81/103 (79)§
2/7 (29)¶
≥3
13/14 (93)
71/86 (83)
1/6 (17)
≥4
13/14 (93)
63/72 (88)
1/5 (20)
≥5
12/13 (92)
59/64 (92)
1/4 (25)
≥6
11/11 (100)
52/55 (95)
1/3 (33)
≥7
10/10 (100)
46/49 (94)
1/3 (33)
≥8
7/7 (100)
42/45 (93)
0/2 (0)
Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001: High vs. intermediate. † P < 0.001. ‡ P < 0.001. § P < 0.01: Intermediate vs. low. ¶ P < 0.01: High vs. low. PE, pulmonary embolism; PTS, patients; Intermed, intermediate; Clin, clinical. Modified and reprinted with permission from Stein et al. [4].
308
BLUK077-Stein
March 12, 2007
CHAPTER 69
16:48
Clinical assessment and prior cardiopulmonary disease for V–Q interpretation
Among patients with prior cardiopulmonary disease, larger numbers of mismatched perfusion defects were required to negate an intermediate likelihood clinical assessment than among patients with no prior cardiopulmonary disease [4]. A low likelihood clinical assessment among patients with prior cardiopulmonary disease indicated a low positive predictive value for PE, irrespective of the number of mismatched defects, although there were few patients in this category. Prior clinical assessment in PIOPED I was made by physicians who were experienced and interested in PE. Readily obtained objective clinical information can be obtained by any physician or physician’s assistant, however, and processed with neural network computer intelligence [6]. This gives a clinical assessment which is as accurate as that of physicians experienced with PE (Chapter 52).
2
3
4
5
References 1 Stein PD, Gottschalk A, Henry JW, Shivkumar K. Stratification of patients according to prior cardiopulmonary disease and probability assessment based on the num-
6
309
ber of mismatched segmental equivalent perfusion defects: approaches to strengthen the diagnostic value of ventilation/perfusion lung scans in acute pulmonary embolism. Chest 1993; 104: 461–467. Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. Stein PD, Henry JW, Gottschalk A. Mismatched vascular defects: an easy alternative to mismatched segmental equivalent defects for the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest 1993; 104: 1468– 1471. Stein PD, Henry JW, Gottschalk A. The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest 1993; 104: 1472–1476. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. Patil S, Henry JW, Rubenfire M, Stein PD. Neural network in the clinical diagnosis of acute pulmonary embolism. Chest 1993; 104: 1685–1689.
BLUK077-Stein
March 28, 2007
70
15:50
CHAPTER 70
Single photon emission computed tomographic perfusion lung scan
Single photon emission computed tomography (SPECT) is a well established and readily available imaging method that is widely used in modern nuclear medicine diagnostics [1]. Provided that a doubleor triple-head camera is used, the acquisition time of a SPECT scan is comparable to that of planar imaging [2]. Radiologists who have used SPECT for the diagnosis of acute pulmonary embolism (PE) suggest that ventilation–perfusion imaging is significantly improved with SPECT [3]. In particular, specificity is thought to be higher with SPECT perfusion lung scans [2, 4] or ventilation–perfusion lung scans [3] compared with planar lung scans. Fewer SPECT perfusion lung scans are interpreted as nondiagnostic compared with planar perfusion lung scans [2]. There are sparse data comparing SPECT perfusion lung scans or SPECT ventilation–perfusion lung scans with planar lung scans using an independent reference standard. A nearly independent composite reference standard was used by Collart et al. [2] (Table 70.1). A SPECT perfusion lung scan may clarify the presence of an apparent stripe sign [5]. A stripe sign is the appearance of normally perfused tissue between the perfusion defect and the pleural surface suggesting a nonpleural-based opacity [6, 7]. Pulmonary embolism is thought to produce a pleural-based defect. Nonpleural-based lesions with a stripe sign are thought to be less likely due to PE [6, 7].
References 1 Reinartz P, Nowak B, Weiss C, Buell U. Acute pulmonary embolism: thin-collimation spiral CT versus planar ventilation–perfusion scintigraphy. Radiology 2004; 232: 621. 2 Collart JP, Roelants V, Vanpee D et al. Is a lung perfusion scan obtained by using single photon emission computed tomography able to improve the radionuclide diagnosis of pulmonary embolism? Nucl Med Commun 2002; 23: 1107– 1113. 3 Lemb M, Pohlabeln H. Pulmonary thromboembolism: a retrospective study on the examination of 991 patients by ventilation/perfusion SPECT using Technegas. Nuklearmedizin 2001; 40: 179–186. 4 Vanninen E, Tenhunen-Eskelinen M, Mussalo H et al. Are three-dimensional surface-shaded SPET images better than planar and coronal SPET images in the assessment of regional pulmonary perfusion? Nucl Med Commun 1997; 18: 423–430. 5 Pace WM, Goris ML. Pulmonary SPECT imaging and the stripe sign. J Nucl Med 1998; 39: 721–723. 6 Sostman HD, Gottschalk A. The stripe sign: a new sign for diagnosis of nonembolic defects on pulmonary perfusion scintigraphy. Radiology 1982; 142: 737–741. 7 Sostman HD, Gottschalk A. Prospective validation of the stripe sign in ventilation–perfusion scintigraphy. Radiology 1992; 184: 455–459.
Table 70.1 SPECT and planar perfusion lung scans in pulmonary embolism. First author
Planar sensitivity
SPECT sensitivity
Planar specificity
SPECT specificity
[Ref]
[n/N (%)]
[n/N (%)]
[n/N (%)]
[n/N (%)]
Collart [2]
12/15 (80)
12/15 (80)
43/55 (78)
53/55 (96)
SPECT, single photon emission computed tomography.
310
BLUK077-Stein
March 11, 2007
71
12:43
CHAPTER 71
Standard and augmented techniques in pulmonary angiography
Introduction Pulmonary angiography is generally accepted as the definitive diagnostic test for pulmonary embolism (PE) (Figure 71.1). Pulmonary angiography has undergone evolutionary development, both in technique and in the identification of angiographic criteria diagnostic of PE. The validity of pulmonary angiography is dependent upon technique. It is usually technically easier to diagnose PE than to confidently exclude it by pulmonary angiography.
rized the angiographic signs into those of major artery occlusion and those of peripheral vessel occlusion [4]. Obstruction of major arteries seemed to be associated with a reduced number of peripheral branches, areas of increased transradiency (oligemia), slow passage
Historical development of pulmonary angiography The first attempt to investigate the appearance of the pulmonary vasculature in the presence of PE was made by Jesser and deTakats in 1941 [1]. They described a reduction of the size of the pulmonary arterial tree and blocked main vessels after embolization of the pulmonary artery of dogs with a barium sulfate suspension. The absence of filling of the pulmonary artery beyond the site of occlusion was also observed by Liberson and Liberson [2] in 1942, in pulmonary angiograms taken after the injection into rabbits of lead fillings mixed with paraffin. The intraluminal filling defect in PE was first described by Lochhead, Roberts, and Dotter in 1952 in dogs embolized with autologous blood clots [3]. They also described incomplete occlusion of vessels, reduced rate of flow through the region of the clot, diminished opacification of distal vessels, and dilation of the pulmonary artery proximal to the embolus. Chrispin, Goodwin, and Steiner, in 1963, based on some right atrial and some pulmonary arterial injections in 7 patients with clinical evidence of PE, catego-
Figure 71.1 Pulmonary angiogram showing intraluminal filling defects (arrows) in left pulmonary artery and its branches.
311
BLUK077-Stein
March 11, 2007
12:43
312
PART III
Diagnosis of acute PE
of contrast material into the capillaries (asymmetrical filling), and failure of the pulmonary veins to fill. Obstruction of smaller vessels was associated with slow disappearance of contrast material from the arteries (prolonged arterial phase), peripheral pruning, and tortuous vessels. Williams and associates in 1963, in addition to previous observations, reported narrowing and irregularity of opacified vessels, which they attributed to circumferential organization of thromboemboli [5]. Smith, Dammin, and Dexter, utilizing postmortem arteriography, showed reduced opacification of small vessels in the periphery of the lung when small vessels were embolized [6]. Sasahara and associates mentioned plaque-like filling defects among the angiographic abnormalities noted in patients with clinical evidence of PE [7]. They noted difficulty of interpretation of pulmonary angiograms in patients with mitral stenosis or chronic lung disease. Dalen et al. [8] obtained selective pulmonary angiograms in dogs embolized with autologous clots (some with radiopaque autologous clots) and observed cutoffs, areas of oligemia, and filling defects to be typical angiographic abnormalities due to pulmonary embolization. Stein and associates divided angiographic signs into those of morphologic significance and those of physiologic significance [9]. Morphologic signs were intraluminal filling defects, abrupt cutoff of an artery, and localized pruning (lack of fine branching). Physiologic signs were oligemia (areas of increased translucency), asymmetrical filling, prolongation of the arterial phase, and bilateral lower-zone filling delay. Absence of fine branching of the pulmonary vessels was subsequently shown to be nonspecific, and may reflect pulmonary vascular destruction such as occurs in emphysema [10, 11]. Abrupt cutoff of vessels may occur with tumor. It became apparent, therefore, that the only abnormality on the pulmonary angiogram that was diagnostic of PE was visualization of the intraluminal filling defect. Other angiographic abnormalities, in a patient with suspected PE who has no prior cardiopulmonary disease, may call attention to suspicious regions.
niques. Pulmonary digital subtraction angiography (DSA) was used in PIOPED II [13–15]. In PIOPED I, techniques of pulmonary angiography were similar to PIOPED II, although cut films were used as was ionic contrast material. The femoral-vein Seldinger technique with a multiple side-holed 4 French size to 8 French size pigtail catheter was used [16]. Small amounts of contrast material (5–8 mL) were injected by hand to check the patency of the inferior vena cava. The catheter was directed into the proximal portion of the pulmonary artery of the lung with the greatest abnormality on the ventilation–perfusion lung scan. The left lung was imaged in the 50◦ right anterior oblique (RAO) and 40◦ left anterior oblique (LAO) projection. The right lung was imaged in the 30◦ RAO and 40◦ LAO projections. Field size was the maximum magnification that allowed visualization of the entire lung in both views. The highest matrix (1024 × 1024 when possible) was used to allow imaging at 6 frames/sec. Low or iso-osmolar 60% contrast material was injected at 20–35 mL/sec for a total of 25–50 mL [12]. If emboli were not identified, superselective injections were made in areas of suspected abnormality. Magnified and coned images were obtained in the same projections [17]. The amount of contrast material for superselective injections was 5–30 mL injected at a rate of 5–20 mL/sec. If no emboli were found in the first lung, or if bilateral angiography in the clinical center was routine, identical techniques were used for the second lung.
Angiographic technique
Criteria for interpretation of acute PE were a partially occlusive filling defect within an arterial branch or a completely occlusive filling defect that leads to termination of the column of contrast agent in a meniscus that outlines the trailing edge of the embolus [21]. A
The techniques used in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED II) [12] may serve as a guide for most recent accepted tech-
Techniques that augment pulmonary DSA Techniques that augment DSA might be employed to reduce the volume of contrast material that is needed and enhance the visualization of small pulmonary emboli in distal branches of the pulmonary artery. Such techniques include cineangiography [18], balloonocclusion cineangiography [14, 19], and wedge arteriography [20] (Figures 71.2–71.4). These techniques are not routinely used, however.
Interpretation of pulmonary DSA
BLUK077-Stein
March 11, 2007
CHAPTER 71
12:43
Pulmonary angiography
313
Figure 71.2 Normal pulmonary wedge arteriogram. Vessels show narrow gradual tapering and numerous fine branches. A background blush of capillary filling and veins draining the segment are shown. Catheter diameter was 2.3 mm. (Reprinted from Stein [20], with permission from Elsevier.)
nonfilling vessel, or hypovascular area were not considered diagnostic, nor were webs, stenoses, or other vessel abnormalities. Criteria for chronic PE are termination of a vessel without a meniscus, mural irregularity suggestive of recanalization, webs, or synechiae [22], as well as focal stenosis and focal calcification.
Quality of pulmonary angiography Digital subtraction angiography was evaluated during the injection of contrast material through a flow
directed balloon catheter [21]. Selective or subselective pulmonary angiograms were obtained during the injection of 20–30 mL of iodinated contrast material at 10–15 mL/sec. Among 211 patients, image quality was excellent in 61.1%, adequate in 37.4%, and poor in 1.4%. The catheter employed was analogous to a Swan-Ganz catheter. The average amount of contrast material was 65 mL. The average duration of the examination was 28 minutes. In PIOPED I, 12 of 1111 (1%) of angiograms were not completed, usually because of complications [23]. Among completed angiograms, 35 of 1099 (3%) were
BLUK077-Stein
March 11, 2007
12:43
314
PART III
Diagnosis of acute PE
Figure 71.3 Wedge arteriogram showing a completely obstructed artery just distal to the catheter (arrow). The diameter of the occluded artery is approximately 2.0 mm. Contrast material has been forced in a retrograde direction due to the obstruction, and is seen along the side of the catheter (arrow). (Reprinted from Stein [20], with permission from Elsevier.)
nondiagnostic. This percentage of nondiagnostic angiograms is similar to the 5% of poor angiograms reported by Dalen and associates following an injection in the main pulmonary artery [24]. The angiographic interpretations that PIOPED I reported, however, were consensus readings, adjudicated if necessary by a panel of readers. Individual angiographers in a clinical environment may not achieve the accuracy of these consensus readings.
Outcome in patients with normal pulmonary angiograms Among the 681 patients in PIOPED I, with negative pulmonary angiograms, the diagnosis of no PE was reversed by the Outcome Classification Committee in 4 (1%) [16]. Each of these patients died within 6 days
of the pulmonary angiogram, and PE was found in each at autopsy. The validity of positive pulmonary angiograms could not be assessed. Pulmonary embolism occurred within 1 year in 6 of 380 (1.6%) patients with suspected PE and normal pulmonary angiograms who received no anticoagulants [25].
Reader agreement Overall agreement on all three categories of interpretation (both readers agreed PE was present, PE absent, or PE uncertain) was 81% [23]. There was closer agreement on the presence of PE than on the absence of PE. Both agreed PE was present or both agreed that PE could not be diagnosed with certainty in 92%. Both readers agreed that PE was absent or both agreed that PE could not be excluded with certainty in 82%.
BLUK077-Stein
March 11, 2007
CHAPTER 71
12:43
315
Pulmonary angiography
Figure 71.4 Wedge arteriogram showing intraluminal filling defects in arteries 1.5–2.0 mm in diameter (arrows). (Reprinted from Stein [20], with permission from Elsevier)
The quality of the angiograms had a greater impact upon the agreement on negativity than on positivity [23]. Agreement on positivity with good, fair, and poor quality angiograms was 93, 90, and 98%, respectively. Agreement on negativity was 88, 77, and 54%, respectively [23].
Pulmonary angiograms showed PE in main or lobar pulmonary arteries as well as smaller orders of branches in 217 of 375 (58%) patients with PE. Among these patients, angiographic readers showed an average co-positivity of 98% [26] (Table 71.1). Average co-positivity was the average
Table 71.1 Pulmonary emboli in main or lobar arteries. Reader 1 Reader 2
PE+
PE−
PE?
Total
PE+
207
2
0
209
PE−
3
0
1
4
PE?
3
1
0
4
213
3
1
217
Total
Data are numbers of patients. Average co-positivity = 1/2 (207/213 + 207/209) × 100 = 98%. PE−, pulmonary embolism absent; PE+, pulmonary embolism present; and PE?, pulmonary embolism uncertain. Reproduced from Stein et al. [26], with permission.
BLUK077-Stein
March 11, 2007
12:43
316
PART III
Diagnosis of acute PE
Table 71.2 Largest pulmonary emboli in segmental arteries. Reader 1 Reader 2
PE+
PE−
PE?
Total
PE+
110
6
3
119
PE−
11
0
1
12
PE?
5
0
0
5
126
6
4
136
Total
Data are numbers of patients. Average co-positivity = 1/2 (110/126 + 110/119) × 100 = 90%. PE−, pulmonary embolism absent; PE+, pulmonary embolism present; and PE?, pulmonary embolism uncertain. Reproduced from Stein et al. [26], with permission.
In PIOPED II, the unweighted kappa statistic [27] for DSA was 0.66 [28].
of the agreement on positivity of reader 1 with reader 2 and reader 2 with reader 1, expressed as percent. Pulmonary angiograms in which the largest order of arteries with PE were segmental branches were shown in 136 of 375 (36%) patients with PE. These patients may have had PE in subsegmental branches as well. Among these patients, average copositivity was 90% [26] (Table 71.2). Angiograms of only 22 of 375 (6%) showed PE limited to subsegmental pulmonary arteries. Readers of these angiograms showed an average co-positivity of 66% (Table 71.3). The angiographers agreed with themselves in 64 of 72 (89%) [26]. (See Chapter 72 for discussion of prevalence in proximal and peripheral pulmonary arteries)
Digital subtraction pulmonary angiography: peripheral injection Peripheral injection of an average of 150 mL of contrast material (range 120–220 mL) in patients with suspected acute PE was compared with conventional pulmonary angiography [29]. The peripheral DSA was satisfactory in 41 of 54 patients (76%). Among patients in whom the conventional pulmonary angiogram showed PE, 26 of 37 (70%) showed PE on peripheral DSA. If the PE was large (Miller index >11) the peripheral DSA showed PE in 16 of 20 (80%). Among patients in whom conventional pulmonary angiography was negative, peripheral DSA was negative in 7 of 15 (47%).
Table 71.3 Largest pulmonary emboli in subsegmental arteries. Reader 1 Reader 2
PE+
PE−
PE?
Total 16
PE+
9
7
0
PE−
2
0
3
5
PE?
1
0
0
1
12
7
3
22
Total
Data are numbers of patients. Average co-positivity = 1/2 (9/12 + 9/16) × 100 = 66%. PE−, pulmonary embolism absent; PE+, pulmonary embolism present; and PE?, pulmonary embolism uncertain. Reproduced from Stein et al. [26], with permission.
BLUK077-Stein
March 11, 2007
CHAPTER 71
12:43
317
Pulmonary angiography
References 1 Jesser JH, de Takats G. Visualization of the pulmonary artery during its embolic obstruction. Arch Surg 1941; 42: 1034–1041. 2 Liberson F, Liberson IR. The use of Diodrast in determining the location and extent of pulmonary embolism: an experimental study. Am J Roetgenol 1942; 48: 352–355. 3 Lochhead RP, Roberts DJ, Jr, Dotter CT. Pulmonary embolism, experimental and angiocardiographic study. Am J Roentgenol 1952; 68: 627–633. 4 Chrispin AR, Goodwin JF, Steiner RE. The radiology of obliterative pulmonary hypertension and thromboembolism. Br J Radiol 1963; 36: 705–714. 5 Williams JR, Wilcox C, Andrews GJ, Burns RR. Angiography in pulmonary embolism. JAMA 1963; 184: 473–476. 6 Smith GT, Dammin GJ, Dexter L. Postmortem arteriographic studies of the human lung in pulmonary embolization. JAMA 1964; 188: 143–151. 7 Sasahara AA, Stein M, Simon M, Littmann D. Pulmonary angiography in the diagnosis of thromboembolic disease. N Engl J Med 1964; 270: 1075–1081. 8 Dalen JE, Mathur VS, Evans H et al. Pulmonary angiography in experimental pulmonary embolism. Am Heart J 1966; 72: 509–520. 9 Stein PD, O’Connor JF, Dalen JE et al. The angiographic diagnosis of acute pulmonary embolism: evaluation of criteria. Am Heart J 1967; 73: 730–741. 10 Stein PD, Leu JD, Welch MH, Guenter CA. Pathophysiology of the pulmonary circulation in emphysema associated with alpha1 antitrypsin deficiency. Circulation 1971; 43: 227–239. 11 Jacobson G, Turner AF, Balchum OJ et al. Pulmonary arteriovenus shunts in emphysema demonstrated by wedge arteriography. Am J Roentgenol 1965; 93: 868–878. 12 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182. 13 Goodman PC, Brant-Zawadzki M. Digital subtraction pulmonary angiography. Am J Roentgenol 1982; 139: 305– 309. 14 Ferris EJ, Holder JC, Lim WN et al. Angiography of pulmonary emboli: digital studies and balloon occlusion cineangiography. Am J Roentgenology 1984; 142: 369–373. 15 van Rooij W-JJ, den Heeten GJ, Sluzewski M. Pulmonary embolism: diagnosis in 211 patients with use of selective
16
17 18
19
20
21
22
23
24
25
26
27 28 29
pulmonary digital subtraction angiography with a flowdirected catheter. Radiology 1995; 195: 793–797. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. Bookstein JJ. Segmental arteriography in pulmonary embolism. Radiology 1969; 93: 1007–1012. Meister SG, Brooks HL, Szucs MM, Banas JS, Jr, Dexter L, Dalen JE. Pulmonary cineangiography in acute pulmonary embolism. Am Heart J 1972; 84: 33–37. Wilson JE, III, Bynum LJ. An improved pulmonary angiographic technique using a balloon-tipped catheter. Am Rev Respir Dis 1976; 114: 1137–1144. Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623. Dotter CT. Acquired Abnormalities of the Pulmonary Arteries in Abrams HL, 3rd edn. Little, Brown, Boston, MA, 1983: 743–761. Nicod P, Peterson K, Levine M et al. Pulmonary angiography in severe chronic pulmonary hypertension. Ann Intern Med 1987; 107: 565–568. Stein PD, Athanasoulis C, Alavi A et al. Complications and validity of pulmonary angiography in acute pulmonary embolism. Circulation 1992; 85: 462–468. Dalen JE, Brooks HL, Johnson LW, Meister SG, Szucs MM, Jr, Dexter L. Pulmonary angiography in acute pulmonary embolism: indications, techniques and results in 367 patients. Am Heart J 1971; 81: 175–185. Henry JW, Relyea B, Stein PD. Continuing risk of thromboemboli among patients with normal pulmonary angiograms. Chest 1995; 107: 1375–1378. Stein PD, Henry JW, Gottschalk A. Reassessment of pulmonary angiography for the diagnosis of pulmonary embolism: relation of interpreter agreement to the order of the involved pulmonary arterial branch. Radiology 1999; 210: 689–691. Landis RJ, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159–174. Stein PD, Goodman LR, Sostman HD. Response to letter. N Engl J Med 2006; 355: 955–956. Musset D, Rosso J, Petitpretz P et al. Acute pulmonary embolism: diagnostic value of digital subtraction angiography. Radiology 1988; 166: 455–459.
BLUK077-Stein
March 28, 2007
72
15:53
CHAPTER 72
Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries
The prevalence of pulmonary embolism (PE) limited to subsegmental pulmonary arteries was evaluated with the pulmonary angiograms obtained in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) [1]. Among all patients with PE, 22 of 375 (6%) had PE limited to subsegmental branches of the pulmonary artery, 136 of 375 (36%) showed PE in segmental as pulmonary arterial branches and 217 of 375 (58%) showed PE in main or lobar pulmonary arteries [1] (Figure 72.1). Therefore, 94% of patients with PE shown by pulmonary angiography had the PE in segmental branches or larger. The prevalence that we observed of PE limited to subsegmental pulmonary arteries (6%) was lower than reported by Oser and associates, 23 of 76 (30%) and by Goodman and associates, 4 of 11 (36%), but was comparable to the frequency reported by Quinn and associates, 2 of 20
(10%) [2–4]. Goodman and associates also reported a higher prevalence of low probability interpretations of the V–Q scan among patients with PE, 5 of 11 (45%) than observed in PIOPED, 59 of 375 (16%) [2]. The patients in PIOPED, therefore, may have had more severe PE, than the patients reported by Goodman and associates. In patients with intermediate probability V–Q scans, the PE was limited to subsegmental pulmonary arteries in 9 of 160 (6%) [1]. In patients with low probability V–Q scans, the PE was limited to subsegmental pulmonary arteries in 10 of 59 (17%) [1]. Among patients with no prior cardiopulmonary disease who had low probability V–Q scans and PE, the prevalence of PE limited to subsegmental pulmonary arteries was higher, 7 of 23 (30%) [1].
Largest artery showing PE (%)
References
70 60 50 40 30 20 10 0
58 36
6 Main or lobar
Segmental
Subsegmental
Figure 72.1 Largest pulmonary artery showing pulmonary embolism (PE) among 375 patients with PE on pulmonary angiogram. (Data from Stein and Henry [1].)
318
1 Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/perfusion lung scans. Chest 1997; 111: 1246–1248. 2 Goodman LR, Curtin JJ, Mewissen MW et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. Am J Roentgenol 1995; 164: 1369–1374. 3 Oser RF, Zuckerman DA, Gutierrez FR, Brink JA. Anatomic distribution of pulmonary emboli at pulmonary angiography: implications for cross-sectional imaging. Radiology 1996; 199: 31–35. 4 Quinn MF, Lundell CJ, Klotz TA et al. Reliability of selective pulmonary arteriography in the diagnosis of pulmonary embolism. AJR 1987; 149: 469–471.
BLUK077-Stein
March 11, 2007
73
12:50
CHAPTER 73
Quantification of pulmonary emboli by conventional and CT angiography
CT angiography index Qanadli and associates described an index for quantifying the severity of pulmonary emboli (PE) by CT angiography [1]. The index is defined as the sum of the products of N × D, where N is the value of the proximal clot site equal to the number of segmental branches arising distally, and D is the degree of obstruction defined as 1 for any amount of partial obstruction (<100%) and 2 for total obstruction. For example, the presence of a PE in a segmental artery is scored as 1 point [2]. This value is multiplied by a factor D, defined as 1 for a partially occlusive PE and 2 for a completely occlusive PE [2]. A PE in a lower lobe artery that gives rise to 5 segmental pulmonary arteries would be scored as N = 5. The extent of the lower lobe artery occlusion would be graded as if each of the segmental arteries from it had the same degree of obstruction as the parent lobar artery. If the lower lobe artery was partially occluded, the score would be 5 × 1 = 5. If the lower lobe artery was totally occluded, the score would be 5 × 2 = 10. Each lung is regarded as having 10 segmental branches: 3 to the upper lobe, 2 to the middle lobe and lingual, and 5 to the lower lobe [2]. Accordingly, the maximum obstruction score for each lung is 20 and for each patient is 40 [2]. The score is converted into a percentage [2].
Prognosis based on CT angiography PE index Preliminary observations with the PE index described by Qanadli and associates [1] showed that a cutoff value of 60% was useful for prediction of survival from acute PE [2]. Among patients with a PE index ≥60%, only 1 of 6 (17%) survived [2]. Among patients with a PE
index <60%, 52 of 53 (98%) survived. The surviving patient with a PE index >60% received thrombolytic therapy. Another patient with a PE index of 92.5% died, even with thrombolytic therapy [2].
Angiographic index: Walsh index The “Walsh” angiographic index is an objective index based on PE shown by conventional pulmonary angiography [3]. The index reflects the number of segmental vessels potentially or actually underperfused because of PE in segmental or larger vessels. The maximum value of the index is 18, each unit representing the equivalent of total or partial occlusion of a segmental pulmonary arterial branch. Each definite total occlusion of a pulmonary vessel receives a score as shown in Table 73.1. Each vessel showing a definite partial occlusion (intraluminal filling defect) receives a score as shown in Table 73.2. A single filling defect that extends into >1 anatomical location receives a score for each location up to but not exceeding the maximum designated for each region. Table 73.1 Score for vessel showing total occlusion. Score for each total Vessel involved
occlusion
Main pulmonary artery
9
Intermediate artery
6
Lobar artery Upper
3
Middle (or lingular)
2
Lower
4
Segmental
1
Data from Walsh et al. [3].
319
BLUK077-Stein
March 11, 2007
12:50
320
PART III
Table 73.2 Score for vessels showing a definite partial occlusion (definite intraluminal filling defect). Score for a filling Anatomical location
defect
Central region (pulmonary
3
Diagnosis of acute PE
scale: normal = 0, mild = 1, moderate = 2, or severe = 3 [3]. The objective angiographic score (Walsh index) and the subjective index showed a close correlation (r = 0.86) [3]. Subjective scores were the means of the readings by three radiologists with an interest and experience in PE.
trunk, main pulmonary arteries, intermediate vessel) Lobar or lingular
2
Segmental
1
Data from Walsh et al. [3].
Abnormalities of a single upper lobe region may receive a score of ≤3, middle lobe or lingual ≤2, and lower lobe ≤4. Obstructions in central anatomical regions (primary trunk, main pulmonary arteries, and intermediate vessels) receive scores according to the vessel involved. If the total score of the lung in question is >4 without considering filling defects in the central region, the central filling defects are ignored. All filling defects in a single central region, whether single or multiple, receive a score of 3, if they are considered. If a single vessel contains both a filling defect and an obstruction, only the obstruction is scored. The sum total of scores for all abnormalities in one lung must not exceed 9. Computing score: Score 1 for each definite segmental obstruction or filling defect. Score each obstruction or filling defect in each lobar (or lingular) region and add to it the scores for segmental abnormalities up to the maximum for the lobe or region in question. Score obstructions in each central region according to the listed criteria and add to it the scores for lobar abnormalities up to a maximum of 9 for each lung. If the total score for a lung is less than 5 at this point, increase the score by 3 if any central region filling defects are present.
Subjective evaluation Subjective evaluation of the angiographic severity of PE was assessed on the basis of a four-point subjective
Miller index The Miller index is composed of an objective score for arterial obstruction and a subjectively determined score for reduction of peripheral perfusion [4]. The right pulmonary artery is assigned 9 segmental arteries: 3 in the right upper lobe, 2 in the right middle lobe, and 4 in the right lower lobe. The left lung is assigned seven segmental arteries: 2 in the left upper lobe, 2 to the lingual, and 3 in the left lower lobe. Partial or complete occlusion of a segmental artery receives a point score of 1. Proximal PE are scored equal to the number of segmental arteries arising distally according to the anatomic subdivisions. The maximal score for obstructions is 16. Reduction of peripheral perfusion is scored by dividing each lung into upper, middle, and lower zones and using a four-point scale: normal perfusion = 0, moderate reduction = 1, severe reduction = 2, and absent perfusion = 3. Maximal score of reduced perfusion in both lungs is 18. The maximal Miller index for both lungs is 34.
References 1 Qanadli SD, El Hajjam M, Vieillard-Baron A et al. New CT index to quantify arterial obstruction in pulmonary embolism: comparison with angiographic index and echocardiography. AJR Am J Roentgenol 2001; 176: 1415– 1420. 2 Wu AS, Pezzullo JA, Cronan JJ, Hou DD, Mayo-Smith WW. CT pulmonary angiography: quantification of pulmonary embolus as a predictor of patient outcome—initial experience. Radiology 2004; 230: 831–835. 3 Walsh PN, Greenspan RH, Simon M et al. An angiographic severity index for pulmonary embolism. Circulation 1973; 47/48(suppl II): 101–108. 4 Miller GA, Sutton GC, Kerr IH, Gibson RV, Honey M. Comparison of streptokinase and heparin in treatment of isolated acute massive pulmonary embolism. BMJ 1971; 2: 681–684.
BLUK077-Stein
March 11, 2007
74
13:30
CHAPTER 74
Complications of pulmonary angiography
Introduction The proportion of patients with pulmonary embolism (PE) who are diagnosed by conventional pulmonary angiography was relatively constant at about 10–20% from 1979 to 2003 [1]. Conventional pulmonary angiography still is the diagnostic reference standard for PE. However, because of the invasive nature of pulmonary angiography, imaging techniques that are relatively noninvasive are generally recommended [2]. Complications of pulmonary angiography have been reported by several groups of investigators. Mills and associates reported major complications (including death) in 43 of 1350 (3%) [3]. Three deaths occurred (0.2%), and each was in a patient with cor pulmonale, elevated right ventricular end-diastolic pressure, and elevated pulmonary artery pressure. Nonfatal major complications among these 1350 patients included cardiac perforation in 14 (1%), major arrhythmias in 11 (1%), successfully treated cardiac arrest in 6 (0.4%), myocardial injury in 6 (0.4%), and significant contrast reaction in 4 (0.3%). In the Urokinase Pulmonary Embolism Trial, among 310 patients with PE, 6 of 310 (2%) patients had nonfatal major complications, exclusive of local complications [4]. There were no deaths. Major arrhythmias occurred in 5 and myocardial perforation in 1 patient. Dalen and associates reported significant complications in 13 of 367 (4%) who underwent angiography because of suspected PE [5]. There was 1 death (0.3%). Hull and associates reported 2 reactions to contrast media, but no deaths among 104 patients [6]. Marsh and associates reported 1 death and 1 nonfatal perforation of the right ventricle among 106 patients [7]. Moses and associates reported major complications of pulmonary angiography in 3 of 298 (1%) patients who had suspected PE [8]. There were 2 deaths and
1 major arrhythmia. Stein and associates reported nonlocal complications in 12 of 122 (10%) patients with suspected PE or other pulmonary disorders [9]. Two remote deaths occurred, which probably were not a direct result of pulmonary arteriography. Ranniger reported nonlocal complications in 8 of 241 (3%) [10]. There were no deaths. Novelline and associates reported no major complications or deaths from pulmonary angiography in 302 patients and only minor arrhythmias in 3 (1%) [11]. Complications of pulmonary angiography were reported in 1111 patients who underwent angiography in the Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) [12]. Major complications were defined as life-threatening, or did not respond promptly to pharmaceutical therapy, or required intensive or prolonged treatment within the hospital. Patients who required cardiopulmonary resuscitation, endotracheal intubation, dialysis, or blood transfusion were defined as having major complications. Only one complication (the most severe complication) was listed for any patient. Nonmajor or minor complications were defined as complications that spontaneously regressed with no apparent residual damage even though prolonged monitoring may have been required. Such complications, in some instances, were important and potentially dangerous events and may have prolonged the hospital stay. The techniques of pulmonary angiography that were employed are relevant. Pulmonary angiograms were performed through a femoral vein using the Seldinger technique with 6–8 French catheters (outside diameter 2–2.7 mm) [13]. Small amounts of contrast material (5–8 mL) were injected by hand to check the patency of the inferior vena cava by fluoroscopy. Seventy-six percent iodinated (ionic) contrast material was used.
321
BLUK077-Stein
March 11, 2007
13:30
322
PART III
Table 74.1 Major complications of angiography among 1111 patients with suspected acute pulmonary embolism. Complication
#PTS (%)
Death
5 (0.5)
Respiratory distress (CPR or intubation)
4 (0.4)
Renal failure (dialysis)
3 (0.3)
Hematoma (transfusion 2 units)
2 (0.1)
Total
14 (1.3)
Only one complication is listed for each patient. CPR, cardiopulmonary resuscitation; PTS, patients. Modified and reproduced from Stein et al. [12], with permission.
Major complications Major complications, including death, occurred in 1.3% of the patients (Table 74.1) [12]. Most of the patients who died or suffered major nonfatal cardiopulmonary complications were in critical condition with severely compromised cardiopulmonary function prior to pulmonary angiography. The severity of the clinical condition usually was not due to massive PE. Death occurred in 0.5% [12]. Although 5 deaths were attributed to pulmonary angiography, the cause of death in 3 patients may not have been due to the
Diagnosis of acute PE
catheterization procedure or pulmonary angiography. At least one of the patients was practically moribund and 4 others had severe underlying cardiac or respiratory disease. The frequency of death (0.5%) was similar to the pooled value reported by others, 9 of 3074 (0.3%) [4–12]. The circumstances associated with respiratory decompensation in 0.4% of the patients who required endotracheal intubation or cardiopulmonary resuscitation are outlined in Table 74.2. Hematoma of the groin with bleeding severe enough to require two units of blood was observed in 0.2% of the patients. One of the two patients who had such bleeding was receiving anticoagulant therapy. Three patients (0.3%) following angiography developed renal dysfunction, presumably acute tubular necrosis, which required dialysis. Serum creatinine in one patient increased from 1.2 mg/100 mL to 6.0 mg/100 mL and in one patient it increased from 2.4 mg/100 mL to 6.0 mg/100 mL. One patient developed a serum creatinine of 2.7 mg/100 mL, which was associated with pulmonary edema due to fluid overload. The patient required endotracheal intubation. In addition to the patients with major renal complications, 10 patients (0.9%) developed an elevation of the serum creatinine and were treated with fluid balance, diuretics, and in one patient, dopamine. These
Table 74.2 Nonmajor or minor complications of angiography among 1111 patients with suspected acute pulmonary embolism. Complication Respiratory distress (prompt response to drugs) Renal dysfunction (responded to drug therapy and fluid balance)
#PTS (%) 4 (0.4) 10 (0.9)
Angina (monitored in coronary care unit)
2 (0.2)
Hypotension (prompt response to drugs/fluids)
2 (0.2)
Pulmonary congestion (prompt response to drugs) Urticaria, itching, or periorbital edema
4 (0.4) 16 (1.4)
Hematoma (not transfused)
9 (0.8)
Arrhythmia (spontaneous conversion or prompt response to drugs)
6 (0.5)
Subintimal contrast Stain
4 (0.4)
Narcotic overdose (treated with naloxone)
1 (0.1)
Nausea and vomiting
1 (0.1)
Right bundle branch block
1 (0.1)
Total Only one complication is listed for each patient. PTS, patients. Modified and reproduced from Stein et al. [12], with permission.
60 (5.4)
BLUK077-Stein
March 11, 2007
CHAPTER 74
13:30
Complications of pulmonary angiography
patients did not require dialysis and the complications, therefore, were defined as minor. Among these patients, 5 with a previously normal serum creatinine developed levels that ranged from 2.1 to 5.9 mg/100 mL. Five patients with a previously abnormal serum creatinine which ranged from 1.5 to 2.7 mg/100 mL developed a further increase of ≥2 mg/100 mL. Renal insufficiency induced by contrast material was a problem, particularly among elderly patients [12, 14]. All pulmonary angiograms were performed with a standard ionic, high-osmolar contrast agent [13]. Whether the frequency of renal insufficiency would have been lower following the use of a lowosmolar contrast agent is undetermined. There is no evidence that the incidence of pronounced renal toxicity is different following the intravascular use of highversus low-osmolar contrast agents [15–17]. Also, the potential clinical advantages in regard to cardiac function have not yet been clearly demonstrated [18, 19].
Nonmajor or minor complications Among the patients with less severe or minor complications, many of the complications raised serious concern, some were potentially life-threatening and some required prolonged monitoring. Minor complications, therefore, were not necessarily trivial (Table 74.2). Two patients had bronchospasm following the injection of contrast material that was severe enough to discontinue the procedure. One required subcutaneous epinephrine. Two patients developed angina during the catheterization and were transferred to the coronary care unit for monitoring. Two patients (in addition to 1 patient whose most prominent problem was angina) became hypotensive.
Relation of complications to clinical characteristics Major complications occurred more frequently among patients sent for angiography from the medical intensive care unit than patients from elsewhere, 5 of 122 (4%) versus 9 of 989 (1%) [12]. The severity of the underlying condition rendered these patients vulnerable to the procedure. Such critically ill patients often present the most vexing diagnostic dilemmas which require precise information for their optimal management, but the risk of angiography is higher in these
323
patients. Clinical judgment is important in this situation. Minor complications occurred with a similar frequency among patients sent from the medical intensive care unit and from elsewhere, 4 of 122 (3%) versus 56 of 989 (6%). The frequency of complications was not related to the presence or absence of PE. Only 4 of 11 patients with major complications, in whom a diagnosis was established, had PE. Major complications occurred in 1% of the patients with PE and in 1% of the patients who did not have PE. Minor complications occurred in 6% of the patients with PE and in 5% of the patients with no PE. Complications, in general, were not related to age, although renal complications occurred more often in elderly patients [12, 14]. Patients with either major or less severe renal dysfunction, in comparison to patients with no renal dysfunction, were 74 ± 13 years versus 57 ± 17 years (mean ± standard deviation). Complications were unrelated to gender. The frequency of complications was not related to pulmonary artery mean pressure [12]. Pulmonary artery mean pressure was measured in patients with major, minor, and no complications and it was 22 ± 14, 19 ± 9, and 23 ± 11 mm Hg, respectively. Patients with major or minor complications had right atrial mean pressures that were no higher than in patients with no complications. The volume of contrast material injected was not significantly larger in patients with major or minor complications than in patients with no complications, 106 ± 73, 185 ± 84, and 181 ± 61 mL, respectively [12]. Even among patients who suffered renal dysfunction, either major or minor (n = 13), the volume of contrast material injected was not significantly greater than among those who had no complications 207 ± 53 mL versus 181 ± 61 mL. The potential complication of myocardial perforation was specifically monitored and none was observed. The absence of myocardial perforation reflects the use of angiographic catheters with a “pigtail” shaped tip. Such catheters also contributed to the reduced incidence of major cardiac arrhythmias occurring during the passage of the catheter through the right ventricle. The risk of pulmonary hypertension has been emphasized by previous investigators [3], but procedural modifications diminished the risk in PIOPED. In the past, most deaths and serious complications were
BLUK077-Stein
March 11, 2007
13:30
324
reported in patients with pulmonary hypertension and an elevated right ventricular end-diastolic pressure [3]. In this setting, an injection of contrast material into the main pulmonary artery may result in death. In PIOPED, the protocol called for selective contrast injections into the right or left pulmonary artery, rather than the main pulmonary artery. Further, in patients with elevated pulmonary artery pressures, the protocol permitted a reduction of the flow rate and the total amount of contrast injected. Such precautions, however, cannot be viewed as a guarantee for the prevention of serious or fatal complications in patients with pulmonary hypertension. Marsh and associates [7] reported death in a patient with severe pulmonary hypertension after a hand injection of 10 mL of contrast material.
PART III
9
10
11
12
13
References 1 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316–1317. 2 Stein PD, Woodard PK,Weg JG et al. Diagnostic pathways in acute pulmonary embolism: recommendations of the PIOPED II Investigators. Am J Med, 2006; 119: 1048– 1055. 3 Mills SR, Jackson DC, Older RA, Heaston DK, Moore AV. The incidence, etiologies, and avoidance of complications of pulmonary angiography in a large series. Radiology 1980; 136: 295–299. 4 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Circulation 1973; 47(suppl II): 38–45. 5 Dalen JE, Brooks HL, Johnson LW, Meister SG, Szucs MM, Jr, Dexter L. Pulmonary angiography in acute pulmonary embolism: indications, techniques, and results in 367 patients. Am Heart J 1971; 81: 175–185. 6 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. 7 Marsh JD, Glynn M, Torman HA. Pulmonary angiography: application in a new spectrum of patients. Am J Med 1983; 75: 763–770. 8 Moses DC, Silver TM, Bookstein JJ. The complementary roles of chest radiography, lung scanning, and selective
14
15
16
17
18
19
Diagnosis of acute PE
pulmonary angiography in the diagnosis of pulmonary embolism. Circulation 1974; 49: 179–188. Stein MA, Winter J, Grollman JH, Jr. The value of the pulmonary-artery-seeking catheter in percutaneous selective pulmonary arteriography. Radiology 1975; 144: 299–304. Ranniger K. Pulmonary arteriography: a simple method for demonstration of clinically significant pulmonary emboli. Am J Roentgenol 1969; 106: 558–562. Novelline RA, Baltarowich OH, Athanasoulis CA, Waltman AC, Greenfield AJ, McKusick KA. The clinical course of patients with suspected pulmonary embolism and a negative pulmonary arteriogram. Radiology 1978; 126: 561–567. Stein, PD, Athanasoulis C, Alavi A et al. Complications and validity of pulmonary angiography in acute pulmonary embolism. Circulation 1992; 85: 462–468. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. Stein PD, Henry JW. Age related complications of pulmonary angiography for acute pulmonary embolism. Am J Geriatr Cardiol 1993; 2: 13–22. Schwab SJ, Hlatky MA, Pieper KS et al. Contrast nephrotoxicity: a randomized controlled trial of a nonionic and an ionic radiographic contrast agent. N Engl J Med 1989; 320: 149–153. Taliercio CP, Vlietstra RE, Ilstrup DM et al. A randomized comparison of the nephrotoxicity of iopamidol and diatrizoate in high risk patients undergoing cardiac angiography. J Am Coll Cardiol 1991; 17: 384– 390. Donadio C, Tramonti G, Giordani R et al. Glomerular and tubular effects of ionic and nonionic contrast media (diatrizoate and iopamidol). Contr Nephrol 1988; 68: 212– 219. Hirshfeld JW, Jr, Wieland J, Davis CA et al. Hemodynamic and electrocardiographic effects of ioversol during cardiac angiography. Comparison with iopamidol and diatrizoate. Invest Radiol 1989; 24: 138–144. Bettmann MA, Higgins CB. Comparison of an ionic with a nonionic contrast agent for cardiac angiography. Results of a multicenter trial. Invest Radiol 1985; 20: S70– S74.
BLUK077-Stein
April 6, 2007
75
17:4
CHAPTER 75
Contrast-enhanced spiral CT for the diagnosis of acute pulmonary embolism before the Prospective Investigation of Pulmonary Embolism Diagnosis
Introduction Case reports and short case series of patients with pulmonary embolism (PE) diagnosed by “conventional” computed tomography (CT), but not necessarily contrast-enhanced CT, have been published since 1978 [1–8]. The first evaluation of contrast-enhanced spiral CT compared with pulmonary angiography was published in 1992 [9]. Since then, dozens of original studies, reviews, commentaries, and technical notes related to contrast-enhanced spiral CT for the diagnosis of acute PE have been published. By 2001, a higher proportion of tests in patients with PE were obtained with spiral CT than with ventilation– perfusion (V–Q) lung scans [10] (Figures 75.1 and 75.2). In spite of extensive literature, the accuracy of spiral CT for the diagnosis of acute PE was not established definitively prior to the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED II) (Chapter 77). This lack of clarity is due, in part, to the difficulty in obtaining prospective comparisons with an established reference standard using the criteria of an investigation free of design-related bias [11]. There have been several systematic reviews of the topic [12–15] and other extensive reviews [16–21]. A European Collaborative Trial of the diagnostic performance of spiral CT (ESTIPEP Trial) was done between 1996 and 1997, but the results have been reported only as an abstract [22]. Divergent results were reported, depending on whether the spiral CT was interpreted by local or expert readers, and whether the reference standard was pulmonary angiography
alone or in combination with ventilation–perfusion lung scans [22]. Review of investigations of diagnostic accuracy of CT was based on inclusion criteria recommended by Lijmer and associates for avoidance of bias in studies of diagnostic tests [11]. Tier 1 was defined as those that met the following requirements: (1) the diagnosis of PE was made on the basis of objective tests; (2) patients were studied consecutively; (3) the study was performed prospectively; (4) the CT was read without knowledge of the results of the reference test; (5) all patients studied were suspected of having PE; (6) the study included patients with and without PE; (7) the decision to perform the reference diagnostic test was made independently of the result of the spiral CT; (8) descriptions of the CT methods were sufficiently detailed to permit replication; (9) patients with a broad spectrum of demographics were investigated; (10) patients with all severities of PE were evaluated, with no restriction on the results of preliminary diagnostic tests. Investigations that met most, but not all of these criteria were defined as Tier 2. Investigations that were retrospective were excluded from this evaluation of diagnostic accuracy.
Diagnostic accuracy, Tier 1 studies One Tier 1 investigation of diagnostic accuracy was with a 2-slice scanner [23]; the others used single detector scanners [9, 24–29] (Table 75.1). Pooled data from Tier 1 studies showed a sensitivity of 76% and specificity of 89% [23–26, 28, 29] (Table 75.2). Two
325
BLUK077-Stein
326
April 6, 2007
17:4
PART III
Diagnosis of acute PE
(a)
(b)
Figure 75.1 (a) Axial plane with 16-slice contrast-enhanced spiral CT showing bilateral pulmonary emboli. (Courtesy of Brian Sabb, DO, Department of Radiology, St. Joseph Mercy-Oakland Hospital, Pontiac, Michigan.) (b) Same patient as Figure 75.1a. Coronal plane.
BLUK077-Stein
April 6, 2007
CHAPTER 75
17:4
Contrast-enhanced spiral CT before PIOPED II
(a)
(b)
Figure 75.2 (a) Coronal plane with 16-slice contrast-enhanced CT showing bilateral pulmonary emboli (PE) with massive PE in the right pulmonary artery. (Courtesy of Brian Sabb, DO, Department of Radiology, St. Joseph Mercy-Oakland Hospital, Pontiac, Michigan.) (b) Same patient as Figure 75.2a. Axial plane.
327
328
1 1 1 2 1 1 1 1
Drucker [24]
Nilsson [25]
Perrier [26]
Qanadli [23]
Remy-Jardin [9]
Remy-Jardin [27]
Ruiz [28]
Van Strijen [29]
5
3
3 or 5
5
5
3
3
5
Collimation (mm)
Reference
V–Q or Angio
Angio
Angio
Angio
Angio
V–Q or Angio
Angio
Angio
standard
† Specificity not stated. PE, pulmonary embolism; Angio, angiography; V–Q, ventilation–perfusion.
* Main, lobar or segmental pulmonary artery.
Detector #
First author [Ref]
Sensitivity
88/128 (69)
21/23 (91)
56/62 (90)
51/74 (69)
30/33 (91)
9/15 (60)
[n/N (%)]
Any PE
92/109 (84)
31/38 (82)
89/95 (94)
88/98 (90)
55/57 (96)
26/32 (81)
[n/N (%)]
Specificity
80/93 (86)
39/43 (91)
18/18 (100)
[n/N (%)]
Sensitivity
25/32 (78)
23/24 (96)
[n/N (%)]
Specificity
Proximal PE*
6/28 (21)†
[n/N (%)]
Sensitivity
Distal PE
April 6, 2007
Table 75.1 Tier 1: sensitivity and specificity of individual investigations.
BLUK077-Stein
17:4
329 97/143 (68) 137/154 (89)
Single-slice CT, 3 or
proximal
5 mm collimation
collimation
collimation
Single-slice CT; 5 mm
199/273 (73)
Single-slice CT 102/130 (78)
255/335 (76)
Single and 2-slice CT
Single-slice CT; 3 mm
[n/N (%)]
Sensitivity
Group
117/141 (83)
174/193 (90)
291/334 (87)
380/429 (89)
[n/N (%)]
Specificity
97/121 (80)
102/121 (84)
199/244 (82)
255/304 (84)
value [n/N (%)]
Positive predictive
117/163 (72)
174/202 (86)
291/365 (80)
380/460 (83)
value [n/N (%)]
Negative predictive
3.98
7.97
5.66
6.66
likelihood ratio
Positive
0.38
0.24
0.31
0.27
likelihood ratio
Negative
29]
[9, 27,
[24, 29]
28]
[25, 26,
28, 29]
[24–26,
28, 29]
[23–26,
Refs
April 6, 2007
Table 75.2 Tier 1 Investigations: pooled data.
BLUK077-Stein
17:4
BLUK077-Stein
April 6, 2007
17:4
330
investigations used pulmonary angiography and ventilation–perfusion lung scans as the reference standard [26, 29]. All others used pulmonary angiography exclusively [22–25, 28]. When investigators reported results according to two observers [24, 28] the results of observer 1 were used.
Single-slice CT using 3 mm and 5 mm collimation Pooled data of all investigations with single-slice CT showed a sensitivity of 73% [24–26, 28, 29] (Table 75.2). Those with 3 mm collimation showed a sensitivity of 78% [25, 26, 28] (Table 75.2). Specificity was 90%. Positive predictive value with single-slice CT using 3 mm collimation was 84% and negative predictive value was 86% (Table 75.2). Likelihood ratios for a positive test and for a negative test were 7.97 and 0.24, respectively (Table 75.2) [positive likelihood ratio = sensitivity/(1–specificity) and negative likelihood ratio = (1–sensitivity)/specificity]. Data with single-slice CT using 5 mm collimation showed a sensitivity of 68% and specificity of 83% [24, 29] (Table 75.2).
Single-slice CT, proximal PE Sensitivity with single-slice CT using 3 or 5 mm collimation for identifying patients with PE in the main, lobar, or segmental pulmonary artery branches was 89% [9, 27, 29] (Table 75.2). In the main PE, sensitivity with 3 mm collimation was 20 of 20 (100%) and specificity was 101 of 101 (100%) [28]. In the lobar arteries, sensitivity was 61 of 67 (91%) and specificity was 277 of 291 (95%) [28].
Single-slice CT, distal PE Data on distal PE are limited to very few patients (Table 75.1). The number of subsegmental pulmonary artery branches with visible PE increases with decreasing section thickness [30]. More subsegmental arteries would be visible with 2 mm collimation and 0.75 sec/revolution than 3 mm collimation and 1.0 sec/revolution [31].
Tier 2 investigations Individual Tier 2 investigations are shown in Table 75.3, and pooled data of sensitivities, specificities, positive and negative predictive values of Tier 2 investigations are shown in Table 75.4 [32–46]. Single-slice CT
PART III
Diagnosis of acute PE
in Tier 2 investigations showed higher values than in Tier 1 investigations [32, 33, 35–43, 45] (Tables 75.3 and 75.4).
4-Slice CT Only two investigations of diagnostic accuracy prior to PIOPED II evaluated multidetector CT, and both were with 4-slice CT [34, 46]. Pooled data from these studies showed a sensitivity of 45 of 46 (98%) and specificity 132 of 141 (94%) [34, 46] (Table 75.3). With 4-slice CT, visualization of the pulmonary arteries in the middle and peripheral lung zones has been shown to be significantly higher than with single-slice CT, although visualization of the central lung zone was ranked equally with single-slice CT and 4-slice CT [47]. Similarly, 4-slice CT with thin-collimation (1.25 mm) showed improved small pulmonary artery visualization compared with single-slice CT [48].
CT in combination with clinical assessment Single-slice CT The apparent objectivity of CT may lull physicians into a sense of complacency [49]. In the diagnostic process, the probability of PE should be considered in combination with the results of CT [49]. A positive CT in patients with a low probability clinical impression is often falsely positive, and a negative CT in combination with a high probability clinical impression is often falsely negative [49]. Using likelihood ratios calculated from the pooled data, the probability of PE after a positive or negative single-slice CT was estimated in patients stratified according to clinical assessment. Calculations of Tier 1 data showed that the posttest probability of PE would be <84% after a positive single-slice CT in patients with a clinical probability of PE <40% (Figure 75.3). In patients with a negative single-slice CT, if the clinical probability of PE was >70%, there would be more than a 36% posttest probability of PE (Figure 75.3). 4-Slice CT Pooled data of only two investigations CT showed that in patients with a positive 4-slice CT and discordantly low clinical probability of <30%, the posttest probability of PE is <90% (Figure 75.4). Conversely, with a negative 4-slice CT, if the clinical probability of PE is discordantly high, >90%, the calculated posttest probability of PE is >17% (Figure 75.4).
331
1 3
1
1 4 1 1 1 1 1 1 1 1 1 1 and 2 1 4
Blachere [32]
Blum [33]
Coche [34]
Garg [35]‡
Goodman [36]
Kim [37]
Mayo [38]
Otmani [39]
Pruszczyk [40]
Sostman [41]
Steiner [42]
Van Rossum [43]
van Strijen [44]
Velmahos [45]
Winer-Muram [46]
2.5
3
3
5
5
5
5
5
3
3
5
5
40/46 (87)
10/13 (75)
Angio + V–Q +
Angio
18/18 (100)
6/15 (40)
117/135 (87)
Angio + V–Q Angio
64/68 (94)
Angio + V–Q
Angio
Clin
17/17 (100)
Angio or V–Q
41/50 (82)
Angio + V–Q V–Q
23/25 (92)
7/11 (64)
4/7 (57)
Angio, V–Q US
Angio
Angio
Angio + V–Q
Angio
discordant
and V–Q
27/28 (96)
64/68 (94)
V–Q + US ± Angio if CT
[n/N (%)]
Sensitivity
Standard
Reference
67/75 (89)
20/22 (91)
106/117 (91)
78/81 (96)
13/15 (89)
6/6 (100)
22/22 (100)
88/93 (95)
75/78 (96)
8/9 (89)
18/18 (100)
65/66 (98)
104/111 (94)
[n/N (%)]
Specificity
†
Main, lobar or segmental. Subsegmental. ‡ Garg had only patients with intermediate V–Q. PE, pulmonary embolism; Angio, angiography; V–Q, ventilation–perfusion; clin, clinical; US, ultrasound.
∗
(mm) 2 or 3
Collimation Detectors
First author
Any PE
4/8 (50)
30/30 (100)
6/7 (86)
7/7 (100)
[n/N (%)]
Sensitivity
−(100)
8/8 (100)
12/13 (92)
3/3 (100)
[n/N (%)]
Specificity
Proximal PE∗
2/7 (29)
1/4 (25)
[n/N (%)]
Sensitivity
−(93)
[n/N (%)]
Specificity
Distal PE†
April 6, 2007
[Ref]
Table 75.3 Tier 2: sensitivity and specificity of individual investigations.
BLUK077-Stein
17:4
45/46 (98) 393/455(86) 276/320 (86) 137/161(85) 139/159 (87) 47/52 (90)
Single and 2-slice CT
Single-slice CT
Single-slice CT; 2 or 3 mm collimation
Single-slice CT; 5 mm collimation
Single-slice CT, proximal
(%)
4-Slice CT
Group
Sensitivity
332 41/44 (93)
127/133 (95)
305/322(95)
432/455 (95)
536/572 (94)
132/141 (94)
Specificity (%)
47/50 (94)
139/145 (96)
137/154 (89)
276/299 (92)
393/429 (92)
45/54 (83)
value (%)
predictive
Positive
41/46 (89)
127/147 (86)
305/329 (93)
432/476 (91)
536/598 (90)
132/133 (99)
value (%)
predictive
Negative
12.9
12.4
17.0
4.1
14.3
16.3
ratio
likelihood
Positive
0.11
0.14
0.16
0.18
0.14
0.02
ratio
likelihood
Negative
[33, 36, 42, 45]
[36, 39–41, 43]
[32, 35, 37, 38, 45]
[32, 35–41, 43, 45]
[32, 35–41, 43–45]
[33, 46]
Refs
April 6, 2007
Table 75.4 Tier 2 Investigations: pooled data.
BLUK077-Stein
17:4
April 6, 2007
CHAPTER 75
17:4
333
Contrast-enhanced spiral CT before PIOPED II
Figure 75.3 Tier 1 data showing calculated probability of pulmonary embolism (PE) after a positive or a negative single-slice CT with 3 mm collimation stratified according to clinical probability. Calculations were based on the likelihood ratio for a positive test or likelihood ratio for a negative test. Patients with a positive CT and clinical probability of PE <40% would have less than an 84% probability of PE, even though the single-slice CT was positive. Patients with a negative CT and a clinical probability of PE >70% would have greater than a 36% probability of PE, even though the single-slice CT was negative.
100
84
80
Post-CT probability of PE (%)
BLUK077-Stein
92
95
97
99.8
68
CT positive
67
99
96
77
60
49 47
40
CT negative
20
3
14
6
9
20
30
36 26
19
0
0
10
40
50
60
70
80
90
99
Clinical probability of PE (%)
Single-slice CT in combination with venous phase imaging of the veins of the pelvis and lower extremities Evaluation of the lower extremities with venous phase imaging in combination with CT of the chest increases the sensitivity for detection of venous thromboembolic disease in patients with suspected acute PE (see Chapter 37). Evaluation of the lower extremities with venous ultrasound in combination with single detector CT pulmonary angiography [50] also improves the ability to exclude PE except if a limited ultrasound technique is used [44].
Outcome studies with single-slice and 2-slice CT Some have suggested that a negative contrastenhanced spiral CT in untreated patients is rarely fol-
lowed by symptomatic PE [36, 44]. Outcome 3 months after a negative single-slice CT or 2-slice CT in untreated patients is shown in Table 75.5 [26, 32, 35, 38, 44, 50–57]. Outcome 3 months after a negative singleslice CT showed PE in 5.4% or fewer patients [26, 32, 35, 38, 44, 50–56] and in many studies fewer than 2% of patients suffered a PE during follow-up [26, 35, 44, 50, 51, 53, 56] (Table 75.5). Retrospective investigations with single-slice CT showed PE following a negative CT in 1.7% or fewer patients [58–64].
Outcome studies with 4-slice and 16-slice CT With 4-slice or 16-slice CT, PE after a negative CT was shown in 1.5% or fewer patients [53, 65–70] (Table 75.6). Most investigations, however, were in patients who had other negative tests in addition to CT
100
Post-CT probability of PE (%)
87
Figure 75.4 Tier 2 data showing calculated probability of pulmonary embolism (PE) after a positive or a negative 4-slice CT stratified according to clinical probability. Patients with a positive CT and clinical probability of PE <30% would have less than a 90% probability of PE, even though the 4-slice CT was positive. Patients with a negative CT and a clinical probability of PE >90% would have greater than a 17% probability of PE, even though the 4-slice CT was negative.
89
91
94
96
97
98
99
99.8
79
80
63
71
CT positive
60
CT negative 40
17
20
0.3
1 20
1
2
2
30
40
50
3
5
8
60
70
80
0
0
10
Clinical probability of PE (%)
90
99
334 1
2
2
1
2
1
1
1
1
Garg [35]
Goodman [56]
Lorut [57]
Mayo [38]
Mussett [50]
Ost [55]
Perrier [26]
Remy-Jardin [53]
Tille-Leblond [51] 2
Van Strijen [44]
1
1
1
1
2
1
1
1
1
2, 3
2, 3
?
3
2, 3
3
2, 3
3
3
5
2, 3
44 low
intermediate
100 low/
100 low
intermediate
100 low/
31
36∗
24∗
intermediate
5 nl 6 low 2
intermediate
26 nl/vl 44 low/
intermediate
100 low/
100 vl/low
14 nl
intermediate
11 nl 89 low/
intermediate
100
0
100
Negative
Negative serial Negative
100
84∗
94
44
45
100
42
100
100
100
77
26
10∗
30
1
ultrasound (%) ultrasound (%) DSA (%)
3
6
3
3
6
3
3
3
3
7
3
6
3
F/U (mos)
DVT
1/246 (0.4)
3/185 (1.6)
0/62 (0)
0/156 (0)
3/71 (4.2)
6/507 (1.1)
2/88 (2.3)
2/117 (1.7)
2/198 (1.0)
0/18 (0)
6/112 (5.4)
4/81 (4.9)
3/100 (3.0)
1/71 (1.4)
3/507 (0.6)
0/117 (0)
PE [n/N (%)] [n/N (%)]
V–Q, ventilation–perfusion; DSA, digital subtraction angiography; PE, pulmonary embolism; DVT, deep venous thrombosis; F/U, follow up; nl, normal; vl, very low.
* Estimated.
1
2
Ferretti [52]
1
1
2
Bourriot [54]
3
Negative D-dimer (%) V–Q (%)
April 6, 2007
1
1
Blachere [32]
probability %
Collimation Clinical
Tier CT slices (mm)
[Ref]
First author
Table 75.5 Outcome following a negative contrast-enhanced 1-slice or 2-slice CT: Tier 1 and 2 investigations.
BLUK077-Stein
17:4
1
1
Perrier [66]
Perrier [67]
335
Collimation
3, 1.25
1.25
1
2
?
3
?
(mm)
Clinical
unlikely
66% likely, 34%
intermediate
low/
intermediate
low/
probability
intermediate
7 nl, 10 low 2
?
V–Q (%)
Negative
100 CTV
93
100
100
?
ultrasound (%)
F/U
3
3
3
6
3
3
9
(mos)
PE
10/1436 (0.7)
0/111 (0)
1/91 (1.0)
1/100 (1.0)
5/292 (1.7)
5/408 (1.2)
1/65 (1.5)
[n/N (%)]
DVT
8/1436 (0.6)
2/111 (1.8)
1/100 (1.0)
2/408 (0.5)
[n/N (%)]
*Patients also had negative venous ultrasonography. V–Q, ventilation–perfusion; PE, pulmonary embolism; DVT, deep venous thrombosis; CTV, computed tomography verous phase; MDCT, multidetector computed tomography, nl = normal.
Christopher [70] 1
1, MDCT
4
Revel [69]
1
4 4
1
4, 16*
1, 82% MDCT 18%
Remy-Jardin [53]1
Prologo [68]
1
Kavanagh [65]
4
CT slices
Tier
First author
[Ref]
April 6, 2007
Table 75.6 Outcome following a negative contrast-enhanced 4-slice or 16-slice CT: Tier 1 and 2 investigations.
BLUK077-Stein
17:4
BLUK077-Stein
April 6, 2007
17:4
336
[53, 65–67, 69, 70]. Even with a “likely” clinical assessment in the majority tested, an outcome study by the Christopher Study Investigators showed only 0.7% PE and 0.6% deep venous thrombosis on 3-month followup [70]. Many had PE excluded by a negative CT with some combination of negative venous ultrasound or venous phase imaging of the leg veins, low or intermediate probability V–Q scan or low or intermediate probability clinical assessment (Tables 75.5 and 75.6). Moores and associates, based on meta-analysis of outcomes in investigations mostly with single-slice CT, recommended concurrent lower extremity imaging before withholding anticoagulation in patients with suspected PE [19, 20]. Quiroz and associates, based on systematic review and meta-analysis of investigations primarily with single-slice CT, also concluded that patients with a negative CT who have a high probability clinical assessment require additional imaging to exclude PE [71]. However, Perrier and associates in 2005 showed a potential ability to exclude PE on the basis of negative multidetector computed tomography angiography (CTA) without lower limb ultrasonography [67]. In untreated patients after a negative CTA and low or moderate clinical probability assessment, Perrier and associates would have shown a risk of only 1.5% for venous thromboembolism (PE and/or deep venous thrombosis) within 3 months if ultrasonography had not been included in the diagnostic workup [67]. Prologo and associates showed, in fact, that PE was rare after a negative single-slice CT alone (0 of 98), and after a negative 4-slice CT alone [1 of 100 (1.0%)] [68] (Table 75.6).
Effect of age on sensitivity and specificity of single-slice CT angiography Among patients ≤59, 60–75, and ≥76 years of age, a trend suggested that the sensitivity of single-slice CTA was higher in the youngest age group, but differences did not reach statistical significance [72]. In patients ≤59 years of age, sensitivity of CTA was 81%, whereas in those 60–75 and ≥76 years of age, sensitivities were 63 and 67%. Specificity in each age group ranged from 86 to 96%.
References 1 Sinner WN. Computed tomographic patterns of pulmonary thromboembolism and infarction. J Comput Assist Tomogr 1978; 2: 395–399.
PART III
Diagnosis of acute PE
2 Sinner WN. Computed tomography of pulmonary thromboembolism. Eur J Radiol 1982; 2: 8–13. 3 DiCarlo LA, Jr, Schiller NB, Herfkens RL, Brundage BH, Lipton MJ. Noninvasive detection of proximal pulmonary artery thrombosis by two-dimensional echocardiography and computerized tomography. Am Heart J 1982; 104: 879–881. 4 Breatnach E, Stanley RJ. CT diagnosis of segmental pulmonary artery embolus. J Comput Assist Tomogr 1984; 8: 762–764. 5 Chintapalli K, Thorsen MK, Olson DL, Goodman LR, Gurney J. Computed tomography of pulmonary thromboembolism and infarction. J Comput Assist Tomogr 1988; 12: 553–559. 6 Shah HR, Buckner CB, Purness GL, Walker CW. Computed tomography and magnetic resonance imaging in the diagnosis of pulmonary thromboembolic disease. J Thorac Imaging 1989; 4: 58–61. 7 Kalebo P, Wallin J. Computed tomography in massive pulmonary embolism. Acta Radiol 1989; 30: 105–107. 8 Godwin JD, Webb WR, Gamsu G, Ovenfors CO. Computed tomography of pulmonary embolism. Am J Roentgenol 1980; 135: 691–695. 9 Remy-Jardin M, Remy J, Wattinne L, Giraud F. Central pulmonary thromboembolism: diagnosis with spiral volumetric CT with the single-breath-hold technique— comparison with pulmonary angiography. Radiology 1992; 185: 381–387. 10 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316– 1317. 11 Lijmer JG, Mol BW, Heisterkamp S et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 1999; 282: 1061–1066. 12 Rathbun SW, Raskob GE, Whitsett TL. Sensitivity and specificity of helical computed tomography in the diagnosis of pulmonary embolism: a systematic review. Ann Intern Med 2000; 132: 227–232. 13 Mullins MD, Becker DM, Hagspiel KD, Philbrick JT. The role of spiral volumetric computed tomography in the diagnosis of pulmonary embolism. Arch Intern Med 2000; 160: 293–298. 14 van Beek EJR, Brouwers EMJ, Song B, Bongaerts AHH, Oudkerk M. Lung scintigraphy and helical computed tomography for the diagnosis of pulmonary embolism: a meta analysis. Clin Appl Thrombosis/Hemostasis 2001; 7: 87–92. 15 Eng J, Krishnan JA, Segal JB et al. Accuracy of CT in the diagnosis of pulmonary embolism: a systematic literature review. Am J Roentgenol 2004; 183: 1819– 1827. 16 Garg K. CT of pulmonary thromboembolic disease. Radiol Clin North Am 2002; 40: 111–122.
BLUK077-Stein
April 6, 2007
CHAPTER 75
17:4
Contrast-enhanced spiral CT before PIOPED II
17 MacDonald SL, Mayo JR. Computed tomography of acute pulmonary embolism. Semin Ultrasound CT MRI 2003; 24: 217–231. 18 Schoepf UJ, Costello P. CT angiography for diagnosis of pulmonary embolism: state of the art. Radiology 2004; 230: 329–337. 19 Moores LK, Jackson WL, Jr, Shorr AF, Jackson JL. Metaanalysis: outcomes in patients with suspected pulmonary embolism managed with computed tomographic pulmonary angiography. Ann Intern Med 2004; 141: 866– 874. 20 Moores LK, Jackson WL, Jr. In response. Ann Intern Med 2005; 142: 802. 21 Patel S, Kazerooni EA. Helical CT for the evaluation of acute pulmonary embolism. Am J Roentgenol 2005; 185: 135–149. 22 Herold CJ, Hahne J, Ghaya B et al. Prospective evaluation of pulmonary embolism: diagnostic performance of spiral CT angiography in the ESTIPEP Trial [Abstract]. Radiology 1999; 213: 126. 23 Qanadli SD, Hajjam ME, Mesurolle B et al. Pulmonary embolism detection: prospective evaluation of dualsection helical CT versus selective pulmonary arteriography in 157 patients. Radiology 2000; 217: 447–455. 24 Drucker EA, Rivitz SM, Shephard JAO et al. Acute pulmonary embolism: assessment of helical CT for diagnosis. Radiology 1998; 209: 235–241. 25 Nilsson T, Soderberg M, Lundqvist G et al. A comparison of spiral computed tomography and latex agglutination D-dimer assay in acute pulmonary embolism using pulmonary arteriography as gold standard. Scand Cardiovasc J 2002; 36: 373–377. 26 Perrier A, Howarth N, Didler D et al. Performance of helical computed tomography in unselected outpatients with suspected pulmonary embolism. Ann Intern Med 2001; 135: 88–97. 27 Remy-Jardin M, Remy J, Dechildre F et al. Diagnosis of pulmonary embolism with spiral CT: comparison with pulmonary angiography and scintigraphy. Radiology 1996; 200: 699–706. 28 Ruiz Y, Caballero P, Caniego JL et al. Prospective comparison of helical CT with angiography in pulmonary embolism: global and selective vascular territory analysis. Interobserver agreement. Eur Radiol 2003; 13: 823– 829. 29 Van Strijen MJ, De Monye W, Kieft GJ et al. Accuracy of single-detector spiral CT in the diagnosis of pulmonary embolism: a prospective multicenter cohort study of consecutive patients with abnormal perfusion scintigraphy. J Thromb Haemost 2005; 3: 17–25. 30 Schoepf UJ, Holzknecht N, Helmberger TK et al. Subsegmental pulmonary emboli: improved detection with thin-collimation multi-detector row spiral CT. Radiology 2002; 222: 483–490.
337
31 Remy-Jardin M, Remy J, Artaud D et al. Peripheral pulmonary arteries: optimization of the spiral CT acquisition protocol. Radiology 1997; 204: 157–163. 32 Blachere H, Latrabe V, Montaudon M et al. Pulmonary embolism revealed on helical CT angiography: comparison with ventilation–perfusion radionuclide lung scanning. AJR Am J Roentgenol 2000; 174: 1041–1047. 33 Blum AG, Delfau F, Grigon B et al. Spiral computed tomography versus pulmonary angiography in the diagnosis of acute pulmonary embolism. Am J Cardiol 1994; 74: 96–98. 34 Coche E, Verschuren F, Keyeux A. Diagnosis of acute pulmonary embolism in outpatients: comparison of thincollimation multi-detector row spiral CT and planar ventilation perfusion scintigraphy. Radiology 2003; 229: 757– 765. 35 Garg K, Welsh CH, Feyerabend AJ et al. Pulmonary embolism: diagnosis with spiral CT and ventilation– perfusion scanning—correlation with pulmonary angiographic results or clinical outcome. Radiology 1998; 208: 201–208. 36 Goodman LR, Curtin JJ, Mewissen MW et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. Am J Roentgenol 1995; 164: 1369–1374. 37 Kim Kun-II, Muller NL, Mayo JR. Clinically suspected pulmonary embolism: utility of spiral CT. Radiology 1999; 210: 693–697. 38 Mayo JR, Remy-Jardin M, Muller NL et al. Pulmonary embolism: prospective comparison of spiral CT with ventilation–perfusion scintigraphy. Radiology 1997; 205: 447–452. 39 Otmani A, Tribouilloy C, Leborgne L et al. [Diagnostic value of echocardiography and thoracic spiral CT angiography in the diagnosis of acute pulmonary embolism]. Ann Cardiol Angeiol (Paris) 1998; 47(10): 707–715. 40 Pruszczyk P, Torbicki A, Pacho R et al. Noninvasive diagnosis of suspected severe pulmonary embolism: transesophageal echocardiography vs spiral CT. Chest 1997; 112: 722–728. 41 Sostman HD, Layish DT, Tapson VF et al. Prospective comparison of helical CT and MR imaging in clinically suspected acute pulmonary embolism. J Magn Reson Imaging 1996; 6: 275–281. 42 Steiner P, Phillips F, Wesner D et al. [Primary diagnosis and follow-up in acute pulmonary embolism: comparison of digital subtraction angiography and spiral CT]. ROFO 1994; 161: 285–291. 43 van Rossum AB, Pattynama PM, Ton ER et al. Pulmonary embolism: validation of spiral CT angiography in 149 patients. Radiology 1996; 201: 467–470. 44 van Strijen MJ, de Mony´e W, Schiereck J et al. Singledetector helical computed tomography as the primary diagnostic test in suspected pulmonary embolism: a
BLUK077-Stein
April 6, 2007
17:4
338
45
46
47
48
49
50
51
52
53
54
55
56
multicenter clinical management study of 510 patients. Ann Intern Med 2003; 138: 307–314. Velmahos GC, Toutouzas KG, Vassiliu P et al. Can we rely on computed tomographic scanning to diagnose pulmonary embolism in critically ill surgical patients? J Trauma 2004; 56(3): 518–525. Winer-Muram HT, Rydberg J, Johnson MS et al. Suspected acute pulmonary embolism: evaluation with multi-detector row CT versus digital subtraction pulmonary arteriography. Radiology 2004; 233: 806–815. Raptopoulos V, Boiselle PM. Multi-detector row spiral CT pulmonary angiography: comparison with singledetector row spiral CT. Radiology 2001; 221: 606–613. Patel S, Kazerooni EA, Cascade PN. Pulmonary embolism: optimization of small pulmonary artery visualization at multi-detector row CT. Radiology 2003; 227: 445–460. Eisner MD. Before diagnostic testing for pulmonary embolism: estimating the prior probability of disease. Am J Med 2003; 114: 232–234. Musset D, Parent F, Meyer G et al. Diagnostic strategy for patients with suspected pulmonary embolism: a prospective multicentre outcome study. Lancet 2002; 360: 1914– 1920. Tillie-Leblond I, Mastora I, Radenne F et al. Risk of pulmonary embolism after a negative spiral CT angiogram in patients with pulmonary disease: 1-year clinical followup study. Radiology 2002; 223: 461–467. Ferretti GR, Bosson J-L, Buffaz P-D et al. Acute pulmonary embolism: role of helical CT in 164 patients with intermediate probability at ventilation–perfusion scintigraphy and normal results at duplex US of the legs. Radiology 1997; 205: 453–458. Remy-Jardin M, Tillie-Leblond I, Szapiro D et al. CT angiography of pulmonary embolism in patients with underlying respiratory disease; impact of multislice CT on image quality and negative predictive value. Eur Radiol 2002; 12: 1971–1978. Bourriot K, Couffinhal T, Bernard V, Montaudon M, Bonnet J, Laurent F. Clinical outcome after a negative spiral CT pulmonary angiographic finding in an inpatient population from cardiology and pneumology wards. Chest 2003; 123: 359–365. Ost D, Rozenshtein A, Saffran L et al. The negative predictive value of spiral computed tomography for the diagnosis of pulmonary embolism in patients with nondiagnostic ventilation–perfusion scans. Am J Med 2001; 110: 16–21. Goodman LR, Lipchik RJ, Kuzo RS, Liu Y, McAuliffe TL, O’Brien DJ. Subsequent pulmonary embolism: risk after a negative helical CT pulmonary angiogram: prospective comparison with scintigraphy. Radiology 2000; 215: 535– 542.
PART III
Diagnosis of acute PE
57 Lorut C, Ghossains M, Horellou M-H et al. A noninvasive diagnostic strategy including spiral computed tomography in patients with suspected pulmonary embolism. Am J Respir Crit Care Med 2000; 162: 1413–1418. 58 Donato A, Scheirer JJ, Atwell MS et al. Clinical outcomes in patients with suspected acute pulmonary embolism and negative helical computed tomographic results in whom anticoagulation was withheld. Arch Intern Med 2003; 163: 2033–2038. 59 Garg K, Sieler H, Welsh CH, Johnston RJ, Russ PD. Clinical validity of helical CT being interpreted as negative for pulmonary embolism: implications for patient treatment. Am J Roentgenol 1999; 172: 1627–1631. 60 Gottsater A, Berg A, Centergard J et al. Clinically suspected pulmonary embolism: is it safe to withhold anticoagulation after a negative spiral CT? Eur Radiol 2001; 11: 65–72. 61 Krestan CR, Klein N, Fleischmann D et al. Value of negative spiral CT angiography in patients with suspected acute PE: analysis of PE occurrence and outcome. Eur Radiol 2004; 14: 93–98. 62 Lombard J, Bhatia R, Sala E. Spiral computed tomographic pulmonary angiography for investigating suspected pulmonary embolism: clinical outcomes. Can Assoc Radiol J 2003; 54: 147–151. 63 Lomis NN, Yoon HC, Moran AG, Miller FJ. Clinical outcomes of patients after a negative spiral CT pulmonary arteriogram in the evaluation of acute pulmonary embolism. J Vasc Interv Radiol 1999; 10: 707–712. 64 Nilsson T, Olausson A, Johnsson H et al. Negative spiral CT in acute pulmonary embolism. Acta Radiologica 2002; 43: 486–491. 65 Kavanagh EC, O’Hare A, Hargaden G, Murray JG. Risk of pulmonary embolism after negative MDCT pulmonary angiography findings. Am J Roentgenol 2004; 182: 499– 504. 66 Perrier A, Roy P-M, Aujesky D et al. Diagnosing pulmonary embolism in outpatients with clinical assessment, D-dimer measurement, venous ultrasound, and helical computed tomography: a multicenter management study. Am J Med 2004; 116: 291–299. 67 Perrier A, Roy P-M, Sanchez O et al. Multi-row computed tomography in suspected pulmonary embolism. N Engl J Med 2005; 352: 1760–1768. 68 Prologo JD, Gilkeson RC, Diaz M, Cummings M. The effect of single-detector CT versus MDCT on clinical outcomes in patients with suspected acute pulmonary embolism and negative results on CT pulmonary angiography. Am J Roentgenol 2005; 184: 1231–1235. 69 Revel MP, Petrover D, Hernigou A, Lefort C, Meyer G, Frija G. Diagnosing pulmonary embolism with fourdetector row helical CT: prospective evaluation of 216 outpatients and inpatients. Radiology 2005; 234: 265–273.
BLUK077-Stein
April 6, 2007
CHAPTER 75
17:4
Contrast-enhanced spiral CT before PIOPED II
70 Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA 2006; 295: 172–179. 71 Quiroz R, Kucher N, Zou KH et al. Clinical validity of a negative computed tomography scan in patients with sus-
339
pected pulmonary embolism: a systematic review. JAMA 2005; 293: 2012–2017. 72 Righini M, Bounameaux H, Perrier A. Effect of age on the performance of single detector helical computed tomography in suspected pulmonary embolism. Thromb Haemost 2004; 91: 296–299.
BLUK077-Stein
March 28, 2007
76
15:58
CHAPTER 76
Methods of PIOPED II
Introduction The Prospective Investigation of Pulmonary Embolism II (PIOPED II) was a prospective multicenter investigation sponsored by the National Institutes of Health, National Heart Lung and Blood Institute of the accuracy of multidetector computed tomography angiography (CTA) alone and combined with venous phase imaging of the pelvic and thigh veins (CTV) for the diagnosis of acute pulmonary embolism (PE) [1]. The PIOPED II Trial was designed with two primary objectives: (1) to determine if multidetector CTA can reliably detect and exclude acute PE and (2) to determine if adding CTV improves the ability to detect and exclude PE. The investigators also determined whether the addition of a validated clinical assessment, the extended Wells score, improved the ability to detect or exclude PE by CTA or CTA/CTV in patients with suspected PE. All the Standards for Reporting Diagnostic Accuracy (STARD) were met [2, 3]. Clinical centers in PIOPED II were University of Calgary, Cornell University, Duke University, Emory University, Henry Ford Hospital, Massachusetts General Hospital, University of Michigan, and Washington University. The data and coordinating center was at George Washington University, and the administrative center was at St. Joseph Mercy-Oakland Hospital in Pontiac, Michigan. All patients aged ≥18 years with clinically suspected acute PE, whether inpatients or outpatients, seen at the eight participating clinical centers between September 2001 and July 2003 were potentially eligible for recruitment. Patients sent for a diagnostic imaging test for PE were identified for recruitment as well as patients for whom the study nurse was aware of a consultation request for suspected PE. Patients were recruited consecutively during periods of staff availability, usually during the daytime on weekdays. Exclusion criteria were an inability to complete tests within 36 hours, critical illness or hemodynamically unstable, ventilatory support, shock or hypotension, myocardial infarction within 1 month, ventricular fib-
340
rillation or sustained ventricular tachycardia within 24 hours, abnormal serum creatinine, chronic renal dialysis (first 14 months of recruitment only), allergy to contrast material, pregnancy, treatment with longterm anticoagulants, thrombolytic therapy planned in next 24 hours, inferior vena cava filter, deep venous thrombosis (DVT) of the upper extremity, prisoners, and patients previously enrolled [4]. All enrolled patients underwent a clinical assessment of the probability of PE, including a Wells extended model score (Chapter 52) [5]. In addition, all patients gave consent to undergo diagnostic testing including CTA/CTV, ventilation–perfusion (V–Q) scanning, venous compression ultrasonography of the lower extremities, and, if necessary, pulmonary digital subtraction angiography (DSA) [4]. For ethical reasons, conventional pulmonary DSA was restricted to patients in whom PE was not conclusively diagnosed or excluded by the noninvasive tests. A composite reference standard was used to diagnose or exclude PE. Diagnosis of PE according to the composite reference standard required one of the following in PIOPED II.
1 High probability V–Q lung scan in a patient with no history of prior PE. 2 Positive pulmonary DSA. 3 Positive venous ultrasound in a patient without prior DVT at that site and a nondiagnostic V–Q scan (not normal and not high probability without prior PE). This was interpreted as a surrogate for the diagnosis of PE. Exclusion of PE according to the composite reference standard required one of the following:
1 Negative DSA 2 Normal V–Q scan 3 Low or very low probability V–Q scan, clinical score by the Wells criteria <2 (Chapter 52, Table 52.1) [5], and negative venous ultrasound. To confirm the accuracy of the diagnoses made according to the composite reference standard, patients in whom PE was excluded by the reference test
BLUK077-Stein
March 28, 2007
CHAPTER 76
15:58
Methods of PIOPED II
Table 76.1 Diagnostic criteria for acute pulmonary embolism by CT. 1 Arterial occlusion—failure to opacify the entire lumen due to a central filling defect (the artery may enlarge compared to peers). 2 Partial filling defect—surrounded by contrast (crosssectional image). 3 “Railway tracking”—there is a small amount of contrast between the central filling defect and the artery wall (in plane, longitudinal image). 4 Peripheral intramural defect—the eccentric filling defect makes an acute angle with the artery wall.
underwent telephone interviews at 3 and 6 months after enrollment. Deaths and new evaluations for venous thromboembolic disease were reviewed by an Outcome Committee. The composite reference test eliminated the risks of pulmonary angiography in patients in whom a diagnosis can be established by noninvasive methods. It paralleled the strategy of management of patients with suspected PE that was used in most medical centers. It eliminated the ethical problem of asking a patient to volunteer for a pulmonary angiogram, an invasive test that carries some risk, although the risk is small. It facilitated recruitment. The disadvantage of including patients diagnosed on the basis of this composite reference test was related to the introduction of some imprecision in the diagnostic reference standard. To minimize this imprecision in the diagnosis or exclusion of PE, application of the noninvasive tests was restricted as follows. A diagnosis of PE on the basis of a high probability V–Q lung scan required that the patient have no history of prior PE. In such patients, based on the results of PIOPED I, the positive predictive value for PE is 91% [6]. The positive predictive value for PE of a high probability V–Q lung scan in a patient with prior PE is 74% [6]. Therefore, patients with prior PE were not used to determine PE on the basis of a V–Q scan [6]. The false positive rate for a high probability V–Q scan in patients with no prior PE has been shown to be 9 of 97 (9.3%) [6]. The false positive rate for compression ultrasound in patients with suspected PE was shown to be 5 of 178 (2.8%) [7]. The false negative rate of conventional pulmonary DSA may be 6 of 380 (1.6%) based on PE within 1 year among untreated patients with a negative DSA [8]. The
341
false negative rate of a normal V–Q scan has been cited as 0 of 21 (0%), 0 of 68 (0%), 0 of 113 (0%) [6, 9, 10], 2 of 586 (0.3%) [11], and 4 of 334 (1.2%) [12]. The combination of a low clinical probability, low or very low probability V–Q scan, and negative leg ultrasound to exclude VTE in this protocol was more stringent than published criteria using a low clinical probability, low or intermediate probability V–Q, and negative leg ultrasound. The latter combination is cited as having a false negative rate for VTE of 9 of 450 (2.0%) [12], 1 of 41 (2.4%) [13], and 3 of 175 (1.7%) using the Geneva score [14]. The diagnosis of DVT by venous compression ultrasound of the legs was a surrogate for the diagnosis of PE. It is unlikely that coincidental DVT would be identified in patients in whom PE was absent. Wells and associates showed that among patients with suspected PE who had normal perfusion lung scan, an initial ultrasound of the lower extremity was positive in 2 of 334 (0.6%) [12]. In those with suspected PE who had a high pretest clinical suspicion and a high probability interpretation of the V–Q lung scan, an initial ultrasound was positive in 35 of 60 (58%). A positive venous compression ultrasound in patients with suspected PE, therefore, was observed 100 times more frequently among patients with PE than in those in whom PE was absent [12]. If venous ultrasound was used as a reference test, the patient did not have a history of prior DVT or prior DVT was in a different location. This reduced the possibility of detecting chronic DVT rather than acute DVT. A negative venous ultrasound alone did not exclude PE, and serial venous ultrasounds were not used in PIOPED II [7, 15, 16]. Although some exactness in the diagnosis of PE may be lost by including patients diagnosed by noninvasive tests, pulmonary angiography itself is not a perfect standard, particularly for the diagnosis of PE in subsegmental pulmonary arteries. In PIOPED I, among 22 patients with PE limited to the subsegmental pulmonary arteries, the average co-positivity of readings by committee readers was 66% [17]. With larger vessels, co-positivity was higher. Among 136 patients whose pulmonary angiograms showed PE in segmental pulmonary arteries, but not in larger orders of arteries, the average co-positivity was 90%. Among 217 patients whose angiograms showed PE in main or lobar pulmonary arteries, the average co-positivity was 98% [17].
BLUK077-Stein
March 28, 2007
15:58
342
The baseline assessment consisted of eligibility screening, patient demographic and clinical characteristics, and medical history. Enrollment preceded imaging. Eligible patients were recruited consecutively during periods of staff availability, usually daytime during weekdays. Each eligible consenting patient had a V–Q lung scan, contrast enhanced spiral CT, venous phase contrast enhanced CT of the veins of the pelvis and thighs (CTV), and venous compression ultrasound of the lower extremities. Some, in addition, underwent DSA of the pulmonary arteries. The V–Q lung scan, spiral CT, and venous compression ultrasound were read locally at the clinical center, and the results of the local reading were available to the physicians caring for the patient. Image interpretations for all diagnostic tests except venous ultrasound were based on agreement of two PIOPED II certified readers from centers other than that at which the image was obtained. Additional readers were used until agreement of two was obtained. Readers were blind to all clinical information and to the results of other imaging tests except chest radiographs, which were included with V–Q scans. Local readings of venous ultrasound were accepted after site visits to validate technique and interpretation. Spiral CTA and conventional DSA required agreement in at least one lobe for a diagnosis of PE, and PE was excluded if two readers agreed that PE was absent. The CTV required agreement in the same leg for DVT. Separate reader consensus was required for both CTA and CTV. For CTA/CTV, PE was diagnosed if there was consensus that either test was positive. Pulmonary embolism was excluded if there was consensus that both CTA and CTV were negative. Patients in whom the CTA or the CTA/CTV was unclassified were excluded from the calculations of sensitivity, specificity, positive and negative predictive values. For calculation of the negative predictive value of CTA in patients with a low probability clinical assessment, only patients with a reference test diagnosis by V–Q scan or conventional pulmonary DSA were included. Values for likelihood ratios for a positive test were calculated as sensitivity/(1 − specificity) and likelihood ratios for a negative test were calculated as (1 − sensitivity)/specificity [18–20]. In a separate analysis, values for the sensitivity and specificity of CTA were adjusted for possible inaccuracy of the composite reference standard, using the
PART III
Diagnosis of acute PE
lowest reported false positive and false negative rates for the tests comprising that standard [6–8, 10, 14]. Calculations were also performed using the highest reported false positive and false negative rates for the composite reference standard tests [5–8, 12].
CT angiography and CT venography The study was done with 4, 8, or 16 head scanners. Scanners with 4 detector arrays [High Speed Advantage (General Electric, Milwaukee, Wisconsin) or Volume Zoom (Siemens Medical Systems, Malvern, Pennsylvania)] were used in 691 patients. Eight detector scanners [Light Speed Ultra (General Electric)] were used in 37 patients, and 16 detector scanners [Light Speed 16 (General Electric), Sensation 16 (Siemens), Aquillon (Toshiba)] in 45 patients. For patients less than 250 lbs scanned on 4-slice equipment, collimation was 1.25 mm, table speed 7.5 mm/rotation, pitch 1.5 (usually between 1.0 and 2.0), voltage 120 kVp, current 400 mA, and rotation time approximately 0.8 seconds. The effective thickness is slightly wider than the nominal thickness. This permitted better visibility of subsegmental vessels [21, 22]. Minor protocol modifications were made for newer scanners. In patients weighing more than 250 lbs, the protocol was modified to decrease the quantum mottle and increase scan quality. Spiral CT of the chest was performed with an injection to scan delay of 20–28 seconds. An injection of 135–150 mL of low osmolar nonionic contrast material (300–320 mg iodine per cc) was made through an arm vein at 4 mL/second. Scans started 2 cm below the lowest diaphragm and proceeded cranially to the top of the lung apex. Because of the thin sections, no overlapping reconstruction was necessary. However, the large patient protocol required 2.5 mm overlapping reconstruction. Work stations were used for primary reading. Images were displayed at 3 different gray scales for interpretation: Lung windows (W = 2000 HU, L = −500 HU), Mediastinal window (W = 450 HU, L = 50 HU), Pulmonary embolus specific setting (wider window and higher level adjusted on the work stations based on degree of contrast enhancement).
BLUK077-Stein
March 28, 2007
CHAPTER 76
15:58
343
Methods of PIOPED II
Table 76.2 Diagnostic criteria for chronic pulmonary embolism by CT
Table 76.3 Diagnostic criteria for acute deep venous thrombosis by spiral CT venography.
1 Complete occlusion but the vessel is smaller than its
1 Complete filling defect. Failure to opacify the entire
peers. 2 Eccentric or crescentic thickening of a vessel wall with obtuse angles to the wall.
lumen due to a central filling defect (the vessel may be enlarged compared to the opposite vein). 2 Partial filling defect surrounded by contrast.
3 Contrast flowing through a thickened, often smaller artery (recanalization). 4 Web or flap within contrast-filled artery. 5 Secondary signs—extensive bronchial or other systemic collateral through that area, accompanying mosaic perfusion pattern on lung windows, or calcification within eccentric thickening.
The main, lobar, segmental, and subsegmental arteries were examined. Each vessel and its branches was interpreted as normal, containing acute PE, containing chronic PE, or uncertain. Complete visualization of a main, lobar, or segmental vessel requires the branch be followed to its bifurcation. Readers scored their degree of diagnostic certainty by using a 3-point scale (present, absent, or uncertain). The diagnostic criteria for acute PE by CT are shown in Table 76.1 and the diagnostic criteria for chronic PE by CT are shown in Table 76.2. An inability to visualize the vessel did not indicate that an embolus is present. Spiral CT of the lung required adjudication if there was disagreement on which lobes contain acute PE. If there was agreement on the presence of PE in one lobe, the case did not require adjudication irrespective of the findings in the other lobes. Central adjudicated readings were performed blindly, without knowledge of the sites of disagreement.
Venous phase spiral CT of the veins of the lower extremities Contrast-enhanced spiral CT of the veins of the lower extremities were performed during the venous phase of the chest spiral CT. No additional contrast was necessary. The deep veins were scanned from the inferior vena cava confluence at the level of the iliac crest through the popliteal veins. The CTV used 7.5 mm collimation, 7.5 mm reconstruction, table speed 30 mm/rotation, pitch 1.5, and an injection to scan delay
of 3 minutes. Current was 180 mA and rotation time was 1 second. In patients ≤250 lbs, voltage was 120 kVp and in patients >250 lbs, it was 140 kVp. Veins of the calves were not studied, because venous phase venography usually fails to opacify the calf veins. The diagnostic criteria for acute DVT by spiral CT venography are shown in Table 76.3 and for chronic DVT in Table 76.4. The criteria for chronic DVT are not well established in the CT literature, but are based on established criteria in the venous ultrasound literature.
Procedures to minimize potential risks Because study patients might get both spiral CT and DSA over a relatively short period of time, the risks of renal dysfunction from contrast material were minimized by the following: Patients with inappropriate plasma creatinine were excluded. Nonionic contrast material was used. (In PIOPED I, high osmolar contrast material was used for pulmonary angiograms.) Patients were hydrated before and after the spiral CT and pulmonary angiogram (Chapter 79). The interval between contrast injections was at least 4 hours to allow excretion of prior contrast material and to allow hydration. To minimize the risk of pulmonary angiography, senior and experienced physicians performed or closely supervised the procedure.
Table 76.4 Diagnostic criteria for deep venous thrombosis of unknown age by spiral CT venography. 1 Complete filling defect, but the vein is smaller than its peers. 2 Contrast flowing through a thickened, often smaller vein. 3 Calcification in vein. 4 Increased collateral vessels (secondary sign)
BLUK077-Stein
March 28, 2007
15:58
344
Time interval for imaging V–Q lung scans, venous compression ultrasound of the lower extremities, pulmonary angiograms, and spiral CTs were obtained within a 36-hour period in 85% of patients, 37–48 hours in 7.5%, and beyond 48 hours in 7.5%. Plain chest radiographs were obtained within 2 hours of the V–Q lung scans if possible, but were required within 12 hours.
V–Q Lung scans—technique Perfusion imaging Macroaggregated human albumin labeled with 99m Tc was used. The aggregated albumin preparation typically contains 90% of particles within the 10–90 μm range. The usual imaging dose is about 0.14 μg/kg albumin. At least 60,000 particles were injected for an adequate study. Perfusion imaging was begun by an intravenous injection of 4 mCi of macroaggregates with the patient in the supine position [23]. After injection, imaging was performed with the patient erect whenever possible. Anterior, posterior, both laterals, and both posterior and anterior obliques were obtained [24, 25] using a low energy all purpose collimator. The posterior view was done first and 750,000 counts were collected and the time for this image was recorded. All other views were done using the same time as the posterior image. Alternatively, 750,000 counts could be collected for all views except the lateral views. For the laterals, 500,000 counts were collected for the lateral with the best perfusion and the time of this image used for the other lateral. If a low energy high-resolution collimator was used, the counts were changed to 600,000 for all views but the laterals, which are done with 400,000 counts as described above. All images were displayed on film as black images on white background about 2 inches in diameter using film no smaller than 9 × 10 inches. The lungs were about 1.25 inches (3 cm) long (superior to inferior). The views were arranged so that contiguous surfaces of the lung were next to each other (e.g., LPO then POST then RPO, etc.). Ventilation imaging Either a xenon gas (133 Xe) or a radioactive aerosol [radioaerosol of 99m Tc diethylenetriamine pent acetic acid (DTPA) or 99m Tc pyrophosphate (PYP)] was used. Whenever possible, all ventilation as well as all perfu-
PART III
Diagnosis of acute PE
sion imaging was done with the patient upright. If upright images were not possible, supine images were accepted.
Xenon ventilation studies The single-breath image was obtained with the patient initially positioned with his/her back to a wide field of view gamma camera (ideally erect, but often supine). A bolus of approximately 15–20 mCi of 133 Xe was injected into the mouthpiece of the spirometer system at a time when the patient begins a maximal inspiration. The patient held his/her breath for the next 15–20 seconds and a single-breath posterior image was obtained. Following this, equilibrium phase imaging was obtained with the patient breathing in a closed spirometer for 4 minutes. During this interval, a posterior and both right and left posterior oblique images were obtained. The washout phase began at the end of this time. The washout phase lasted at least 5 minutes. Images were taken every 60 seconds during the course of the washout. Posterior oblique views approximately halfway throughout the washout process enhanced the ability to detect and locate xenon retention in the anterior–posterior plane. If the study was done on a dual-headed camera, oblique views in both the equilibrium and washout phases were replaced by simultaneous anterior and posterior views. The study was displayed on film using 2 inch black images on a clear background using film no smaller than 9 × 10 inches. Radioaerosol inhalation studies The radioaerosol most often employed was 99m Tcdiethylenetriamine pentacetate (DTPA). 99m Tcpyrophosphate has also been used and had the advantage of a larger residence time, which was especially important in smokers [26]. Routinely, preperfusion aerosol scanning was obtained. Eight view aerosol images were obtained comparable to the perfusion scan images. The count rate for the aerosol imaging was no more than 25% of the perfusion scan count rate and ideally 20% or less. Aerosol images were made by first obtaining a 200,000–250,000 count posterior image. The time to collect this view was recorded. If any central deposition was noted, the counts collected were about 250,000. All other images were made using the same time as the posterior image, or using the same counts for all images except the laterals timed as described above for the perfusion
BLUK077-Stein
March 28, 2007
CHAPTER 76
15:58
345
Methods of PIOPED II
scan. The images were displayed using the same display technique as used for the perfusion lung scan, and were displayed either above or below the perfusion scan or on a separate sheet of film. Contiguous lung surfaces were displayed next to each other as was done for the perfusion scan. There was no readily apparent difference between the various agents used for ventilation scans from a diagnostic standpoint [27], current practice suggests that about half of nuclear physicians use xenon and half use a radioaerosol. In PIOPED II, each ventilation technique was used.
Chest radiograph Upright 6-foot posterior–anterior chest and lateral radiographs were usually obtained within 2 hours of the
V–Q lung scan but always within 12 hours. If the patient was unable to sit or stand, an anterior–posterior supine chest radiograph was obtained.
Lung scan interpretation It was acceptable to interpret a perfusion alone but only if it was normal. It was also acceptable to interpret a perfusion scan with a chest X-ray and no ventilation scan. In general, the revised PIOPED criteria for high probability were maintained; the intermediate probability criteria and low probability have been revised. The very low probability criteria were not used in PIOPED II. Readings that would have been interpreted as “very low probability” were included among
Table 76.5 V–Q scan diagnostic criteria. High probability
≥ 2 large (>75% of a segment) mismatched perfusion defects substantially larger than corresponding ventilation or radiographic abnormalities, or without any ventilation or radiographic abnormalities. May have ≥2 mismatched moderate segmental (≥25 and ≤75% of a segment) plus 1 large mismatched segmental defect, or ≥4 mismatched moderate segmental perfusion defects.
Intermediate probability
0.5–1.5 mismatched perfusion defects. This may be 1 large plus 1 moderate mismatched perfusion defect, or 1–3 moderate mismatched segmental perfusion defects. Difficult to categorize as high or low. Solitary moderate or large segmental size triple match in lower lobe (zone). Multiple opacities with associated perfusion defects.
Low probability
A single large or moderate size matched segmental defect. >3 small segmental perfusion defects (<25% of a segment) with a normal chest radiograph. Probable PE mimic (absent perfusion in entire lung, solitary lobar mismatch, mass or other chest X-ray lesion(s) causing mismatch(s)). Moderate-sized pleural effusion (> costophrenic angle but <1/3 of pleural cavity) with no other perfusion defect in either lung. Marked heterogeneous perfusion.
Very low probability
Nonsegmental lesion (e.g., prominent hilum, cardiomegaly, elevated diaphragm, linear atelectasis, costophrenic angle effusion with no other perfusion defect in either lung). Perfusion defect smaller than radiographic lesion. ≥2 V–Q matched defects with regionally normal chest X-ray and some areas of normal perfusion elsewhere in the lungs. 1–3 small segmental perfusion defects. Solitary triple matched defect in the mid or upper lung zone confined to a single segment. Stripe sign present around the perfusion defect (best tangential view). Pleural effusion ≥1/3 of the pleural cavity with no other perfusion defect in either lung.
Normal
No perfusion defects present. Perfusion outlines exactly the shape of the lungs as seen on the chest radiograph (hilar and aortic impressions may be seen, chest radiograph and/or ventilation study may be abnormal, e.g., scoliosis).
BLUK077-Stein
March 28, 2007
15:58
346
“low probability.” The final interpretation of each ventilation–perfusion scan was done using techniques comparable to those used in the original PIOPED study. Each central reader received a chest radiograph and a ventilation–perfusion scan from one of the centers (not his own) to interpret with no other data available. Using PIOPED II criteria, he/she assessed the V–Q scan. A second central reader similarly assessed the same V–Q scan. For purposes of PIOPED II, a positive diagnosis required that both readers must make a high probability interpretation. If not, the V–Q scan was sent to a third central reader and majority interpretation ruled. Similarly, a normal interpretation required at least two central readers in agreement. Any reading other than normal or high probability was considered “nondiagnostic” for purposes of the study, and the data saved for ultimate analysis. For exclusion of PE, however, a low probability interpretation in combination with a negative venous ultrasound of the lower extremities, and a low probability clinical assessment by the Wells criteria also excluded PE. Each central reader also had the ability to provide an intuitive percent probability of the V–Q scan for pulmonary embolism in each case and in addition to provide an idea of where in the category his reading lies. For example, if the interpretation is “intermediate probability” the central reader indicated if he thought it was at the low, middle, or high end of that probability category. The PIOPED II criteria are shown in Table 76.5.
Venous compression ultrasound Venous duplex imaging of the lower extremities used B-mode real time venous compression in combination with color Doppler. Guidelines for the diagnosis of acute venous thrombosis are shown in Table 76.6.
Digital subtraction pulmonary angiography
PART III
Diagnosis of acute PE
Table 76.7 Diagnostic criteria for acute pulmonary embolism by digital subtraction angiography. 1 Partial occlusive filling defect within an arterial branch. 2 Completely occlusive filling defect that leads to termination of the column of contrast agent in a meniscus that outlines the trailing edge of the embolus.
location of defects both on the V–Q scan and on the spiral CT. The angiographer had access to the results of both V–Q scan and spiral CT of the lung in order to study in detail areas of suspected PE. The following guidelines were provided for bilateral examinations. If the angiogram was positive on side 1 and the spiral CT was also positive on side 1, the angiographer may stop, although the angiographer was encouraged to continue to side 2, if he/she felt comfortable in doing so. If the angiogram was negative on side 1 but the spiral CT showed a PE on side 1, then the angiographer did sub/superselective injections at the site of the apparent PE shown by spiral CT. If the sub/superselective injections were negative, the angiographer studied side 2. If the sub/superselective injections were positive, then the angiogram was discontinued if desired. If both the angiogram and spiral CT were normal on side 1, the angiographer studied side 2. If the angiogram was positive on side 1 but the spiral CT was negative, the angiographer was encouraged to do side 2, particularly if the spiral CT or V–Q scan showed PE on side 2. Criteria for interpretation of acute PE by DSA are shown in Table 76.7 [28] and criteria for chronic PE by DSA are shown in Table 76.8 [13].
If an angiogram was required by protocol, the angiographer determined which side to do first based on the Table 76.6 Diagnostic criteria for acute deep venous thrombosis by venous ultrasound. Noncompressibility of the vein in combination with: 1 Enlarged size of vein.
Table 76.8 Diagnostic criteria for chronic pulmonary embolism by digital subtraction angiography. 1 Termination of vessel without meniscus. 2 Mural irregularity suggestive of recanalization, webs, or synechiae.
2 Hypoechoic vein lumen.
3 Focal stenosis.
3 Lack of significant collaterals.
4 Focal calcification.
BLUK077-Stein
March 28, 2007
CHAPTER 76
15:58
Methods of PIOPED II
References 1 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 2 Bossuyt PM, Reitsma JB, Bruns DE et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med 2003; 138: W1–W12. 3 Bossuyt PM, Reitsma JB, Bruns DE et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD Initiative. Ann Intern Med 2003; 138: 40–44. 4 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173– 182. 5 Wells PS, Anderson DR, Rodger M et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001; 135: 98–107. 6 PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 7 Turkstra F, Kuijer PMM, van Beek JR et al. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781. 8 Henry JW, Relyea B, Stein PD. Continuing risk of thromboemboli among patients with normal pulmonary angiograms. Chest 1995; 107: 1375–1378. 9 van Beek EJR, Kuyer PMM, Schenk BE et al. A normal perfusion lung scan in patients with clinically suspected pulmonary embolism. Frequency and clinical validity. Chest 1995; 108: 170–173. 10 Kipper MS, Moser KM, Kortman KE et al. Long-term follow-up of patients with suspected pulmonary embolism and a normal lung scan. Perfusion scans in embolic suspects. Chest 1982; 82: 411–415. 11 Hull RD, Raskob GE, Ginsberg JS et al. A noninvasive strategy for the treatment of patients with suspected pulmonary embolism. Arch Intern Med 1994; 154: 289– 297. 12 Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. 13 Nicod P, Peterson K, Levine M et al. Pulmonary angiography in severe chronic pulmonary hypertension. Ann Intern Med 1987; 107: 565–568.
347
14 Perrier A, Miron MJ, Desmarais S et al. Using clinical evaluation and lung scan to rule out suspected pulmonary embolism: is it a valid option in patients with normal results of lower-limb venous compression ultrasonography? Arch Intern Med 2000; 160: 512–516. 15 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. 16 Huisman MV, Buller HR, ten Cate JW et al. Serial impedance plethysmography for suspected deep venous thrombosis in outpatients. The Amsterdam general practitioner study. N Engl J Med 1986; 314: 823–828. 17 Stein PD, Henry JW, Gottschalk A. Reassessment of pulmonary angiography for the diagnosis of pulmonary embolism: relation of interpreter agreement to the order of the involved pulmonary arterial branch. Radiology 1999; 210: 689–691. 18 Lijmer JG, Mol BW, Heisterkamp S et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 1999; 282: 1061–1066. 19 Jaeschke R, Guyatt GH, Sackett DL, for The EvidenceBased Medicine Working Group. Users’ guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? JAMA 1994; 271: 703–707. 20 Sox HC. Commentary. Ann Intern Med 2004; 140: 602. 21 Remy-Jardin M, Remy J. Spiral CT angiography of the pulmonary circulation. Radiology 1999; 212: 615–636. 22 Remy-Jardin M, Remy J, Artaud D et al. Peripheral pulmonary arteries: optimization of the spiral CT acquisition protocol. Radiology 1997; 204: 157–163. 23 Gottschalk A, Alderson PO, Sostman HD. Nuclear medicine techniques and applications. In: Murray JF & Nadel JA, eds. Textbook of Respiratory Medicine, 2nd edn. WB Saunders, Philadelphia, 1994: 682–710. 24 Mack JF, Wellman HN, Saenger EL et al. Clinical experience with oblique views in pulmonary perfusion camera imaging in normal and pathological anatomy. Radiology 1969; 92: 897–902. 25 Caride VJ, Puri S, Slavin JD et al. The usefulness of posterior oblique views in perfusion lung imaging. Radiology 1976; 121: 669–671. 26 Krasnow AZ, Isitman AT, Collier BD et al. Diagnostic applications of radioaerosols in nuclear medicine. In: Freeman LM, ed. Nuclear Medicine Annual 1993. Raven Press, New York, 1993: 123–193. 27 Alderson PO, Biello DR, Gottschalk A et al. Tc-99 mDTPA aerosol and radioactive gases compared as adjuncts to perfusion scintigraphy in patients with suspected pulmonary embolism. Radiology 1984; 153: 515–521. 28 Dotter CT. Acquired abnormalities of the pulmonary arteries. In: Abrams HL, ed. Angiography, 3rd edn. Little Brown, Boston, 1983: 743–761.
BLUK077-Stein
March 28, 2007
77
15:59
CHAPTER 77
Multidetector spiral CT of the chest for acute pulmonary embolism: results of the PIOPED II trial
Introduction The Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II Trial) was designed with two primary objectives: (1) to determine if multidetector CT angiography (CTA) can reliably detect and exclude acute pulmonary embolism (PE) and (2) to determine if adding CT venous phase imaging (CTV) improves the ability to detect and exclude PE [1]. The methods of PIOPED II were discussed in Chapter 76. In this chapter, the results of PIOPED II are described [1]. During the 23-month recruitment period, 824 patients were enrolled and had an adequate reference test (Figure 77.1). Most were judged to have a low or moderate probability of PE based upon the expanded Wells criteria (Table 77.1) [1, 2]. Based upon the composite reference standard, PE was diagnosed in 192 of 824 (23%) of the patients who received a reference diagnosis. The basis for the diagnosis according to the reference standard is shown in Table 77.2. Among 632 patients in whom PE was excluded by the reference standard, the 592 patients who also had an interpretable CTA were followed in 6 months. Among these, 590 did not receive anticoagulants. The clinical courses over 6 months suggested an initially unrecognized PE in 2 of 590 (0.3%).
was 19.6 and the likelihood ratio for a negative test was 0.18. The positive predictive value was 150 of 175 (86%) and the negative predictive value was 567 of 598 (95%) (Table 77.5).
Positive predictive value according to largest pulmonary artery branch with PE Positive predictive values were 116 of 120 (97%) for PE in a main or lobar artery, 32 of 47 (68%) for a segmental vessel, and 2 of 8 (25%) for a subsegmental branch (Table 77.6).
Sensitivity, specificity, likelihood ratios and predictive values of CTA/CTV The CTA/CTV was of insufficient quality for a conclusive interpretation in 87 of 824 patients (11%) (Table 77.4). Among the 737 patients with an adequate CTA/CTV, the sensitivity of CTA/CTV for PE was 164 of 183 (90%) and specificity was 524 of 554 (95%) (Table 77.5, Figures 77.2 and 77.3). The likelihood ratio for a positive test was 16.5 and the likelihood ratio for a negative test was 0.11. The positive predictive value was 164 of 194 (85%) and the negative predictive value was 524 of 543 (97%).
Sensitivity, specificity, likelihood ratios and predictive values of CTA
CT results and clinical assessment
The CTA was of insufficient quality for a conclusive interpretation in 51 of 824 patients (6.2%) (Table 77.3). Among the 773 (94%) with an adequate CTA, the sensitivity of CTA for PE was 150 of 181 (83%) and specificity was 567 of 592 (96%) (Tables 77.3–77.5, Figures 77.2 and 77.3). The likelihood ratio for a positive test
Among patients with a positive CTA and high or intermediate probability prior clinical assessment, the respective positive predictive values for PE were 22 of 23 (96%) and 93 of 101 (92%) (Table 77.7). Among patients with a positive CTA and discordantly low clinical probability, the positive predictive value was
348
BLUK077-Stein
March 28, 2007
CHAPTER 77
15:59
349
Multidetector spiral CT: results of PIOPED II
Screened patients with suspected PE N = 7284
Not eligible 4022 Reasons not eligible, in order of frequency*: Inability to complete testing within 36 hours (1360) Abnormal creatinine (1350) Long-term anticoagulant use (976) Critical illness (802) Ventilatory support (595) Allergic to contrast agents (272) MI within 1 month (229) Could not confirm not pregnant (184) Inferior vena caval filter in situ (169) PE not suspected (67) Upper extremity DVT (57) Previously enrolled in the study (31) VF or sustained VT within 24 hours (30) Shock or hypotension (29) Thrombolytic therapy planned in next 24 hours (20) Age less than 18 years (14) Prisoner (8)
Eligible 3262
Not enrolled 2172 Reasons not enrolled: Patient or clinical team refused (1117) Unable to complete protocol (767) Unspecified reasons (288)
* Reasons are not mutually exclusive (some patients had more than one)
Enrolled 1090
Ref Dx obtained and CT done 824
No Ref Dx 238
CT not done 28
Reasons no diagnosis obtained: Noninvasive testing not performed (63) Inconclusive noninvasive testing and did not undergo pulmonary DSA (175)
PE (by Ref Dx) 192
No PE (by Ref Dx) 632
Figure 77.1 Recruitment diagram for patients in the PIOPED II trial. CT, contrast-enhanced spiral computed tomography; PE, pulmonary embolism; Ref Dx, diagnosis
using the composite reference standard. (Reproduced from Stein et al. [1], with permission.)
BLUK077-Stein
March 28, 2007
15:59
350
PART III
Diagnosis of acute PE
Table 77.1 Demographic characteristics, clinical probability, other illnesses, and presenting signs and symptoms.
Characteristic
Ref Dx obtained and CT done (n = 824)* [n/N (%)]
Female
507/824 (62)
Age (years) (mean ± SD) White*
51.4 ± 16.9
Black
244/823 (30)
Outpatients (including nursing homes and
754/822 (92)
535/823 (65)
rehabilitation centers) Smoking
411/822 (50)
Congestive heart failure
127/823 (15)
Current asthma
125/814 (15)
Chronic obstructive pulmonary disease
68/821 (8)
Current pneumonia
39/773 (5)
Surgery within 3 months
124/824 (15)
Cancer
132/816 (16)
Central venous instrumentation
57/819 (7)
Dyspnea
610/821 (74)
Pleuritic pain
465/619 (75)
Cough
330/820 (40)
Calf pain
203/817 (25)
Hemoptysis
37/816 (5)
Tachypnea (≥20/min)
404/815 (50)
Crackles
151/817 (18)
Tachycardia (>100/min)
147/818 (18)
Calf tender to palpation
71/218 (33)
Swollen calf (>1 cm)
234/646 (36)
Low (0–1.5) Clinical probability
500/796 (63)
Moderate (2–6) Clinical probability
252/796 (32)
High (6.5–12.5) Clinical probability
44/796 (6)
PaO2 (mm Hg) ≥80
83/230 (36)
70–79
39/230 (17)
60–69
44/230 (19)
50–59
41/230 (18)
<50
23/230 (10)
PaCO2 (mm Hg) ≥40
87/228 (38)
36–39
50/228 (22)
<36
91/228 (40)
* Asian/Pacific Islanders, Hispanics, and Native American/Eskimo/Inuit were ≤4% in all groups. PaO2 denotes partial pressure of oxygen in arterial blood while breathing room air. PaCO2 denotes partial pressure of carbon dioxide in arterial blood while breathing room air. Trauma, hemiparesis, diaphoresis, and pleural friction rub were <10% in all groups. Reproduced from Stein et al. [1], with permission.
BLUK077-Stein
March 28, 2007
CHAPTER 77
15:59
351
Multidetector spiral CT: results of PIOPED II
Table 77.2 Basis for diagnosis and exclusion of pulmonary embolism in patients evaluated by computed tomographic pulmonary angiography (n = 824). PE (n = 192) [N (%)] Digital subtraction angiography Ventilation–perfusion scan* Positive leg ultrasound, no prior DVT at same
No PE (n = 632) [N (%)]
33 (17)
192 (30)
109 (57)
146 (23)
50 (26)
—
site, nondiagnostic V–Q Low or very low probability V–Q, low clinical
—
294 (47)
probability† and negative ultrasound * High probability V–Q scan in a patient with no prior PE for diagnosis. Normal V–Q scan for exclusion of PE. † Clinical score by Wells criteria <2 (24). PE, pulmonary embolism; V–Q, ventilation–perfusion lung scan; DVT, deep venous thrombosis. Reproduced from Stein et al. [1], with permission.
Table 77.5 Multidetector CTA and CTA/CTV for acute PE.
Table 77.3 Results of CTA. REF STD PE +
REF STD PE −
CTA [n/N (%)]
CTV [n/N (%)]
Inadequate quality
51/824 (6)
87/824 (11)
Sensitivity
150/181 (83)
164/183 (90)
Specificity
567/592 (96)
524/554 (95)
Positive predictive value
150/175 (86)
164/194 (85)
Negative predictive value
567/598 (95)
524/543 (97)
Likelihood ratio for
19.6
16.5
0.18
0.11
Total
CTA +
150
25
175
CTA −
31
567
598
CTA ?
11
40
51
Total
192
632
824
REF STD PE +, pulmonary embolism (PE) present by the composite reference standard; REF STD PE −, PE absent by the composite reference standard; CTA +, CTA positive for PE; CTA −, CTA negative for PE; CTA ?, CTA of insufficient quality for conclusive interpretation. CTA, computed tomographic pulmonary angiography. Reproduced from Stein et al. [1], with permission.
positive test Likelihood ratio for negative test PE, pulmonary embolism; CTA, computed tomographic pulmonary angiography; CTV, venous phase venogram. Data from Stein et al. [1].
Table 77.4 Results of CTA/CTV. REF STD PE −
100 Total
CTA/CTV +
164
30
194
CTA/CTV −
19
524
543
CTA/CTV ?
9
78
87
192
632
824
Total
REF STD PE +, pulmonary embolism (PE) present by the composite reference standard; REF STD PE −, PE absent by the composite reference standard; CTA/CTV +, either CTA or CTV positive; CTA/CTV −, both CTA and CTV negative; CTA/CTV?, either CTA or CTV negative and the other not done or of insufficient quality for conclusive interpretation. CTA, computed tomographic pulmonary angiography; CTV, venous phase venogram. Reproduced from Stein et al. [1], with permission.
Sensitivity (%)
REF STD PE +
80
90% 83%
60 40 20 0 CTA
CTA/CTV
Figure 77.2 Sensitivity of multislice computed tomographic pulmonary angiography (CTA) and CTA in combination with venous phase imaging (CTA/CTV). (Data from Stein et al. [1].)
BLUK077-Stein
March 28, 2007
15:59
352
PART III
only 22 of 38 (58%); 42% of the CTA readings were falsely positive. In patients with a negative CTA, and low or intermediate probability prior clinical assessment, the negative predictive value for exclusion of PE was 158 of 164 (96%) and 121 of 136 (89%), respectively (Table 77.8). Among patients with a negative CTA and discordantly high clinical probability, the negative predictive value was only 9 of 15 (60%); 40% of the CTA readings were falsely negative. To avoid bias [3], negative predictive values in patients with a low clinical probability were based entirely on a DSA (digital subtraction angiography) or V–Q scan as the reference test. Among patients with a positive CTA/CTV (either a positive CTA or positive CTV) and high or intermediate probability prior clinical assessment, the respective positive predictive values for PE were 27 of 28 (96%) and 100 of 111 (90%) (Table 77.7). Among patients with a positive CTA/CTV and discordantly low clinical probability, the positive predictive value was only 24 of 42 (57%); 43% of the CTA/CTV readings were falsely positive. Among patients with a negative CTA/CTV (a negative CTA and negative CTV), and low or intermediate probability prior clinical assessment, the negative predictive value for exclusion of PE was 146 of 151 (97%) and 114 of 124 (92%), respectively (Table 77.8). Among patients with a negative CTA/CTV and discordantly high clinical probability, the negative predictive value was only 9 of 11 (82%); 18% of the CTA/CTV readings were falsely negative. As with CTA, to avoid bias [3], negative predictive values in patients with a low clinical probability were based entirely on a DSA or V–Q scan as the reference test.
Specificity (%)
100 80
96%
95%
60 40 20 0 CTA
CTA/CTV
Figure 77.3 Specificity of multislice computed tomographic pulmonary angiography (CTA) and CTA in combination with venous phase imaging (CTA/CTV). (Data from Stein et al. [1].)
Diagnosis of acute PE
Table 77.6 Positive predictive value according to largest vessel with PE. Largest vessel
Positive predictive
showing PE
value [n/N (%)]
Main pulmonary artery or lobar
116/120 (97)
Segmental
32/47 (68)
Subsegmental
2/8 (25)
PE, pulmonary embolism. Data from Stein et al. [1].
Calculations of posttest probability using pretest probability and likelihood ratios according to Bayes’ theorem are shown in Figure 77.4. The calculations show that with a discordantly low clinical assessment and positive CTA or CTA/CTV, there is a high proportion of false positive images. Conversely with a discordantly high probability clinical assessment and negative CTA or CTA/CTV there is a high proportion of false negative images. The curves showing posttest probability of CTA and CTA/CTV are similar and nearly overlap. With an intermediate probability clinical assessment, which typically is about a 30–40% clinical probability of PE (see Chapter 52), CTA/CTV gives a somewhat better posttest probability than CTA alone.
Location of venous thrombi Among 105 patients with a positive CTV, thrombi were shown in the inferior vena cava or pelvic veins alone in 3 (3%), thigh veins alone in 89 (85%), and both in 13 (12%) (Table 77.9).
Sensitivity analysis If the composite reference standard is not assumed to be an absolute standard for the diagnosis of PE, but is considered to have its own false positive and false negative rates, the diagnostic accuracy of CTA and CTA/CTV is altered slightly. Sensitivity analysis using the lowest reported false positive and false negative rates of the components of the composite reference test gave an adjusted sensitivity of CTA of 150 of 178 (84%) and adjusted sensitivity of CTA/CTV of 164 of 179 (92%). Using the highest reported false positive and false negative rates, the adjusted value of the sensitivity of CTA was 150 of 182 (82%) and the adjusted sensitivity of CTA/CTV was unchanged at 164 of 183 (90%). Specificities changed ≤1%.
BLUK077-Stein
March 28, 2007
CHAPTER 77
15:59
353
Multidetector spiral CT: results of PIOPED II
Table 77.7 Positive predictive values of CTA and CTA/CTV in relation to prior clinical assessment. High probability
Intermediate probability
Low probability clinical
(Wells score >6)
clinical (Wells score 2–6)
(Wells score <2)
[PE+/CT+ (%)]
[PE+/CT+ (%)]
[PE+/CT+ (%)]
CTA positive
22/23 (96)
93/101 (92)
22/38 (58)
CTA OR CTV positive
27/28 (96)
100/111 (90)
24/42 (57)
PE, pulmonary embolism; CTA, computed tomographic pulmonary angiography; CTV, venous phase venogram. Reproduced from Stein et al. [1], with permission.
Complications Complications of 1095 CTAs were a mild allergic reaction (itching, swollen eyelid, or vomiting) in 4 (0.4%), urticaria in 1 (0.1%), and moderately severe extravasation of contrast material into the antecubital fossa in 2 (0.2%). One patient with diabetes mellitus suffered a transient episode of acute renal failure characterized by an increase of his serum creatinine from 1.3 to 2.9 mg/dL after a CTA/CTV followed in 22 hours by a DSA. The elevated creatinine returned to normal with intravenous fluids. No other complications were reported with 209 DSAs or with any other reference tests. No other elevations of creatinine were attributed to the procedures. Serum creatinine levels were typically checked daily in hospitalized patients, but were not required by protocol and typically were not obtained in outpatients following CTAs and DSAs. Clinically obtained creatinine levels >2 times baseline were recorded in the PIOPED II database. These data show that multidetector CTA/CTV had a higher sensitivity (90%) than CTA alone (83%) with comparable specificity, about 95% for both. Sensitivity was based on the number of patients who had conclu-
sive interpretations of CTA or CTA/CTV. The sensitivity of diagnostic imaging with CTA and CTA/CTV would be lower if patients with inconclusive interpretations due to poor image quality were included. The reported data are primarily with 4-slice CT. Not enough patients with 8- or 16-slice scanners were studied to determine whether accuracy improved with more advanced scanners. Both multislice CTA/CTV and multislice CTA alone would require additional testing to diagnose or exclude PE if the prior clinical probability were discordant with the imaging results. In PIOPED II, there were 711 patients with adequate quality CTV and venous ultrasound. Both CTVs and ultrasounds were negative in 598 of 711 (84%), both were positive in 81 of 711 (11%), CTV was discordantly positive in 17 (2%), and discordantly negative in 15 (2%) [4]. Overall, there was over 95% (679/711) agreement between CTV and ultrasound. This suggests that venous ultrasound and CT venography are equally valid in the diagnosis of patients with suspected PE. Strengths of the PIOPED II investigation include incorporation of all of the standards for reporting diagnostic accuracy (STARD criteria) [5, 6]. The composite
Table 77.8 Negative predictive values of CTA and CTA/CTV in relation to prior clinical assessment. High probability
Intermediate probability
Low probability clinical
(Wells score >6)
clinical (Wells score 2–6)
(Wells score <2)
[PE−/CT− (%)]
[PE−/CT− (%)]
[PE−/CT− (%)]
CTA negative
9/15 (60)
121/136 (89)
158/164 (96)
CTA and CTV negative
9/11 (82)
114/124 (92)
146/151 (97)
To avoid bias for calculation of the negative predictive value in patients with a low probability prior clinical assessment, only patients with a reference test diagnosis by V–Q scan or conventional pulmonary digital subtraction angiography were included. PE, pulmonary embolism; CTA, computed tomographic pulmonary angiography; CTV, venous phase venogram. Reproduced from Stein et al. [1], with permission.
BLUK077-Stein
March 28, 2007
15:59
PART III
120 100
CTA pos CTA/CTV pos
80 60 40
CTA neg
20
CTA/CTV neg
90 10 0
80
70
60
50
40
30
20
10
0 0
Posttest probability after CTA or CTA/CTV (%)
354
Pretest probability (%) Figure 77.4 Posttest probability of multislice computed tomographic pulmonary angiography (CTA) and CTA in combination with venous phase imaging (CTA/CTV) according to pretest clinical assessment. Curves were drawn on the basis of calculations of clinical probability and likelihood ratios according to Bayes’ theorem. (Data from Stein et al. [1].)
Diagnosis of acute PE
agnosis including respiratory motion artifact, image noise, pulmonary artery catheter, flow-related artifact, window settings, streak artifact, lung algorithm artifact, partial volume artifact, stair step artifact, partial volume averaging effect in lymph nodes, vascular bifurcation, misidentification of view, mucus plug, perivascular edema, localized increase in vascular resistance, pulmonary artery stump, in situ thrombosis, primary pulmonary artery sarcoma, and tumor emboli [9]. PIOPED II did not assess the performance of non-subspecialist radiologists. Art and skill play an important role [8]. However, there are many expert clinical radiologists outside of academic centers who can achieve high accuracy in CT interpretation [8]. Some have recommended outsourcing readings of CT by international teleradiology, but this is controversial [10].
References reference test was shown to be robust by sensitivity analysis and by the generally benign outcome of patients with a negative reference test. For readers of CTA, the single unweighted kappa statistic [7] for agreement between the first two readers was 0.73 [8]. For CTV, the kappa for acute DVT was 0.73. Computed tomography, therefore, provided substantial inter-observer agreement. To determine whether having two readers influenced the results of PIOPED II, the sensitivity and specificity of the first central reader of each patient were evaluated [8]. The results were similar to the results of consensus readings. The sensitivity of CTA based on interpretations of the first central reader was 80% and specificity was 95%. The sensitivity of CTA/CTV based on interpretations of the first central reader was 89% and specificity was 94%. The readers of CT images were experienced university radiologists. There are many causes of misdi-
Table 77.9 Location of venous thrombi on CTV. Site
Prevalence [n/N (%)]
IVC or pelvic veins only
3/105 (3)
Thigh veins only
89/105 (85)
Pelvic and thigh
13/105 (12)
CTV, venous phase venogram; IVC, inferior vera cana. Data from Stein et al. [1].
1 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 2 Wells PS, Anderson DR, Rodger M et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001; 135: 98–107. 3 Sackett DL. Bias in analytic research. J Chronic Dis 1979; 32: 51–63. 4 Goodman LR, Beemath A, Stein PD, Sostman HD, Wakefield T, Woodard PK. Evaluating deep venous thrombosis: lessons from PIOPED II, submitted. 5 Bossuyt PM, Reitsma JB, Bruns DE et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med 2003; 138: W1–W12. 6 Bossuyt PM, Reitsma JB, Bruns DE et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD Initiative. Ann Intern Med 2003; 138: 40–44. 7 Landis RJ, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159–174. 8 Stein PD, Goodman LR, Sostman HD. Response to letter. New Engl J Med 2006; 354: 955–956. 9 Wittram C, Maher MM, Yoo AJ et al. CT angiography of pulmonary embolism: diagnostic criteria and causes of misdiagnosis. Radiographics 2004; 24: 1219–1238. 10 Wachte RM. International Teleradiology. N Engl J Med 2006; 354: 662–663.
BLUK077-Stein
April 6, 2007
78
16:22
CHAPTER 78
Outcome studies of pulmonary embolism versus accuracy
It is important to distinguish between investigations of accuracy of a diagnostic test and studies of outcome based on the results of a negative test [1]. The results do not equate [1]. Results of outcome studies may be entirely different than results of accuracy studies. Differences need not imply a methodological error. Outcome studies provide an appropriate measurement for clinical management, while accuracy is needed for research purposes. In the Prospective Investigation of Pulmonary Embolism II (PIOPED II), among patients with an intermediate probability clinical assessment, negative multidetector contrastenhanced computed tomographic (CT) angiograms of the chest failed to identify 11% of patients with pulmonary embolism (PE) [2]. This is compatible with results of well-performed outcome studies that showed recurrent PE in 1% or fewer untreated patients with a negative CT pulmonary angiogram providing that one recognizes that recurrent PE occurs in only a fraction of patients with acute PE [1]. The Christopher Group, for example, showed PE on 3-month followup of untreated patients with a negative CT angiogram in only 10 of 1436 (0.7%) [3]. An additional 8 of 1436 (0.6%) had deep venous thrombosis (DVT) on followup. Perrier and associates showed PE at 3 months in 3 of 294 (1.0%) of patients with low or intermediate probability clinical assessments, and an additional 4 of 294 (1.4%) had DVT, some of which was shown by compression ultrasound during initial testing [4]. The following data support the fact that outcome studies do not show recurrent PE in all untreated patients. It is reasonable to assume that the PEs were small in the 11% of patients with an intermediate probability clinical assessment in PIOPED II in whom CT angiograms were falsely negative. The positive predictive value of PE in main or lobar arteries was 97%, in segmental pulmonary arteries 68%, and in subsegmental pulmonary artery branches 25% [2]. It is unlikely,
therefore, that large PE in main or lobar pulmonary arteries were not identified. In the prior era of late diagnoses of PE based on clinical findings, Hermann and associates reported a 36% frequency of fatal recurrent PE [5]. There was, in addition, a 21% frequency of nonfatal recurrent PE among untreated patients with clinically diagnosed PE [5]. The PE was severe in these patients. Mortality from the initial PE was 37% [5]. With early diagnosis, the risk of recurrent PE is lower. In PIOPED I, 20 patients did not receive treatment because the diagnosis was not established until later when central readers identified the PE on pulmonary angiograms [6]. Among these 20 untreated patients, fatal PE occurred in 1 (5%) and 1 patient had a nonfatal PE on follow-up [6]. Therefore, 10% of patients with untreated mostly mild PE suffered a PE on follow-up. The rate of recurrence of PE from data on the outcome of patients with suspected PE in whom treatment was withheld following nondiagnostic ventilation– perfusion lung scans and normal serial noninvasive leg tests can be calculated. For example, among 711 patients with suspected PE and nondiagnostic ventilation–perfusion lung scans evaluated by Hull and associates, 3-month outcome of untreated patients with negative serial impedance plethysmography showed PE in 4 of 627 (0.6%) [7]. Among 711 patients with nondiagnostic V–Q scans, 22% (156 patients) would have had PE based on results of PIOPED I [8]. Since 84 patients were identified with impedance plethysmography and treated, the group of 627 who were followed with no anticoagulant therapy would have included 74 patients with PE. On follow-up, 4 of 74 (5.4%) showed recurrent PE. Similar calculations and similar results are shown in data of Wells et al. [9]. Among 702 patients with suspected PE, nondiagnostic ventilation–perfusion lung scans and a low or moderate pretest clinical probability
355
BLUK077-Stein
April 6, 2007
16:22
356
of PE, outcome at 3 months in those with negative serial compression ultrasound showed PE in 3 of 665 (0.5%) [9]. If, as in PIOPED I, 22% of patients with nondiagnostic ventilation–perfusion lung scans had PE, then 154 of 702 would have had PE. Among these, 37 patients were identified with venous compression ultrasound and treated. The group of 665 who were followed with no anticoagulant therapy, therefore, would have included 117 patients with PE. On follow-up, 3 of 117 (2.6%) showed recurrent PE. Smith et al. showed PE during an average of 8.4 months follow-up in 1 of 173 untreated patients (0.6%) with low probability ventilation–perfusion lung scans [10]. Assuming that 14% of patients with low probability lung scans would have had PE, as in PIOPED I, the rate of PE on follow-up would have been 1 of 24 (4.2%). Finally, Kahn et al. showed no PE on 1-year follow-up of 90 patients with a low probability interpretation of ventilation–perfusion lung scans [11]. If 13 of these patients would have had PE, based on the rate shown in PIOPED I, the rate of recurrent PE would have been 0 of 13 (0%). In these five investigations, therefore, recurrent PE is estimated to have occurred in 0, 2.6, 4.2, 5.4, and 10% of untreated patients in whom the initial PE was mild [6, 7, 9–11]. In PIOPED II, among 31 patients in whom the CT angiogram did not show PE, the Outcome Committee did not identify any PE on 3-month follow-up (95% CI = 0–12.2). Based on the above calculations, perhaps as many as 3 patients would have shown PE on follow-up, which is well within the 95% confidence interval of this observation. If 5.4% of untreated patients develop symptomatic recurrent PE, then, in the outcome studies by Perrier et al. [4], and in the Christopher study [3], if a CTA failed to identify 18% of patients with PE, outcome would have shown symptomatic PE on follow-up in <1%. If 10% of untreated patients develop symptomatic recurrent PE, then if a CT angiogram failed to identify 10% of patients with PE, 1% would have shown PE on follow-up. The percentage of patients with PE who were not identified by CT angiography in PIOPED II, therefore, was within the range calculated
PART III
Ventilation–perfusion lung scans
to have been present in the outcome studies of Perrier et al. [4], and in the Christopher study [3].
References 1 Stein PD, Beemath A, Goodman LR, et al. Outcome studies of pulmonary embolism versus accuracy: they do not equate? Thromb Haemost 2006; 96: 107–108. 2 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 3 van Belle A, Buller HR, Huisman MV et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA 2006; 295: 172–179. 4 Perrier A, Roy P-M, Sanchez O et al. Multidetectorrow computed tomography in suspected pulmonary embolism. N Engl J Med 2005; 352: 1760–1768. 5 Hermann RE, Davis JH, Holden WD. Pulmonary embolism: a clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961; 102: 19–28. 6 Stein PD, Henry JW, Relyea B. Untreated patients with pulmonary embolism: outcome, clinical and laboratory assessment. Chest 1995; 107: 931–935. 7 Hull RD, Raskob GE, Ginsberg JS et al. Noninvasive strategy for the treatment of patients with suspected pulmonary embolism. Arch Intern Med 1994; 154: 289–297. 8 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 9 Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. 10 Smith R, Maher JM, Miller RI, Alderson PO. Clinical outcomes of patients with suspected pulmonary embolism and low-probability aerosol-perfusion scintigrams. Radiology 1987; 164: 731–733. 11 Kahn D, Bushnell DL, Dean R, Perlman SB. Clinical outcome of patients with a ‘low probability’ of pulmonary embolism on ventilation–perfusion lung scan. Arch Intern Med 1989; 149: 377–379.
BLUK077-Stein
March 12, 2007
79
17:13
CHAPTER 79
Contrast-induced nephropathy
Radiocontrast nephropathy is a common cause of hospital-acquired renal failure and has been associated with increased in-hospital mortality and length of stay [1, 2]. In the Prospective Investigation of Pulmonary Embolism I (PIOPED I), among 1111 patients who underwent pulmonary angiography with average doses of 181–207 mL of high osmolar, ionic 76% contrast material, 3 (0.3%) developed acute renal failure and required dialysis [3]. The preangiography serum creatinine values in these patients were 1.2, 2.4, and 2.7 mg/dL. In addition, 10 patients (0.9%) developed an elevation of the serum creatinine, and were treated with fluid balance, diuretics, and in 1 patient, dopamine. Among these patients, 5 had previously normal creatinine levels that increased to 2.1– 5.9 mg/dL and 5 had abnormal levels that ranged from 1.5 to 2.7 mg/dL and increased ≥2 mg/dL. Patients who developed contrast-induced nephropathy were older (average age 74 years) than those who did not (average age 57 years). In PIOPED II, only 1 of 824 patients who had CT angiography (0.1%) developed renal failure [4]. This patient had diabetes mellitus, and the serum creatinine increased from 1.3 to 2.9 mg/dL after a CT pulmonary angiogram followed in 22 hours by a standard digital subtraction pulmonary angiogram. The elevated serum creatinine returned to normal with intravenous fluids. Nonionic contrast material was used in PIOPED II [4]. Patients with abnormal serum creatinine levels were excluded. Nonionic contrast material appears to be less nephrotoxic [5] and generally better tolerated [6] than ionic contrast material, although no difference of nephrotoxicity has been reported [7] (Table 79.1). The risks of renal dysfunction from contrast material in patients with serum creatinine levels ≥1.1 mg/dL can be reduced with prophylactic hydration [8] (Table 79.1). Hydration with sodium bicarbonate before contrast exposure has been reported to be more effective than hydration with sodium chloride [8]. An isotonic solution of sodium bicarbonate 3 mL/kg per hour for
1 hour before and 1 mL/kg per hour for 6 hours after the administration of contrast material is recommended [8, 9]. Nonsteroidal anti-inflammatory drugs and dipyridamole should be discontinued as early as possible before CT angiography (Table 79.1) [10]. Results with angiotensin converting enzyme inhibitors have been equivocal [10]. In PIOPED II, they were discontinued as early as possible before CT angiography. Metformin (Glucophage) should be discontinued before CT angiography because if contrast-induced renal failure occurs, metformin accumulation in body tissues could cause lactic acidosis [11]. Metformin, however, does not cause renal failure. In emergency or urgent situations, if renal function is normal, the study may proceed with little risk. If renal function is abnormal or unknown, metformin should be discontinued, and hydration as well as other precautions listed above should be taken [11]. Therapy with metformin can be resumed when renal function has been shown to be normal [10, 11]. Prophylactic administration of the antioxidant acetylcysteine (Mucomist) initially showed promise, along with hydration, for the prevention of contrastagent-induced reductions of renal function in patients with chronic renal insufficiency [12, 13]. Several authors, however, cast doubt on the benefit of
Table 79.1 Recommendations for prevention of contrast-induced nephropathy. All patients
r r r r
Use nonionic contrast material Discontinue nonsteroidal anti-inflammatory drugs Discontinue dipyridamole Discontinue metformin—to prevent lactic acidosis
Patients with serum creatinine ≥1.1 mg/dL
r
Hydrate with isotonic sodium bicarbonate 3 mL/kg/hr × 1 hr before contrast 1 mL/kg/hr × 6 hr after contrast
357
BLUK077-Stein
March 12, 2007
17:13
358
N-acetylcysteine, as individual studies and several meta-analyses reached conflicting conclusions [9]. Fendolopam increases both medullary and renal cortical blood flow [14]. Its use in high-risk patients appeared to minimize the likelihood of radiocontrast nephropathy [15]. Further clinical trials in high-risk populations have been recommended [10].
PART III
8
9 10
References 1 Levy EM, Viscoli CM, Horwitz RI. The effect of acute renal failure on mortality. A cohort analysis. JAMA 1996; 275: 1489–1494. 2 McCullough PA, Wolyn R, Rocher LL, Levin RN, O’Neill WW. Acute renal failure after coronary intervention: incidence, risk factors, and relationship to mortality. Am J Med 1997; 103: 368–375. 3 Stein PD, Athanasoulis C, Alavi A et al. Complications and validity of pulmonary angiography in acute pulmonary embolism. Circulation 1992; 85: 462–468. 4 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 1996; 354: 2317–2327. 5 Harris KG, Smith TP, Cragg AH, Lemke JH. Nephrotoxicity from contrast material in renal insufficiency: ionic versus nonionic agents. Radiology 1991; 179: 849–852. 6 Barrett BJ, Carlisle EJ. Metaanalysis of the relative nephrotoxicity of high- and low-osmolality iodinated contrast media. Radiology 1993; 188: 171–178. 7 Schwab SJ, Hlatky MA, Pieper KS et al. Contrast nephrotoxicity: a randomized controlled trial of a nonionic and
11
12
13
14
15
Ventilation–perfusion lung scans
an ionic radiographic contrast agent. N Engl J Med 1989; 320: 149–153. Merten GJ, Burgess WP, Gray LV et al. Prevention of contrast-induced nephropathy with sodium bicarbonate: a randomized controlled trial. JAMA 2004; 291: 2328– 2334. Chertow GM. Prevention of radiocontrast nephropathy: back to basics. JAMA 2004; 291: 2376–2377. Waybill MM, Waybill PN. Contrast media-induced nephrotoxicity: identification of patients at risk and algorithms for prevention. J Vasc Interv Radiol 2001; 12: 3–9. Heupler FA, Jr. Guidelines for performing angiography in patients taking metformin. Members of the Laboratory Performance Standards Committee of the Society for Cardiac Angiography and Interventions. Cathet Cardiovasc Diagn 1998; 43: 121–123. Tepel M, van der Giet M, Schwarzfeld C, Laufer U, Liermann D, Zidek W. Prevention of radiographic-contrastagent-induced reductions in renal function by acetylcysteine. N Engl J Med 2000; 343: 180–184. Safirstein R, Andrade L, Vieira JM. Acetylcysteine and nephrotoxic effects of radiographic contrast agents—a new use for an old drug. N Engl J Med 2000; 343: 210– 212. Kien ND, Moore PG, Jaffe RS. Cardiovascular function during induced hypotension by fenoldopam or sodium nitroprusside in anesthetized dogs. Anesth Analg 1992; 74: 72–78. Madyoon H, Croushore L, Weaver D, Mathur V. Use of fenoldopam to prevent radiocontrast nephropathy in high-risk patients. Catheter Cardiovasc Interv 2001; 53: 341–345.
BLUK077-Stein
March 28, 2007
80
16:0
CHAPTER 80
Radiation exposure and risk
Introduction Radiation for diagnostic tests for pulmonary embolism is not trivial. In low-yield settings, consideration of alternate diagnostic imaging approaches that minimize exposure to radiation should be considered [1]. Relevant definitions are as follows [2, 3]: Exposure [roentgens (R)] = amount of energy produced by X-ray or gamma radiation in a cubic centimeter of air. It is determined by measuring ionization in air caused by the X-ray beam. Units: 1 roentgen (R) = 2.58 × 10−4 coulomb/kg (C/kg) of air Absorbed dose (joules/kg) is the energy deposited in a volume of tissue by the radiation beam divided by the mass of the tissue. Units: 1 joule/kg = 1 gray (Gy) = 100 rad 1 mGy = 0.1 rad Equivalent dose [sievert (Sv)] = absorbed dose (Gy) × radiation weighting factor. Equivalent dose is a modification of absorbed dose that includes weighting factors to account for the different biological effects of various sources of radiation. Biological effects depend on type of radiation delivered. For X-ray, weighting factor = 1 Therefore, for X-ray, absorbed dose (Gy) = equivalent dose (Sv) Units: 1 sievert (SV) = 100 roentgen equivalents in man (rem) 1 mSv = 0.1 rem For X-ray and gamma photons, 1 rem = 1 rad Effective dose (Sv) = equivalent dose × tissue specific weighting factor. Effective dose is a measurement that estimates the whole body dose that will be required to cause the same random effects risk as the partial body dose that was actually delivered in a localized radiologic procedure. The tissue-specific weighting factor accounts for the risk of cancer induction for a specific organ. “Effective dose” is the sum of the products of the equivalent dose
times the weighting factor for all irradiated organs. Tissue weighting factors are gonads, 0.2; red bone marrow, colon, lung, stomach, 0.12; bladder, breast, liver esophagus, thyroid 0.05; skin, bone surface, 0.01; remainder, 0.05 [4]. The biological effects of radiation vary for differing types of and energy radiation, including X-ray, gamma ray, alpha particles, beta particles, and neutrons. The biological effects depend on the rate of energy deposition within the cell. For example, alpha particles are only weakly penetrating and can be stopped by thin paper [5]. However, they can produce more biological damage than x-radiation because all of the energy may be lost within the space of a single cell. The direct measurement of effective dose is not strictly possible in diagnostic radiology [5]. The measurement needs to take into account the degree of penetration of the X-ray beam, and the penetration of irradiated organs. The dose–length–product (DLP) gives a measure related to effective dose. It is the product of the entrance dose and the length of tissue irradiated. The entrance dose is measured by means of an ionization chamber affixed to the collimator assembly on the X-ray tube through which the X-ray beam passes. The chamber is attached to an electronic readout. For spiral CT, the effective dose of radiation can be estimated from the DLP displayed on the console of some brands of equipment.
Effective (whole body) dose with CT Published literature for the effective dose of CT pulmonary angiography ranges from 1.6 to 8.3 mSv [3, 6–9] (Table 80.1). Much of the literature is based on observations in only a few patients (Table 80.1). Multidetector CT resulted in a dose profile approximately 27% higher than single detector CT [10].
359
BLUK077-Stein
March 28, 2007
16:0
360
PART III
Diagnosis of acute PE
Table 80.1 Effective whole body dose. First author
Number of
Number of
Effective dose
[Ref]
slice CT
patients studied
(mSv) (range)
Kuiper [6]
4
27
4.2 (2.2–6.0)
O’Neil [7]
1
16
1.6 (1.4–1.9)
Mayo [3]
1
7.0
Pitch 1
Mayo [3]
1
3.5
Pitch 2
Resten [8]
1
6.4
Phantom
Wittram [9]
4
10
8.3
Chest
4
10
5.7
Venous phase pelvis and legs
Comment
CT pulmonary angiography
CT venous phase imaging Wittram [9]
The effective dose of venous phase imaging of the pelvis and thighs is based on only 10 patients studied with 4-slice CT [9] (Table 80.1).
Effective (whole body) dose radiation with digital subtraction angiography The effective dose of radiation to which patients are exposed with pulmonary digital subtraction angiography (DSA) ranges from 3.2 to 30.1 mSv [3, 6–9] (Table 80.2).
The effective (whole body) dose with pulmonary artery and lateral chest radiograph The effective dose with a chest pulmonary artery (PA) and lateral radiograph is 0.07 mSv [11] (Table 80.3). Most of the effective dose is from the lateral projection, 0.06 mSv, with only a small portion from the PA, about 0.02 mSv [11].
Table 80.2 Digital subtraction pulmonary angiography: effective whole body dose. First author
Number of
Effective dose
[Ref]
patients studied
(mSv) (range)
Kuiper [6]
12
7.1 (3.3–17.3)
O’Neil [7]
3
3.2 (2.3–4.1)
Mayo [3]
6.0
Resten [8] Wittram [9]
Comment
28 10
Phantom
30.1
Table 80.3 Chest radiograph: effective whole body dose. First author
Effective dose
[Ref]
(mSv) (range)
Comment
Huda [11]
0.07
PA and lateral
Huda [11]
0.02
PA
Huda [11]
0.06
Lateral
PA, pulmonary artery.
The effective (whole body) dose with a perfusion lung scan The effective dose with a perfusion lung scan is 0.8 mSv [7] (Table 80.4), and 1.2–1.95 mSv with a ventilation–perfusion lung scan combination [7, 9] (Table 80.5). In order to have an intuitive appreciation of the effective dose of radiation with various diagnostic tests,
Table 80.4 Perfusion lung scan: effective whole body dose. First author
Number of
Effective dose
[Ref]
patients studied
(mSv) (range)
O’Neil [7]
30
0.8
BLUK077-Stein
March 28, 2007
CHAPTER 80
16:0
361
Radiation exposure and risk
Table 80.5 Ventilation–perfusion lung scan: effective whole body dose. First author
Number of
Effective dose
[Ref]
patients studied
(mSv) (range)
O’Neil [7]
30
1.2
Wittram [9]
10
1.95
the estimated equivalent number of chest radiographs for each test is shown in Table 80.6. The effective dose of the individual diagnostic tests was based on values shown in Tables 80.1–80.5.
Background radiation and allowable yearly effective dose in radiation workers The average yearly effective dose of natural background radiation is 2.5 mSv [5]. Regarding occupational exposure, the International Commission of Radiological Protection recommends that the effective dose of radiation should not exceed 50 mSv in any single year and should not exceed 100 mSv in 5 years [4].
Risks The risk of inducing a malignancy by exposure to radiation is estimated to be 5 cases of malignancy for
each 100,000 individuals exposed to an effective dose of 1 mSv [4]. The risk is estimated to be higher in infants and children and lower in the elderly [4]. The data on which these figures are based are extrapolated from relatively high levels of radiation exposure [ 4]. It is uncertain whether there is a linear relation in lower levels of radiation [12]. If one assumes a linear proportion with a low level of radiation, an effective whole body dose of 8 mSv, as might occur with a CT pulmonary angiogram, would lead to 4 malignancies in 10,000 exposed patients and a CT pulmonary angiogram with CT venous phase imaging might lead to as many as 7 malignancies in 10,000 exposed patients. It is assumed that no lower threshold exists for radiation-induced carcinogenesis [4]. Female breast radiation exposure is a particular concern. Using single-slice CT and 4-slice CT, the calculated dose for a 60 kg woman was ≥20 mGy per breast [13] (Table 80.7). Others calculate a breast exposure of 10 mGy [14]. According to the International Commission on Radiological Protection the breast dose in CT of the thorax may be as much as 30–50 mGy [15]. The average glandular breast dose for single-view mammography must not exceed 3 mGy [16]. On average, it is about 3 mGy for 2-view mammography [1, 13]. Women with a cumulative absorbed dose of >100 mGy from repeated fluoroscopic examinations for tuberculosis had a relative risk from breast cancer
Table 80.6 Estimated equivalent number of PA and lateral chest radiographs for diagnostic background radiation and industrial exposure.
Examination
Effective (whole body)
Equivalent number of PA
dose (mSv)
and lateral chest radiographs
References
Chest PA and lateral
0.07
1
Perfusion lung scan
0.8
11
[11] [7]
Ventilation–perfusion lung scan
1.2–2.0
17–28
[9]
CT chest
1.6–8.3
23–119
[3, 6–9]
CTV
5.7
81
[9]
Pulmonary DSA
3.2–30.1
46–430
[3, 6–9]
Natural yearly background
2.5
36
[5]
50
714
[4]
20
286
[4]
radiation Allowable yearly maximal exposure in radiation workers Average allowable exposure in radiation workers PA, pulmonary artery; CTV, computed tomographic venous phase imaging; DSA, digital subtraction angiography. Modified from Stein PD, Woodard PK, Weg JG, et al. Diagnostic pathways in acute pulmonary embolism: Recommendations of the PIOPED II investigators. Am J Med 2006; 119: 1048–1055 with permission from Elsevier Inc.
BLUK077-Stein
March 28, 2007
16:0
362
PART III
Table 80.7 Female breast radiation.
Diagnosis of acute PE
Table 80.8 Fetal exposure. First author
CT chest
V–Q scan
Perfusion
Examination
Glandular breast dose (mGy)
References
[Ref]
(mGy)
(mGy)
scan (mGy)
0.25–0.36
0.21–0.30
1-view mammogram
≤3
[16]
Hurwitz [22]
0.24–0.66
2-view mammogram
3
[1, 13]
Cook [14]
0.01
CT chest
10–50
[13–15]
Perfusion lung scan
0.28
[14]
0.12
V–Q = ventilation–perfusion.
PA, pulmonary artery.
of 1.36 as compared to those exposed to less than 100 mGy [17]. The risk was age-dependent. For girls aged 10–14 years, the relative risk was 4.46. For women over 35 years of age, the relative risk was 1.10.
Genetic effects Very few quantitative data are available on radiogenic mutations in human germ cells, particularly from lowdose exposure [4]. Serious genetic disease includes inherited ill health, handicaps, disabilities, and shortened lifespan. Genetic disease may be manifest at birth or may not become evident until some time in adulthood. The frequency of radiation-induced genetic impairment is relatively small in comparison with the magnitude of detriment associated with spontaneously arising genetic disease [19, 20]. Genetic studies carried out on the children of atomic bomb survivors in Japan showed no difference in any of the indicators of genetic damage including untoward pregnancies, survival of children to their mid-20s, malignancies, chromosomal rearrangements, mutations affecting protein charge or function, growth and development of infants, and sexratio shifts [21]. Absorbed doses were as high as 4000– 5000 mGy [21]. Radiation is a well-known teratogenic agent [4]. The developing fetus is much more sensitive to radiation than the mother [4].
Fetal exposure Some indicate the radiation dose to the fetus from 16slice CTA, 0.24–0.47 mGy at 0 months, and 0.61–0.66 mGy at 3 months, is of the same magnitude as a V–Q scan (0.36 mGy at 0 months and 0.32 mGy at 3 months) or a perfusion scan alone (0.21 mGy at 0 months and 0.30 mGy at 3 months) [22] (Table 80.8). Others in-
dicate that the absorbed dose to the fetus is less with CTA than a perfusion scan (0.01 mGy versus 0.12 mGy) [14]. However, the absorbed dose to the breast was higher with CTA (10–50 mGy) [13–15] than for a perfusion scan (0.28 mGy) [14].
Methods for decreasing radiation dose 1 Reduction of tube current. For plane chest CT, it was shown that the usual dose of 80–300 mAs could be reduced to 10–50 mAs [12]. 2 Increase table increment. The ratio of table increment to slice collimation (pitch) can be increased, thereby reducing exposure time and exposure level. 3 Reduction of tube voltage. This is often not possible [12]. An unfavorable ratio of absorbed radiation and signal-to-noise ratio may result. Tube voltage less than 110–120 kV appears not to be useful [12]. 4 Use discontinuous images for CT venous phase imaging and scan from the acetabulum to the tibial plateau rather than from the iliac crest (see Chapter 37).
References 1 Nickoloff EL, Alderson PO. Radiation exposures to patients from CT: reality, public perception, and policy. Am J Roentgenol 2001; 177: 285–287. 2 U.S. Food and Drug Administration. Whole body scanning using computed tomography (CT): radiation quantities and units. http://www.fda.gov/CDRH/CT/rqu. html. Updated April 17, 2002. 3 Mayo JR, Aldrich J, M¨uller NL. Radiation exposure at chest CT: a statement of the Fleischner society. Radiology 2003; 228: 15–21. 4 Annals of the International Commission of Radiological Protection (ICRP) International Commission on Radiological Protection. ICRP Publications 60: 1990
BLUK077-Stein
March 28, 2007
CHAPTER 80
5
6
7
8
9
10
11
12
16:0
Radiation exposure and risk
Recommendations of the International Commission on Radiological Protection, 60. Ann ICRP 1991; 21: 1–3. Robinson A. Radiation protection and patient doses in diagnostic radiology. In: Grainger RG & Allison D, eds. Grainger and Allison’s Diagnostic Radiology, 3rd edn. Churchill Livingstone, New York, 1997: 169–183. Kuiper JW, Geleijns J, Matheijssen NAA, Teeuwisse W, Pattynama PMT. Radiation exposure of multi-row detector spiral computed tomography of the pulmonary arteries: comparison with digital subtraction pulmonary angiography. Eur Radiol 2003; 13: 1496–1500. O’Neill J, Murchison JT, Wright L, Williams J. Effect of the introduction of helical CT on radiation dose in the investigation of pulmonary embolism. Br J Radiol 2005; 78: 46–50. Resten A, Mausoleo F, Valero M, Musset D. Comparison of doses for pulmonary embolism detection with helical CT and pulmonary angiography. Eur Radiol 2003; 13: 1515–1521. Wittram C, Liu B, Callahan RJ, Hales C, Quinn DA, Stein PD. An estimate of the radiation dose received per patient for the investigation of pulmonary venous thromboembolism based on the PIOPED II data [Abstract 4408707]. Presented at the Radiol Soc North Amer. Chicago, November 2005. Thornton FJ, Paulson EK, Yoshizumi TT, Frush DP, Nelson RC. Single versus multi-detector row CT: comparison of radiation dose and dose profile. Acad Radiol 2003; 10: 379–385. Huda W, Sourkes AM. Radiation doses from chest X-rays in Manitoba (1979 and 1987). Radiat Prot Dosim 1989; 28: 303–308. Diederich S, Lenzen H. Radiation exposure associated with imaging of the chest. Comparison of different radiographic and computed tomography techniques. Cancer 2000; 89(suppl): 2457–2460.
363
13 Parker MS, Hui FK, Camacho MA. Female breast radiation exposure during CT pulmonary angiography. Am J Roentgenol 2005; 185: 1228–1233. 14 Cook JV, Kyriou J. Radiation from CT and perfusion scanning in pregnancy. BMJ 2005; 331: 350. 15 International Commission on Radiological Protection. ICRP Publication 87: managing patient dose in computed tomography, 87. Ann ICRP 2002; 30: 4. 16 Am College Radiol (ACR). ACR Guideline for the Performance of Screening Mammography. ACR, Reston, VA, 2004: 317–328. 17 Miller AB, Howe GR, Sherman GJ et al. Mortality from breast cancer after irradiation during fluoroscopic examinations in patient being treated fro tuberculosis. N Engl J Med 1989; 321: 1285–1289. 18 U.S. Environmental Protection Agency, Office of Radiation Programs. Background Document on Radioactive and Mixed Waste Incineration. Vol. II: Risks of Radiation Exposure. EPA, 1991. EPA 520/1–91-010–2. 19 United Nations Scientific Committee on the effects of atomic radiation, ionizing radiation sources and biological effects. 1982 report to the General Assembly, with annexes, Sales No. E.82.IX.8, United Nations, New York, 1982. 20 United Nations Scientific Committee on the effects of atomic radiation. Sources, effects and risks of ionizing radiation. 1988 report to the General Assembly, Sales No. 88.IX.7, United Nations, New York, 1988. 21 Sankaranarayanan K. Emerging perspectives in radiation genetic risk estimation. In: Proc Sci Seminar held in Luxemburg 29 November 1999. European Commission, Radiation Protection 123. Genetic susceptibility and new evolutions on genetic risk, 2000. 22 Hurwitz LM, Yoshizumi T, Reiman RE et al. Radiation dose to the fetus from body MDCT during early gestation. Am J Roentgenol 2006; 186: 871–876.
BLUK077-Stein
April 6, 2007
81
16:23
CHAPTER 81
Magnetic resonance angiography for the diagnosis of acute pulmonary embolism
Introduction Many patients with suspected acute pulmonary embolism (PE) have a relative contraindication to CT pulmonary angiography. Among patients with suspected acute PE in the Prospective Investigation of Pulmonary Diagnosis II (PIOPED II), 18.6% had an elevated creatinine, 3.9% were allergic to iodinated contrast material, and 4.7% of women were pregnant. One or more of these relative contraindications to an imaging procedure that would expose the patient to ionizing radiation or iodinated contrast material was present in 24% of patients with suspected acute PE. Although often patients with relative contraindications take the risk of CT or DSA (digital subtraction angiography) because of the importance of having a definitive diagnosis, such patients could benefit from safer diagnostic testing with contrast-enhanced magnetic resonance angiography, perhaps in combination with venous phase imaging of the veins of the lower extremities if it were shown to be sufficiently accurate.
Contrast-enhanced MRA for acute PE Detailed review of the literature [1] identified only 3 investigations in which gadolinium-enhanced magnetic resonance angiography (Gd-MRA) was used to diagnose acute PE and met the following criteria: (1) an objective diagnostic reference test was ordered independently of the results of the Gd-MRA, (2) GdMRA images were acquired three-dimensionally, and (3) sensitivity and specificity were calculated and data were shown [2–4] (Table 81.1). The diagnostic ref-
364
erence standard for PE in these 3 investigations was pulmonary angiography and 1.5 T systems were used. The sensitivity of Gd-MRA ranged from 77 to 100% and the specificity ranged from 95 to 98% (Table 81.1). More recently, however, Blum et al., using single-slice and multislice CT, pulmonary DSA, as well as clinical probability, D-dimer and venous ultrasound among 89 patients with suspected PE, showed sensitivities of 31 and 71% (Reader 1 and Reader 2) and specificities of 85 and 92% [5]. Inter-observer agreement was low (Kappa = 0.16). Numerous other investigations evaluated various aspects of MRA for PE, but did not meet these inclusion criteria [6–30]. Pulmonary Gd-MRA of normal pulmonary vasculature is illustrated in Figures 81.1 and 81.2. A Gd-MRA showing PE is illustrated in Figures 81.3 and 81.4. An iron-based contrast agent was used to show PE by MRA in Figure 81.5. Total examination time for Gd-MRA of the pulmonary arteries is about 15 minutes. This includes scout scanning for position of the images and the test bolus.
Table 81.1 Sensitivity and specificity of Gd-MRA for PE. Sensitivity
Specificity
[n/N (%)]
[n/N (%)]
Reference
8/8 (100)
21/22 (95)
Meaney et al. [2]
27/35 (77)
81/83 (98)
Oudkerk et al. [3]
11/13 (85)
22/23 (96)
Gupta et al. [4]
−/−(71)
−/−(92) Reader 1
Blum et al. [5]
−/−(31)
−/−(85) Reader 2
Gd-MRA, gadolinium-enhanced magnetic resonance angiography; PE, pulmonary embolism.
BLUK077-Stein
April 6, 2007
CHAPTER 81
16:23
365
MR angiography for diagnosis of acute PE
Figure 81.1 Multiplanar reconstruction of normal pulmonary arteries using gadolinium-enhanced magnetic resonance angiography (Gd-MRA). (Courtesy of Pamela K. Woodard, MD, Department of Radiology, Mallinckrodt Institute of Radiology, Washington University, St. Louis, Missouri.)
Gd-MRA: diagnostic criteria for PE The diagnostic criteria for acute PE on Gd MRA are:
1 A partially occlusive intraluminal filling. This may be shown as “railway tracking,” i.e., a small amount of contrast material between the central filling defect and the arterial wall or, in cross-sectional images, as a filling defect surrounded by contrast material.
Figure 81.2 Same patient as Figure 81.1 showing gadolinium-enhanced magnetic resonance angiography (Gd-MRA) of normal pulmonary arteries and pulmonary veins in left posterior oblique projection. (Courtesy of Pamela K. Woodard, MD, Department of Radiology, Mallinckrodt Institute of Radiology, Washington University, St. Louis, Missouri.)
2 Complete arterial occlusion with termination of the column of contrast material in a meniscus that outlines the trailing edge of the embolus [31].
Gd-MRA venous phase imaging “One stop shopping” with venous phase imaging of the proximal veins of the lower extremity in combination
BLUK077-Stein
April 6, 2007
16:23
366
PART III
Diagnosis of acute PE
Figure 81.3 Gadolinium-enhanced magnetic resonance angiography (Gd-MRA) showing intraluminal filling defects from pulmonary emboli (arrows). (Courtesy of
David P. Naidich, MD, Department of Radiology, New York University Medical Center, New York, New York.)
with GD-MRA of the chest has been recommended [32]. No additional gadolinium is necessary; the gadolinium injected for the pulmonary MR angiography is used for venous imaging. Following an antecubital injection of gadopentetate dimeglumine, the femoral vein was satisfactorily imaged in 20 of 20
(100%) and the iliac veins were satisfactorily imaged in 7 of 7 (100%) [33]. The use of MR venography of the lower extremities in combination with Gd-MR images of the pulmonary arteries creates a comprehensive study for thromboembolism comparable to the combination of
Figure 81.4 Same patient as in Figure 81.3 showing pulmonary emboli (arrows) in transverse plane. (Courtesy of David P. Naidich, MD, Department of Radiology, New York University Medical Center, New York, New York.)
BLUK077-Stein
April 6, 2007
CHAPTER 81
16:23
367
MR angiography for diagnosis of acute PE
Figure 81.5 Contrast-enhanced MRA using a single bolus of an iron-based R ) showing PE contrast agent (Clariscan (arrow). (Courtesy of Pamela K. Woodard, MD, Department of Radiology, Mallinckrodt Institute of Radiology, Washington University, St. Louis, Missouri.)
contrast-enhanced pulmonary spiral CT of the pulmonary arteries in combination with venous phase CT of the veins of the lower extremities (CTV) [34– 36]. To decrease the duration of the examination, the pelvic veins need not be investigated. In the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II), isolated pelvic vein DVT was shown in only 3 of 105 patients (3%) who had DVT identified by venous phase CT imaging of the veins of the pelvis and thighs [36]. Among 161 patients with DVT at autopsy, each of 7 patients with thrombi in the common iliac vein and each of 22 with thrombi in the external iliac vein also had DVT in the femoral vein [37]. The external iliac vein, however, showed thrombi in 12 of 161 patients (7%) without femoral vein involvement [37] (see Chapter 1).
MRA without contrast material Preliminary experience using MRA for acute PE without a contrast agent, in small case series, showed either
a high sensitivity, but suboptimal specificity [9, 25, 26] or vice versa [10].
MRA perfusion imaging Some have used Gd-MRA to show pulmonary perfusion based on filling of small vessels [15]. This is analogous to the perfusion phase of a pulmonary angiogram, which has been shown to be useful, but nonspecific [38]. It is also analogous to a perfusion lung scan. Most perfusion imaging has been used for physiologic investigations [39–41]. Investigations that focused on detection of perfusion defects from PE have been preliminary and showed inconclusive results [42–44]. Newer time-resolved three-dimensional contrast-enhanced MR angiography sequences that can provide both pulmonary angiography and perfusion in a single breath-hold sequence have been developed [45]. To date, however, these types of sequences are not commercially available on all vendor systems.
BLUK077-Stein
April 6, 2007
16:23
368
Direct thrombus imaging with MRI Acute thrombi can be directly visualized by looking for the T1 bright methemoglobin [46]. The technique appears promising [47–49], but has not been thoroughly investigated. Likewise, fibrin binding contrast agents are currently in preclinical trials [50].
Complications Severe reactions to magneto-pharmaceuticals are rare, 1 in 350,000–450,000 [51]. In the 2 reported cases of anaphylactoid reactions, the patients had a history of reactive airway disease [51]. Any adverse reaction to gadolinium contrast material has been reported in 0.06% [52], but perhaps as many as 1% may have mild allergic reactions (hives) according to the package insert [53]. Typically, 5–10% of the patients who have MRI feel claustrophobic [54]. Nausea, headache, and dizziness occur in 5% or less of patients [53]. In June 2006, the US Food and Drug Administration (FDA) issued an alert indicating that 25 cases of nephrogenic systemic fibrosis or nephrogenic fibrosing dermopathy (NSF/NFD) had been reported in patients with advanced renal failure (patients requiring dialysis or with a glomerular filtration rate ≤15 mL/min) who received the gadolinium-containing contrast agent, OmniscanR for MRA [55]. The dose required for adequate quality MRA is 2 or 3 times the FDA approved dose used for gadolinium-enhanced magnetic resonance imaging. Over the 4-year period that these 25 cases were reported, about 20 million patients received OmniscanR [56]. Since 1997, when NSF/NFD was first recognized, over 215 cases have been reported worldwide [57]. In December, 2006 the FDA reported that NSF/NFD occurred in patients with moderate renal disease (glomerular filtration rate <60 mL/min/1.73 m2 ) as well as end-stage renal disease (glomerular filtration rate <15 mL/min/1.73 m2 ) or on dialysis following use of gadolinium-containing contrast agents [58, 59]. All patients in whom information was available had received gadolinium-containing contrast agents either as a single dose or high dose [58]. The NSF/NFD began 2 days to 18 months after exposure [58]. Although NSF/NFD was reported with only 3 of 5 FDA approved gadolinium-containing contrast agents, the FDA believes that there is a potential for NSF/NFD to occur
PART III
Diagnosis of acute PE
with any of the approved gadolinium-containing contrast agents [58]. Patients with NSF/NFD have swelling and tightening of the skin, usually limited to the extremities, making it difficult to flex or extend the joints [55, 56]. The condition may develop over a period of days to several weeks [56]. In approximately 5%, of patients, the course is fulminate [56]. The NSF/NFD may result in fibrosis of body organs resulting in organ dysfunction or death [55].
References 1 Stein PD, Woodard PK, Hull RD et al. Gadoliniumenhanced magnetic resonance angiography for detection of acute pulmonary embolism: an in-depth review. Chest 2003; 124: 2324–2328. 2 Meaney JF, Weg JG, Chenevert TL, Stafford-Johnson D, Hamilton BH, Prince MR. Diagnosis of pulmonary embolism with magnetic resonance angiography. N Engl J Med 1997; 336: 1422–1427. 3 Oudkerk M, van Beek EJ, Wielopolski P et al. Comparison of contrast-enhanced magnetic resonance angiography and conventional pulmonary angiography for the diagnosis of pulmonary embolism: a prospective study. Lancet 2002; 359: 1643–1647. 4 Gupta A, Frazer CK, Ferguson JM et al. Acute pulmonary embolism: diagnosis with MR angiography. Radiology 1999; 210: 353–359. 5 Blum A, Bellou A, Guillemin F, Douek P, LaprevoteHeully MC, Wahl D, for GENEPI study group. Performance of magnetic resonance angiography in suspected acute pulmonary embolism. Thromb Haemost 2005; 93: 503–511. 6 Kruger S, Haage P, Hoffmann R et al. Diagnosis of pulmonary arterial hypertension and pulmonary embolism with magnetic resonance angiography. Chest 2001; 120: 1556–1561. 7 Loubeyre P, Revel D, Douek P et al. Dynamic contrastenhanced MR angiography of pulmonary embolism: comparison with pulmonary angiography. Am J Roentgen 1994; 162: 1035–1039. 8 Laissy JP, Bancal C, Sekkal S et al. Pulmonary MR angiography at 1.0 T: early results with k-space segmented and post-contrast TurboFLASH two-dimensional timeof-flight sequences. Eur J Radiol 1995; 20: 9–15. 9 Moody AR, Liddicoat A, Krarup K. Magnetic resonance pulmonary angiography and direct imaging of embolus for the detection of pulmonary emboli. Invest Radiol 1997; 32: 431–440. 10 Sostman HD, Layish DT, Tapson VF et al. Prospective comparison of helical CT and MR imaging in clinically
BLUK077-Stein
April 6, 2007
CHAPTER 81
11
12
13
14
15
16
17
18
19
20
21
22
23
16:23
MR angiography for diagnosis of acute PE
suspected acute pulmonary embolism. J Magn Reson Imaging 1996; 6: 275–281. Goyen M, Laub G, Ladd ME et al. Dynamic 3D MR angiography of the pulmonary arteries in under four seconds. J Magn Reson Imaging 2001; 13: 372–377. Haage P, Bucker A, Kruger S et al. [Radial k-scanning for real-time MR imaging of central and peripheral pulmonary vasculature]. ROFO 2000; 172: 203–206. Drevet D, Revel D, Marx P, Amiel M, Pinet F. [Obstruction of pulmonary arteries. Contribution of the scanner and MRI. Apropos of 10 cases]. Ann Radiol 1991; 34: 349– 354. Posteraro RH, Sostman HD, Spritzer CE, Herfkens RJ. Cine-gradient-refocused MR imaging of central pulmonary emboli. Am J Roentgenol 1989; 152: 465–468. Berthezene Y, Croisille P, Wiart M et al. Prospective comparison of MR lung perfusion and lung scintigraphy. J Magn Reson Imaging 1999; 9: 61–68. Amundsen T, Kvaerness J, Jones RA et al. Pulmonary embolism: detection with MR perfusion imaging of lung–a feasibility study. Radiol 1997; 203: 181–185. Amundsen T, Torheim G, Kvistad KA et al. Perfusion abnormalities in pulmonary embolism studied with perfusion MRI and ventilation-perfusion scintigraphy: an intra-modality and inter-modality agreement study. J Magn Reson Imaging 2002; 15: 386–394. Nakagawa T, Sakuma H, Murashima S, Ishida N, Matsumura K, Takeda K. Pulmonary ventilation–perfusion MR imaging in clinical patients. J Magn Reson Imaging 2001; 14: 419–424. Hatabu H, Gaa J, Kim D, Li W, Prasad PV, Edelman RR. Pulmonary perfusion and angiography: evaluation with breath-hold enhanced three-dimensional fast imaging steady-state precession MR imaging with short TR and TE. Am J Roentgenol 1996; 167: 653–655. Hatabu H, Tadamura E, Levin DL et al. Quantitative assessment of pulmonary perfusion with dynamic contrast-enhanced MRI. Magn Reson Med 1999; 42: 1033– 1038. Bergin CJ, Sirlin CB, Hauschildt JP et al. Chronic thromboembolism: diagnosis with helical CT and MR imaging with angiographic and surgical correlation. Radiology 1997; 204: 695–702. Bergin CJ, Hauschildt J, Rios G, Belezzuoli EV, Huynh T, Channick RN. Accuracy of MR angiography compared with radionuclide scanning in identifying the cause of pulmonary arterial hypertension. Am J Roentgenol 1997; 168: 1549–1555. Kreitner KF, Ley S, Kauczor HU et al. [Contrast media enhanced three dimensional MR angiography of the pulmonary arteries in patients with chronic recurrent pulmonary embolism—comparison with selective intraarterial DSA]. ROFO 2000; 172: 122–128.
369
24 Wolff K, Bergin CJ, King MA et al. Accuracy of contrastenhanced magnetic resonance angiography in chronic thromboembolic disease. Acad Radiol 1996; 3: 10–17. 25 Schiebler ML, Holland GA, Hatabu H et al. Suspected pulmonary embolism: prospective evaluation with pulmonary MR angiography. Radiology 1993; 189: 125–131. 26 Grist TM, Sostman HD, MacFall JR et al. Pulmonary angiography with MR imaging: preliminary clinical experience. Radiology 1993; 189: 523–530. 27 Erdman WA, Peshock RM, Redman HC et al. Pulmonary embolism: comparison of MR images with radionuclide and angiographic studies. Radiology 1994; 190: 499–508. 28 Hatabu H, Gefter WB, Axel L et al. MR imaging with spatial modulation of magnetization in the evaluation of chronic central pulmonary thromboemboli. Radiology 1994; 190: 791–796. 29 Gefter WB, Hatabu H, Dinsmore BJ et al. Pulmonary vascular cine MR imaging: a noninvasive approach to dynamic imaging of the pulmonary circulation. Radiology 1990; 176: 761–770. 30 White RD, Winkler ML, Higgins CB. MR imaging of pulmonary arterial hypertension and pulmonary emboli. Am J Roentgenol 1987; 149: 15–21. 31 Dotter CT. Acquired abnormalities of the pulmonary arteries. In: Abrams HL, ed. Angiography, 3rd edn. Little Brown, Boston, 1983: 743–761. 32 Obernosterer A, Aschauer M, Portugaller H, Koppel H, Lipp RW. Three-dimensional gadolinium-enhanced magnetic resonance angiography used as a “One-stop Shop” imaging procedure for venous thromboembolism: a pilot study. Angiology 2005; 56: 423–430. 33 Fraser DG, Moody AR, Davidson IR, Martel AL, Morgan PS. Deep venous thrombosis: diagnosis by using venous enhanced subtracted peak arterial MR venography versus conventional venography. Radiology 2003; 226: 812– 820. 34 Loud PA, Katz DS, Klippenstein DL, Shah RD, Grossman ZD. Combined CT venography and pulmonary angiography in suspected thromboembolic disease: diagnostic accuracy for deep venous evaluation. AJR 2000; 174: 61– 65. 35 Cham MD, Yankelevitz DF, Shaham D et al., for The Pulmonary Angiography-Indirect CT Venography Cooperative Group. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216: 744–751. 36 Stein PD, Fowler SE, Goodman LR et al., Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354(22): 2317–2327. 37 Havig O. Deep vein thrombosis and pulmonary embolism. An autopsy study with multiple regression analysis of possible risk factors. Chapters 2, 4. Pulmonary thromboembolism. Acta Chir Scand 1977; 478(suppl): 4– 11, 24–37.
BLUK077-Stein
April 6, 2007
16:23
370
38 Stein PD, O’Connor JF, Dalen JE et al. The angiographic diagnosis of acute pulmonary embolism: evaluation of criteria. Am Heart J 1967; 73: 730–741. 39 Roberts DA, Rizi RR, Lipson DA et al. Dynamic observation of pulmonary perfusion using continuous arterial spin labeling in a pig model. J Magn Reson Imaging 2001; 14: 175–180. 40 Levin DL, Chen Q, Zhang M, Edelman RR, Hatabu H. Evaluation of regional pulmonary perfusion using ultrafast magnetic resonance imaging. Magn Reson Med 2001; 46: 166–171. 41 Mai VM, Chen Q, Bankier AA et al. Effect of lung inflation on arterial spin labeling signal in MR perfusion imaging of human lung. J Magn Reson Imaging 2001; 13: 954– 959. 42 Amundsen T, Kvaerness J, Jones RA et al. Pulmonary embolism: detection with MR perfusion imaging of lung—a feasibility study. Radiology 1997; 203: 181–185. 43 Berthezene Y, Croisille P, Wiart M et al. Prospective comparison of MR lung perfusion and lung scintigraphy. J Magn Reson Imaging 1999; 9: 61–68. 44 Zheng J, Leawoods JC, Nolte M et al. Combined MR proton lung perfusion/angiography and helium ventilation: potential for detecting pulmonary emboli and ventilation defects. Magn Reson Med 2002; 47: 433–438. 45 Swan JS, Carroll TJ, Kennell TW et al. Time-resolved three-dimensional contrast-enhanced MR angiography of the peripheral vessels. Radiology 2002; 225: 43– 52. 46 Fraser DG, Moody AR, Morgan PS, Martel AL, Davidson I. Diagnosis of lower-limb deep venous thrombosis: a prospective blinded study of magnetic resonance direct thrombus imaging. Ann Intern Med 2002; 136: 89– 98. 47 Kelly J, Hunt BJ, Moody A. Images in cardiovascular medicine. Deep vein thrombosis demonstrated by magnetic resonance direct thrombus imaging. Circulation 2003; 107: 2165. 48 Moody AR. Magnetic resonance direct thrombus imaging. J Thromb Haemost 2003; 1: 1403–1409.
PART III
Diagnosis of acute PE
49 Moody AR. Direct imaging of deep-vein thrombosis with magnetic resonance imaging. Lancet 1997; 350: 1073. 50 Spuentrup E, Buecker A, Katoh M et al. Molecular magnetic resonance imaging of coronary thrombosis and pulmonary emboli with a novel fibrin-targeted contrast agent. Circulation 2005; 111: 1377–1382. 51 Carr JJ. Magnetic resonance contrast agents for neuroimaging. Safety issues. Neuroimaging Clin N Am 1994; 4: 43–54. 52 Cochran ST, Bomyea K, Sayre JW. Trends in adverse events after IV administration of contrast media. Am J Roentgenol 2001; 176: 1385–1388. 53 Package Insert: Magnevist. Berlex Laboratories, Wayne, NJ. Revised May 2000. 54 Kilborn LC, Labbe EE. Magnetic resonance imaging scanning procedures: development of phobic response during scan and at one-month follow-up. J Behav Med 1990; 13: 391–401. 55 U.S. Food and Drug Administration, Healthcare professional sheet. Gadolinium-containing contrast agents for magnetic resonance imaging (MRI) (marketed as Omniscan, OptiMARK, Magnevist, ProHance and MultiHance). http://www.fda.gov/cder/drug/InfoSheets/HCP/ gccaHCP.htm 6/23/06. 56 Flaten H. Omniscan safety update. June 6, 2006. http:// www.amershamhealth-us.com/omniscan/safety/index .html. 57 Cowper SE. International Center for Nephrogenic Fibrosing Dermopathy Research (ICNFDR). http://www. pathmax.com/dermweb. Last updated January 5, 2007. 58 FDA Public Health Advisory. Update on Magnetic resonance imaging (MRI) contrast agents containing gadolinium and nephrogenic fibrosing dermopathy. http:// www.fda.gov/cder/drug/advisory/gadolinium agents 20061222.htm. 59 FDA Information for Healthcare professionals. Gadolinium-based contrast agent for magnetic resonance imaging scans (marketed as Omniscan, Optiscan, OptiMARK, Magnevist, ProHance, and MultiHance). update 12/2006.
BLUK077-Stein
March 12, 2007
82
17:33
CHAPTER 82
Serial noninvasive leg tests in patients with suspected pulmonary embolism
Introduction In patients with suspected pulmonary embolism (PE), serial noninvasive leg tests would permit a diagnosis of thromboembolic disease or a safe exclusion of thromboembolic disease in 71% [1]. Among patients with PE, compression ultrasound was positive in 43 of 149 (29%) [2]. The technique of ultrasound, however, was limited because only the common femoral and popliteal veins were evaluated and only a single examination was performed [2]. Wells and associates, using more extensive imaging of the leg veins and serial studies, showed deep venous thrombosis (DVT) in 73 of 169 patients (43%) with high probability ventilation– perfusion (V–Q) scans [3]. Impedance plethysmography showed DVT among patients with PE, in 21 of 37 (57%) [4] and 36 of 83 (43%) [5].
Outcome of patients with suspected PE who had negative serial impedance plethysmograms Among patients with suspected PE, and a nondiagnostic V–Q scan, impedance plethysmography on day
1 showed DVT in 68 of 711 (9.6%) (Table 82.1) [6]. Serial impedance plethysmography over 14 days showed DVT in an additional 16 of the remaining 643 (2.5%) who had normal tests on day 1. Among patients with DVT shown by impedance plethysmography, 68 of 84 (81%) were identified on the first examination. Followup in 3 months showed PE in 4 of 627 (0.6%) and DVT in 8 of 627 (1.3%) [6]. All patients had normal cardiorespiratory reserve.
Outcome of patients with suspected PE and negative serial venous ultrasonograms Among 248 patients with suspected PE, a nondiagnostic V–Q scan and moderate probability clinical assessment, bilateral compression ultrasonography on day 1 showed DVT in 19 of 248 (7.7%) (Table 82.1) [3]. Serial compression ultrasonography on days 3, 7, and 14 showed DVT in an additional 7 of the remaining 229 (3.1%) who had normal tests on day 1. Among patients with DVT shown by compression ultrasound, 19 of 26 (73%) were identified on the first examination.
Table 82.1 Patients with suspected pulmonary embolism and nondiagnostic ventilation–perfusion scans. Leg test
Leg test positive
positive (day 1)
(3–14 days)
Clinical
[DVT/suspected
[DVT/suspected
PE (3-month
DVT (3-month
Test
assessment
PE (%)]
PE (%)]
follow-up)
follow-up)
References
Impedance
—
68/711 (9.6)
16/643 (2.5)
4/627 (0.6)
8/627 (1.3)
Hull et al. [6]
plethysmography Compression ultrasound
Intermediate
19/248 (7.7)
7/229 (3.1)
1/248 (0.4)
0/248 (0)
Wells et al. [3]
Low
4/454 (0.9)
7/450 (1.6)
2/443 (0.5)
0/443 (0)
Wells et al. [3]
DVT, deep venous thrombosis; PE, pulmonary embolism.
371
BLUK077-Stein
March 12, 2007
17:33
372
Follow-up in 3 months showed PE in 1 of 248 (0.4%). Asymptomatic DVT was not detected in any of these patients. Among 454 patients with suspected PE, a nondiagnostic V–Q scan and low probability clinical assessment, bilateral compression ultrasonography on day 1 showed DVT in 4 of 454 (0.9%) (Table 82.1) [3]. Serial compression ultrasonography on days 3, 7, and 14 showed DVT in an additional 7 of the remaining 450 (1.6%) who had normal tests on day 1. Among patients with DVT shown by compression ultrasound only 4 of 11 (36%) were identified on the first examination. Follow-up in 3 months showed PE in 2 of 443 (0.5%). Asymptomatic DVT was not detected in any of these patients.
References 1 Stein PD, Hull RD, Pineo G. Strategy that includes serial noninvasive leg tests for diagnosis of thromboembolic dis-
PART III
2
3
4
5
6
Diagnosis of acute PE
ease in patients with suspected acute pulmonary embolism based on data from PIOPED. Arch Intern Med 1995; 155: 2101–2104. Turkstra F, Kuijer PM, van Beek EJ, Brandjes DP, ten Cate JW, Buller HR. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781. Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. Hull RD, Hirsh J, Carter CJ et al. Diagnostic value of ventilation–perfusion lung scanning in patients with suspected pulmonary embolism. Chest 1985; 88: 819– 828. Hull RD, Raskob GE, Ginsberg JS et al. Noninvasive strategy for the treatment of patients with suspected pulmonary embolism. Arch Intern Med 1994: 154: 289–297.
BLUK077-Stein
April 6, 2007
83
16:26
CHAPTER 83
Predictive value of diagnostic approaches to venous thromboembolism
The positive and negative predictive values of any test depend on the prevalence of disease in the population studied. Investigations of diagnostic tests in pulmonary embolism (PE) had the following prevalences
of PE: Wells et al., 18% [1], Christopher Group 20% [2], PIOPED II, 23% [3], Perrier et al., 25% [4], and PIOPED I, 28% [5]. In this chapter, the positive and negative predictive values of tests that have been considered
Table 83.1 Positive predictive value (%) [(PE present/test positive) × 100] of tests for diagnosis of acute pulmonary embolism. n/N
%
Ref
CTA positive, any clin prob
150/175
86
CTA positive, high clin prob
22/23
96
[3] [3]
CTA positive, intermed clin prob
93/101
92
[3] [3]
CTA positive, low clin prob
22/38
58
CTA/CTV positive, any clin prob
164/194
85
[3]
CTA/CTV positive, high clin prob
27/28
96
[3]
CTA/CTV positive, intermed clin prob
100/111
90
[3]
CTA/CTV positive, low clin prob
24/42
57
[3]
CTA positive, main or lobar PA
116/120
97
[3]
CTA positive, segmental PA
32/47
68
[3]
CTA positive, subsegmental PA
2/8
25
[3]
Positive pulmonary DSA
100 (assumed)
V–Q high prob, any clin prob
103/118
87
V–Q high prob, no prior PE
88/97
91
[5] [5]
V–Q high prob, high clin prob
28/29
97
[5]
V–Q high prob, intermed clin prob
70/80
88
[5]
V–Q high prob, low clin prob
5/9
56
[5]
V–Q intermed prob, any clin prob
104/345
30
[5]
V–Q intermed prob, high clin prob
27/41
66
[5]
V–Q intermed prob, intermed clin prob
66/236
28
[5]
V–Q intermed prob, low clin prob
11/68
16
[5]
V–Q low prob, any clin prob
40/296
14
[5]
V–Q low prob, high clin prob
6/15
40
[5] (Continued )
373
BLUK077-Stein
April 6, 2007
16:26
374
PART III
Diagnosis of acute PE
Table 83.1 Positive predictive value (%) [(PE present/test positive) × 100] of tests for diagnosis of acute pulmonary embolism. (Continued ) n/N
%
Ref
V–Q low prob, intermed clin prob
30/191
16
[5]
V–Q low prob, low clin prob
4/90
4
[5]
V–Q very low prob, any clin prob
38/422
9
[14]
V–Q near normal/normal, any clin prob
5/128
4
[5]
V–Q near normal/normal, high clin prob
0/5
0
[5]
V–Q near normal/normal, intermed clin prob
4/62
6
[5]
V–Q near normal/normal, low clin prob
1/61
2
[5]
Positive venous ultrasound, suspected PE
173/178
97
[6]
CTA, computed tomographic pulmonary angiography; clin, clinical; prob, probability; intermed, intermediate; CTV, computed tomographic venous phase imaging; PA, pulmonary artery; DSA, digital subtraction pulmonary angiography; V–Q, ventilation–perfusion scan; PE, pulmonary embolism.
Table 83.2 Negative predictive value (%) [(PE absent/test negative) × 100] of tests for exclusion of acute pulmonary embolism. n/N
%
Ref
Neg D-dimer (<500 ng/mL), quantitative rapid ELISA*
98
[7]
Neg D-dimer (<500 ng/mL), quantitative rapid ELISA, low clin prob (Wells)
98
[7]
Neg D-dimer (<500 ng/mL), quantitative rapid ELISA, intermed clin prob (Wells)
93
[7]
Neg D-dimer (<500 ng/mL), quantitative rapid ELISA, high clin prob (Wells)
80
[7]
Neg D-dimer, whole blood agglutination*
93
[7]
Neg D-dimer, whole blood agglutination, low clin prob (Wells)
94
[7]
Neg D-dimer, whole blood agglutination, intermed clin prob (Wells)
84
[7]
Neg D-dimer, whole blood agglutination, high clin prob (Wells)
60
[7]
95
[3]
CTA neg, any clin prob
567/598
CTA neg, high clin prob
9/15
60
[3]
CTA neg, intermed clin prob
121/136
89
[3]
CTA neg, low clin prob
158/164
96
[3]
CTA/CTV both neg, any clin prob
524/543
97
[3] [3]
CTA/CTV both neg, high clin prob
9/11
82
CTA/CTV both neg, intermed clin prob
114/124
92
[3]
CTA/CTV both neg, low clin prob
146/151
97
[3]
Neg (normal) V–Q scan
1116/1122
99
[1, 5, 8–10]
Low clin prob, low or intermed prob V–Q, neg ultrasound
653/666
98
[1, 11, 12]
Neg DSA
374/380
98
[13]
*Estimate based on assumed prevalence of PE of 20%. Neg, negative; ELISA, enzyme-linked immunosorbent assay; clin, clinical; prob, probability; intermed, intermediate; CTA, computed tomographic pulmonary angiography; CTV, computed tomographic venous phase imaging; V–Q, ventilation– perfusion scan; DSA, digital subtraction pulmonary angiography.
BLUK077-Stein
April 6, 2007
CHAPTER 83
16:26
375
Predictive values of diagnostic tests for VTE
are summarized in Tables 83.1 and 83.2 [3–14]. These tables permit a rapid assessment of the validity of various diagnostic approaches.
8
References 1 Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. 2 van Belle A, Buller HR, Huisman MV et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA 2006; 295: 172–179. 3 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 4 Perrier A, Roy PM, Sanchez O et al. Multidetectorrow computed tomography in suspected pulmonary embolism. N Engl J Med 2005; 352: 1760–1768. 5 PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 6 Turkstra F, Kuijer PMM, van Beek EJR, Brandjes DPM, ten Cate JW, Buller HRl. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781. 7 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary
9
10
11
12
13
14
embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. van Beek EJR, Kuyer PMM, Schenk BE, Brandjes DP, ten Cate JW, B¨uller HR. A normal perfusion lung scan in patients with clinically suspected pulmonary embolism. Frequency and clinical validity. Chest 1995; 108: 170– 173. Kipper MS, Moser KM, Kortman KE, Ashburn WL. Longterm follow-up of patients with suspected pulmonary embolism and a normal lung scan. Perfusion scans in embolic suspects. Chest 1982; 82: 411–415. Hull RD, Raskob GE, Ginsberg JS et al. A noninvasive strategy for the treatment of patients with suspected pulmonary embolism. Arch Intern Med 1994; 154: 289– 297. Wells PS, Anderson DR, Rodger M et al. Excluding PE at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001; 135: 98–107. Perrier A, Miron MJ, Desmarais S et al. Using clinical evaluation and lung scan to rule out suspected pulmonary embolism: is it a valid option in patients with normal results of lower-limb venous compression ultrasonography? Arch Intern Med 2000; 160: 512–516. Henry JW, Relyea B, Stein PD. Continuing risk of thromboemboli among patients with normal pulmonary angiograms. Chest 1995; 107: 1375–1378. Gottschalk A, Stein PD, Sostman HD et al. Very low probability interpretation of ventilation–perfusion lung scans in combination with low probability clinical assessment reliably excludes pulmonary embolism: Data from PIOPED II (submitted for publication).
BLUK077-Stein
April 14, 2007
84
14:53
CHAPTER 84
Diagnostic approaches to acute pulmonary embolism
Introduction Many pathways for the diagnosis of acute pulmonary embolism (PE) have been recommended. With the development and verification of new diagnostic tests, some of the previously recommended approaches have become outdated [1–3]. There is no single recommended diagnostic pathway. A common feature of virtually all such pathways is that the selection of the sequence of diagnostic tests depends on the clinical probability of PE, the condition of the patient, local availability of diagnostic tests, risks of iodinated contrast material, radiation exposure, as well as cost [4]. Diagnostic management recommendations were formulated based on the results of PIOPED II [1], Wells et al. [2], Perrier et al. [3], and the Christopher Study [5], albeit with continued reliance on the physician’s opinion for some clinical circumstances [4]. Recommendations included both evidence-based recommendations and opinions based on information available at the time and both are subject to revision as further data become available [4]. The approach to diagnosis may be individualized, and depends upon the needs of the patient and availability of diagnostic tests. Considerations include risk, exposure to radiation, rapidity of obtaining the test, need for assessment of ancillary information, comfort, patient characteristics such as allergy, shock, ventilatory support, and potential therapy including thrombolytic agents or thrombectomy. Charges for individual tests based on a single community hospital are shown in Table 84.1.
Clinical assessment The PIOPED II investigators recommended that a clinical assessment should be made in all patients prior to imaging [4]. Physicians with experience in PE have shown similar results with empirical assessment [6–8] and by objective assessment using the Wells criteria
376
[2, 6, 9–11], the original Geneva criteria [7, 11], and the revised Geneva criteria [12] (Chapter 52). Since reproducible and more objective measures are likely to be more robust when applied by nonexperts, the PIOPED II investigators recommended assessment by one of these objective methods [2, 6, 7, 9–12], without obviating the value of informed clinical judgment.
Patients with low probability clinical assessment The PIOPED II investigators recommended a D-dimer quantitative rapid enzyme-linked immunosorbent assay (ELISA) in all patients with a low probability clinical assessment. Recommendations for the application of D-dimer rapid ELISA in combination with clinical assessment are shown in Figure 84.1 [13]. Systematic review showed that the quantitative rapid ELISA in general showed the most clinically useful values among the various D-dimer assays with a sensitivity of 95%. When used in combination with a low probability objective clinical assessment, which ranges from 4 to 15% [2, 6, 7, 9–12], the posttest probability of PE ranges from 0.7 to 2% [13, 14]. The D-dimer is more likely to be positive in hospitalized patients than in outpatients due to comorbid conditions [15]. Based on the evidence, the PIOPED II investigators suggested Table 84.1 Charges (including physicians’ fees) at a community hospital. Procedure
Charges (dollars)
Pulmonary angiography
6106
Contrast-enhanced spiral CT
1739
Ventilation–perfusion lung scan
917
Ultrasound, both legs
631
D-dimer (rapid ELISA)
24
ELISA, enzyme-linked immunosorbent assay.
BLUK077-Stein
April 14, 2007
CHAPTER 84
14:53
377
Diagnostic approaches to acute PE
Suspect pulmonary embolism
Figure 84.1 Pathway for D-dimer in combination with clinical assessment. If clinical assessment is low or moderate probability, and D-dimer rapid ELISA is negative, PE would be excluded. If clinical assessment is high probability, further testing is necessary irrespective of the results of D-dimer testing. Pos, positive; PE, pulmonary embolism; ELISA, enzyme-linked immunosorbent assay. (Reproduced from Stein et al. [4], with permission.)
Clinical low or moderate
Clinical high
D-dimer rapid ELISA negative
D-dimer rapid ELISA positive
No treatment
Further tests
that no further testing is necessary if the D-dimer is normal in a patient with a low probability clinical assessment. Additional testing with venous ultrasound or gadolinium-enhanced magnetic resonance venography [16] is optional and may be of value in some clinical settings. An abnormal D-dimer has no specific diagnostic value: it merely indicates the need for further testing if PE is suspected. If the D-dimer is positive, imaging with multidetector contrast-enhanced computed tomography (CT) was recommended for most patients [4]. The PIOPED II investigators had mixed opinions on whether CT of the chest (CTA) with objective clinical assessment is sufficient, or whether venous phase imaging (CTV) should be included (Figure 84.2). The majority preferred the combination CTA/CTV. A CTA alone had a sensitivity of 83% in PIOPED II and should not be used without clinical assessment or venous phase imaging [1]. A consideration was radiation exposure, and this may be more relevant in patients of reproductive age. The radiation exposure can be reduced by obtaining venous phase imaging only of the proximal leg veins, and omitting the iliac veins and inferior vena cava. In PIOPED II, among patients who showed thrombi on CTV, the iliac veins or the inferior vena cava showed thrombi in the absence of femoral or popliteal vein thrombi in only 3 of 105 (3%) [1]. In PIOPED II, among patients with a low probability clinical assessment, if CTA was negative, PE was present in 4% and if CTA/CTV was negative, PE was present in 3% [1] (Figure 84.2). In the Christopher Study, 1.3% of untreated patients with a normal CTA
Further tests
who had either a likely clinical probability or an abnormal D-dimer, had PE or DVT on 3-month follow-up [5]. Perrier et al. showed that among patients with a low or moderate clinical probability and positive Ddimer, if PE has been ruled out on the sole basis of a negative CTA, 1.5% would have had PE or DVT on 3-month follow-up [3]. Treatment, therefore, among such patients is unnecessary. In PIOPED II, if CTA was positive in a patient with a low probability clinical assessment, PE was present in 58% and with a positive CTA/CTV, PE was present in 57% [1]. However, if the CTA showed PE in a main or lobar pulmonary artery, PE was present in 97% [1]. Therefore, in patients with main or lobar artery emboli on CTA, treatment is indicated regardless of clinical probability. If the largest vessel showing PE by CTA was in a segmental branch, PE was present in 68%, and if PE was shown by CTA in a subsegmental branch, PE was present in 25% of patients, but data are sparse in the subsegmental group [1]. The certainty of the CT diagnosis should be re-assessed and if still in doubt, additional imaging is required. The PIOPED investigators believe that the CTA or CTA/CTV should be repeated if image quality is poor, unless the patient has renal impairment. Other options include a ventilation–perfusion (V–Q) lung scan (or perfusion lung scan alone if the chest radiograph is normal or nearly normal), a single venous ultrasound examination in those evaluated by CTA only, serial venous ultrasound examinations, or pulmonary digital subtraction angiography (DSA) [4]. The selection of the test depends on clinical judgment. Published data
BLUK077-Stein
April 14, 2007
14:53
378
PART III
regarding the performance of these options in specific settings are insufficient to permit evidence-based recommendations and accordingly clinical judgment is critical.
Patients with a moderate probability clinical assessment The PIOPED II investigators recommended a D-dimer test by the rapid ELISA method in all patients with a moderate probability clinical assessment [4] (Figure 84.3). Pooled values of objectively measured moderate clinical probabilities of PE have ranged from 29 to 38% [2, 6–8, 10, 11, 14]. The posttest probability of PE with a 30% clinical probability of PE is 5% with a negative rapid ELISA [13, 14]. No further testing was considered necessary if the rapid ELISA D-dimer is negative in a patient with a moderate probability clinical assessment, but a venous ultrasound or magnetic resonance venography was considered optional [4].
Diagnosis of acute PE
If the D-dimer plus clinical assessment does not exclude PE, imaging with CTA or CTA/CTV was recommended (Figure 84.3) [4]. In patients with a moderate probability clinical assessment, treatment with anticoagulants while awaiting the outcome of diagnostic tests may be appropriate, particularly if the tests cannot be obtained immediately [17]. As in patients with low probability clinical assessment, the PIOPED II investigators had mixed opinions on whether CTA with clinical assessment was adequate, or whether CTV should be included. Most preferred the combination of CTA/CTV. Among patients in PIOPED II with a moderate clinical probability assessment, if CTA was negative, PE was present in 11% and if CTA/CTV was negative, PE was present in 8% [1]. An outcome study of untreated patients after a negative CTA and low or moderate clinical assessment would have shown a risk of only 1.5% for venous thromboembolism (PE and/or deep venous thrombosis) within 3 months [3]. Even
Low probability clinical assessment Positive D-dimer fails to exclude PE
CTA or CTA/CTV
CTA neg, NPV 96% CTA/CTV neg, NPV 97%
CTA pos, PPV 58% CTA/CTV pos, PPV 57%
No treatment
Segmental, PPV 68% Subsegmental, PPV 25%
Main or lobar PE PPV 97%
Treat
Options: Repeat CTA or CTA/CTV if poor quality If CTA only, ultrasound or MRV V−Q scan or Q scan only DSA Serial ultrasound Figure 84.2 Pathway for diagnosis with CTA or CTA/CTV following testing with D-dimer in combination with low probability clinical assessment. CTA, contrast-enhanced multidetector computed tomographic pulmonary angiography; CTV, contrast-enhanced multidetector
computed tomographic venous phase imaging of the veins of the lower extremities; DSA, digital subtraction angiography; MRV, magnetic resonance venography; NPV, negative predictive value; PPV, positive predictive value. (Reproduced from Stein et al. [4], with permission.)
BLUK077-Stein
April 14, 2007
CHAPTER 84
14:53
379
Diagnostic approaches to acute PE
Moderate probability clinical assessment Positive D-dimer rapid ELISA
CTA or CTA/CTV
CTA neg, NPV 89% CTA/CTV neg, NPV 92%
No
CTA pos, PPV 92% CTA/CTV pos, PPV 90%
Treat
treatment
Option if CTA only, ultrasound or MRV
Figure 84.3 Pathway for diagnosis with CTA or CTA/CTV following testing with D-dimer in combination with moderate probability clinical assessment. CTA, contrast-enhanced multidetector computed tomographic pulmonary angiography; CTV, contrast-enhanced multidetector computed tomographic venous phase
with a “likely” clinical assessment, in the majority of patients another outcome study showed only 1.3% venous thromboembolism on 3-month follow-up [5]. Treatment is unnecessary in patients with a negative CTA or CTA/CTV, but based on the false negative rate of CTA in PIOPED II, a venous ultrasound or magnetic resonance venography in those with a negative CTA alone was recommended. The predictive values with intraluminal filling defects in branches of various orders are as described in the paragraph on low probability clinical assessment. Options for further imaging are the same and selection of the diagnostic approach again depends on the specific clinical situation [4]. If the CTA was positive in a patient with a moderate probability clinical assessment, PE was present in 92% in PIOPED II and with a positive CTA/CTV, PE was present in 90% [1]. Treatment was recommended [4].
imaging of the veins of the lower extremities; DSA, digital subtraction angiography; MRV, magnetic resonance venography; NPV, negative predictive value; PPV, positive predictive value. (Reproduced from Stein et al. [4], with permission.)
Patients with a high probability clinical assessment For patients with a high clinical suspicion of PE, treatment with anticoagulants is recommended while awaiting the outcome of diagnostic tests [17]. A negative D-dimer does not exclude PE in a patient with a high probability clinical assessment. An objectively measured high probability clinical assessment ranges from 59 to 74% [2, 6–12]. Patients with a negative D-dimer and high probability clinical assessment will have PE in >15%. Therefore, measurement of D-dimer serves no useful purpose in such patients. The PIOPED II investigators recommended CTA or CTA/CTV [4] (Figure 84.4). The choice depends on the patient’s age, reproductive status, clinical evidence of DVT, and clinical judgment. Most PIOPED II investigators recommended CTA/CTV [4].
BLUK077-Stein
April 14, 2007
14:53
380
PART III
Diagnosis of acute PE
High probability clinical assessment
CTA or CTA/CTV Figure 84.4 Pathway for diagnosis with CTA or CTA/CTV in patients with a high probability clinical assessment. CTA, contrast-enhanced multidetector computed tomographic pulmonary angiography; CTV, contrast-enhanced multidetector computed tomographic venous phase imaging of the veins of the lower extremities; DSA, digital subtraction angiography; MRV, magnetic resonance venography; NPV, negative predictive value; PPV, positive predictive value. (Reproduced from Stein et al. [4], with permission.)
CTA neg, NPV 60% CTA/CTV neg, NPV 82%
CTA pos, PPV 96% CTA/CTV pos, PPV 96%
Treat Options: Repeat CTA or CTA/CTV if poor quality If CTA only, ultrasound or MRV V−Q scan or Q scan only DSA Serial ultrasound
In PIOPED II, if CTA was negative in a patient with a high probability assessment, PE was present in 40% and if CTA/CTV was negative, PE was present in 18% [1]. The certainty of a negative CTA or CTA/CTV should be re-assessed and if the diagnosis is still in doubt, additional imaging is required [4]. The PIOPED investigators believe that the CTA or CTA/CTV should be repeated if image quality is poor, unless the patient has renal impairment [4]. A venous ultrasound or magnetic resonance venography was recommended in those with a negative CTA in whom CTV was not performed or was technically inadequate. Other options include serial venous ultrasound examinations [18] or pulmonary DSA. Serial venous ultrasound studies in patients with suspected PE who had a low or intermediate probability V–Q scan and low or intermediate probability clinical assessment showed PE in 3% on day 1, and in an additional 2% of patients during days 3 through 14 [9]. On 3-month follow-up of 665 untreated patients, 0.5% showed PE [9]. Some recommended a ventilation– perfusion (V–Q) lung scan or perfusion lung scan alone if the chest radiograph is normal or nearly normal [4]. The proportion of patients with a nondiagnostic V–Q scan is lower in those with a normal chest radiograph than in those with an abnormal chest radiograph [19, 20] and since PIOPED I, the proportion of patients with a nondiagnostic V–Q scan and a normal chest radiograph has been reported to be only
9% [20]. The selection of the test depends on clinical judgment. If either the CTA or CTA/CTV was positive in a patient with a high probability clinical assessment, PE was present in 96% in PIOPED II [1]. We recommend treatment.
Optional pathways, all patients A venous ultrasound may be obtained prior to imaging with CTA or CTA/CTV, thereby potentially eliminating the necessity for further imaging if the ultrasound is positive. In patients with suspected PE, a single venous ultrasound was positive in 13% of those with suspected PE and a positive D-dimer and it was positive in 42% of those shown to have PE or DVT [21]. Others showed a positive single venous ultrasound in 15% of those with suspected PE and in 29% of those shown to have proven PE [22]. If ultrasound is negative, only a CTA is required [4].
Patients with allergy to iodinated contrast material Alternative diagnostic pathways are available for patients in whom the physician believes that iodinated contrast material poses an increased risk or is contraindicated. Patients with mild to moderate iodine allergies may be pretreated with steroids, and then
BLUK077-Stein
April 14, 2007
CHAPTER 84
14:53
381
Diagnostic approaches to acute PE
Suspect PE (all PTS)
V−Q NL/NRNL PE 4%
Clinical intermediate
Clinical low
V−Q low 4% PE NO R X
V−Q inter 16 % PE
V−Q high 56 % PE
V−Q low 16 % PE
V−Q inter 28 % PE
Clinical high
V−Q inter 88 % PE
V−Q low 40 % PE
V−Q inter 66 % PE
V−Q high 96 % PE RX
Uncertain DX 72% of all PTS CTA/CTV Figure 84.5 Diagnostic pathway based on ventilation–perfusion (V–Q) lung scans showing positive predictive values obtained in PIOPED I [8]. PE, pulmonary embolism; PTS, patients; NL/NRNL, normal/near normal; INTER, intermediate; CTA, contrast-enhanced
multidetector computed tomographic pulmonary angiography; CTV, contrast-enhanced multidetector computed tomographic venous phase imaging of the veins of the lower extremities; RX, treatment; DX, diagnosis.
imaged with iodinated contrast material. If clinical assessment and D-dimer fail to exclude PE, venous ultrasound is recommended. If the venous ultrasound is positive, treatment is indicated and if negative, further tests are necessary. If further testing is needed, a V–Q scan or a perfusion lung scan alone, if the chest radiograph is normal or nearly normal, is recommended [4]. In PIOPED I, a low probability V–Q scan combined with a low probability clinical assessment showed PE in only 4% and treatment was not necessary [8] (Figure 84.5). The validity of perfusion scans alone was similar to the validity of V–Q scans [23]. A high probability V–Q scan in a patient with a high probability clinical assessment showed PE in 96% [8], and treatment is recommended. With other combinations, PE was present in 16–88% and further evaluation is recommended. Further eval-
uation may include serial venous ultrasound [9, 18], or gadolinium-enhanced CTA [24]. The purpose of PIOPED III is to evaluate the sensitivity and specificity of gadolinium-enhanced magnetic resonance imaging. Preliminary investigations suggest that it may be useful [25–28].
Patients with impaired renal function Alternative diagnostic pathways for patients with impaired renal function include clinical assessment in combination with a quantitative rapid ELISA D-dimer followed by venous ultrasound if PE is not excluded. If the venous ultrasound is positive, treatment is indicated and if negative, further tests are necessary [4]. The PIOPED II investigators recommended a V–Q
BLUK077-Stein
April 14, 2007
14:53
382
scan [4, 8], or a perfusion lung scan alone if the chest radiograph is normal or nearly normal [23]. Further evaluation may include serial venous ultrasound [9, 18], gadolinium-enhanced CTA [24], or perhaps gadolinium-enhanced magnetic resonance imaging [25–28]. Rarely, however, nephrogenic systemic fibrosis/nephrogenic fibrosing dermopathy (NSF/NFD) occurs in patients with impaired renal function following gadolinium-containing contrast agents (Chapter 81). In PIOPED II, only 1 of 824 patients who had a CTA (0.1%) developed renal failure [1]. This patient’s serum creatinine increased from 1.3 /to 2.9 mg/dL after a CTA followed in 22 hours by a DSA and returned to normal with intravenous fluids. Nonionic contrast material was used in PIOPED II [1]. Patients with abnormal serum creatinine levels were excluded. If the creatinine clearance is only somewhat elevated, whether to proceed with CTA or CTA/CTV depends on clinical judgment. A consideration in the risk of renal dysfunction is the load and type of contrast material. Nonionic contrast material appears to be less nephrotoxic [29] and generally better tolerated [30] than ionic contrast material, although some reported no difference in nephrotoxicity [31]. The risks of renal dysfunction from contrast material in patients with renal insufficiency can be reduced with prophylactic hydration. This and other measures are discussed in Chapter 79 [32–35].
Women of reproductive age The PIOPED II investigators recommended clinical assessment with D-dimer testing in women of reproductive age [4]. A negative rapid ELISA in a patient with a low or moderate clinical probability assessment excludes PE. If D-dimer is positive, or if clinical assessment is high probability for PE, venous ultrasound as the next diagnostic test is optional, and would eliminate in some patients the need for radiographic imaging. To minimize breast radiation a V–Q scan or a perfusion scan alone, if the chest radiograph is normal or nearly normal, was recommended by the majority of the PIOPED investigators [4] (Figure 84.5). In PIOPED I, a V–Q scan in patients with a normal chest radiograph was diagnostic (high probability or normal/nearly normal) in 52% of patients who had suspected PE [19]. More recently, pulmonary scintigraphy was shown to be diagnostic in 91% of patients with suspected PE and a normal chest radiograph [20].
PART III
Diagnosis of acute PE
A CTA with venous ultrasound is an acceptable alternative [4]. If CTV is deemed necessary for some clinical reason, it is advisable to start at the acetabulum, to reduce gonadal irradiation [4], since only 3 of 105 (3%) patients in PIOPED II with a positive CTV had isolated iliac vein or inferior vena cava thrombosis [1]. Female breast radiation exposure is a concern, but the risk of death from undiagnosed PE far exceeds the risk of radiation-induced malignancy [4]. The absorbed dose to the breast with CTA and lung scan is described in Chapter 80 [36–38]. The risk–benefit ratio clearly favors imaging until a diagnosis is obtained, but it seems prudent to attempt to keep exposure to ionizing radiation as low as reasonably achievable [4].
Pregnant patients In pregnant women, clinical assessment in combination with D-dimer testing was recommended [4] even though the D-dimer may be positive due to the pregnancy [39]. If D-dimer is positive or if clinical assessment is high probability, venous ultrasound is recommended before imaging tests with ionizing radiation. If the venous ultrasound is negative, some recommend a ventilation–perfusion scan (Figure 84.5) (or perfusion scan alone if the chest radiograph is normal or nearly normal), and some recommend CTA as had been recommended by others [40] and is done by many others [41]. Magnetic resonance imaging is an alternative diagnostic possibility that requires further validation [25–28] although adequate and wellcontrolled studies of gadopentetate dimeglumine have not been conducted in pregnant women [42]. It is not known to what extent gadopentetate dimeglumine is excreted in human milk [42]. Some indicate that the radiation dose to the fetus from 16-slice CTA, 0.24–0.47 mGy at 0 months and 0.61–0.66 mGy at 3 months, is of the same magnitude as a V–Q scan, 0.25–0.36 mGy at 0 months and 0.31– 0.32 mGy at 3 months, or a perfusion scan alone, 0.21 at 0 months and 0.30 at 3 months mGy [43]. Others indicate that the absorbed dose to the fetus is less with CTA than a perfusion lung scan (0.01 mGy versus 0.12 mGy) [36].
Patients in extremis Bedside echocardiography in combination with bedside leg ultrasonography are recommended as rapidly
BLUK077-Stein
April 14, 2007
CHAPTER 84
14:53
383
Diagnostic approaches to acute PE
obtained bedside tests [4]. Right ventricular enlargement or poor right ventricular function, in a proper clinical setting, can be interpreted as resulting from PE. A positive venous ultrasound in the appropriate clinical setting also indicates the need for immediate treatment. A combination of a negative bedside echocardiogram and ultrasound test indicates the need for CTA if it is clinically feasible to move the patient to the CT suite. A portable perfusion scan is recommended by some, while others recommend immediate transfer to an interventional catheterization laboratory. When the patient stabilizes, appropriate imaging studies should be performed to diagnose or exclude PE. The sensitivity of transthoracic echocardiography for right ventricular enlargement or dysfunction in patients with massive PE or unstable patients, combining data from three case series, was 33 of 33 (100%) [44–46]. If any 2 of the following 3 assessments were positive (clinical probability high, echocardiogram, and ultrasound), the sensitivity for massive pulmonary embolism was 33 of 34 (97%) and the negative predictive value was 98% [47]. Recommendations for clinical assessment
r Clinical assessment should be made prior to imaging. r Clinical assessment should be made by an objective method. Recommendations for patients with low probability clinical assessment
r Perform a D-dimer rapid ELISA. r No further testing is required if D-dimer is normal. r If D-dimer is positive, CTA/CTV is recommended by most (77%) PIOPED II investigators. r CTV of only the femoral and popliteal veins is recommended to reduce radiation. r If CTA or CTA/CTV is negative, treatment is unnecessary. r With main or lobar PE on CTA, treatment is indicated. r With segmental or subsegmental PE the certainty of the CT diagnosis should be re-assessed. r The CTA or CTA/CTV should be repeated if image quality is poor. r In patients with segmental or subsegmental PE, pulmonary scintigraphy, a single venous ultrasound in those evaluated by CTA only, serial venous ultrasound examinations, or pulmonary DSA are optional.
Recommendations for patients with a moderate probability clinical assessment
r A D-dimer rapid ELISA should be obtained. r If D-dimer rapid ELISA is negative no further testing is necessary, but a venous ultrasound or MRV is optional. r If D-dimer is positive, CTA/CTV is recommended by most (77%) PIOPED II investigators. r Treatment with anticoagulants while awaiting the outcome of diagnostic tests may be appropriate, particularly if the tests cannot be obtained immediately [17]. r If CTA or CTA/CTV are negative, no treatment is necessary, but a venous ultrasound is recommended for those with a negative CTA alone. r If CTA or CTA/CTV are positive, treatment is recommended. r With segmental or subsegmental PE the certainty of the CT diagnosis should be re-assessed and options followed according to recommendations for patients with a low probability clinical assessment.
Recommendations for patients with a high probability clinical assessment
r D-dimer testing need not be done because a negative D-dimer in a patient with a high probability clinical assessment may not exclude PE. r Treat with anticoagulants while awaiting the outcome of diagnostic tests [17]. r Most PIOPED II (77%) investigators recommend CTA/CTV. r If CTA is negative and CTA/CTV was not done or was technically inadequate, a venous ultrasound or MRV is recommended. r If CTA or CTA/CTV are negative, other options include serial venous ultrasound examinations, pulmonary DSA, and pulmonary scintigraphy. r If CTA or CTA/CTV are positive, treatment is recommended. Recommendations for optional pathways
r A venous ultrasound prior to imaging with CTA or CTA/CTV is optional and may guide treatment if positive. Recommendations for patients with allergy to iodinated contrast material
r D-dimer with clinical assessment is recommended to exclude PE. r Patients with mild iodine allergies may be treated with steroids prior to the CT imaging.
BLUK077-Stein
April 14, 2007
14:53
384
r Venous ultrasound and pulmonary scintigraphy are recommended as alternative diagnostic tests in patients with severe iodine allergy. r Serial venous ultrasound and gadoliniumenhanced CTA are options. Recommendations for patients with impaired renal functions
r D-dimer with clinical assessment is recommended to exclude PE. r Venous ultrasound is recommended and if positive, treatment is indicated. r Pulmonary scintigraphy is recommended if venous ultrasound is negative. r Serial venous ultrasound and gadoliniumenhanced CTA are options but NSF/NFD is a risk with gadolinium-containing contrast agents in patients with impaired renal function. Recommendations for women of reproductive age
r If D-dimer rapid ELISA is positive, venous ultrasound as the next diagnostic test is optional. r Pulmonary scintigraphy is recommended by some (31%) PIOPED investigators as the next imaging test, but most (69%) recommend CT angiography. r A CTA with venous ultrasound is an acceptable alternative. r If a CTV is deemed necessary, it is advisable to start at the acetabulum to reduce gonadal irradiation. Recommendations for pregnant patients
r D-dimer with clinical assessment should be obtained. r If D-dimer is positive, venous ultrasound is recommended before imaging tests with ionizing radiation. r Most (69%) PIOPED II investigators recommend pulmonary scintigraphy, and some (31%) recommend a CTA. Recommendations for patients in extremis
r Bedside echocardiography in combination with bedside leg ultrasonography are recommended as rapidly obtainable bedside tests. r Right ventricular enlargement or poor right ventricular function, in a proper clinical setting, can be interpreted as resulting from PE. r A positive venous ultrasound in the appropriate clinical setting also indicates PE. r A portable perfusion scan is recommended by some (38%).
PART III
Diagnosis of acute PE
r Immediate transfer to an interventional catheterization laboratory is recommended by some (38%). r A combination of a negative bedside echocardiogram and venous ultrasound indicate the need for CTA if feasible. r When the patient stabilizes, appropriate imaging studies should be performed.
References 1 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 2 Wells PS, Anderson DR, Rodger M et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001; 135: 98–107. 3 Perrier A, Roy PM, Sanchez O et al. Multidetectorrow computed tomography in suspected pulmonary embolism. N Engl J Med 2005; 352: 1760–1768. 4 Stein PD, Woodard PK, Weg JG et al. Diagnostic pathways in acute pulmonary embolism: recommendations of the PIOPED II Investigators. Am J Med 2006; 119: 1048– 1055. 5 van Belle A, Buller HR, Huisman MV et al. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA 2006; 295: 172– 179. 6 Sanson BJ, Lijmer JG, Mac Gillavry MR, Turkstra F, Prins MH, Buller HR, for the ANTELOPE-Study Group. Comparison of a clinical probability estimate and two clinical models in patients with suspected pulmonary embolism. Thromb Haemost 2000; 83: 199–203. 7 Wicki J, Perneger TV, Junod AF, Bounameaux H, Perrier A. Assessing clinical probability of pulmonary embolism in the emergency ward. A simple score. Arch Intern Med 2001; 161: 92–97. 8 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 9 Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. 10 Wells PS, Anderson DR, Rodger M et al. Derivation of a simple clinical model to categorize patients probability
BLUK077-Stein
April 14, 2007
CHAPTER 84
11
12
13
14 15
16 17
18
19
20
21
22
23
14:53
385
Diagnostic approaches to acute PE
of pulmonary embolism: increasing the models utility with the SimpliRED D-Dimer. Thromb Haemost 2000; 83: 416–420. Chagnon I, Bounameaux H, Aujesky D et al. Comparison of two clinical prediction rules and implicit assessment among patients with suspected pulmonary embolism. Am J Med 2002; 113: 269–275. Le Gal G, Righini M, Roy PM et al. Prediction of pulmonary embolism in the emergency department: the revised Geneva score. Ann Intern Med 2006; 144: 165– 171. Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. Sox HC. Commentary. Ann Intern Med 2004; 140: 602. Schrecengost JE, LeGallo RD, Boyd JC et al. Comparison of diagnostic accuracies in outpatients and hospitalized patients of D-dimer testing for the evaluation of suspected pulmonary embolism. Clin Chem 2003; 49: 1483– 1490. Sostman HD. MRA for diagnosis of venous thromboembolism. Q J Nucl Med 2001; 45: 311–323. Buller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease: the seventh ACCP Conference on antithrombotic and thrombolytic therapy. Chest 2004; 126(3 suppl): 401S–428S. Stein PD, Hull RD, Pineo G. Strategy that includes serial noninvasive leg tests for diagnosis of thromboembolic disease in patients with suspected acute pulmonary embolism based on data from PIOPED. Prospective Investigation of Pulmonary Embolism Diagnosis. Arch Intern Med 1995; 155: 2101–2104. Stein PD, Alavi A, Gottschalk A et al. Usefulness of noninvasive diagnostic tools for diagnosis of acute pulmonary embolism in patients with a normal chest radiograph. Am J Cardiol 1991; 67: 1117–1120. Forbes KP, Reid JH, Murchison JT. Do preliminary chest X-ray findings define the optimum role of pulmonary scintigraphy in suspected pulmonary embolism? Clin Radiol 2001; 56: 397–400. Perrier A, Roy PM, Aujesky D et al. Diagnosing pulmonary embolism in outpatients with clinical assessment, D-dimer measurement, venous ultrasound, and helical computed tomography: a multicenter management study. Am J Med 2004; 116: 291–299. Turkstra F, Kuijer PM, van Beek EJ, Brandjes DP, ten Cate JW, Buller HR. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781. Stein PD, Terrin ML, Gottschalk A, Alavi A, Henry JW. Value of ventilation/perfusion scans versus perfusion
24
25
26
27
28
29
30
31
32
33 34
35
36 37
scans alone in acute pulmonary embolism. Am J Cardiol 1992; 69: 1239–1241. Remy-Jardin M, Bahepar J, Lafitte JJ et al. Multi-detector row CT angiography of pulmonary circulation with gadolinium-based contrast agents: prospective evaluation in 60 patients. Radiology 2006; 238: 1022–1035. Stein PD, Woodard PK, Hull RD et al. Gadolinium enhanced magnetic resonance angiography for detection of acute pulmonary embolism: and in depth review. Chest 2003; 124: 2324–2328. Meaney JF, Weg JG, Chenevert TL, Stafford-Johnson D, Hamilton BH, Prince MR. Diagnosis of pulmonary embolism with magnetic resonance angiography. N Engl J Med 1997; 336: 1422–1427. Oudkerk M, van Beek EJ, Wielopolski P et al. Comparison of contrast-enhanced magnetic resonance angiography and conventional pulmonary angiography for the diagnosis of pulmonary embolism: a prospective study. Lancet 2002; 359: 1643–1647. Gupta A, Frazer CK, Ferguson JM et al. Acute pulmonary embolism: diagnosis with MR angiography. Radiology 1999; 210: 353–359. Harris KG, Smith TP, Cragg AH, Lemke JH. Nephrotoxicity from contrast material in renal insufficiency: ionic versus nonionic agents. Radiology 1991; 179: 849– 852. Barrett BJ, Carlisle EJ. Metaanalysis of the relative nephrotoxicity of high- and low-osmolality iodinated contrast media. Radiology 1993; 188: 171–178. Schwab SJ, Hlatky MA, Pieper KS et al. Contrast nephrotoxicity: a randomized controlled trial of a nonionic and an ionic radiographic contrast agent. N Engl J Med 1989; 320: 149–153. Merten GJ, Burgess WP, Gray LV et al. Prevention of contrast-induced nephropathy with sodium bicarbonate: a randomized controlled trial. JAMA 2004; 291: 2328– 2334. Chertow GM. Prevention of radiocontrast nephropathy: back to basics. JAMA 2004; 291: 2376–2377. Waybill MM, Waybill PN. Contrast media-induced nephrotoxicity: identification of patients at risk and algorithms for prevention. J Vasc Interv Radiol 2001; 12: 3–9. Heupler FA, Jr. Guidelines for performing angiography in patients taking metformin. Members of the Laboratory Performance Standards Committee of the Society for Cardiac Angiography and Interventions. Cathet Cardiovasc Diagn 1998; 43: 121–123. Cook JV, Kyriou J. Radiation from CT and perfusion scanning in pregnancy. BMJ 2005; 331: 350. Parker MS, Hui FK, Camacho MA. Female breast radiation exposure during CT pulmonary angiography. Am J Roentgenol 2005; 185: 1228–1233.
BLUK077-Stein
April 14, 2007
14:53
386
38 International Commission on Radiological Protection. ICRP Publication 87: Managing patient dose in computed tomography, 87. Ann ICRP 2002; 30: 4. 39 Eichinger S. D-dimer testing in pregnancy. Pathophysiol Haemost Thromb 2003; 33: 327–329. 40 Matthews S. Imaging pulmonary embolism in pregnancy: what is the most appropriate imaging protocol? Br J Radiol 2006; 79: 441–444. 41 Schuster ME, Fishman JE, Copeland JF, Hatabu H, Boiselle PM. Pulmonary embolism in pregnant patients: a survey of practices and policies for CT pulmonary angiography. Am J Roentgenol 2003; 181: 1495–1498. 42 Package Insert: Magnevist. Berlex Laboratories, Wayne, NJ. Revised May 2000. 43 Hurwitz LM, Yoshizumi T, Reiman RE et al. Radiation dose to the fetus from body MDCT during early gestation. Am J Roentgenol 2006; 186: 871–876.
PART III
Diagnosis of acute PE
44 Cheriex EC, Sreeram N, Eussen YF, Pieters FA, Wellens HJ. Cross sectional Doppler echocardiography as the initial technique for the diagnosis of acute pulmonary embolism. Br Heart J 1994; 72: 52–57. 45 Mansencal N, Redheuil A, Joseph T et al. Use of transthoracic echocardiography combined with venous ultrasonography in patients with pulmonary embolism. Int J Cardiol 2004; 96: 59–63. 46 Rudoni RR, Jackson RE, Godfrey GW, Bonfiglio AX, Hussey ME, Hauser AM. Use of two-dimensional echocardiography for the diagnosis of pulmonary embolus. J Emerg Med 1998; 16: 5–8. 47 Grifoni S, Olivotto I, Cecchini P et al. Utility of an integrated clinical, echocardiographic, and venous ultrasonographic approach for triage of patients with suspected pulmonary embolism. Am J Cardiol 1998; 82: 1230– 1235.
BLUK077-Stein
March 12, 2007
18:31
IV Prevention and PART IV
treatment of deep venous thrombosis and pulmonary embolism
387
BLUK077-Stein
March 12, 2007
85
18:31
CHAPTER 85
New and old anticoagulants
Warfarin Vitamin K antagonists, such as warfarin, inhibit the production of prothrombin, factor VII, factor IX, factor X, and the natural anticoagulants, protein C and protein S (Figure 85.1). Vitamin K antagonists, in addition to depletion of vitamin K coagulation factors, also cause synthesis of proteins that interfere with coagulation [1, 2].
International normalized ratio The prothrombin time ratio as well as the prothrombin time are dependent upon the thromboplastin reagent [3]. An international normalized ratio (INR), based upon a World Health Organization standard thromboplastin, was defined, which permits reporting of the prothrombin time ratio in a standardized fashion. This permits comparison of the anticoagulant effect, irrespective of the local thromboplastin reagent used for measuring the prothrombin time. The INR is calculated as follows: INR = (prothrombin time ratio)ISI , where ISI is international sensitivity index of the thromboplastin reagent used for measuring the prothrombin time. The prothrombin time ratio (the ratio of patient prothrombin time to control prothrombin time) is measured in the local laboratory. The ISI is a measure of the responsiveness of a given thromboplastin to a reduction of the vitamin K-dependent coagulation factor, compared with the international reference preparation [3]. The value of the ISI for commercial thromboplastins is usually indicated on the reagent packages, and therefore, is easily reported by the laboratory. The INR is calculated by raising the prothrombin time ratio to the ISI power. For example, if the patient prothrombin time is 18 seconds and the control prothrombin time is 12 seconds, the prothrombin time ratio is 1.5. If the ISI is 2, then INR = 1.52 = 2.25. The ISI of commercial thromboplastin reagents varies according to manufacturer, and it varies from
batch to batch. Most thromboplastins used in the United States vary from 1.8 to 2.8 [4]. Most thromboplastins used in the United Kingdom, and in many parts of Scandinavia and The Netherlands, have ISI values of 1.0–1.1 [3]. The range of INR recommended for treatment of deep venous thrombosis (DVT) or acute pulmonary embolism (PE) is 2.0–3.0. This is considered “less intense warfarin” compared with the range of INR recommended for preventing thrombosis with some types of mechanical prosthetic heart valves (INR 2.5– 3.5) [5].
Administration of warfarin The prothrombin time may become prolonged before a full anticoagulant effect is reached. This is a result of a rapid reduction of factor VII, which has a half-life of 6–7 hours [6]. Full anticoagulant activity is delayed 72–96 hours after the administration of warfarin because the half-lives of prothrombin (60 hours), factor IX (24 hours), and factor X (48–72 hours) are considerably longer than the half-life of factor VII [7]. Protein C and protein S are vitamin K-dependent natural anticoagulants that are also affected by warfarin. The potential exists for the early anticoagulant effect of warfarin to be counteracted by a reduction of protein C, because the half-life is short (8–11 hours), similar to factor VII [3, 8]. Warfarin-induced skin necrosis has been attributed to this reduction in protein C [3]. Warfarin therapy should begin with a maintenance dose, or perhaps twice the predicted maintenance dose [9]. Larger loading doses are not recommended [3]. If there is an urgent need for antithrombotic therapy, heparin is indicated, and heparin is recommended to be continued 4–5 days after the prothrombin time is in the therapeutic range [3]. Warfarin should not be used during the first trimester of pregnancy because of the risk of embryopathy or fetal bleeding [3, 10]. If possible, it should be avoided throughout pregnancy [3]. Heparin is
389
BLUK077-Stein
March 12, 2007
18:31
390
PART IV
Prevention and Treatment of DVT and PE
(a) Intrinsic pathway
Extrinsic pathway
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa
XIa
XI
IXa
IX
TF/VIIa complex X VIIIa Xa Va IIa thrombin
II prothrombin
Ia fibrin
I fibrinogen (b)
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa
XI IX
TF/VIIa complex
AT III
XIa IXa
AT III
X VIIIa AT III
Xa
Va Activated protein C IIa thrombin
II prothrombin
Protein C + thrombomodulin Protein S
AT III I fibrinogen Figure 85.1 (a) Simplified diagram of coagulation cascade. (b) Simplified coagulation cascade including naturally occurring anticoagulants. Activated protein C blocks the
Ia fibrin action of factors Va and VIIIa, indicated by dotted break lines in arrows. Antithrombin III (AT III) blocks the action of factors XIIa, Xia, IXa, Xa, and IIa.
BLUK077-Stein
March 12, 2007
CHAPTER 85
18:31
391
New and old anticoagulants
Table 85.1 Major bleeding (3 months) in patients* treated with vitamin K antagonists with an INR of 2.0–3.0. Number of patients Author [Ref]
[n/N (%)]
Das [13]
0/55 (0)
Gonzalez-Fajardo [14]
2/80 (2.5)
Hull [15]
2/239 (0.8)
Hull [16]
17/368 (4.6)
Kakkar [17]
1/246 (0.4)
Pooled
22/988 (2.2)
* All patients treated for venous thromboembolic disease. INR, international normalized ratio.
preferred if anticoagulants are indicated in pregnancy [3]. Warfarin, in nursing mothers, appears not to induce an anticoagulant effect in the breast-fed infant [11, 12].
Risk of major hemorrhage with warfarin The risk of major hemorrhagic complications among patients with thromboembolic disease treated with vitamin K antagonists (INR = 2.0–3.0) on average was 2.2% in 3 months (Table 85.1) [13–17]. Among patients with atrial fibrillation treated with warfarin at an INR of 2.0–3.0, the rate of bleeding was 4.2%/year [18– 21] (Table 85.2). The definition of major bleeding varied, but in general was defined as overt bleeding associated with a reduction of hemoglobin >2 gm/dL, blood transfusion >2 units, intracerebral bleed, retroperitoneal bleed, pericardial bleed, bleeding that required a surgical intervention, bleeding into a major joint, or bleeding into the eye.
Unfractionated heparin Unfractionated heparin has equal anti-Xa and antiIIa activity [22] (Figure 85.2). Unfractionated heparin binds to antithrombin III forming a complex that blocks factor Xa and inhibits thrombin in the fluid phase but does not inhibit clot-bound thrombin [23]. The heparin–antithrombin III complex also inhibits factor IXa, Xia, and XIIa [24, 25]. Heparin is sensitive to inactivation by platelet factor 4. Unfractionated heparin binds to platelets, endothelial cells, plasma proteins, and macrophages, and this nonspecific binding limits its ability to bind to the plasma cofactor antithrombin III. Less than 30% of unfractionated heparin is available for blocking the activity of factor Xa [26]. Unfractionated heparins have a rate of thrombocytopenia of 9 of 332 (2.7%) [27].
Dosage The unit of heparin is measured in animals using a biologic assay [28]. Units of unfractionated heparin may vary as much as 50% on a weight basis and therefore unfractionated heparin is properly prescribed by units, not weight [28]. For the treatment of DVT, either intravenous [9, 29] or subcutaneous [30, 31] unfractionated heparin has been adequate, provided that the activated partial thromboplastin time (APTT) has been within the therapeutic range. A blood level of heparin between 0.2 and 0.4 U/mL (protamine titration) inhibits thrombus propagation [32, 33]. In general, an APTT more than 1.5 times control or 1–5 times mean normal values corresponded to a blood level of 0.2 U/mL [28, 34–36]. Failure to achieve an APTT >1.5 times control in the initial treatment of proximal DVT was associated with a risk of recurrent DVT or PE of 20–25% [9]. If the APTT was
Table 85.2 Major bleeding per year in patients* treated with warfarin with an INR of 2.0–3.0. Author [Ref]
Number of patients
Major bleeding (%/year)
Gullov [18, 19]
170
1.1
Pengo [20]
153
2.6
Matchar [21]
572
6
Matchar [21]
593
7
Pooled
1488
* All patients treated for atrial fibrillation. INR, international normalized ratio.
4.2
BLUK077-Stein
March 12, 2007
18:31
392
PART IV
Prevention and Treatment of DVT and PE
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa XI
XIa IX
TF/VIIa complex
X
IXa
ATIII/Heparin
Heparin
VIIIa Xa
II prothrombin
Va
IIa thrombin
I fibrinogen
Ia fibrin
Figure 85.2 Sites of action of unfractionated heparin. Heparin combines with antithrombin III (AT III) to form an ATIII/heparin complex that primarily blocks the action of factors Xa, and IIa and to a lesser extent, factors XIIa, Xia, and IXa.
greater than 1.5 times control, recurrent DVT or PE was infrequent [37]. Heparin requirements are usually greatest in the first few days after an acute thromboembolic event [38–40]. Therapy, therefore, should be monitored closely during that period. Unfortunately, the APTT does not always correlate reliably with blood heparin levels or the antithrombotic activity [28]. The APTT can be shortened by increased levels of clotting factors such as factor VIII [28]. The anticoagulant effect of unfractionated heparin can be suppressed by heparin binding proteins in plasma [41]. A trend suggested that patients whose blood level of unfractionated heparin was measured directly had fewer recurrences of DVT or PE and less bleeding [32]. If the hospital laboratory does not measure heparin blood levels directly, it would be useful to determine the range of the APTT that corresponds to blood levels of heparin between 0.2 and 0.4 U/mL [28]. With fixed blood levels of heparin, values of the APTT may change if the APTT reagent is changed or a different batch is used [28]. The therapeutic range of the APTT ratio in many laboratories is higher than 1.5–2.5 times the mean of a normal reference range [33]. No international reference of APTT reagent exists that allows de-
velopment of a normalized ratio for heparin analogous to the INR for warfarin. It seems prudent, therefore, to establish a reference range in each laboratory [28].
Risk of major hemorrhage with unfractionated heparin The risk of major hemorrhagic complications (defined as with warfarin) among patients treated with therapeutic doses of unfractionated heparin for thromboembolic disease, whether intermittent intravenous, continuous intravenous infusion or subcutaneous, in doses over 24 hours ranging from 29,180 to 40,320 U was 3.4 to 7.8% [9, 30, 31, 42]. The average risk of major bleeding with therapeutic unfractionated heparin for thromboembolic disease in contemporary studies that used a continuous intravenous infusion or high-dose subcutaneous injection was approximately 5% [43]. These risks have been stratified according to whether the patients were at high risk of bleeding or low risk of bleeding [44]. Among patients at high risk (surgery within previous 14 days, history of peptic ulcer disease, gastrointestinal or genitourinary tract bleeding, disorders predisposing to bleeding, thrombotic studies
BLUK077-Stein
March 12, 2007
CHAPTER 85
18:31
393
New and old anticoagulants
within 14 days, or platelet count <150 × 109 /L), the frequency of major bleeding from heparin was 10.8% [44]. Among patients at low risk of bleeding, the frequency of major bleeding with heparin was 1.1%.
U/24 hours [53]. Most osteoporosis occurred in patients treated with unfractionated heparin in doses of at least 20,000 U/day for more than 6 months [52].
Oral heparin (SNAC heparin) Other complications of therapy with unfractionated heparin Thrombocytopenia occurred in 15.6% of patients who received bovine unfractionated heparin and 5.8% who received porcine unfractionated heparin [45]. Frequent monitoring of the platelet count, therefore, is essential. Thrombocytopenia may occur within hours after beginning therapy with unfractionated heparin, although it usually occurs between 3 and 15 days after beginning heparin [45, 46]. The platelet count usually returns to baseline levels within 4 days of stopping therapy [47], but persistent thrombocytopenia may occur. Thrombocytopenia is uncommon (<1%) in patients who receive low-dose heparin [48, 49]. Heparininduced thrombocytopenia can be complicated by thrombosis caused by platelet-rich thrombi [50]. An immunologic basis for heparin-induced thrombocytopenia and thrombosis has been proposed [51]. Rarely, unfractionated heparin may induce skin necrosis. Histologic features are consistent with a hypersensitive angiitis [52]. Osteoporosis is unlikely in patients treated with unfractionated heparin for <3 months at doses <20,000
Combining unfractionated heparin with a carrier molecule, sodium N-[8(2-hydroxybenzoyl) amino] caprylate (SNAC) has markedly increased the gastrointestinal absorption of heparin [54]. The drug appears to be efficacious, based on preclinical and a limited number of clinical investigations [54]. Bleeding was comparable when compared with subcutaneously administered unfractionated heparin and with lowmolecular-weight heparins (LMWHs) [54].
Low-molecular-weight heparin Low-molecular-weight heparins act by binding to antithrombin III and catalyzing the inactivation of factor Xa (Figure 85.3). Low-molecular-weight heparins have 90% bioavailability for blocking the activity of factor Xa [26]. They have 2–4 times greater anti-factor Xa activity than anti-factor IIa activity [22]. Use of LMWH does not require monitoring of factor Xa levels. Lowmolecular-weight heparin activity would not be shown with the APTT. Low-molecular-weight heparin can have a potent antithrombotic effect causing only a minimal change in the APTT [55]. They have a predictable
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa TF/VIIa complex
XI
XIa IX
X
IXa
VIIIa Figure 85.3 Sites of action of low-molecular-weight heparin (LMWH). The LMWH combines with antithrombin III (AT III) to form an ATIII/LMWH complex that primarily blocks the action of factor Xa and to a lesser extent, factor IIa.
Xa
II prothrombin
ATIII/LMWH
IIa Va thrombin I fibrinogen
Ia fibrin
LMWH
BLUK077-Stein
March 12, 2007
18:31
394
PART IV
Prevention and Treatment of DVT and PE
Table 85.3 Low-molecular-weight heparins for prophylaxis and treatment of deep venous thrombosis and pulmonary embolism.
Drug (trade name)
FDA approved for
FDA approved for
FDA approved
DVT prophylaxis
DVT treatment
for PE treatment
Enoxaparin (LovenoxR )
Yes
Yes
Yes
Tinzaparin (InnohepR )
No
Yes
No
Dalteparin (FragminR )
Yes
No
No
Nadroparin (FraxiparinR )
No
No
No
Certoparin (AlphaparinR )
No
No
No
Bemiparin (ZiborR )
No
No
No
Reviparin (ClivarineR )
No
No
No
Data on LMWH from Thomson MICROMEDEX [58]. FDA, Food and Drug Administration; DVT, deep venous thrombosis; PE, pulmonary embolism.
anticoagulable effect and bleeding complications are less than with unfractionated heparin. Low-molecularweight heparins in comparison to unfractionated heparin have a reduced ability to inhibit thrombin, reduced nonspecific bindings to proteins, and reduced interaction with platelets. Low-molecular-weight heparins bind only weakly to the endothelial surface. The biological consequences of reduced binding of LMWH to platelets (in comparison to unfractionated heparin) is a lower production of heparin-dependent antibody, which results in a lower rate of thrombocytopenia [56]. Among 330 patients treated with LMWH, none developed thrombocytopenia [27]. Reduced binding to osteoblasts results in less activation of osteoclasts, thereby producing less osteopenia [56]. Low-molecular-weight heparins are listed in Table 85.3, and approved indications by the Food and Drug Administration for prophylaxis of DVT and treatment of DVT and PE are shown. The anticoagulant response observed with a given dose was highly correlated with
Table 85.4 Major bleeding in patients during initial treatment with low-molecular-weight heparin (Enoxaparin 1 mg/kg twice daily) for deep venous thrombosis or pulmonary embolism. Major bleeding Author [Ref]
[n/N (%)]
Simonneau [62]
0/67 (0)
Levine [61]
5/247 (2)
Decousus [63]
7/195 (3.6)
Merli [64]
4/312 (1.3)
Pooled
16/821 (1.9)
body weight [57]. Low-molecular-weight heparin is effective when given in standard doses (Factor Xa units per kilogram body weight) without laboratory monitoring [28]. Low-molecular-weight heparin administered subcutaneously twice a day was as effective and safe as continuous intravenous unfractionated heparin in the initial treatment of DVT [42, 59–61]. Major bleeding rates during initial treatment with LMWH are shown in Table 85.4. The long-term rate of recurrent DVT was comparable with standard therapy and with unfractionated heparin [42]. Low-molecular-weight heparin administered subcutaneously once a day was also shown to be as effective as continuous intravenous unfractionated heparin in the initial treatment of DVT [62]. Low-molecular-weight heparin appears useful for treating patients with uncomplicated DVT in an outpatient setting [61]. Findings of clinical trials apply only to the particular LMWH evaluated [28]. Properties of a particular LMWH cannot be extrapolated to a different LMWH.
Heparinoids Danaparoid (Orgaran) consists of a mixture of heparin sulfate (83%), dermatan sulfate, and chondroitin sulfate [65, 66]. Only a small fraction (4%) of heparin sulfate component in danaparoid contains the pentasaccharide sequence, common to heparin and LMWHs, that has a selective inhibition of factor Xa via its high affinity to antithrombin III [67] (Table 85.5). The net effect is a more selective inhibition of factor Xa than heparin or LMWHs [67].
Danaparoid [65, 66, 68]
Oral Oral
Razaxaban [70, 72]
Rivaroxaban [73]
395 Direct thrombin inhibitor
IV Oral
Argatroban [70]
SC
Nematode anticoagulant
IV SC
Protein C [74, 75]
Soluble thrombomodulin
Va and VIIIa inhibitor
Va and VIIIa inhibitor
Va and VIIIa inhibitor
Xa and VIIa/TF inhibitor
Xa and VIIa/TF inhibitor
2–3 days
8–11 hr
23 min
50 hr
—
4 hr
45 min
25 min
60 min
5.7–9.2 hr
17 hr
80 hr
17 hr
25 hr
Half life
Clinical trials
Available
Available
Clinical trials
Clinical trials
Available
Available
Available
Available
Clinical trials
Clinical trials
Clinical trials
Available
Discontinued in US
Status
—
No
No
—
—
No
Yes (with HIT)
No
No
—
—
—
Yes
Yes
for DVT prophylaxis
FDA approved
—
No
No
—
—
No
No
No
No
—
—
—
Yes
No
DVT treatment
FDA approved for
FDA, Food and Drug Administration; TF, tissue factor; IV, intravenous; SC, subcutaneous; hr, hours; min, minutes; HIT, heparin-induced thrombocytopenia.
[76]
IV
Activated protein C [70, 75]
Factor Va and VIIIa inhibitors
Peptide c2 [70]
IV
Tifocogin [74]
inhibitors
Tissue factor pathway
Ximelegatran [70]
Direct thrombin inhibitor
IV
Direct thrombin inhibitor
IV or SC
Bivalirudin [70]
Direct thrombin inhibitor
Direct Xa inhibitor
Direct Xa inhibitor
ATIII/idraparinux Xa inhibitor
ATIII/fondaparinux Xa inhibitor
ATIII/danaparoid Xa inhibitor
of action
Mechanism
Hirudin [70]
Direct thrombin inhibitors
SC SC
Idraparinux [70, 71]
IV
Fondaparinux [69, 70]
Oligosaccharides
Route
—
No
No
—
—
No
No
No
No
—
—
—
Yes
No
for PE treatment
FDA approved
March 12, 2007
Heparinoids
Drug [Ref]
Table 85.5 Mechanism of action, route of administration, half-life, and status of newer antithrombotic agents.
BLUK077-Stein
18:31
BLUK077-Stein
March 12, 2007
18:31
396
PART IV
Prevention and Treatment of DVT and PE
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa XI
XIa IX
TF/VIIa complex
IXa
X
VIIIa
Xa
ATIII/Oligosaccharide
Fondaparinux Idraparinux
II prothrombin
Va
I fibrinogen
IIa thrombin
Ia fibrin
Figure 85.4 Site of action of the oligosaccharides, fondaparinux, and idraparinux. These combine with antithrombin III (AT III) to form an ATIII/oligosaccharide complex that blocks the action of factor Xa.
Danaparoid has been used for the prophylaxis of deep vein thrombosis (DVT) and an alternative anticoagulant in patients who develop heparin-induced thrombocytopenia (HIT) from heparin therapy [68]. Compared with heparin and LMWHs, danaparoid has almost no effect on physiologic platelet function and has low cross-reactivity with heparin-induced antibodies (HIT-IgG) against platelets [67]. Therefore, danaparoid is an alternative over heparin to avoid the complication of HIT and minimize the risk of bleeding. Danaparoid is FDA approved for DVT prophylaxis, but is no longer available in the United States [77].
Oligosaccharides Fondaparinux (ArixtraR ) and idraparinux are pentasaccharides that are indirect factor Xa inhibitors. They form a complex with antithrombin III, which produces a conformational change in antithrombin
resulting in an increase in factor Xa inhibition [78] (Figure 85.4, Table 85.5). Fondaparinux is administered subcutaneously once daily whereas idraparinux is being evaluated for once weekly injections [71, 79]. Both of these drugs are manufactured through chemical synthesis. No testing of antithrombotic levels is necessary. Idraparinux is a more highly sulfated derivative of fondaparinux and has a long half-life of 130 hours similar to that of antithrombin III [71]. Fondaparinux is not metabolized in the liver, it has few drug interactions, although its use in patients with renal impairment is contraindicated [69]. Unlike the heparins, fondaparinux and the other factor Xa inhibitors do not affect platelet function or react with heparin–platelet factor 4 antibodies, thus reducing the risk of heparininduced thrombocytopenia [78]. Fondaparinux has been shown to be effective in the prevention of venous thromboembolism (VTE) following abdominal surgery, surgery for hip fracture, hip replacement, and
BLUK077-Stein
March 12, 2007
CHAPTER 85
18:31
397
New and old anticoagulants
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa TF/VIIa complex
XI
XIa IX
X
IXa
VIIIa Xa
Figure 85.5 Site of action of the direct factor Xa inhibitors, razaxaban and rivaroxaban. These directly block the action of factor Xa.
knee replacement and it was effective for the treatment of DVT and PE [80]. Rates of major bleeding, 2–3% after hip or knee replacement, were comparable to rates of bleeding with LMWH [80]. Fondaparinux was also effective in the prevention of VTE in older acute medical patients, with no increase in major bleeding when compared to placebo [81].
Direct factor Xa inhibitors The direct factor Xa inhibitors bind and block the activity of factor Xa without the cofactor antithrombin III (Figure 85.5, Table 85.5). Razaxaban (DPC 906) is an orally active agent investigated in a phase II trial of thromboprophylaxis following knee arthroplasty. With the lower doses tested, if was effective and bleeding was comparable to the rate seen with LMWH [82]. Rivaroxaban is also an orally active factor Xa inhibitor being investigated for prevention of DVT in patients undergoing knee replacement surgery. It has shown short-term efficacy comparable to that of LMWH [83].
Direct thrombin inhibitors Direct thrombin inhibitors bind directly to thrombin (factor IIa) to inhibit its role in the coagulation pathway (Figure 85.6, Table 85.5). Thrombin bound to fibrin or fibrin degradation products is resistant to
II prothrombin Va
Razaxaban Rivaroxaban IIa thrombin
I fibrinogen
Ia fibrin
inhibition by the heparin/antithrombin complex, but is susceptible to inactivation by direct thrombin inhibitors [23, 84]. This allows an inhibition of fibrinbound thrombin and fluid phase thrombin resulting in greater attenuation of thrombus growth [85]. Direct thrombin inhibitors are not susceptible to neutralization by platelet factor 4, unlike heparin, allowing the anticoagulant effect to be maintained in the vicinity of platelet-rich thrombi [85]. And they also produce a predictable anticoagulant response because they do not bind to plasma proteins [86]. There are three parenterally administered direct thrombin inhibitors: hirudin, argatroban, and bivalirudin. Hirudin and argatroban are approved for the treatment of patients with HIT, whereas bivalirudin is licensed as an alternative to heparin in patients undergoing percutaneous coronary interventions [74]. Ximelagatran (Exanta) is an orally administered direct thrombin inhibitor [74]. Hirudin has a high affinity for binding thrombin, thereby inhibiting thrombin from interacting with any substrates. The hirudin–thrombin complex is almost irreversible [78]. There is no available antidote should bleeding occur [78]. Hirudin is excreted through the kidney, and should not be used in patients with renal insufficiency [79, 82]. Treatment is generally monitored with the APTT or the ecarin clotting time when higher doses of hirudin are used [78]. Bivalirudin is a synthetic hirudin that binds thrombin with much less affinity, producing only a transient
BLUK077-Stein
March 12, 2007
18:31
398
PART IV
Prevention and Treatment of DVT and PE
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa TF/VIIa complex
XI
XIa IX
X
IXa
VIIIa Xa
II IIa prothrombin Va thrombin I fibrinogen
Hirudin Bivalirudin Agratroban Ximelegatran Ia fibrin
inhibition of thrombin [91]. As a result of this lower affinity and the ability of thrombin to cleave bivalirudin, substrates are able to compete for the binding site of thrombin [70]. Therefore, the shorter halflife (Table 85.5) may confer bivalirudin with a better safety profile than hirudin [70]. In contrast to hirudin, renal excretion is not the major route of bivalirudin clearance [92]. It seems likely that bivalirudin is also degraded by endogenous peptidases and therefore may be safer than hirudin in patients with renal impairment [85]. Argatroban is similar to bivalirudin with regard to its binding of thrombin being reversible and of less affinity than hirudin. Its half-life is somewhere between that of hirudin and bivalirudin (Table 85.5). Argatroban is relatively easily monitored since it prolongs the APTT in a dose-dependent manner [70]. Unlike either hirudin or bivalirudin, it is extensively metabolized in the liver and plasma levels are not influenced by renal function [93, 94]. This makes argatroban a preferable choice in renally impaired patients, while limiting its use in those patients with hepatic dysfunction. Ximelagatran is different from all of the other direct thrombin inhibitors due to its oral bioavailability. Due to its oral administration ximelagatran is a prodrug (inactive precursor of a drug, converted into its active form in the body by normal metabolic processes) that is cleaved to form melagatran, the active form that actually inhibits thrombin.
Figure 85.6 Site of action of the direct thrombin inhibitors, hirudin, bivalirudin, agratroban, and ximelagatran. These directly block the action of thrombin (factor IIa).
The fact that ximelagatran is an oral anticoagulant draws comparison to vitamin K antagonists such as warfarin. It has potential advantages over warfarin which include: (1) rapid onset of action, (2) no food or drug interactions, (3) wide therapeutic window, and (4) short half-life [76]. Ximelagatran’s rapid onset of action obviates the need for a parenteral anticoagulant when initiating therapy in patients with thrombosis or at high risk of thrombosis. It also has a predictable anticoagulant response since there are no food or drug interactions. The wide therapeutic window allows for fixed doses without routine coagulation monitoring and the short half-life potentially reduces the need for an antidote. Considering that melagatran, the active form of ximelagatran, is eliminated via the kidneys and its use would be limited in renally impaired patients [95].
Factor VIIa/tissue factor inhibitors Factor VIIa/tissue factor (TF) complex inhibitors target the initiation step in the extrinsic pathway of the coagulation cascade. Drugs in this class include tifacogin, a recombinant form of tissue factor pathway inhibitor (TFPI) [74] and nematode anticoagulant peptide c2. One of the proposed mechanisms of action of these drugs is that they bind and inactivate factor VIIa and they also bind and inactivate factor Xa (Figure 85.7, Table 85.5). The resulting TFPI/factor Xa complex then
BLUK077-Stein
March 12, 2007
CHAPTER 85
18:31
399
New and old anticoagulants
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa
XI
XIa IX
TF/VIIa complex
IXa
X TFPI/Xa TFPI Tifacogin Nematode anticoagulant peptide c2
VIIIa Xa II prothrombin
Va
I fibrinogen
IIa thrombin
Ia fibrin
Figure 85.7 Sites of action of the tissue pathway factor inhibitors (TFPI), tifacogin and nematode anticoagulant peptide c2. These combine with factor Xa to form a Xa/TFPI
complex that blocks the action of the tissue factor/VIIa complex. Tifacogin and nematode anticoagulant peptide c2 also directly block the action of tissue factor (TF).
inhibits factor VIIa within the factor VIIa/TF complex [96]. Nematode anticoagulant peptide c2 has a half-life of 50 hours after subcutaneous injection [97]. Both drugs showed promising results in phase II trials.
drotrecogin alfa), protein C, or soluble thrombomodulin [76]. Recombinant-activated protein C and protein C have antithrombotic, anti-inflammatory, and profibrinolytic properties [98]. Although we are focusing on anticoagulant drugs, it must be noted that the inflammatory and procoagulant host responses to infection are closely related [99]. Therefore, both protein C and recombinant-activated protein C are drugs that can treat both the procoagulant and inflammatory responses associated with severe sepsis. Clinical trials have confirmed the benefit of recombinant-activated protein C in sepsis and has shown promising results in the treatment of meningococcemia. Soluble thrombomodulin binds thrombin and converts it from a procoagulant enzyme into a potent activator of protein C [76] (Figure 85.8b, Table 85.5). Soluble thrombomodulin has been evaluated in a phase II clinical trial with patients undergoing elective hip arthroplasty and apparently reduced the rate of DVT in a dose-dependent fashion [100, 101]. With a half-life
Factor Va and VIIIa inhibitors Factor Va and VIIIa inhibitors target two cofactors involved in the coagulation cascade, as opposed to the active enzymes. One potential advantage to targeting these cofactors, as opposed to the active enzymes, is increased safety and an increased therapeutic window. By focusing on the cofactors, factor Va and VIIIa inhibitors simply tone down and soften the clotting cascade signal, whereas targeting the active enzymes shuts down or prevents the signal from producing the end product of thrombin. Both factors Va and VIIIa are inactivated by activated protein C [76] (Figure 85.8a, Table 85.5). Strategies aimed at enhancing this protein C anticoagulant pathway include administration of recombinant human-activated protein C (Xigris,
BLUK077-Stein
March 12, 2007
18:31
400
PART IV
Prevention and Treatment of DVT and PE
(a)
Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa XI
XIa IX
TF/VIIa complex
IXa
X VIIIa Xa Va
Activated protein C
II prothrombin
IIa thrombin
I fibrinogen
Ia fibrin
(b) Prekallikrein High-molecular-weight kininogen XII Kallikrein XIIa
Tissue factor (TF) VIIa
XI
TF/VIIa complex
XIa IX
X
IXa
VIIIa Xa Va II prothrombin
Activated protein C II a thrombin
I fibrinogen Figure 85.8 (a) Sites of action of the Va and VIIIa inhibitor, activated protein C. Activated protein C acts directly on factors Va and VIIIa. (b) Sites of action of the Va and VIIIa
Protein C
Soluble thrombomodulin
Ia fibrin
IIa/thrombomodulin
inhibitor, soluble thrombomodulin. It acts on thrombin (IIa) to form a IIa/thrombomodulin complex that catalyzes the activation of protein C.
BLUK077-Stein
March 12, 2007
CHAPTER 85
18:31
401
New and old anticoagulants
of 2–3 days (Table 85.5) soluble thrombomodulin offers the advantage of 2–3 times per week dosing. 14
References 1 Hemker HC, Muller AD. Kinetic aspects of the interaction of blood clotting enzymes. V. The reaction mechanism of the extrinsic clotting system as revealed by the kinetics of one-stage estimations of coagulation enzymes. Thromb Diath Haemorrh 1968; 19: 368–382. 2 Hemker HC, Muller AD. Kinetic aspects of the interaction of blood-clotting enzymes. VI. Localization of the site of blood-coagulation inhibition by the protein induced by vitamin K absence (PIVKA). Thromb Diath Haemorrh 1968; 20: 78–87. 3 Hirsh J, Dalen JE, Deykin D et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest 1992; 102(suppl): 312S–326S. 4 Bussey HI, Force RW, Bianco TM et al. Reliance on prothrombin time ratios causes significant errors in anticoagulation therapy. Arch Intern Med 1992; 152: 278–282. 5 Salem DN, Stein PD, Al-Ahmad A et al. Antithrombotic therapy in valvular heart disease–native and prosthetic: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126(suppl): 457S– 482S. 6 O’Reilly RA, Aggeler PM. Determinants of the response to oral anticoagulant drugs in man. Pharmacol Rev 1970; 22: 35–96. 7 Hellemans J, Vorlat M, Verstraete M. Survival time of prothrombin and factors VII, IX, and X after complete synthesis blocking doses of coumarin derivatives. Br J Haematol 1963; 9: 505–512. 8 Vigano S, Mannucci PM, Solinas S et al. Decrease in protein C antigen and formation of an abnormal protein soon after starting oral anticoagulant therapy. Br J Haematol 1984; 57: 213–220. 9 Hull RD, Raskob GE, Hirsh J et al. Continuous intravenous heparin compared with intermittent subcutaneous heparin in the initial treatment of proximal-vein thrombosis. N Engl J Med 1986; 315: 1109–1114. 10 Hall JAG, Pauli RM, Wilson KM. Maternal and fetal sequelae of anticoagulation during pregnancy. Am J Med 1980; 68: 122–140. 11 McKenna R, Cole ER, Vasan U. Is warfarin sodium contraindicated in the lactating mother? Pediatrics 1983; 103: 325–327. 12 Lao TT, DeSwiet M, Letsky SE et al. Prophylaxis of thromboembolism in pregnancy: an alternative. Br J Obstet Gynecol 1985; 92: 202–206. 13 Das SK, Cohen AT, Edmondson RA, Melissari E, Kakkar VV. Low-molecular-weight heparin versus warfarin for
15
16
17
18
19
20
21
22
23
24
prevention of recurrent venous thromboembolism: a randomized trial. World J Surg 1996; 20: 521–527. Gonzalez-Fajardo JA, Arreba E, Castrodeza J et al. Venographic comparison of subcutaneous low-molecular weight heparin with oral anticoagulant therapy in the long-term treatment of deep venous thrombosis. J Vasc Surg 1999; 30: 283–292. Hull R, Pineo GF, Mah A et al. Long-term low molecular weight heparin treatment versus oral anticoagulant therapy for proximal deep vein thrombosis [Abstract]. Blood 2000; 96: 449a. Hull R, Pineo GF, Mah A et al. A randomized trial evaluating long-term low-molecular weight heparin therapy for three months versus intravenous heparin followed by warfarin [Abstract]. Blood 2003; 100: 556a. Kakkar VV, Gebska M, Kadziola Z, Saba N, Carrasco P. Low-molecular-weight heparin in the acute and long-term treatment of deep vein thrombosis. Thromb Haemost 2003; 89: 674–680. Gullov AL, Koefoed BG, Petersen P. Bleeding during warfarin and aspirin therapy in patients with atrial fibrillation: the AFASAK 2 study. Atrial fibrillation aspirin and anticoagulation. Arch Intern Med 1999; 159: 1322–1328. Gullov AL, Koefoed BG, Petersen P et al. Fixed minidose warfarin and aspirin alone and in combination vs adjusted-dose warfarin for stroke prevention in atrial fibrillation: second Copenhagen atrial fibrillation, aspirin, and anticoagulation study. Arch Intern Med 1998; 158: 1513–1521. Pengo V, Zasso A, Barbero F et al. Effectiveness of fixed minidose warfarin in the prevention of thromboembolism and vascular death in nonrheumatic atrial fibrillation. Am J Cardiol 1998; 82: 433–437. Matchar DB, Samsa GP, Cohen SJ, Oddone EZ, Jurgelski AE. Improving the quality of anticoagulation of patients with atrial fibrillation in managed care organizations: results of the managing anticoagulation services trial. Am J Med 2002; 113: 42–51. Padilla A, Gray E, Pepper DS, Barrowcliffe TW. Inhibition of thrombin generation by heparin and low molecular weight (LMW) heparins in the absence and presence of platelet factor 4 (PF4). Br J Haematol 1992; 82: 406– 413. Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound thrombin is protected from inhibition by heparin-antithrombin III but is susceptible to inactivation by antithrombin III-independent inhibitors. J Clin Invest 1990; 86: 385–391. Rosenberg RD, Bauer KA. The heparin-antithrombin system; a natural anticoagulant mechanism. In: Colman RW, Hirsh J, Marder VJ et al., eds. Hemostasis and Thrombosis; Basic Principles and Clinical Practice, 3rd edn. JB Lippincott, Philadelphia, PA, 1994: 837–860.
BLUK077-Stein
March 12, 2007
18:31
402
25 Lindahl U, Lidholt K, Spillmann D, Kjellen L. More to “heparin” than anticoagulation. Thromb Res 1994; 75: 1–32. 26 Hirsh J, Levine MN. Low molecular weight heparin. Blood 1992; 79: 1–17. 27 Warkentin TE, Levine MN, Hirsh J et al. Heparininduced thrombocytopenia in patients treated with lowmolecular-weight heparin or unfractionated heparin. N Engl J Med 1995; 332: 1330–1335. 28 Hyers TM, Hull RD, Weg JG. Antithrombotic therapy for venous thromboembolic disease. Chest 1995; 108(suppl): 335S–351S. 29 Brandjes DPM, Heijboer H, Buller HR et al. Acenocoumarol and heparin compared with acenocoumarol alone in the initial treatment of proximal vein thrombosis. N Engl J Med 1992; 327: 1485–1489. 30 Doyle DJ, Turpie AG, Hirsh J et al. Adjusted subcutaneous heparin or continuous intravenous heparin in patients—with acute deep vein thrombosis: a randomized trial. Ann Intern Med 1987; 107; 441–445. 31 Pini M, Pattacini C, Quintavalla R et al. Subcutaneous vs intravenous heparin in the treatment of deep venous thrombosis—a randomized clinical trial. Thromb Haemost 1990; 64: 222–226. 32 Levine MN, Hirsh J, Gent M et al. A randomized trial comparing activated thrombo-plastin time with heparin assay in patients with acute venous thromboembolism requiring large daily doses of heparin. Arch Intern Med 1994; 154: 49–56. 33 Brill-Edwards P, Ginsberg JS, Johnston M et al. Establishing a therapeutic range for heparin therapy. Ann Intern Med 1993; 119: 104–109. 34 Gitel SN, Wesler S. The antithrombotic effects of warfarin and heparin following infusions of tissue thromboplastin in rabbits: clinical implication. J Lab Clin Med 1979; 94: 481–488. 35 Wesler S, Reiner L, Freiman DG et al. Serum-induced thrombosis: studies of its induction and evolution under controlled conditions in vivo. Circulation 1959; 20: 864– 874. 36 Chiu HM, Hirsh J, Yung WL et al. Relationship between the anticoagulant and antithrombotic effects of heparin in experimental venous thrombosis. Blood 1977; 49: 171–184. 37 Basu D, Gallus A, Hirsh J et al. A prospective study of the value of monitoring heparin treatment with the activated partial thromboplastin time. N Engl J Med 1972; 287: 324–327. 38 Simon TL, Hyers TM, Gaston JP et al. Heparin pharmokinetics: increased requirements in pulmonary embolism. Br J Haemotol 1978; 39: 111–120. 39 Hirsh J, van Aken WG, Gallus AS et al. Heparin kinetics in venous thrombosis and pulmonary embolism. Circulation 1976; 53: 691–695.
PART IV
Prevention and Treatment of DVT and PE
40 Cipolle RJ, Seifert RD, Neilan BA et al. Heparin kinetics: variables related to disposition and dosage. Clin Pharmacol Ther 1981; 29: 387–393. 41 Young E, Prius M, Levine MN et al. Heparin binding to plasma proteins: an important mechanism for heparin resistance. Thromb Haemost 1992; 67: 639– 643. 42 Hull RD, Raskob GE, Pineo GF et al. Subcutaneous low-molecular-weight heparin compared with continuous intravenous heparin in the treatment of proximalvein thrombosis. N Engl J Med 1992; 326: 975– 982. 43 Levine MN, Hirsh J, Landefeld S et al. Hemorrhagic complications of anticoagulant treatment. Chest 1992; 102(suppl): 352S–363S. 44 Hull RD, Raskob GE, Rosenbloom D et al. Heparin for 5 days as compared with 10 days in the initial treatment of proximal venous thrombosis. N Engl J Med 1990; 322: 1260–1264. 45 King DJ, Kelton JG. Heparin-associated thrombocytopenia. Ann Intern Med 1984; 100: 535–540. 46 Warkentin TE, Kelton JG. Heparin-induced thrombocytopenia. In: Creger WP, ed. Annual Review in Medicine, Vol. 40. Annual Review Medicine Inc., California, 1989: 40: 31–44. 47 Hirsh J, Raschke R, Warkentin TE et al. Heparin: mechanism of action, pharmacoki-netics, dosing considerations, monitoring, efficacy and safety. Chest 1995; 108(suppl): 258S–275S. 48 Galle PC, Muss HB, McGrath KM et al. Thrombocytopenia in two patients treated with low-dose heparin. Obstet Gynecol 1978; 52(supp): 9S–11S. 49 Phillips YY, Copley JB, Stor RA. Thrombocytopenia and low-dose heparin. South Med J 1983; 76: 526–528. 50 Glock Y, Szmil E, Boudjema B et al. Cardiovascular surgery and heparin-induced thrombocytopenia. Int Angio 1988; 7: 238–245. 51 Rhodes GR, Dixon RH, Silver D. Heparin-induced thrombocytopenia: eight cases with thrombotichemorrhagic complications. Ann Surg 1977; 186: 752– 758. 52 Hirsh J, Dalen JE, Deykin D et al. Heparin: mechanism of action, pharmacokinetics, dosing considerations, monitoring, efficacy, and safety. Chest 1992; 102(suppl): 337S–351S. 53 Hull R, Delmore T, Carter C et al. Adjusted subcutaneous heparin versus warfarin sodium in the long term treatment of venous thrombosis. N Engl J Med 1982; 306: 189–194. 54 Arbit E, Goldberg M, Gomez-Orellana I, Majuru S. Oral heparin: status review. Thromb J 2006; 4: 6 [Epub ahead of print]. 55 Turpie AG. Pharmacology of the low-molecular-weight heparins. Am Heart J 1998; 135: S329– S335.
BLUK077-Stein
March 12, 2007
CHAPTER 85
18:31
403
New and old anticoagulants
56 Hirsh J. Low-molecular-weight heparin for the treatment of venous thromboembolism. Am Heart J 1998; 135: S336–S342. 57 Matzsch T, Bergqvist D, Hedner U et al. Effects of an enzymatically depolymerized heparin as compared with conventional heparin in healthy volunteers. Thromb Haemost 1987; 57: 97–101. 58 Thomson MICROMEDEX. 1974–2006 Healthcare series. Vol. 128. www.micromedex.trinity-health.org. Last accessed June 20, 2006. 59 Prandoni P, Vigo M, Cattelan AM et al. Treatment of deep venous thrombosis by fixed doses of a lowmolecular-weight heparin (CY216). Haemostasis 1990; 20(suppl): 220–223. 60 Prandoni P, Lensing AWA, Buller HR et al. Comparison of subcutaneous low molecular weight heparin with intravenous standard heparin in proximal vein thrombosis. Lancet 1992; 339: 441–445. 61 Levine M, Gent M, Hirsh J et al. A comparison of low-molecular-weight heparin administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis. N Engl J Med 1996; 334: 677–681. 62 Simonneau G, Charbonnier B, Decousus H et al. Subcutaneous low molecular weight heparin compared with continuous intravenous unfractionated heparin in the treatment of proximal deep vein thrombosis. Arch Intern Med 1993; 153: 1541–1546. 63 Decousus H, Leizorovicz A, Parent F et al., Prevention du Risque d’Embolie Pulmonaire par Interruption Cave Study Group. A clinical trial of vena caval filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. N Engl J Med 1998; 338: 409–415. 64 Merli G, Spiro TE, Olsson CG et al. Subcutaneous enoxaparin once or twice daily compared with intravenous unfractionated heparin for treatment of venous thromboembolic disease. Ann Intern Med 2001; 134: 191–202. 65 Danhof M, de Boer A, Magnani HN, Stiekema JC. Pharmacokinetic considerations on Orgaran (Org 10172) therapy. Haemostasis 1992; 22: 73–84. 66 Nurmohamed MT, Fareed J, Hoppensteadt D, Walenga JM, ten Cate JW. Pharmacological and clinical studies with Lomoparan, a low molecular weight glycosaminoglycan. Semin Thromb Hemost 1991; 17: 205–213. 67 de Valk HW, Banga JD, Wester JW et al. Comparing subcutaneous danaparoid with intravenous unfractionated heparin for the treatment of venous thromboembolism. A randomized controlled trial. Ann Intern Med 1995; 123: 1–9. 68 Shalansky K. Danapariod for heparin-induced thrombocytopenia. Drug Therapeutics Newsletter 1998; 5: 4–5. 69 Samama MM, Gerotziafas GT. Evaluation of the pharmacological properties and clinical results of the syn-
70 71
72
73
74
75
76
77
78
79
80
81
82
thetic pentasaccharide (fondaparinux). Thromb Res 2003; 109: 1–11. Bates SM, Weitz JI. Emerging anticoagulant drugs. Arterioscler Thromb Vasc Biol 2003; 23: 1491–1500. Herbert JM, Herault JP, Bernat A et al. Biochemical and pharmacological properties of SANORG 34006, a potent and long-acting synthetic pentasaccharide. Blood 1998; 91: 4197–4205. Caprini JA. Update on anticoagulants: new mechanisms and new options. Vascular Web.org http://www. vascularweb.org/ CONTRIBUTION PAGES/Research/ Past Research Initiatives Programs/2004 Research Initiatives/2004 Invited Papers/Update on Anticoagulants New Mechanisms and New Options Cap.html. Accessed June 9, 2006. Kubitza D, Becka M, Wensing G, Voith B, Zuehlsdorf M. Safety, pharmacodynamics, and pharmacokinetics of BAY 59–7939—an oral, direct Factor Xa inhibitor— after multiple dosing in healthy male subjects. Eur J Clin Pharmacol 2005; 61: 873–880. Weitz JI, Hirsh J, Samama MM. New anticoagulant drugs: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 265S– 286S. Okajima K, Koga S, Kaji M et al. Effect of protein C and activated protein C on coagulation and fibrinolysis in normal human subjects. Thromb Haemost 1990; 63: 48–53. Bates SM, Weitz JI. New anticoagulants: beyond heparin, low-molecular-weight heparin and warfarin. Br J Pharmacol 2005; 144: 1017–1028. Danaparoid. In: PDR [intranet database]. Version 5.1. r Greenwood Village, Colo: Thomson Micromedex Healthcare Series. Vol. 128. www.micromedex.trinityhealth.org. Last accessed /June 19/, 2006. Nutescu EA, Shapiro NL, Chevalier A, Amin AN. A pharmacologic overview of current and emerging anticoagulants. Cleve Clin J Med 2005; 72: S2–S6. Turpie AG, Gallus AS, Hoek JA. Pentasaccharide Investigators. A synthetic pentasaccharide for the prevention of deep-vein thrombosis after total hip replacement. N Engl J Med 2001; 344: 619–625. Fondaparinux. In: PDR [intranet database]. Version r 5.1. Greenwood Village, Colo: Thomson Micromedex Healthcare Series. Vol. 128. www.micromedex.trinityhealth.org. Last accessed June 19, 2006. Cohen AT, Davidson BL, Gallus AS et al. ARTEMIS Investigators. Efficacy and safety of fondaparinux for the prevention of venous thromboembolism in older acute medical patients: randomised placebo controlled trial. BMJ 2006; 332: 325–329. A Phase II Randomized, Double-Blind, Five-Arm, Parallel Group, Dose-Response Study to a New Oral Directly-Acting Factor Xa Inhibitor, Razaxaban, for the
BLUK077-Stein
March 12, 2007
18:31
404
83
84
85
86
87 88
89
90
91
PART IV
Prevention of Deep Vein Thrombosis in Knee Replacement Surgery—on Behalf of the Razaxaban Investigators [Abstract 41]. American Society of Hematology http://www.eurekalert.org/pub releases/2003-12/asohnai120503.php. Turpie AG, Fisher WD, Bauer KA et al. BAY 59–7939: an oral, direct factor Xa inhibitor for the prevention of venous thromboembolism in patients after total knee replacement. A phase II dose-ranging study. J Thromb Haemost 2005; 3: 2479–2486. Weitz JI, Leslie B, Hudoba M. Thrombin binds to soluble fibrin degradation products where it is protected from inhibition by heparin-antithrombin but susceptible to inactivation by antithrombin-independent inhibitors. Circulation 1998; 97: 544–552. Weitz JI, Buller HR. Direct thrombin inhibitors in acute coronary syndromes: present and future. Circulation 2002; 105: 1004–1011. Bates SM, Weitz JI. The mechanism of action of thrombin inhibitors. J Invasive Cardiol 2000; 12(Suppl F): 27F– 32F. Stringer KA, Lindenfeld J. Hirudins: antithrombin anticoagulants. Ann Pharmacother 1992; 26: 1535–1540. Walenga JM, Pifarre R, Fareed J. Recombinant hirudin as an antithrombotic agent. Drugs Future 1990; 14: 267– 280. Lefevre G, Duval M, Gauron S et al. Effect of renal impairment on the pharmacokinetics and pharmacodynamics of desirudin. Clin Pharmacol Ther 1997; 62: 50– 59. Greinacher A, Lubenow N. Recombinant hirudin in clinical practice: focus on lepirudin. Circulation 2001; 103: 1479–1484. Di Nisio M, Middeldorp S, Buller HR. Direct thrombin inhibitors. N Engl J Med 2005; 353: 1028–1040.
Prevention and Treatment of DVT and PE
92 Robson R. The use of bivalirudin in patients with renal impairment. J Invasive Cardiol 2000; 12(Suppl F): 33F– 36F. 93 Fitzgerald D, Murphy N. Argatroban: a synthetic thrombin inhibitor of low relative molecular mass. Coron Artery Dis 1996; 7: 455–458. 94 Hursting MJ, Alford KL, Becker JC et al. Novastan (brand of argatroban): a small-molecule, direct thrombin inhibitor. Semin Thromb Hemost 1997; 23: 503–516. 95 Gustafsson D, Nystrom J, Carlsson S et al. The direct thrombin inhibitor melagatran and its oral prodrug H 376/95: intestinal absorption properties, biochemical and pharmacodynamic effects. Thromb Res 2001; 101: 171–181. 96 Panteleev MA, Zarnitsina VI, Ataullakhanov FI. Tissue factor pathway inhibitor: a possible mechanism of action. Eur J Biochem 2002; 269: 2016–2031. 97 Lee AY, Vlasuk GP. Recombinant nematode anticoagulant protein c2 and other inhibitors targeting blood coagulation factor VIIa/tissue factor. J Intern Med 2003; 254: 313–321. 98 Bernard GR, Vincent JL, Laterre PF et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001; 344: 699–709. 99 Esmon CT, Taylor FB, Jr, Snow TR. Inflammation and coagulation: linked processes potentially regulated through a common pathway mediated by protein C. Thromb Haemost 1991; 66: 160–165. 100 Hirsh J, O’Donnell M, Weitz JI. New anticoagulants. Blood 2005; 105: 453–463. 101 Kearon C, Comp P, Douketis J, Royds R, Yamada K, Gent M. Dose–response study of recombinant human soluble thrombomodulin (ART-123) in the prevention of venous thromboembolism after total hip replacement. J Thromb Haemost 2005; 3: 962–968.
BLUK077-Stein
March 12, 2007
86
17:48
CHAPTER 86
Prevention of deep venous thrombosis and pulmonary embolism
Introduction The risks of deep venous thrombosis (DVT) are multiple; many are listed in Table 86.1. Pooled data indicate a risk of DVT following general surgery without prophylaxis of 19% (Table 86.2) [1]. The prevalence of DVT without prophylaxis following total hip replacement, hip fracture, and knee replacement ranges from 48 to 64% based on venography (Tables 86.3– 86.5) [1]. Newer and older antithrombotic agents are discussed in Chapter 85. Aspirin showed little or no bene-
Table 86.1 Risk factors for pulmonary embolism and deep venous thrombosis. Surgery Burns Fracture of lower extremity Immobility Bed rest Malignancy Prior DVT or PE Stroke
fit in the prevention of DVT following general surgery, hip replacement, hip fracture, and knee replacement surgeries (Tables 86.3–86.6). Pooled data on effects of treatment following general surgery, based on fibrinogen uptake studies are shown in Table 86.6. The results of these investigations, however, have been questioned because of lack of sensitivity and specificity of the fibrinogen uptake tests [2].
Intermittent pneumatic compression Intermittent pneumatic compression is effective in reducing the frequency of DVT (Tables 86.3, 86.4, and 86.6). Following neurosurgery, intermittent pneumatic compression is particularly useful because there is no risk of bleeding with this procedure. The number of patients enrolled in trials of prophylaxis using intermittent pneumatic compression is considerably fewer than in trials of low-molecular-weight heparin (LMWH) [1]. Intermittent pneumatic compression and LMWH have not been directly compared in prospective investigations.
Heart failure Myocardial infarction Increasing age Pregnancy and the postpartum period
Table 86.2 Venous thromboembolism in general surgery patients without thromboprophylaxis (pooled data).
Estrogen-containing oral contraceptives Chronic obstructive pulmonary disease
End point
Acute medical illness
Incidence [n/N (%)]
DVT (confirmed venography)
288/1507 (19)
Air travel >6 hours
Proximal DVT
83/1206 (7)
Nephrotic syndrome
All PE
82/5091 (1.6)
Obesity
Fatal PE
48/5547 (0.9)
Sickle cell disease Prolonged central venous catheterization DVT, deep venous thrombosis; PE, pulmonary embolism.
DVT, deep venous thrombosis; PE, pulmonary embolism. Reprinted and modified with permission from Geerts et al. [1].
405
BLUK077-Stein
March 12, 2007
17:48
406
PART IV
Prevention and Treatment of DVT and PE
Table 86.3 Prevention of deep venous thrombosis after total hip replacement surgery (pooled data).* Prophylaxis
Combined
Total DVT
Proximal DVT†
regimen
enrollment
(prevalence, %)
(prevalence, %)
Placebo/control
626
54
27
Elastic stockings
290
42
26
Aspirin
473
40
11
Low-dose heparin
1016
30
19
Warfarin
1828
22
5
423
20
14
IPC Recombinant hirudin
1172
16
4
LMWH
6216
16
6
Danaparoid
441
16
4
Adjusted-dose heparin
293
14
10
* Pooled DVT rates determined by routine contrast venography from randomized trials. † The denominators for proximal DVT may be slightly different than for total DVT, since some studies did not report proximal DVT rates. LMWH, low-molecular-weight heparin; IPC, intermittent pneumatic compression; DVT, deep venous thrombosis. Reprinted and modified with permission from Geerts et al. [1].
Table 86.4 Prevention of deep venous thrombosis after total knee replacement surgery (pooled data).* Prophylaxis
Combined
Total DVT
Proximal DVT†
regimen
enrollment
(prevalence, %)
(prevalence, %)
Placebo/control
199
64
15
Elastic stockings
145
61
17
Aspirin Warfarin
443
56
9
1294
47
10 11
Low-dose heparin
236
43
Venous foot pump
172
41
2
1740
31
6
110
28
7
LMWH IPC
* Pooled DVT rates determined by routine contrast venography from randomized trials. †The denominators for proximal DVT may be slightly different than for total DVT, since some studies did not report proximal DVT rates. LMWH, low-molecular-weight heparin; IPC, intermittent pneumatic compression; DVT, deep venous thrombosis. Reprinted and modified with permission from Geerts et al. [1].
Table 86.5 Prevention of deep venous thrombosis after surgery for hip fracture (pooled data).* Prophylaxis regimen
Combined enrollment
Total DVT (prevalence, %)
Placebo/control
381
48
Aspirin
171
34
59
27
LMWH/heparanoids
437
27
Warfarin
239
24
Low-dose heparin
* Pooled DVT rates determined by routine contrast venography from randomized trials. LMWH, low-molecular-weight heparin; DVT, deep venous thrombosis. Reprinted with permission from Geerts et al. [1].
BLUK077-Stein
March 12, 2007
CHAPTER 86
17:48
407
Prevention of DVT and PE
Table 86.6 Prevention of deep venous thrombosis after general surgery (pooled data).* Prophylaxis
Number
Number of
regimen
of patients
patients with DVT
Incidence (%) 25
4,310
1,084
Aspirin
Untreated controls
372
76
20
Elastic stockings
196
28
14
10,339
784
8
9,364
595
6
132
4
3
Low-dose heparin LMWH IPC
* Pooled data from randomized trials using fibrinogen leg scanning as the primary outcome. LMWH, low-molecular-weight heparin; IPC, intermittent pneumatic compression; DVT, deep venous thrombosis. Reprinted and modified with permission from Geerts et al. [1].
Graded compression elastic stockings Graded compression elastic stockings following general surgery and following elective neurosurgery seemed to reduce the frequency of DVT, although such elastic stockings failed to produce much benefit following hip replacement and total knee replacement surgeries (Tables 86.3, 86.4, and 86.6). The data regarding elastic stockings should be interpreted guardedly, however, because high-risk patients were specifically excluded in some studies [1], and patients with malignant disease have not been evaluated in sufficient numbers. Presumably, the use of graded compression elastic stockings with appropriate antithrombotic drugs would give better protection against DVT than either approach alone, but combined prophylactic measures have not been studied [1].
r r r r
r
Recommendations for prevention of venous thromboembolism The following recommendations for prevention of DVT and PE (pulmonary embolism) are from the Seventh American College of Chest Physicians Conference on Antithrombotic and Thrombolytic Therapy and reprinted with permission from Geerts et al. [2]. Grades of recommendations are defined in Table 86.7.
General recommendations r It is recommend that mechanical methods of prophylaxis be used primarily in patients who are at high risk of bleeding (Grade 1C+) or as an ad-
junct to anticoagulant-based prophylaxis (Grade 2A). Careful attention ensuring the proper use of, and optimal compliance with the mechanical device is recommended (Grade 1C+). Aspirin alone should not be used as prophylaxis against venous thromboembolism (VTE) for any patient group (Grade 1A). Clinicians should consult the manufacturer’s suggested dosing guidelines for each of the antithrombotic agents (Grade 1C). Renal impairment and function should be taken into account when deciding on doses of LMWH, fondaparinux, the direct thrombin inhibitors, and other antithrombotic drugs that are cleared by the kidneys, particularly in elderly patients and those who are at high risk for bleeding (Grade 1C+). Special caution should be exercised when using anticoagulant prophylaxis in patients undergoing neuraxial anesthesia or analgesia (Grade 1C+).
General, vascular, gynecologic, and urologic surgery General surgery r The use of specific prophylaxis other than early and persistent mobilization is recommended against in low-risk general surgery patients, defined as those who are undergoing a minor procedure, are <40 years of age, and have no additional risk factors (Grade 1C+). r Moderate-risk general surgery patients are defined as those patients undergoing a non-major procedure, and are between the ages of 40 and
BLUK077-Stein
March 12, 2007
17:48
408
PART IV
Prevention and Treatment of DVT and PE
Table 86.7 Current approach to grades of recommendations. Grade of
Clarity of
Methodological strength
recommendation
risk/benefit
of supporting evidence
Implications
1A
Clear
RCTs without important limitations
Strong recommendation; can apply to most patients in most circumstances without reservation
1C+
Clear
No RCTs but strong RCT results can be unequivocally extrapolated, or
Strong recommendation: can apply to most patients in most circumstances
overwhelming evidence from observational studies 1B
Clear
RCTs with important limitations
Strong recommendations; likely to apply to most patients
(inconsistent results, methodological flaws) 1C
Clear
Observational studies
Intermediate- strength recommendation: may change when stronger evidence is available
2A
Unclear
RCTs without important limitations
Intermediate-strength recommendation; best action may differ depending on circumstances or patients’ or societal values
2C+
Unclear
No RCTs but strong RCT results can
Weak recommendation: best action may
be unequivocally extrapolated or
differ depending on circumstances or
overwhelming evidence from
patients’ or societal values
observational studies 2B
Unclear
RCTs with important limitations (inconsistent results, methodological flaws)
2C
Unclear
Observational studies
Weak recommendation; alternative approaches likely to be better for some patients under some circumstances Very weak recommendations: other alternatives may be equally reasonable
RCT, randomized clinical trials. Reprinted with permission from Guyatt et al. [3].
60 years or have additional risk factors, or those patients who are undergoing major operations and are <40 years of age with no additional risk factors. Prophylaxis with low-dose unfractionated heparin (LDUH), 5000 U bid, or LMWH, ≤3400 U once daily is recommended (Enoxaparin 1 mg is approximately 100 Factor Xa units) (both Grade 1A). r Among higher-risk general surgery patients, defined as those patients undergoing non-major surgery and are >60 years of age or have additional risk factors, or patients undergoing major surgery who are >40 years of age or have additional risk factors, thromboprophylaxis with LDUH, 5000 U tid, or LMWH, >3400 U daily is recommended (both Grade 1A).
r In high-risk general surgery patients with multiple risk factors, pharmacologic methods (i.e., LDUH, tid, or LMWH, >3400 U daily) should be combined with the use of graduated compression stockings and/or intermittent pneumatic compression is recommended (Grade 1C+). r In general surgery patients with a high risk of bleeding, mechanical prophylaxis with properly fitted compression stockings or intermittent pneumatic compression, should be used at least initially until the bleeding risk decreases (Grade 1A). r In selected high-risk general surgery patients, including those who have undergone major cancer surgery, post-hospital discharge prophylaxis with LMWH is suggested (Grade 2A).
BLUK077-Stein
March 12, 2007
CHAPTER 86
17:48
409
Prevention of DVT and PE
Vascular surgery r In patients undergoing vascular surgery who do not have additional thromboembolic risk factors, it is suggested that clinicians do not routinely use thromboprophylaxis (Grade 2B). r For patients undergoing major vascular surgical procedures who have additional thromboembolic risk factors, prophylaxis with LDUH or LMWH is recommended (Grade 1C+). Gynecologic surgery r For gynecologic surgery patients undergoing brief procedures of ≤30 minutes for benign disease, the use of specific prophylaxis other than early and persistent mobilization is not recommended (Grade 1C+). r For patients undergoing laparoscopic gynecologic procedures, in whom additional VTE risk factors are present, the use of thromboprophylaxis with one or more of the following is recommended: LDUH, LMWH, intermittent pneumatic compression, or graduated compression stockings (all Grade 1C). r The use of thromboprophylaxis in all major gynecologic surgery patients is recommended (Grade 1A). r For patients undergoing major gynecologic surgery for benign disease, without additional risk factors, LDUH, 5000 U bid is recommended (Grade 1A). Alternatives include once-daily prophylaxis with LMWH, ≤3400 U/d (Grade 1C+), or intermittent pneumatic compression started just before surgery and used continuously while the patient is not ambulating (Grade 1B). r For patients undergoing extensive surgery for malignancy, and for patients with additional VTE risk factors, routine prophylaxis with LDUH, 5000 U tid (Grade 1A), or higher doses of LMWH (i.e., >3400 U/d) is recommended (Grade 1A). r Alternative considerations include intermittent pneumatic compression alone continued until hospital discharge (Grade 1A), or a combination of LDUH or LMWH plus mechanical prophylaxis with graduated compression stockings or intermittent pneumatic compression (all Grade 1C). r For patients undergoing major gynecologic procedures, prophylaxis should be continued until discharge from the hospital (Grade 1C). r For patients who are at particularly high risk, including those who have undergone cancer surgery
and who are >60 years of age or have previously experienced a VTE, continuing prophylaxis for 2–4 weeks after hospital discharge is suggested (Grade 2C). Urologic surgery r In patients undergoing transurethral or other low-risk urologic procedures, the use of specific prophylaxis other than early and persistent mobilization is not recommended (Grade 1C+). r For patients undergoing major, open urologic procedures, routine prophylaxis with LDUH twice daily or three times daily is recommended (Grade 1A). r Acceptable alternatives include prophylaxis with intermittent pneumatic compression and/or graduated compression stockings (Grade 1B) or LMWH (Grade 1C+). r For urologic surgery patients who are actively bleeding or are at very high risk for bleeding, the use of mechanical prophylaxis with graduated compression stockings and/or intermittent pneumatic compression at least until the bleeding risk decreases, is recommended (Grade 1C+). r For patients with multiple risk factors, a combination of graduated compression stockings and/or intermittent pneumatic compression with LDUH or LMWH is recommended (Grade 1C+).
Laparoscopic surgery r Routine thromboprophylaxis in these patients, other than aggressive mobilization is not recommended (Grade 1A). r For patients undergoing laparoscopic procedures and who have additional thromboembolic risk factors, the use of thromboprophylaxis with one or more of the following is recommended: LDUH, LMWH, intermittent pneumatic compression, or graduated compression stockings (Grade 1C+).
Orthopedic surgery Elective hip arthroplasty r For patients undergoing elective total hip replacement, the routine use of one of the following three anticoagulants is recommended: 1 LMWH (at a usual high-risk dose, started 12 hours before surgery or 12–24 hours after surgery,
BLUK077-Stein
March 12, 2007
17:48
410
or 4–6 hours after surgery at half the usual highrisk dose and then increasing to the usual highrisk dose the following day); 2 Fondaparinux (2.5 mg started 6–8 hours after surgery); or 3 Adjusted-dose vitamin K antagonist started preoperatively or the evening after surgery (INR target, 2.5; INR range, 2.0–3.0) (all Grade 1A). r The use of fondaparinux over LMWH and vitamin K antaonist, or the use of LMWH over vitamin K antagonist have not been recommended, because a relatively low value on the prevention of venographic thrombosis and a relatively high value on minimizing bleeding complications are placed. r The use of aspirin, dextran, LDUH, graduated compression stockings, intermittent pneumatic compression, or venous foot pump as the only method of thromboprophylaxis in these patients is recommended against (Grade 1A). Elective knee arthroplasty r For patients undergoing elective total knee arthroplasty, routine thromboprophylaxis using LMWH (at the usual high-risk dose), fondaparinux, or adjusted-dose vitamin k antagonist (INR target, 2.5; INR range, 2.0–3.0) is recommended (all Grade 1A). r The use of fondaparinux over LMWH and vitamin K antagonist, or LMWH over vitamin K antagonist have not been recommended, because a relatively low value on the prevention of venographic thrombosis and a relatively high value on minimizing bleeding complications were placed. r The optimal use of intermittent pneumatic compression is an alternative option to anticoagulant prophylaxis (Grade 1B). r The use of any of the following as sole methods of thromboprophylaxis: aspirin (Grade 1A); LDUH (Grade 1A); or venous foot pump (Grade 1B) are recommended against.
PART IV
longed or complicated procedure, thromboprophylaxis with LMWH is suggested (Grade 2B). Hip fracture surgery r For patients undergoing hip fracture surgery, the routine use of fondaparinux (Grade 1A), LMWH at the usual high-risk dose (Grade 1C+), adjusted-dose vitamin K antagonist (INR target, 2.5; INR range, 2.0–3.0) (Grade 2B), or LDUH (Grade 1B) is recommended. r The use of aspirin alone is recommended against (Grade 1A). r If surgery is likely to be delayed, prophylaxis with either LDUH or LMWH be initiated during the time between hospital admission and surgery (Grade 1C+) is recommended. r If anticoagulant prophylaxis is contraindicated because of a high risk of bleeding, mechanical prophylaxis is recommended (Grade 1C+).
Other prophylaxis issues in major orthopedic surgery r For major orthopedic surgical procedures, a deci-
r
r
r Knee arthroscopy r It is suggested that clinicians do not routinely use thromboprophylaxis in these patients, other than early mobilization (Grade 2B). r For patients undergoing arthroscopic knee surgery who are at a higher than usual risk, based on preexisting VTE risk factors or following a pro-
Prevention and Treatment of DVT and PE
r
sion about the timing of the initiation of pharmacologic prophylaxis is recommended to be based on the efficacy-to-bleeding tradeoffs for that particular agent (Grade 1A). For LMWH, there are only small differences between starting preoperatively or postoperatively, and both options are acceptable (Grade 1A). The routine use of Doppler ultrasonography screening at the time of hospital discharge in asymptomatic patients following major orthopedic surgery is recommended against (Grade 1A). It is recommended that patients undergoing total hip replacement, total knee arthroplasty, or hip fracture surgery receive thromboprophylaxis with LMWH (using a high-risk dose), fondaparinux (2.5 mg daily), or a vitamin K antagonist (INR target, 2.5; INR range, 2.0–3.0) for at least 10 days (Grade 1A). It is recommended that patients undergoing total hip replacement or hip fracture surgery be given extended prophylaxis for up to 28–35 days after surgery (Grade 1A). The recommended options for total hip replacement include LMWH (Grade 1A), a vitamin K antagonist (Grade 1A), or fondaparinux (Grade 1C+).
BLUK077-Stein
March 12, 2007
CHAPTER 86
17:48
Prevention of DVT and PE
r The recommended options following hip fracture surgery are fondaparinux (Grade 1A), LMWH (Grade 1C+), or a vitamin K antagonist (Grade 1C+).
Elective spine surgery r For spinal surgery patients with no additional risk factors, the routine use of any thromboprophylaxis modality, apart from early and persistent mobilization, is recommended against (Grade 1C). r It is recommended that some form of prophylaxis be used in patients undergoing spinal surgery who exhibit additional risk factors such as advanced age, known malignancy, presence of a neurologic deficit, previous VTE, or an anterior surgical approach (Grade 1B). r For patients with additional risk factors, any of the following prophylaxis options is recommended: postoperative LDUH alone (Grade 1C+); postoperative LMWH alone (Grade 1B); or perioperative intermittent pneumatic compression alone (Grade 1B). Other considerations include perioperative graduated compression stockings alone (Grade 2B), or perioperative intermittent pneumatic compression combined with graduated compression stockings (Grade 2C). In patients with multiple risk factors for VTE, it is recommended to combine LDUH or LMWH with graduated compression stockings and/or intermittent pneumatic compression (Grade 1C+).
Isolated lower extremity injuries r It is suggested that clinicians do not routinely use thromboprophylaxis in patients with isolated lower extremity injuries (Grade 2A).
Neurosurgery r It is recommended that thromboprophylaxis be routinely used in patients undergoing major neurosurgery (Grade 1A). r The use of intermittent pneumatic compression with or without graduated compression stockings in patients undergoing intracranial neurosurgery is recommended (Grade 1A). r Acceptable alternatives to the above options are prophylaxis with LDUH (Grade 2B) or postoperative LMWH (Grade 2A).
411
r The combination of mechanical prophylaxis (i.e., graduated compression stockings and/or intermittent pneumatic compression) and pharmacologic prophylaxis (i.e., LDUH or LMWH) in high-risk neurosurgery patients is suggested (Grade 2B).
Trauma, spinal cord injury, burns Trauma r All trauma patients with at least one risk factor for VTE should receive thromboprophylaxis, if possible (Grade 1A). r In the absence of a major contraindication, it is recommended that clinicians use LMWH prophylaxis starting as soon as it is considered safe to do so (Grade 1A). r Mechanical prophylaxis with intermittent pneumatic compression, or possibly with graduated compression stockings alone, is recommended to be used if LMWH prophylaxis is delayed or if it is currently contraindicated due to active bleeding or a high risk for hemorrhage (Grade 1B). r Doppler ultrasound screening in patients who are at high risk for VTE (e.g., the presence of a spinal cord injury, lower extremity or pelvic fracture, major head injury, or an indwelling femoral venous line), and who have received suboptimal prophylaxis or no prophylaxis is recommended (Grade 1C). r The use of inferior vena cava filters as primary prophylaxis in trauma patients is not recommended (Grade 1C). r The continuation of thromboprophylaxis until hospital discharge, including the period of inpatient rehabilitation is recommended (Grade 1C+). Continuation of prophylaxis after hospital discharge with LMWH or a vitamin K antagonist (INR target, 2.5; INR range, 2.0–3.0) in patients with major impaired mobility is suggested (Grade 2C). Acute spinal cord injury r It is recommended that thromboprophylaxis be provided for all patients with acute spinal cord injuries (Grade 1A). r The use of LDUH, graduated compression stockings, or intermittent pneumatic compression as single prophylaxis modalities is recommended against (Grade 1A).
BLUK077-Stein
March 12, 2007
17:48
412
r In patients with acute spinal cord injury, prophylaxis with LMWH, to be commenced once primary hemostasis is evident is recommended (Grade 1B). The combined use of intermittent pneumatic compression and either LDUH (Grade 2B) or LWMH (Grade 2C) as alternatives to LMWH is suggested. r The use of intermittent pneumatic compression and/or graduated compression stockings is recommended when anticoagulant prophylaxis is contraindicated early after injury (Grade 1C+). r The use of an inferior vena cava filter as primary prophylaxis against PE is recommended against (Grade 1C). r During the rehabilitation phase following acute spinal cord injury, the continuation of LMWH prophylaxis or conversion to an oral vitamin K antagonist (INR target, 2.5; INR range, 2.0–3.0) is recommended (Grade 1C). Burns r It is recommended that burn patients with additional risk factors for VTE, including one or more of the following: advanced age, morbid obesity, extensive or lower extremity burns, concomitant lower extremity trauma, use of a femoral venous catheter, and/or prolonged immobility, receive thromboprophylaxis, if possible (Grade 1C+). r If there are no contraindications, the use of either LDUH or LMWH, starting as soon as it is considered safe to do so is recommended (Grade 1C+).
Medical conditions r In acutely ill medical patients who have been admitted to the hospital with congestive heart failure or severe respiratory disease, or who are confined to bed and have one or more additional risk factors, including active cancer, previous VTE, sepsis, acute neurologic disease, or inflammatory bowel disease, prophylaxis with LDUH (Grade 1A) or LMWH (Grade 1A) is recommended. r In medical patients with risk factors for VTE, and in whom there is a contraindication to anticoagulant prophylaxis, the use of mechanical prophylaxis with compression stockings or intermittent pneumatic compression (Grade 1C+) is recommended.
PART IV
Prevention and Treatment of DVT and PE
Cancer patients r It is recommended that cancer patients undergoing surgical procedures receive prophylaxis that is appropriate for their current risk state (Grade 1A). Refer to the recommendations in the relevant surgical subsections. r It is recommended that hospitalized cancer patients who are bedridden with an acute medical illness receive prophylaxis that is appropriate for their current risk state (Grade 1A). Refer to the recommendations in the section dealing with medical patients. r It is suggested that clinicians do not routinely use prophylaxis to try to prevent thrombosis related to long-term indwelling central venous catheters in cancer patients (Grade 2B). Specifically, it is suggested that clinicians do not use LMWH (Grade 2B), and the use of fixed-dose warfarin (Grade 1B) for this indication is recommended against.
Critical care r It is recommended that, on admission to a critical care unit, all patients be assessed for their risk of VTE. Accordingly, most patients should receive thromboprophylaxis (Grade 1A). r For patients who are at high risk for bleeding, mechanical prophylaxis with graduated compression stockings and/or intermittent pneumatic compression until the bleeding risk decreases (Grade 1C+) is recommended. r For ICU patients who are at moderate risk for VTE (e.g., medically ill or postoperative patients), the use of LDUH or LMWH prophylaxis (Grade 1A) is recommended. r For patients who are at higher risk, such as that following major trauma or orthopedic surgery, LMWH prophylaxis (Grade 1A) is recommended.
Long distance travel r The following general measures for long-distance travelers (i.e., flights of >6 hour duration) are recommended: avoidance of constrictive clothing around the lower extremities or waist; avoidance of dehydration; and frequent calf muscle stretching (Grade 1C).
BLUK077-Stein
March 12, 2007
CHAPTER 86
17:48
413
Prevention of DVT and PE
r For long-distance travelers with additional risk factors for VTE, the general strategies listed above are recommended. If active prophylaxis is considered, because of the perceived increased risk of venous thrombosis, the use of properly fitted, below-knee graduated compression stockings providing 15–30 mm Hg of pressure at the ankle (Grade 2B), or a single prophylactic dose of LMWH injected prior to departure (Grade 2B) is suggested. r The use of aspirin for VTE prevention associated with travel (Grade 1B) is recommended against.
References 1 Geerts WH, Heit JA, Clagett GP et al. Prevention of venous thromboembolism. Chest 2001; 119: 132S–175S. 2 Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the seventh ACCP Conference on antithrombotic and thrombolytic therapy. Chest 2004; 126: 338S–400S. 3 Guyatt G, Schunemann HJ, Cook D, Jaeschke R, Pauker S. Applying the grades of recommendation for antithrombotic and thrombolytic therapy: the seventh ACCP Conference on antithrombotic and thrombolytic therapy. Chest 2004; 126: 179S–187S.
BLUK077-Stein
March 12, 2007
87
18:11
CHAPTER 87
Treatment of deep venous thrombosis and acute pulmonary embolism
Introduction In the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) among 399 patients (73% treated only with anticoagulants), the cumulative rate of recurrent pulmonary embolism (PE) was 4.0% in 1 week and 8.3% in 1 year [1]. Among patients who had a recurrent PE, 27% were fatal [1]. The cumulative mortality from PE was 1.3% in 1 day, 2.0% in 1 week, 2.3% in 2 weeks, and 2.5% in 1 year [1]. Among patients who died of PE, 90% died of recurrent PE [1]. Pulmonary embolism carries a higher risk of recurrent fatal PE in treated patients than the risk of fatal PE in treated patients with DVT (deep venous thrombosis). In an investigation by Douketis and associates, fatal recurrent PE within the first year occurred in 1.5% of patients treated for PE and fatal PE within the first year occurred in 0.4% of patients treated for DVT [2]. Acute PE in patients with isolated deep calf-vein thrombosis has been reported [3, 4]. It is now recommended that isolated deep calf-vein thrombosis should be treated with heparin followed by oral anticoagulants [5]. However, the safety of following untreated patients with normal serial noninvasive leg tests obtained over a period of 14 days [6] has also been shown [7–9]. The importance of immediate treatment, while awaiting confirmatory tests, has been stated over the years [10] and emphasized in recent literature [11]. Most members of the American College of Chest Physicians Consensus Committee on Pulmonary Embolism believed that the risk of recurrent untreated PE during the period of diagnostic tests exceeds the risk of bleeding from heparin, and therefore recommended immediate administration of heparin [6]. Some reserve the immediate administration for those with poor cardiopulmonary reserve [6]. The decision of whether to
414
treat before the diagnosis is confirmed by objective tests depends on the likelihood and severity of suspected acute PE, the cardiopulmonary status and the potential risk of bleeding with heparin [6]. Initial therapy with heparin for proximal DVT prevents late PE or recurrent DVT as well as early events [12–14]. Among patients with proximal DVT treated only with a vitamin K antagonist, symptomatic recurrences occurred in 20% compared with 6.7% treated with unfractionated heparin (UFH) followed by a vitamin K antagonist [13]. The frequency of late PE or recurrent DVT as well as early VTE (venous thromboembolism) was higher in patients who received only a vitamin K antagonist [13]. Similarly, SC (subcutaneous) UFH for the initial treatment of proximal DVT, which induced an initial anticoagulant response below the target range in majority of the patients, resulted in a high frequency of recurrent VTE [12]. The VTE was virtually confined to patients with a subtherapeutic anticoagulant response [12]. In contrast, continuous IV (intravenous) UFH induced a therapeutic response in the majority of patients. This resulted in a low frequency of recurrent VTE, which was limited to those with an initial subtherapeutic anticoagulant response [12]. Most recurrences were during heparin therapy or relatively early after its discontinuation [12]. There were no episodes of recurrent VTE during long-term therapy with warfarin in patients in whom initial heparin therapy had been in the prescribed therapeutic range [12]. Intravenous initial treatment with UFH for 5–7 days is as effective as treatment with UFH for longer durations [15, 16]. Meta-analysis in 2000 comparing IV UFH and LMWH (low-molecular-weight heparin) for the initial treatment of acute DVT showed no difference in the rate of recurrent VTE or major bleeding
BLUK077-Stein
March 12, 2007
CHAPTER 87
18:11
Treatment of DVT and acute PE
[17]. No apparent differences were observed among the different LMWHs [17]. Patients who received only heparin for several days, not followed by a vitamin K antagonist, suffered high rates (20%) of recurrent DVT or an extension of DVT [18]. Patients treated only 4–6 weeks with a vitamin K antagonist had higher rates of recurrent DVT than those treated 3–6 months [19–21]. The long-term outcome was assessed in 355 patients with acute DVT treated with heparin followed by vitamin K antagonists (INR range, 2.0–3.0) for 3 months and venous compression stockings for 2 years [22]. The cumulative rate of postthrombotic syndrome was 17% at 1 year, 23% at 2 years, 28% at 5 years, and 29% at 8 years [22]. The very mildest cases were not included [22]. Higher proportions (50–89%) showed the postthrombotic syndrome among those who received only anticoagulants in randomized controlled trials of thrombolytic therapy [23–25]. Severe postthrombotic syndrome after 1 year, defined by some as healed or active ulceration and additional signs of congestion [24] occurred in about 3% in 1 year, according to Prandoni and associates [22], and in ≤1% in control groups of randomized trials [24, 25]. The cumulative rate of recurrent DVT or PE was 4.9% in 3 months, 8.6% in 6 months, 17.5% in 2 years, 24.6% in 5 years, and 30.3% in 8 years [22]. Patients with transient risk factors for DVT such as surgery or recent trauma had about a third to half the rate of recurrences [22]. Treatment for 1–2 years reduced the risk of recurrent DVT while on therapy [26, 27] but not after therapy was discontinued [27]. Low-intensity warfarin therapy (INR range, 1.5–1.9) was less effective than therapy at an INR of 2.0–3.0 (recurrent VTE 1.9%/yr versus 0.6%/yr) with similar rates of major bleeding (0.96%/yr with low INR versus 0.93%/yr with standard INR) [28]. Long-term (2.7 year) warfarin at a higher intensity (INR range, 3.1–4.0) did not reduce the rate of recurrent VTE compared with patients treated at an INR of 2.0–3.0 [29]. Long-term (6 months) treatment of DVT with LMWH (dalteparin) was compared with a vitamin K antagonist after initial treatment with dalteparin [30]. Recurrent VTE occurred in 8.0% treated with long-term dalteparin and 15.7% treated with a vitamin K antagonist [30]. Major bleeding occurred in 6 and 4%, respectively [30]. Others compared 3 months of LMWH (tinzaparin) with a vitamin K antagonist
415
for treatment of proximal DVT [31] after both groups were initially treated with tinzaparin. Similar rates of recurrent VTE occurred in both groups [31]. Home treatment of patients with proximal DVT for initial treatment with LMWH followed by a vitamin K antagonist has been shown to be safe [32–34]. Home treatment of PE has not been tested, and at this time, is not appropriate. Based on the observation that home treatment of DVT was as safe as treatment in hospitals with IV infusions of UFH [32–34], it appears that bed rest is unnecessary. Subsequently, small studies showed that the prevalence of silent PE identified by ventilation– perfusion lung scans was not reduced by bed rest among patients treated with anticoagulants [35, 36].
Recommendations for treatment of venous thromboembolism The following recommendations for treatment of DVT and PE are from the Seventh American College of Chest Physicians Conference on Antithrombotic and Thrombolytic Therapy and reprinted with permission from Buller et al. [11]. Grades of recommendations are defined in Table 86.7 of Chapter 86.
Treatment of deep venous thrombosis
Initial treatment of acute DVT of the leg r For patients with objectively confirmed DVT, short-term treatment with SC LMWH or IV UFH or SC UFH (all Grade 1A) is recommended. r For patients with a high clinical suspicion of DVT, treatment with anticoagulants while awaiting the outcome of diagnostic tests (Grade 1C+) is recommended. r In acute DVT, initial treatment with LMWH or UFH for at least 5 days (Grade 1C) is recommended. r Vitamin K antagonists should be initiated together with LMWH or UFH on the first treatment day and discontinuation of heparin when the INR is stable and >2.0 (Grade 1A). IV UFH for the initial treatment of DVT r If IV UFH is chosen, it should be administered by continuous infusion with dose adjustment to achieve and maintain an activated partial thromboplastin time (APTT) prolongation
BLUK077-Stein
March 12, 2007
18:11
416
corresponding to plasma heparin levels from 0.3 to 0.7 IU/mL anti-Xa activity by the amidolytic assay (Grade 1C+). r In patients requiring large daily doses of UFH without achieving a therapeutic APTT, the measurement of the anti-Xa level for dose guidance (Grade 1B) is recommended.
SC UFH for the initial treatment of DVT r In patients with acute DVT, it is recommended that SC-administered UFH can be used as an adequate alternative to IV UFH (Grade 1A). r For patients who receive SC UFH, an initial dose of 35,000 U/24 hr SC, with subsequent dosing to maintain the APTT in the therapeutic range (Grade 1C+) is recommended. LMWH for the initial treatment of DVT r In patients with acute DVT, initial treatment with LMWH SC once or twice daily over UFH as an outpatient, if possible (Grade 1C), and as inpatient, if necessary, (Grade 1A) is recommended. r In patients with acute DVT treated with LMWH, routine monitoring with anti-factor Xa level measurements (Grade 1A) is recommended against. r In patients with severe renal failure, the use of IV UFH over LMWH (Grade 2C) is suggested. Systematically administered thrombolysis in the initial treatment of DVT r In patients with DVT, the routine use of IV thrombolytic treatment (Grade 1A) is recommended against. r In selected patients, such as those with massive ileofemoral DVT at risk of limb gangrene secondary to venous occlusion, IV thrombolysis (Grade 2C) is suggested. Catheter-directed thrombolysis in the initial treatment of DVT r In patients with DVT, the routine use of catheterdirected thrombolysis (Grade 1C) is recommended against. r It is suggested that this treatment should be confined to selected patients such as those requiring limb salvage (Grade 2C).
PART IV
Prevention and Treatment of DVT and PE
Catheter extraction or fragmentation and surgical thrombectomy for the initial treatment of DVT r In patients with DVT, the routine use of venous thrombectomy (Grade 1C) is recommended against. r In selected patients such as patients with massive ileofemoral DVT at risk of limb gangrene secondary to venous occlusion, venous thrombectomy is recommended (Grade 2C). Vena caval interruption for the initial treatment of DVT r For most patients with DVT, the routine use of a vena cava filter in addition to anticoagulants (Grade 1A) is recommended against. r The placement of an inferior vena caval filter is suggested in patients with a contraindication for, or a complication of anticoagulant treatment (Grade 2C), as well as in those with recurrent thromboembolism despite adequate anticoagulation (Grade 2C). Nonsteroidal anti-inflammatory agents for the initial treatment of DVT r For the initial treatment of DVT, the use of nonsteroidal anti-inflammatory agents (Grade 2B) is recommended against. Immobilization r For patients with DVT, it is recommended to ambulate as tolerated (Grade 1B). Long-term treatment of acute DVT of the leg
Vitamin K antagonists for the long-term treatment of DVT r For patients with a first episode of DVT secondary to a transient (reversible) risk factor, long-term treatment with a vitamin K antagonist for 3 months over treatment for shorter periods (Grade 1A) is recommended.* This recommendation applies both to patients with proximal vein
* This recommendation ascribes a relatively high value to preventing recurrent thromboembolic events and a relatively low value on bleeding and cost.
BLUK077-Stein
March 12, 2007
CHAPTER 87
r r r
r
r
r r
r
r
18:11
417
Treatment of DVT and acute PE
thrombosis, and to patients with symptomatic DVT confined to the calf veins. For patients with a first episode of idiopathic DVT, treatment with a vitamin K antagonist at least 6– 12 months (Grade 1A) is recommended. It is suggested that patients with first-episode idiopathic DVT be considered for indefinite anticoagulant therapy (Grade 2A).* For patients with DVT and cancer, LMWH for the first 3–6 months of long-term anticoagulant therapy (Grade 1A) is recommended. For these patients, anticoagulant therapy is recommended indefinitely or until the cancer is resolved (Grade 1C). For patients with a first episode of DVT who have documented antiphospholipid antibodies or who have two or more thrombophilic conditions (e.g., combined factor V Leiden and prothrombin 20210 gene mutations), 12 months of treatment is recommended (Grade 1C+). Indefinite anticoagulant therapy is suggested in these patients (Grade 2C).* For patients with a first episode of DVT who have documented deficiency of antithrombin, deficiency of protein C or protein S, or the factor V Leiden or prothrombin 20210 gene mutation, homocysteinemia, or high factor VIII levels (>90th percentile of normal), treatment for 6–12 months is recommended (Grade 1A). Indefinite therapy for patients with idiopathic thrombosis is suggested (Grade 2C).* For patients with two or more episodes of objectively documented DVT, indefinite treatment is suggested (Grade 2A). The dose of vitamin K antagonist should be adjusted to maintain a target INR of 2.5 (INR range, 2.0 and 3.0) for all treatment durations (Grade 1A). High-intensity vitamin K antagonist therapy (INR range, 3.1–4.0) (Grade1A) and lowintensity therapy (INR range, 1.5–1.9) compared to INR range of 2.0–3.0 (Grade 1A) is recommended against. In patients who receive indefinite anticoagulant treatment, the risk–benefit of continuing such treatment should be reassessed in the individual patient at periodic intervals (Grade 1C). Repeat testing with compression ultrasonography for the presence or absence of residual thrombosis or measurement of plasma D-dimer (Grade 2C) is suggested.
LMWH for the long-term treatment of DVT r For most patients with DVT and cancer, treatment with LMWH for at least the first 3–6 months of long-term treatment (Grade 1A) is recommended. The regimens of LMWH that have been established to be effective for long-term treatment in randomized trials are dalteparin, 200 IU/kg body weight qd for 1 month, followed by 150 IU/kg qd thereafter, or tinzaparin at 175 IU/kg body weight SC qd. The postthrombotic syndrome
Elastic stockings for the prevention of the postthrombotic syndrome r The use of an elastic compression stocking with a pressure of 30–40 mm Hg at the ankle during 2 years after an episode of DVT is recommended (Grade 1A). Physical and drug treatment of the postthrombotic syndrome r A course of intermittent pneumatic compression for patients with severe edema of the leg due to postthrombotic syndrome (Grade 2B), or the use of elastic compression stockings for patients with mild edema of the leg due to the postthrombotic syndrome (Grade 2C), is suggested. r In patients with mild edema due to postthrombotic syndrome, administration of rutosides (Grade 2B), is suggested. Initial treatment of acute PE
IV UFH or LMWH for the initial treatment of PE r For patients with objectively confirmed nonmassive PE, short-term treatment with SC LMWH, or IV UFH is recommended (both Grade 1A). r Patients with a high clinical suspicion of PE should be treated with anticoagulants while awaiting the outcome of diagnostic tests (Grade 1C+). r In patients with acute nonmassive PE, LMWH over UFH is recommended (Grade 1A). r In acute nonmassive PE, initial treatment should be with LMWH or UFH for at least 5 days (Grade 1C).
BLUK077-Stein
March 12, 2007
18:11
418
r In patients with acute nonmassive PE treated with LMWH, routine monitoring with anti-factor Xa levels is recommended against (Grade 1A). r In patients with severe renal failure, IV UFH should be chosen over LMWH (Grade 2C). r If IV UFH is chosen, administration by continuous infusion with dose adjustment to achieve and maintain an APTT prolongation corresponding to plasma heparin levels from 0.3 to 0.7 IU/mL anti-Xa activity by the amidolytic assay (Grade 1C+) is recommended. r In patients requiring large daily doses of UFH without achieving a therapeutic APTT, the measurement of the anti-Xa level for dose guidance should be done (Grade 1B). r Vitamin K antagonist should be initiated together with LMWH or UFH on the first treatment day and discontinuation of heparin when the INR is stable and >2.0 (Grade 1A). Systemically and locally administered thrombolytic drugs for the initial treatment of PE r For most patients with PE, it is recommended to clinicians not to use systemic thrombolytic therapy (Grade 1A). In selected patients, systemic administration of thrombolytic therapy (Grade 2B) is suggested. For patients who are hemodynamically unstable, the use of thrombolytic therapy (Grade 2B) is suggested. r We suggest clinicians not to use local administration of thrombolytic therapy via a catheter (Grade 1C). r For patients with PE who receive thrombolytic regimens, it is suggested to use thrombolytic regimens with a short infusion time over those with prolonged infusion times (Grade 2C). Catheter extraction or fragmentation for the initial treatment of PE r For most patients with PE, the use of mechanical approaches (Grade 1C) is recommended against. In selected highly compromised patients who are unable to receive thrombolytic therapy or whose critical status does not allow sufficient time to infuse thrombolytic therapy, the use of mechanical approaches (Grade 2C) is suggested. Pulmonary embolectomy for the initial treatment of PE r For most patients with PE, pulmonary embolectomy (Grade 1C) is recommended against. In se-
PART IV
Prevention and Treatment of DVT and PE
lected highly compromised patients who are unable to receive thrombolytic therapy or whose critical status does not allow sufficient time to infuse thrombolytic therapy, pulmonary embolectomy is suggested (Grade 2C).
Vena caval interruption for the initial treatment of PE r In PE patients with a contraindication for, or a complication of anticoagulant treatment, as well as in those with recurrent thromboembolism despite adequate anticoagulation, placement of an inferior vena caval filter (both Grade 2C) is suggested. Long-term treatment of acute PE
Vitamin K antagonists for the long-term treatment of PE r Patients with a first episode of PE secondary to a transient (reversible) risk factor, should be placed on long-term treatment with a vitamin K antagonist for at least 3 months (Grade 1A). r Patients with a first episode of idiopathic PE should be treated with a vitamin K antagonist for at least 6–12 months (Grade 1A). r Patients with first-episode idiopathic PE should be considered for indefinite anticoagulant therapy (Grade 2A).* r Patients with PE and cancer should be treated with LMWH for the first 3–6 months of long-term anticoagulant therapy (Grade 1A). These patients should then receive anticoagulant therapy indefinitely or until the cancer is resolved (Grade 1C). r Patients with a first episode of PE who have documented antiphospholipid antibodies or who have two or more thrombophilic conditions (e.g., combined factor V Leiden and prothrombin 20210 gene mutations) should be treated for 12 months (Grade 1C+). For these patients, we suggest indefinite anticoagulant therapy (Grade 2C).* r Patients with a first episode of PE who have documented deficiency of antithrombin, deficiency of protein C or protein S, or the factor V Leiden or prothrombin 20210 gene mutation, homocysteinemia, or high factor VIII levels (>90th percentile of normal), should be placed on treatment for 6–12 months (Grade 1A). Indefinite therapy for patients with idiopathic PE is suggested (Grade 2C).*
BLUK077-Stein
March 12, 2007
CHAPTER 87
18:11
419
Treatment of DVT and acute PE
r Patients with two or more episodes of objectively documented PE should be treated indefinitely (Grade 2A). r The dose of vitamin K antagonist should be adjusted to maintain a target INR of 2.5 (INR range, 2.0 and 3.0) for all treatment durations (Grade 1A). High-intensity VKA therapy (INR range, 3.1–4.0) (Grade 1A) and low-intensity therapy (INR range, 1.5–1.9) compared to INR range of 2.0–3.0 (Grade 1A) is recommended against. r In patients who receive indefinite anticoagulant treatment, the risk–benefit of continuing such treatment should be reassessed in the individual patient at periodic intervals (Grade 1C). LMWH for the long-term treatment of PE r For most patients with PE and concurrent cancer, treatment with LMWH for at least the first 3–6 months of long-term treatment is recommended (Grade 1A). The LMWH regimens that have been established to be effective for long-term treatment are dalteparin, 200 IU/kg body weight qd for 1 month followed by 150 IU/kg qd thereafter, and tinzaparin at 175 IU/kg body weight SC qd. Chronic thromboembolic pulmonary hypertension
Pulmonary thromboendarterectomy, vitamin K antagonists, and caval filter for the treatment of chronic thromboembolic pulmonary hypertension r In selected patients with chronic thromboembolic pulmonary hypertension, i.e., patients with central disease under the care of an experienced surgical/medical team, pulmonary thromboendarterectomy is recommended (Grade 1C). r Life-long treatment with vitamin K antagonists is recommended to an INR of 2.0–3.0 to be administered following pulmonary thromboendarterectomy, and also be administered to patients with chronic thromboembolic pulmonary hypertension who are ineligible for pulmonary thromboendarterectomy (Grade 1C). r Placement of a vena cava filter before or at the time of pulmonary thromboendarterectomy for patients with chronic thromboembolic pulmonary hypertension is suggested (Grade 2C).
Superficial thrombophlebitis
Treatment for superficial thrombophlebitis r For patients with superficial thrombophlebitis as a complication of an infusion, topical diclofenac gel (Grade 1B) or oral diclofenac (Grade 2B) is suggested. r For patients affected by spontaneous superficial thrombophlebitis, intermediate dosages of UFH or LMWH for at least 4 weeks is suggested (Grade 2B). Acute upper-extremity DVT
IV UFH or LMWH for the initial treatment of upper extremity DVT r Patients with acute upper-extremity DVT should be initially treated with UFH (Grade 1C+) or LMWH (Grade 1C+). Thrombolytic therapy for the initial treatment of upper extremity DVT r In selected patients with acute upper-extremity DVT, e.g., in those with a low risk of bleeding and symptoms of recent onset, a short course of thrombolytic therapy for initial treatment is suggested (Grade 2C). Catheter extraction, surgical thrombectomy, or superior vena cava filter for the initial treatment of upper extremity DVT r In selected patients with acute upper-extremity DVT, e.g., those with failure of anticoagulant or thrombolytic treatment and persistent symptoms, surgical embolectomy (Grade 2C) or catheter extraction (Grade 2C) is suggested. r In selected patients with acute upper-extremity DVT, e.g., those in whom anticoagulant treatment is contraindicated, a superior vena cava filter (Grade 2C) could be considered for initial treatment. Anticoagulants for the long-term treatment of upper extremity DVT r Patients with acute upper-extremity DVT should be treated long-term with a vitamin K antagonist (Grade 1C+). As for acute DVT of the leg, a similar process should be considered for determining the duration of vitamin K antagonist treatment.
BLUK077-Stein
March 12, 2007
18:11
420
Elastic bandages for the long-term treatment of upper extremity DVT r In patients with upper-extremity DVT who have persistent edema and pain, elastic bandages for symptomatic relief is suggested (Grade 2C). Subsequent to the recommendations of the American College of Chest Physicians [11], a clinical practice guide was published by the American College of Physicians and the American Academy of Family Physicians [37]. Their recommendations were similar. They suggested the use of LMWH whenever possible. Patients with PE could possibly be treated as outpatients. Compression stockings should be used routinely for 1 year. Anticoagulants should be maintained 3 to 6 months for VTE caused by transient risk factors and >12 months for recurrent VTE. Use of LMWH is safe and efficacious for long-term treatment of VTE [37].
References 1 Carson JL, Kelley MA, Duff A et al. The clinical course of pulmonary embolism. N Engl J Med 1992; 326: 1240– 1245. 2 Douketis JD, Kearon C, Bates S, Duku EK, Ginsberg JS. Risk of fatal pulmonary embolism in patients with treated venous thromboembolism. JAMA 1998; 279: 458–462. 3 Pellegrini VD, Jr, Langhans MJ, Totterman S et al. Embolic complications of calf thrombosis following total hip arthroplasty. J Arthroplasty 1993; 8: 449–457. 4 Lagerstedt CI, Olsson C-G, Fagher BO et al. Need for long term anticoagulant treatment in symptomatic calf-vein thrombosis. Lancet 1985; 2: 515–518. 5 Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 338S–400S. 6 ACCP Consensus Committee on Pulmonary Embolism. Opinions regarding the diagnosis and management of venous thromboembolism. Chest 1996; 109: 233–237. 7 Huisman MV, Buller HR, ten Gate JW et al. Serial impedance plethysmography for suspected deep venous thrombosis in outpatients. The Amsterdam general practitioner study. N Engl J Med 1986; 314: 823–828. 8 Huisman MV, Buller HR, ten Gate JW. Utility of impedance plethysmography in the diagnosis of recurrent deep-vein thrombosis. Arch Intern Med 1988; 148: 681–683. 9 Hull RD, Raskob GE, Carter CJ. Serial impedance plethysmography in pregnant patients with clinically suspected deep-vein thrombosis. Clinical validity of negative findings. Ann Intern Med 1990; 112: 663–667.
PART IV
Prevention and Treatment of DVT and PE
10 Stein PD, Willis PW, III. Diagnosis and treatment of acute pulmonary embolism. Part II: Management. Family Pract Recert 1983; 5: 79–96. 11 Buller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126(suppl): 401S–428S. 12 Hull RD, Raskob GE, Hirsh J et al. Continuous intravenous heparin compared with intermittent subcutaneous heparin in the initial treatment of proximalvein thrombosis. N Engl J Med 1986; 315: 1109– 1114. 13 Brandjes DP, Heijboer H, Buller HR, de Rijk M, Jagt H, ten Cate JW. Acenocoumarol and heparin compared with acenocoumarol alone in the initial treatment of proximal-vein thrombosis. N Engl J Med 1992; 327: 1485– 1489. 14 Hull RD, Raskob GE, Brant RF, Pineo GF, Valentine KA. The importance of initial heparin treatment on long-term clinical outcomes of antithrombotic therapy. The emerging theme of delayed recurrence. Arch Intern Med 1997; 157: 2317–2321. 15 Gallus A, Jackaman J, Tillett J, Mills W, Wycherley A. Safety and efficacy of warfarin started early after submassive venous thrombosis or pulmonary embolism. Lancet 1986; 2: 1293–1296. 16 Hull RD, Raskob GE, Rosenbloom D et al. Heparin for 5 days as compared with 10 days in the initial treatment of proximal venous thrombosis. N Engl J Med 1990; 322: 1260–1264. 17 Dolovich LR, Ginsberg JS, Douketis JD, Holbrook AM, Cheah G. A meta-analysis comparing low-molecularweight heparins with unfractionated heparin in the treatment of venous thromboembolism: examining some unanswered questions regarding location of treatment, product type, and dosing frequency. Arch Intern Med 2000; 160: 181–188. 18 Lagerstedt CI, Olsson CG, Fagher BO, Oqvist BW, Albrechtsson U. Need for long-term anticoagulant treatment in symptomatic calf-vein thrombosis. Lancet 1985; 2: 515–518. 19 Research Committee of the British Thoracic Society. Optimum duration of anticoagulation for deep-vein thrombosis and pulmonary embolism. Lancet 1992; 340: 873–876. 20 Schulman S, Rhedin AS, Lindmarker P et al., for Duration of Anticoagulation Trial Study Group. A comparison of six weeks with six months of oral anticoagulant therapy after a first episode of venous thromboembolism. N Engl J Med 1995; 332: 1661–1665. 21 Levine MN, Hirsh J, Gent M et al. Optimal duration of oral anticoagulant therapy: a randomized trial comparing
BLUK077-Stein
March 12, 2007
CHAPTER 87
22
23
24
25
26
27
28
29
18:11
Treatment of DVT and acute PE
four weeks with three months of warfarin in patients with proximal deep vein thrombosis. Thromb Haemost 1995; 74: 606–611. Prandoni P, Lensing AW, Cogo A et al. The long-term clinical course of acute deep venous thrombosis. Ann Intern Med 1996; 125: 1–7. Arnesen H, Heilo A, Jakobsen E, Ly B, Skaga E. A prospective study of streptokinase and heparin in the treatment of deep vein thrombosis. Acta Med Scand 1978; 203: 457– 463. Schweizer J, Kirch W, Koch R et al. Short- and longterm results after thrombolytic treatment of deep venous thrombosis. J Am Coll Cardiol 2000; 36: 1336–1343. Schweizer J, Elix H, Altmann E, Hellner G, Forkmann L. Comparative results of thrombolysis treatment with rt-PA and urokinase: a pilot study. VASA 1998; 27: 167– 171. Kearon C, Gent M, Hirsh J et al. A comparison of three months of anticoagulation with extended anticoagulation for a first episode of idiopathic venous thromboembolism. N Engl J Med 1999; 340: 901–907. Agnelli G, Prandoni P, Santamaria MG et al., Warfarin Optimal Duration Italian Trial Investigators. Three months versus one year of oral anticoagulant therapy for idiopathic deep venous thrombosis. N Engl J Med 2001; 345: 165–169. Kearon C, Ginsberg JS, Kovacs MJ et al. Comparison of low-intensity warfarin therapy with conventionalintensity warfarin therapy for long-term prevention of recurrent venous thromboembolism. N Engl J Med 2003; 349: 631–639. Crowther MA, Ginsberg JS, Julian J et al. A comparison of two intensities of warfarin for the prevention of recurrent thrombosis in patients with the antiphospholipid antibody syndrome. N Engl J Med 2003; 349: 1133–1138.
421
30 Lee AY, Levine MN, Baker RI et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med 2003; 349: 146–153. 31 Hull R, Pineo GF, Mah A et al. A randomized trial evaluating long-term low-molecular-weight heparin therapy for three months versus intravenous heparin followed by warfarin sodium [Abstract]. Blood 2002; 100: 148a. 32 Levine M, Gent M, Hirsh J et al. A comparison of low-molecular-weight heparin administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis. N Engl J Med 1996; 334: 677–681. 33 Koopman MM, Prandoni P, Piovella F et al., for The Tasman Study Group. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecularweight heparin administered at home. N Engl J Med 1996; 334: 682–687. 34 The Columbus Investigators. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med 1997; 337: 657–662. 35 Schellong SM, Schwarz T, Kropp J, Prescher Y, BeuthienBaumann B, Daniel WG. Bed rest in deep vein thrombosis and the incidence of scintigraphic pulmonary embolism. Thromb Haemost 1999; 82(suppl): 127–129. 36 Aschwanden M, Labs KH, Engel H et al. Acute deep vein thrombosis: early mobilization does not increase the frequency of pulmonary embolism. Thromb Haemost 2001; 85: 42–46. 37 Snow V, Qaseem A, Barry P et al. Management of venous thromboembolism: a clinical practice guideline from the American College of Physicians and the American Academy of Family Physicians. Ann Intern Med 2007; 146: 204–210.
BLUK077-Stein
March 28, 2007
88
16:0
CHAPTER 88
Withholding treatment of patients with acute pulmonary embolism who have a high risk of bleeding provided and negative serial noninvasive leg tests
Patients with suspected acute pulmonary embolism (PE) and nondiagnostic ventilation–perfusion lung scans who have negative serial compression ultrasound or impedance plethysmography of the lower extremities have been shown not to require treatment with anticoagulants provided cardiorespiratory reserve is adequate [1] or the pretest clinical probability of PE is low or moderate [2]. Such untreated patients have a low (0.5–0.6%) rate of PE on follow-up [1, 2]. Among patients with suspected PE who had negative serial noninvasive leg tests, one can readily calculate that some, in fact, had PE that did not recur. This suggested a possible strategy of withholding treatment in patients with documented nonmassive acute PE if serial compression ultrasound of the lower extremities were negative [3]. Such a strategy would be particularly advantageous in patients who have a high risk of bleeding [3]. The concept that anticoagulants may be safely withheld in patients with proven PE providing serial noninvasive leg tests are negative and cardiorespiratory reserve is adequate is an untested approach [3]. Previous strategies that employed serial noninvasive leg tests were applicable to patients with suspected PE, not proven PE [1–3]. Management of fully diagnosed PE on the basis of serial noninvasive leg tests is a transition from prior strategies of management. The estimated frequency of PE during 3-month follow-up among untreated patients with nonmassive PE who had negative serial noninvasive leg tests was between 3 and 9%. The estimated frequency of fatal PE was 1% [3]. The calculation that led to this estimate,
422
based on studies of negative serial impedance plethysmography are as follows: Among 711 patients with suspected PE and nondiagnostic ventilation–perfusion lung scans evaluated by Hull and associates, serial impedance plethysmography showed proximal deep venous thrombosis (DVT) in 84 patients [1]. The sensitivity of a single impedance plethysmogram for DVT in patients with PE was 43 and 57% (average 50%) [4, 5]. Among patients with suspected PE, serial impedance plethysmograms were positive in an additional 2.5% [1]. Assuming the sensitivity of serial impedance plethysmograms was 53%, the number of patients with PE in Hull and associates study, therefore, would have been 158. This number of patients with PE is about the same as would be calculated assuming the positive predictive value of nondiagnostic V–Q scans is 22%, as it was in PIOPED [6]. Since 84 patients were identified with impedance plethysmography and treated, the group of 627 who were followed with no anticoagulation treatment included 74 patients with PE. On follow-up, 4 of 74 (5.4%) (95% CI = 1.5– 13.3%) showed recurrent PE; 1 of 74 (1.4%) was fatal and 3 of 74 (4.1%) were nonfatal. Calculations based on serial compression ultrasonography were as follows: Among 702 patients with suspected PE, nondiagnostic ventilation–perfusion lung scans and a low or moderate pretest clinical probability of PE evaluated by Wells and associates, serial compression ultrasound showed proximal DVT in 37 patients [2]. The average sensitivity of a single compression ultrasonography for DVT in patients with PE is 37% [2, 7], and an estimated additional 2% would
BLUK077-Stein
March 28, 2007
CHAPTER 88
16:0
Withholding treatment in patients with negative serial leg tests
become positive with serial testing [7]. The number of patients with PE in the study by Wells and associate would have been 95. Since 37 patients were identified with venous compression ultrasound and treated, the group of 665 who were followed with no anticoagulant therapy included 58 patients with PE. On follow-up, 3 of 58 (5.2%) (95% CI = 1.1–14.4%) showed recurrent PE. Calculations based on the 22% positive predictive value of nondiagnostic V–Q scans in PIOPED suggest that a larger number of patients, 154 patients, would have had PE. Of these 117 would have been followed off treatment. The rate of PE in 3 months would have been 3 of 117 (2.6%). The frequency of fatal and nonfatal PE among untreated patients with negative serial noninvasive leg tests is comparable to rates of recurrent PE in patients treated with anticoagulants and in patients treated with inferior vena cava filters. The frequency of fatal recurrent PE among patients with PE treated with anticoagulants, based on pooled data from prospective trials, was 18 of 1325 (1%) [1, 8–10]. The frequency of nonfatal recurrent PE was 16 of 1325 (1%). The frequency of recurrent PE among patients with newer designs of inferior vena cava filters was 1–4% [11–14]. A randomized trial of various inferior vena cava filters (Vena Tech LGM, titanium Greenfield, Cardial and Bird’s Nest) in the prevention of PE showed a frequency of nonfatal PE of 2 of 182 (1%) [15]. There was no fatal recurrent PE. A variety of complications occur with inferior vena cava filters [16]. Symptomatic inferior vena cava occlusion, in a randomized trial, occurred in 16 of 182 (9%) [15]. Inferior vena cava thrombosis following filter placement has been reported in 2–30% [11, 12, 17]. The frequency of major bleeding from unfractionated heparin in patients with PE who had a high risk of bleeding was 12 of 111 (11%) [18]. Major bleeding was defined as overt bleeding associated with a reduction of hemoglobin ≥2 g/dL, blood transfusion ≥2 units, intracranial bleed, retroperitoneal bleed, bleeding into a major prosthetic joint [19, 20]. Because of the complications of therapy, treatment with either anticoagulants or an inferior vena cava filter may result in high rates of adverse events in patients with PE who are at a high risk of bleeding. The patients in whom calculations supportive of this strategy were based, on average, had PE that was not massive. Patients with PE in PIOPED who had nondiagnostic ventilation–perfusion lung scans, on average, had less extensive PE than those with high probability
423
interpretations [21]. Patients with massive PE and patients with PE who have poor cardiorespiratory reserve should not be considered for this strategy. The strategy of withholding treatment of acute PE if serial ultrasonography is negative is based on the observation that PE arises from DVT of the lower extremities in most patients. The safety of withholding treatment in patients with suspected PE, providing serial leg tests are negative, supports this observation [1, 2]. Autopsy showed DVT of the lower extremities in 83–91% of patients with PE [20, 21]. Occasionally, however, other sites may be the source of the thromboemboli. The heart has been reported as a source of PE in 8–17% [22, 23]. Thrombosis confined to the iliac vein, without involvement of the veins of the lower extremity, is uncommon, 1 of 149 patients with DVT shown by dissection at autopsy [24]. Thrombosis confined to the inferior vena cava is equally uncommon (1 of 149) [24]. Deep venous thrombosis of the upper extremities due to the insertion of central venous catheters may be a source of PE [25, 26]. Withholding treatment of acute PE, providing serial noninvasive leg tests are negative and cardiopulmonary reserve is adequate, is one possible strategy of management of patients with a high risk of bleeding or other contraindication to anticoagulants. The data on which this strategy is based are a synthesis derived from results of different investigations. This strategy has not been tested. The choice of this or other strategies depends on the potential risks of treatment and are a matter of clinical judgment.
References 1 Hull RD, Raskob GE, Ginsberg JS et al. Noninvasive strategy for the treatment of patients with suspected pulmonary embolism. Arch Intern Med 1994; 154: 289–297. 2 Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. 3 Stein PD, Hull RD, Raskob GE. Withholding treatment in patients with acute pulmonary embolism who have a high risk of bleeding and negative serial noninvasive leg tests. Am J Med 2000; 109: 301–306. 4 Hull RD, Hirsh J, Carter CJ et al. Diagnostic value of ventilation–perfusion lung scanning in patients with suspected pulmonary embolism. Chest 1985; 88: 819–828. 5 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically
BLUK077-Stein
March 28, 2007
16:0
424
6
7
8
9
10
11
12 13
14
15
suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. Turkstra F, Kuijer PM, van Beek EJ, Brandjes DP, ten Cate JW, Buller HR. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781. Simonneau G, Sors H, Charbonnier B et al., for The THESEE Study Group. A comparison of low-molecular-weight heparin with unfractionated heparin for acute pulmonary embolism. Tinzaparine ou Heparine Standard: Evaluations dans l’Embolie Pulmonaire. N Engl J Med 1997; 337: 663–669. Stein PD, Henry JW, Relyea B. Untreated patients with pulmonary embolism: outcome, clinical and laboratory assessment. Chest 1995; 107: 931–935. The Columbus Investigators. Low-molecular weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med 1997; 337: 657–662. Mohan CR, Hoballah JJ, Sharp WJ, Kresowik TF, ChienTai L, Corson JD. Comparative efficacy and complications of vena caval filters. J Vasc Surg 1995; 21: 235– 246. Ray CE, Jr, Kaufman JA. Complications of inferior vena cava filters. Abdom Imaging 1996; 21: 368–374. Sullivan TM, Martinez BD, Lemmon G, Schwartz RA, Bondy B. Clinical experience with the Greenfield Filter in 193 patients and description of a new technique for operative insertion. J Am Coll Surg 1994; 178: 117– 122. Nicholson AA, Ettles DF, Paddon AJ, Dyet JF. Long-term follow-up of the Bird’s Nest IVC filter. Clin Radiol 1999; 54: 759–764. Decousus H, Leizorovicz A, Parent F et al. A clinical trial of vena caval filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. Prevention du Risque d’Embolie Pulmonaire par Inter-
PART IV
16
17
18
19
20
21
22
23 24
25
26
Prevention and Treatment of DVT and PE
ruption Cave Study Group. N Engl J Med 1998; 338: 409– 415. Bergqvist D. The role of vena caval interruption in patients with venous thromboembolism. Prog Cardiovasc Dis 1994; 37: 25–37. Crochet DP, Brunel P, Trogrlic S, Grossetete R, Auget JL, Dary C. Long-term follow-up of Vena Tech-LGM filter: predictors and frequency of caval occlusion. J Vasc Interv Radiol 1999; 10: 137–142. Hull RD, Raskob GE, Rosenbloom D et al. Heparin for 5 days as compared with 10 days in the initial treatment of proximal venous thrombosis. N Engl J Med 1990; 322: 1260–1264. Hull RD, Delmore T, Carter C et al. Adjusted subcutaneous heparin versus warfarin sodium in the long term treatment of venous thrombosis. N Engl J Med 1982; 306: 189–194. Hull RD, Hirsh J, Jay R et al. Different intensities of oral anticoagulant therapy in the treatment of proximal vein thrombosis. N Engl J Med 1982; 307: 1676–1681. Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/perfusion lung scans. Chest 1997; 111: 1246– 1248. Byrne JJ, O’Neil EE. Fatal pulmonary emboli. A study of 130 autopsy-proven fatal emboli. Am J Surg 1952; 83: 47–49. Short DS. A survey of pulmonary embolism in a general hospital. BMJ 1952; 1: 790–796. Gibbs NM. Venous thrombosis of the lower limbs with particular reference to bed-rest. Br J Surg 1957; 45: 209– 236. Monreal M, Raventos A, Lerma R et al. Pulmonary embolism in patients with upper extremity DVT associated to venous central lines—a prospective study. Thromb Haemost 1994; 72: 548–550. Dollery CM, Sullivan ID, Bauraind O, Bull C, Milla PJ. Thrombosis and embolism in long-term central venous access for parenteral nutrition. Lancet 1994; 344: 1043– 1045.
BLUK077-Stein
March 12, 2007
89
18:16
CHAPTER 89
Thrombolytic therapy in acute pulmonary embolism
History Streptokinase, the first thrombolytic agent to be discovered, was reported by Tillett in 1933 [2]. Experimental studies demonstrated that it could quickly induce a fibrinolytic state capable of dissolving experimentally induced thrombi. Early clinical investigations uncovered two problems: pyrogenicity and inactivation of streptokinase by preexisting antibodies. The first clinical report of its use in patients with pulmonary embolism (PE) was that of Browse and James in 1964 [3]. They reported the results of the infusion of streptokinase in four patients with a clinical diagnosis of acute PE. All four patients, including two with hypotension, recovered and were discharged. Urokinase derived from human urine was found to have thrombolytic capability. The first clinical report of its use in patients with a variety of thromboembolic disorders, including six patients with venous thromboembolism, was by Hansen et al. in 1961 [4].
Infusion directly into the pulmonary artery
Standard doses infused into the pulmonary artery We are aware of only one investigation in which an infusion of thrombolytic agent directly into the pulmonary artery was compared with an intravenous infusion of the same thrombolytic agent [8]. An infusion of rt-PA 50 mg over 2 hours directly into the pulmonary artery (19 patients) was not more effective than an intravenous injection of rt-PA 50 mg (15 patients) [8]. The respective changes in the angiographic severity score were −9 and −10. Some of the patients received a repeat dose of 50 mg rt-PA over 5 hours.
Table 89.1 Thrombolytic regimens. FDA approved*
Administration of thrombolytic agents FDA-approved regimens and unapproved regimens Regimens for the intravenous infusion of thrombolytic agents are rt-PA 100 mg administered over 2 hours, streptokinase administered over 24 hours, and urokinase administered over 12–24 hours [5] (Table 89.1). Shorter infusions of streptokinase and of urokinase as well as other thrombolytic agents have been used [6, 7]. These regimens, not approved by the FDA, are also shown in Table 89.1.
t-PA
100 mg/2 hr
Streptokinase
250,000 U/30 min 100,000 U/hr × 24 hr
Urokinase
4,400 U/kg/10 min 4,400 U/kg/hr × 12–24 hr
Alteplase
100 mg/2 hr
Dose or drugs not FDA approved* Urokinase
3,000,000 U/2 hr
Streptokinase
1,500,000 U/1–2 hr
Reteplase
10 U, repeat 10 U in 30 min
Saruplase
80 mg/30 min
Staphylokinase
20 mg/30 min
Tenecteplase
Single bolus in 5 sec† 30–50 mg depending on weight†
Anistreplase
5 or 10 mg × 3 doses
* MICROMEDEX(R) Healthcare Series [6].
This chapter is updated, but largely reproduced, with permission, from Stein and Dalen [1].
† Dose based on acute myocardial infarction. Reproduced with permission from Stein and Dalen [1].
425
BLUK077-Stein
March 12, 2007
18:16
426
PART IV
Prevention and Treatment of DVT and PE
Table 89.2 Rate of resolution: lytic agents versus heparin.
Lytic
Number of patients Heparin
Resolution
Time after
agent
Lytic
start lytic (hr)
Lytic
UK
64
62
24
SK (IP)
11
12
72
−13.3
−2.8†
Tibbutt et al. [9]
SK
14
10
72
−11.3
−3.4¶
Ly [11]
rt-PA
9
4
24
10%
0%‡
rt-PA
33
25
24
34.4%
12.0%§
Levine et al. [13]
rt-PA
20
16
2
−3.5
−0.1¶
Dalla-Volta et al. [14]
rt-PA
46
55
24
24.1%
14.6%
Heparin 8.3%*
1.5%#
References UPET [10]
PIOPED [12]
Goldhaber et al. [15]
* Mean percent resolution in scan deficit (significant). †
Change in pulmonary angiogram (P < 0.01 to P < 0.001). Percent improvement in mismatched scan deficit (NS). § Percent of patients showing > 50% improvement of perfusion scan (P = 0.026). ¶ Change in pulmonary angiographic severity score (P < 0.01). #Proportion of lung showing improved perfusion (P < 0.0001). UK, urokinase; SK, streptokinase; IP, intrapulmonary artery; HEP, heparin. Reproduced with permission from Stein and Dalen [1]. ‡
One randomized trial compared an infusion of streptokinase into the main pulmonary artery with a heparin infusion into the main pulmonary artery [9]. Streptokinase was infused as a loading dose of 600,000 U over 30 minutes followed by 100,000 U/hr for 72 hours. The mean angiographic score decreased by 13.3 in the streptokinase arm and it decreased by 2.8 in the heparin arm (P < 0.001) (Table 89.2).
Low-doses infused into the pulmonary artery Several case series evaluated an infusion of low doses of a lytic agent directly into the pulmonary artery, but there was no comparison with other routes of administration or other drugs [16–20]. In a case series by Leeper and associates, in which streptokinase in low doses was infused into the pulmonary artery in combination with heparin, the perfused lung showed a reduction of the angiographic severity index score [16]. Even though the dose of streptokinase was only 5–9% of the FDA-approved dose for systemic use, severe bleeding occurred in 2 of 7 patients (28.6%). In a case series by Gallus and associates, low doses of streptokinase infused into the pulmonary artery showed considerable lysis in all 7 who had posttreatment angiograms [17]. Severe bleeding occurred in 2 of 13 patients (15.4%) [17]. Vujic and associates, in 3 patients, used even lower doses of streptokinase (5500–10,000 U) in combination with heparin infused over 16–30 hours [18]. There
was significant angiographic improvement and no major bleeding. Others also showed improvement in most and no bleeding following an intrapulmonary artery infusion of low doses of streptokinase [19] or urokinase [20].
Rate and extent of resolution of PE Resolution with urokinase Thrombolytic therapy causes a more rapid rate of resolution of PE that occurs with natural processes or with heparin. This was demonstrated in the Urokinase Pulmonary Embolism Trial (UPET) [10]. Patients with PE documented by pulmonary angiography were randomized to a 12-hour infusion of urokinase or to intravenous heparin. Eighty-two patients received urokinase (2000 U/lb loading dose over 10–15 minutes followed by 2000 U/lb/hr for 12 hours) and 78 patients received heparin alone. Mean percentage resolution on perfusion lung scans 24 hours after starting urokinase was 24.1%, compared with 8.3% among patients who received heparin alone. By day 14, mean percentage resolution was comparable in both groups (55.4% versus 56.2%). At the end of 1 year, mean percentage resolution remained comparable in both groups (78.8% versus 77.2%). Among patients with no prior cardiopulmonary disease, 88% of patients treated with urokinase showed more than 90% resolution of the perfusion lung scan, versus 91% of those treated only
BLUK077-Stein
March 12, 2007
CHAPTER 89
18:16
427
Thrombolytic therapy in acute PE
with anticoagulants. Among patients with prior cardiopulmonary disease, the percentage of patients who showed more than 90% resolution of the perfusion lung scan was 77% with urokinase and 72% with anticoagulants.
Resolution with streptokinase In the Urokinase–Streptokinase Embolism Trial, a second phase of the UPET, 167 patients with angiographically documented PE were randomized to treatment with 12 hours of urokinase as in the Phase 1 trial, or 24 hours of urokinase or 24 hours of streptokinase [21]. Angiographic follow-up showed no significant differences among the three regimens. Perfusion lung scans, however, showed that 24 hour urokinase resulted in significantly more resolution than 24 hours of streptokinase. All 3 regimens were more effective in accelerating resolution of PE than heparin alone was in the Phase 1 trial [21]. Comparisons of rate of resolution with heparin We are aware of 7 randomized trials in which the rate of resolution of PE with thrombolytic agents was com-
pared with heparin. Three of these were with streptokinase or urokinase [9–11] and 4 were with rt-PA [12–15] (Table 89.2). Different methods were used for comparing resolution of the angiogram or perfusion lung scan. Thrombolytic agents in all investigations in which there were 10 or more patients in each arm showed a more rapid rate of resolution of PE than with heparin alone.
Comparisons of resolution with rt-PA, urokinase, and streptokinase We found 5 investigations comparing the rate of resolution of PE following treatment with rt-PA with the rate of resolution following urokinase or streptokinase [22–26] (Table 89.3). The rt-PA was given in a dose of 100 mg over 2 hours. At 2 hours, resolution was greater with rt-PA than with streptokinase or urokinase administered over 12–24 hours [22, 24, 25]. However, resolution at the time of completion of the infusion of urokinase or streptokinase (12–24 hours) was comparable with rt-PA [22, 24, 25]. When streptokinase or urokinase were administered over 2 hours, resolution at 2 hours was comparable with rt-PA [23, 26].
Table 89.3 Resolution rate: rt-PA 100 mg/2 hr versus other thrombolytic agents.
Dose of other
Number of patients
thrombolytic agent
rt-PA
Other
Urokinase
22
23
2,000 U/lb bolus
Resolution rt-PA
Angiographic improvement at 2 hr 82%
2,000 U/lb/hr for 24 hr
Other
References Goldhaber et al. [22]
48% (P = 0.008)
Lung scan perfusion at 24 hr Both groups equal (NS)
Urokinase
34
29
TPR change at 2 hr −42%
4,400 U/kg bolus 4,400 U/kg/hr for 12 hr
Angiographic severity score at 12 hr −5.9
Streptokinase
25
25
−7.5 (NS)
TPR change at 2 hr
250,000 U/15 min
−42%
1,200,000 U/12 hr
TPR change at 12 hr −48%
Urokinase
42
45
−40% (NS)
Patients showing angiographic Improvement at 2 hr
2,000,000 U/110 min
79%
Streptokinase
TPR change at 2 hr 23
43
−38%
rt-PA dose was 100 mg/2 hr in all investigations. TPR, total pulmonary resistance; UK, urokinase; SK, streptokinase. Reproduced with permission from Stein and Dalen [1].
Meneveau et al. [25]
−13% (P < 0.001)
1,000,000 U/10 min
1,500,000 U/2 hr
Meyer et al. [24]
−21% (P < 0.0001)
Goldhaber et al. [23]
67% (NS) −31%(NS)
Menevau et al. [26]
BLUK077-Stein
March 12, 2007
18:16
428
PART IV
Prevention and Treatment of DVT and PE
Table 89.4 Mortality of patients not in shock: thrombolytic agents versus heparin. Number of patients Heparin
Mortality
Lytic agent
Lytic
Lytic [n (%)]
Heparin [n (%)]
References
Urokinase
73
73
2 (2.7)
6 (8.2)
UPET [10]
Urokinase
20
10
0 (0)
0 (0)
Marini et al. [29]
rt-PA
9
4
1 (11.1)
0 (0)
PIOPED [12]
rt-PA
33
25
1 (3.0)
0 (0)
Levine et al. [13]
rt-PA
20
16
2 (10.0)
1 (6.3)
Dalla-Volta et al. [14]
rt-PA
46
55
0 (0)
2 (3.6)
Goldhaber et al. [15]
rt-PA
118
138
4 (3.4)
3 (2.2)
Konstantinides et al. [30]
Total
319
321
10 (3.1)
12 (3.7)*
*All differences, lytic vs. heparin, not significant. Reproduced with permission from Stein and Dalen [1].
Resolution of functional abnormalities Although residual perfusion defects in the UPET were comparable after 1 year among small numbers of patients treated with thrombolytic agents and patients treated with anticoagulants alone, the pulmonary diffusing capacity (DLCO) after 1 year was higher among patients who received thrombolytic therapy [27]. Among 21 patients treated with anticoagulants compared with 19 patients treated with thrombolytic agents, the DLCO was 72% predicted versus 93% predicted, respectively [27]. Follow-up a mean of 7.4 years later of 12 patients treated with urokinase or streptokinase and 11 patients treated with heparin showed that mean pulmonary artery pressure was higher in the heparin group (22 mm Hg versus 17 mm Hg) as was pulmonary vascular resistance (351 dyne sec/cm5 versus 171 dyne sec/cm5 ) [28]. Exercise increased the pulmonary artery mean pressure and pulmonary vascular resistance in the heparin group, but not in the thrombolytic therapy group. Among patients treated with urokinase or streptokinase, 8 of 12 (67%) were New York Heart Association (NYHA) Class 0 to I, compared with 3 of 11 (27%) in the heparin therapy arm.
Mortality with thrombolytic therapy Mortality of patients not in shock Among symptomatic patients with acute PE who were not in shock, the mortality with heparin was comparable to the mortality with thrombolytic therapy, irre-
spective of whether treatment was with urokinase administered over 12 hours, urokinase 12 hours/day for 3 days, or rt-PA [10, 12–15, 29, 30] (Table 89.4). The largest trial, the UPET showed a 2-week mortality of 2 of 73 (2.7%) treated with urokinase and 6 of 73 (8.2%) treated with heparin (NS) [10]. Regarding more recent trials, pooled data from 5 randomized trials that compared rt-PA with heparin showed a mortality of 8 of 226 (3.5%) among patients treated with rt-PA and 6 of 238 (2.5%) among patients treated with heparin (NS) [12–15, 30] (Table 89.4). Among all trials, mortality with thrombolytic therapy was 10 of 319 (3.1%) versus 12 of 321 (3.7%) in patients treated with heparin alone (NS). In a registry of patients with nonmassive PE, defined as systolic blood pressure (≥90 mmHg), all cause mortality within 90 days was 56 of 266 (21%) in those treated with thrombolytic therapy and 283 of 2018 (14%) in those not treated with thrombolytic agents [31]. Meta-analysis reached the same conclusion: compared with intravenous heparin thrombolytic therapy does not appear to reduce the mortality from acute PE in unselected patients [32]. The results were homogeneous and unaffected by the formulation of the thrombolytic agent, the clinical severity of PE, the extent of vascular obstruction determined radiologically or the methodological quality of the included trials [32]. The authors cautioned, however, that these negative results should be interpreted with caution due to a potential lack of statistical power [32]. A more recent meta-analysis also found no reduction in the mortality rate of unselected patients with acute PE [33].
BLUK077-Stein
March 12, 2007
CHAPTER 89
18:16
429
Thrombolytic therapy in acute PE
Table 89.5 Hypotensive patients (systolic arterial pressure <90 mm Hg) treated with thrombolytic agents or no thrombolytic agents. Event within 90 days
Thrombolytic therapy [n/N (%)]
No thrombolytic therapy [n/N (%)]
Mortality from PE
15/33 (46)
40/73 (55)
Recurrent PE
4/33 (12)
9/73 (12)
Bleeding
8/33 (24)
11/73 (15)
PE, pulmonary embolism. Data from Kucher et al. [31].
Mortality in hypotensive patients and patients in shock Among patients in shock in the UPET, 2-week mortality among those treated with urokinase and those treated with anticoagulants was 4 of 9 (44%) versus 1 of 5 (20%), respectively (NS) [10]. In a trial of patients with massive PE, all of whom were in shock, 4 of 4 randomized to streptokinase 1,500,000 U in 1 hour improved within the first hour after treatment and all survived [34]. Among 4 patients randomized to heparin alone, none survived (P = 0.02). The likelihood of obtaining such comparative data in the future is remote. Meta-analysis of 5 trials that included patients with unstable PE (but all included patients were not unstable) showed a trend toward lower death rates in patients treated with thrombolytic agents, 8 of 128 (6.3%) than in patients treated with heparin, 16 of 126 (12.7%) (NS) [33]. Counterintuitively, among hypotensive patients with PE (systolic arterial pressure <90 mm Hg), 90-day mortality was similar in a registry of patients treated
with thrombolytic therapy and those who did not receive thrombolytic therapy, 46% versus 55% [32] (Table 89.5). The rate of recurrent PE was the same in both groups (12%).
Mortality in patients with right ventricular dilatation or dysfunction Among stable patients with baseline right ventricular hypokinesis, the mortality was 0 of 18 (0%) among patients randomized to rt-PA compared with 2 of 18 (11.1%) randomized to heparin alone (NS) [15] (Table 89.6). In a retrospective investigation of 64 patients with right ventricular dysfunction who received thrombolytic therapy, compared with 64 who received heparin alone, the mortality was also comparable in both groups [35] (Table 89.6). A multicenter registry showed a higher 30-day mortality rate among patients not in shock with right ventricular enlargement, afterload stress or pulmonary hypertension who were treated with anticoagulants (11.1%) than among such patients treated with thrombolytic agents (4.7%) [36] (Table 89.6). More recently, among stable patients with
Table 89.6 Recurrent PE and All-cause mortality in patients with right ventricular dysfunction or dilatation and not in shock: thrombolytic agents vs. heparin. Recurrent PE Number of patients Lytic agent
Lytic
Heparin
All-cause Mortality
Lytic
Heparin
Lytic
Heparin
[n (%)]
[n (%)]
[n (%)]
[n (%)]
References
rt-PA
18
18
0 (0)
5 (27.8)*
0 (0)
2 (11.1)
Goldhaber et al. [15]
rt-PA,
64
64
3 (4.7)
3 (4.7)
4 (6.3)
0 (0)
Hamel et al. [35]
169
550
—
—
8 (4.7)
61 (11.1)
Konstantinides et al. [36*]
urokinase saruloplase various
*P < 0.02. PE, pulmonary embolism. Reproduced with permission from Stein and Dalen [1].
BLUK077-Stein
March 12, 2007
18:16
430
PART IV
right ventricular enlargement enrolled in PIOPED II who were treated with anticoagulants and/or an inferior vena cava filter, the in-hospital mortality from PE was 0 of 76 (0%) and the in-hospital all cause mortality was 2 of 76 (2.6%) (Stein PD, Beemath A, Matta F, et al., unpublished data from PIOPED II). These patients were not hypotensive, critically ill, or on ventilatory support and did not have a myocardial infarction within the previous month or ventricular tachycardia or ventricular fibrillation within the previous 24 hours. Konstantinides and associates reported the clinical course of stable patients with submassive acute PE who were treated either with rt-PA plus heparin or heparin alone [30]. Right ventricular dysfunction was present in 31% of patients in both treatment arms. Mortality was comparable in both groups (Table 89.4). “Rescue thrombolysis” was required in 23% of the patients in the heparin arm [30].
Prevention and Treatment of DVT and PE
current PE confirmed by lung scan, pulmonary angiogram, or contrast-enhanced spiral CT, based on pooled data, was shown in 5 of 217 (2.3%) patients treated with rt-PA and in 6 of 234 (2.6%) patients treated with heparin alone (NS) [13–15, 30].
Recurrent PE with urokinase In the UPET, recurrent PE by 14 days was comparable in both arms of the study [10]. Recurrent PE documented by both clinical evidence and perfusion lung scan was observed in 5 of 82 (6.1%) in the urokinase arm and in 7 of 78 (9.0%) in the heparin arm. After a mean of 7.4 years, recurrent PE, occurred in 2 of 19 (10.5%) in the thrombolytic therapy arm and 4 of 21 (19.0%) in the heparin arm (P < 0.05) [28].
Recurrent PE with right ventricular dysfunction or dilatation The suggestion that thrombolytic therapy may be indicated in patients with right ventricular dysfunction was proposed by Goldhaber et al. [15]. Among stable patients with baseline right ventricular hypokinesis, 18 patients had been randomized to rt-PA followed by heparin and 18 patients had been randomized to heparin alone. In this subgroup of patients with right ventricular dysfunction, there was no recurrent PE among patients treated with rt-PA, but 5 of 18 (27.8%) treated with heparin alone suffered clinical or objective evidence of recurrent PE (P < 0.02) (Table 89.6). In 2 of those with suspected recurrent PE, the diagnosis of recurrent PE was confirmed by a perfusion lung scan. In a retrospective investigation of 64 patients with right ventricular dysfunction who received thrombolytic therapy, compared with 64 who received heparin alone, the rate of recurrent PE was the same in both groups, 3 of 64 (4.7%) [35] (Table 89.6). Meta-analysis of trials that included patients with unstable PE (but all patients were not unstable) showed recurrence with thrombolytic agents in 5 of 128 (3.9%) versus 9 of 126 (7.1%) in patients treated with heparin (NS) [33].
Recurrent PE with rt-PA Recurrent PE in randomized trials in which rt-PA was compared with heparin is shown in Table 89.7. Re-
Meta-analysis: unselected trials Compared with intravenous heparin thrombolytic therapy does not appear to reduce the recurrence from
Recurrent PE with thrombolytic therapy
Table 89.7 Recurrent pulmonary embolism rt-PA vs. heparin: randomized trials. Number of patients rt-PA
Heparin
Recurrent PE objective test rt-PA [n (%)]
Heparin [n (%)]
References
33
25
0 (0)
0 (0)
Levine et al. [13]
20
16
1 (5)
0 (0)
Dalla-Volta et al. [14] Goldhaber et al. [15]
46
55
0 (0)
2 (3.6)
118
138
4 (3.4)
4 (2.9)
Konstatinides et al. [30]
217
234
5 (2.3)
6 (2.6)*
Total
*Not significant. PE, pulmonary embolism. Reproduced with permission from Stein and Dalen [1].
BLUK077-Stein
March 12, 2007
CHAPTER 89
18:16
431
Thrombolytic therapy in acute PE
Definition of major bleeding We define major bleeding as bleeding associated with a reduction of hemoglobin ≥2 g/dL, blood transfusion ≥2 units, intracerebral bleed, retroperitoneal bleed, pericardial bleed, bleeding that required a surgical intervention, bleeding into a major joint, or bleeding into the eye [37]. We also define a reduction of hematocrit >10 points as major bleeding.
acute PE in unselected patients [32]. As with mortality, the results were unaffected by the formulation of the thrombolytic agent, the clinical severity of PE, the extent of vascular obstruction determined radiologically, or the methodological quality of the included trials [32]. However, these negative results should be interpreted with caution due to a potential lack of statistical power [32]. A more recent meta-analysis also found no reduction in the rate of recurrent PE in unselected patients [33].
Major bleeding among patients in registries Among 169 stable patients with PE (19% diagnosed by pulmonary angiography) who received thrombolytic therapy in a multicenter registry, major bleeding occurred in 21.9% [36] (Table 89.8). Intracranial hemorrhage occurred in 1.2%. Death from complications of diagnostic procedures or from therapy occurred in 0.6%. In another multicenter registry, the International Cooperative Pulmonary Embolism Registry, major bleeding was reported in 21.7% patients who received thrombolytic therapy and intracranial hemorrhage occurred in 3.0% [38]. In a review of 7 investigations with t-PA or urokinase, in which high-risk patients were excluded, major bleeding occurred in 61 of 312 patients (19.6%) [40]. Bleeding at the site of catheterization occurred in 34 (10.9%). Intracranial hemorrhage occurred in 5 patients (1.6%).
Major bleeding with thrombolytic therapy Overview The overwhelming necessity for a cautious and prudent use of thrombolytic therapy in patients with PE is the risk of major bleeding. Narrow indications of the recommendations for thrombolytic therapy in patients with PE are based on high rates of bleeding. Over three decades of experience have shown that high rates of bleeding occur with thrombolytic therapy for PE in spite of modifications of the type of thrombolytic agent, the dose, the rate of administration, and care in avoiding arterial punctures and venipunctures in patients.
Table 89.8 Major bleeding with various thrombolytic agents in treatment of PE.* Major catheter Patients
Number
Angiography
Major bleed
site bleed
[n (%)]
[n (%)]
[n (%)]
Lytic agent
References
37 (21.9)
?
Any
Konstantinides et al. (36) Goldhaber et al. [38]
Hemodynamically 169
32 (19)
stable registry Major PE registry RV dilatation,
304 64
58 (19) ?
66 (21.7)
?
Any
10 (15.6)
?
t-PA, Urokinase,
dysfunction
Saruplase
registry Randomized
23
23 (100)
11 (47.8)
?
Urokinase
Randomized
46
46 (100)
6 (13.0)
?
Urokinase
Goldhaber et al. [23]
Randomized
20
20 (100)
1 (5.0)
0
Urokinase
Marini et al. [29]
Randomized
129
129 (100)
16 (12.4)
2 (1.6)
Urokinase
UKEP [39]
Randomized
29
29 (100)
8 (27.6)
2 (6.9)
Urokinase
Meyer et al. [24]
Randomized
43
43 (100)
3 (7.0)
2 (4.7)
Streptokinase
Menevau et al. [26]
Randomized
4
0 (0)
0 (0)
0
Streptokinase
Jerjes-Sanchez et al. [34]
Randomized
25
25 (100)
3 (12.0)
2 (8.0)
Streptokinase
Menevau et al. [25]
*Only investigations since the 1980s are included. PE, pulmonary embolism; RV, right ventricular. Reproduced with permission from Stein and Dalen [1].
Goldhaber et al. [22]
BLUK077-Stein
March 12, 2007
18:16
432
PART IV
Prevention and Treatment of DVT and PE
Table 89.9 Major bleeding with rt-PA in randomized trials of treatment of PE. Angiography
Major bleed
Major catheter site bleed
Number
[n (%)]
[n (%)]
[n (%)]
With heparin
References
34*
34 (100)
4 (11.8)
0 (0)
Yes
Verstraete et al. [8]
9
9 (100)
1 (11.1)
1 (11.1)†
Yes
PIOPED [12]
20
20 (100)
3 (15.0)
0 (0)
No
Dalla-Volta et al. [14]
22
22 (100)
4 (18.2)
?
No
Goldhaber et al. [22]
44
44 (100)
9 (20.5)
?
No
Goldhaber et al. [23]
46
6 (13)
4 (8.7)
1 (2.2)
No
Goldhaber et al. [15]
33
22 (67)
0 (0)
0 (0)
No
Levine [13]
34
34 (100)
7 (20.6)
2 (5.9)
No
Meyer et al. [24]
87
60 (69)
14 (16.1)
5 (5.7)
No
Goldhaber et al. [42]
23
23 (100)
5 (21.7)
4 (17.4)
No
Meneveau et al. [26]
25
25 (100)
4 (16.0)
3 (12.0)
Yes
Meneveau et al. [25]
53
53 (100)
8 (15.1)
5 (9.4)
No
Sors et al. [41]
In all trials, only patients at low risk of bleeding were included. Investigations were excluded if we could not determine the rate of major bleeding according to our definition. *Intrapulmonary and intravenous. † Many sites including site of catheter insertion. PE, pulmonary embolism. ?, unknown. Reproduced with permission from Stein and Dalen [1].
Major bleeding with rt-PA Among patients in randomized trials who received rt-PA, 63 of 430 (14.7%) suffered major bleeding [8, 12–15, 22–26, 41, 42] (Table 89.9). Among those who received rt-PA not accompanied by simultaneous heparin, pooled data showed major bleeding in 54 of 362 (14.9%) [13–15, 22–24, 26, 41, 42]. In those who received simultaneous heparin, the rate of major bleeding was not higher, 9 of 68 (13.2%) [8, 12, 25]. The criteria for major bleeding were the criteria described above. Patients with a high risk of bleeding were excluded. Among those who had pulmonary angiograms, major bleeding at the insertion site occurred in 21 of 286 (7.3%) [8, 12–15, 24–26, 41, 42] (Table 89.9). In a review of patients in randomized trials that used rt-PA and in whom the diagnosis of PE was made by pulmonary angiography, Levine found major bleeding in only 19 of 227 (8.4%) [43]. The definition of major bleeding that he used did not include a 10-point drop in hematocrit or a transfusion of 2 units of blood. Konstantinides and associates reported major bleeding with rt-PA in only 1 of 118 (0.8%) [30]. They defined major bleeding as fatal bleeding, hemorrhagic stroke, or a drop in the hemoglobin concentration of at least 4 g/dL.
Among patients who received urokinase in randomized trials in the 1980s or more recently, the rate of major bleeding based on pooled data was 42 of 247 (17.0%) [22–24, 29, 39] (Table 89.8). The criteria for major bleeding were the criteria described above. Patients with a high risk of bleeding were excluded. All patients had pulmonary angiograms. Pooled data from randomized trials showed comparable rates of major bleeding among patients treated with rt-PA (14.7%) [8, 12–15, 22–26, 41, 42] (Table 89.9) and patients treated with urokinase (17.0%) [22– 24, 29, 39] (Table 89.8). Only sparse data are available in recent trials of patients treated with streptokinase [25, 26, 34] (Table 89.8).
Major bleeding: meta-analyses Meta-analysis of randomized controlled trials comparing thrombolytic agents with intravenous heparin in patients with acute PE showed a higher rate of major hemorrhage with thrombolytic therapy, 33 of 241 (13.7%) compared with 17 of 220 (7.7%) with heparin [32]. Major bleeding was defined as intracranial or retroperitoneal hemorrhage or other bleeding requiring blood transfusion or surgery [32]. Results were largely unaffected by the formulation of the thrombolytic agent [32].
BLUK077-Stein
March 12, 2007
CHAPTER 89
18:16
433
Thrombolytic therapy in acute PE
Table 89.10 Intracranial hemorrhage with thrombolytic agents in treatment of PE. Patients
Number
Major PE hemodynamically
169
ICH [n (%)] 2 (1.2)
Fatal ICH [n (%)]
Lytic agent
References
1 (0.6)
Any
Konstantinides et al. [36]
stable registry Major PE registry
304
9 (3.0)
—
Any
Goldhaber et al. [38]
PE trials pooled
312
6 (1.9)
2 (0.6)
rt-PA, urokinase
Kanter et al. [46]
PE trials pooled
559
12 (2.1)
9 (1.6)
rt-PA
Dalen et al. [45]
ICH, intracranial hemorrhage; PE, pulmonary embolism. Reproduced with permission from Stein and Dalen [1].
Another meta-analysis showed a nonstatistically significant increase in major bleeding in patients treated with thrombolytic therapy, 34 of 374 (9.1%) compared with 23 of 374 (6.1%) in patients treated with heparin [33]. Major bleeding was not defined. The rate of nonmajor bleeding (undefined) was higher in patients treated with thrombolytic therapy, 53 of 233 (22.7%) versus 22 of 221 (10.0%) [33].
Major bleeding at the site of catheter insertion Severe bleeding at the site of catheter insertion had been a problem when virtually all patients with PE of severity sufficient to require thrombolytic therapy were diagnosed on the basis of a pulmonary angiogram. Now that most patients are diagnosed by CT pulmonary angiography [44], this source of bleeding has been eliminated. It remains a problem, however, in those who require catheter-tip pulmonary thromboembolectomy in combination with thrombolytic therapy (see Chapter 92). Among all patients in randomized trials since the 1980s, irrespective of whether treatment was with rt-PA, urokinase, or streptokinase, major bleeding at the site of catheter insertion was reported in 29 of 532 (5.4%) [8, 12–15, 24–26, 29, 34, 39, 41, 42] (Tables 89.8 and 89.9). Major bleeding at the site of catheter insertion among patients treated with rt-PA was 21 of 286 (7.3%) [8, 12–15, 24–26, 41, 42] (Table 89.9). In the UPET, severe bleeding from the venous cutdown site among patients treated with urokinase occurred in 8 of 82 (9.8%) [10]. Isolated bleeding at the site of the arterial puncture was rare. Severe bleeding unrelated to the venous cutdown in the UPET occurred in 14 of 82 (17.1%). Intracranial hemorrhage Dalen and associates, based on pooled data in 559 patients, showed an incidence of intracranial hemorrhage
of 2.1% among patients treated with rt-PA for acute PE [45] (Table 89.10). The incidence of fatal intracranial hemorrhage with rt-PA was 1.6%. Based on data from registries or pooled data, a comparable rate of intracranial hemorrhage was shown with various thrombolytic agents [36, 38, 46] (Table 89.10). Others, based on meta-analysis, found lower rates of intracranial hemorrhage than previous analyses showed [33]. The rate of intracranial hemorrhage was 2 of 374 (0.5%) in patients treated with thrombolytic agents [33]. This rate was similar to the rate of intracranial hemorrhage in patients treated with heparin, 1 of 374 (0.3%) [33].
Diagnostic tests before thrombolytic therapy Multidetector CT pulmonary angiography, particularly in patients with PE in the main or lobar pulmonary, has now been shown to have a high positive predictive value, 97%, and can be readily used to establish a diagnosis of PE prior to thrombolytic therapy [47]. Prior to the use of CT pulmonary angiography, conventional pulmonary angiography had been recommended to confirm the diagnosis of PE prior to thrombolytic therapy [48]. A high probability ventilation–perfusion lung scan in a patient in whom the clinical suspicion is also high was also considered satisfactory evidence to allow the administration of thrombolytic therapy [48] and it is still satisfactory for this purpose.
Categories of patients in whom thrombolytic therapy has been considered Thrombolytic therapy has been considered for four categories of patients with PE: (1) all patients with PE,
BLUK077-Stein
March 12, 2007
18:16
434
(2) those who are stable with right ventricular dysfunction, (3) those who are hemodynamically unstable, and (4) those with massive PE who are undergoing resuscitation for cardiopulmonary arrest.
Thrombolytic therapy for all patients with PE The possibility that all patients with PE should receive thrombolytic therapy at one time had been suggested on the basis of physiological follow-up evaluation of patients who had been enrolled in the UPET [27]. Subsequent investigation by the same group of investigators showed a physiological and clinical benefit of thrombolytic therapy in many of these patients after 7 years [28]. Even so, thrombolytic therapy for all patients with PE is not recommended, because the bleeding rate is high, and the mortality among patients treated with heparin alone is very low [49]. Thrombolytic therapy for stable patients with PE and right ventricular dysfunction At one major university center, thrombolytic therapy is given to all patients with PE who have right ventricular dysfunction unless contraindicated, in spite of sparse data as to its efficacy [50]. The need for a randomized trial in such patients, however, is recognized [35, 50]. In our opinion [1], data are insufficient to recommend thrombolytic therapy in such patients, the risk of bleeding would seem to exceed the benefits. Patients with PE who are hemodynamically unstable Hemodynamic instability is a generally accepted indication for the administration of thrombolytic therapy [49]. Intuitively, the indication seems valid [1]. There are, however, no definitive trials to prove the validity of thrombolytic therapy in PE, even for unstable patients. One randomized controlled trial was halted for ethical reasons after 4 desperately ill patients with PE who received thrombolytic therapy survived, whereas 4 treated only with heparin died [34]. Our opinion is that thrombolytic therapy is indicated in patients who are hemodynamically unstable provided there are no contraindications, even though the effectiveness has not been established by randomized controlled trials [1].
PART IV
Prevention and Treatment of DVT and PE
Thrombolytic therapy for patients with massive PE who are being resuscitated from cardiopulmonary arrest Several case reports and case series have shown survival among patients with cardiopulmonary arrest who received thrombolytic therapy during the resuscitative effort [51]. This is a heroic measure that may benefit some patients. A Consensus Committee on Pulmonary Embolism in 1996 cautioned that there should be some objective evidence of PE, such as right ventricular dysfunction on an echocardiogram, or a positive noninvasive leg test [48]. Others on the committee felt that the risks of a misdiagnosis are so great that a definitive diagnosis by pulmonary angiography (or now CT pulmonary angiography) is mandatory [48]. Our opinion is that if there is no contraindication, thrombolytic therapy is indicated in patients in cardiopulmonary arrest in whom the diagnosis appears to be PE [1].
References 1 Stein PD, Dalen JE. Thrombolytic therapy in acute pulmonary embolism. In: Dalen JE, ed. Venous Thromboembolism. Marcel Dekker, New York, 2003: 253–269. 2 Tillett WS, Garner RL. The fibrinolytic activity of hemolytic streptococci. J Exp Med 1933; 58: 485–502. 3 Browse NL, James DCO. Streptokinase and pulmonary embolism. Lancet 1964; 18: 1039–1043. 4 Hansen PF, Jorgensen M, Kjeldgaard NO, Ploug J. Urokinase—an activator of plasminogen from human urine. Experiences with intravenous application on twenty-two patients. Angiology 1961; 12: 367–371. 5 Goldhaber SZ. The current role of thrombolytic therapy for pulmonary embolism. Semin Vasc Surg 2000; 13: 217– 220. 6 MICROMEDEX(R) Healthcare Series. Vol. 128 expires 6/2006. Thomson Healthcare, Inc. http://micromedex. trinity-health.org. Accessed May 30, 2006. 7 Goldhaber SZ. Pulmonary embolism thrombolysis: do we need another agent? Am Heart J 1999; 138: 1–2. 8 Verstraete M, Miller GA, Bounameaux H et al. Intravenous and intrapulmonary recombinant tissue-type plasminogen activator in the treatment of acute massive pulmonary embolism. Circulation 1988; 77: 353– 360. 9 Tibbutt DA, Davies JA, Anderson JA et al. Comparison by controlled clinical trial of streptokinase and heparin in treatment of life-threatening pulmonary embolism. BMJ 1974; 1: 343–347.
BLUK077-Stein
March 12, 2007
CHAPTER 89
18:16
Thrombolytic therapy in acute PE
10 A National Cooperative Study. The Urokinase Pulmonary Embolism Trial. Circulation 1973; 47(suppl II): II-1–II108. 11 Ly B, Arnesen H, Eie H, Hol R. A controlled clinical trial of streptokinase and heparin in the treatment of major pulmonary embolism. Acta Med Scand 1978; 203: 465– 470. 12 PIOPED Investigators. Tissue plasminogen activator for the treatment of acute pulmonary embolism. Chest 1990; 97: 528–533. 13 Levine M, Hirsh J, Weitz J et al. A randomized trial of a single bolus dosage regimen a recombinant tissue plasminogen activator in patients with acute pulmonary embolism. Chest 1990; 98: 1473–1479. 14 Dalla-Volta S, Palla A, Santolicandro A et al. PAIMS 2: Alteplase combined with heparin versus heparin in the treatment of acute pulmonary embolism. Plasminogen activator Italian multicenter study 2. J Am Coll Cardiol 1992; 20: 520–526. 15 Goldhaber SZ, Haire WD, Feldstein ML et al. Alteplase versus heparin in acute pulmonary embolism: randomized trial assessing right-ventricular function and pulmonary perfusion. Lancet 1993; 341: 507–511. 16 Leeper KV, Popovich J, Lesser BA et al. Treatment of massive acute pulmonary embolism. The use of low doses of intrapulmonary arterial streptokinase combined with full doses of systemic heparin. Chest 1988; 93: 234–240. 17 Gallus AS, Hirsch J, Cade JF, Turpie AGG, Walker IR, Gent M. Thrombolysis with a combination of small doses of streptokinase and full doses of heparin. Semin Thromb Hemost 1975; 2: 14–32. 18 Vujic I, Young JWR, Gobien RP, Dawson WT, Liebscher L, Shelley BE, Jr. Massive pulmonary embolism: treatment with full heparinization and topical low-dose streptokinase. Radiology 1983; 148: 671–675. 19 Ambrose JE, Venditto M, Dickerson WH. Local fibrinolysis for the treatment of massive pulmonary embolism: efficacy of streptokinase infusion through pulmonary arterial catheter. J Am Osteopath Assoc 1985; 85: 97– 101. 20 Edwards IR, MacLean KS, Dow JD. Low-dose urokinase in major pulmonary embolism. Lancet 1973; 2: 409–413. 21 A Cooperative Study. Urokinase–streptokinase embolism trial, Phase 2 results. JAMA 1974; 229: 1606–1613. 22 Goldhaber SZ, Kessler CM, Heit J et al. Randomized controlled trial of recombinant tissue plasminogen activator versus urokinase in the treatment of acute pulmonary embolism. Lancet 1988; 2: 293–298. 23 Goldhaber SZ, Kessler CM, Heit JA et al. Recombinant tissue-type plasminogen activator versus a novel dosing regimen of urokinase in acute pulmonary embolism: a randomized controlled multicenter trial. J Am Coll Cardiol 1992; 20: 24–30.
435
24 24. Meyer G, Sors H, Charbonnier B et al. Effects of intravenous urokinase versus alteplase on total pulmonary resistance in acute massive pulmonary embolism: a European multicenter double-blind trial. J Am Coll Cardiol 1992; 19: 239–245. 25 Meneveau N, Schiele F, Vuillemenot A et al. Streptokinase vs alteplase in massive pulmonary embolism. A randomized trial assessing right heart haemodynamics and pulmonary vascular obstruction. Eur Heart J 1997; 18: 1141–1148. 26 Meneveau N, Schiele F, Metz D et al. Comparative efficacy of a two-hour regimen of streptokinase versus alteplase in acute massive pulmonary embolism: immediate clinical and hemodynamic outcome and one-year follow-up. J Am Coll Cardiol 1998; 31: 1057–1063. 27 Sharma GVRK, Burleson VA, Sasahara AA. Effect of thrombolytic therapy on pulmonary capillary blood volume in patients with pulmonary embolism. N Engl J Med 1980; 303: 842–845. 28 Sharma GVRK, Folland E, McIntyre KM, Sasahara AA. Long-term benefit of thrombolytic therapy in patients with pulmonary embolism. Vasc Med 2000; 5: 91–95. 29 Marini C, Di Ricco G, Rossi G, Rindi M, Palla R, Giuntini C. Fibrinolytic effects of urokinase and heparin in acute pulmonary embolism: a randomized clinical trial. Respiration 1988; 54: 162–173. 30 Konstantinides S, Geibel A, Heusel G, Heinrich F, Kasper W, for the Management Strategies and Prognosis of Pulmonary Embolism-3 Trial Investigators. Heparin plus alteplase compared with heparin alone in patients with submassive pulmonary embolism. N Engl J Med 2002; 347: 1143–1150. 31 Kucher N, Rossi E, De Rosa M, Goldhaber SZ. Massive pulmonary embolism. Circulation 2006; 113: 577–582. 32 Thabut G, Thabut D, Myers RP et al. Thrombolytic therapy of pulmonary embolism: a meta-analysis. J Am Coll Cardiol 2002; 40: 1660–1667. 33 Wan S, Quinlan DJ, Agnelli G, Eikelboom JW. Thrombolysis compared with heparin for the initial treatment of pulmonary embolism: a meta-analysis of the randomized controlled trials. Circulation 2004; 110: 744– 749. 34 Jerjes-Sanchez C, Ramirez-Rivera A, de Lourdes Garcia M et al. Streptokinase and heparin versus heparin alone in massive pulmonary embolism: a randomized controlled trial. J Thromb Thrombolysis 1995; 2: 227–229. 35 Hamel E, Pacouret G, Vincentelli D et al. Thrombolysis or heparin therapy in massive pulmonary embolism with right ventricular dilatation. Results from a 128-patient monocenter registry. Chest 2001; 120: 120–125. 36 Konstantinides S, Geibel A, Olschewski M et al. Association between thrombolytic treatment and the prognosis of hemodynamically stable patients with major pulmonary
BLUK077-Stein
March 12, 2007
18:16
436
37
38
39
40
41
42
43
embolism. Results of a multicenter registry. Circulation 1997; 96: 882–888. Stein PD, Hull RD, Raskob G. Risks for major bleeding from thrombolytic therapy in patients with acute pulmonary embolism. Consideration of noninvasive management. Ann Intern Med 1994; 121: 313–317. Goldhaber SZ, Visani L, De Rosa M, for ICOPER. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999; 353: 1386–1389. UKEP Study Research Group. The UKEP study: Multicentre clinical trial on two local regimens of urokinase in massive pulmonary embolism. Euro Heart J 1987; 8: 2–10. Mikkola KM, Patel SR, Parker JA, Grodstein F, Goldhaber SZ. Increasing age is a major risk factor for hemorrhagic complications after pulmonary embolism thrombolysis. Am Heart J 1997; 134: 69–72. Sors H, Pacouret G, Azarian R, Meyer G, Charbonnier B, Simmonneau G. Hemodynamic effects of bolus vs 2-h infusion of alteplase in acute massive pulmonary embolism. A randomized controlled multicenter trial. Chest 1994; 106: 712–717. Goldhaber SZ, Agnelli G, Levine MN. Reduced dose bolus alteplase vs conventional alteplase infusion for pulmonary embolism thrombolysis. An international multicenter randomized trial. Chest 1994; 106: 718–724. Levine MN. Thrombolytic therapy for venous thromboembolism. Clin Chest Med 1995; 16: 321–328.
PART IV
Prevention and Treatment of DVT and PE
44 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316–1317. 45 Dalen JE, Alpert JS, Hirsh J. Thrombolytic therapy for pulmonary embolism. Is it effective? Is it safe? When is it indicated? Arch Intern Med 1997; 157: 2550–2556. 46 Kanter DS, Mikkola KM, Patel SR, Parker JA, Goldhaber SZ. Thrombolytic therapy for pulmonary embolism. Frequency of intracranial hemorrhage and associated risk factors. Chest 1997; 111: 1241–1245. 47 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 48 ACCP Consensus Committee on Pulmonary Embolism. Opinions regarding the diagnosis and management of venous thromboembolism. Chest 1996; 109: 233– 237. 49 Hyers TM, Agnelli G, Hull RD et al. Antithrombotic therapy for venous thromboembolic disease. Chest 2001; 119(suppl): 176S–193S. 50 Goldhaber SZ. Thrombolysis in pulmonary embolism. A large-scale clinical trial is overdue. Circulation 2001; 104: 2876–2878. 51 Bailen MR, Cuadra JAR, de Hoyos EA. Thrombolysis during cardiopulmonary resuscitation in fulminant pulmonary embolism: a review. Crit Care Med 2001; 29: 2211–2219.
BLUK077-Stein
March 28, 2007
90
16:2
CHAPTER 90
Thrombolytic therapy for deep venous thrombosis
Thrombolytic therapy for deep venous thrombosis (DVT) has been given to reduce the long-term complications of the postthrombotic syndrome (pain, swelling, skin discoloration, or ulceration) [1]. There is good evidence that thrombolytic therapy accelerates the lysis of venous thrombi [1]. Evidence that thrombolytic therapy reduces the prevalence of the postthrombotic syndrome is less certain [1]. Severe postthrombotic syndrome after 1 year, defined by some as healed or active ulceration and additional signs of congestion [2], occurred in about 3% [3] or fewer [2, 4]. In selected patients with massive ileofemoral DVT, such as those at risk of limb gangrene, thrombolytic therapy is recommended [5].
Both with streptokinase and urokinase, the frequency of the postthrombotic syndrome was lower than among controls (Table 90.1) [2, 8]. Pooled data that included therapy with both drugs showed the postthrombotic syndrome in 57% of the treated patients and 78% of the controls (Table 90.2, Figure 90.2) [2, 8]. Any bleeding with systemic full-dose streptokinase, urokinase, and t-PA ranged from 8 to 19% (Table 90.3) [2, 6–10]. On average, any bleeding occurred in 16% of the patients treated with full-dose systemic thrombolytic therapy compared with 6% of the anticoagulant-treated controls (Table 90.4) [2, 6– 10]. Early pulmonary embolism (PE) among those who received thrombolytic therapy occurred in 8% and among controls in 1% (Table 90.4) [2, 7, 8].
Systemic thrombolytic therapy
Local–regional thrombolytic therapy
Among patients who received full-dose systemic streptokinase or systemic urokinase in randomized clinical trials, complete early lysis of the thrombus occurred in 39 and 34%, respectively, compared with 5 and 2% among those who were treated only with anticoagulants (Table 90.1) [2, 6, 7]. Systemic tissue plasminogen activator (t-PA) appeared to be less effective, with only 6% showing complete early lysis (Table 90.1) [9]. Altogether, pooled data of patients who received full-dose systemic thrombolytic therapy showed complete early lysis in 28% compared with 3% of the anticoagulant-treated controls (Table 90.2, Figure 90.1) [2, 6–10]. Results with local regional and catheter-directed thrombolytic therapy are also shown in Table 90.1 and 90.2 [11–19]. Patients who received low-dose systemic streptokinase or low-dose urokinase were not included among these patients because lowdose therapy appeared to be ineffective and did not reduce bleeding [20, 21].
With local–regional thrombolytic therapy, the thrombolytic agent is infused into the region of the thrombus, but not directly into the thrombus [2, 4]. Among patients in randomized clinical trials who received a full dose of urokinase or t-PA by local–regional infusion, any early thrombolysis occurred in 72 and 66%, respectively, and complete early thrombolysis occurred in 20% among both treated groups compared with 2% among anticoagulant-treated controls (Table 90.1, Figure 90.1) [2]. Both with urokinase and t-PA, the frequency of postthrombotic syndrome was somewhat lower among treated patients than controls (Table 90.1) [2, 4]. Pooled data that included therapy with both drugs showed postthrombotic syndrome in 72% of the treated patients and 86% of the controls (Table 90.2) [2, 4]. Any bleeding with full-dose urokinase and t-PA, administered by local–regional infusion, ranged from 4 to 5%, compared with none among controls (Table 90.3)
Introduction
437
40/50 (80) 51/93 (55)
t-PA
438 33/50 (66)
t-PA
10/50 (20)
10/50 (20)
11/54 (20)
10/50 (20)
24/55 (44)
Control
20/20 (100)
—
47/52 (90)
t-PA
—
— [17, 18]
[13, 14]
[11, 12]
[2]
[2]
[9, 10]
[2]
[6–8]
Reference
*1 year after treatment. † 1–3 years after treatment. t-PA, tissue plasminogen activator. Reprinted from Alesh et al. [19].
31/34 (91)
Urokinase
Catheter-directed with angioplasty, stent, or thrombectomy
t-PA
Catheter-directed no angioplasty, stent, or thrombectomy
36/50 (72)
Urokinase
Local–regional
39/51 (76)
Streptokinase
Urokinase
Systemic
Thrombolytic
16/28 (57)
99/123 (80)
9/10 (90)
10/50 (20)
10/50 (20)
3/53 (6)
17/50 (34)
32/82 (39)
Thrombolytic
—
—
—
1/50 (2)
1/50 (2)
0/12 (0)
1/50 (2)
4/85 (5)
Control
lysis [n/N (%)]
[n/N (%)]
[17]
[13–16]
[11]
[2]
[2]
[9]
[2]
[2, 6, 7]
Reference
58/68 (85)
— —
16/52 (31)†
—
59/68 (87)
1/10 [10]†
—
48/72 (67)* 55/72 (76)*
—
41/46 (89) —
9/18 (50)
4/17 (24)*
Control
32/46 (70)*
Thrombolytic
(mild to severe) [n/N (%)]
Postthrombotic syndrome
[17, 18]
[14]
[2, 4]
[2, 4]
[2]
[8]
Reference
March 28, 2007
Thrombolytic agent
Complete (100%) early
Any early lysis
Table 90.1 Pooled data of therapeutic results using full-dose systemic, local–regional, or catheter-directed thrombolytic therapy.
BLUK077-Stein
16:2
439 78/86 (91)
Catheter-directed with angioplasty,
*1 year after treatment. † 1–3 years after treatment. Reprinted from Alesh et al. [19] with permission.
stent, or thrombectomy
stent, or thrombectomy
20/20 (100)
Catheter-directed no angioplasty, —
—
20/100 (20)
45/159 (28)
130/194 (67) 69/100 (69)
Systemic
Control
Thrombolytic
[13, 14, 17, 18]
[11, 12]
[2]
[2, 6–10]
Reference
115/151 (76)
9/10 (90)
20/100 (20)
52/185 (28)
Thrombolytic
—
—
2/100 (2)
5/147 (3)
Control
lysis [n/N (%)]
[n/N (%)]
Local–regional
Full-dose thrombolytic agent
Complete (100%) early
Any early lysis
[13–17]
[11]
[2]
[2, 6, 7, 9]
Reference
—
17/62 (27)†
117/136 (86) —
50/64 (78)
36/63 (57)* 103/144 (72)* —
Control
Thrombolytic
(mild to severe) [n/N (%)]
Postthrombotic syndrome
[14, 17, 18]
[2, 4]
[2, 8]
Reference
March 28, 2007
Table 90.2 Summary of pooled data using full-dose systemic, local–regional, or catheter-directed thrombolytic therapy.
BLUK077-Stein
16:2
BLUK077-Stein
March 28, 2007
16:2
440
PART IV
Prevention and Treatment of DVT and PE
100
Complete early lysis (%)
90% 80
76%
60
40 28% 20%
20 3% n = 147
n = 185
n = 100
n = 10
n = 151
Systemic
Local− regional
Cathdirected
Cath-directed +/− adjunct*
0 Anticoag only
Figure 90.1 Pooled data showing percent of patients with deep venous thrombosis (DVT) in whom complete early thrombolysis resulted with anticoagulant therapy, systemic thrombolytic therapy, local–regional administration, catheter-directed alone, and catheter-directed
administration with or without adjunct balloon angioplasty, stents, or thrombectomy. Asterisk indicate, angioplasty, stent, or thrombectomy. (Data from Alesh et al. [19].)
[2, 4]. On average, bleeding (both minor and major) occurred in 5% of the patients treated with full-dose local–regional thrombolytic therapy compared with 0% among anticoagulant-treated controls (Table 90.4) [2, 4]. Early PE did not occur in these patients (Table 90.4) [2].
Catheter-directed thrombolytic therapy Catheter-directed thrombolytic therapy potentially may provide an improved long-term outcome and may be preferable to systemically administered
Postthrombotic syndrome (%)
100 84% 80
72% 57%
60
40 27% 20
0
n = 200
n = 63
n = 144
Anticoag only
Systemic
Local− regional
Figure 90.2 Pooled data showing the percent of patients with deep venous thrombosis (DVT) in whom complete postthrombotic syndrome developed following anticoagulant therapy, systemic thrombolytic therapy, local–regional administration, and catheter-directed administration with or without adjunct balloon
n = 62 Cath-directed +/− adjunct*
angioplasty, stents, or thrombectomy. The prevalence of postthrombotic syndrome in anticoagulant-treated controls is the average of prevalences in controls for systemic thrombolysis and controls for local–regional thrombolysis. (Data from Alesh et al. [19].)
BLUK077-Stein
March 28, 2007
CHAPTER 90
16:2
441
Thrombolytic therapy for DVT
Table 90.3 Pooled data of complications with full-dose systemic, local–regional, or catheter-directed thrombolytic therapy. Any bleeding* [n/N (%)] Thrombolytic agent
Early PE [n/N (%)]
Thrombolytic
Control
Reference
Thrombolytic
Control
Reference
Streptokinase
20/104 (19)
11/106 (10)
[2, 6–8]
6/80 (8)
2/91 (2)
[2, 7, 8]
Urokinase
4/50 (8)
0/50 (0)
[2]
4/50 (8)
0/50 (0)
[2]
t-PA
16/94 (17)
2/54 (4)
[9, 10]
—
—
Systemic
Local–regional Urokinase
3/73 (4)
0/73 (0)
[2, 4 ]
0/50 (0)
0/50 (0)
[2]
t-PA
4/73 (5)
0/73 (0)
[2, 4]
0/50 (0)
0/50 (0)
[2]
[11, 12]
4/20 (20)†
—
[11, 12]
Catheter-directed no angioplasty, stent, or thrombectomy t-PA
6/20 (30)
—
Catheter-directed with angioplasty, stent, or thrombectomy Urokinase
30/111 (27)
—
[13, 14, 16]
1/123 (1)‡
—
[13–16]
t-PA
19/52 (37)
—
[17, 18]
1/52 (2)‡
—
[17, 18]
*Includes both major and minor bleeding. † Three of 4 patients were found on routine survey with scintiscans. ‡ Inferior vena cava filters were inserted in 8 patients treated with urokinase [2, 3] and in 1 treated with t-PA [8]. PE, pulmonary embolism; t-PA, tissue plasminogen activator. Reprinted from Alesh et al. [19] with permission.
thrombolytic agents [22, 23]. With catheter-directed thrombolytic therapy, the tip of the catheter is placed inside the thrombus and the thrombolytic agent is administered directly into the thrombus [15]. The only randomized clinical trial of catheter-directed thrombolytic therapy was with streptokinase administered to 18 patients [24]. Complete thrombolysis after 1 week was observed in 11 of 18 (61%) versus 0 of 17 (0%) anticoagulant-treated controls [24]. No major bleeding occurred. Postthrombotic syndrome was not
reported. However, normal late venous function after 6 months was reported in 72% of the treated patients and in only 12% of the anticoagulant-treated controls [24]. Because immediate thrombolysis was not reported in these patients, the investigation was excluded from the pooled data. Many of the investigators who used catheterdirected thrombolysis used balloon angioplasty, stents, or thrombectomy in addition [19]. Only two case series were reported of catheter-directed thrombolytic
Table 90.4 Summary of pooled data of complications with full-dose systemic, local–regional, or catheter-directed thrombolytic therapy.
Any full-dose thrombolytic agent
Any bleeding* [n/N (%)] Thrombolytic
Control
Early PE [n/N (%)] Reference
Thrombolytic
Control
Reference
Systemic
40/248 (16)
13/210 (6)
[2, 6–10]
10/130 (8)
2/141 (1)
[2, 7, 8]
Local–regional
7/146 (5)
0/146 (0)
[2, 4 ]
0/100 (0)
0/100 (0)
[2]
Catheter-directed no angioplasty,
6/20 (30)
—
[11, 12]
4/20 (20)†
—
[11, 12]
49/163 (30)
—
[13-18]
2/175 (1)‡
—
[13–18]
stent, or thrombectomy Catheter-directed with angioplasty, stent, or thrombectomy *Includes both major and minor bleeding. † Three of 4 patients were found on routine survey with scintiscans. ‡ Inferior vena cava filters were inserted in 8 patients treated with urokinase [2, 3] and in 1 treated with t-PA [8]. PE, pulmonary embolism; t-PA, tissue plasminogen activator. Reprinted from Alesh et al. [19] with permission.
BLUK077-Stein
March 28, 2007
16:2
442
therapy alone [11, 12]. Both used t-PA. Complete early thrombolysis, reported in only 10 patients, was achieved in 90% (Table 90.1, Figure 90.1) [11]. The prevalence of the postthrombotic syndrome was not reported. None of the patients suffered major bleeding and 30% suffered minor bleeding (Table 90.3) [11, 12]. Early recurrent PE occurred in 20% [11, 12]. Three of these patients were asymptomatic and detected by survey with ventilation–perfusion lung scans after treatment. Systematic review [19] showed that the majority of the case series of catheter-directed administration of thrombolytic agents included patients who were also treated with balloon angioplasty, stents, or thrombectomy [13–18]. Among patients treated with t-PA, 38% had such adjunct therapy [17, 18]. Complete early opening with t-PA, often followed by angioplasty, insertion of stents, or thrombectomy was produced in 57% [17] and the postthrombotic syndrome was reported in 31% (Table 90.1) [17, 18]. Major bleeding occurred in 12% and minor bleeding occurred in 25% [17, 18]. Doses of thrombolytic agents and durations of therapy varied. Among 123 patients treated with catheter-directed urokinase, approximately 60% were treated, in addition, with balloon angioplasty, stents, or thrombectomy [13–16]. Complete early opening with urokinase, often followed by angioplasty, stents, or thrombectomy, was shown in 80% [13–16]. The incidence of postthrombotic syndrome was reported in only one series [14]. Among 10 patients there was one reported case of postthrombotic syndrome that occurred in 1–3 years [14]. Pooled data of patients treated with catheterdirected thrombolytic therapy, either with t-PA or urokinase, often followed by adjunct therapy, based on systematic review [19] showed complete early opening in 76% [13–17] and postthrombotic syndrome in 27% [14, 17, 18] (Table 90.2, Figure 90.1). Complete early opening of occluded veins was accomplished more frequently with catheter-directed thrombolysis alone (90%) [11] or with catheter-directed thrombolysis often followed by adjunct therapy (76%) [13– 17] than with a systemic infusion (28%) [2, 6, 7, 9] or local–regional administration (20%) [2]. Any early thrombolysis also occurred at higher rates with catheter-directed administration [11, 12] and catheterdirected administration often followed by adjunct therapy [13, 14, 17, 18] than with systemic [2, 6–10]
PART IV
Prevention and Treatment of DVT and PE
or local–regional [2] administration (Tables 90.1 and 90.2). The prevalence of postthrombotic syndrome was lower with catheter-directed administration often followed by adjunct therapy, 27% [14, 17, 18], compared with 57 [2, 8] and 72% [2, 4] (Table 90.2, Figure 90.2). The prevalence of the postthrombotic syndrome in anticoagulant-treated control patients was higher than reported in a survey of patients with DVT [3]. This probably reflects the randomized selection of anticoagulant-treated controls from a group of patients who had DVT of sufficient severity to warrant thrombolytic therapy. Any bleeding with catheter-directed thrombolytic therapy alone occurred in 30% [11, 12] and with catheter-directed therapy followed by adjunct therapy, it also occurred in 30% [13–18]. With catheterdirected therapy alone, all bleeding was minor. With catheter-directed therapy followed by adjunct therapy, in 36 of 49 bleeding was minor. Early recurrent PE was not more frequent with catheter-directed thrombolysis in combination with adjunct therapy [13–18] than in patients treated with systemic thrombolysis [2, 7, 8] (Table 90.4). In patients with massive DVT who require thrombolytic therapy, catheter-directed thrombolytic therapy may be more beneficial than systemic or local–regional administration [19]. An advantage of catheter-directed thrombolytic therapy is that it lends itself to angioplasty, stent insertion, and thrombectomy following the administration of thrombolytic agents if the thrombolysis is inadequate [19].
References 1 Watson LI, Armon MP. Thrombolysis for acute deep vein thrombosis. Cochrane Database Syst Rev 2004; (4): CD002783. 2 Schweizer J, Kirch W, Koch R et al. Short- and longterm results after thrombolytic treatment of deep venous thrombosis. J Am Coll Cardiol 2000; 36: 1336–1343. 3 Prandoni P, Lensing AW, Cogo A et al. The long-term clinical course of acute deep venous thrombosis. Ann Intern Med 1996; 125: 1–7. 4 Schweizer J, Elix H, Altmann E, Hellner G, Forkmann L. Comparative results of thrombolysis treatment with rt-PA and urokinase: a pilot study. VASA 1998; 27: 167– 171. 5 Buller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease: the Seventh ACCP Conference on
BLUK077-Stein
March 28, 2007
CHAPTER 90
6
7
8
9
10
11
12
13
14
16:2
Thrombolytic therapy for DVT
Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 401S–428S. Common HH, Seaman AJ, Rosch J, Porter JM, Dotter CT. Deep vein thrombosis treated with streptokinase or heparin. Follow-up of a randomized study. Angiology 1976; 27: 645–654. Kakkar VV, Flanc C, Howe CT, O’Shea M, Flute PT. Treatment of deep vein thrombosis: a trial of heparin, streptokinase, and arvin. BMJ 1969; 1: 806–810. Arnesen H, Heilo A, Jakobsen E, Ly B, Skaga E. A prospective study of streptokinase and heparin in the treatment of deep vein thrombosis. Acta Med Scand 1978; 203: 457–463. Goldhaber SZ, Meyerovitz MF, Green D et al. Randomized controlled trial of tissue plasminogen activator in proximal deep venous thrombosis. Am J Med 1990; 88: 235–240. Turpie AG, Levine MN, Hirsh J et al. Tissue plasminogen activator (rt-PA) vs heparin in deep vein thrombosis. Results of a randomized trial. Chest 1990; 97: 172S– 175S. Horne MK, III, Mayo DJ, Cannon RO, III, Chen CC, Shawker TH, Chang R. Intraclot recombinant tissue plasminogen activator in the treatment of deep venous thrombosis of the lower and upper extremities. Am J Med 2000; 108: 251–255. Chang R, Cannon RO, III, Chen CC et al. Daily catheterdirected single dosing of t-PA in treatment of acute deep venous thrombosis of the lower extremity. J Vasc Interv Radiol 2001; 12: 247–252. Raju S, Fountain T, McPherson SH. Catheter-directed thrombolysis for deep venous thrombosis. J Miss State Med Assoc 1998; 39: 81–84. Patel NH, Stookey KR, Ketcham DB, Cragg AH. Endovascular management of acute extensive iliofemoral deep venous thrombosis caused by May–Thurner syndrome. J Vasc Interv Radiol 2000; 11: 1297–1302.
443
15 Molina JE, Hunter DW, Yedlicka JW. Thrombolytic therapy for iliofemoral venous thrombosis. Vasc Surg 1992; 26: 630–637. 16 Bjarnason H, Kruse JR, Asinger DA et al. Iliofemoral deep venous thrombosis: safety and efficacy outcome during 5 years of catheter-directed thrombolytic therapy. J Vasc Interv Radiol 1997; 8: 405–418. 17 Ly B, Njaastad AM, Sandbaek G, Solstrand R, Rosales A, Slagsvold CE. Catheter-directed thrombolysis of iliofemoral venous thrombosis. Tidsskr Nor Laegeforen 2004; 124: 478–480. 18 Verhaeghe R, Stockx L, Lacroix H, Vermylen J, Baert AL. Catheter-directed lysis of iliofemoral vein thrombosis with use of rt-PA. Eur Radiol 1997; 7: 996– 1001. 19 Alesh I, Kayali F, Beemath A, Janjua M, Patel NR, Stein PD. Catheter-directed thrombolysis (intra-thrombus injection) in treatment of deep venous thrombosis: a systematic review. Catherization and Cardiovascular Interventions (In Press). 20 Schulman S, Granqvist S, Juhlin-Dannfelt A, Lockner D. Long-term sequelae of calf vein thrombosis treated with heparin or low-dose streptokinase. Acta Med Scand 1986; 219: 349–357. 21 Kiil J, Carvalho A, Sakso P, Nielsen HO. Urokinase or heparin in the management of patients with deep vein thrombosis? Acta Chir Scand 1981; 147: 529–532. 22 Meissner MH. Thrombolytic therapy for acute deep vein thrombosis and the venous registry. Rev Cardiovasc Med 2002; 3(suppl 2): S53–S60. 23 Semba CP, Dake MD. Catheter-directed thrombolysis for iliofemoral venous thrombosis. Semin Vasc Surg 1996; 9: 26–33. 24 Elsharawy M, Elzayat E. Early results of thrombolysis vs anticoagulation in iliofemoral venous thrombosis: a randomised clinical trial. Eur J Vasc Endovasc Surg 2002; 24: 209–214.
BLUK077-Stein
April 6, 2007
91
17:5
CHAPTER 91
Inferior vena cava filters: trends in use, complications, indications, and use of retrievable filters
Introduction Since the first percutaneous umbrella inferior vena cava (IVC) filter was described by Mobin-Uddin in 1969 [1] several have been designed for percutaneous insertion and approved by the United States Food and Drug Administration [2] (Table 91.1). They differ in outer diameter of the delivery system, maximal IVC diameter into which they can be inserted, hook design, retrievability, biocompatibility, and filtering efficiency [3–10]. Improved technology in the fabrication of IVC filters has made them less thrombogenic, smaller, easier to insert percutaneously, safer, in some instances retrievable [11–14], and capable of insertion at the bedside [15, 16]. This has led to a broadening of the indications for insertion [12, 17]. There is uniform agreement [18] that an IVC filter should be inserted in a patient with proximal deep venous thrombosis (DVT) or pulmonary embolism (PE), if (1) anticoagulants are contraindicated, (2) PE has recurred while on adequate anticoagulant therapy, and (3) PE is so severe that any recurrent PE may be fatal. Insertion of an IVC filter is also strongly recommended in patients following pulmonary embolec-
tomy [19]. It is felt that routine insertion of an IVC filter is not indicated only on the basis of a continuing predisposition for DVT [18]. In special circumstances, however, this may be the best approach. Some have recommended prophylactic insertion of IVC filters for high-risk patients with DVT, severe pulmonary hypertension, and minimal cardiopulmonary reserve [20]. Additional indications include patients with chronic recurrent PE and pulmonary hypertension and patients undergoing thromboendarterectomy [21]. Broader indications (patients with poor cardiopulmonary reserve in whom even a small recurrent PE might be fatal and patients who show a free floating thrombus in the IVC) now account for 46–65% of IVC insertions [12]. More liberal recommendations by some include prophylaxis in patients with cancer [22], trauma [23, 24], burns [25], acetabular fracture [26], hip or knee replacement in patients with a history of thromboembolism [27], or indeed, prophylaxis in all patients with DVT or PE, especially if over 65 years of age [28]. Many of the indications for insertion of IVC filters are a matter of opinion [29]. Complications of IVC filter insertion include improper anatomic placement of the filter (7%)
Table 91.1 Inferior vena cava filters approved by the United States Food and Drug Administration for permanent percutaneous insertion. Stainless steel Greenfield filter
Medi-tech/Boston Scientific, Natick, Massachusetts
Titanium Greenfield filter (TGF)
Medi-tech/Boston Scientific, Natick, Massachusetts
Vena Tech-LGM filter
B. Braun Medical, Bethlehem, Pennsylvania
Vena Tech LP filter
B. Braun Medical, Bethlehem, Pennsylvania
Simon Nitinol filter
Nitinol Medical Technologies; Bard, Boston, Massachusetts
Bird’s nest filter
Cook Inc., Bloomington, Indiana
TrapEase filter
Cordis Endovascular, Warren, New Jersey
Data from Siskin et al. [2].
444
BLUK077-Stein
April 6, 2007
CHAPTER 91
17:5
445
Inferior vena cava filters
Figure 91.2 Greenfield filter 2 years after insertion, embolized on first post-operative day following laparotomy. Transverse CT image of heart shows filter in right ventricle. Courtesy of Malik McKany, MD and Einas Joseph, MD, St. Joseph Mercy Hospital Oakland Hospital, Pontiac, Michigan.
Figure 91.1 Mal-positioned Simon nitinol filter partially lodged in the right renal vein.
(Figure 91.1), migration (2–3%), Figure 91.2 angulation of the filter (2%), caval stenosis or filter narrowing (2%), caval occlusion (2–9%), air embolism (1%), penetration of the caval wall (1%), lower extremity edema (13–26%), and sequelae of venous stasis (27%) [11, 17, 23, 30]. Deep venous thrombosis at the puncture site has been reported in 8–25% [31–34] and in a series of 17 patients, 41% developed DVT at the puncture site [35]. Additional complications include filter deformation, filter fracture, insufficient opening of the filter, and erosion of the caval wall [30]. Three percent of the patients with Greenfield filters had a PE on follow-up and 1% had a fatal PE [30]. In a review of investigations since 1994 of trauma patients who had a filter inserted prophylactically, 2% had a PE [23]. More data are needed on the relative safety and efficacy of IVC filters [36]. In a registry of patients, those with massive PE, defined as systolic arterial pressure <90 mm Hg, 90-day
survival was higher in a few treated with IVC filters, 91%, compared with those who did not receive IVC filters, 57% (Table 91.2) [37]. The improved survival related to a lower rate of recurrent PE among those who received IVC filters, 0 of 11 (0%) versus 13 of 97 (13%). Among patients with nonmassive PE (systolic arterial pressure ≥90 mm Hg), 90-day survival was lower in those who received IVC filters compared with those who did not, 79% versus 86% [37]. Some observed that the prophylactic use of IVC filters in patients following trauma failed to decrease the overall rate of PE [38]. Others observed that the use of an IVC filter in patients with DVT or PE did not decrease the rate of rehospitalization for recurrent DVT or PE compared to patients treated with anticoagulants Table 91.2 90-day survival and use of inferior vena cava filters. Survived/PE (%) IVC filter
No IVC filter
Massive PE
10/11 (91)
55/97 (57)
Nonmassive PE
179/227 (79)
1769/2057 (86)
PE, pulmonary embolism; IVC, inferior vena cava. Data are from Kucher et al. [37].
BLUK077-Stein
April 6, 2007
17:5
No. of IVC filters
446
PART IV
Prevention and Treatment of DVT and PE
25000 DVT
20000
PE
15000
No PE or DVT
10000 5000 1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
0
Years
In 1999, 22,000 IVC filters were inserted in hospitalized patients with DVT, 18,000 with PE, and 9000 in patients who did not have a coded diagnosis of DVT or PE. In 1999, 45% of IVC filter insertions were in patients with DVT alone, 36% were in patients with PE, and 19% of insertions were in patients who presumably were at high risk but did not have DVT or PE listed as a discharge code. Between 1979 and 1985 some of the coded procedures may have been open surgical ligation, plication, or insertion of an IVC filter [43]. After 1985, virtually all of the coded vena cava procedures were transvenous insertion of an IVC filter [43]. Only a few case reports of insertion of a filter in the superior vena cava were reported prior to 1999 [44–46]. The number of IVC filters identified by our analysis of the NHDS database [42] corresponds closely to the estimate of 30,000–40,000 IVC filters inserted yearly, based on calculations by the industry [47]. The percentage of patients with PE who had IVC filters inserted increased from 0.7% in the triennial period 1979–1981 to 12% in 1997–1999 (Figure 91.4) [42]. The percentage of patients with DVT who had
20 15 PE
10
DVT
5 0 1997−1999
1994−1996
Years
1991−1993
1988−1990
1985−1987
1982−1984
1979−1981
Filters insertion rate per 100 patients with PE or DVT
alone [39]. There is a sparsity of studies comparing the use of IVC filters with anticoagulants [36, 40]. With more liberal indications, increased use of IVC filters has followed [13, 17, 41]. According to some, there is a potential for unwarranted insertion [17]. Proper selection of patients for IVC filter insertion is an important challenge [23]. The nonselective use of IVC filters is associated with unacceptable morbidity and mortality [17]. Based on data from the National Hospital Discharge Survey (NHDS) database we evaluated trends over 21 years in the use of IVC filters in patients with PE, patients with DVT but no PE, and patients with no diagnosis of DVT or PE who presumably were at high risk [42]. An analysis of records of Medicare patients indicates that the use of IVC filters is increasing in elderly patients [41]. There are no other reports for the United States on trends in the use of IVC filters. The number of IVC filters inserted yearly over the 21-year period of observation increased from 2000 in 1979 to 49,000 in 1999 (Figure 91.3) [42]. The number inserted yearly began to increase sharply in 1988.
Figure 91.3 Number of IVC filters inserted between 1979 and 1999 in patients with pulmonary embolism (PE), deep venous thrombosis (DVT) alone, and prophylactically (no PE or DVT). (Reprinted with permission from Stein et al. [42]. Copyright 2004, American Medical Association. All rights reserved.)
Figure 91.4 Rates of insertion of IVC filters/100 patients with pulmonary embolism (PE) and rates/100 patients with deep venous thrombosis (DVT) alone. (Reprinted with permission from Stein et al. [42]. Copyright 2004, American Medical Association. All rights reserved.)
April 6, 2007
CHAPTER 91
17:5
447
Inferior vena cava filters 15
NE
10
MW S
5
W
0 1999
1997
1995
1993
1991
1989
1987
In July and October 2003 and March 2004, 3 retrievable IVC filters were approved by the Food and Drug Administration (FDA): Recovery (Bard Peripheral Vas-
1985
Retrievable IVC filters
1983
IVC filters inserted increased from 0.2% in 1979–1981 to 6% in 1997–1999 (Figure 91.4). The percentage of patients with PE who had IVC filters inserted increased linearly over the 21-year period of observation as did the percentage of patients with DVT who had IVC filters (Figure 91.4). Trends comparable to the national trends that we observed have been shown in a study of the experience of a university hospital [43]. In the triennial period 1997–1999, 12% of the patients hospitalized with PE had IVC filters and 6% of the patients hospitalized with DVT had IVC filters (Figure 91.4) [42]. Over the 21-year period of observation, the percentage of patients in whom IVC filters were inserted, both for PE and for DVT, was comparable in men and women, black and white patients, elderly patients (70 years or older), and younger patients (20–69 years). Regional differences in the use of IVC filters for PE and/or DVT were shown (Figure 91.5). The use of IVC filters over the 21-year period of observation was more frequent in northeastern states than in western states (Figure 91.5) [42]. Regional differences in the use of IVC filters in Medicare patients, with greater use in the New England states than on the Pacific coast [41] had previously been shown. This is a further evidence of differences of opinion regarding the indication for IVC filters. Differences in the rate of IVC filter insertion between the United States and Sweden [30] suggest differences of opinion across national borders.
1981
1979
Figure 91.5 Rates of insertion of IVC filters/100 patients with deep venous thrombosis (DVT) and/or pulmonary embolism (PE) according to the region in the United States. The regions are northeast (NE), midwest (MW), south (S), and west (W). (Reprinted with permission from Stein et al. [42]. Copyright 2004, American Medical Association. All rights reserved.)
Insertion rate per 100 patients with DVT or PE
BLUK077-Stein
Years
cular, Inc., Tempe, AZ), G¨unther Tulip (Cook, Inc., Bloomington, IN), and OptEase (Cordis Endovascular, Warren, NJ). With the ready availability of retrievable IVC filters, physicians have an option in addition to anticoagulants in the management of venous thromboembolic disease. Fundamental data on the safety of insertion, occurrence of venous thromboembolism (VTE) while the IVC filter is in place, and the safety of retrieval have been sparsely reported. We reviewed this information [48]. A systematic review identified 6 prospective case series in which 4 different types of retrievable IVC filters were used in 284 patients [48]. Among patients in whom percutaneous removal of the filter was attempted, the filter was successfully removed in 144 of 159 (91%), but there was obvious heterogeneity in the data, so this figure is clearly an estimate. Studies were included in Tier 1 if they met all of the following criteria: (1) The study was performed prospectively; (2) the indication for retrievable IVC filter insertion was stated; (3) the duration of insertion and number retrieved were stated; (4) complications of IVC filters during insertion, while the filter was in place, and at retrieval were stated (Figure 91.6); (5) a broad population of patients and broad range of indications for IVC filters were studied; and (6) the approach for insertion and retrieval were stated. Investigations that failed to examine a broad spectrum of indications or investigated a selected population were defined as Tier 2. Investigations that were retrospective or did not state results according to the type of filter were defined as Tier 3. No investigation compared the outcome of patients treated with retrievable IVC filters to those treated without filters. We identified six Tier 1 [49–54] and
BLUK077-Stein
April 6, 2007
17:5
448
Figure 91.6 Migrated OptEase retrievable filter lodged in the tricuspid valve. Frontal plane view (top). Transverse plane view (bottom). Courtesy of Adnan H, Matta, MD and Majd Alnas, MD, St. Joseph Mercy Hospital Oakland Hospital, Pontiac, Michigan.
PART IV
Prevention and Treatment of DVT and PE
three Tier 2 investigations [55–57]. There were also nine retrospective investigations [10, 14, 58–64] and two in which the results according to the type of filter were not stated [65, 66] (Tier 3). Subsequently, an additional retrospective case series in trauma patients was reported [67]. Among Tier 1 investigations, 4 different types of filters were inserted in 284 patients [49–54]. The mean age of patients in these studies ranged from 51 to 62 years. Indications for filter insertion were a contraindication to (or complication from) anticoagulants in a patient with VTE, 121 (43%); surgery in a patient with VTE, 61 (21%); massive, chronic, or recurrent PE, DVT with high risk of PE, or patient with VTE and poor cardiopulmonary reserve, 53 (19%); VTE and failure of anticoagulants, 29 (10%); trauma, 7 (2%); and thrombolysis or thrombectomy, 4 (1%). In 1 patient (0.4%), the filter was inserted because of a paradoxical embolism, and in 8 patients (3%) the indication was not known. Typical durations of insertion were 1–14 days [51–53], but durations as long as 55 days and 134 days were reported (Table 91.3) [49, 50]. Among Tier 1 investigations, 5 patients with G¨unther Tulip filters required immediate retrieval of the filter and reinsertion of another filter because of technical problems or errors of manipulation (Table 91.4) [53, 54]. One of these filters was removed surgically through a venotomy of the femoral vein [53]. The design of the prototype insertion mechanism was subsequently modified to prevent problems that occurred in 2 of the patients [53]. A massive hematoma at the site of insertion occurred in a patient who had a coagulation disorder (Table 91.4) [52]. Outcome events during the period of filter insertion among Tier 1 patients are shown in Table 91.5. Thrombi trapped in the filter were frequent with most types of filters [49, 51–54]. Anticoagulants were used routinely after insertion only with the Tempofilter [50]. Among patients in whom percutaneous removal of the filter was attempted, the filter was successfully removed in 144 of 159 (91%). Surgery was necessary to remove the filter from 1 patient (1%) [50], and in 14 patients (9%) filters could not be removed because of large trapped thrombi [54]. Complications of retrieval of filters in 144 patients were thrombosis and DVT at the removal site in 1 patient (1%) and minor bleeding in 1 patient (1%) [52]. Reasons filters were not retrieved percutaneously are shown in Table 91.6.
BLUK077-Stein
April 6, 2007
CHAPTER 91
17:5
449
Inferior vena cava filters
Table 91.3 Tier 1 studies: type of retrievable filter, number inserted, duration of insertion, number removed.
Filter
Number
Number
Proportion
Duration of insertion
inserted*
removed
removed (%)
(days) [mean range]
Reference [51]
¨ Gunther Tulip
15
9
60
9 (5–13)
¨ Gunther Tulip
71
33
46
8 (1–13)*
[54]
¨ Gunther Tulip
83
2
2
– (6, 11)†
[53]
Recovery Nitinol
32
24
75
53 (5–134)
[49]
¨ Gunther
17
17
100
7 (3–14)
[52]
Tempofilter
66
59
89
30 (1–55)
[50]
* An additional 2 filters were replaced immediately due to complications of insertion [54]. In 2 patients with phlegmasia cerulea dolens, 2 consecutive filters were required, one before femoral venotomy and one after insertion of ileocaval stents [54]. † An additional 3 filters were replaced immediately or within 5 days due to complications of insertion [53].
Tier 2 investigations described the use of retrievable filters in 95 trauma patients with G¨unther Tulip filters inserted over a mean of 14 days [56, 67], 65 patients with Prothia filters (Prothia, Montrouge, France) inserted over a mean of 5 days in patients who were undergoing thrombolytic therapy [57], and in 4 pregnant women with Tempo filters (B. Braun Celsa, Chasseneuil Cedex, France) inserted over a mean of 21 days [55]. The only reported complications of insertion were moderate to severe hematoma at the puncture site in 1 of 65 (2%) receiving thrombolytic therapy and mild infection at the insertion site in 4 of 65 (6%) of these patients [57]. During the period of insertion,
thrombus trapped within the filter was observed with each type of filter [55–57]. No complications of retrieval were reported with any of these filters. Retrieval with the Gunther Tulip filter was successful in 24 of 25 attempts [67]. Tier 3 investigations (retrospective, or not stating results according to the type of filter) reported the use of retrievable filters in 626 patients in whom 444 filters were retrieved (Table 91.7) [10, 14, 58–66]. In 2 of 444 patients (0.5%), a PE occurred during filter retrieval [61, 65]. Intramural hemorrhage was also observed in 2 of 444 patients (0.5%) [10]. No other complications of retrieval were reported.
Table 91.4 Tier 1 investigations: complications of insertion according to the retrievable filter. Complications of Filter
No. of patients
¨ Gunther Tulip
169
Recovery Nitinol
32
¨ Gunther
17
Tempofilter
66
Complications
insertion (number)
Reference
Manipulation error†
2
[53, 54]
Insertion mechanism defect*
2
[53]
Failure of filter to expand†
1
[54]
0
[49]
[51, 53, 54]
[52] Massive hematoma
1
Moderate insertion site thrombosis/DVT
1
Infection, insertion site
2
Reopening of incision
3
[50]
* Prototype was subsequently modified; 1 patient required surgical removal, 1 required percutaneous removal. † Percutaneous removal required. Reprinted from Stein et al. [48], with permission from Elsevier.
BLUK077-Stein
April 6, 2007
17:5
450
PART IV
Prevention and Treatment of DVT and PE
Table 91.5 Tier 1 investigations: outcome during period of insertion according to the type of retrievable filter. Number (%)*
Filter
No. of patients
Outcome
¨ Gunther Tulip
169
Fatal PE
1 (1)
Nonfatal PE
2 (1)
IVC thrombosis
Reference [51, 53, 54]
7 (4)
Thrombus trapped in filter
16 (9)
Filter migration, asymptomatic
5 (3)
Filter tilt
20 (12)
Leg edema, moderate
1 (1)
IVC penetration, no clinical
2 (1)
consequence Recovery Nitinol
¨ Gunther
Tempofilter
32
17
66
Thrombus trapped in filter
7 (22)
Filter migration, asymptomatic
1 (3)
Filter tilt
2 (6)
Thrombus trapped in filter
3 (18)
Leg edema, moderate
1 (6)
IVC thrombosis
2 (12)
IVC penetration, no clinical
4 (6)
[49]
[52]
[50]
consequence Filter thrombosis
10 (15)
Filter migration
5 (8)
* Percentages were calculated on the basis of the number of patients with the indicated type of IVC filter. Reprinted from Stein et al. [48], with permission from Elsevier. Table 91.6 Tier 1 investigations: number of inferior vena cava filters retrieved percutaneously and reasons for nonretrieval. Successful
Not
Number
percutaneous
percutaneously
Reason not percutaneously
Filter
of patients
retrieval
retrieved
retrieved
Reference
¨ Gunther Tulip
169
44
125
19 deaths
[51, 53, 54]
14 large trapped thrombi 5 continuing contraindication to anticoagulants 6 patients unstable 81 intended for permanent use Recovery Nitinol
32
24
8
3 deaths
[49]
1 intraoperative removal during another procedure 4 continuing indication ¨ Gunther
17
17
0
Tempofilter
66
59
7
[52] 2 deaths 2 moribund 1 surgical retrieval 1 continuing indication 1 refused retrieval
Reprinted from Stein et al. [48], with permission from Elsevier.
[50]
BLUK077-Stein
April 6, 2007
CHAPTER 91
17:5
451
Inferior vena cava filters
Table 91.7 Tier 3 studies: type of retrievable filter, number inserted, duration of insertion, and number removed.
Filter
Number
Number
Proportion
Duration of insertion
inserted
removed
removed (%)
(days) [mean range]
Reference
¨ Gunther Tulip
9
9
100
— (6–12)
[66]
¨ Gunther Tulip
88
70
80
13 (6–20)
[59]
¨ Gunther Tulip
91
52
57
9 (2–25)
[14]
¨ Gunther Tulip
10
7
70
9 (3–13)
[63]
¨ Gunther Tulip,
50
50
100
7 (1–12)
[65]
¨ Gunther, Antheor ¨ Gunther
50
50
100
¨ Gunther,
45
45
100
— 10 (5–15)
5 (—)
[61] [62]
Antheor, Prolyser Antheor
2
2
100
Antheor
106
106
100
5 (—)
[60] [61]
Antheor
44
—
—
—
[64]
Angiocor
11
—
—
—
[64]
Amplatz
16
1
6
Amplatz
52
5
10 —
—
[64]
16
100
10 (5–15)
[60]
—
Cook
12
Filcard
16
Filcard Prolyser
—
1
1
100
32
32
100
—
[58]
— (1–16)
[10]
5 (—)
[66] [61]
Reprinted from Stein et al. [48], with permission from Elsevier.
References 1 Mobin-Uddin K, McLean R, Jude JR. A new catheter technique of interruption of inferior vena cava for prevention of pulmonary embolism. Am Surg 1969; 35: 889–894. 2 Siskin GP, Bartholomew K. Inferior vena cava filters. http://emedicine.com/radio/topic762.htm. Last updated December 1, 2005. 3 Rousseau H, Perreault P, Otal P et al. The 6-F nitinol TrapEase inferior vena cava filter: results of a prospective multicenter trial. J Vasc Interv Radiol 2001; 12: 299– 304. 4 Greenfield LJ, Michna BA. Twelve-year clinical experience with the Greenfield vena caval filter. Surgery 1988; 104: 706–712. 5 Greenfield LJ, Cho KJ, Proctor M et al. Results of a multicenter study of the modified hook-titanium Greenfield filter. J Vasc Surg 1991; 14: 253–257. 6 Roehm JOF, Johnsrude IS, Barth MH, Gianturco C. The bird’s nest inferior vena cava filter: progress report. Radiology 1988; 168: 745–749. 7 Ricco JB, Crochet D, Sebilotte P et al. Percutaneous transvenous caval interruption with the “LGM” filter: early results of a multicenter trial. Ann Vasc Surg 1988; 3: 242–247.
8 Simon M, Athanasoulis CA, Kim D et al. Simon nitinol inferior vena cava filter: initial clinical experience. Radiology 1989; 172: 99–103. 9 Fobbe F, Dietzel M, Korth R et al. Gunther vena caval filter: results of long-term follow-up. Am J Roentgenol 1988; 151: 1031–1034. 10 Epstein DH, Darcy MD, Hunter DW et al. Experience with the Amplatz retrievable vena cava filter. Radiology 1989; 172: 105–110. 11 Dorfman GS. Percutaneous inferior vena caval filters. Radiology 1990; 174: 987–992. 12 Ballew KA, Philbrick JT, Becker DM. Vena cava filter devices. Clin Chest Med 1995; 16: 295–305. 13 Whitehill TA. Current vena cava filter devices and results. Semin Vasc Surg 2000; 13: 204–212. 14 Millward SF, Oliva VL, Bell SD et al. Gunther tulip retrievable vena cava filter: results from the registry of the Canadian Interventional Radiology Association. J. Vasc Interv Radiol 2001; 12: 1053–1058. 15 Matsumura JS, Morasch MD. Filter placement by ultrasound technique at the bedside. Semin Vasc Surg 2000; 13: 199–203. 16 Sing RF, Jacobs DG, Heniford BT. Bedside insertion of inferior vena cava filters in the intensive care unit. J Am Coll Surg 2001; 192: 570–576.
BLUK077-Stein
April 6, 2007
17:5
452
17 Arnold TE, Karabinis VD, Mehta V, Dupont EL, Matsumoto T, Kerstein MD. Potential of overuse of the inferior vena cava filter. Surg Gynecol Obstet 1993; 177: 463– 467. 18 ACCP Consensus Committee on Pulmonary Embolism. Opinions regarding the diagnosis and management of venous thromboembolic disease. Chest 1996; 109: 233– 237. 19 Grassi CJ, Goldhaber SZ. Interruption of the inferior vena cava for prevention of pulmonary embolism: transvenous filter devices. Herz 1989; 14: 182–191. 20 Rohrer MJ, Scheidler MG, Wheeler HB, Cutler BS. Extended indications for placement of an inferior vena cava filter. J Vasc Surg 1989; 10: 44–50. 21 Hyers TM, Agnelli G, Hull RD et al. Antithrombotic therapy for venous thromboembolic disease. Chest 2001; 119: 176S–193S. 22 Cohen JR, Grella L, Citron M. Greenfield filter instead of Heparin as primary treatment for deep venous thrombosis or pulmonary embolism in patients with cancer. Cancer 1992; 70: 1993–1996. 23 Greenfield LJ, Proctor MC, Michaels AJ, Taheri PA. Prophylactic vena caval filters in trauma: the rest of the story. J Vasc Surg 2000; 32: 490–497. 24 Carlin AM, Tyburski JG, Wilson RF, Steffes C. Prophylactic and therapeutic inferior vena cava filters to prevent pulmonary emboli in trauma patients. Arch Sur 2002; 137: 521–527. 25 Still J, Friedman B, Furman S et al. Experience with the insertion of vena caval filters in acutely burned patients. Am Surg 2000; 66: 277–279. 26 Webb LX, Rush PT, Fuller SB. Meredith JW. Greenfield filter prophylaxis of pulmonary embolism in patients undergoing surgery for acetabular fracture. J Orthop Trauma 1992; 6: 139–145. 27 Golueke PJ, Garrett WV, Thompson JE, Smith BL, Talkington CM. Interruption of the vena cava by means of the Greenfield filter: expanding the indications. Surgery 1988; 103: 111–117. 28 Fink JA, Jones BT. The Greenfield filter as the primary means of therapy in venous thromboembolic disease. Surg Gyenecol Obstet 1991; 172: 253–256. 29 Girard P, Stern J-B, Parent F. Medical literature and vena cava filters so far so weak. Chest 2002; 122: 963–967. 30 Bergqvist D. The role of venal caval interruption in patients with venous thromboembolism. Prog Cardiovasc Dis 1994; 37: 25–37. 31 Greenfield LJ, Tauscher JR, Marx MV. Evaluation of a new percutaneous stainless steel Greenfield filter by intravascular ultrasonography. Surgery 1991; 109: 722–729. 32 Murphy TP, Dorfman GS, Yedlicka JW et al. LGM vena cava filter. Objective evaluation of early results. J Vasc Interv Radiol 1991; 2: 107–115.
PART IV
Prevention and Treatment of DVT and PE
33 Dorfman GS, Cronan JJ, Paolella LP et al. Iatrogenic changes at the venotomy site after percutaneous placement of the Greenfield filter. Radiology 1989; 173: 159– 162. 34 Mewissen MW, Erickson SJ, Foley WD et al. Thrombosis at venous insertion sites after inferior vena caval filter placement. Radiology 1989; 173: 155–157. 35 Kantor A, Glanz S, Gordon D, Sclafani SJ. Percutaneous insertion of the Kimray–Greenfield filter: incidence of femoral vein thrombosis. Am J Roentgenol 1987; 149: 1065–1066. 36 Haire WD. Vena caval filters for the prevention of pulmonary embolism. N Engl J Med 1998; 338: 463–464. 37 Kucher N, Rossi E, De Rosa M, Goldhaber SZ. Massive pulmonary embolism. Circulation 2006; 113: 577– 582. 38 McMurtry AL, Owings JT, Anderson JT, Battistella FD, Gosselin R. Increased use of prophylactic vena cava filters in trauma patients failed to decrease overall incidence of pulmonary embolism. J Am Coll Surg 1999; 189: 314– 320. 39 White RH, Zhou H, Kim J, Romano PS. A populationbased study of the effectiveness of inferior vena cava filter use among patients with venous thromboembolism. Arch Intern Med 2000; 160: 2033–2041. 40 Decousus H, Leizorovicz A, Parent F et al. A clinical trial of vena caval filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. N Engl J Med 1998; 338: 409–415. 41 Walsh DB, Birkmeyer JD, Barrett JA, Kniffin WD, Cronenwett JL, Baron JA. Use of inferior vena cava filters in the Medicare population. Ann Vasc Surg 1995; 9: 483– 487. 42 Stein PD, Kayali F, Olson RE. Twenty-one-year trends in the use of inferior vena cava filters. Arch Intern Med 2004; 164: 1541–1545. 43 Athanasoulis CA, Kaufman JA, Halpern EF, Waltman AC, Geller SC, Fan C-M. Inferior vena caval filters: review of a 26-year single-center clinical experience. Radiology 2000; 216: 54–66. 44 Spence LD, Gironta MG, Malde HM, Mickolick CT, Geisinger MA, Dolmatch BL. Acute upper extremity deep venous thrombosis: safety and effectiveness of superior vena caval filters. Radiology 1999; 210: 53–58. 45 Ascher E, Hingorani A, Mazzariol F, Jacob T, Yorkovich W, Gade P. Clinical experience with superior vena caval Greenfield filters. J Endovasc Surg 1999; 6: 365–369. 46 Ascer E, Gennaro M, Lorensen E, Pollina RM. Superior vena caval Greenfield filters: indications, techniques, and results. J Vasc Surg 1996; 23: 498–503. 47 Magnant JG, Walsh DB, Juravsky LI, Cronenwett JL. Current use of inferior vena cava filters. J Vasc Surg 1992; 16: 701–706.
BLUK077-Stein
April 6, 2007
CHAPTER 91
17:5
Inferior vena cava filters
48 Stein PD, Alnas M, Skaf E et al. Outcome and complications of retrievable inferior vena cava filters. Am J Cardiol 2004; 94: 1090–1093. 49 Asch M. Initial experience in humans with a new retrievable inferior vena cava filter. Radiology 2002; 225: 835– 844. 50 Bovyn G, Gory P, Reynaud P, Ricco JB. The Tempofilter: a multicenter study of a new temporary caval filter implantable for up to six weeks. Ann Vasc Surg 1997; 11: 520–528. 51 Millward S, Bhargava A, Aquino J et al. G¨unther Tulip filter: preliminary clinical experience with retrieval. J Vasc Interv Radiol 2000; 11: 75–82. 52 Millward S, Bormanis J, Burbridge B, Markman S, Peterson R. Preliminary clinical experience with the G¨unther temporary inferior vena cava filter. J Vasc Interv Radiol 1994; 5: 863–868. 53 Neuerburg J, G¨unther R, Vorwerk D et al. Results of a multicenter study of the retrievable Tulip vena cava filter: early clinical experience. Cardiovasc Intervent Radiol 1997; 20: 10–16. 54 Wicky S, Doenz F, Meuwly J, Portier F, Schnyder P, Denys A. Clinical experience with retrievable G¨unther Tulip vena cava filters. J Endovasc Ther 2003; 10: 994–1000. 55 Ferraro F, D’ignazio N, Matarazzo A, Rusciano G, Iannuzzi M, Belluomo Anello C. Thromboembolism in pregnancy: a new temporary caval filter. Minerva Anestesiol 2001; 67: 381–385. 56 Offner P, Hawkes A, Madayag R, Seale F, Maines C. The role of temporary inferior vena cava filters in critically ill surgical patients. Arch Surg 2003; 138: 591–595. 57 Thery C, Asseman P, Amrouni N et al. Use of a new removable vena cava filter in order to prevent pulmonary embolism in patients submitted to thrombolysis. Eur Heart J 1990; 11: 334–341.
453
58 Darcy M, Cardella J, Hunter D et al. Experience with the Amplatz retrievable vena caval filter. Radiology 1986; 161: 611–614. 59 de Gregorio M, Gamboa P, Gimeno M et al. The G¨unther Tulip retrievable filter: prolonged temporary filtration by repositioning within the inferior vena cava. J Vasc Interv Radiol 2003; 14: 1259–1265. 60 Ferrer M, Blanquer J, Mu˜noz J, Gil J, Blasco L, Nu˜nez C. Thrombolysis and temporary vena cava filter placement in venous thromboembolic disease. J Interv Radiol 1999; 14: 152–156. 61 Lorch H, Welger D, Wagner V et al. Current practice of temporary vena cava filter insertion: a multicenter registry. J Vasc Interv Radiol 2000; 11: 83–88. 62 Lorch H, Zwaan M, Siemens H, Wagner T, Kagel C, Weiss H. Temporary vena cava filters and ultrahigh streptokinase thrombolysis therapy: a clinical study. Cardiovasc Intervent Radiol 2000; 23: 273–278. 63 Yamagami T, Kato T, Iida S, Tanaka O, Nishimura T. Retrievable vena cava filter placement during treatment for deep venous thrombosis. Br J Radiol 2003; 76: 712–718. 64 Zwaan M, Lorch H, Kulke C et al. Clinical experience with temporary vena caval filters. J Vasc Interv Radiol 1998; 9: 594–601. 65 Linsenmaier U, Rieger J, Schenk F, Rock C, Mangel E, Pfeifer K. Indications, management, and complications of temporary inferior vena cava filters. Cardiovasc Intervent Radiol 1998; 21: 464–469. 66 Ponchon M, Goffette P, Hainaut P. Temporary vena caval filtration preliminary clinical experience with removable vena caval filters. Acta Clin Belg 1999; 54: 223–228. 67 Allen TL, Carter JL, Morris BJ, Harker CP, Stevens MH. Retrievable vena cava filters in trauma patients for highrisk prophylaxis and prevention of pulmonary embolism. Am J Surg 2005; 189: 656–661.
BLUK077-Stein
March 28, 2007
92
16:2
CHAPTER 92
Catheter-tip embolectomy in the management of acute massive pulmonary embolism
Introduction Catheter-tip embolectomy has evolved since the first suction embolectomy catheter used by Greenfield and associates [1] and the first catheter-tip fragmentation device tested by Stein and associates in dogs [2]. Currently, there are three categories of catheter interventional techniques for removing pulmonary em-
boli (PE) or for decreasing clot burden: (1) aspiration thrombectomy, (2) fragmentation, and (3) rheolytic thrombectomy [3–5]. Although originally it was thought that catheter embolectomy or fragmentation could substitute for thrombolytic therapy, limited data on the use of such catheters alone suggests that they are generally used as an adjunct to thrombolytic therapy (Table 92.1) [5]. All the devices appear to be useful in
Table 92.1 Pooled data of clinical success and major bleeding in patients who underwent catheter embolectomy. Technique and catheter type
Total patients
Clinical success* n/N (%)
Major bleeding (%)
Aspiration technique Greenfield/PEC catheter No lytics: steel cap [6–8]
27
25/27 (93)
5 (19)
Plastic cap [8, 9–12]
62
46/62 (74)
1 (2)
Systemic lytics [11, 13]
9
9/9 (100)
0 (0)
Local lytics [10]
9
9/9 (100)
0 (0)
Systemic and local lytics [14]
1
1/1 (100)
0 (0)
Fragmentation Pigtail/Angio catheters No lytics [15, 16]
3
2/3 (67)
0 (0)
21
15/21 (71)
0 (0)
121
115/121 (95)
30
24/30 (80)
3 (10)
No lytics [28, 29]
8
7/8 (88)
1 (13)
Local lytics [28, 29]
6
6/6 (100)
0 (0)
Systemic lytics [17] Local lytics [11, 15, 18–25, 41] Systemic and local lytics [16, 26, 27]
2 (2)
Catheter-tip impeller
Rheolytic technique Rheolytic angiojet catheter No lytics [4, 30, 31]
8
6/8 (75)
0 (0)
Local lytics [32, 33]
23
20/23 (87)
0 (0)
12
11/12 (92)
Hydrolyser catheter Local lytics [34, 35] Systemic and local lytics [36]
8
8/8 (100)
* Relief of acute signs and symptoms resulting from a partial or complete removal of thrombi. Reprinted from Skaf et al. [5], with permission from Elsevier.
454
0 (0) 0 (0)
BLUK077-Stein
March 28, 2007
CHAPTER 92
16:2
455
Catheter-tip embolectomy in acute PE
the management of acute massive PE. The limited data suggest that manual fragmentation with ordinary angiographic catheters, usually the pigtail catheter, may be as effective as fragmentation with technically advanced catheter-tip devices.
Aspiration thrombectomy The aspiration thrombectomy technique uses sustained syringe suction applied to a vacuum suction cup at the tip of the catheter to securely hold the embolus while it is removed via the venotomy [6–8, 37–40]. The Greenfield catheter and similar catheters use this technique [37–41]. Relief of acute signs and symptoms resulting from a partial or complete removal of thrombi with the steel cup, earlier version of the Greenfield catheter, was shown in 25 of 27 patients (93%) (Table 92.1) [6–8]. Major bleeding, however, occurred in 19%. With the newer, more maneuverable plastic cup suction catheter, clinical success was reported in a somewhat lower proportion of patients, 46 of 62 (74%), but rate of major bleeding was also lower, 2% (Table 92.1) [8, 9, 12]. When thrombolytic agents were administered in combination with the use of aspiration catheters, the success rate among 19 patients was 100% [10, 11, 13, 14] and no major bleeding was reported in these patients.
Fragmentation with standard angiographic catheters Fragmentation of PE has been done by manually breaking the clot with an angiographic catheter [37–41]. Pigtail catheters have been used more often than other catheters [37–41]. The pigtail catheter is advanced to the site of the thrombus and then while advancing and withdrawing, the catheter is rotated to fragment the thrombus [37–41]. Fragmentation of proximal emboli by manual manipulation of angiographic catheters, in combination with thrombolytic agents, appears to have been useful. With the administration of local and/or systemic thrombolytic agents, relief of acute signs and symptoms resulting from a partial or complete removal of thrombi was reported in 154 of 172 patients (90%) with major bleeding in 5 (3%) (Table 92.1) [9–11, 15–27, 41].
Fragmentation with catheter-tip impellar The Amplatz catheter has an impeller housed in a capsule at the tip, driven by a compact air turbine that homogenizes the thrombus by maceration and fragmentation (Figure 92.1). A metal capsule protects the vessel from the rotating impellar [37–41]. Relief of acute
Blood Flow
Homogenized Thrombus
Impellar Protective Capsule
Figure 92.1 Diagram (not to scale) showing the mechanism of action of the Amplatz catheter. An impeller housed in a capsule at the tip of the catheter homogenizes
the thrombus. (Reproduced from Skaf et al. [5], with permission from Elsevier Inc.)
BLUK077-Stein
March 28, 2007
16:2
456
PART III
signs and symptoms resulting from a partial or complete removal of thrombi was reported in 7 of 8 (88%) with 1 case of major bleeding (Table 92.1) [28, 29]. When thrombolytic agents were administered in combination with the Amplatz catheter, such clinical success was reported in 6 of 6 (100%) (Table 92.1) [28, 29] and none of these patients had major bleeding.
Rheolytic technique The rheolytic thrombectomy technique uses a dual lumen catheter: a small lumen is for delivery of pulsatile pressurized saline and a larger effluent or exhaust lumen drains the thrombus. High-velocity saline jets are injected or pumped toward the effluent lumen. This produces an area of low pressure at the tip of the catheter (Venturieffect) (Figure 92.2), thereby fragmenting the thrombus and allowing it to be evacuated throughthe effluent lumen [37–41]. The AngioJet, Hydrolyser, and Oasis catheters are examples of catheters demonstrating this mechanism of action [37–41].
Low Pressure (Venturi Effect)
Pressurized Saline
Exhaust Lumen (Suction)
Figure 92.2 Diagram (not to scale) showing the mechanism of action of a double lumen catheter-tip device that functions on the basis of the Venturi effect (rheolytic catheter). The narrow injection lumen ends in a hairpin loop at the distal end of the catheter to direct a single high-speed retrograde fluid jet into the larger evacuation lumen. As the jet crosses the side hole, the resulting Venturi effect fragments the adjacent thrombus and entrains the debris into the evacuation lumen, which is connected to a vacuum bag. (Reproduced from Skaf et al. [5], with permission from Elsevier Inc.)
Prevention and Treatment of DVT and PE
With a rheolytic AngioJet catheter relief of acute signs and symptoms resulting from a partial or complete removal of thrombi was reported in 6 of 8 patients (75%), none of whom had major bleeding (Table 92.1) [4, 30, 31]. When used in combination with the local injection of local thrombolytic agents, 20 of 23 (87%) showed such clinical success (Table 92.1) [32, 33]. With the hydrolyser catheter used in combination with local and/or systemic thrombolytic agents, relief of acute signs and symptoms resulting from a partial or complete removal of thrombi was observed in 19 of 20 patients (95%), none of whom had major bleeding (Table 92.1) [34–36].
References 1 Greenfield LJ, Kimmell GO, McCurdy WC, III. Transvenous removal of pulmonary emboli by vacuum-cup catheter technique. J Surg Res 1969; 9: 347–352. 2 Stein PD, Sabbah HN, Basha MA, Popovich J, Jr, Kensey KR, Nash JE. Mechanical disruption of pulmonary emboli in dogs with a flexible rotating-tip catheter (Kensey catheter). Chest 1990; 98: 994–998. 3 Goldhaber SZ. Integration of catheter thrombectomy into our armamentarium to treat acute pulmonary embolism. Chest 1998; 114: 1237–1238. 4 Koning R, Cribier A, Gerber L et al. A new treatment for severe pulmonary embolism: percutaneous rheolytic thrombectomy. Circulation 1997; 96: 2498–2500. 5 Skaf E, Beemath A, Siddiqui T, et al. Catheter-tip embolectomy in the management of acute massive pulmonary embolism: a systematic review. Am J Cardiol (In press). 6 Greenfield LJ, Bruce TA, Nichols NB. Transvenous pulmonary embolectomy by catheter device. Ann Surg 1971; 174: 881–886. 7 Greenfield LJ, Zocco J. Intraluminal management of acute massive pulmonary thromboembolism. J Thorac Cardiovasc Surg 1979; 77: 402–410. 8 Greenfield LJ, Proctor MC, Williams DM, Wakefield TW. Long-term experience with transvenous catheter pulmonary embolectomy. J Vasc Surg 1993; 18: 450–457. 9 Timsit JF, Reynaud P, Meyer G, Sors H. Pulmonary embolectomy by catheter device in massive pulmonary embolism. Chest 1991; 100: 655–658. 10 Tajima H, Murata S, Kumazaki T et al. Manual aspiration thrombectomy with a standard PTCA guiding catheter for treatment of acute massive pulmonary thromboembolism. Radiat Med 2004; 22: 168–172. 11 Moore JH, Jr, Koolpe HA, Carabasi RA, Yang SL, Jarrell BE. Transvenous catheter pulmonary embolectomy. Arch Surg 1985; 120: 1372–1375.
BLUK077-Stein
March 28, 2007
CHAPTER 92
16:2
457
Catheter-tip embolectomy in acute PE
12 Scoggins WG, Greenfield LJ. Transvenous pulmonary embolectomy for acute massive pulmonary embolism. Chest 1977; 71: 213–216. 13 Hiramatsu S, Ogihara A, Kitano Y et al. Clinical outcome of catheter fragmentation and aspiration therapy in patients with acute pulmonary embolism. J Cardiol 1999; 34: 71–78. 14 Nakasaki Y, Tsujiyama S, Higo M et al. A case of acute massive pulmonary embolism successfully treated with transvenous pulmonary embolectomy by catheter. Kokyu To Junkan 1989; 37: 1363–1366. 15 Brady AJB, Crake T, Oakley CM. Percutaneus catheter fragmentation and distal dispersion of proximal pulmonary embolus. Lancet 1991; 338: 1186–1189. 16 Schmitz-rode T, Janssens U, Schild HH, Basche S, Hanrath P, Gunther RW. Fragmentation of massive pulmonary embolism using a pigtail rotation catheter. Chest 1998; 114: 1427–1436. 17 Prokubovski VI, Kapranov SA, Bobrov BY. Endovascular rotary fragmentation in the treatment of massive pulmonary thromboembolism. Sosud Khir 2003; 9: 31–39. 18 De Gregorio MA, Gimeno MJ, Mainar A et al. Mechanical and enzymatic thrombolysis for massive pulmonary embolism. J Vasc Interv Radiol 2002; 13: 163– 169. 19 Fava M, Loyola S, Flores P, Huete I. Mechanical fragmentation and pharmacologic thrombolysis in massive pulmonary embolism. J Vasc Interv Radiol 1997; 8: 261– 266. 20 Schmitz-rode T, Janssens U, Hanrath P, Gunther RW. Fragmentation of massive pulmonary embolism using a pigtail rotation catheter: possible complication. Eur Radiol 2001; 11: 2047–2049. 21 Stock KW, Jacob AL, Schnabel KJ, Bongartz G, Steinbrich W. Massive pulmonary embolism: treatment with thrombus fragmentation and local fibrinolysis with recombinant human-tissue plasminogen activator. Cardiovasc Intervent Radiol 1997; 20: 364–368. 22 Murphy JM, Mulvihill N, Mulcahy D, Foley B, Smiddy P, Molloy MP. Percutaneous catheter and guidewire fragmentation with local administration of recombinant tissue plasminogen activator as a treatment for massive pulmonary embolism. Eur Radiol 1999; 9: 959– 964. 23 Rafique M, Middlemost S, Skoularigis J, Sareli P. Simultaneous mechanical clot fragmentation and pharmacologic thrombolysis in acute massive pulmonary embolism. Am J Cardiol 1992; 69: 427–430. 24 Tajima H, Murata S, Kumazaki T et al. Hybrid treatment of acute massive pulmonary thromboembolism: mechanical fragmentation with a modified rotating pigtail catheter, local fibrinolytic therapy, and clot aspiration followed by
25
26
27
28
29
30
31
32
33
34
35
36
37
systemic fibrinolytic therapy. Am J Roentgenol 2004; 183: 589–595. Manthey J, Frohlich G, Mautner JP, Munderloh KH, Zimmermann R. Mechanical recanalization and local thrombolysis in a patient with fulminant pulmonary embolism and craniocerebral trauma. Anasthesiol Intensivmed Notfallmed Schmerzther 1994; 29: 446–449. Schmitz-rode T, Janssen U, Duda SH, Erley CM, Gunther RW. Massive pulmonary embolism: percutaneous emergency treatment by pigtail rotation catheter. J Am Coll Cardiol 2000; 36: 375–380. Sigmund M, Rubart M, Vom Dahl J, Uebis R, Hanrath P. Successful treatment of massive pulmonary embolism by combined mechanical and thrombolytic therapy. J Interv Cardiol 1991; 4: 63–68. Muller-Hulsbeck S, Brossmann J, Jahnke T et al. Mechanical thrombectomy of major and massive pulmonary embolism with use of the Amplatz thrombectomy device. Invest Radiol 2001; 36: 317–322. Uflacker R, Strange C, Vujic I. Massive pulmonary embolism: preliminary results of treatment with the Amplatz thrombectomy device. J Vasc Interv Radiol 1996; 7: 519– 528. Voigtlander T, Rupprecht HJ, Nowak B et al. Clinical application of a new rheolytic thrombectomy catheter system for massive pulmonary embolism. Catheter Cardiovasc Interv 1999; 47: 91–96. Chiam P, Kwok V, Johan BA, Chan C. Major pulmonary embolism treated with a rheolytic thrombectomy catheter. Singapore Med J 2005; 46: 479–482. Zeni PT, Jr, Blank BG, Peeler DW. Use of rheolytic thrombectomy in treatment of acute massive pulmonary embolism. J Vasc Interv Radiol 2003; 14: 1511–1515. Siablis D, Karnabatidis D, Katsanos K, Kagadis GC, Zabakis P, Hahalis G. AngioJet rheolytic thrombectomy versus local intrapulmonary thrombolysis in massive pulmonary embolism: a retrospective data analysis. J Endovasc Ther 2005; 12: 206–214. Fava M, Loyola S, Huete I. Massive pulmonary embolism: treatment with the hydrolyser thrombectomy catheter. J Vasc Interv Radiol 2000; 11: 1159–1164. Michalis LK, Tsetis DK, Rees MR. Case report. Percutaneous removal of pulmonary artery thrombus in a patient with massive pulmonary embolism using the hydrolyser catheter: the first human experience. Clin Radiol 1997; 52: 158–161. Reekers JA, Baarslag HJ, Koolen MG, Van Delden O, van Beek EJ. Mechanical thrombectomy for early treatment of massive pulmonary embolism. Cardiovasc Intervent Radiol 2003; 26: 246–250. Fava M, Loyola S. Applications of percutaneous mechanical thrombectomy in pulmonary embolism. Tech Vasc Interv Radiol 2003; 6: 53–58.
BLUK077-Stein
March 28, 2007
16:2
458
38 Meyer G, Koning R, Sors H. Transvenous catheter embolectomy. Semin Vasc Med 2001; 1: 247–252. 39 Cho KJ, Dasika NL. Catheter technique for pulmonary embolectomy or thrombofragmentation. Semin Vasc Surg 2000; 13: 221–235. 40 Sharafuddin MJ, Hicks ME. Current status of percutaneous mechanical thrombectomy. Part II: Devices and
PART III
Prevention and Treatment of DVT and PE
mechanisms of action. J Vasc Interv Radiol 1998; 9: 15– 31. 41 Wong PS, Singh SP, Watson RD, Lip GY. Management of pulmonary thromboembolism using catheter manipulation: a report of four cases and review of the literature. Postgrad Med J 1999; 75: 737– 741.
BLUK077-Stein
March 28, 2007
93
16:3
CHAPTER 93
Pulmonary embolectomy
Introduction Pulmonary embolectomy is suggested in selected highly compromised patients who are unable to receive thrombolytic therapy or whose critical status does not allow sufficient time to infuse thrombolytic agents [1]. For most patients with pulmonary embolism (PE), pulmonary embolectomy is not indicated [1]. The following criteria were recommended [1]:
1 massive PE, angiographically documented, if possible; 2 hemodynamic instability (shock) despite heparin therapy and resuscitative efforts; and 3 failure of thrombolytic therapy or a contraindication to its use. Case series that included data from 1961 to 2006 were reviewed [2]. Mortality ranged from 6 to 64% [5–46] (Table 93.1). Pooled data showed an average mortality of 389 of 1300 (30%) [5–46]. Among case series that included data on patients operated from 1961 to 1984, the average mortality was 338 of 1047 (32%) [5–28, 30–33, 36, 37] and in case series that
included data on patients operated from 1985 to 2006, 51 of 253 (20%) [29, 34, 35, 38–46]. The most frequent indication for pulmonary embolectomy was hemodynamic instability, which was reported in 74% of patients [5–46]. Other indications were cardiac arrest (32%) and contraindications to thrombolytic therapy (19%) [5–46] (Table 93.1). Among patients who suffered a cardiac arrest prior to pulmonary embolectomy, the operative mortality was 188 of 317 (59%) compared with a mortality of 201 of 983 (20%) patients who did not have a preoperative cardiac arrest [5–13, 15, 17, 19, 23–25, 27, 28, 30–33, 35–37, 39, 41–45]. Mortality from pulmonary embolectomy showed a loose correlation with preoperative cardiac arrest (Figure 93.1). Only a few case series reported mortalities ≤10% [7, 40, 42, 44, 45]. These case series included only a small proportion of patients (0–18%) who suffered a cardiac arrest prior to pulmonary embolectomy [7, 42, 44, 45], whereas on average, 32% of patients who underwent pulmonary embolectomy had a preoperative cardiac arrest [5–13, 15, 17, 19, 23–25, 27, 28, 30–33, 35–37, 39, 41–45]. Among patients in studies that reported a
75
Figure 93.1 Mortality following pulmonary embolectomy shown in relation to the percentage of patients who had preoperative cardiac arrest. A loose but significant correlation was shown (P = 0.0012). (Reprinted from Stein et al. [2], with permission from Elsevier.)
Mortality (%)
50
25
0 0
25
50
75
100
Cardiac arrest (%)
459
460 30 13
Meyns [30]
Leitz [31]
15/44 (34%)
44 11
Bauer [28]
Biglioli [29] 7/13 (54%)
12/30 (40%)
—
—
23/134 (17%)
24/96 (25%)
25/71 (35%)
—
16/27 (59%)
71
Gray [23]
16
25
Lund [22]
—
—
27
29
Stalpaert [21]
Schmid [27]
23
Jaumin [20]
19/55 (35%)
—
Boulafendis [26]
55
Clarke [19]
96
28
Savelyev [18]
0/17 (0%)
—
134
17
Soyer [17]
Meyer [24]
46
Satter [16]
34/39 (87%)
—
10/20 (50%)
5/11 (45%)
9/17 (53%)
2/17 (12%)
2/11 (18%)
5/26 (19%)
3/11 (27%)
Kieny [25]
39
Mattox [15]
23
11 33
De Weese [10]
Miller [11]
Bottzauw [14]
4/24 (17%)
17
Reul [9]
24
17
Berger [8]
20
11
Heimbecker [7]
Tschirkov [12]
26
Clarke [6]
with preoperative
—
6/12 (50%)
—
7/15 (47%)
8/16 (50%)
—
11/23 (48%)
14/24 (58%)
16/25 (64%)
—
—
—
16/19 (84%)
—
0 (0%)
—
22/34 (65%)
—
5/10 (50%)
2/4 (50%)
6/6 (100%)
5/5 (100%)
2/9 (22%)
2/2 (100%)
0/2 (0%)
5/5 (100%)
3/3 (100%)
cardiac arrest
Total
6/13 (46%)
6/30 (20%)
3/11 (27%)
9/44 (20%)
12/27 (44%)
5/16 (31%)
21/134 (16%)
36/96 (38%)
21/71 (30%)
5/25 (20%)
10/29 (34%)
7/23 (30%)
20/55 (36%)
12/28 (43%)
5/17 (29%)
17/46 (37%)
22/39 (56%)
6/23 (26%)
8/20 (40%)
7/24 (29%)
7/33 (21%)
7/11 (64%)
6/17 (35%)
4/17 (24%)
1/11 (9%)
13/26 (50%)
6/11 (55%)
mortality
13/13 (100%)
7/44 (16%)
—
—
3/134 (2%)
10/96 (10%)
16/71 (23%)
—
—
—
—
—
2/17 (12%)
—
—
—
—
—
—
—
—
—
—
—
—
to thrombolytics
Contraindications
7/13 (54%)
12/30 (40%)
—
15/44 (34%)
16/27 (59%)
—
23/134 (17%)
24/96 (25%)
25/71 (35%)
—
—
—
19/55 (35%)
—
0/17 (0%)
—
34/39 (87%)
—
10/20 (50%)
4/24 (17%)
6/33 (18%)
5/11 (45%)
9/17 (53%)
2/17 (12%)
2/11 (18%)
5/26 (19%)
3/11 (27%)
arrest
Cardiac
Indications for embolectomy Hemodynamic
—
30/30 (100%)
3/11 (27%)
28/44 (64%)
—
—
57/134 (43%)
43/96 (45%)
71/71 (100%)
—
—
23/23 (100%)
35/55 (64%)
—
6/17 (35%)
—
39/39 (100%)
23/23 (100%)
—
24/24 (100%)
33/33 (100%)
11/11 (100%)
17/17 (100%)
17/17 (100%)
11/11 (100%)
21/26 (81%)
—
instability
March 28, 2007
Glassford [13]
6/33 (18%)
11
Cooley [5]
Preoperative
(Ref.)
cardiac arrest
No. of patients
First author
Mortality in patients
Table 93.1 Indications and outcome of patients undergoing pulmonary embolectomy.
BLUK077-Stein
16:3
14/36 (39%)
461 1300
Total
330/1022 (32%)
—
0/14 (0%)
2/13 (15%)
0/15 (0%)
—
Reprinted from Stein et al. [2], with permission from Elsevier.
18
Sukhija [46]
6/47 (13%)
47 14
13 11
Yalamanchili [42]
Dauphine [43]
Leacche [44]
15
Ando [41]
Meneveau [45]
4/11 (36%)
29
19/40 (48%)
Aklog [40]
—
14/41 (34%)
40
36 41
Doerge [36]
Doerge [37]
13/25 (52%)
—
17
25
Jakob [35]
6/34 (18%) 31/50 (62%)
Ullmann [39]
13
Gulba [34]
Preoperative cardiac arrest
Chartier [38]
34 50
Stulz [33]
patients
Laas [32]
No. of
(Ref.)
188/317 (59%)
—
0/0 (0%)
2/6 (33%)
3/4 (75%)
1/2 (50%)
0 (0%)
—
12/19 (63%)
—
9/14 (64%)
8/14 (57%)
4/13 (31%)
—
19/31 (61%)
—
cardiac arrest
with preoperative
Mortality in patients Total
389/1300 (30%)
2/18 (11%)
1/14 (7%)
3/47 (6%)
3/11 (27%)
1/13 (8%)
4/15 (27%)
3/29 (10%)
14/40 (35%)
8/17 (47%)
12/41 (29%)
9/36 (25%)
6/25 (24%)
3/13 (23%)
23/50 (46%)
15/34 (44%)
mortality
104/553 (19%)
—
0/14 (0%)
21/47 (45%)
5/11 (45%)
—
—
12/29 (41%)
—
—
10/41 (24%)
5/36 (14%)
—
—
—
—
to thrombolytics
Contraindications
330/1022(32%)
—
0/14 (0%)
6/47 (13%)
4/11 (36%)
2/13 (15%)
0/15 (0%)
—
19/40 (48%)
—
14/41 (34%)
14/36 (39%)
13/25 (52%)
—
31/50 (62%)
6/34 (18%)
arrest
Cardiac
Indications for embolectomy Hemodynamic
772/1039 (74%)
18/18 (100%)
14/14 (100%)
12/47 (26%)
11/11 (100%)
13/13 (100%)
12/15 (80%)
—
32/40 (80%)
—
41/41 (100%)
36/36 (100%)
25/25 (100%)
13/13 (100%)
19/50 (38%)
34/34 (100%)
instability
March 28, 2007
First author
Table 93.1 (Continued )
BLUK077-Stein
16:3
BLUK077-Stein
March 28, 2007
16:3
462
mortality ≤10%, 59% of patients had hemodynamic instability [7, 42, 44, 45], whereas on average, 74% of patients had hemodynamic instability [5–46]. However, mortality did not correlate with the proportion of patients who had preoperative instability. In a prospective investigation in which patients who failed thrombolytic therapy (defined as persistent clinical instability and right ventricular dysfunction) physicians selected to treat the patients either with a second course of thrombolytic therapy or with pulmonary embolectomy [45]. Mortality tended to be lower among those who underwent embolectomy, 1 of 14 (7%) versus 10 of 26 (38%) who had a second course of thrombolytic therapy [45]. It appears that, in spite of a generally high mortality in patients undergoing pulmonary embolectomy, it may have lifesaving potential in some instances.
References 1 Buller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126(suppl): 401S–428S. 2 Stein PD, Alnas M, Beemath A, Patel NR. Pulmonary embolectomy for acute pulmonary embolism: a systematic review. Am J Cardiol (In press). 3 Bossuyt PM, Reitsma JB, Bruns DE et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med 2003; 138: W1–W12. 4 Lijmer JG, Mol BW, Heisterkamp S et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 1999; 282: 1061–1066. 5 Cooley DA, Beall AC, Jr. Embolectomy for acute massive pulmonary embolism. Surg Gynecol Obstet 1968; 126: 805–810. 6 Clarke DB, Abrams LD. Pulmonary embolectomy with venous inflow-occlusion. Lancet 1972; 1: 767–769. 7 Heimbecker RO, Keon WJ, Richards KU. Massive pulmonary embolism: a new look at surgical management. Arch Surg 1973; 107: 740–746. 8 Berger RL. Pulmonary embolectomy with preoperative circulatory support. Ann Thorac Surg 1973; 16: 217–227. 9 Reul GJ, Beall AC, Jr. Emergency pulmonary embolectomy for massive pulmonary embolism. Circulation 1974; 50: 236–241. 10 De Weese JA. The role of pulmonary embolectomy in venous thromboembolism. J Cardiovasc Surg 1976; 17: 348–353.
PART IV
Prevention and Treatment of DVT and PE
11 Miller GA, Hall RJ, Paneth M. Pulmonary embolectomy, heparin, and streptokinase: their place in the treatment of acute massive pulmonary embolism. Am Heart J 1977; 93: 568–574. 12 Tschirkov A, Krause E, Elert O, Satter P. Surgical management of massive pulmonary embolism. J Thorac Cardiovasc Surg 1978; 75: 730–733. 13 Glassford DM, Jr, Alford WC, Jr, Burrus GR, Stoney WS, Thomas CS, Jr, Pulmonary embolectomy. Ann Thorac Surg 1981; 32: 28–32. 14 Bottzauw J, Vejlsted H, Albrechtsen O. Pulmonary embolectomy using extracorporeal circulation. Thorac Cardiovasc Surg 1981; 29: 320–322. 15 Mattox KL, Feldtman RW, Beall AC, Jr, DeBakey ME. Pulmonary embolectomy for acute massive pulmonary embolism. Ann Surg 1982; 195: 726–731. 16 Satter P. Pulmonary embolectomy with the aid of extracorporeal circulation. Thorac Cardiovasc Surg 1982; 30: 31–35. 17 Soyer R, Brunet AP, Redonnet M, Borg JY, Hubscher C, Letac B. Follow-up of surgically treated patients with massive pulmonary embolism—with reference to 12 operated patients. Thorac Cardiovasc Surg 1982; 30: 103– 108. 18 Savelyev VS. Massive pulmonary embolism: embolectomy or thrombolysis? Int Angiol 1985; 4: 137–140. 19 Clarke DB, Abrams LD. Pulmonary embolectomy: a 25 year experience. J Thorac Cardiovasc Surg 1986; 92: 442– 445. 20 Jaumin P, Moriau M, el Gariani A et al. Pulmonary embolectomy: clinical experience. Acta Chir Belg 1986; 86: 123–125. 21 Stalpaert G, Suy R, Daenen W, Flameng W et al. Surgical treatment of acute, massive lung embolism: results and follow-up. Acta Chir Belg 1986; 86: 118–122. 22 Lund O, Nielsen TT, Schifter S, Roenne K. Treatment of pulmonary embolism with full-dose heparin, streptokinase or embolectomy—results and indications. Thorac Cardiovasc Surg 1986; 34: 240–246. 23 Gray HH, Morgan JM, Paneth M, Miller GA. Pulmonary embolectomy for acute massive pulmonary embolism: an analysis of 71 cases. Br Heart J 1988; 60: 196–200. 24 Meyer G, Tamisier D, Sors H et al. Pulmonary embolectomy: a 20-year experience at one center. Ann Thorac Surg 1991; 51: 232–236. 25 Kieny R, Charpentier A, Kieny MT. What is the place of pulmonary embolectomy today? J Cardiovasc Surg 1991; 32: 549–554. 26 Boulafendis D, Bastounis E, Panayiotopoulos YP, Papalambros EL. Pulmonary embolectomy (answered and unanswered questions). Int Angiol 1991; 10: 187–194. 27 Schmid C, Zietlow S, Wagner TO, Laas J, Borst HG. Fulminant pulmonary embolism: symptoms, diagnostics,
BLUK077-Stein
March 28, 2007
CHAPTER 93
28
29
30
31
32
33
34
35
36
37
16:3
Pulmonary embolectomy
operative technique, and results. Ann Thorac Surg 1991; 52: 1102–1105; discussion 1105–1107. Bauer EP, Laske A, von Segesser LK, Carrel T, Turina MI. Early and late results after surgery for massive pulmonary embolism. Thorac Cardiovasc Surg 1991; 39: 353–356. Biglioli P, Alamanni F, Spirito R, Arena V. From deep venous thrombosis to pulmonary embolism. Cardiologia 1991; 36: 195–201. Meyns B, Sergeant P, Flameng W, Daenen W. Surgery for massive pulmonary embolism. Acta Cardiol 1992; 47: 487–493. Leitz KH, Tsilimingas N, Reichert K. Acute pulmonary embolism—are vena cava interruption procedures and extracorporeal circulation always necessary? Langenbecks Arch Chir Suppl Kongressbd 1992: 511–517. Laas J, Schmid C, Albes JM, Borst HG. Surgical aspects of fulminant pulmonary embolism. Z Kardiol 1993; 82: 25–28. Stulz P, Schlapfer R, Feer R, Habicht J, Gradel E. Decision making in the surgical treatment of massive pulmonary embolism. Eur J Cardiothorac Surg 1994; 8: 188–193. Gulba DC, Schmid C, Borst HG, Lichtlen P, Dietz R, Luft FC. Medical compared with surgical treatment for massive pulmonary embolism. Lancet 1994; 343: 576–577. Jakob H, Vahl C, Lange R, Micek M, Tanzeem A, Hagl S. Modified surgical concept for fulminant pulmonary embolism. Eur J Cardiothorac Surg 1995; 9: 557–560. Doerge HC, Schoendube FA, Loeser H, Walter M, Messmer BJ. Pulmonary embolectomy: review of a 15-year experience and role in the age of thrombolytic therapy. Eur J Cardiothorac Surg 1996; 10: 952–957. Doerge HC, Schoendube FA, Voss M, Seipelt R, Messmer BJ, Surgical therapy of fulminant pulmonary embolism: early and late results. Thorac Cardiovasc Surg 1999; 47: 9–13.
463
38 Chartier L, B´era J, Delomez M et al. Free-floating thrombi in the right heart: diagnosis, management, and pronostic indexes in 38 consecutive patients. Circulation 1999; 99: 2779–2783. 39 Ullmann M, Hemmer W, Hannekum A. The urgent pulmonary embolectomy: mechanical resuscitation in the operating theatre determines the outcome. Thorac Cardiovasc Surg 1999; 47: 5–8. 40 Aklog L, Williams CS, Byrne JG, Goldhaber SZ. Acute pulmonary embolectomy: a contemporary approach. Circulation 2002; 105: 1416–1419. 41 Ando M. Surgical treatment for acute massive pulmonary thromboembolism. Nippon Rinsho 2003; 61: 1769– 1774. 42 Yalamanchili K, Fleisher AG, Lehrman SG et al. Open pulmonary embolectomy for treatment of major pulmonary embolism. Ann Thorac Surg 2004; 77: 819– 823. 43 Dauphine C, Omari B. Pulmonary embolectomy for acute massive pulmonary embolism. Ann Thorac Surg 2005; 79: 1240–1244. 44 Leacche M, Unic D, Goldhaber SZ et al. Modern surgical treatment of massive pulmonary embolism: results in 47 consecutive patients after rapid diagnosis and aggressive surgical approach. J Thorac Cardiovasc Surg 2005; 129: 1018–1023. 45 Meneveau N, Seronde MF, Blonde MC et al. Management of unsuccessful thrombolysis in acute massive pulmonary embolism. Chest 2006; 129: 1043–1050. 46 Sukhija R, Aronow WS, Lee J et al. Association of right ventricular dysfunction with in-hospital mortality in patients with acute pulmonary embolism and reduction in mortality in patients with right ventricular dysfunction by pulmonary embolectomy. Am J Cardiol 2005; 95: 695– 696.
BLUK077-Stein
January 10, 2007
94
5:20
CHAPTER 94
Chronic thromboembolic pulmonary hypertension and pulmonary thromboendarterectomy
Chronic thromboembolic pulmonary hypertension is a complication of pulmonary embolism (PE) that is caused by incomplete resolution of the PE [1] (Figure 94.1). If untreated, the pulmonary hypertension progresses to right heart failure and death [1]. It has been estimated that there are 2500 new cases of chronic thromboembolic pulmonary hypertension each year in the United States [2]. The predominant symptom of chronic thromboembolic pulmonary hypertension is unexplained dyspnea on exertion [3]. An asymptomatic period of several months or years between acute PE and dyspnea on exertion from chronic thromboembolic pulmonary hypertension is common [2, 3]. Worsening of chronic thromboembolic pulmonary hypertension may be due to recurrent PE, in situ thrombosis, or remodeling of the small distal vessels pulmonary arteries similar to that seen in primary pulmonary hypertension [2, 4]. The pathophysiological events in the progression of pulmonary hypertension have not been well defined [2]. It seems that remodeling of the small distal vessels pulmonary arteries may play a major role [2, 4]. Conventional pulmonary angiography is the key diagnostic procedure in patients with suspected chronic thromboembolic pulmonary hypertension [3, 5] and is safe when performed by experienced angiographers [6]. Many patterns of thrombus organization and recanalization require experience for proper identification [3]. Contrast-enhanced spiral CT, however, is essential to exclude rare conditions that may present with similar signs and symptoms such as fibrous mediastinitis, mediastinal carcinoma, and pulmonary artery sarcoma [4].
464
Early chronic thromboembolic pulmonary hypertension is important because the surgical mortality in patients who have progressed to dyspnea at rest is substantially greater than among those with less severe symptoms [7]. If the perfusion lung scan remains unchanged for 6–8 weeks after PE, the organized thrombotic residuals persist indefinitely and such patients merit consideration for surgical relief [7]. Pulmonary thromboendarterectomy reduces the pulmonary artery resistance and improves dyspnea and exercise capacity. Surgical thromboendarterectomy bears no resemblance to acute pulmonary embolectomy [3]. The intraluminal material is composed of fibrous tissue that is inseparable from the pulmonary arterial intima and, therefore, inaccessible to thrombectomy or dilatation [4]. A true endarterectomy is required [4]. The procedure should be performed only in experienced centers [5]. In 2003, it was estimated that there were about 10 centers around the world where pulmonary thromboendarterectomy is systematically performed [8]. In the hands of an experienced diagnostic and surgical team, the results can be curative [9]. The location and extent of the proximal thromboembolic obstruction are the most critical determinants of operability [2]. Occluding thrombi must involve the main, lobar, or proximal segmental arteries [2]. At a highly experienced center, mortality from pulmonary thromboendarterectomy was 7% in 457 patients [10]. Elsewhere, in case series that included 14–68 patients, mortality ranged from 5 to 24% [11– 19]. Long-term results after thromboendarterectomy for chronic pulmonary hypertension show persistent
BLUK077-Stein
January 10, 2007
CHAPTER 94
5:20
Pulmonary thromboendarterectomy in chronic PE
(a)
465
(b)
(c)
Figure 94.1 (a) Left: Scout film, coronal section of CT angiogram showing markedly dilated pulmonary arteries. (b) Right: Coronal section of CT pulmonary angiogram showing dilated pulmonary arteries and intraluminal
symptomatic [20–22].
and
hemodynamic
improvements
References 1 Kim NH, Fesler P, Channick RN et al. Preoperative partitioning of pulmonary vascular resistance correlates with early outcome after thromboendarterectomy for chronic thromboembolic pulmonary hypertension. Circulation 2004; 109: 18–22.
filling defects indicative of pulmonary hypertension and chronic pulmonary embolism. Patient also has a ventricular septal defect. (c) Bottom: Transverse section CT showing dilated pulmonary arteries and intraluminal filling defects.
2 Fedullo PF, Auger WR, Kerr KM, Rubin LJ. Chronic thromboembolic pulmonary hypertension. N Engl J Med 2001; 345: 1465–1472. 3 Moser KM, Auger WR, Fedullo PF, Jamieson SW. Chronic thromboembolic pulmonary hypertension: clinical picture and surgical treatment. Eur Respir J 1992; 5: 334– 342. 4 Dartevelle P, Fadel E, Mussot S et al. Chronic thromboembolic pulmonary hypertension. Eur Respir J 2004; 23: 637–648.
BLUK077-Stein
January 10, 2007
5:20
466
PART IV
5 Doyle RL, McCrory D, Channick RN, Simonneau G, Conte J. Surgical treatments/interventions for pulmonary arterial hypertension: ACCP evidence-based clinical practice guidelines. Chest 2004; 126: 63S–71S. 6 Pitton MB, Duber C, Mayer E, Thelen M. Hemodynamic effects of nonionic contrast bolus injection and oxygen inhalation during pulmonary angiography in patients with chronic major-vessel thromboembolic pulmonary hypertension. Circulation 1996; 94: 2485–2491. 7 Moser KM. Venous thromboembolism. Am Rev Respir Dis 1990; 141: 235–249. 8 Cerveri I, D’Armini AM, Vigano M. Pulmonary thromboendarterectomy almost 50 years after the first surgical attempts. Heart 2003; 89: 369–370. 9 Moser KM, Auger WR, Fedullo PF. Chronic major-vessel thromboembolic pulmonary hypertension. Circulation 1990; 81: 1735–1743. 10 Jamieson SW, Kapelanski DP. Pulmonary endarterectomy. Curr Probl Surg 2000; 37: 165–252. 11 Masuda M, Nakajima N. Our experience of surgical treatment for chronic pulmonary thromboembolism. Ann Thorac Cardiovasc Surg 2001; 7: 261–265. 12 Menzel T, Wagner S, Mohr-Kahaly S et al. [Reversibility of changes in left and right ventricular geometry and hemodynamics in pulmonary hypertension. Echocardiographic characteristics before and after pulmonary thromboendarterectomy]. Z Kardiol 1997; 86: 928– 935. 13 Gilbert TB, Gaine SP, Rubin LJ, Sequeira AJ. Short-term outcome and predictors of adverse events following pulmonary thromboendarterectomy. World J Surg 1998; 22: 1029–1032. 14 Miller WT, Jr, Osiason AW, Langlotz CP, Palevsky HI. Reperfusion edema after thromboendarterectomy: radio-
graphic patterns of disease. J Thorac Imaging 1998; 13: 178–183. Dartevelle P, Fadel E, Chapelier A et al. Angioscopic videoassisted pulmonary endarterectomy for post-embolic pulmonary hypertension. Eur J Cardiothorac Surg 1999; 16: 38–43. Ando M, Okita Y, Tagusari O, Kitamura S, Nakanishi N, Kyotani S. Surgical treatment for chronic thromboembolic pulmonary hypertension under profound hypothermia and circulatory arrest in 24 patients. J Card Surg 1999; 14: 377–385. Mares P, Gilbert TB, Tschernko EM et al. Pulmonary artery thromboendarterectomy: a comparison of two different postoperative treatment strategies. Anesth Analg 2000; 90: 267–273. Rubens F, Wells P, Bencze S, Bourke M. Surgical treatment of chronic thromboembolic pulmonary hypertension. Can Respir J 2000; 7: 49–57. D’Armini AM, Cattadori B, Monterosso C et al. Pulmonary thromboendarterectomy in patients with chronic thromboembolic pulmonary hypertension: hemodynamic characteristics and changes. Eur J Cardiothorac Surg 2000; 18: 696–701. Archibald CJ, Auger WR, Fedullo PF et al. Long-term outcome after pulmonary thromboendarterectomy. Am J Respir Crit Care Med 1999; 160: 523–528. Kramm T, Mayer E, Dahm M et al. Long-term results after thromboendarterectomy for chronic pulmonary embolism. Eur J Cardiothorac Surg 1999; 15: 579–583. Menzel T, Wagner S, Kramm T et al. Pathophysiology of impaired right and left ventricular function in chronic embolic pulmonary hypertension: changes after pulmonary thromboendarterectomy. Chest 2000; 118: 897–903.
15
16
17
18
19
20
21
22
Prevention and Treatment of DVT and PE
BLUK077-Stein
April 15, 2007
15:53
Index
absorbed dose, radiation, defined, 359 activated partial thromboplastin time (APTT), 391–3, 397–8, 415–16, 418 acute cor pulmonale, electrocardiogram, 206–7 adenosine triphosphate (ATP), 27 African Americans, 17, 47, 49–50, 66, 72–83, 91, 94–5, 110, 113–16 age and, 93–5, 101–2, 107, 125, 198, 224–5 infants and children, 66–7 air travel, 119–20 prophylaxis, 120 alpha-1 antitrypsin deficiency, 104–5 Alaskan Natives, 83 allergy to contrast material, 380–81, 383–4 alveolar-arterial oxygen difference, 60, 103–4, 219, 222–28 relation to partial pressure of oxygen in arterial blood, 226 relation to pulmonary artery pressure, 226–7 relation to ventilation-perfusion lung scan, 226–7 alveolar dead-space, 234–5 American College of Chest Physicians recommendations for prevention of venous thromboembolism, 407–12 American College of Chest Physicians recommendations for treatment of venous thromboembolism, 415–20 American Indians, 83 amino-terminal B-type natriuretic peptide (NT-proBNP), 253 Amplatz catheter, 455–6 angiographic catheter, 455 angiography, 319–20 AngioJet catheter, 456 ankle asymmetry, 139 anteroseptal infarction, electrocardiographic simulation, 207 anticoagulants, 389–401 direct factor Xa inhibitors, 397 direct thrombin inhibitors, 397–8 factor Va and VIIIa inhibitors, 399–401 factor VIIa/tissue factor inhibitors, 398–9 heparin, 99, 120, 128–9, 132–3, 414–15, 426–34, 459 heparinoids, 394–6 low-molecular-weight heparin, 393–4 natural anticoagulants, 389 oligosaccharides, 396–7 oral heparin, 393 unfractionated heparin, 391–4
warfarin, 81, 132, 389, 391, 415 withholding, 422–3 antiphospholipid syndrome, 133 antithrombin III, 391, 393, 396 antithrombin III deficiency, 128, 418 antithrombotic prophylaxis, 38, 99–100 age and, 54–5 APTT. see activated partial thromboplastin time argatroban, 397–8 arterial blood gases, 221–8 ascending venography, 171 Asians, 18, 49–50, 76–81 aspiration thrombectomy, 454–5 aspirin, 99, 120, 133, 407 asthma, 107 relative risk for venous thromboembolism according to age, 107 atelectasis, 58, 200, 216–18, 232 atrial fibrillation, 206 atrial flutter, 206 atrial natriuretic peptide, 26 atrioventricular block, 206 autoimmune disease, 133 autopsy deep venous thrombosis, 6–11 collateral veins, 11 forward versus retrograde, 10 locations, 8 pathology, 10–13, phlebothrombosis, 10 thrombophlebitis, 10 valve pockets, 13 pulmonary embolism, 3–5 fatal, 3 organization, 5 prevalence, 3–5 small, 3 unsuspected, 3, 5 axillary vein, 37, 41 azygos vein, dilation of, 216 B-type natriuretic peptide (BNP), 26–7, 253 balloon angioplasty for deep venous thrombosis, 441–2 bivalirudin, 397–8 biventricular dilatation, 96
467
BLUK077-Stein
April 15, 2007
15:53
468
bleeding anticoagulants, 391 thrombolytic therapy, 431–3 body mass index (BMI), 123 brachiocephalic vein, 37–9, 41–2 breast radiation with CT angiography, 361–2, 382 burns, prophylaxis, 412 calf asymmetry, 139–42 Canadian extended clinical model for assessment of pulmonary embolism, 239 cancer, 37, 39, 41, 85–92, 133–4 death from pulmonary embolism, 90–92 incidence of venous thromboembolism according to gender, 90 incidence of venous thromboembolism according to race, 91 incidence of venous thromboembolism according to type of cancer, 89–90 mechanisms of thrombosis, 86 prophylaxis, 412 risks from radiation exposure, 361–2 carcinoma of the bladder, 88–9 carcinoma of the pancreas, 86–9 cardiac arrest, 459 cardiac troponins, 25–6 cardiomegaly, 200, 218–19 cardiopulmonary disease, no prior clinical assessment, 183–89 electrocardiogram, 206–14 chest radiograph, 216–9 ventilation-perfusion lung scan, 294–309 case fatality rate, 19–22, 54 massive pulmonary embolism, 21 population mortality rate, 21–22 treated, 20–21 pulmonary embolism compared with deep venous thrombosis, 21 present era, 32–3 prior years, 31–2, untreated, 19–20 pulmonary embolism compared with deep venous thrombosis, 21–2 catheter insertion site, bleeding with thrombolytic therapy, 433 catheter-directed thrombolytic therapy, 416, 438–42 catheter-tip embolectomy, 418, 454–6 aspiration thrombectomy, 455 catheter-tip fragmentation, 418, 454–6 angiographic catheters, with, 455 impellar catheter, 455–6 rheolytic, 456 Caucasians, 17, 47, 49–50, 66, 72–83, 91, 94–5, 113–16 central pulmonary embolism prevalence, 318 CT diagnosis, 352 central venous catheter and deep venous thrombosis, 37–9, 41–2, 67
Index
cesarean section, 114–17 chemotherapy and venous thromboembolism, 86 chest radiograph, 216–19, 345 abnormal, 200–201, 216–19, 267–9 abnormalities related to pulmonary artery pressure, 218–9 abnormalities related to partial pressure of oxygen in arterial blood, 219 age and, 58–9 effective dose of radiation, 360–61 children, 66–7 Christopher Study, 336, 355 chronic pulmonary embolism, angiographic and CT criteria, 342–3, 346 chronic obstructive pulmonary disease (COPD), 101–105, 203–4, 267, 278–9 clinical findings and laboratory tests for venous thromboembolism, 103–4 rates of venous thromboembolism, 101–5 relative risk of venous thromboembolism according to age, 102 wedge angiograms, 104–5 chronic thromboembolic pulmonary hypertension, 419, 464–5 circulatory collapse, 185, 192, 197–202, 229 clinical assessment of pulmonary embolism, 183–9 all patients irrespective of prior cardiopulmonary disease combinations of findings, 187–8 predisposing factors, 184 relation to right sided pressures, 190–1 sensitivity of signs, 188 signs, 185–188 symptoms, 185–6 syndromes, 184–5 critically ill patients, 203–4 patients with no prior cardiopulmonary disease according to presenting syndromes, 197–202 combinations of findings, 194 dyspnea, hypoxia, normal chest radiograph, 194 predisposing factors, 192–3 relation to right sided pressures, 190–1 sensitivity of signs, 188 signs, 192–4 symptoms, 192, 194 syndromes, 192–3 clinical assessmsnt of lower extremities, 139–42 plus ultrasound, 147 clinical assessment and ventilation-perfusion lung scan, 275–7, 304–9 clinical characteristics of pulmonary embolism, 183–9, 192–202 clinical models and empirical assessment deep venous thrombosis, 144–5 pulmonary embolism, 239–41, 376 coagulation cascade, 128–9, 131, 133, 390, 398–9 coagulation factors, 77–81 coagulopathy in children, 67 collateral veins, deep venous thrombosis, 10
BLUK077-Stein
April 15, 2007
15:53
Index
compression stocking, 406, 415 compression ultrasound. see ultrasonography; compression computed tomographic (CT) scanners, 260–61 computed tomographic angiography age and sensitivity, 61, 336 clinical assessment, with, 330 collimation, 330, 342 complications, 353–4, 357–8 diagnostic approach, 377–8 effective dose of radiation, 359–61 elderly, 55, 61 gadolinium-enhanced, 382, 384 in chronic pulmonary embolism, 464–5 likelihood ratios, 333, 348, 351, 354 methods, 342–3 multidetector, 330, 340–3, 348–54, 433, 348–54 negative predictive value, 348, 351–3 outcome with, 333–6, 355–6 plus CT venography, 161, 166, 333, 340–43 positive predictive value, 332, 348, 351–3 procedures to minimize risk, 343 prior to PIOPED II, 325–36 results of PIOPED II, 348–54 sensitivity, 328–32, 348, 351–2 specificity, 328–32, 348, 351–2 severity index, 319–20 single slice, 343, 464–5 computed tomographic venography ascending venography, 171 complications, 353–54 diagnostic criteria, 172 diagnostic pathways, 377–84 methods, 171–2, 343 plus CT angiography, 161, 166, 177 results, 172–3, 348–54 venous phase imaging, 161, 166, 171–2 with clinical assessment, 377–84 congestive heart failure, 26–77 consolidation, chest radiograph, 217 contraceptives, oral, 122–3 contrast venography, see venography contrast-enhanced magnetic resonance angiography, see magnetic resonance angiography contrast-enhanced spiral computed tomography, see computed tomographic angiography contrast-induced nephropathy, 357–8 COPD. see chronic obstructive pulmonary disease coronary blood flow, 213–14 costophrenic angle, 218, 291 creatinine levels, 357, 382 critically ill, 203–4, 412 CT venography, see computed tomographic venography CT scanners, 260–61 CT venous phase imaging, see computed tomographic venography CTV, see computed tomography venography cyanotic heart disease, 28
469
D-dimer, for deep venous thrombosis, 149–54 for pulmonary embolism, 243–7 in diagnostic approach to pulmonary embolism, 376–84 in thrombotic/fibrinolytic process, 254 with clinical assessment for deep venous thrombosis, 158–9 with compression ultrasound for deep venous thrombosis, 160 with P-selectin and microparticles for deep venous thrombosis, 179 with alveolar dead-space for pulmonary embolism, 234–5 with clinical probability for pulmonary embolism, 250–52 with natriuretic peptide for pulmonary embolism, 253 dalteparin, 415 danaparoid, 394–6 dead-space volume, 234 deep venous thrombosis (DVT) autopsy, 6–13 brachiocephalic veins, 41–42 computed tomography (CT), 171–3 contrast venography, 161–2 D-dimer, 149–54, 179 plus clinical assessment, 158–9 plus ultrasound, 160 empirical assessment, 144–145 impedance plethysmography (IPG), 168–9, 371 lower extremities, 139–42 magnetic resonance angiography (MRA), 175–7 P-selectin and microparticles, 179 pathology, 5–13 postmortem angiography, 9 scoring system for deep venous thrombosis, 144–145 serial noninvasive leg tests, 371–2 superior vena cava, 41–42 upper extremity, 37–39, 419–20 ultrasound, 147, 160, 164–6, 346, 371–2 dermatan sulfate cofactor, 132 diagnostic accuracy, CT angiography, 325–36, 348–54 diagnostic approach, 376–84 allergy to contrast material, 380–1 clinical assessment, 376–80 low probability, 376–8 moderate probability, 378–9 high probability, 379–80 D-dimer, 376–7 computed tomography, 376–84 impaired renal function, 381–2 in extremis optional pathways, 380 pregnancy, 382 serial leg tests, 378, 380 venous phase computed tomography, 374–84 ventilation-perfusion lung scans, 378, 380–1 women of reproductive age, 382
BLUK077-Stein
April 15, 2007
15:53
470
diagnostic imaging, trends, 260–1 according to race, 72–5 according to gender, 69 digital subtraction angiography (DSA), 312–13, 340–3, 346, 360–61 direct factor Xa inhibitors, 397 direct thrombin inhibitors, 397–8 DLCO. see pulmonary diffusing capacity of carbon monoxide dorsiflexion, 139 dose-length-product, radiation, 359 DTPA, see 99m Tc-diethylenetriamine pentaacetate dual lumen catheter, embolectomy, 456 duplex ultrasonography, 164, 171 duration of hospitalization, 70, 74 DVT. see deep venous thrombosis dysfibrinogenemia, 132 dysplasminogenemia, 132 dyspnea, 55–8, 185–9, 192–5, 197–202, 253, 464 ecarin clotting time, 397 ECG. see electrocardiogram echocardiogram, 255–8, 382–4 right ventricular enlargement or dysfunction, 255–6 unstable patients, 257 visualization of emboli, 255 with venous ultrasound, 257 economy class syndrome, 119 elastic stockings, 407 elderly, 52–63 antithrombotic prophylaxis, 54–55 blood gases, 60 case fatality rate, 54 chest radiograph, 58 clinical assessment, 61 combinations of clinical findings, 58 CT angiography, 61 electrocardiogram, 59–60 diagnosis, 55–64 diagnostic tests, 61–3 predisposing factors, 55 pulmonary hemorrhage/infarction syndrome, 61 rates of diagnosis, 52–54 signs, 56–8 symptoms, 56 syndromes, 56 ventilation-perfusion lung scan, 61 effective (whole body) dose, 359–61 electrocardiogram, 59–60, 199–201, 206–14 coronary blood flow, 212–214 definitions of abnormalities, 206 disappearance of abnormalities, 211 hypoxemia, 213 ELISA. see enzyme-linked immunosorbent assay embolectomy, 459–62 emphysema, 208, 267 empirical assessment deep venous thrombosis, 145
Index
pulmonary embolism, 239–41, 376 enoxaparin, 120, 126 enzyme-linked immunosorbent assay (ELISA), 149–54, 158–9, 243–7, 250–2, 254 equivalent dose, radiation, defined, 359 estrogen-containing oral contraceptives, 122–3 dose, 122 duration, 123 obesity, 123 risks, 122–3 smoking, 123 surgery, 123 European Collaborative Trial of the Diagnostic Performance of Spiral CT (ESTIPEP Trial), 325 exposure, radiation, defined, 359 extremis, clinical assessment of patients in, 382–4 factor V Leiden, 77, 83, 129–31, 418 factor Va and VIIIa inhibitors, 399–401 factor VIIa/tissue factor inhibitors, 398–9 factor VIII, 131, 389, 418 factor X, 389 factor Xa, 397–8 factor XI, 131–2, 389 fatality rate. see case fatality rate femoral-vein Seldinger technique, 312 fever, 229–30 fibrinogen, 133 fibrinogen uptake test, 169 fibrinolytic system activation, 254 flight-related deep venous thrombosis, 119 focal oligemia, 218 fondaparinux, 396 forward thrombosis, 10 fragmentation, catheter embolectomy, 454–6 gadolinium-enhanced magnetic resonance angiography (Gd-MRA), 364–8, 382 gadolinium-enhanced magnetic resonance venography (Gd-MRV), 175–7, 377 gadopentetate dimeglumine, 110, 382 gastric carcinoma, 85 Gd-MRA. see gadolinium-enhanced magnetic resonance angiography Gd-MRV. see gadolinium-enhanced magnetic resonance venography gender, 16–18, 68–70, 90, 94–5, 125–6 Geneva clinical model, pulmonary embolism assessment, 239–41, 251–2 Greenfield catheter, 455 Greenfield filter, 444–5 G¨unther Tulip filter, 448–51 heart disease, 93–6 heart failure, 93–6, 218 fatal pulmonary embolism with, 95–6 relative risk of venous thromboembolism with, 93–4 other heart disease, 96 hemidiaphragm, elevated, chest radiograph, 216–18
BLUK077-Stein
April 15, 2007
15:53
471
Index
hemodynamic instability, 429, 434, 459–62 hemoptysis, 55–6, 61, 185, 192, 199, 232 hemoptysis/pleuritic pain syndrome, 61, 109, 197–202, 216–19, 229–30, 232–3 hemorrhage with anticoagulants, 391–3 with thrombolytic therapy, 431–3 hemorrhagic stroke, 98–100, 432 hemostasis, 109 heparin, 99, 120, 128–9, 132–3, 393, 414–15, 426–34, 459 heparin cofactor A, 132 heparin cofactor II deficiency, 132 heparin cofactor II-thrombin complex, 132 heparin-induced thrombocytopenia, 133, 393 heparinoids, 394–6 high probability clinical assessment, 379–80, 383 high probability ventilation-perfusion scan, 272–7, 280–86, 345–6 hilar arteries, chest radiograph, 216, 218 hirudin, 397–8 history and physical examination deep venous thrombosis, 139–42 no prior cardiopulmonary disease, 183–9 all patients with pulmonary embolism, 192–6 Homans’ sign, 139 homocystenemia, 132, 418 Hydrolyser catheter, 456 hypercoagulable syndrome, 128–34 acquired, 133–4 anti-B2 -glycoprotein I antibodies, 133 anticardiolipid antibodies, 133 antiphospholipid syndrome, 130, 133 dysfibrinogenemia, 133 heparin-induced thrombocytopenia, 133 lupus anticoagulant antibodies, 133 myeloproliferative disorders, 133 malignancy, 130, 133–4 diagnostic approach, 134 inherited, 128–32 antithrombin III deficiency, 128, 130 dysfibrinogemias, 130, 132 factor V Leiden, 129–31 factor VIII elevation, 130–1 factor XI elevation, 130–2 heparin cofactor II deficiency, 130, 132 hyperhomocystenemia, 130, 132 plasmin generation disorder, 132 protein C deficiency, 128–30 protein S deficiency, 129–30 prothrombin 20210A mutation, 130–1 sticky platelet syndrome, 133 hyperemic zones, chest radiograph, 218 hyperhomocystenemia, 132 hypocapnia, 227 hypoxemia, 194, 203–4, 212–14, 227, 267 I-125-fibrinogen uptake, 31, 162 idraparinux, 396
immobilization, 184, 192 impedance plethysmography (IPG), 168–9, 355, 371, 422 outcome, 168–9 incidence in hospitals, 16–18 deep venous thrombosis, 16–18 gender, 16–18 pulmonary embolism, 16–18 race, 16–18 venous thromboembolism, 16–18 infants, 66–7 inferior vena cava filters, 24, 418, 423, 430, 444–51 approved types, 444, 447 complications, 444–5, 448–9 retrievable, 447–51 trends in use, 445–7 infiltrate, chest radiograph, 216, 218, 232 inflammation of the vein. see venous inflammation intensive care unit, clinical assessment, 203–4 intermittent pneumatic compression, 405–7, 417 internal jugular vein thrombosis, 37, 41 International Commission of Radiological Protection, 361 International Cooperative Pulmonary Embolism Registry, 431 international normalized ratio (INR), 81, 389 international sensitivity index (ISI), 389 International Stroke Trial Collaboration Group, 99 interstitial edema, chest radiograph, 218 intracranial hemorrhage with thrombolytic therapy, 431, 433 intraluminal filling defect contrast venography, 162 CT angiogram, 341 CT venogram, 343 pulmonary angiogram, 311–12 quantification, 319–20 intrinsic pathway, 131 iodinated contrast material, allergy to, 380–81 ischemia, myocardial, 25 ischemic stroke, 98–100 isolated deep calf-vein thrombosis, treatment, 414 isolated dyspnea syndrome, 197–202, 230 krypton ventilation, 264 late dead-space fraction, 234–5 latex turbidimetric assay, 179 left-axis deviation, 207–8, 211 leukemia, 86 leukocytosis, 232–3 local-regional thrombolytic therapy for deep venous thrombosis, 437–42 low probability clinical assessment diagnostic approach, 376–8, 383 clinical models, 239–41 low probability ventilation-perfusion scan, 272–7, 282–3, 288, 345–6 low-molecular-weight heparin, 393–4, 414–20 lymphoma, 86
BLUK077-Stein
April 15, 2007
15:53
472
macroaggregated human albumin, 262, 344 magnetic resonance angiography complications, 368 diagnostic criteria, 365 direct thrombus imaging, 368 perfusion imaging, 367 pulmonary embolism, 364–8 sickle cell disease, 110 without contrast material, 367 magnetic resonance imaging (MRI), 368 magnetic resonance venography, 175–7 complications, 177 delayed gadolinium enhanced, 175–7, 365–7 diagnostic criteria, 177 gadolinium enhanced, 175 upper extremities, 177 without contrast material, 175 malignant neoplasm, 85–86, 88, 133 massive pulmonary embolism, 3–5, 21–2, 31, 206–11, 216–19, 255–8, 434, 454–5, 459 matched ventilation-perfusion abnormalities, 290–91 men, 68–70 duration of hospitalization, 70 rates of diagnosis, 68–9 ultrasonography, 69 venography, 69 ventilation-perfusion lung scan, 69 microparticles, 179 Miller index, 320 mismatched segmental perfusion defects, 294–309 mismatched vascular perfusion defect, 299–309 moderate probability clinical assessment diagnostic approach, 378–9, 383 clinical models, 239–41 mortality rate from pulmonary embolism compared with deep venous thrombosis, 414 in-hospital, 24–5, 255 pulmonary embolectomy, 459–62 population, 21–2, 77–81 regional, 47–51 seasonal, 44–45 with thrombolytic therapy, 428–30 MRA. see magnetic resonance angiography MRV. see magnetic resonance venography multidetector contrast-enhanced CT, 330, 348–54, 433, 377 2-slice CT, 325, 331–4 4-slice CT, 330–36 16-slice CT, 333 myeloid metaplasia, 133 myeloproliferative disease, 86, 91,133 myocardial infarction, 25–6, 206–7, 212–3 myocardial ischemia, 25, 27, 212–13 myocardial perforation, pulmonary angiography, 323 myoglobin, 26 natriuretic peptides, 26–7, 253 negative predictive value of diagnostic tests, 374–5 nematode anticoagulant peptide c2, 398–9
Index
neonatal deep venous thrombosis, 67 neoplastic vascular compression, ventilation-perfusion lung scan, 267 nephrogenic systemic fibrosis or nephrogenic fibrosing dermopathy (NSF/NFD), 177, 368, 382, 384 nephrotic syndrome, risk of deep venous thrombosis, 184 nephrotoxicity, iodinated contrast material, 357–8, 382 neural network, 236–8 neutrophils, 232 nonionic low-osmolar contrast material, 161–2 nonsegmental perfusion abnormalities, 288–9 normal ventilation-perfusion lung scan, 272–7, 345–6 NT-pro-B-type natriuretic peptide, 253 Oasis catheter, 456 obesity, 123, 125–6 oligemia chest radiograph, 216, 218 pulmonary angiogram, 311–12 oligosaccharides, 396–7 oral contraceptives, 122–3 oral heparin, 393 organization rate of thrombi, 5–7 orthopedic surgery prophylaxis, 409–12 orthopnea, 185–6, 189, 192 osteopenia, 394 osteoporosis, 393 outcome studies versus accuracy, 355–6 P pulmonale, 206–8 P-selectin, 179 partial pressure of carbon dioxide in arterial blood (Pa CO2 ), 222–4, 227, 234–5 plasminogen activator inhibitor-1 (PAI-1), 254 partial pressure of oxygen in arterial blood ( Pa O2 ), 60, 190–91, 194, 200–201, 211–14, 219, 221–8 Pacific Islanders, 18, 49–50, 76–81 pathology, 5–13 pentasaccharides, 396 perfusion defects, 288–309, 367, 428 perfusion lung scan, 35–6, 262, 271, 280–81, 344, 360–61 pH in arterial blood, 222 phlebitis, 11 phlebothrombosis, 10–13 phosphatidylserine, 179 physiological dead-space, 234–5 pigtail catheter, thrombus fragmentation, 455 Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED, also PIOPED I), 183–95, 192, 197–9, 202, 237–8, 271–8, 284–6, 288 Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II) clinical characteristics of patients in, 183–95, 197–9, 202 criteria for very low probability of PE, 292 diagnostic approaches of, 162, 164, 376–84 DVT and, 161–2, 172–3 methods of, 340–46
BLUK077-Stein
April 15, 2007
15:53
Index
objectives of, 348 results of, 348–54 Prospective Investigation of Pulmonary Embolism Diagnosis III (PIOPED III), 381 Pisa model for clinical assessment, 239–41 PISA-PED study, perfusion lung scans, 281 plain chest radiograph. see chest radiograph planar imaging, perfusion lung scan, 310 plasmin generation disorder, 132 plasminogen activator inhibitor type-1 (PAI-1), 254 plastic cup suction catheter, aspiration thrombectomy, 455 platelets, sticky, 132 pleonemia, chest radiograph, 216 pleural effusion chest radiograph, 216–18 lung scan, 291–3 pleural-based opacity, 218 pleuritic pain, 55–6, 58, 61, 185–8, 190, 192, 199, 202, 232 pneumonia, 230, 232 pneumothorax, chest radiograph, 216 polycythermia vera, 133 population mortality rate, 21–2, 77–81 positive predictive value of diagnostic tests, 373–4 postmortem contrast venography, 162 postthrombotic syndrome, 415, 417, 437, 442 postvenography thrombosis, 162 predictive value of diagnostic tests, 373–5 predisposing factors, 184, 192–3, 197 pregnancy, 113–17, 382, 384 age, 113–4 deep venous thrombosis, 113–117 mode of delivery, 114–5 race, 113–5 rate of diagnosis, 113 pulmonary embolism, 114–117 preperfusion ventilation scan, 263–4 prevalence, at autopsy, 3–6 prevalence in central and subsegmental arteries, 318 prevention of deep venous thrombosis and pulmonary embolism, 405–13 burns, 412 cancer, 412 critical care, 412 elastic stockings, 406–7 intermittent pneumatic compression, 405–7, 417 medical conditions, 412 surgery, 407–12 general, 407–8 gynecologic, 409 hip arthroplasty, 409–10 hip fracture, 410 knee arthroplasty, 410 knee arthroscopy, 410 laparoscopic, 409 lower extremity, 411 neurosurgery, 411
473
orthopedic, other 410–1 spine, 411 urologic, 409 vascular, 409 trauma, 411–2 travel, 412–3 pro-B-type natriuretic peptide (pro BNP), 253 probability assessment number of perfusion defects, 294–303 in diagnostic pathway, 376–84 prognosis of pulmonary embolism biochemical markers, 25–8 cardiac troponins, 25 myoglobin, 26 natriuretic peptides, 26–7 prognostic model, 24–5 untreated, 31–33 uric acid, 27–8 with right ventricular enlargement, 24 without right ventricular enlargement, 24 prominent central pulmonary artery, 218–19 protein C, 128–9, 389, 399, 418 protein S, 129, 389, 418 Prothia filter, 449 prothrombin, 389 prothrombin 20210A mutation, 131, 418 prothrombin gene, 131 prothrombin time ratio, 389 proximal pulmonary artery, chest radiograph, 218 pseudo left-axis deviation, 208 pseudoinfarction pattern, 207–8, 211 pulmonary angiography complications, 55, 321–4, 357–8 digital subtraction, 346 history, 311–2 in elderly, 55, 63 in sickle cell disease, 110 outcome with normal, 314 peripheral injection, 316 quality, 313–4 quantification by, 319–20 reader agreement, 314–6 severity indices, 319–320 techniques in, 311–16 trends in use, 260–1 wedge angiography, 312–5 pulmonary artery branch and lung scan, 282–3 pulmonary artery hypertension pattern chest radiograph, 218 pulmonary artery mean pressure, 190–91, 218–19 related to clinical characteristics, 190–91 related to chest radiographic abnormalities 218–19 pulmonary artery trunk, chest radiograph, 216, 218 pulmonary artery tumors and lung scan, 267 pulmonary artery, dilation, chest radiograph, 218 pulmonary consolidation and lung scan, 267 pulmonary diffusing capacity of carbon monoxide (DLCO), 36, 428
BLUK077-Stein
April 15, 2007
15:53
474
pulmonary edema chest radiograph, 216, 218 lung scan, 267 pulmonary embolectomy, 28, 418, 444, 459–62 pulmonary hemmorhage/infarction syndrome, 109, 197–202, 216–19, 229–30, 232–3, 267, 292 pulmonary hypertension, 109–11, 218, 323–4, 429, 444 pulmonary parenchyma abnormalities, 58, 200–201, 216–18 pulmonary thromboendarterectomy, 419, 444, 464–5 pulmonary vascularity, decreased, chest radiograph, 218 pulmonary vasculitis, lung scan, 267 pulmonary venous hypertention pattern, chest radiograph, 218 purine nucleotide degradation, 27–8 Q wave, 206–12 QRS axis, 207–11 quantitative latex assay, 149–54, 158–9, 244–247, 253 quantitative rapid ELISA, 149–54, 158–60, 243–7, 250–51, 376–84 race African Americans, 17, 47, 49–50, 66, 72–83, 91, 94–5, 110, 113–16 American Indians/Alaskan Natives, 83 Asians/Pacific Islanders, 18, 49–50, 76–81 Caucasians, 17, 47, 49–50, 66, 72–83, 91, 94–5, 113–16 incidence and mortality, 72–5 use of diagnostic tests, 73–75 use of medical facilities, 75 radiation as risk factor, 86 radiation exposure, 359–362, 382 absorbed dose, 359 background radiation, 361 chest radiograph, 360–1 computed tomography, 359 definitions, 359 digital subtraction angiography, 360–1 effective dose, 359 exposure, 359 equivalent dose, 359 fetal exposure, 362 genetic effects, 362 methods for decreasing radiation, 362 radiation workers, 361 risks, 361–2 perfusion lung scan, 360–1 ventilation-perfusion scan, 361 radioactive fibrinogen scintiscans, deep venous thrombosis 31, 162 radioaerosol inhalation, 264–5, 344–5 radionuclide lung scans, see ventilation-perfusion lung scans rales, 187, 199–201 razaxaban, 397 recurrent pulmonary embolism, 32–3, 355–6, 414, 430, 444 serial leg tests, 33 untreated patients, 33
Index
regional rates of diagnosis and mortality, 47–51 mortality, 47–51 rates of diagnosis, 47–50 renal failure, 55, 63, 357, 368, 381–2, 384,398, 407, 416, 418 resolution of pulmonary embolism, 35–36 with prior cardiopulmonary disease, 35–6 without prior cardiopulmonary disease, 35–6 restrictive disease, lung scan, 267 reticulonodular disease, lung scan, 267 retrievable inferior vena cava filters, 447–51 retrograde thrombosis, 10 retroperitoneal hemorrhage, thrombolytic therapy, 432 rheolytic thrombectomy, 454, 456 right atrial enlargement, electrocardiogram, 208 right atrial mean pressure, related to clinical characteristics, 190–91 right bundle branch block, 206–11 right ventricular dysfunction, 24–6, 255–8, 429–30, 434 right ventricular enlargement, 24–6, 96, 213, 255–8, 429–30 right ventricular failure, 19, 32 right ventricular to left ventricular dimension ratio (RV/LV), 24 right-axis deviation, 206–211 risk factors, 405 rivaroxaban, 397 recombinant tissue plasminogen activator (rt-PA), 427–34 S wave, 206 S1 Q3 T3 complex, 206–8, 211 S1 S2 S3 pattern, 208 scintigraphy, see ventilation-perfusion lung scans seasonal, 44–5 incidence, 44 mortality, 44–5 segmental pulmonary arteries, prevalence of pulmonary embolism, 318–20 semiquantitative latex assay, 149–54, 243–7 serial noninvasive leg tests in suspected venous thromboembolism, 371–2 in pulmonary embolism and high risk bleeding, 422–3 shock, 429, 434, 459–62 sickle cell disease, 109–11, 267 signs deep venous thrombosis, 139–43 pulmonary embolism all patients, irrespective of prior cardiopulmonary disease, 192–4 patients with no prior cardiopulmonary disease, 185–8 according to age, 56–8 according to presenting syndromes, 199 silent pulmonary embolism, 16 single photon emission computed tomography (SPECT), 310 single-detector CT, 325, 330–336 smoking, and oral contraceptives, 123 SNAC heparin, {sodium N-[8(2-hydroxybenzoyl) amino] caprylate}, 393 spiral CT, 171, 177, 262, 325–36, 342–4, 346, 348–54
BLUK077-Stein
April 15, 2007
15:53
Index
ST segment, 59–60, 206–13 Standards for Reporting Diagnostic Accuracy (STARD), 340 stasis, venous, 119 stents, for deep venous thrombosis, 441–2 sticky platelet syndrome, 132–3 stocking, compression, 406, 415 streptokinase, 425–34, 437–8, 441 stripe sign, lung scan, 291, 310 stroke, 98–100 incidence of venous thromboembolism, 98–9 fatal pulmonary embolism, 99–100 subclavian vein thrombosis, 37, 41 subsegmental pulmonary emboli, prevalence 318, CT diagnosis, 352 superior vena cava (SVC) chest radiograph, 216 syndrome, 41–2 thrombosis, 37–9 surgery, antithrombotic prophylaxis, 407–12 symptoms pulmonary embolism, 56, 185–8, 192–4, 197–8 all patients, irrespective of prior cardiopulmonary disease, 192, 194 patients with no prior cardiopulmonary disease, 185–8 according to age, 56–7 according to presenting syndromes, 197–8 according to age, 56 systemic lupus erythematosus, antiphospholipid syndrome, 133 systemic thrombolytic therapy, 437–42 T wave, 59–60, 206–12 tachycardia, 185–7, 194–5, 199 tachypnea, 55–8, 185–9, 192–5, 199, 202 99m Tc-diethylenetriamine pentaacetate (DTPA), 264–5, 344 Tecumseh Community Health Study, 68 Tempofilter, 448–9 thrombectomy catheter, pulmonary embolism, 454–6 deep venous thrombosis, 441–2 thromboendarterectomy, pulmonary, 419, 444, 464–5 thrombin, 397–8 thrombin time (TT), 132 thrombocytopenia, 133, 393, 396 thromboendarterectomy, 464–5 thrombolysis and uric acid, 28 thrombolytic agents, 425–8 thrombolytic processes, natural, 35 thrombolytic therapy acute pulmonary embolism, 425–34 bleeding, 431–3 cardiopulmonary arrest diagnostic tests, 433 direct infusion, 425–6 indications, 433–4 meta-analysis, 430–1 normotensive patients, 428–9 recurrent pulmonary embolism, 430
475
regimens, 425 resolution, 426–8 right ventricular dilation or dysfunction, 429–30 shock, 429 deep venous thrombosis, 437–42 angioplasty, with, 441–2 catheter-directed, 438–42 local-regional, 437–40 systemic, 437–42 stent, with, 441–2 thrombectomy, with, 441–2 thrombophilia, 128–134 thrombophlebitis, 10–13, 55, 93, 419 thromboplastin, 389 thrombosis in situ, sickle cell disease, 109–10 thrombotic/fibrinolytic process, 254 tifacogin, 398 tinzaparin, 415 tissue factor, 179, 390–400 tissue factor pathway inhibitor (TFPI), 398–9 tissue plasminogen activator (tPA), 254, 425–34, 437–8, 441–2 toxic endophlebitis, 11 transvenous pacemakersand upper extremity deep venous thrombosis, 37–8 trauma antithrombotic prophylaxis, 411–2 travel, 119–20, 412–3 treatment, deep venous thrombosis and acute pulmonary embolism antithrombotic therapy, 414–20 catheter-directed thrombolysis, 416, 418, 438–42 catheter-tip embolectomy, 418, 454–6 catheter-tip fragmentation, 418 chronic thromboembolic hypertension, 419 inferior vena cava filter, 418, 444–52 negative serial leg tests and high risk of bleeding, 422–3 pulmonary embolectomy, 418, 459–62 thrombolytic therapy, 418, 425–34, 437–442 upper extremity deep venous thrombosis, 419 withholding anticoagulants if high risk of bleeding and negative serial leg tests, 422–3 troponin C, 25 troponin I, 25–6 troponin T, 25–6 tumor associated inflammatory cell procoagulants and/or cytokines, 86 tumor cell procoagulants and/or cytokines, 86 ultrasonography, 147, 160–61, 164–6, 257, 341, 346, 355–6, 422–3 accuracy, 164–5 bedside, 382–4 calf veins, 165 compression, 164–6 comparison with CT venous phase imaging, 166 diagnostic criteria, 164 equipment, 164 in pulmonary embolism, 165
BLUK077-Stein
April 15, 2007
15:53
476
ultrasonography (continued ) outcome studies, 166 venous, 257, 260, 336, 341, 346, 355–6, 380–84, 423 unfractionated heparin, 391–4, 414–19, 423 unsuspected pulmonary embolism at autopsy, 3–6 untreated deep venous thrombosis, 31–33 pulmonary embolism, 31–3, 422–3 recurrent pulmonary embolism, 31–3 upper extremity deep venous thrombosis, 37–9 brachioephalic vein, 39, 41–2 superior vena cava, 39, 41–2 treatment, 419 uric acid, 27–8 urokinase, 425–34, 437–8, 441–2 Urokinase Pulmonary Embolism Trial (UPET), 35–6, 183, 188, 194, 202, 216, 230, 321, 426, 428 Urokinase-Streptokinase Embolism Trial, 230, 427 use of diagnostic tests, 73–5 use of medical facilities, 75 V-Q lung scan. see ventilation-perfusion lung scan valve pockets, 13 vascular defects ventilation-perfusion lung scans, 299–309 vascular occlusion disease, ventilation-perfusion lung scans, 267 venography, 9, 161–162, 171 adequacy of visualization, 162 complications, 162 correlation with I-125 fibrinogen, 162 correlation of postmortem venography with dissection, 162 diagnostic criteria, 162 methods, 161–2 trends in use, 161 venous inflammation, 10–13 venous phase computed tomography, 161, 171–2, 348–354, 377–84 venous ultrasound. see ultrasonography; venous ventilation imaging, 262–4, 271, 344 ventilation-perfusion lung scan assessment of, 294–309 causes for mismatch, 267 chronic obstructive pulmonary disease and, 278–9 clinical assessment and, 275, 6 criteria for interpretation, 267–9, 284–286, 288–93, 294–7 critically ill, 204 effective dose of radiation, 360–61 high probability, 165, 200–202, 272–7, 280–86, 345–6 in main or lobar pulmonary arteries, 282 intermediate probability, 200, 202, 204, 275–6, 282–3, 288, 345–6 interpretation of, 267–9 low probability, 202, 204, 272–7, 282–3, 288, 345–6 matched ventilation-perfusion defects, 290 methods in PIOPED II, 344–6 no prior cardiopulmonary disease and, 278–9 nonsegmental abnormalities, 268 normal chest radiograph and, 278–9
Index
perfusion imaging, 262, 280 pleural effusion, 291 prior cardiopulmonary disease and, 278–9 radioaerosol inhalation, 344–5 resolution following pulmonary embolism results of PIOPED, 271–277 revised criteria, 284–286 relation to alveolar-arterial oxygen difference, 226–7 segmental branch emboli, 282 segmental equivalent perfusion defects mismatched, 271, 294–7 stratified according to prior cardiopulmonary disease, 298–303 with clinical assessment, 304–9 single photon emission computed tomographic perfusion scan, 310 small perfusion defects, 291 stripe sign, 291 subsegmental branch emboli, 282 echniques of, 262–5, 344–6 trends in use of, 61, 63, 69, 260–61 triple match, 291 vascular defects mismatched, 294–7 stratified according to prior cardiopulmonary disease, 298–303 with clinical assessment, 304–9 ventilation imaging, 262–5 very low probability, 288–93, 345–6 xenon ventilation, 344 ventilatory support and clinical assessment, 203–4 ventricular dysfunction, prognosis, 24 Venturi effect, embolectomy catheter, 456 very low probability ventilation-perfusion scan, 288–93, 345–6 vitamin K antagonists, 389, 414–15, 417–19 volumetric capnography, 234 Walsh index, 319–20 warfarin, 81, 132, 389, 391, 415 Weibel-Palade bodies, 179 Wells score deep venous thrombosis, 144–5, 147, 159–60 pulmonary embolism, 234–5, 239–41, 251–2, 340 Westermark’s sign, chest radiograph, 218 white blood cell (WBC) count, 232–3 whole blood agglutination assay, 149–54, 158–60, 243–7, 250–51 whole body dose radiation, 359–61 withholding treatment of, 422–3 women, 68–70 duration of hospitalization, 70 rates of diagnosis, 68–9 ultrasonography, 69 venography, 69 ventilation-perfusion lung scan, 69 women of reproductive age, diagnostic approach, 382, 384 xenon ventilation lung scan, 262–4, 344–5 ximelagatran, 397–8