Urinary Tract Infections (Infectiology, Vol. 1)

EEEE:L01:BIF01XXXVO.77

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infections

7.4.1997

EEEE:L01:BIF01XXXVO.77

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Infectiology Vol. 1

Series Editor

ABC

Tom Bergan, Oslo

Basel W Freiburg W Paris W London W New York W New Delhi W Bangkok W Singapore W Tokyo W Sydney

EEEE:L01:BIF01XXXVO.77

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infections

Volume Editor

Tom Bergan, Oslo

23 figures and 34 tables, 1997

ABC

Basel W Freiburg W Paris W London W New York W New Delhi W Bangkok W Singapore W Tokyo W Sydney

EEEE:L01:BIF01XXXVO.77

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Infectiology

Library of Congress Cataloging-in-Publication Data Urinary tract infections/volume editor, Tom Bergan. (Infectiology; vol. 1) Includes bibliographical references and index. 1. Urinary tract infections. I. Bergan, Tom. II. Series. [DNLM: 1. Urinary Tract Infections. WJ 151 U7622 1997] RC901.8.U752 1997 616.6--dc21 ISBN 3–8055–6440–6 (hardcover: alk. paper)

Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher. © Copyright 1997 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) Printed in Switzerland on acid-free paper by Thür AG Offsetdruck, Pratteln ISBN 3–8055–6440–6

EEEE:L01:BIF01XXXVO.77

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Contents

VII Preface 1

Treatment of Acute Uncomplicated Urinary Tract Infection Stamm, W.E. (Seattle, Wash.)

8

Management of Acute Uncomplicated Pyelonephritis Nicolle, L.E. (Winnipeg)

14

Uncomplicated Acute Pyelonephritis Bailey, R.R. (Christchurch)

19

Complicated Urinary Tract Infections Kumazawa, J.; Matsumoto, T. (Fukuoka)

27

Urinary Tract Infection in the Renal Transplant Recipient Tolkoff-Rubin, N.E. (Boston, Mass.); Rubin, R.H. (Cambridge, Mass.)

34

Urinary Tract Infection and the Immunocompromised Host: UTI in Renal Transplant Patients Tolkoff-Rubin, N.E.; Rubin, R.H. (Boston, Mass.)

37

Bacteriuria in Male Patients Infected with Human Immunodeficiency Virus Type 1 Relation to Immune Status (CD4+ Cell Count) and the Influence of Pneumocystis carinii Pneumonia Prophylaxis on Incidence and Resistance Development van Dooyeweert, D.A.; Schneider M.M.E.; Borleffs, J.C.C.; Hoepelman, A.I.M. (Utrecht)

46

Urinary Tract Infections in Young Men Stamm, W.E. (Seattle, Wash.)

48

Urinary Tract Infection in Renal Transplant Patients Krcˇméry, S. (Bratislava)

V

EEEE:L01:BIF01XXXVO.77

FF: W1FP

E1: IN01

$$$

52 Urinary Tract Infections in Patients on Anticancer Drugs Matsumoto, T.; Sakumoto, M.; Kotoh, S.; Kumazawa, J. (Fukuoka) 57 Urinary Tract Infections in Patients on Glucocorticosteroid Therapy Ludwig, E. (Budapest) 60 Chronic Bacterial Prostatitis – A Clinical Reevaluation of Old Woes Weidner, W.; Ludwig, M.; Schiefer, H.-G. (Giessen) 67 Relevance of NCCLS Breakpoints for Susceptibility as Applied to Urinary

Tract Infections Jones, R.N. (Iowa City, Iowa) 74 Antibacterial Activity of Antibacterial Agents in Urine: An Overview of

Applied Methods Naber, K.G. (Straubing) 84 Response of Escherichia coli to Fleroxacin in an in vitro Model of the

Urinary Bladder Kawada, Y.; Kanimoto Y. (Gifu) 89 The Role of the Animal Model in the Study of Prostatitis Nickel, J.C. (Kingston) 98 Tropism in Bacterial Infections: Urinary Tract Infections Roberts, J.A. (Covington, La.) 106 Experimental Prostatitis: Impact, Trends and Future Weidner, W. (Giessen) 109 Adherence and the Pathogenesis of Urinary Tract Infection Connell, H.; Svanborg, C.; Hedges, S.; Agace, W.; Hedlund, M; Svensson, M. (Lund); Benson, M.; Jodal, U. (Göteborg) 118 Identification of a Novel Antibacterial Factor from Human Urine Connell, H.; Sabharwal, H.; Persson, L. (Lund); Zasloff, M. (Plymouth Meeting, Pa.); Svanborg, C. (Lund) 125 Virulence Profile of Uropathogenic Escherichia coli in Patients with

Nonobstructive Chronic Pyelonephritis Fünfstück, R.; Jacobsohn, N. (Jena); Tschäpe, H. (Wernigerode); Stein, G. (Jena) 133 Urinary Tract Infection: Some Research Priorities Ronald, A.R.; Sanche, S.E. (Winnipeg) 138 Author Index 139 Subject Index

Contents

VI

EEEE:L01:BIF01XXXVO.77

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Preface

This is the first volume of a new series of books named Infectiology. The purpose is to focus on new and improved knowledge that is appearing with increasing frequency within this important field of medicine. Infections involve all branches of medicine and, consequently, have implications for the whole spectrum of disease categories. Infections are particularly frequent among children and represent one of the untoward events which disturbs success in technical advances within surgery. This branch of medicine is special in that the laboratory science, microbiology and pharmacy involved in chemical advances of new antimicrobial agents play such a vital role in the understanding of which processes are taking place and in the discovery of new drugs for the treatment of patients with infectious diseases. The idea to publish our first volume of Infectiology on the topic of Urinary Tract Infections grew out of a session held by the International Society of Chemotherapy Commission on Urinary Tract Infections, presented at their 19th International Congress in Montreal. This book compiles and interprets recent research data and presents current concepts of the pathogenesis, prevention and treatment of UTIs. Readers will find an overview of modern methods of diagnosis and new antibacterial agents, and will be provided with useful recommendations for the choice of antibiotic and the duration of treatment in different forms of uncomplicated and complicated UTIs. Consequently, there is no doubt that the series Infectiology will potentially represent a valuable source of information within one of the most vibrant fields of basic science integrated with applied medicine. Tom Bergan, Editor

VII

PIAM:L01:BIF01002XA.88

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 1–7

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Treatment of Acute Uncomplicated Urinary Tract Infection Walter E. Stamm Division of Allergy and Infectious Diseases, University of Washington School of Medicine, Seattle, Wash., USA

Much of the recent literature on treatment of uncomplicated tract infections (UTI) has focused on the use of short-course therapy [1]. The goals of short-course therapy are to improve patient compliance, to reduce the costs of treating uncomplicated UTI, to reduce the number of adverse reactions encountered, to eradicate infection with the same efficacy achieved with longer courses of therapy, to reduce the likelihood of recurrence, and to prevent, if possible, the emergence of antimicrobial resistance in microorganisms from patients who are treated for uncomplicated UTI. The first studies which looked at short-course treatment of UTI were done more than 20 years ago by Drs. Brumfitt, Ronald, Bailey and colleagues. The characteristics of an antimicrobial that can be used for short-course therapy of acute uncomplicated UTI should include: (1) activity in vitro against most of the anticipated pathogens (namely Escherichia coli, Staphylococcus saprophyticus, and occasionally Proteus, Klebsiella, or enterococci); (2) relatively high and prolonged urinary concentrations; (3) no loss of antimicrobial activity in urine; (4) an agent that is rapidly bactericidal, and (5) if possible, an agent that is active at subinhibitory concentrations (perhaps by inhibiting bacterial adherence). Recent studies suggest that the effects of an antimicrobial on the vaginal and fecal flora may be quite important in determining long-term cure and one would thus like an agent that minimally alters the normal vaginal flora yet eradicates E. coli or other uropathogens from the vaginal flora [1]. Finally, an antimicrobial used for short-term treatment should minimize selection of resistant strains. No antimicrobial satisfactory fulfills all of these objectives, but there are several that fulfill most. Table 1 compares some of the commonly used antimicro-

PIAM:L01:BIF01002XA.88

FF: W1FP

E1: IN01

$$$

Table 1. Drugs for treatment of uncomplicated UTI Sulfonamide

Nitrofurantoin

Amoxicillin

Amoxicillin TMPclavulanic SMX acid

Fluoroquinolone

Spectrum

25%

15%

30%

5%

5–15%

! 2%

Urine level

high

high

high

high

high

high

T1/2 – urine

M

M

S

S

L

L

Gut/vagina

P

–

P

P

G

G

Side effects

++

++

++

+++

++

++

Cost

low

low

low

high

low

high

M = Medium; S = short; L = long; P = poor; – = negative; G = good; ++ = infrequent; +++ = moderately frequent.

bials for treatment of uncomplicated UTI with respect to the properties previously listed. Trimethoprim (TMP)-sulfamethoxazole (SMX) and the fluoroquinolones best fulfill many of the criteria just outlined. In parts of North America, TMP and TMP-SMX are still active in vitro against 95% of the organisms that cause acute uncomplicated UTI, while in other parts of the world, in vitro resistance may be as high as 15–30% or even higher [1]. In most parts of the world, fluoroquinolones remain very active against the vast majority of uropathogens causing uncomplicated UTI (! 2% resistant). Both these classes of compounds achieve high urinary levels that are present in the urine for long periods of time and they exert excellent activity against uropathogens colonizing the vaginal and fecal flora while not effecting the anaerobic vaginal flora. They differ in cost, with TMP and TMP-SMX being relatively inexpensive and the fluoroquinolones being relatively expensive. One of the original goals of short-course therapy was to localize the infection based on response to therapy; that is, the concept was developed that one could divide acute uncomplicated UTI in women into two groups, those who truly had infection limited to the bladder (acute cystitis) in which a very high cure rate could be achieved with single-dose therapy, and those patients who also had occult pyelonephritis. In these individuals, the cure rate was theoretically lower with single-dose therapy and it was thought that longer treatment was indicated [2]. While early studies utilizing the antibody-coated bacteria test suggested that this paradigm might be valid, subsequent studies have demonstrated high cure rates in patients with positive ACB tests even when single-dose treatment with TMP or TMP-SMX were given [2, 3].

Stamm

2

PIAM:L01:BIF01002XA.88

FF: W1FP

E1: IN01

$$$

Table 2. Meta-analysis of short-term therapy for acute UTI in women [data from 4] Reviewed 28 controlled trials For all antibiotics, single-dose therapy (SDT) less effective than 3-day or 65-day therapy Considerable differences by drug ß-Lactam least effective for SDT, more effective as 65-day therapy TMP-SMX more effective than ß-lactams for SDT, but still more effective as 3-day therapy Side effects increase with TMP-SMX given for 63 days

Table 3. Meta-analysis of single-dose therapy (SDT) for acute UTI in women (data from [5]) Reviewed 25 controlled treatment trials SDT less effective than 67-day therapy at 3–14 days or 4–6 weeks’ follow-up SDT cure rates especially low with ß-lactams Side effects less with SDT than with longer regimens

Two excellent reviews have been published which emphasize many of the trends that have been observed in individual studies. Norrby [4] reviewed 28 controlled trials of short-term therapy for acute uncomplicated UTI (table 2) and concluded that for most antibiotics where significant numbers of comparative trials were available, single-dose therapy was less effective than 3 days, 5 days, or longer therapy. There were important differences by individual drugs, however; thus, the ß-lactams were the least effective when given as single-dose therapy and were much more effective when given for 5 days or more. In contrast, TMP-SMX or TMP along were more effective than the ß-lactams when given for single-dose therapy, but they were more effective still when given for 3 days. Individual studies, as well as this review, demonstrated that side effects do increase in prevalence with all compounds as the duration of therapy lengthens, but this was especially true for TMP-SMX when the duration was longer than 3 days. A second review and meta-analysis done by Leibovici and Wysenbeek [5] (table 3) reviewed 25 treatment trials and reached conclusions similar to Norrby [4]. Several disadvantages attributed to single-dose therapy in practice should also be mentioned. From the patient’s perspective, symptoms sometimes persist after the treatment period has ended and patients, therefore, sometimes feel that not enough therapy have been given. In addition, some studies suggest that there

Treatment of Acute Uncomplicated Urinary Tract Infection

3

PIAM:L01:BIF01002XA.88

FF: W1FP

E1: IN01

$$$

Table 4. Predictors of therapeutic failure with short-course therapy Patient factors Age (postmenopausal) History of recent UTI Diaphragm/spermicide use Non-Black race (?)

Microbial factors Colony count 6105/ml (?) S. saprophyticus, non-E. coli Antibiotic resistance Virulence determinants

Table 5. Treatment regimens for acute uncomplicated cystitis Drug

Dose, mg Dose regimen

Cost, USD

TMP

200

q12 h ! 3 days

1.95

TMP-SMX

160/800

q12 h ! 3 days

2.94

Norfloxacin

400

q12 h ! 3 days

9.30

Ciprofloxacin

500

q12 h ! 3 days

10.80

Ofloxacin

200

q12 h ! 3 days

9.80

Nitrofurantoin

100

q6 h ! 3 days

1.74

may actually be slower resolution of symptoms with single-dose therapy as compared with longer treatment duration. Several studies that have been done over the last decade have looked at predictors of therapeutic failure with short-course therapy (either with single-dose therapy or with 3 days of therapy). Among patient-related factors (table 4), a history of recent UTI, age 1 65, diaphragmspermicide use, and non-Black race have all been related to an increased risk of subsequent failure [1, 3–6]. In at least one study, a higher colony count (i.e., 6105 colony-forming units (cfu)/ml) as compared with infections that were characterized by 102–104 cfu/ml was associated with a higher failure rate [3]. However, this was not confirmed in at least one other study. Antimicrobially resistant organisms, defined either on the basis of MIC or on the basis of species (i.e., S. saprophyticus or non-E. coli UTI, which are typically more resistant to antibiotics) have been associated with higher failure rates and in one study urovirulent E. coli has been associated with higher failure rates. Table 5 lists the regimens that I would regard as optimal for treatment of acute uncomplicated UTI in women. Single-

Stamm

4

PIAM:L01:BIF01002XA.88

FF: W1FP

E1: IN01

$$$

dose regimens certainly work as well, but for the reasons outlined would be second choices, in my mind, for most patients. Few trials have directly compared one 3-day form of therapy versus another. In a recently published study, TMP-SMX when given for 3 days had higher cure rates (82%) than nitrofurantoin (61%), cefadroxil (66%), or amoxacillin (67%) [7]. This may have been due to the fact that prolonged eradication of E. coli from the vaginal flora was achieved to a greater degree with TMP-SMX as compared with the other agents. An important issue in the selection of a treatment regimen for uncomplicated UTI is cost. This study calculated and compared the total cost of each treatment regimen, including costs for bringing patients back to treat failures or costs that were incurred treating side effects. The outcomes were somewhat different than one might have predicted since the TMP-SMX and ofloxacin regimens were the least expensive while the overall costs associated with amoxicillin, nitrofurantoin and cefadroxil were higher. While the drug acquisition costs of these latter agents are much less, the overall cost of management is actually greater because the cure rates were lower and because side effect prevalences were higher with these compounds. Further assessment of the costs and cost effectiveness associated with various treatment regimens is needed. An important current issue is whether short-course therapy with fluoroquinolones should be recommended for uncomplicated UTI. Certainly, fluoroquinolones provide high cure rates that with single-dose therapy approximately equal those with TMP or TMP-SMX. With many fluoroquinolones, 3-day therapy is even more effective [4, 5]. Failures after fluoroquinolone short-course therapy have been observed mainly with S. saprophyticus infection. Some of the fluoroquinolones, based on long periods of high drug concentration in urine, can be used as a once-a-day regimen. The major arguments against fluoroquinolones are their cost and whether their use for uncomplicated UTI may increase the risk of selection of fluoroquinolone resistance organisms. In regions where TMP-SMX resistance is high among uropathogens causing uncomplicated UTI, the fluoroquinolones should probably be used for empiric therapy. In places where TMP and TMP-SMX can still be effectively used, they should perhaps remain the first empiric therapy for uncomplicated UTI in order to avoid fostering emergence of resistance to the fluoroquinolones. In conclusion, studies over the last two decades have shown the lesser effectiveness of single-dose therapy (especially with amoxicillin and the oral cephalosporins) as compared with 3 days of therapy. Single-dose therapy will cure the majority of the patients, but higher cure rates can be achieved with 3-day regimens of TMP, TMP-SMX, and the fluoroquinolones. With these drugs, 3 days of treatment appears to be equally effective to 7 days and more effective than singledose therapy. Side effects are reduced with single-dose therapy or with 3 days of

Treatment of Acute Uncomplicated Urinary Tract Infection

5

PIAM:L01:BIF01002XA.88

FF: W1FP

E1: IN01

$$$

Table 6. Future considerations Develop means to stratify by renal involvement Use of fluoroquinolone for acute UTI: Emergence of resistance Which drug and for how long Better define patient factors predicting treatment failure Identify factors promoting recurrence Importance of subinhibitory effects of antimicrobials Management strategies, cost-effectiveness

therapy with most compounds. The overall antimicrobial effect of the treatment given on the vaginal flora is important in long-term cure and early reinfection from the vaginal reservoir may take place. Considering all of these factors, 3 days of therapy would seem to be best at this point in time and TMP, TMP-SMX and fluoroquinolones would be the recommended current therapy. However, there are relatively few comparative trials of various 3-day therapy regimens. In the future (table 6), it would be very useful to develop a test that actually allows us to stratify patients by renal involvement because we still do not understand what proportion of treatment failures are actually due to occult renal infection versus other factors. Further data are needed to better assess the use of fluoroquinolones for treatment of acute uncomplicated UTI vis-a-vis the frequency of emergence of resistance both in uropathogens and in other flora of the patient being treated. There are few studies to determine which fluoroquinolone should be used and for how long. We need to better define patient factors predicting treatment failure and recurrence. Some antimicrobials, doubtlessly, have effects other than bacterial killing that may play a role in cure, but are, as yet, not very well delineated. Finally, an important area for future studies is different management strategies and their cost effectiveness. For example, should we be utilizing short-course regimens entirely empirically or should such courses be used in conjunction with laboratory testing. If so, should culture and/or urinalysis be used to better identify who is actually infected before providing empiric therapy? These are important questions in the current era of cost consciousness.

Stamm

6

PIAM:L01:BIF01002XA.88

FF: W1FP

E1: IN01

$$$

References 1 2 3

4 5 6

7

Stamm WE, Hooton TM: Management of urinary tract infections in adults. N Engl J Med 1993; 329:1328–1334. Ronald AR, Boutros P, Mourtada H: Bacteriuria localization and response to single-dose therapy in women. JAMA 1976;235:1854–1856. Fihn SD, Johnson C, Roberts PL, Running K, Stamm WE: Trimethoprim-sulfamethoxazole for acute dysuria in women: a single-dose or 10-day course: A double-blind, randomized trial. Ann Intern Med 1988;108:350–357. Norrby SR: Short-term treatment of uncomplicated lower urinary tract infections in women. Rev Infect Dis 1990;12:458–467. Leibovici L, Wysenbeek AJ: Single-dose antibiotic treatment for symptomatic urinary tract infections in women: A meta-analysis of randomized trials. Q J Med 1991;285:43–57. Iravani A, Tice AD, McCarty J, et al: Short-course ciproflox treatment of acute uncomplicated urinary tract infection in women. The minimum effective dose. Arch Intern Med 1995;155:485– 494. Hooton TM, Winter C, Tiu F, Stamm WE: Randomized comparative trial and cost analysis of 3-day antimicrobial regimens for treatment of acute cystitis in women. JAMA 1995;273:41–45.

Prof. Dr. Walter E. Stamm, Division of Allergy and Infectious Diseases, University of Washington School of Medicine, M/C Box 356523, 1959 NE Pacific Street, Seattle, WA 98195 (USA)

Treatment of Acute Uncomplicated Urinary Tract Infection

7

PIAM:L01:BIF01003XA.85

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 8–13

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Management of Acute Uncomplicated Pyelonephritis Lindsay E. Nicolle Section of Infectious Diseases, Departments of Medicine and Medical Microbiology, University of Manitoba, Winnipeg, Man., Canada

Acute uncomplicated (nonobstructive) pyelonephritis is a clinical syndrome characterized by fever, chills, flank pain or tenderness and a positive urine culture (6104 cfu/ml) with associated pyuria [1]. This syndrome occurs virtually only in women, and most frequently in those from 18 to 40 years of age. Women who experience acute nonobstructive pyelonephritis also usually experience acute uncomplicated urinary infection [2], and the same genetic and behavioral factors are likely associated with both syndromes. Acute uncomplicated pyelonephritis is a common disease. Despite this, there are virtually no population-based studies which describe the incidence. In addition, the relative risk for selected subgroups, such as diabetics, who may have both a greater incidence as well as severity of disease, is not well established [3]. Descriptive information provided in case series of women presenting with pyelonephritis is provided in table 1 [4–10].

Microbiology Escherichia coli is the infecting organism isolated from 85 to 90% of episodes of acute uncomplicated pyelonephritis [2, 9, 10]. For the small number of episodes from which other organisms are isolated, Klebsiella pneumoniae, Staphylococcus saprophyticus and Proteus mirabilis are the usual infecting organisms. From 70 to 100% of E. coli isolated from episodes of acute uncomplicated pyelonephritis have phenotypic expression of P-fimbriae [Gal(·1-4)Gal pilus] [11]. This virulence factor occurs at substantially lower rates in E. coli isolated from

PIAM:L01:BIF01003XA.85

FF: W1FP

E1: IN01

$$$

Table 1. Reported case series of pyelonephritis Reference

Country

Locale

Subjects

Diabetic

Complicated

Bacteremia

Hospitalized

n

%

n

%

n

%

n

101 E. coli only 61

8

8

16

16

27

27

NS

6

10

27

44

9

15

‘Most’

90

NS

NS

2/61

NS

NS

15/104 14

%

Otto et al. [4]

Sweden

Jernelius et al. [5]

Sweden

Ward et al. [6] Israel et al. [7] Ikaheimo et al. [8] Safrin et al. [9] Pinson et al. [10]

USA

University ambulatory care NS Randomized trial Emergency

USA

Emergency

147

Finland

University hospital General hospital Emergency university hospital

49

4

8

8

16

15

31

NS

194

17

9

14

7

21

11

95/173 55

111

10

9

21

19

10/32

31

25

USA USA

3.21

31 60

NS = Not stated. 1 Only subjects not initially hospitalized.

other clinical syndromes of urinary infection, including asymptomatic bacteriuria, cystitis, and complicated pyelonephritis. Other potential E. coli virulence factors such as hemolysin, aerobactin production, and selected serotypes may occur with increased frequency in pyelonephritis strains, but are not as consistently associated with this clinical syndrome as P-fimbria expression [11]. The specific mechanism by which P-fimbriae mediate acute pyelonephritis is not known, although expression of this pilus is associated with a greater inflammatory response [12].

Host Response Acute uncomplicated pyelonephritis is a systemic illness characterized by both a local urinary and systemic immune and inflammatory response. Serum IL-6, IL-8 and C-reactive protein are elevated [13, 14] and a systemic antibody increase to diverse antigenic components of the infecting uropathogen occurs [15, 16]. The local urinary response includes pyuria, elevated urinary IL-6, IL-8 and

Pyelonephritis

9

28

PIAM:L01:BIF01003XA.85

FF: W1FP

E1: IN01

$$$

Table 2. Management issues for consideration in acute uncomplicated pyelonephritis Selection of antimicrobial therapy Duration of antimicrobial therapy Decision for hospitalization Investigation of potential genitourinary abnormalities

other cytokines [13, 14], as well as a local urine antibody response. The extent to which the local or systemic humoral response may protect against subsequent infection is not well studied.

Management of Acute Uncomplicated Pyelonephritis Most women with acute uncomplicated pyelonephritis are otherwise healthy and respond promptly to appropriate antimicrobial therapy. There is limited short-term and virtually no long-term morbidity. A number of management issues, however, warrant further consideration (table 2). Selection of Antimicrobial Therapy Antimicrobial therapy for acute uncomplicated pyelonephritis may be either oral or parenteral. Frequently, initial parenteral therapy is given as a single dose or as long as 48–72 h, then changed to oral therapy to complete the antimicrobial course once the patient is clinically stable or improving [17]. There are few comparative studies of outcome with oral therapy alone. Trimethoprim/sulfamethoxazole was superior to amoxicillin in curing infection in one study, with cure rates of 88 and 56%, respectively [18]. Several studies have reported that quinolones are superior to cephalosporins for oral therapy of pyelonephritis [19]. Trimethoprim/sulfamethoxazole or quinoline antimicrobials should be considered current drugs of choice for oral therapy. ß-Lactam antibiotics should be avoided, if possible. For parenteral therapy, an aminoglycoside with or without ampicillin remains standard therapy [20]. Many other parenteral antimicrobials are effective, including trimethoprim/sulfamethoxazole, quinolones and several cephalosporins. There are few comparative studies, however, which have specifically measured the relative efficacy, benefits, adverse events, and costs of different parenteral regimens.

Nicolle

10

PIAM:L01:BIF01003XA.85

FF: W1FP

E1: IN01

$$$

Duration of Therapy The current recommended duration of therapy is for 10–14 days [17]. In a study of pivmecillinam/pivampicillin, bacteriologic cure was achieved in only 28% of patients with 1 week of therapy compared with 69% with 3 weeks [5]. About one third of patients enrolled in this study had underlying complicating factors. In a comparative study of duration of oral regimens, 6 weeks of trimethoprim/sulfamethoxazole therapy was not more efficacious than 2 weeks for the treatment of pyelonephritis in women, with cure rates of 83 and 90%, respectively [18]. There have been reports of successful treatment of pyelonephritis with therapy duration as short as 5 days [21]. These much shorter regimens require further evaluation in comparative studies with greater study numbers enrolled before they can be enthusiastically recommended. Hospitalization or Outpatient Management? There is substantial discretionary judgement in the decision to hospitalize women presenting with acute uncomplicated pyelonephritis [9]. Currently, hospitalization is recommended for patients with hemodynamic instability, pregnant women, when there is diagnostic uncertainty, and where compliance is not assured. From 50 to 70% of women presenting with the syndrome of acute pyelonephritis can be managed without hospitalization [6, 7, 9, 10]. A common current approach is to treat women in the emergency department or observation unit with an initial short course of parenteral therapy [6, 7], with the decision about need for further parenteral therapy or hospitalization made after the initial 24 h of therapy and observation. Case series have all reported over 90% cure rates and no excess morbidity in nonhospitalized women managed with this approach [6, 7, 9, 10]. Hospitalization is not necessary for most women, and physician judgement with respect to the need for hospitalization seems appropriate in reported series. The present climate promoting health care cost minimization and deinstitutionalization of care suggests the outpatient management of pyelonephritis should be further explored to ensure optimal use of this option. Investigation for Underlying Genitourinary Abnormalities Women with acute uncomplicated pyelonephritis have, by definition, a normal genitourinary tract. However, individuals with underlying structural or functional abnormalities of their genitourinary tract may also present with the clinical syndrome of pyelonephritis. Thus, some individuals may require investigation to identify abnormalities which may be corrected surgically to prevent recurrence of infection. There is no current consensus on what characteristics require genitourinary investigation. Reported discriminatory factors to identify women with underlying

Pyelonephritis

11

PIAM:L01:BIF01003XA.85

FF: W1FP

E1: IN01

$$$

abnormalities include: fever beyond 72 h of appropriate antimicrobial therapy [22], and early relapse or infection following therapy [23]. Early investigation is likely warranted in any individual who presents with a clinical syndrome of acute pyelonephritis with septic shock. These individuals have an increased likelihood of obstruction and may require early and aggressive intervention. Women who are hemodynamically stable at presentation and respond promptly to antimicrobial therapy without recurrence are unlikely to have underlying genitourinary abnormalities and do not warrant further investigation [17]. Where investigations are appropriate, the safest, most cost-effective, initial test is likely an ultrasound examination. This is a rapid, noninvasive, and reliable method for determining the presence of obstruction and other abnormalities. If the ultrasound is normal, further investigations are not indicated unless the clinical course remains atypical.

References 1 2 3 4

5 6 7 8

9 10 11 12 13

14

Rubin RH, Shapiro ED, Andriole VT, Davis RJ, Stamm WE: Evaluation of new anti-infective drugs for the treatment of urinary tract infection. Clin Infect Dis 1992;15(suppl 1):216–227. Stamm WE, McKevitt M, Roberts P, White NJ: Natural history of recurrent urinary tract infections in women. Rev Infect Dis 1991;13:77–84. Patterson JE, Andriole VT: Bacterial urinary tract infections in diabetics. Infect Dis Clin North Am 1995;9:25–51. Otto G, Sandberg T, Marklund B-I, Ulleryd P, Svanborg C: Virulence factors and pap genotype in Escherichia coli isolates from women with acute pyelonephritis with or without bacteremia. Clin Infect Dis 1993;17:448–456. Jernelius H, Zbornick J, Bauer C-A: One or three weeks’ treatment of acute pyelonephritis. Acta Med Scand 1988;223:469–477. Ward G, Jordon RC, Severana HW: Treatment of pyelonephritis in an observation unit. Ann Emerg Med 1991;20:258–261. Israel RS, Lowenstein SR, Marx JA, Mazio-McLain J, Svobodia L, Ramger S: Management of acute pyelonephritis in an emergency department observation unit. Ann Emerg Med 1991;20:253–257. Ikaheimo R, Siitonen A, Karkkainen U, Mustonen J, Heiskanen T, Makela TH: Communityacquired pyelonephritis in adults: Characteristics of E. coli isolates in bacteremic and nonbacteremic patients. Scand J Infect Dis 1994;26:289–296. Safrin S, Seigil D, Black D: Pyelonephritis in adult women: Inpatient versus outpatient therapy. Am J Med 1988;85:793–798. Pinson AG, Philbrick JT, Lindbeck GH, Schorling JB: ED management of acute pyelonephritis in women: A cohort study. Am J Emerg Med 1994;12:271–278. Johnson E Jr: Virulence factors in Escherichia coli urinary tract infection. Clin Microbiol Rev 1991; 4:80–128. de Man P, Jodal U, Lincoln K, Svanborg Eden C: Bacterial attachment and inflammation in the urinary tract. J Infect Dis 1988;158:29–35. Hedges S, Stenquist K, Lidin-Janson G, Martinell J, Sandberg T, Svanborg C: Comparison of urine and serum concentration of interleukin-6 in women with acute pyelonephritis or asymptomatic bacteriuria. J Infect Dis 1992;166:653–656. Jacobson SH, Hylander B, Wretlind B, Brauner A: Interleukin-6 and interleukin-8 in serum and urine in patients with acute pyelonephritis in relation to bacterial. Virulence-associated traits and renal function. Nephron 1994;67:172–179.

Nicolle

12

PIAM:L01:BIF01003XA.85

15 16 17 18 19 20

21 22 23

FF: W1FP

E1: IN01

$$$

Sobel JD: Pathogenesis of urinary tract infections: Host defences. Infect Dis Clin North Am 1987;1: 751–772. Nicolle LE, Brunka J, Ujack E, Bryan L: Antibodies to major outer membrane proteins of Escherichia coli in urinary infection in the elderly. J Infect Dis 1989;160:627–633. Stamm WE, Hooton TM: Management of urinary tract infections in adults. N Engl J Med 1993; 329:1328–1334. Stamm WE, McKevitt M, Counts GW: Acute renal infection in women: Treatment with trimethoprim-sulfamethoxazole or ampicillin for two or six weeks. Ann Intern Med 1987;106:341–345. Pinson AG, Philbrick JT, Lindbeck GH, Schorling JB: Oral antibiotic therapy for acute pyelonephritis: A methodologic review of the literature. J Gen Intern Med 1992;7:544–553. Johnson JR, Lyons MF II, Pearce W, Gorman P, Roberts PL, White N, Brust P, Olsen R, Gnann JW Jr, Stamm WE: Therapy for women hospitalized with acute pyelonephritis. A randomized trial of ampicillin versus trimethoprim-sulfamethoxazole for 14 days. J Infect Dis 1991;163:325–330. Bailey RR, Lynn KL, Robson RA, Peddie BA, Smith A: Comparison of ciprofloxacin with netilmicin for the treatment of acute pyelonephritis. N Z Med J 1992;105:102–103. Kanel KT, Kroboth FJ, Schwentker FN, Lecky JW: The intravenous pyelogram in acute pyelonephritis. Arch Intern Med 1988;148:2144–2148. Sanberg T, Stokland E, Brolin I, Lidin-Janson G, Svanborg Eden C: Selective use of excretory urography in women with acute pyelonephritis. J Urol 1989;141:1290–1294.

Dr. Lindsay E. Nicolle, Health Sciences Centre, MS675D-820 Sherbrook Street, Winnipeg, Man. R3A 1R9 (Canada)

Pyelonephritis

13

BOIS:L01:BIF01004XA.88

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 14–18

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Uncomplicated Acute Pyelonephritis Ross R. Bailey Department of Nephrology, Christchurch Hospital, Christchurch, New Zealand

A large number of prospective, controlled studies have been conducted on the treatment of women with bacterial cystitis, as well as on patients with various categories of complicated urinary tract infections. There has, however, been a dearth of studies in women with uncomplicated acute pyelonephritis, particularly those requiring parenteral antimicrobial therapy. Such patients are predominantly young, sexually active women who frequently require hospitalization, because of the severity of their clinical illness and an inability to tolerate oral medication. With appropriate treatment, the majority of patients with uncomplicated acute pyelonephritis respond rapidly to parenteral antimicrobial therapy, rehydration and pain relief and are usually well enough to be discharged home within 2–3 days. In our own unit during the 3-year period 1992–1994, 163 patients were admitted with acute pyelonephritis, of whom 144 (88%) were women. Of these 144 women, 126 (88% of the women and 77% of the total) had uncomplicated acute pyelonephritis (fig. 1) [1]. There is currently some debate as to whether all patients with acute pyelonephritis warrant urinary tract investigation. We believe that all patients hospitalized because of acute pyelonephritis warrant ultrasonography of the urinary tract, primarily to exclude urinary tract obstruction. This is clearly essential for all such patients enrolled in prospective treatment studies so as to enable the acute pyelonephritis to be classified as uncomplicated or complicated. Of great interest in patients with acute pyelonephritis are the findings on DMSA scan during the acute phase. In our own department, 37 of 81 (46%) consecutive women hospitalized with acute pyelonephritis had one or more perfusion

BOIS:L01:BIF01004XA.88

FF: W1FP

E1: IN01

$$$

➙

Fig. 1. A suggested algorithm for the management of women with suspected acute pyelonephritis (APN).

defects on a DMSA scan done on, or shortly after admission. The majority of these defects had resolved within 3 months. Patients in whom the perfusion defect had persisted subsequently had an intravenous urogram. The latter invariably demonstrated the characteristic features of either reflux nephropathy or obstructive nephropathy which had clearly predated the episode of acute pyelonephritis. Perhaps patients enrolled in prospective treatment studies of acute pyelonephritis should have a DSMA scan at enrollment?

Uncomplicated Acute Pyelonephritis

15

BOIS:L01:BIF01004XA.88

FF: W1FP

E1: IN01

$$$

Table 1. Prospective, randomized treatment studies of acute pyelonephritis Authors

Drugs

Duration days

Enrolled

14 14

Cured n

%

44 41

25/27 35/36

93 97

Johnson et al. [14]

Ampicillin/gentamicin CTM/gentamicin

Bailey et al. [5]

Netilmicin Ciprofloxacin

5 5

22 21

15/17 15/17

88 88

Mouton et al. [15]

Lomefloxacin CTM

14 14

33 30

20/20 16/18

100 89

CTM = Co-trimoxazole. Data from MEDLINE via CD Plus from January 1991 to March 1995.

In our unit, we have undertaken a series of controlled therapeutic studies in hospitalized patients with acute pyelonephritis [2–6]. The duration of treatment has been 5 days and the results, especially with the aminoglycosides, have been excellent. Some reviewers, however, have ignored these studies and have continued to promote much longer courses of treatment which have little scientific basis and have been handed down as historical dogma following the repetitious citing of anecdotal reports. The authoritative General Guidelines for the Evaluation for New Anti-Infective Drugs for the Treatment of Urinary Tract Infection have recommended a 2-week oral course of treatment [7]. This conclusion, however, appeared to come from one single study [8] where the patients were clearly not characteristic of those with acute pyelonephritis as they were able to tolerate oral antimicrobial therapy and did not warrant hospitalization. Stamm and Hooton [9] in discussion stated that ‘shorter regimens (e.g., 5–7 days) are often effective in patients whose fever abates rapidly, but they have not been evaluated in wellcontrolled trials’. Pinson et al. [10] undertook a literature review from 1965 to 1991 and reviewed 10 randomized studies, 9 of which compared oral drugs and the tenth compared an oral with an intravenously administered drug. Once again, these were not studies which enrolled the typical patient who requires a short period of hospitalization [11]. Pinson et al. [10] stated that ‘until further data is available, we conclude that patients with acute pyelonephritis should be treated for 10–14 days’. The duration of therapy for patients with uncomplicated acute pyelonephritis was reviewed briefly by this discussant at the 2nd International Symposium on

Bailey

16

BOIS:L01:BIF01004XA.88

FF: W1FP

E1: IN01

$$$

Clinical Evaluation of Drug Efficacy in UTI [12], and in more detail at the 3rd International Symposium [13]. The same arguments in favor of shorter curative regimens of treatment still apply. To assess what new science had become available since the latter symposium a Medline search was undertaken on 06 April 1995 for all controlled treatment studies of acute pyelonephritis published between January 1991 and March 1995. Only 3 prospective, randomized trials were published and these are summarized in table 1 [5, 14, 15]. Two of these studies had unacceptably high numbers of dropouts, mainly because of patients infected with resistant pathogens or of failure to return for follow-up. From the literature search, there were several other articles proposing intravenous/oral switch therapy to enable an earlier discharge from hospital for patients with acute pyelonephritis [16, 17]. Our own studies have shown that by using intravenous/ oral switch therapy these patients can be discharged earlier from hospital [5, 6]. This is the greatest possible saving of costs that can be made in the management of acute pyelonephritis [18].

References 1 2 3

4 5 6

7 8

9 10 11

12

Bailey RR, Lynn KL, Robson RA, Smith AH, Maling TMJ, Turner JG: DMSA renal scans in adults with acute pyelonephritis. Clin Nephrol 1996;46:99–104. Bailey RR, Lynn KL, Peddie BA, Swainson CP: Comparison of netilmicin with cefoperazone for the treatment of severe or complicated urinary tract infections. Aust N Z J Med 1985;15:22–26. Bailey RR, Lynn KL, Peddie BA, Walker RJ, Swainson CP: Comparison of netilmicin with ceftriaxone for the treatment of severe or complicated urinary tract infections. N Z Med J 1986;99:459– 461. Bailey RR, Lynn KL, Robson RA, Peddie BA: Comparison of aztreonam (Azactam) and netilmicin in the treatment of severe or complicated urinary tract infections. Clin Trials J 1989;26:288–294. Bailey RR, Lynn KL, Robson RA, Peddie BA, Smith A: Comparison of ciprofloxacin with netilmicin for the treatment of acute pyelonephritis. N Z Med J 1992;105:102–103. Bailey RR, Begg EJ, Smith AH, Robson RA, Lynn KL, Chambers ST, Barclay ML, Hornibrook J: Prospective, randomized, controlled study comparing two dosing regimens of gentamicin/oral ciprofloxacin switch therapy for acute pyelonephritis. Clin Nephrol 1996;46:183–186. Rubin RH, Shapiro ED, Andriole VT, Davis RJ, Stamm WE: Evaluation of new anti-infective drugs for the treatment of urinary tract infection. Clin Infect Dis 1992;15(suppl 1):216–227. Stamm WE, McKevitt M, Counts GW: Acute renal infection in women: Treatment with trimethoprim-sulfamethoxazole or ampicillin for two or six weeks: A randomized trial. Ann Intern Med 1987;106:341–345. Stamm WE, Hooton TM: Management of urinary tract infections in adults. N Engl J Med 1993; 329:1328–1334. Pinson AG, Philbrick JT, Lindbeck GH, Schorling JB: Oral antibiotic therapy for acute pyelonephritis: A methodologic review of the literature. J Gen Intern Med 1992;7:544–553. Bailey RR: Uncomplicated acute pyelonephritis in women; in Weissenbacher ER, Jackel C (eds): International Standard Book for Infectious Diseases in Obstetrics, Gynecology, Dermatology, Urology and Clinical Immunology. Munich, Medifact, 1996. Bailey RR: Treatment of uncomplicated acute pyelonephritis; in Ohkoshi M, Naber KG (eds): International Consensus Discussion on Clinical Evaluation of Drug Efficacy in Urinary Tract Infection. Infection 1992;20:S176–S177.

Uncomplicated Acute Pyelonephritis

17

BOIS:L01:BIF01004XA.88

13

14

15

16 17

18

FF: W1FP

E1: IN01

$$$

Bailey RR: What should the duration of antimicrobial treatment be for patients with uncomplicated acute pyelonephritis (Round Table Discussion); in Ohkoshi M, Naber KG, Kawada Y (eds): Duration of Antimicrobial Treatment and the Use of Drug Combinations for the Treatment of Uncomplicated Acute Pyelonephritis. Infection 1994;22:S50–S52. Johnson JR, Lyons MF, Pearce W, et al: Therapy for women hospitalized with acute pyelonephritis: A randomized trial of ampicillin versus trimethoprim/sulfamethoxazole for 14 days. J Infect Dis 1991;163:325–330. Mouton Y, Ajana F, Chidiac C, Capron MH, Home P, Masquelier AM: A multicenter study of lomefloxacin and trimethoprim/sulfamethoxazole in the treatment of uncomplicated acute pyelonephritis. Am J Med 1992;92:87S–90S. Ward G, Jorden RC, Severance HW: Treatment of pyelonephritis in an observation unit. Ann Emerg Med 1991;20:258–261. Caceres VM, Stange KC, Kikano GE, Zyzanski SJ: The clinical utility of a day of hospital observation after switching from intravenous to oral antibiotic therapy in the treatment of pyelonephritis. J Fam Pract 1994;39:337–340. Safrin S, Siegel D, Black D: Pyelonephritis in adult women: Inpatient versus outpatient therapy. Am J Med 1988;85:793–798.

Prof. Ross R. Bailey, Department of Nephrology, Christchurch Hospital, Christchurch (New Zealand)

Bailey

18

PIAM:L01:BIF01005XA.85

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 19–26

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Complicated Urinary Tract Infections Joichi Kumazawa, Tetsuro Matsumoto Department of Urology, Faculty of Medicine, Kyushu University, Fukuoka, Japan

Diagnosis of urinary tract infections (UTI) is made on the bases of symptoms, pyuria, bacteriuria, blood analysis, and radiologic or ultrasound examinations [1]. Pyuria and bacteriuria are quite important not only for the diagnosis, but also for the evaluation of treatment efficacy. In Japanese criteria for the evaluation of drug efficacy, pyuria and bacteriuria are now changing in order to harmonize with American and European guidelines.

Predisposing Factors of Complicated UTI Japanese urologists classify UTI into acute uncomplicated cystitis, acute uncomplicated pyelonephritis, complicated UTI including cystitis and pyelonephritis, bacterial prostatitis, and urethritis. Complicated UTI entails urinary infections combined with predisposing factors which decrease antibacterial defense mechanisms of the urinary tract. Predisposing factors are indwelling catheter, functional and anatomical abnormalities in the urinary tract, including bladder outlet obstruction, urolithiasis, hydronephrosis, vesicoureteric reflux, urogenital cancer, neurogenic bladder dysfunction, foreign bodies, chemical or radiation injuries of uroepithelium and indwelling catheter. The first approach to control complicated UTI is to precisely diagnose the predisposing factors, and treat or control them with urologic surgical intervention. American and European guidelines add as predisposing factors the presence of indwelling catheter, intermittent catherization, over 100 ml residual urine, obstructive uropathy, vesicoureteral reflux, ileal loops, azotemia and renal transplantation [2].

PIAM:L01:BIF01005XA.85

FF: W1FP

E1: IN01

$$$

Table 1. Difference of criteria for predisposing factors of complicated UTI in American and Japanese guidelines IDSA/FDA guidelines

Japanese criteria

Presence of indwelling catheter Intermittent catheterization Residual urine Obstructive uropathy VUR or other urologic abnormalities Ileal loop Azotmeia Renal transplantation

Presence of indwelling catheter Neurogenic bladder dysfunction Obstructive uropathy VUR or other urologic abnormalities Chemical or radiation cystitis

IDSA = Infectious Diseases Society of America; FDA = Food and Drug Administration; VUR = vesicoureteral reflux.

Table 2. Difference of entry criteria in complicated UTI IDSA guidelines

European guidelines

New Japanese guidelines

Age

Not fixed

Not fixed

620 years

Presence of symptoms

Dysuria Urgency Frequency Suprapubic pain Fever Flank pain CVA tenderness

Same as IDSA/FDA guidelines

Not specified

Underlying disease

Indwelling catheter Intermittent catheterization Residual urine (1 100 ml) Obstructive uropathy VUR or urologic abnormality Azotemia Renal transplantation

Same as IDSA/FDA guidlines

Urinary underlying disease

Pyuria, WBC/mm3

610

610

610

Bacteriuria, cfu/ml

6105

6105

1 104 or 105

CVA = Cost vertebral angle; VUR = vesicoureteral reflux.

Kumazawa/Matsumoto

20

PIAM:L01:BIF01005XA.85

FF: W1FP

E1: IN01

$$$

Fig. 1. Patient age distribution of complicated UTI in Japan.

The Japanese UTI committee criteria listed as underlying factors of complicated UTI the presence of indwelling catheter, neurogenic bladder dysfunction, obstructive uropathy, vesicoureteral reflux, chemical or radiation injuries of the uroepithelium. Azotemia and renal transplantation should not be included among the underlying diseases, because they involve general impairment of defense mechanism of infection such as impaired phagocytes and/or lymphocytes (table 1). Such impairment is frequently severe and the evaluation of the efficacy of antimicrobial drugs may be difficult. Japanese criteria exclude diabetes mellitus, when it is not adequately controlled, because it also entails impaired phagocytic function. Furthermore, urinary tract modifications such as ileal loop or colon pouch are also excluded, because bacteriuria and pyuria always accompany such conditions.

Patients’ Background of Complicated UTI Table 2 demonstrates the differences in the inclusion criteria between American, European and Japanese guidelines. Patient age is not part of the inclusion criteria in the American or European guidelines. Japanese criteria included patients who are 20 years or older. This is due to patient age distribution of complicated UTI (fig. 1). Very few patients have been assessed in recent trials of com-

Complicated UTI

21

PIAM:L01:BIF01005XA.85

FF: W1FP

E1: IN01

$$$

Fig. 2. Complicated UTI in patients with ($) and without ()) symptoms.

plicated UTIs in Japan. Symptoms are frequently mild or lacking in patients with complicated UTI (fig. 2). Symptoms should not be specified at evaluation of drug efficacy in complicated UTI. Figure 3 shows distribution of bacteria isolated from patients with uncomplicated and complicated cystitis in our hospital. The incidence of Escherichia coli differs in uncomplicated versus complicated cystitis. Entercoccus faecalis and Pseudomonas aeruginosa are frequently isolated from complicated cystitis. Complicated pyelonephritis is caused by various kinds of bacteria such as E. coli, Citrobacter spp., Proteus spp. and P. aeruginosa. Some isolates are highly resistant to several antimicrobials. In addition, biofilm is thought frequently to cover the urinary epithelium. A biofilm is frequently present around indwelling catheters, ureteral stents and struvite stones. Antimicrobial agents used to treat UTI do not become accumulated inside the urinary tract or loci of inflammation in part due to renal dysfunction, abscess formation, scarring and biofilm formation.

Treatment of Complicated UTI Treatment and control of complicated UTI is difficult because of predisposing factors in the urinary tract or the causative bacteria which are frequently associated with a mixed infection and/or resistance to several antimicrobial agents. Until predisposing factors are completely resolved, true cure is not possible in

Kumazawa/Matsumoto

22

PIAM:L01:BIF01005XA.85

FF: W1FP

E1: IN01

$$$

Fig. 3. Bacteria isolated from patients with (a) uncomplicated (left) and complicated (right) cystitis and (b) uncomplicated (left) and complicated (right) pyelonephritis.

complicated UTI. Various kinds of antimicrobial agents including cephems, carbapenems and fluoroquinolones have been developed and became commercially available during recent years. These drugs have a higher antimicrobial activity and a wider antimicrobial spectrum. Combination therapy with several antimicrobials may effectively cure complicated UTI. Duration of therapy should be

Complicated UTI

23

PIAM:L01:BIF01005XA.85

FF: W1FP

E1: IN01

$$$

Table 3. New treatment for complicated UTI

Newer antimicrobial agents Cephems, carbapenems, fluoroquinolones Combination therapy ß-Lactamase inhibitors + ß-lactams Aminoglycosides + ß-lactams Macrolides + fluoroquinolones

short term to prevent emergence of resistant strains. Treatment of complicated UTI is difficult when underlying diseases continue. Acute exacerbations of complicated pyelonephritis, evidenced by symptoms such as high-grade fever and flank pain, should be treated by parenteral drugs until the body temperature is decreased. Almost all patients with complicated UTI are treated by oral drugs. Combination therapy with ß-lactamase inhibitiors and ß-lactams, aminoglycoside and ß-lactams, and macrolides and fluoroquinolones have effectively treated complicated UTI (table 3). These new strategies allow short-term therapy and render good clinical results in complicated UTI. Newer cephems with stronger activities against Gram-positive bacteria have been developed recently. Thus, cefpirome exhibits therapeutically relevant activity against, e.g., enterococci. We have compared cefpirome to ceftazidime. Clinical effectiveness was similar when the same doses were compared. The eradication rate of Gram-positive bacteria, especially of enterococci, was significantly higher in the cefpirome group. We found cefpirome more suitable for complicated UTI [3]. Among carbapenem antibiotics, panipenem was developed in Japan. This is slightly more active against Gram-positive bacteria than imipenem. Panipenem showed a slightly higher efficacy rate than imipenem in a comparative study on complicated UTI. The bacterial eradication rates of various bacterial species were similar for imipenem and panipenem. Carbapenem antibiotics are thought to be suitable for the treatment of acute relapse of complicated pyelonephritis [4]. Various kinds of new quinolone antimicrobials have been developed recently. Long half-life antimicrobials like fleroxacin and sparfloxacin were developed in Japan. Levofloxacin, an active optical isomer of ofloxacin, was developed in Germany. These quinolones were suitable for complicated UTI without fever. Sparfloxacin is more active against various kinds of urinary isolates, but its absorption and urinary excretion were less. The antimicrobial activities of fleroxacin and ofloxacin were similar. Its superiority was a longer half-life and a higher urinary excretion. Once-a-day therapy was equally effective in complicated UTI.

Kumazawa/Matsumoto

24

PIAM:L01:BIF01005XA.85

FF: W1FP

E1: IN01

$$$

Levofloxacin is twice as active in vitro as ofloxacin. Once-a-day treatment of 300 mg sparfloxacin or fleroxacin were as effective as 600 mg of enoxacin in a comparative study on complicated UTI. A comparative study of levofloxacin and ofloxacin showed the same results for doses of 300 and 600 mg. Side effects and abnormal laboratory results occurred at a low rate in the levofloxacin group [5–7]. We tried a higher dose of levofloxacin in catheter-associated UTI. A 600-mg dose of levofloxacin was more effective in this type of infection whereas its safety was similar to the 300-mg dose [8, 9]. Combination chemotherapy has frequently been tried in serious infections. One of the best combinations was aminoglycosides and ß-lactams. Isepamicin as an aminoglycoside and piperacillin as a ß-lactam were frequently used to treat complicated catheter-associated UTI. Combination therapy of isepamicin and piperacillin proved more active than monotherapy. The order of administration was thought to be important to treat complicated UTI. We compared three dosing regimens, namely isepamicin prior to piperacillin, concomitant administration, and piperacillin prior to isepamicin in catheter-associated UTI. The most effective regimen was isepamicin prior to piperacillin, indicating the superior first exposure effect of aminoglycoside [9]. Recently, some macrolides have had a preventive or destructive effect on biofilm formation. Combination chemotherapy of macrolides and fluoroquinolones have been known to be quite effective in infections related to biofilm formation. Ciprofloxacin combined with clarithromycin was significantly effective in complicated UTI. New antimicrobial agents have been more active and have a wider spectrum in various bacterial species. These were quite effective in complicated UTI as in other types of infection. However, strains resistant to these antimicrobials emerged in urinary isolates. We will use these antimicrobials shortly and select effective dosing regimens. Combination chemotherapy is one effective form of treatment in complicated UTI. We will also try more active usage of these antimicrobials, remembering that the cure of complicated UTI is to treat completely the predisposing factors of the urinary tract.

References 1 2

3

Ohkoshi M: Criteria for evaluation of clinical efficacy of antimicrobial agents on urinary tract infection, ed 3. Jpn J Antibiot 1987;40:2149–2191. Rubin RH, Shapiro ED, Andriole VT, Stamm WE: General guideline for the evaluation of new anti-infective drugs for the treatment of urinary tract infection. Clin Infec Dis 1992;15(suppl 1): 216–227. Naide Y, Aso Y, Oshi M, Kumamoto Y, Hirose T, et al: Comparative study of cefpirome and ceftazidime in complicated urinary tract infections. Hinyoukika Kiyo 1991;37:447–464.

Complicated UTI

25

PIAM:L01:BIF01005XA.85

4

5 6

7 8

9

FF: W1FP

E1: IN01

$$$

Kumazawa J, Matsumoto T, Kumamoto Y, Hirose T, Aso Y, et al: Phase III comparative clinical trial of panipenem/betamipron with imipenem/cilastatin sodium for the treatment of complicated urinary tract infections. Nishinihon J Urol 1992;54:254–271. Kawada Y, Kumamoto Y, Aso Y, Oshi M, Machida T, et al: Comparative study of sparfloxacin and enoxacin in complicated urinary tract infections. Chemotherapy 1991;39(suppl 4):571–588. Kawada Y, Kumamoto Y, Orikasa S, Aso Y, Machida T, Saito I, et al: Comparative study on fleroxacin and ofloxacin in complicated urinary tract infections. Chemotherapy 1990;38(suppl 2): 571–590. Kawada Y, Kumamoto Y, Aso Y, Machida T, Saito I, et al: Comparative study on levofloxacin and ofloxacin in complicated urinary tract infections. Chemotherapy 1992;40(suppl 3):230–248. Matsumoto T, Ogata N, Mizunoe Y, Kumazawa J, Tahara H, et al: Clinical study of two dosing regimens of levofloxacin in catheter-associated complicated urinary tract infections and the effects of catheter exchange. Nishinihon J Urol 1992;54:299–310. Matsumoto T, Kumazawa J, Nishimura N, Kumamoto Y, Arakawa S, et al: Combination chemotherapy with isepamicin and piperacillin; therapeutic effect of three dosing regimens. Int J Exp Clin Chemother 1995;6:111–117.

Dr. Tetsuro Matsumoto, Department of Urology, Faculty of Medicine, Kyushu University, 3-1-1, Maidashi, Higashi-ku, Fukuoka 812 (Japan)

Kumazawa/Matsumoto

26

PIAM:L01:BIF01006XA.89

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 27–33

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infection in the Renal Transplant Recipient Nina E. Tolkoff-Rubin a, Robert H. Rubin b a

b

Hemodialysis and CAPD Units, Massachusetts Gernal Hospital, and Harvard Medical School, Boston, Mass., and Transplantation Infectious Disease, Massachusetts General Hospital; Center for Experimental Pharmacology and Therapeutics, Harvard-MIT Division of Health Sciences and Technoloy, and Gordon and Marjorie Osborne Chair of Health Sciences and Technology, Cambridge, Mass., USA

Urinary tract infection (UTI) has traditionally been the most common cause of bacterial and fungal infection in the renal transplant recipient, occurring in 35–79% of allograft recipients not receiving antimicrobial prophylaxis [1–8]. The consequences of such infections have been considerable, with 60% of the Gram-negative bacteremias that have occurred in these patients being due to UTI, and with at least some evidence linking such infections to allograft injury that persists long after eradication of the microbial invaders with antimicrobial therapy [1, 9, 10]. Whereas in the early days of renal transplantation, technical complications with the ureteral anastomosis and residual infection in native kidneys played an important role in the pathogenesis of UTI, today these problems have been largely eliminated – by better surgery and eradication of infection pretransplant. However, the incidence of UTI in the absence of prophylaxis remains in the 35–45% range [1, 8, 11]. The purpose of this review is to present current concepts of the pathogenesis, clinical impact, prevention, and treatment of UTI in the renal transplant recipient; an entitiy of continuing clinical importance.

Pathogenesis of UTI in the Renal Transplant Recipient As with all infectious diseases, the pathogenesis of UTI involves the interaction between the host’s defenses and the invading organism’s virulence factors. In the case of UTI, the most important host defenses against the initiation of infec-

PIAM:L01:BIF01006XA.89

FF: W1FP

E1: IN01

$$$

tion are not the specific immune response or the mobilization of phagocytic leukocytes; rather, the most important host defenses can be characterized as anatomical and mechanical – an anatomically normal urinary tract, without obstruction or an indwelling foreign body such as a catheter, a normally emptying bladder such that residual pools of stagnant urine are not present, and a kidney that has not been injured by trauma or some other disease process. According to this analysis, then, three major host factors play an important role in the pathogenesis of UTI in the renal transplant patient: the posttransplant urinary catheter, the physical and immunologic trauma the transplanted kidney endures in the peritransplant period, and the poor bladder function after the catheter is removed in a significant number of renal transplant recipients, particularly diabetics and those whose native kidney failure was related to congenital anatomical abnormalities of the urinary tract. In the case of the urinary catheter, with current practice dictating the removal of the bladder catheter in less than a week, overt UTI is unusual while the catheter is still in place. However, the catheter tips not infrequently become contaminated, and provide a reservoir from which infection is derived. In animal models, the combination of bacteria inoculated into the bladder and trauma to the kidney will result in pyelonephritis, whereas bladder infection without renal trauma results only in a transient cystitis. In the transplant recipient, it would seem logical that the harvesting, transport, and then transplantation of the kidney allograft, followed by whatever immunologic trauma that is engendered by the rejection process, would likewise increase the susceptibility of the kidney to invasive infection. Such infection, once contracted, would then, presumably, be exacerbated by the exogenous immunosuppressive therapy. It is important to emphasize, however, that immunosuppression is a secondary amplification factor in the pathogenesis of these infections, with the anatomical/ mechanical factors being the most important in determining host susceptibility [1, 12–16]. The aforementioned events are further exacerbated if a technical complication such as a ureteral leak or an area of renal infarction (usually due to a problem with an accessory renal artery) occurs. In the case of a ureteral leak, the need for nephrostomy tubes, surgical drains, ureteral stents, and/or indwelling catheters makes infection almost inevitable. In the case of renal infarcts, even such relatively nonvirulent, commensal organisms as Staphylococcus epidemidis or diphtheroids can cause symptomatic, even bacteremic, and, commonly, relapsing infections that are virtually impossible to eliminate with antimicrobial therapy (although such infections can be rendered asymptomatic with chronic, suppressive therapy [1, 8]. There has long been speculation that the type of ureteral anastomosis performed might have an important influence on the occurrence and consequences of UTI in the renal transplant recipient. Ureteropyelostomy has the theoretical

Tolkoff-Rubin/Rubin

28

PIAM:L01:BIF01006XA.89

FF: W1FP

E1: IN01

$$$

advantage of utilizing the recipient’s own ureter, if it is normal, and thus having a nonrefluxing urinary tract and the ability to remove the bladder catheter within 24 h of surgery; the disadvantage of this approach is that this form of anastomosis is technically more difficult. The ureteroneocystostomy anastomosis is technically easier to perform, but requires the bladder catheter to remain in place for a longer period of time and not infrequently is associated with vesicoureteral reflux which could exacerbate the effects of any UTI on the allograft. In point of fact, in skilled hands, and provided antimicrobial prophylaxis (see below) is administered, there is currently little evidence that the type of anastomosis employed has an important effect on the incidence and severity of UTI [1, 8]. Over the past two decades, important information has been gathered on the molecular pathogenesis of UTI that has delineated the critical role of bacterial virulence factors in the causation of UTI. It is now apparent that there are a few clones of uropathogenic Escherichia coli (and presumably other uropathogens) that possess a series of virulence factors that mediate their success as invaders of the anatomically normal urinary tract. The most important such virulence factors are bacterial surface adhesins. The best studied of these are the so-called p-pili, which mediate attachment of the bacteria to specific receptors on the uroepithelium. Careful studies in children have demonstrated that pyelonephritis in individuals with anatomically normal urinary tracts usually requires the introduction of such uropathogenic bacteria; in contrast, children with anatomical abnormalities can develop pyelonephritis due to nonvirulent bacteria [17–20]. Although not well studied in renal transplant patients, the epidemiologic characteristics of UTI in these patients suggest a smiliar series of events. In the first few months following transplant, anatomical and technical issues related to the urinary tract are the dominant features in the pathogenesis of UTI. Reflecting this, a relatively wide range of organisms can cause infection. After this time period (e.g., more than 3 months posttransplant), although anatomical/technical issues may still be an issue, the pattern of infection more commonly resembles that observed in the general population, and it is likely that the preponderance of infections in this time period in patients with anatomically normal urinary tracts is due to uropathogenic clones of bacteria, particularly E. coli.

Clinical Effects of UTI in the Renal Transplant Patient The clinical effects of UTI in the renal transplant patient can be divided into two categories: the direct and the indirect effects. When considering the direct effects, it is important to recognize that just as the pathogenesis of UTI in the first few months is different than in the later period, the clinical manifestations are different as well. UTI occurring in the first 3 months posttransplant is frequently

UTI in Renal Transplantation

29

PIAM:L01:BIF01006XA.89

FF: W1FP

E1: IN01

$$$

associated with overt pyelonephritis, bacteremia, and a high rate of relapse when treated with a conventional course of antibodies (particularly ß-lactams). In contrast, UTI, occurring at a later time, unless associated with an anatomical problem such as an obstructing stone, a malfunctioning ileal loop, or a poorly emptying bladder, is rarely associated with overt pyelonephritis or bacteremia, can be successfully managed with a conventional 10- to 14-day course of antibiotics, and can usually be managed as an outpatient [1, 8]. The indirect effects of UTI, although incompletely characterized at present, are of increasing interest. Cytokines and growth factors elaborated by the patient in response to UTI are capable of a number of effects. In addition to the nonspecific stimulatory effects of endotoxin on the immune system, these cytokines can modulate the display of major histocompatibility complex (MHC) antigens on the allograft, perhaps contributing to the host’s immunologic response to the foreign tissue. Since cytokines play an important role in the pathogenesis of the rejection process, the ‘excessive’ generation of cytokines in response to infection could drive the rejection process harder [1]. In addition, there has been speculation that molecular mimicry, structural homology between bacterial antigens and MHC antigens, also could increase the host’s immune response to the allograft. Elevations of serum creatinine during symptomatic UTI are common, and present data from biopsies in asymptomatic patients in the early posttransplant period suggest that infection-induced allograft injury may be more important thant previously suggested [1, 8]. An additional effect of UTI-induced cytokine release is the effect of these cytokines in modulating the course of other infections. In particular, German investigators have shown convincingly that tumor necrosis factor, a cytokine regularly released in the course of symptomatic UTI, is a prime factor in the reactivation of latent cytomegalovirus (CMV) [21, 22]. Once reactivated, CMV is an important pathogen in these patients, producing not only direct clinical effects but itself modulating host response, allograft injury, and the course of other infections [1]. In sum then, there is increasing evidence that the indirect effects of UTI, mediated by cytokines, are of clinical importance, both in influencing the host response to the allograft, and in influencing the course of other clinically important infections that can impact on the renal allograft recipient.

Clinical Management of UTI in Renal Transplant Recipients The clinical management of UTI in the renal transplant recipient has two components: prevention of infection and treatment of infection when prevention fails. Over the past 15 years it has become apparent that UTI and urosepsis can be effectively prevented with low doses (a single tablet at bedtime) of trimethoprim/

Tolkoff-Rubin/Rubin

30

PIAM:L01:BIF01006XA.89

FF: W1FP

E1: IN01

$$$

sulfamethoxazole, trimethoprim, or a fluoroquinolone such as ciprofloxacin. Failures of such prophylactic programs usually connote an anatomical problem, and should always trigger an evaluation of the urinary tract for correctable anatomical abnormalities [1, 23–26]. Low-dose trimethoprim-sulfamethoxazole therapy (we use one tablet which contains 80 mg trimethoprim and 400 mg sulfamethoxazole, at bedtime) remains the regimen of choice for this purpose, as it simultaneously also prevents Pneumocystis carinii pneumonia (the incidence of this infection falling from approximately 10% to essentially zero), listeriosis, and nocardiosis. At this dose, adverse interactions with cyclosporine are rare (although synergistic nephrotoxicity occurs at higher doses). However, trimethoprim-sulfamethoxazole use can be associated with other toxicities – rash, Stevens-Johnson’s syndrome, bone marrow toxicity (particularly in patients receiving azathioprine), and, rarely, interstitial nephritis. For patients who cannot tolerate trimethoprim-sulfamethoxazole, ciprofloxacin (or other fluoroquinolones) can be substituted, with the addition of some other form of Pneumocystis prophylaxis [1, 23–26]. The optimal duration of prophylaxis is unknown, with most transplant groups continuing this therapy for 6–12 months, unless patients have continuing anatomic risk factors that might predispose to a continuing risk of UTI. In addition, more than one episode of UTI off prophylaxis should trigger the reinstitution of an effective prophylactic regimen following a search for a correctable anatomic abnormality [1]. Therapy of symptomatic UTI in renal transplant patients is similar to that prescribed for nontransplant patients with UTI, with one notable exception – short-course therapy, single-dose or 3-day regimens have not been studied in this population. Accordingly, the standard of care is a 10- to 14-day course of an effective antibiotic, usually a fluoroquinolone. The key issue in choosing the appropriate antibiotic is to remember the potential for adverse interactions with cyclosporine. Thus, therapy with aminoglycosides, or high-dose trimethoprim-sulfamethoxazole (1160 mg trimethoprim/800 mg sulfamethoxazole per day) should be avoided [27].

Candiduria in the Renal Transplant Recipient One additional form of UTI is that caused by Candida. Although asymptomatic candiduria is frequently observed also in the general population, in the renal transplant patient it can be particularly dangerous because of the potential for developing obstructing candidal fungal balls. This occurs most commonly in diabetic patients with poorly functioning bladders. Once such candidal fungal balls develop, ascending candidal pyelonephritis and candidal sepsis

UTI in Renal Transplantation

31

PIAM:L01:BIF01006XA.89

FF: W1FP

E1: IN01

$$$

will ensue. For this reason, we preemptively treat even asymptomatic candiduria with oral fluconazole, or with low-dose amphotericin plus flucytosine, if this fails [12].

References 1

2 3 4 5 6 7 8

9 10 11 12 13 14 15 16 17 18

19 20 21

Rubin RH: Infection in the organ transplant recipient; in Rubin RH, Young LS (eds): Clinical Approach to Infection in the Compromised Host, ed 3. New York, Plenum Medical, 1994, pp 629–705. Hinman F Jr, Schmaelzle JF, Belzer FO: Urinary tract infections and renal homotransplantation. II. Posttransplantation bacterial invasion. J Urol 1969;101:673–679. Leigh DA: The outcome of urinary tract infections in patients after human cadaveric renal transplantation. Br J Urol 1970;101:453–456. Martin DC: Urinary tract infection in clinical renal transplantation. Arch Surg 1969;99:474–476. Bennett WM, Beck CH Jr, Young HH et al: Bacteriuria in the first month following renal transplantation. Arch Surg 1970;101:453–456. Prout GR Jr, Hume DM, Lee HM, et al: Some urological aspects of 93 consecutive renal homotransplants in modified recipients. J Urol 1967;97:409–425. Ramsey DE, Finch WT, Birtch AG: Urinary tract infections in kidney transplant recipients. Arch Surg 1979;114:1022–1025. Rubin RH, Fang LST, Cosimi AB, et al: Usefulness of the antibody-coated bacteria assay in the management of urinary tract infection in the renal transplant patient. Transplantation 1979;27: 18–20. Myerowitz RL, Medeiros AAM, O’Brien TF: Bacterial infection in renal homotransplant recipients: A study of fifty-three bacteremic episodes. Am J Med 1972;53:308–314. Nielsen HE, Korsager B: Bacteremia after renal transplantation. Scand J Infect Dis 1977;9:111– 117. Pearson JC, Amend WJ Jr, Vincenti FG, et al: Posttransplantation pyelonephritis: Factors producing low patient and transplant morbidity. J Urol 1980;123:153–156. Rubin RH: Infectious disease complications of renal transplantation. Kidney Int 1993;44:221– 236. Burleson RL, Brennan AM, Scruggs BF: Foley catheter tip cultures: A valuable diagnostic aid in the immunosuppressed patient. Am J Surg 1977;133:723–725. Schaeffer AJ: Catheter-associated bacteriuria in patients in reverse isolation. J Urol 1982;128:752– 754. Schaeffer AJ, Chmiel J: Urethral meatal colonization in the pathogenesis of catheter-associated bacteriuria. J Urol 1983;130:1096–1099. Heptinstall RH: Experimental pyelonephritis: A comparison of blood-borne and ascending patterns of infection. J Pathol 1965;89:71–80. Suanborg C, de Man P, Sandberg T: Renal involvement in urinary tract infection. Kidney Int 1991; 39:541–548. Otto G, Sandberg T, Marklund BI, et al: Virulence factors and pap genotype in Escherichia coli isolates from women with acute pyelonephritis, with and without bacteremia. Clin Infect Dis 1993; 17:448–453. O’Hanley P, Low D, Romero I, et al: Gal-Gal binding and hemolysin phenotype and genotype associated with uropathogenic E. coli. N Engl J Med 1985;313:414–419. Johnson JR, Roberts PL, Stamm WE: P. fimbriae and other virulence factors in Escherichia coli urosepsis: Assocation with patients’ characteristics. J Infect Dis 1987;156:225–230. Stein J, Volk HD, Liebenthal C, et al: Tumour necrosis factor-alpha stimulates the activity of the human cytomegalovirus major immediate early enhancer/promoter in immature monocytic cells. J Gen Virol 1993;74:2333–2338.

Tolkoff-Rubin/Rubin

32

PIAM:L01:BIF01006XA.89

22 23

24

25

26

27

FF: W1FP

E1: IN01

$$$

Docke WD, Prosch S, Fietze E, et al: Cytomegalovirus reactivation and tumour necrosis factor. Lancet 1994;343:268–269. Tolkoff-Rubin NE, Cosimi AB, Russell PS, et al: A controlled study of trimethoprim-sulfamethoxazole prophylaxis of urinary tract infections in renal transplant recipients. Rev Infect Dis 1982;4: 616–618. Fox BC, Sollinger HW, Belzer FO, et al: A prospective, randomized, double-blind study of trimethoprim-sulfamethoxazole for prophylaxis of infection in renal transplantation: Clinical efficacy, absorption of trimethoprim-sulfamethoxazole, effects on the microflora, and the cost benefit of prophylaxis. Am J Med 1990;89:255–274. Maki DG, Fox BC, Kuntz J, et al: A prospective, randomized, double-blind study of trimethoprimsulfamethoxazole for prophylaxis of infection in renal transplantation. Side effects or trimethoprim-sulfamethoxazole, interaction with cyclosporine. Lab Clin Med 1992;119:11–24. Hibberd PL, Tolkoff-Rubin NE, Doran M, et al: Trimethoprim-sulfamethoxazole compared with ciprofloxacin for the prevention of urinary tract infection in renal transplant recipients. Online Curr Clin Trials, August 11, 1992 (Doc No 15). Rubin RH, Tolkoff-Rubin NE: Antimicrobial strategies in the care of organ transplant recipients. Antimicrob Agents Chemother 1993;37:619–624.

Dr. Nina E. Tolkoff-Rubin, Hemodialsis and CAPD Unit, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114 (USA)

UTI in Renal Transplantation

33

PIAM:L01:BIF01007XA.79

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 34–36

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infection and the Immunocompromised Host: UTI in Renal Transplant Patients Nina E. Tolkoff-Rubin, Robert H. Rubin Transplantation and Dialysis Units, Massachusetts General Hospital and Harvard Medical School, Boston, Mass., USA

All urinary tract infections (UTIs) in renal transplant recipients should be regarded as complicated. The frequency of UTI following renal transplantation may be as high as 79% [1, 2]. Most UTIs are acquired early after transplantation and are related to the use of indwelling urinary catheters in the immediate postoperative period. The incidence is reduced if the catheter is removed early. Catheter-related UTIs are not prevented by prophylactic therapy with cotrimoxazole [3]. Up to 12% of these ‘early’ UTIs may develop a complicating Gram-negative bacteremia [1]. Documented UTIs acquired from the donor kidney are rare. A small proportion of patients, usually women, on regular dialysis treatment may have an asymptomatic UTI when admitted for transplantation. This is more likely in patients who have a pre-existing urological problem such as urinary calculi or a neurogenic bladder. Major urological complications after renal transplantation, such as urinary leakage or obstruction, are uncommon, but are usually complicated by bacteriuria. The incidence of UTI is lower in recipients who have had an antireflux procedure during the ureteric anastamosis, while the type of immunosuppression does not alter the risk of UTI. Asymptomatic bacteriuria or recurrent UTIs do not affect either patient or graft survival. However, acute pyelonephritis involving the transplanted kidney may occasionally cause rapid, but reversible, graft dysfunction [4]. There has been some suggestion that transplant rejection may be precipitated by an Enterococcus faecalis UTI, but this has not been substantiated [2]. Occasionally the transplanted kidney may be infected by an unusual or fastidious

PIAM:L01:BIF01007XA.79

FF: W1FP

E1: IN01

$$$

Table 1. Number of UTIs documented during the first 12 weeks after renal transplantation in the last 62 consecutive patients Males (n = 34)

Females (n = 28)

Patients with UTI at time of transplant

3 (P. aeruginosa 1; Klebsiella sp. 1; S. epidermidis 1)

2 (both Streptococcus group B)

Patients with a posttransplant UTI in first 3 weeks

13 (8)

10 (3)

(Number of these with posttransplant urological complication)

E. coli E. faecalis Acinetobacter spp. S. epidermidis P. mirabilis P. aeruginosa

1 3 3 2 2 2

5 (1 Acinetobacter sp.) 4 0 1 0 0

Patients followed daily for 12 weeks

19

16

Number of these with first posttransplant UTI between 3 and 12 weeks

1 (S. epidermidis)

3 (E. coli 2; Citrobacter sp. 1)

pathogen. If there is any doubt about the interpretation of the culture of a voided urine specimen then urine should be taken by suprapubic aspiration. Most catheter-related UTIs are due to nosocomial organisms. There has been considerable interest recently in the bacteriology of the biofilm that forms on catheters and stents [5]. After graft function has stabilized, most UTIs are not usually associated with a urinary tract abnormality. However, if there was a documented early relapse, a focus of infection within the native kidneys, the grafted kidney or elsewhere in the urinary tract such as the prostate gland or a urinary calculus should be suspected. An audit on 30 consecutive renal transplants was undertaken in this department in 1988 [6]. Three (all women) of these 30 patients had a UTI present at the time of transplantation. A UTI was documented in 8 of the 21 men and 4 of the 9 women within the first 5 days after transplantation when the urinary catheter was in place. Nineteen (63%) of these 30 patients had at least one documented UTI in the first 3 months. Following that audit it became our routine practice to remove the urinary catheter on day 3, rather than day 5, after transplantation.

UTT in Renal Transplantation

35

PIAM:L01:BIF01007XA.79

FF: W1FP

E1: IN01

$$$

A second audit completed in January 1995 studied the last 62 consecutive renal transplant patients (table 1). Five (8%) of these 62 patients had a UTI at the time of transplantation. Three of these 5 were males with a bladder-emptying problem. An additional 23 patients had a UTI within the first 3 weeks after transplantation and these infections were invariably related closely to urethral catheterization. Eleven of these 23 had required prolonged bladder or nephrostomy drainage. The majority of these infections in the males were with nosocomial organisms (table 1). Of these 62 patients, 35 were followed with daily urine cultures for 12 weeks and an additional 4 patients (3 women) developed a UTI. Soon after this we started leaving a double J ureteric stent in place for 28 days [7]. For those patients with stable graft function the management of a UTI should be the same as for any other complicated UTI. We treat such patients for 5 days. UTI involving the native kidneys is uncommon, but should be considered if the patient fails a standard 5-day curative regimen. These patients pose a diagnostic and therapeutic dilemma. As it is difficult to achieve adequate antibiotic concentrations in poorly perfused native kidneys, surgical intervention may be required to eradicate a persisting focus of infection. Thirteen percent of our transplant patients have experienced recurrent UTIs, which have been successfully controlled with conventional long-term, low-dose, prophylactic drug regimens such as nitrofurantoin 50 mg nocte, trimethoprim 100 mg nocte or norfloxacin 200 mg nocte.

References 1

2 3

4 5 6 7

Tolkoff-Rubin NE, Cosimi AB, Russell PS, et al: A controlled study of trimethoprim-sulfamethoxazole prophylaxis of urinary tract infection in renal transplant recipients. Rev Infect Dis 1982;4: 614–618. Kunin CM (ed): Detection, Prevention and Management of Urinary Tract Infections. Philadelphia, Lea & Febiger, 1987, pp 1–447. Fox BC, Sollinger HW, Belzer FO, Maki DG: A prospective, randomized, double-blind study of trimethoprim-sulfamethoxazole for prophylaxis of infection in renal transplantation: Clinical efficacy, absorption of trimethoprim-sulfamethoxazole, effects on the microflora, and the cost-benefit of prophylaxis. Am J Med 1990;89:255–274. Yang CW, Kim YS, Choi EJ, et al: Acute pyelonephritis of graft kidney and renal failure in renal transplant recipients. Kidney Int 1995;47:362. Reid G, Denstedt JD, Kang YS, Lam D, Naus C: Microbial adhesion and biofilm formation on ureteral stents in vitro and in vivo. J Urol 1992;148:1592–1594. Bailey RR: Urinary tract infections in specific conditions; in Sidabutar RP (ed): Proceedings 8th Asian Colloquium of Nephrology, Jakarta 1989, pp 175–185. Nicholson ML, Veitch PS, Donnelly PK, Bell PR: Urological complications of renal transplantation: The impact of double J ureteric stents. Ann R Coll Surg 1991;73:316–321.

Drs. Nina E. Tolkoff-Rubin and Robert H. Rubin, Massachusetts General Hospital, 32 Fruit St., Boston, MA 02114 (USA)

Tolkoff-Rubin/Rubin

36

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 37–45

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Bacteriuria in Male Patients Infected with Human Immunodeficiency Virus Type 1 Relation to Immune Status (CD4+ Cell Count) and the Influence of Pneumocystis carinii Pneumonia Prophylaxis on Incidence and Resistance Development Daniëlle A. van Dooyeweert a, Margriet M.E. Schneider a, Jan C.C. Borleffs a, Andy I.M. Hoepelman a, b a

b

Department of Medicine, Division of Infectious Diseases and AIDS, University Hospital Utrecht, and Eijkman Winkler Laboratory for Medical Microbiology, University Hospital Utrecht, The Netherlands

Young, sexually active women often experience symptomatic urinary tract infections (UTIs). In contrast, infections in men between the ages of 15 and 50 are uncommon [1, 2]. Most UTIs in males occur in babies with urological anomalies, in older men with prostatic hypertrophy, or after invasive urogenital procedures [1–3]. Reports from the United States that UTIs are more common in homosexual men than in heterosexual men have not been confirmed in Europe [4, 5]. In the late 1980s, however, we noted the occurrence of several UTIs in men infected with human immunodeficiency virus type 1 (HIV-1). The prevalence of such infections among patients infected with HIV-1 is unknown; however, others have also suggested the incidence to be high [6, 7]. Moreover, it is not known whether UTI is generally more common in immunocompromised patients, renal transplant patients excepted [2, 8]. Because of this, we analyzed the occurence of bacteriuria in men infected with HIV in relation to their immune status (CD4+ cell count) for a 2-year period until February 1990. Thereafter, we prescribed Pneumocystis carinii pneumonia (PCP) prophylaxis in all patients with a CD4+ cell count !200/mm3. Moreover, for this meeting we analyzed retrospectively the effect of three types of PCP prophylaxis on the incidence of bacteriuria and development of resistance.

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

Methods Patients and Methods [cf. 9] In brief, all eligible patients (n = 130) were men who attended the Department of Internal Medicine (Section of Immunology and Infection) at the University Hospital Utrecht, The Netherlands, between December 1987 and February 1990. They were eligible for both studies if they met the following criteria: HIV infection documented by the enzyme-linked immunosorbent assay with Western blot confirmation and had no indwelling catheter or recent urologic manipulation. Informed consent was obtained from all patients. No primary prophylaxis for PCP had been prescribed before February 1990. Thereafter, primary PCP prophylaxis was started with either cotrimoxazole at a low (480 mg) or a high dose (960 mg) daily, or pentamidine aerosol monthly in patients with CD4+ cell count ! 200 mm3 [10]. The first study was terminated in February 1990. In the first study, urinary cultures were obtained during the first visit and every 6 months thereafter, and when signs or symptoms of UTI were noted, or the patients had fever of unknown origin. An episode was considered symptomatic in the presence of fever 638.0 ° C (without the presence of other infection sites) and signs (frequency, urgency, dysuria) or symptoms (flank pain) of UTI. We divided the patients into three groups according to their CD4+ cell count: group I ! 200 mm3, (n = 47); group II 200–500/mm3 (n = 27) and group III 1 500/mm3 (n = 24). Grouping was based on the CD4+ cell count at the time of a positive urinary culture or, when no positive cultures were found, on the last CD4+ cell count noted. Subsequently, to study the influence of primary and secondary PCP prophylaxis (960 mg) on bacteriuria and TMP/SMZ resistance (2nd study), we cultured during 4 months (December 1994 until March 1995) all men (n = 103) receiving PCP prophylaxis at least once. We also checked files for documented bacteriuria in the past. Laboratory Studies Urine for urinary culture and urinalysis was obtained by the midstream clean-catch method. A urinary culture was considered positive if 6105 CFU/ml were grown for Gramnegative rods and 6104 CFU/ml for Enterococcus faecalis. When a positive urinary culture was found in an asymptomatic patients, the test was repeated and only considered positive if it yielded growth of the same microorganism. Bacteria were identified by standard bacteriological and serological methods [11]. Relapsing infections (n = 1) were differentiated from reinfections and were considered to belong to the same episode of infection. Leukocyturia was noted if five or more leukocytes were seen in a high-power microscopic field of urinary sediment. In the retrospective study, TMP/SMZ resistance was defined as an MIC 1 4 mg/ 76/liter. Statistical Analysis The effect of PCP prophylaxis on the occurrence of bacteriuria and resistance development was compared by Kaplan-Meier curves. Data were analyzed with SYSTAT and with the primer of biostatistics program [12].

van Dooyeweert/Schneider/Borleffs/Hoepelman

38

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

Results Relationship between Bacteriuria and CD4+ Cell Count Thirty of the 130 eligible patients were excluded because of insufficient follow-up information with respect to urinary cultures (n = 13), CD4+ cell count (n = 11) or both (n = 6). Moreover, 1 patient with hemophilia, urethral stenosis and recurrent UTI, a CD4+ cell count of 104/mm3, and a nephrectomy for renal abscesses in the past, was excluded. Of the 5 hemophilic patients, he was the only one who presented with a UTI. One patient with a CD4+ cell count of 243/mm3 was excluded because he presented with a urosepsis due to hephrolithiasis. Thus, data from 98 patients were analyzed for the first study. The patients in all groups were comparable in age, follow-up period (p = 0.39), sexual behavior and Karnofsky scale of performance at the end of the follow-up period. There was significant difference between groups in the number of admissions (p = 0.011) and the number of days the patients were hospitalized (p = 0.02). Eighty-nine (91%) were either homo- or bisexual. Table 1 shows the results of the urinary cultures. Of the 47 patients in group I, 30% (14) had at least one period of bacteriuria. There were 21 episodes, 13 (62%) of which were symptomatic including 7 with fever of unknown origin. Leukocyturia was present in 5 (24%) episodes. Infecting microorganisms were: Escherichia coli (n = 8); E. faecalis (n = 6); Klebsiella pneumoniae (n = 3); ßhemolytic streptococci group B (n = 2); Staphylococcus aureus (n = 1); Enterobacter aerogenes (n = 1) and Proteus mirabilis (n = 2). In two urinary cultures two species were grown. One patient with E. faecalis developed bacteremia. In group II, 3 (11%) patients had bacteriuria, with 5 episodes. Two (40%) of these episodes were symptomatic. Leukocyturia was present in 3 (60%). Infecting microorganisms were E. coli (n = 3), Acinetobacter anitratum (n = 1), and ß-hemolytic streptococci group B (n = 1). No patients in group III had bacteriuria.

Table 1. Results of urinary cultures CD4+

Patients Patients with a positive urinary culture Episodes Symptomatic Asymptomatic Leukocyturia

Bacteriuria in Males Infected with HIV-1

! 200/mm3

200–500/mm3 1 500/mm3

47 14 21 13 8 5

27 3 5 2 3 3

24 0 – – – –

39

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

Fig. 1. Relationship between the CD4+ cell counts and bacteriuria.

When a UTI was diagnosed, therapy with a quinolone was prescribed (for 2 weeks) and cultures were repeated at 2 weeks, 6 weeks, 3 months and every 6 months thereafter. One patient with a relapse was treated for 6 weeks with ciprofloxacin. The rate of bacteriuria per patient-month, 4 (group I) versus 2 (group II), differed significantly (p ! 0.001). Moreover, with logistic regression analysis, a significant relationship between CD4+ cell count and bacteriuria was found (p = 0.00003) (fig. 1). No relationship to anal intercourse (p = 0.48), Karnofsky score (p = 0.2), days spent in hospital (p = 0.3), or age (p = 0.36) was found, however. A significant difference was documented between groups in the number of admissions and days spent in hospital. However, bacteriuria developed in only 3 of 17 (18%) patients, and 4 of 26 (15%) episodes during hospitalization, no significant relationship between number of admissions or days spent in hospital was found (with logistic regression). All but one (an IV drug user) period of bacteriuria occurred in homosexual males. Analysis of our data did not change significantly after deleting this individual, nor after deleting all the 9 heterosexual men from the analysis.

van Dooyeweert/Schneider/Borleffs/Hoepelman

40

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

Table 2. Influence of PCP prophylaxis on the incidence of UTI

Patients CD4+ count Cultures Follow-up, weeks Positive cultures Months Patients with positive cultures Number TMP/SMZ-resistant Time to resistance, weeks

TMP/SMZ (480)

TMP/SMZ (960)

52 71 (5–193) 113 62 (6–120) 12 14 9 (17%) 12 (86%) 4–154

23 45 (0–205) 59 63 (4–240) 7 8 5 (22%) 7 (88%) 1–88

Pentamidine

23 42 (2–198) 102 98 (22–240) 14 18 9 (32%) 11 (61%) 1–171

There were no significant differences in sexual behavior between the three groups. In group I, 7 (50%) patients with bacteriuria currently practiced rectalinsertive intercourse (active, passive or both); the sexual behavior of 3 (21%) other patients was unknown. Of the patients without bacteriuria, 17 (52%) practiced anal intercourse, while the sexual techniques of 8 (24%) others were unknown. In group II all patients with a positive urinary culture practiced rectal-insertive intercourse. Fifteen (63%) patients without bacteriuria also had anal intercourse, while the sexual behavior of 3 (13%) was unknown. In group III, 16 (66%) patients had anal sexual contacts and 3 (13%) had unknown sexual techniques. All patients who practiced anal intercourse claimed to use condoms. In group I, 6 patients reported a history of UTI, and 3 of them developed bacteriuria during the study period. In group II, none, and in group III, 2 patients remembered a previous UTI. A total of 23 (24%) patients used zidovudine (AZT) at the time of the study, 21 of them in group I. Of these 21, 8 (38%) were bacteriuric, not significant when compared to nonbacteriuric patients who used AZT. Three (14%) patients were granulocytopenic and 6 (23%) episodes were accompanied by granulocytopenia (!500/mm3) (NS). In group II none of the patients with bacteriuria used AZT. Secondary PCP prophylaxis was used by 6 patients in group I. None of them had a positive urinary culture. Effect of PCP Prophylaxis on the Incidence of Bacteriuria Previously, we have shown that TMP/SMZ is more effective as primary PCP prophylaxis than pentamidine [10]. Both, however, remain in use in our hospital. For secondary PCP prophylaxis we use a high (960 mg) dose of TMP/SMZ. We now analyzed the possible effect of these regimens on bacteriuria. Table 2 shows that the number of patients with a positive culture using pentamidine, 9 (32%),

Bacteriuria in Males Infected with HIV-1

41

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

Fig. 2. Probability of survival without a positive urinary culture as a function of time.

was not significantly different (p = 0.207) from patients using low-dose, 9 (17%), or high-dose TMP/SMZ, 5 (22%). When we compare the group of patients using low-dose TMP/SMZ with those using TMP/SMZ (960 mg), no difference was found either (fig. 2.). Although, not significantly different more strains (87%) in patients receiving TMP/SMZ were resistant than strains cultured from patients receiving pentamidine (61%) (table 2, fig. 3). One should remember that all the patients have been exposed previously to TMP/SMZ.

Discussion Infection of the urinary tract is usually considered a female problem. The occurrence of UTI in men has, however, been studied in different patient groups [1–5, 13]. Two studies on the incidence and prevalence of UTI among homosexual men have yielded conflicting results [4, 5]. In our clinic, we noticed an unusual number of UTI in patients infected with HIV-1. In the prospective study, published previously [9], a clear relationship between bacteriuria in males infected with HIV-1 and their CD4+ cell count was documented (fig. 1). In the group of patients with a CD4+ cell count !200/mm3, 30% developed bacteriuria, while in only 11% of the patients with CD4+ cells between 200 and 500/mm3, was a positive urinary culture found. No bacteriuria

van Dooyeweert/Schneider/Borleffs/Hoepelman

42

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

Fig. 3. Probability of survival without a TMP/SMZ-resistant microorganism as a function of time.

was found in patients with a CD4+ cell count 1 500/mm3 (table 1). Moreover, the rate of bacteriuria per patient-month, 4 (group I) versus 2 (group II), differed significantly (p ! 0.001). Although infections are among the main events in immunocompromised hosts, UTI is a rare symptom of immunodeficiency (excluding renal transplant patients) [8]. A relationship between mental status, incontinence of bladder and bowel and mobility has been found in the elderly [13]. In our study, however, no significant difference in the Karnofsky performance score or the mental status between the groups, nor between patients with or without bacteriuria was found [9]. Moreover, in an ongoing activities-of-daily living study, all patients with bacteriuria claimed complete continence and were both mobile and independent [9]. Hospital admissions are known to predispose to nosocomial infections, among them UTI. A significant difference between hospital admissions and days spent in hospital was found between the groups. However, 22 of 26 (85%) episodes of bacteriuria were not associated with hospitalization and with logistic regression analysis no such relation was found. Therefore, hospitalization does not account for the association found between CD4 cell count and bacteriuria. The pathogenesis, therefore, is still unclear. A relationship between symptoms, carriage of pathogens causing the ‘gay bowel syndrome’, and HIV infection with a low CD4+ cell count has been found

Bacteriuria in Males Infected with HIV-1

43

PIAM:L01:BIF01008XA.87

FF: W1FP

E1: IN01

$$$

in homosexual men [14, 15]. Many extraintestinal-occurring-opportunistic infections were also found among these men. Moreover, it was not clear from these studies whether this association was directly related to other independent variables, such as extensive anal intercourse. Stamey [15] noted that male sex partners of women with vaginal colonization by Gram-negative bacilli may develop transient urethral colonization with the same organisms. Although generally believed to be unimportant in men, presumed heterosexual transmission of UTI through intercourse has been reported [16]. Moreover, using tissue cultures from dogs, Schwarz [17] showed that labelled E. coli inoculated into a mucosal lesion of the distal sigmoid colon were recovered from both kidneys. The pathway he suggested, either via lymphatic channels directly to the kidney or through lymphatics into the bloodstream and then to the kidney, is, however, contradictory. In this study, we did not find any differences in current sexual behavior between the bacteriuric and the nonbacteriuric patients. Moreover, current anal intercourse was not found to be a risk factor. Although the follow-up period in the three groups was no significantly different, there may have been differences in the frequency of exposures through anal intercourse. However, a considerable number of patients with bacteriuria did not practice anal intercourse and all patients with bacteriuria claimed to use condoms. Knowing the route of transmission of HIV in homosexual men, it seems justified to conclude that all patients practiced unprotected anal intercourse in the past. It is known that the prostate gland is a source for endogenous reinfections [1, 2, 16]. Unexpectedly, there was no difference in the rate of bacteriuria between patients using TMP/SMZ and those using pentamidine as PCP prophylaxis (table 2, fig. 2). This may be partially explained by the fact that people are infected with resistant microorganisms. A recent article [18] showed the same. In summary, we conclude that patients infected with HIV-1 and having a CD4+ cell count !200 mm3 are at increased risk for bacteriuria. These UTI may, however, remain undiagnosed, since only 60% were symptomatic in our study. PCP prophylaxis with trimethoprim/sulfamethoxazole does not significantly influence the rate of bacteriuria, probably due to the development of resistance.

References 1 2 3 4

Kunin CM: Detection, Prevention and Management of Urinary Tract Infections, ed 4. Philadelphia, Lea & Febiger, 1987. Lipsky BJ: Urinary tract infections in men. Ann Intern Med 1989;110:138–150. Burbige KA, Retik AB, Colodny AH, Bauer SB, Lebowitz R: Urinary tract infections in boys. J Urol 1984;132:541–542. Barnes RC, Roddy RE, Daifuku R, Stamm WE: Urinary tract infection in sexually active homosexual men. Lancet 1986;i:171–173.

van Dooyeweert/Schneider/Borleffs/Hoepelman

44

PIAM:L01:BIF01008XA.87

5 6 7

8 9 10

11 12 13

14 15 16 17

18

FF: W1FP

E1: IN01

$$$

Wilson APR, Tovey SJ, Adler MW, Gruneberg RN: Prevalence of urinary tract infection in homosexual and heterosexual men. Genitourin Med 1986;62:189–190. de Pinho F, Maranconi DV, Moreira MB, Rodrigues KM, Paz NA, Ramos-Filho CF, Schechter M: Possible high frequency of bacteriuria among Brazilian AIDS patients. AIDS 1991;5:342. Omar de Rosa S, Lopes GS, da Costa JLF, Lacerda MR, Haringer J, Faria RB: Urinary tract infections in men with HIV infection. Results of a pilot uncontrolled study (abstracts). FB 535. Sixth International Conference on AIDS, San Francisco 1990, p 211. Rubin RH, Young LS (eds): Clinical Approach to Infection in the Compromised Host, ed 2. New York, Plenum Medical, 1988. Hoepelman AIM, Van Buren M, Van Den Broek J, Borleffs JCC: Bacteriuria in men with HIV-1 is related to their immune status (CD4+ cell count). AIDS 1992;6:179–184. Schneider MME, Hoepelman AIM, Eeftinck Schattenkerk JKM, Nielsen TL, Van Der Graaf Y, Frissen JPHJ, Van Der Ende IME, Kolsters AFP, Borleffs JCC: A controlled trial of aerosolized pentamidine or trimethoprim-sulfamethoxazole as primary prophylaxis against Pneumocystis carinii pneumonia in patients with human immunodificieny virus infection. N Engl J Med 1992;327: 1836–1841. Hoepelman AIM, Bakker LJ, Verhoef J: Carumonam compared with gentamicin for treatment of complicated urinary tract infections. Antimicrob Agents Chemother 1988;32:473–476. Glantz SA: Primer of Biostatistics, The Program. New York, McGraw Hill, 1988. Nicolle LE, Henderson E, Bjornson J, McIntyre M, Harding GKM, MacDonell JA: The association of bacteriuria with resident characteristics and survival in elderly institutionalized men. Ann Intern Med 1987;106:682–686. Smith PD, Lane HC, Vee JG, et al: Intestinal infections in patients with the acquired immunodeficiency syndrome. Ann Intern Med 1988;108:328–333. Stamey TA: Pathogenesis and treatment of urinary tract infections. Baltimore, Willimas & Wilkins, 1980, pp 342–429. Wong ES, Stamm WE: Sexual acquisition of urinary tract infection in a man. JAMA 1983;250: 3087–3088. Schwarz H: Renal invasion by E. coli via a mucosal lesion of the sigmoid colon: A demonstration utilizing methods of autoradiography and group-specific serologic typing. Invest Urol 1968;6:98– 113. Evans JK, McGwan A, Hillman RJ, Forster GE: Incidence of symptomatic urinary tract infections in HIV-seropositive patients and the use of cotrimoxazole as prophylaxis against Pneumocystis carinii pneumonia. Genitourin Med 1995;71:120–122.

Dr. Andy I.M. Hoepelman, Department of Medicine, University Hospital, PO Box 85500, NL–3508 GA Utrecht (The Netherlands) E-mail: [email protected]

Bacteriuria in Males Infected with HIV-1

45

PIAM:L01:BIF01009XA.91

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 46–47

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infections in Young Men Walter E. Stamm Division of Allergy and Infectious Diseases, University of Washington School of Medicine, Seattle, Wash., USA

Recent data suggest that HIV infection, particularly when the CD4 count falls below 200 cells/mm3, may be associated with an increased risk of bacteriuria and symptomatic urinary tract infection in men [1]. These studies are clearly of interest, but additional studies are needed to clarify whether the increased risk of urinary tract infection is definitely related to immunodeficiency or to other factors. Other potential factors of importance would include the increased rates of hospitalization, catheterization, and the generally declining functional status that usually accompanies advanced illness. Certainly, these factors have been related to the risk of urinary tract infection in elderly patients. In addition to HIV infection, other factors identified in recent years appear to predispose otherwise healthy young men to acute uncomplicated urinary tract infection. In a case-control study done by Barnes and our group in the early 1980s [2], acute uncomplicated urinary tract infection in sexually active young men was associated with homosexuality. Although insufficient data were available to associate specific sexual practices with urinary tract infection in the study, it was proposed that insertive rectal intercourse likely predisposed these male patients to Escherichia coli that was acquired sexually. Of interest was the fact that 8 of the 11 E. coli strains available for testing were of urovirulent phenotypes, expressing P adhesins, hemolysin, aerobactin, and belonging to urovirulent O, K and H serogroups. Thus, it was proposed that sexual exposure via rectal intercourse to urovirulent strains predisposed to uncomplicated urinary tract infections. Also of interest was the fact that these patients in some cases presented with typical urinary tract symptoms (dysuria, urgency, frequency, pyuria, and hematuria), whereas other cases presented more as if they had urethritis with complaints of urethral discharge and urethritis shown by urethral Gram-stain. In a second study done by our group [3], Spach and co-workers performed a retrospective case-control study of men with urinary tract symptoms who had urine cultures obtained. These men attended the same clinic as those men studied

PIAM:L01:BIF01009XA.91

FF: W1FP

E1: IN01

$$$

by Barnes et al. [2]. In this study, 26 bacteriuric cases were compared with 52 symptomatic but culture-negative controls. Lack of circumcision was present in 31% of the cases but only 12% of controls (p = 0.04). When the analysis was performed just on those patients having infections with Gram-negative rods, 8 (42%) of 19 of the infected men were uncircumcised versus 8 (12%) of the 52 controls (p = 0.004). In this study, 19% of the cases and 19% of the controls were homosexual men. Of those tested (approx. 40%), only 1 man was HIV-positive. As in the first study, the majority of the E. coli strains tested (10 of 14) expressed the urovirulence properties cited above. Finally, in another study done at the University of Washington, Krieger et al. [4] described the occurrence of acute uncomplicated urinary tract infection in 38 healthy university men attending the student health clinic. These patients complained of typical symptoms of lower urinary tract infection and in 93% of cases E. coli was isolated. The strains were not typed for virulence characteristics. Eighty-seven percent of the men were circumcised, all claimed to be heterosexual, and all were sexually active with female partners. Eighty-four percent of the patients had no history of a genitourinary abnormality, and all who returned for follow-up had responded to 10–14 days of therapy with antimicrobials. Eleven had IVPs and uroflow studies performed and all were within normal limits. Taken together, the above studies suggest that although infrequent, urinary tract infections of an uncomplicated nature occur in sexually active young men. The infecting organisms are probably acquired by either rectal or vaginal intercourse. Interestingly, the infections are generally caused by urovirulent E. coli clones similar to those causing pyelonephritis in women. Lack of circumcision may increase the risk of infection. Clinically, the infections may present like a urinary tract infection or in some cases like nongonococcal urethritis. Although not rigorously studied, most of these patients appear to respond to 10–14 days of antimicrobial therapy and urological studies may not be generally warranted. Further studies of these issues would be of interest. References 1 2 3 4

Hoepelman AIM, van Buren M, van den Broek J, Borleffs JCC: Bacteriuria in men infected with HIV-1 is related to their immune status (CD4+ cell count). AIDS 1992;6:179–184. Barnes RC, Daifuku R, Roddy RE, Stamm WE: Urinary tract infection in sexually active homosexual men. Lancet 1986;i:171–173. Spach DH, Stapleton AE, Stamm WE: Lack of circumcision increases the risk of urinary tract infection in young men. JAMA 1992;267:679–681. Krieger JN, Ross SO, Simonsen JM: Urinary tract infection in healthy university men. J Urol 1993; 149:1046–1048. Prof. Dr. Walter E. Stamm, Division of Allergy and Infectious Diseases, University of Washington School of Medicine, M/C Box 356523, 1959 NE Pacific Street, Seattle, WA 98195 (USA)

Urinary Tract Infections in Young Men

47

PIAM:L01:BIF01010XA.89

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 48–51

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infection in Renal Transplant Patients Silvester Krcˇméry Department of Medicine, Old Town University Hospital, Bratislava, Slovak Republic

Urinary tract infection (UTI) is the most common form of infection affecting renal transplant recipients. Its incidence has been reported to vary from 35 to 79% [1, 2]. UTI during the first few months posttransplant is frequently associated with overt pyelonephritis, bacteremia, allograft dysfunction and a high rate of relapse. Acute pyelonephritis after kidney transplantation is often accompanied by fever and chills, allograft tenderness and deteriorating renal function. In a diagnostic differentiation from a rejection episode, urine culture, microscopic examination of the urine and allograft ultrasonography are helpful, in some cases also fine-needle graft biopsy is necessary.

Predisposing Factors of UTI in Renal Transplant Recipients What are the predisposing factors of UTI in renal transplant recipients? (1) Decreased immunologic competence due to immunosuppressive therapy leads to a weakened inflammatory response to a microbial invasion. (2) Nosocomial threats include invasive instrumentation during surgery, indwelling urinary catheter, and environmental contamination with antibiotic resistant pathogens (e.g., Pseudomonas aeruginosa, Legionella, Aspergillus). (3) Secondary bladder dysfunction and/or reflux into graft urinary tract. (4) Urologic complications (e.g., urinary leak, fistula, lymphocele with ureteral obstruction). (5) Renal function impairment (allograft dysfunction). (6) Bacteria in the donor kidney (rare). In general, the treatment guidelines for UTI following renal transplantation comprise immediate initiation of antibacterial or antifungal chemotherapy, while

PIAM:L01:BIF01010XA.89

FF: W1FP

E1: IN01

$$$

Table 1. UTI following renal transplant – empiric antimicrobial chemotherapy First choice

Alternatives

Mild to moderate illness

Fluoroquinolones PO Trimethoprim/sulfamethoxazole PO

Oral cephalosporins Amoxicillin/clavulanate PO Ampicillin/sulbactam PO

Severe illness

Fluoroquinolones IV Ureidopenicillins IV (azlocillin, piperacillin/tazobactam) Third-generation cephalosporins IV (ceftazidim, cefoperazone, etc.)

Ticarcillin/clavulanate Aztreonam1 Imipenem

1

Gram-negative infections with allergy or resistance to penicillins or cephalosporins.

predisposing factors and allograft functional parameters are considered. Possibly nephrotoxic substances should be avoided and chemotherapy should be individually prolonged (a 2- to 6-week course may be necessary in some cases). Subsequent long-term chemoprophylaxis is quite effective in preventing UTI relapses. Certain antimicrobial agents have the highest rate of drug interactions with cyclosporine A: (1) Pharmacokinetic interactions (via the hepatic cytochrome P450 enzyme system) lead either to enzymatic enhancement and decreased cyclosporine serum levels (e.g., rifampin), or to enzymatic down-regulation and increased cyclosporine serum levels (e.g. macrolides, ketoconazole, itraconazole). Appropriate cyclosporine A dose adjustment is necessary when one of these or analogous antimicrobial agents is used. (2) Idiosyncratic nephrotoxic reactions causing oliguric acute renal failure may occur with amphotericin B, aminoglycosides, vancomycin, itraconazole, and trimethoprim-sulfamethoxazole. Recommended antibacterial regimens suitable for UTI therapy in renal transplant patients are (table 1): (1) fluoroquinolones, (2) cephalosporins and (3) penicillins (alone or potentiated by ß-lactamase inhibitors). Monobactams or carbapenems are possible alternatives. Concerning fluoroquinolones, good results have been observed with ciprofloxacin and fleroxacin [3]. In our experience, pefloxacin is excellent due to its low nephrotoxicity, oncedaily dosing (no dose reduction is necessary in renal impairment due to predominant extrarenal metabolism), nearly 100% oral bioavailability, and long-lasting therapeutic urine levels (up to 5–7 days after a single oral dose) [4]. Increasing emphasis is now being placed on prophylactic and pre-emptive strategies – long-term UTI prophylaxis in renal transplant recipients: (1) Lowdose trimethoprim-sulfamethoxazole (480 mg p.o. at bedtime) provides also effec-

Urinary Tract Infection in Renal Transplant Patients

49

PIAM:L01:BIF01010XA.89

FF: W1FP

E1: IN01

$$$

tive prophylaxis against Pneumocystis carinii, Nocardia spp. and Listeria monocytogenes. The possibility of idiosyncratic reactions with cyclosporine A exists [5]. (2) Fluoroquinolones are better tolerated than, and at least as effective in UTI prevention as cotrimoxazole, but with a higher incidence or Pneumocystis pneumonia (combination with aerosolized pentamidine needs further evaluation).

Fungal UTI in Renal Transplant Recipients and Their Risk Factors Factors associated with funguria are: (1) diabetes mellitus; (2) urinary bladder dysfunction; (3) indwelling urinary catheter; (4) vaginal mycosis and other forms of mucocutaneous candidiasis; (5) prolonged antimicrobial chemotherapy, and (6) decreased immunologic competence. Every documented episode of candidal UTI (asymptomatic candiduria, candidal cystitis or pyelonephritis) requires effective systemic antifungal chemotherapy. Candidal infections, even when initially associated with minimal symptoms, should be approached aggressively in transplant patients [1]. Fluconazole is active against Candida albicans and C. tropicalis, but usually ineffective against C. krusei and C. glabrata (although our own experience is that C. glabrata sepsis may be successfully treated with fluconazole). Fluconazole can be administered orally, or intravenously. It is a triazole antifungal agent with a very low affinity for human cytochrome P450 and consequently, not expected to interact with drugs, including cyclosporine, metabolized through this system [6]. Minor adjustments in cyclosporine dose may be necessary when given with fluconazole. In our own experience, prolonged treatment duration has been safe in transplant recipients. Itraconazole and ketoconazole both down-regulate the P450 cytochrome system. Addition of low-dose ketoconazole (100–200 mg/day regimen) to cyclosporine-treated kidney transplant recipients allowed a decreased dose of cyclosporine A by 20–30%; this is not only cost-saving, but may have a favourable effect on graft and patient outcome. A significantly lesser incidence of fungal infections has been reported in these patients [7]. Amphotericin B should be used with caution in renal transplant recipients due to its nephrotoxicity predominantly in fluconazole-resistant C. krusei and C. glabrata. Liposomal amphotericin B is an effective alternative [8]. In our experience, urine typically becomes culture-negative after 1–4 days of dosing at 50 mg/day. The very low incidence of toxicity with this preparation is particularly useful in renal transplant patients on cyclosporine immunosuppression with impaired renal function.

Krcˇméry

50

PIAM:L01:BIF01010XA.89

FF: W1FP

E1: IN01

$$$

References 1 2 3

4

5

6 7

8

Rubin RH: Infectious disease complications of renal transplantation. Kidney Int 1993;44:221– 236. Kuriyama M, Nagai T, Uno H, Nishida Y, Ishihara S, Kobayashi K, Takahashi Y, Saito A, Kawada Y: Urinary tract infections after kidney transplantation. Acta Urol 1991;37:1173–1179. Grekas D, Thanos V, Dioudis C, Alivanis P, Tourkantonis A: Treatment of urinary tract infections with ciprofloxacin after renal transplantation. Int J Clin Pharmacol Ther Toxicol 1993;31:309– 311. Hofbauer H, Kinzig M, Kresken M, Naber KG, Reiz A, Sörgel F, Wiedemann B: Urine bactericidal titers of norfloxacin and pefloxacin in volunteers receiving a single oral dose of 800 mg. Abstr 34th ICAAC, Orlando 1994. Fox BC, Sollinger HW, Belzer FO, Maki DG: A prospective, randomized, double-blind study of trimethoprim-sulfamethoxazole for prophylaxis of infection in renal trnasplantation. Am J Med 1990;89:255–274. Gericke KR: Possible interactions between warfarin and fluconazole. Pharmacotherapy 1993;15: 508–509. Sobh M, El-Agroudy A, Moustafa F, El-Bedewy M, Ghoneim M: Co-administration of ketoconazole in cyclosporine-treated kidney transplant recipients: A prospective randomized study. Nephrol Dial Transplant 1995;10:1072–1073. Ralph ED, Barber KR, Grant CW: Liposomal amphotericin B: An effective, nontoxic preparation for the treatment of urinary tract infections caused by Candida. Am J Nephrol 1991;11:118–122.

Dr. Silvester Krcˇméry, Department of Medicine, Old Town University Hospital, Bezrucova 5, SL–815 26 Bratislava (Slovak Republic)

Urinary Tract Infection in Renal Transplant Patients

51

PIAM:L01:BIF01011XA.88

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 52–56

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infections in Patients on Anticancer Drugs Tetsuro Matsumoto, Misao Sakumoto, Shuji Kotoh, Joichi Kumazawa Department of Urology, Faculty of Medicine, Kyushu University, Fukuoka, Japan

Anticancer drugs impair the defense mechanisms against infections including urinary tract infections (UTIs) [1]. Defense mechanisms of UTIs are cellular, humoral, and others depend on the normal urinary tract. One of the most important defenses is polymorphonuclear leukocytes, which are decreased in number and function by anticancer drugs [2]. Urological anticancer therapy combines drugs to achieve increase of anticancer effect accompanied by a decreased toxicity. The main combination chemotherapies against urologic cancers are M-VAC and CMV (see below) therapy for transitional cell carcinoma, and VAB-6 and PVB for testicular cancer. Intravesical instillation of anticancer drugs is usually performed to prevent recurrence of superficial bladder cancer. Cystitis may be due to injury of the uroepithelium following frequent catheterization. We examined the leukocyte count and function in cancer patients undergoing anticancer chemotherapy, and evaluated the incidence of infectious complications during anticancer chemotherapy and in particular intravesical instillation.

Materials and Methods Leukocyte chemiluminescence response: Leukocyte function was investigated in 8 patients (mean age 64.7) undergoing M-VAC (methotrexate, vinblastine, Adriamycin and cis-platinum) therapy by using luminol-dependent chemiluminescence response of leukocytes stimulated by formyl-methionyl-leucyl-phenylalanine (FMLP). Leukocytes were isolated from peripheral blood of patients by using Ficoll-Hypaque, indicating that the chemiluminescence detected the phagocytic and killing function of each leukocyte.

FF: W1FP

E1: IN01

$$$

➤ ➤ ➤ ➤ ➤ ➤ ➤

PIAM:L01:BIF01011XA.88

Fig. 1. Peripheral leukocyte counts in patients undergoing two courses of M-VAC therapy. Each value was mean B SD of 8 patients.

UTIs in patients on anticancer drugs: Infectious complicationswere analyzed in 82 patients who were treated by various combination anticancer chemotherapy regimens. 50 patients of M-VAC therapy, 7 of CMV (cis-platinum, methotrexate and vinblastine) therapy, 6 of VAB-6 (cis-platinum, vinblastine, actinomycin D, pepleomycin and cyclophosphamide) therapy, 8 of PVB (cis-platinum, vinblastine and bleomycin) therapy and 11 of other regimens. Intravesical instillation therapy: Intravesical instillation therapy to prevent recurrence of superficial bladder cancer was performed for 81 courses in 76 patients (33 courses of Adriamycin, 20 of epirubicin and 28 of mitomycin C plus cytosine arabinoside) during the recent 3 years. Instillation was performed 5–30 times in each course of therapy.

Results Chemiluminescence: M-VAC therapy was studied in two courses, each completed within 3 weeks. The second course was assisted by granulocyte colonystimulating factor (GCSF), but the first course was not. White cell counts in peripheral blood was reduced on days 8 and 15, and back to normal on day 30. The second course of chemotherapy was performed after the leukocyte counts had recovered (fig. 1). GCSF was administered 7 times every day from day 8. Chemi-

UTI and Anticancer Drugs

53

PIAM:L01:BIF01011XA.88

FF: W1FP

E1: IN01

$$$

✽

✽

✽

✽

Fig. 2. Chemiluminescence response of peripheral leukocytes from patients undergoing the first course of M-VAC therapy. Each value was mean B SD of 8 patients.

Table 1. Incidence of infectious complications in patients undergoing anticancer chemotherapy

Leukocytes at nadir

Patients infected

noninfected

n

%

64,000 3,000 2,000 1,000 ! 1,000

3 1 4 7 4

23 6 27 23 50

10 15 11 23 4

Total

19

23

63

luminescence response was significantly reduced on days 8 and 15, and gradually recovered on days 29–37 (fig. 2). This finding suggested that M-VAC therapy reduced not only the leukocyte count but also the leukocyte function. During the second course of M-VAC therapy the leukocyte chemiluminescence response was significantly reduced, but it normalized promptly after the administration of GCSF (fig. 3). This finding suggests that GCSF increases leukocyte count and normalizes leukocyte function impaired by anticancer drugs.

Matsumoto/Sakumoto/Kotoh/Kumazawa

54

PIAM:L01:BIF01011XA.88

FF: W1FP

E1: IN01

$$$

✽

✽ ✽

➤

➤

➤

➤

➤

➤

➤

✽

Fig. 3. Chemiluminescence response of peripheral leukocytes from patients undergoing the second course of M-VAC therapy. Each value was mean B SD of 6 patients.

UTI in patients on anticancer drugs: Infectious complications were observed in 19 (23%) patients. When the leukocyte count at nadir was over 1,000/mm3, infections were observed in 6–27% of the patients. However, 50% of the patients were infected when the leukocyte count was reduced below 1,000/mm3 (table 1). Pyelonephritis and cystitis were observed in 3 (3.7%) and 8 (9.8%) patients, respectively. All patients suffering from pyelonephritis had undergone radical cystectomy and ileal conduit before anticancer chemotherapy. The bacteria isolated from urines was Pseudomonas aeruginosa in pyelonephritis and Enterobacteriaceae (Escherichia coli or Enterobacter cloacae) in cystitis. Gram-positive organisms were isolated in other infectious complications. UTIs in patients undergoing intravesical instillation: One (1.2%) case of pyelonephritis and 6 (7.4%) cases of cystitis were observed during intravesical instillation therapy. The patient with pyelonephritis had a neurogenic bladder and vesicoureteral reflux. P. aeruginosa was isolated from 3 patients.

Discussion Clearly, granulocytopenic cancer patients are at risk of infections, not only because of a lowered number of phagocytic cells, but also because of changes in the permeability of physical defense barriers. The cancer patient is a complex

UTI and Anticancer Drugs

55

PIAM:L01:BIF01011XA.88

FF: W1FP

E1: IN01

$$$

compromised host [1]. The importance of granulocytopenia as a risk factor for developing serious infectious complications in the patients with cancer was initially elucidated by Bodey et al. [3]. They showed that a drop of the granulocyte count below 1,000/mm3 led to a 12% incidence of fever or infection. In our series, infectious complications occurred when leukocyte counts were below 1,000/mm3; then infections were found in half of the patients. The incidence of UTIs on the anticancer chemotherapy is controversial. Sickles et al. [4] reported an incidence of 26% UTI in 344 patients. Recently, 2–7% of all infectious complications associated with anticancer chemotherapy were UTIs [5]. UTIs, pyelonephritis and cystitis were observed in 3.7–9.8% among our patients. Pyelonephritis followed radical cystectomy and ileal conduit operations before anticancer chemotherapy. Hydronephrosis, reflux of urine from conduit and colonized bacteria in the conduit intestine may relate to the upper UTI. Leukocyte counts and functions were reduced by anticancer chemotherapy, but these factors were promptly normalized by concomitant use of GCSF. In conclusion, when leukocyte count is below 1,000/mm3, GCSF should be used together with anticancer chemotherapy to decrease the risk of infectious complications, including UTIs.

References 1 2 3 4 5

Pizzo PA: Granulocytopenia and cancer therapy. Past problem, current solution, future challenges. Cancer 1984;54:2649–2661. Measley R, Levison ME: Host defense mechanisms in the pathogenesis of urinary tract infection. Med Clin North Am 1991;75:275–286. Bodey GP, Buckley M, Sathe YS, Freireich EJ: Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intern Med 1966;64:328–340. Sickles EA, Greene WH, Wiernik PH: Clinical presentation of infection in granulocytopenic patients. Arch Intern Med 1975;135:715–719. Korzeniowsky OM: Urinary tract infection in the impaired host. Med Clin North Am 1991;75: 391–404.

Dr. Tetsuro Matsumoto, Department of Urology, Faculty of Medicine, Kyushu University, 3-1-1, Maidashi, Higashi-ku, Fukuoka 812 (Japan)

Matsumoto/Sakumoto/Kotoh/Kumazawa

56

PIAM:L01:BIF01012XA.89

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 57–59

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infections in Patients on Glucocorticosteroid Therapy Endre Ludwig Péterfy Teaching Hospital, Budapest, Hungary

Steroid therapy is listed among the complicating factors increasing the risk for urinary tract infections (UTIs). The consensus regarding the deleterious effect of steroids on host defense is in general based on in vitro studies and clinical observations [1]. The increased susceptibility to infections is mainly related to: (a) impaired function and diminished chemotactic activity of neutrophils; (b) reduced production of cytokines; (c) reduced macrophage activity, and (d) altered function of lymphocytes. It seems probable that cell-mediated immunity does not play an essential role in host defense of UTIs. Antibodies may decrease kidney tissue damage and may prevent colonization and recurrences [2], but the clinical relevance of these factors is uncertain. When considering the impact of steroid administration on UTIs, the following three questions should be considered: (1) Is the incidence of UTIs in patients on steroid therapy above the average? (2) Is the clinical manifestation of UTI different (more severe) in these patients compared to the situation in those without steroids? (3) Is there a need to develop special strategies for the prevention and treatment of UTIs in patients with steroid therapy? Unfortunately, the lack of data and an array of problems in the evaluation of results in this group of patients characterizes this problem. The number of controlled studies addressing any of the above questions is very small or nil and available results are conflicting. Only a few clinical situations warrant steroid monotherapies without concomitant medication. Many patients on steroid therapy also receive other immunosuppressive agents such as cyclosporines or cytostatics, etc,

PIAM:L01:BIF01012XA.89

FF: W1FP

E1: IN01

$$$

making assessment of the contribution of each factor rather challenging. The greatest proportion of patients with long-lasting systemic steroid therapy are suffering from rheumatoid arthritis, a disease which itself may cause immunosuppression due to impaired monocyte chemotactic response and depressed bactericidal function [3]. Thus, in answering the questions above, one has to resort rather to clinical experience and impressions than on well-conducted clinical trials.

Incidence of UTIs in Patients on Steroid Therapy In a prospective study, Burry [4] showed that the incidence of asymptomatic bacteriuria was 15.6% in patients with rheumatoid arthritis when they were on steroid therapy as compared to 5.6% without steroid medication. Lawson and McLean [5] in autopsied patients with rheumatoid arthritis found pyelonephritis in 47.5%, which is significantly higher than the rate of 10–20% in unselected cases. In a recently published prospective study, Karnik et al. [6] compared the effects of high-dose dexamethasone (ca. 100 mg parenterally for 3 days) and placebo in the adjunctive therapy of subarachnoidal hemorrhage. 107 patients were included in the dexamethasone group and 64 patients in the control arm. The only adverse effect of dexamethasone therapy was the higher incidence of UTIs (14 vs. 4.6%). The pathogen species were not shown. Several studies list steroid administration as a predisposing factor for UTIs caused by Candida, Pseudomonas or Enterobacteriaceae species. Corticosteroid therapy might have contributed to the development of Mycobacterium fortuitum infection in a 73-year-old patient treated for bronchial asthma [7].

Clinical Manifestations of UTIs in Patients with Steroid Therapy There are no convincing data on the more severe nature of UTIs in patients with steroid therapy. Papillary necrosis is one of the manifestations of a severe renal infection. Harvald [8] found a high incidence of papillary necrosis in patients treated with steroids, but others [5] could not confirm this finding. Detailed analysis of the clinical nature and treatment results of UTIs have not appeared in relevant publications [4, 6], although Karnit et al. [6] mentioned that cure from UTIs occurred without any problem.

Ludwig

58

PIAM:L01:BIF01012XA.89

FF: W1FP

E1: IN01

$$$

The Need of a Special Strategy for the Prevention and Treatment of UTIs in Patients on Steroids No study in the English or German literature that could suggest a significantly different approach to UTIs regarding prevention and treatment strategies in patients on steroid therapy could be found – except for some general statements. The lack of data by itself indicates that steroid monotherapy does not profoundly influence the occurrence of UTIs. The majority of patients on steroids are exposed to several other risk factors, such as other immunosuppressive drugs, catheterization, and neutropenia. This would seem to obviate any need to identify the precise effect of steroids on immunosuppression. In summary, steroid administration appears to increase the incidence of UTIs and promotes the incidence of UTIs caused by opportunistic pathogens. However, there are scant data on the basis of which one should initiate the development of specific strategies for prevention and treatment of UTIs in this cohort.

References 1 2 3 4 5 6 7 8

Van der Meer WMJ: Defects in host defense mechanisms; in Rubin RH, Young LS (eds): Clinical Approach to Infection in the Compromised Host. New York, Plenum Medical, 1994, pp 1–52. Korzeniowski OM: Urinary tract infection in the impaired host. Med Clin North Am 1991;75: 391–404. Mowat AG, Baum J: Chemotaxis of polymorphonuclear leucocytes with rheumatoid arthritis. J Clin Invest 1971;59:2541–2549. Burry HC: Bacteriuria in rheumatoid arthritis. Ann Rheum Dis 1973;32:208–211. Lawson AAH, McLean N: Renal diseases and drug therapy in rheumatoid arthritis. Ann Rheum Dis 1966;25:441–445. Karnik R, Valentin A, Prainer Ch, Stöllberger C, Slany J: Zum Problem der Steroidtherapie bei Subarachnoidalblutungen. Wien Klin Wochenschr 1990;102:1–4. Oren B, Raz R, Hass H: Urinary Mycobacterium fortuitum infection. Infection 1990;18:105–106. Harvald B: Papillitis necrotisans renalis. Nord Med 1962;67:274–278.

Dr. Endre Ludwig, Péterfy Teaching Hospital, Péterfy S u 14, H–1441 Budapest (Hungary)

UTIs in Glucocorticosteroid-Treated Patients

59

BOIS:L01:BIF01013XA.86

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 60–66

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Chronic Bacterial Prostatitis – A Clinical Reevaluation of Old Woes Wolfgang Weidner a, Martin Ludwig a, Hans-Georg Schiefer b a b

Urologische Klinik, Institut für Medizinische Mikrobiologie, Justus-Liebig-Universität Giessen, Deutschland

Prostatitis continues to be a major clinical enigma. About every second man will experience symptoms of prostatitis during his lifetime. The term ‘prostatitis’ implies an inflammatory disease, but true bacterial infection is detected in only 5–10% of patients. Essentially, three clinical types of real inflammatory prostatitis are widely accepted: acute and chronic bacterial and nonbacterial prostatitis. Chronic bacterial prostatitis (CBP) is a disease characterized by relatively asymptomatic periods between episodes of recurrent bacteriuria. It is impossible to diagnose by physical examination only. It is characterized by small numbers of bacteria in the prostatic fluid and may be difficult to eradicate by antimicrobial therapy. Most patients are asymptomatic until they have recurrent urinary tract infections, a hallmark of this condition. Bacterial lower urinary tract localizations are a prerequisite for antimicrobial therapy.

Etiology and Pathogenesis Bacteria causing both acute and CBP are identical in species, serotype and incidence to those causing common urinary tract infections, with Escherichia coli predominating. Data from three recent studies underline that prostatitis is a clinically important, but uncommon, bacterial infectious disease [1–4]. In these carefully performed examinations of patients with a history of chronic prostatitis, 5–7% [2–4] and 9% [1] met the criteria of CBP, the prostatitis histogram of the four-specimen test being used as the determining factor.

BOIS:L01:BIF01013XA.86

FF: W1FP

E1: IN01

$$$

Gram-positive organisms, such as Enterococcus faecalis and Staphylococcus epidermidis, are uncommon causes of bacterial prostatitis. Pseudomonas and other nosocomial strains should be suspected if the patient has recently had instrumentation, particularly in a hospital setting. The pathogenesis of bacterial prostatitis is unknown. Presumably, ascending urethral infection, after vaginal or rectal inoculation of the urinary meatus during sexual intercourse, plays an important role. Direct extension or lymphatic spread of fecal microorganisms, as well as hematogenous spread, are also possible, but minor routes of infection. Some investigators have postulated that reflux of bacteria inside the prostatic ducts is essential for pathogenesis [5]. The typical symptom of CBP is relapsing bacteriuria, in which the same pathogen is repeatedly detected. In a prospective study, the authors compared the results of the four-specimen technique in a group of 597 patients with symptoms of chronic prostatitis with the findings of 48 men without symptoms. Before examination, urinary tract infection, urethritis, and acute bacterial prostatitis had been excluded in all men. Whereas significant bacteriuria could not be found in any case in the control group, it was verified in 26 (4.4%) patients of the symptomatic group, in spite of normal leukocyte and bacteria counts before examination. With regard to all patients with CBP, 25% reveal periods of significant bacteriuria [5].

Classification and Bacteriological Diagnosis In order to properly identify the site of inflammation correctly, the clinician has to analyze one drop of EPS for increased numbers of white blood cells and lipid-laden macrophages. Provided the urethral and midstream urine samples show low numbers of PML, over 10 PML/hpf in EPS is considered diagnostic of prostatic inflammation [6]. The causes of elevated white blood cell count may be transient or permanent. Transient causes include premature ejaculation, urethritis, acute cystitis, or prostatitis. Chronic or long-standing inflammation is due to chronic bacterial prostatitis or to ‘nonbacterial’ prostatitis, which may be due to as yet unrecognized microorganisms. Krieger and McGonagle [1] analyzed the prostatic inflammatory response with respect to infections by Gram-positive or Gram-negative bacteria: cases with proven enterobacterial infection revealed a mean of 44 PML/hpf, while patients in whom only Gram-positive bacteria were found, had normal leukocyte counts in EPS. Recently, Wright et al. [7] confirmed the association between increased numbers of white blood cells and clinical symptomatic episodes with urinary tract infection.

Bacterial Prostatitis

61

BOIS:L01:BIF01013XA.86

FF: W1FP

E1: IN01

$$$

Table 1. Symptoms in patients with the prostatitis syndrome [from 11] Leading symptoms

Clinical features

Typical signs of inflammation

Burning urethral sensations when voiding; discharge; prostatorrhea; pyospermia

Diffuse anogenital symptoms

Pressure behind pubic bone; perineal pressure; tension in testes and epididymides; penile pain; inguinal pain; anorectal dysesthesia; lower abdominal discomfort

Voiding disturbances

Difficult urination; stranguria; frequency; nocturia

Sexual dysfunction

Loss of libido; erectile dysfunction; ejaculatory dysfunction (pain during or after orgasm)

Other symptoms

Myalgia

Diagnosis of CBP must be confirmed by quantitative bacteriologic localization studies. The diagnostic method of choice is the ‘four-specimen test’, in which approximately equal urine samples are quantitatively compared before and after prostate massage, also including a drop of EPS. A ‘prostatitis histogram’ of quantifiable pathogens is pathognomonic, i.e. a tenfold higher number of bacteria in VB3, as compared to the first voided urine (VB1). With this technique, CBP can be clearly diagnosed. The question is whether this technique really provides useful information for every patient suspected of having chronic prostatitis. Considering that between 50% [3] and 80% [1] of patients with symptoms of prostatitis attending a special outpatient clinic suffer from prostatodynia, we now limit this examination to those men in whom repeated evidence of increased leukocyte numbers in EPS has been established [8]. Ejaculate analysis is also done in these men to get further information about whether the inflamed prostate is part of a generalized infection of the male accessory glands. It should be kept in mind that for correct bacteriologic evaluation of ejaculate specimens, urethral inflammation or upper urinary tract infection must have been ruled out; only then does the bacteriologic study of semen specimens give significant results, that is to say, evidence of bacteriospermia (1103 cfu/ml) in half of the patients with CPB in contrast to 6.8% with NBP; with only one exception, bacterial species isolated from the ejaculate were identical to those from EPS [2]. These results underline the biological significance of bacteriospermia in men with CBP.

Weidner/Ludwig/Schiefer

62

BOIS:L01:BIF01013XA.86

FF: W1FP

pH value LDH 5 LDH 1 IgA, IgG IgM

E1: IN01

$$$

- Specific gravity - Prostatic antibacterial factor (PAF) - Cations (zinc, magnesium, calcium) - Citric acid - Cholesterol - Enzymes (acid phosphatase, lysozyme)

Fig. 1. Alterations in composition of prostatic secretions found in CBP patients.

Clinical Features Unlike acute bacterial prostatitis, which is easily diagnosed by acute symptoms and clinical findings, the history and physical examination of patients may suggest the diagnosis, but most signs and symptoms of CBP and nonbacterial prostatitis are identical. The clinical features of CBP are highly variable (table 1). Although so many develop chronic prostatitis following an initial bout of acute bacterial prostatitis, many have no history of acute prostatitis. Especially in men over the age of 40, recurrent epididiymitis is often noted. Although a tender, spongy prostate is often found, these findings are not specifically diagnostic of CBP [5]. Furthermore, alterations in the composition of prostatic secretions have been assumed to be diagnostic in patients with prostatitis [9]. In CBP, these alterations are sufficiently distinct to suggest an accompanying generalized secretory dysfunction of the gland (fig. 1). The prostate gland is also known to be capable of a systemic and local immune response to invading microorganisms. In CBP that has been cured by antiobiotic therapy, antigen-specific IgA in prostatic secretions – particularly secretory IgA – is elevated for almost 2 years, and antigen-specific IgG for 6 months, before both immunoglobulins slowly decrease to the previous values [10].

Bacterial Prostatitis

63

BOIS:L01:BIF01013XA.86

FF: W1FP

E1: IN01

$$$

Symptoms Symptoms include typical inflammatory signs, but also voiding disturbances and signs of sexual dysfunction [11]. Symptoms and complaint complexes have been evaluated by Brähler [12] in a highly sophisticated way, thus providing the possibility in the future to ask for ‘true’ prostatitis symptoms, using a special questionnaire. Urodynamically effective changes are found in approximately 33–43% of patients with prostatitis-like symptoms. Changes in the areas of the bladder neck, resulting from congenital or acquired pathological changes in the urethra with disturbances of the laminar urinary flow may cause prostatitis-like symptoms such as alguria and difficult micturition [5, 13]. Prostatic Calculi Prostatic calculi are known to increase with growing age and occur independently of the kind of prostatic disease. They are estimated to originate from thickened prostatic secretions and reflux of urine. The detection of prostatic calculi in chronic prostatitis is of clinical importance, because they are considered as being one of the main causes of therapeutic failure. They hinder the diffusion of antibiotics through the prostate gland tissue and, therefore, by serving as a nidus for pathogens, lead to recurrent prostatic infections. In a prospective study, we compared the sonographic findings in 88 patients with CBP and NBP with the results of 53 men suffering from prostatodynia. The frequency of solitary calculi was identical in both groups, whereas a significantly increased number of calculi and diffuse type calcifications were demonstrated [14]. Prostatovesiculitis The seminal vesicles can be figured near the upper prostatic pole and often stretch out as far as the lateral bladder regions. By transrectal prostatic sonography, a wide variety of size and shape can be seen. Asymmetry in the case of inflammation is considered either a consequence of enlargement due to ductal obstruction of the infected gland, or a result of seminal vesicle shrinking by obliteration. It indicates participation of the lower urogenital tract in the inflammatory process, particularly in prostatitis [14]. Prostatitis and Benign Prostatic Hyperplasia (BPH) Some attempts have also been made to clarify the diagnosis by perineal biopsy of ‘typical’ prostatitis areas under ultrasonographic guidance order to get a specific histology and to evaluate particular microorganisms for adequate classification [3, 15]. The question is whether there are ‘typical inflammatory findings’ under transrectal ultrasonography [16]. At present, we disbelieve in such findings.

Weidner/Ludwig/Schiefer

64

BOIS:L01:BIF01013XA.86

FF: W1FP

E1: IN01

$$$

Following the experience of Helpap [17], the ‘chronic prostatitis’ associated with paraurethral nodular prostatic hyperplasia has no significance ‘as an independent disease, with the possible exception of acute exacerbation’. In patients without BPH, periglandular prostatitis may occur in a moderate form in up to 24% of patients and in a severe form in up to 13.5%, data which are not very different from those suffering from BPH.

Conclusion Prostatitis is a widespread inflammatory disease in urologic practice. Inflammation of the prostate gland may be due to bacterial or ‘nonbacterial’ causes. Healthy men show only minimal signs of inflammation in prostatic secretions. Evaluation of a patient with prostatitis requires determination of white blood cell counts in the prostatic fluid, so that patients with elevated leukocyte numbers, i.e., suffering from true prostatitis, can be separated from those without prostatitis. Then, the search for etiologically involved bacteria is indicated using the criteria pointed out above.

References 1 2 3

4 5

6

7 8 9 10

11

Krieger JN, McGonagle LA: Diagnostic considerations and interpretation of microbiological findings for evaluation of chronic prostatitis. J Clin Microbiol 1989;27:2240–2244. Weidner W, Jantos C, Schiefer HG, Haidl G, Friedrich HJ: Semen parameters in men with and without proven chronic prostatitis. Arch Androl 1991;26:173–183. Weidner W, Schiefer HG, Krauss H, Jantos C, Friedrich HJ, Altmannsberger M: Chronic prostatitis: A thorough search for etiologically involved microorganisms in 1,461 patients. Infection 1991; 19(suppl 3):119–125. Weidner W, Ludwig M, Brähler E, Schiefer HG: Diagnosis of 656 patients with prostatitis syndromes. J Urol 1995;153(suppl):A328. Weidner W, Ludwig M: Diagnostic management in chronic prostatitis; in Weidner W, Madsen PO, Schiefer H-G (eds): Prostatitis: Etiopathology, Diagnosis and Treatment. New York, Springer, 1994, pp 49–65. Schaeffer AJ: Etiology pathogenesis and inflammatory reactions in chronic bacterial prostatitis; in Weidner W, Madsen PO, Schiefer HG (eds): Prostatitis: Etiopathology, Diagnosis and Treatment. Heidelberg, Springer, 1994, pp 151–157. Wright ET, Chmiel JS, Grayhack JT, Schaeffer AJ: Prostatic fluid inflammation in prostatitis. J Urol 1994;152:2300–2303. Weidner W: Prostatitis-diagnostic criteria, classification of patients and recommendations for therapeutic trials. Infection 1992;20(suppl 3):227–231. Meares EM Jr: Prostatitis; in Chisholm GT, Fair WR (eds): Scientific Foundations of Urology, ed 3. London, Heinemann, 1990, pp 373–378. Shortliffe LMD, Wehner N, Stamey TA: Use of a solid-phase radioimmunoassay and formalinfixed whole bacterial antigen in the detection of antigen-specific immunoglobulin in prostatic fluid. J Clin Invest 1981;67:790–799. Weidner W: Moderne Prostatitisdiagnostik. München, Zuckschwerdt, 1984.

Bacterial Prostatitis

65

BOIS:L01:BIF01013XA.86

12

13 14

15 16 17

FF: W1FP

E1: IN01

$$$

Brähler E: Complaint complexes and psychosomatic aspects; in Weidner W, Madsen PO, Schiefer H-G (eds): Prostatitis: Etiopathology, Diagnosis and Treatment. Heidelberg, Springer, 1994, pp 40–48. Blacklock NJ: The anatomy of the prostate: Relationship with prostatic infection. Infection 1991;19 (suppl 3):111–114. Ludwig M, Weidner W, Schroeder-Printzen I, Zimmermann O, Ringert R-H: Transrectal prostatic sonography as a useful diagnostic means for patients with chronic prostatitis or prostatodynia. Br J Urol 1994;73:664–668. Doble A, Walter MM, Harris JRW, Taylor-Robinson D, Witherow R: Intraprostatic antibody deposition in chronic abacterial prostatitis. Br J Urol 1991;65:598–605. Christiansen E, Purvis K: Diagnosis of chronic abacterial prostato-vesiculitis by rectal ultrasonography in relation to symptoms and findings. Br J Urol 1991;67:173–176. Helpap B: Pathology of chronic non-specific prostatitis; in Vahlensieck W, Rutishauser G (eds): Benign Prostate Diseases. Stuttgart, Thieme, 1992, pp 33–48.

Prof. Dr. Wolfgang Weidner, Urologische Klinik, Klinikstrasse 29, D–35392 Giessen (Germany)

Weidner/Ludwig/Schiefer

66

BOIS:L01:BIF01014XA.87

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 67–73

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Relevance of NCCLS Breakpoints for Susceptibility as Applied to Urinary Tract Infections Ronald N. Jones Medical Microbiology Division, Department of Pathology, University of Iowa College of Medicine, Iowa City, Iowa, USA

The National Committee for Clinical Laboratory Standards (NCCLS) has published antimicrobial susceptibility testing standardized methods and associated interpretive criteria for susceptibility since the 1970s. These documents are the most widely used methodology guidelines in the United States and the world. Specific methods are suggested that have high accuracy, reproducibility, and predictive value of clinical responses when performing disk diffusion, agar or broth dilution, and anaerobic procedures [1–3]. Additional NCCLS guidelines, M23-A [4], have been drafted outlining the laboratory and clinical data necessary for the NCCLS Expert Subcommittee to determine interpretive criteria and quality control limits. The manufacturers of the antimicrobial agents use M23-A to present a comprehensive information base for the NCCLS decision-making process.

NCCLS Document Tables The Table 1 of each NCCLS antimicrobial susceptibility testing method document [1, 2] lists those agents that have acceptable, predictive tests. Furthermore, these drugs are categorized into reporting groups (A, B, C, D), indicating whether a laboratory should test as a routine and report results (group A) or test and/or selectively report results (groups B and C). The group D antimicrobials are those drugs specifically used only for urinary tract infections (UTI) indexed by a given genus group (examples: oral carbenicillin, trimethoprim, sulfonamides, some fluoroquinolones, and some oral cephems). Although these groups are listed in

BOIS:L01:BIF01014XA.87

FF: W1FP

E1: IN01

$$$

Table 1. Clinical efficacy of 100 mg nitrofurantoin qid in the treatment of UTI (665 patients; 38 medical centers). Year

Location

Patients studied

Bacteriological cure rate, %

1987 1983 1983 1982 1982 1981 1980 1979 1977

USA Mexico USA USA USA USA Mexico UK India

153 20 256 108 26 46 17 23 16

89 90 94 95 92 91 88 100 100

NCCLS documents, the batteries are only suggestions and the ultimate choice of tested and reported antimicrobials is a local laboratory and medical care setting issue. The NCCLS Table 1 (multiple types) contains interpretive criteria (mg/l MICs; zones in mm) that indicate susceptible, intermediate, and resistant categories. These criteria were selected by a combination of databases that include: serum/‘tissue’ kinetics and concentrations, antimicrobial spectrum, drug-organism MIC and zone diameter population distributions, toxicity information (if appropriate), and the bacterial eradication/clinical cure and improvement rates correlated to the results of the standardized (NCCLS) tests. For the infrequently analyzed UTI-specific compounds other breakpoint selection parameters ‘have been’ applied including a MIC that is approximately 10-fold below the mean urinary tract level and/or a MIC that would predict a clinical efficacy of 685%. Breakpoints for a large number of drugs (29 since 1975) have been considered for systemic and UTI indications, but only three drugs (nitrofurantoin, norfloxacin, fosfomycin) have had more limited UTI claims. The nitrofurantoin breakpoints (^32 mg/l = susceptible and 64 mg/l = intermediate) were validated by comprehensive analysis including a literature clinical review (table 1) and also correlations of nitrofurantoin in vitro test results to clinical success. Similarly, the norfloxacin breakpoints were re-addressed in the light of the package insert recommendation for susceptibility (^16 mg/l) that was significantly greater than that found in the NCCLS Tables (table 2) [3, 4]. Examination of the initial, reported norfloxacin regression experiments [5] in 1983, indicated a discord in the correlate MIC that would equal the recommended susceptible

Jones

68

BOIS:L01:BIF01014XA.87

FF: W1FP

E1: IN01

$$$

Table 2. Example of clinical response criteria applied to an early fluoroquinolone (norfloxacin) having a UTI-specific indicationa MIC or disk test results

Number tested

Bacterial eradication, %

Disk 621 mm 17–20 mm 13–16 mm ^12 mm

860 102 52 10

97 85b 77 80

MIC ^1 mg/l 2 mg/l 4 mg/l 68 mg/l

225 12 13 3

92 75b 54 67

a NCCLS Agenda Book, January 1988. NDA, post-NDA and open-label clinical trial for uncomplicated and complicated UTI through 1987. b NCCLS breakpoints for UTI organisms, see text.

breakpoint zone of 617 mm (e.g. ^4 vs. ^16 mg/l). UTI trial results for norfloxacin confirmed that a 621-mm zone diameter predicted a high probability of therapeutic success (97%; table 1), but organisms with MICs for norfloxacin of 62 mg/l were more refractory (54–75% success) to therapy. A consensus opinion by the NCCLS Subcommittee leads to the current norfloxacin interpretive breakpoints of 617 mm that predicts 685% clinical success, with a MIC correlate of ^4 mg/l by regression analyses [5, 6]. This decision was strengthened by later norfloxacin open-label UTI trial success rates of 97% for patient pathogens having zone diameters of 617 mm (1,665 cases). The MIC database was too limited to direct the breakpoint ‘titration’ process (only 16 cases with MICs of 64 mg/l), consequently the ^4 mg/l that correlated to the 617-mm breakpoint for susceptibility seemed valid. More recently, preliminary clinical and microbiology data were presented to the NCCLS for fosfomycin UTI trials in the United States that utilized, for the first time, a single-dose regimen (3 g). Clinical success was demonstrated in 686% (644 cases) of patients with pathogens having fosfomycin MICs of ^4 mg/l (table 3). However, lower, unacceptable rates of pathogen eradication (58–78%) were observed for isolates having fosfomycin MICs up to the break-

NCCLS UTI Breakpoints

69

BOIS:L01:BIF01014XA.87

FF: W1FP

E1: IN01

$$$

Table 3. Example of clinical response criteria applied to fosfomycin trial results (3 g single dose)a Fosfomycin MIC, mg/l

Patients Response

^2 4b 8 16 32 64 6128

540 20 22 19 18 14 11

cured

failed

463 18 16 11 14 10 7

77 2 6 8 4 4 4

% success 86 90 73 58 78 71 64

a

From NCCLS Agenda Book, June 1995. Breakpoint MIC with acceptable clinical efficacy (685%). b

point MICs of ^128 mg/l used in other nations for ‘multidose’ therapy. If a lower breakpoint of ^4 mg/l were selected as representing susceptibility, the proposed disk diffusion test method for this drug would not be usable because of unacceptable high error rates [7]. The selection of the appropriate breakpoint awaits further information from the manufacturer and continued discussion by the NCCLS Subcommittee and the US FDA.

UTI Breakpoints Using NCCLS Systemic Criteria The fluoroquinolone compounds represent an excellent group of therapeutic agents for UTIs and are good examples of the ongoing NCCLS process of applying the same breakpoint to multiple clincal indications. Table 4 lists the pharmacokinetic features of four widely used quinolones at commonly prescribed dosing schedules. Pharmacodynamic considerations suggested by Forrest et al. [8] indicate that highest clinical success with fluoroquinolones (intravenous regimen, systemic disease) can be expected when the area under the curve (AUC) divided by the infecting pathogens MIC was 6125 mg W h/l. Other authors have determined similar values in animal models (6100 mg W h/l) with static or minimal effect at approximately 30–35 mg W h/l. This latter statistic should be sufficient for therapy of UTIs where the drug concentrations are markedly elevated (table 4), e.g. an elevated urine AUC.

Jones

70

BOIS:L01:BIF01014XA.87

FF: W1FP

E1: IN01

$$$

Table 4. Fluoroquinolone pharmacokinetic parameters Drug (usual dose)

T1/2, h Cmax, mg/l

AUC mg W h/l

Norfloxacin (400 mg ! 2) Ciprofloxacin (500 mg ! 2) Ofloxacin (400 mg ! 2) Lomefloxacin (400 mg ! 1)

3–4 4 6 8

5–7 7–10 29–35 26–37

1.5 2.0–2.9 5.2–11.0 3.0–5.2

Urine conc. mg/l 80–250 400–500 300–700 1300

% in urine 30 30 90 90

Table 5. Success rates of lomefloxacin therapy of UTI caused by S. saprophyticus and all pathogens indexed by organism MICa MIC, mg/l

Strains, n

Bacteriologic eradication S. saprophyticus all species

^0.5 1 2 4 68 a b

402 (0)b 13 (1) 32 (6) 10 (2) 6 (0)

– 100 100 100 –

96 92 81 90 67

NCCLS Agenda Book, January 1991. S. saprophyticus strains in parentheses.

When this ‘30–35 mg W h/l rule’ was used for the currently available fluoroquinolones, the results dictated the following breakpoint MICs (rounded to nearest log2 dilution): norfloxacin at ^0.25 mg/l; ciprofloxacin at ^0.5 mg/l; ofloxacin at ^2 mg/l, and lomefloxacin at ^1 mg/l. Other parameters were then utilized by the NCCLS Subcommittee to adjust these calculated breakpoints such as UTI response rates (norfloxacin, lomefloxacin), systemic infection response rates (ciprofloxacin), and organism MIC population analyses (all fluoroquinolones). An example of the latter is shown in table 5 showing that the Staphylococcus saprophyticus strains have a lomefloxacin MIC mode at 2 mg/l (range 1–4 mg/l). All S. saprophyticus strains were eradicated by single-dose (400 mg) lomefloxacin therapy, but the ^1 mg/l breakpoint for susceptibility assigned by pharmacodynamic criteria alone would have predicted some clinical failures. The elevation of the

NCCLS UTI Breakpoints

71

BOIS:L01:BIF01014XA.87

FF: W1FP

E1: IN01

$$$

lomefloxacin breakpoint MIC to ^2 mg/l minimized the error for this species and similar problems that occurred with Pseudomonas aeruginosa strains (data on file, NCCLS Agenda Book, 1991). The clinical success rate of lomefloxacin was very acceptable for all organisms having MICs of ^4 mg/l (table 5).

Conclusions The NCCLS tables separate drugs according to clinical indications, so antimicrobial agents with UTI-only claims are found in a Section D. When breakpoints are assigned, pharmacokinetic and pharmacodynamic considerations and other microbiologic features will lead to the most accurate selection. Clinical results and urine pharmacokinetic characteristics are principally used for drugs having unique UTI application, whereas the systemic qualities of the drug will dictate the breakpoints for antimicrobials with multiple clinical indications. UTI-specific breakpoints appear to be problematic. Lumpkin [9] stated in 1994 that rational dosing schedules with decreased frequency and amount are becoming more important to the pharmaceutical industry. Therefore, registry agencies such as the US FDA ‘have been interested in exploring some dose-ranging trial designs and in working with manufacturers to answer some of these basic questions’ [9]. By such investigations, a common breakpoint could be assigned for both systemic and UTI indications, but, where possible, the appropriate dose for the treatment of UTI could be significantly reduced (in length of therapy and/or volume of dosing) without compromising efficacy. The flaws of the current systems of breakpoint selection is the ‘self-fulfilling prophecy’ concept. Preliminary, usually conservative breakpoints are selected for clinical trials, leading to only rare treated patients having pathogens with elevated MICs or corresponding small-zone diameters. This limits the data necessary to accurately ‘titrate’ the in vivo response with the in vitro test results. Early phase I and II, controlled application of pharmacodynamic and dose-ranging investigations [9] may lead to more accurate finalized breakpoints predicting susceptibility and acceptable clinical responses (685%). The other factors that should be considered in this process include the unique qualities of the urine as a body fluid site of infection: pH, inoculum, divalent cation concentrations, host factors, and organism mechanisms of infecting adhesion. To date, breakpoints used by the NCCLS have generally predicted a high level of potential clinical success because the breakpoints are lower (directed by systemic infection claims) than might be appropriate for a UTI. It is hoped that future selections of breakpoints for susceptibility can be more standardized via international consensus [1–4], common criteria, and more rational use of early dose-ranging investigations [9].

Jones

72

BOIS:L01:BIF01014XA.87

FF: W1FP

E1: IN01

$$$

References 1 2

3

4

5

6

7 8

9

National Committee for Clinical Laboratory Standards (NCCLS): Performance standards for antimicrobial disk susceptibility tests: Approved Standard M2-A5. Villanova/PA, NCCLS, 1993. National Committee for Clinical Laboratory Standards (NCCLS): Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Approved Standard M7-A3. Villanova/ PA, NCCLS, 1993. National Committee for Clinical Laboratory Standards (NCCLS): Methods for antimicrobial susceptibility testing of anaerobic bacteria. Approved Standard M11-A3. NCCLS publ M11-A3. Villanova/PA, NCCLS, 1993. National Committee for Clinical Laboratory Standards (NCCLS): Development of in vitro susceptibility testing criteria and quality control guidelines. Approved Guideline M23-A. Villanova/PA, NCCLS, 1994. Shungu DL, Weinberg E, Gadebusch HH: Tentative interpretive standards for disk diffusion susceptibility testing with norfloxacin (MK-0366, AM-715). Antimicrob Agents Chemother 1983;23: 256–260. Cormican MG, Jones RN: Re-evaluations of disk diffusion testing interpretive criteria for lomefloxacin and norfloxacin using fluoroquinolone-resistant isolates. Diagn Microbiol Infect Dis 1995;21: 227–230. Pfaller MA, Barry AL, Fuchs PC: Evaluation of disk susceptibility testing of fosfomycin trimethamine. Diagn Microbiol Infect Dis 1993;17:67–70. Forrest A, Nix DE, Ballow CH, Goss TF, Birmingham MC, Schentag JJ: Pharmacodynamics of intravenous ciprofloxacin in seriously ill patients. Antimicrob Agents Chemother 1993;37:1073– 1081. Lumpkin MM: Clinical trials from a regulator’s perspective. Infection 1994;22(suppl 1):61–65.

Dr. Ronald N. Jones, Department of Pathology, 5232 RCP, University of Iowa College of Medicine, Iowa City, Iowa 52242 (USA)

NCCLS UTI Breakpoints

73

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 74–83

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Antibacterial Activity of Antibacterial Agents in Urine: An Overview of Applied Methods Kurt G. Naber Urologic Clinic, Elisabeth Hospital, Straubing, Germany

The activity of antibacterial agents is usually measured under standardized conditions in nutrient broth or on agar enriched with a defined concentration of nutritive substances. Various methods can be applied, e.g. the determination of the minimal inhibitory (MIC) or bactericidal (MBC) concentrations of an antibacterial agent, the rapidity or degree of bacterial killing and the time period without regrowth, etc. The advantage of using artificial, but well-defined media is obvious: under the same methodological conditions, results become reproducible and, thus, readily comparable from one laboratory to the other. Since, however, the antibacterial activity of an antibacterial agent depends on the strain to be tested and its metabolic state within a specific environment on the one hand, and on the actual antibacterial concentration reaching the pathogen including possible interaction with other substances present on the other hand, the results obtained in artificial media cannot be transferred to clinical situation without cation. Infections occur in parts of the body with very different environmental conditions and with varying accessibility for antibacterial agents. Thus, recommendations concerning therapeutic indications and dosage regimens derived from in vitro results and pharmacokinetic studies in usually healthy volunteers alone are not sufficient. They have, finally to be evaluated in clinical, preferably comparative studies. Clinical studies, however, if performed according to valid protocols with enough patients to produce statistically assessable results, consume much time and manpower and, therefore, become rather expensive. If, in addition, the final results do not meet expectations compared with standard treatment, the whole concept of the clinical study might subsequently be doubted as far as economic or ethical aspects are concerned. The most economic and promising approach would

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

probably be to supply additional preclinical information on the possible efficacy of the antibacterial agent concerned for a given infection. Since, obviously, the antibacterial activity of antibacterial agents in urine is of great importance to the outcome of treatment of urinary tract infections (UTIs), several methods have been developed to determine the antibacterial activity of antibacterials in urine. The following intends to give an overview of various methods available and their possible relevance.

Category of Methods The various methods can be classified as follows: (1) In vitro tests using urine, usually pooled urine, as medium (broth or agar) to determine which MIC or MBC are, respectively, able to inhibit or to kill bacteria during incubation for 18–24 h. (2) In vitro pharmacokinetic studies, mostly used as a ‘bladder’ model, measuring the bacterial elimination from a system containing urine, artificial urine, or broth with varying antibacterial concentrations imitating the pharmacokinetics of urinary concentrations after drug administration, and the different status of urodynamics with rhythmic bladder emptying and various amounts of residual urine. (3) Ex vivo/in vitro studies, i.e. urine from treated volunteers or patients is obtained in time intervals and for this urine the bactericidal or bacteriostatic titers, i.e. the highest dilution factor which is able to inhibit growth or to kill bacteria during incubation for 18–24 h, or the kinetics of bacterial elimination after inoculation with various uropathogens can be determined. (4) In vivo studies: the kinetics of pathogen elimination from the urine of patients with UTIs are studied after administration of an antibacterial agent. The results obtained by such preclinical trials can support the choice of the right dosage regimen when setting up a protocol for a clinical study. This approach is important because dose-finding studies are usually performed with a number of patients which is usually too small to allow statistical tests of equivalence. Pilot studies in principle inherently entail the problem of possibly using either too low a dose or of achieiving an unacceptably high rate of adverse events in at least some of the patient groups (i.e. dose sizes).

Minimal Inhibitory and Bactericidal Concentrations in Urine The MICs in urine expressed as ratios of the MICs in broth of ß-lactams, aminoglycosides, cotrimoxazole and nitrofurantoin for Escherichia coli, Klebsiella spp., Proteus mirabilis and Pseudomonas aeruginosa have been published by Helm [1, 2]. Table 1 shows that MICs of ß-lactams, like ampicillin and cefazolin,

Antibacterial Activity in Urine

75

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

Table 1. Relative activity1 of antibacterial agents in urine against uropathogens [2]

Ampicillin Cefazolin Gentamicin Cotrimoxazole Nitrofurantoin

E. coli

Klebsiella spp.

1–1–1 1–2–2 1/128–1/64–1/16 1/16–1/8–1/8 1–2–2

– 1/4–1/4–1/2 1/64–1/32–1/8 1/16–1/8–1/4 1–2–2

P. mirabilis !1/64 !1/512 1/16–1/16–1/4 !1/64 2

P. aeruginosa Carbenicillin Azlocillin Gentamicin Tobramycin

1/2–1–2 1/32–1/2–1 1/64–1/16 1/128–1/32

1 Activity measured as MIC in urine in relation to MIC determined according to standard susceptibility methods.

or cotrimoxazole against P. mirabilis are significantly higher in urine than in standard laboratory test media. Such circumstances may probably explain some treatment failures. The antibacterial activity of nitrofurantoin is not much influenced by urine as a medium. Regarding aminoglycosides, a reduced activity in urine, compared to nutrient broth is widely recognized and has also bein demonstrated by Helm [1, 2] for the four bacterial species tested. Goi et al. [3] compared the MICs of cefmenoxime, cefotaxime, and latamoxef for 11 clinical pathogens which had caused complicated UTIs. For this purpose, the bacteria in urine with pH 5.5, 7.0, and 8.5 and inocula varying from 103 to 107 CFU/ml were suspended, and results compared to those of the standard susceptibility test method using Mueller-Hinton agar. It is notable that MICs determined under standard conditions corresponded to the MICs in urine only under certain conditions. These conditions, however, varied from strain to strain and from cephalosporin to cephalosporin. In certain cases, a low inoculum and in other cases a high inoculum gave MICs similar to that obtained by standard methods. Neutral and alkaline urine pH corresponded better with the standard method than pH 5.5. In general, a higher inoculum produced higher MICs, as already expected. Considering inoculum and pH together, the MICs of a specific cephalosporin obtained for a given bacterial strain varied from 3 dilution steps (latamoxef for E. coli) up to more than 12 dilution steps, i.e. 1 1,000-fold (e.g. cefotax-

Naber

76

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

ime and latamoxef for Enterobacter cloacae). In other words, if the recommended break points of MICs had been applied, some strains could have been rated as either susceptible or resistant, to cefotaxime and latamoxef, depending on the test conditions. This was mainly true for ß-lactamase producing Staphylococcus aureus, Morganella morganii, Serratia marcescens, E. cloacae, and P. aeruginosa. Urine inhibits the activity of aminoglycosides. The decreased activity has been correlated with a low pH and a high osmolarity caused by high salt and glucose concentrations [4–9]. This may explain occasional treatment failures. The presence of betaines in culture media, especially glycine betaine and proline betain, increases osmotolerance and induces rapid bacterial growth at high osmolarities [10]. Since glycine betaine and proline betaine are found at significant concentrations in urine, Peddie and Chambers [11] investigated the effects of betaines and urine on the antibacterial activity of of gentamicin on E. coli. Concentrated urine and low pH rendered high MICs which were further increased by glycine betaine, which had a more marked impact than sodium chloride or glucose. These results suggest a synergism by the osmoprotective effects of glycine betaine and the inhibitory effect of sodium chloride or glucose against aminoglycosides. Many studies on quinolones show that MICs are higher in urine than standard test media. Leigh et al. [12] investigated the MICs and MBCs of lomefloxacin, ofloxacin, norfloxacin, pefloxacin, and ciprofloxacin in Mueller-Hinton broth and in urines at pH 5 and 7 against two strains of P. aeruginosa and E. coli. Urine increased the MICs and MBCs by several dilution steps, particularly at pH 5. A high MgSO4 concentration also attributes to high MICs, whereas other urine components caused only little additional MIC increases [13, 14]. Besides the MIC, the postantibiotic effect is also influenced by human urine, and in particular its pH [15]. Since the effect of urine might be different for the various quinolones, it should be mandatory to determine the MICs of a new quinolone, not only by standard methods but also in urine and at different pHs against selected uropathogens, as has been done by several authors [16–19].

In vitro Pharmacodynamic Studies (Bladder Model) In vitro models of the bladder [20] have been used to study interactions between urine flow, bladder washout, residual urine, bacterial growth, and antibiotic inhibition under varying urinary antibiotic concentrations according to simulation of human pharmacokinetics. A number of workers have described devices with varying degrees of sophistication [21–26].

Antibacterial Activity in Urine

77

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

The bladder model of Greenwood and O’Grady [24] at 37 ° C uses a dense bacterial culture which is diluted at a rate which simulates the flow of ureteric urine accompanied by intermittent ‘micturition’ episodes, which removes liquid culture aliquots. The inflow can be altered to simulate normal, reduced or increased ureteric flow rates; the frequency of ‘micturition’ can be readily varied. Bacteria are exposed to changing concentrations of antibacterial agents to simulate the consequences of renal excretion. The bacterial growth and the response of the culture to a drug are continuously monitored by photometric turbidity measurements. The authors prefer broth to urine since it is more convenient to work with and also provides more reproducible growth conditions. This bladder model was used for example to help to elucidate (a) optimal dosage of ß-lactam antibiotics, (b) the clinical relevance of synergy between trimethoprim and sulfamethoxazole and between penicillins and clavulanic acid, and (c) the factors influencing the emergence of resistance to quinolones [27]. Anderson et al. [25] subsequently developed a similar model which differed in two respects: (1) the bacterial population was determined at intervals by viable counts, which is a more reliable approach, especially for dense suspensions of organisms, and (2) urine was chosen as medium instead of laboratory media because (1) many organisms grow more slowly in urine than in broth and (b) the activity of various antibiotics is altered by urine compared to traditional laboratory media [25]. Whereas the earlier in vitro bladder models approximately imitate uncomplicated cystitis, Nishimura et al. [28] developed an in vitro kidney-bladder model imitating severely complicated UTI with foreign bodies and biofilm formation. These authors studied the significance of the peak antimicrobial concentration and time above MIC; they concluded that the two factors necessary to obtain proper antimicrobial effects were (a) a certain high antibiotic concentration and (b) maintenance of inhibitory concentrations in a certain period of time. According to this model, the efficacy of clarithromycin in inhibiting biofilm formation was also studied [29]. Whereas ciprofloxacin was able to eradicate P. aeruginosa from the bladder model culture medium after 32 h, only a combination of ciprofloxacin and clarithromycin prevented bacterial regrowth. Alginate, a major component of P. aeruginosa biofilm, was increased slightly on day 3 when ciprofloxacin was given alone, but decreased, to below the detection limit, on days 5 and 7 when ciprofloxacin was combined with clarithromycin.

Ex vivo/in vitro Studies In such studies, urine is collected after administration of antibiotic agents. Casellas et al. [30] collected urine of children at 0, 2, 4, 8, 16, 24 and 48 h after the first dose of cefpodoxime proxetil (10 mg/kg b.w.). The urine was sterilized by

Naber

78

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

Table 2. Urinary bactericidal titer (UBT) of pefloxacin and norfloxacin in female volunteers (A, B, C) after a single oral dose of 800 mg: UBT was determined as the reciprocal value of the highest bactericidal dilution against the strains tested: urine was diluted either with antibiotic-free urine (urine) of each individual subject compared to Mueller-Hinton broth (MHB) [42] Time h

Volunteer A

Volunteer B

Volunteer C

pefloxacin

norfloxacin

pefloxacin

norfloxacin

pefloxacin

norfloxacin

urine MHB

urine MHB

urine MHB

urine MHB

urine

urine

32 512 64 61,024 128 61,024 16 512

256 61,024 64 61,024 16 256 4 64

256 256 128 64

512 61,024 512 61,024 61,024 61,024 256 61,024 61,024 61,024 512 256 64 512 512 512 64 64 8 64 256 256 32 32

E. coli ATCC 25922 Nal-S

0–6 6–12 12–24 24–48

E. coli Nal-R

0–6 6–12 12–24 24–48

4 4 0 0

64 128 256 64

K. pneumoniae Nal-R

0–6 6–12 12–24 24–48

4 8 8 2

S. saprophyticus

E. faecalis

61,024 61,024 61,024 61,024

MHB

MHB

16 61,024 4 128 0 64 0 0

32 256 64 61,024 32 256 8 128

32 32 0 0

512 256 64 0

32 64 32 32

128 128 64 64

64 32 8 4

128 32 8 4

64 64 28 32

32 8 2 0

512 128 32 0

32 256 64 61,024 32 256 16 128

64 32 4 0

512 256 64 0

64 64 128 32

64 64 64 32

128 64 8 4

64 32 8 2

0–6 6–12 12–24 24–48

8 61,024 8 128 8 128 4 32

8 2 4 0

16 8 4 1

16 32 16 8

64 128 128 32

32 8 2 0

128 128 16 0

4 2 2 2

32 64 32 16

4 2 1 0

16 16 4 0

0–6 6–12 12–24 24–48

1 4 2 1

4 2 0 0

128 16 4 0

16 61,024 16 61,024 16 256 2 32

2 16 0 0

128 64 16 0

8 8 8 2

32 32 32 16

8 2 1 0

32 8 2 0

16 16 32 8

filtration and time-killing curves of E. coli were studied. At 8 and 24 h, all the urines showed a 3 to 5 log10 reduction of the initial inoculum, this occurred with all strains including a TEM-1 hyperproducer strain. Dubini and Riviera [31] studied the elimination of E. coli, P. mirabilis and K. pneumoniae using actual urine and an in vitro bladder model [24]. Whereas fosfomycin, norfloxacin and pipemidic acid eliminated the bacteria within 2–4 h, cotrimoxazole only reduced the bacterial load. Interestingly, the bacteria still present in the urine containing cotrimoxazole remained susceptible to this antibacterial agent. The highest possible dilutions exhibiting inhibitory or bactericidal activity after incubation for 18 h provide valuable additional quantitative information and can be considered analogues to the method estimating the bactericidal activi-

Antibacterial Activity in Urine

79

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

ty of serum when tested against the patients’ own infecting organisms, as first described by Schlichter et al. [32, 33]. No standard medium has been used to dilute the urine. For example, Klastersky et al. [34] used saline, Guibert et al. [35–37] phosphate buffer, Cruziani et al. [38] a 1:1 mixture of saline and broth, Soussy et al. [39] and Bergan [40] nutrient broth, whereas Wiedemann et al. [41] used broth with an addition of 10 mg/l glucose-6-phosphate. Since the diluent medium may influence bacterial growth and antibacterial activity, our group [42, 43] has studied urinary bactericidal titers of norfloxacin, pefloxacin and fleroxacin in volunteer urine diluted with antibiotic-free urine of each individual. It could be demonstrated that two to three dilution steps more are found for Mueller-Hinton broth than urine as diluent (table 2), although the clinical relevance of these findings remains to be established.

In vivo Studies To assess the effectiveness of an antibacterial agent directly, the kinetics of pathogen elimination from the urine of patients with a current UTI can be studied by sampling urine at various time intervals after drug administration. Acar [44] quantitated bacteriuria in 156 therapeutic courses and concluded that counts reduced within 24 h to below 100/ml reflect susceptibility. The ‘intermediate’ isolates, with higher MICs, are eliminated more slowly. Treatment failures occurred when the infecting organism was replaced by a resistant species or selection of a resistant variant of the original strain. Helm et al. [45] used a membrane filtration technique, and demonstrated in patients with uncomplicated UTIs that ß-lactam antibiotics (bacampicillin and amoxycillin) decreased the colony count faster than cotrimoxazole within the first 4 h. After 24 h, far more samples of patients treated with the former two were sterile than when treated with cotrimoxazole. During the first 8 h, cefroxadine killed more rapidly than cephalexin [46]. In 5 patients, fosfomycin-resistant bacteria were detected during and immediately after the end of therapy, without causing an infection, however [47].

Conclusion Various in vitro, ex vivo/in vitro, and in vivo methods and models determining the activity of antibacterial agents in urine have been described. None of these methods is so far standardized or officially accepted for the development of new agents for the treatment of UTIs. All the methods should provide predictive infor-

Naber

80

BOIS:L01:BIF01015XA.77

FF: W1FP

E1: IN01

$$$

mation faster and less expensively than animal experiments or clinical studies. They should also guide the potential value of new agents and should help in predicting optimal dose regimens [48]. Inherent limitations of all these methods, however, should be kept in mind.

References 1 2 3 4 5

6 7 8 9 10

11 12 13 14 15

16 17

18

19

Helm EB, Stille W: Resistenzbestimmungen von Antiobiotika in Körperflüssigkeiten. Immun Infekt 1978;6:217–222. Helm EB: Wirkungskinetik von Chemotherapeutika bei Harnwegsinfektionen; in Stille W (ed): Kurzzeittherapie von Harnwegsinfektionen. München, Zuckschwerdt, 1983, pp 14–21. Goi H, Watanabe T, Hara T, Ishii T, Kazuno Y, Inouye S: Antibacterial activity of cefminox in human urine. Jpn Antibiot 1988;41:25–36. Mou TW: Effect of urine pH on the antibacterial activity of antibiotics and chemotherapeutic agents. J Urol 1962:87:978–987. Medeiros AA, O’Brien TF, Wacker WEC, Yulug NF: Effect of salt concentration on the apparent in vitro susceptibility of Pseudomonas and other Gram-negative bacilli to gentamicin. J Infect Dis 1971;124:S59–S64. Beggs WH, Andrews FA: Role of ionic strength in salt antagonism of aminoglycoside action on Escherichia coli and Pseudomonas aeruginosa. J Infect Dis 1976;134:500–504. Minuth JN, Musher DM, Thorsteinsson SB: Inhibition of the antibacterial activity of gentamicin by urine. Infect Dis 1976;133:14–21. Papapetropoulou M, Papavassiliou J, Legakis NJ: Effect of the pH and osmolality of urine on the antibacterial activity of gentamicin. J Antimicrob Chemother 1983;12:571–575. Papapetropoulou M, Giamarellou H: The antibacterial activity of gentamicin in the glucosuric environment. Drugs Exp Clin Res 1985;11:793–795. Chambers S, Kunin CM: The osmoprotective properties of urine for bacteria: The protective effect of betaine and human urine against low pH and high concentrations of electrolytes, sugars and urea. J Infect Dis 1985;152:1308–1316. Peddie BA, Chambers ST: Effects of betaines and urine on the antibacterial activity of aminoglycosides. J Antimicrob Chemother 1993;31:481–488. Leigh DA, Tait S, Walsh B: Antibacterial activity of lomefloxacin. J Antimicrob Chemother 1991; 27:589–598. Machka K, Braveny I: Inhibitorische Wirkung verschiedener Faktoren auf die Aktivität von Norfloxacin. Fortschr Antimikrob Antineoplast Chemother 1984;3:557–562. Ratcliffe NT, Smith JT: The mechanism of reduced activity of 4-quinolone agents in urine. Fortschr Antimikrob Antineoplast Chemother 1984;3:563–569. Zhanel GG, Karlowsky JA, Davidson RJ, Hoban DJ: Influence of human urine on the in vitro activity and postantibiotic effect of ciprofloxacin against Escherichia coli. Chemotherapy 1991;37: 218–223. Tanaka M, Hoshino K, Ishida H, Sato K, Hayakawa I, Osada Y: Antimicrobial activity of DV7751a, a new fluoroquinolone. Antimicrob Agents Chemother 1993;37:2122–2118. Guinea J, Robert M, Gargallo-Viola D, Xicota MA, Garcia J, Tudela E, Esteve M, Coll R, Pares M, Roser R: In vitro and in vivo antibacterial activities of E-4868, a new fluoroquinolone with a 7azetidin ring substituent. Antimicrob Agents Chemother 1993;37:868–874. Gargallo-Viola D, Esteve M, Llovera S, Roca X, Guinea J: In vitro and in vivo antibacterial activities of E-4497, a new 3-amine-3-methyl-azetidinyl tricyclic fluoroquinolone. Antimicrob Agents Chemother 1991;35:442–447. Wise R, Ashby JP, Andrews JM: In vitro activity of PD 127,391, an enhanced-spectrum quinolone. Antimicrob Agents Chemother 1988;32:1251–1256.

Antibacterial Activity in Urine

81

BOIS:L01:BIF01015XA.77

20

21 22

23 24 25 26

27 28

29

30

31 32 33 34 35 36

37 38

39 40 41

FF: W1FP

E1: IN01

$$$

Grasso S, Meinardi G, de Carneri I, Tamassia V: New in vitro model to study the effect of antibiotic concentration and rate of elimination on antibacterial activity. Antimicrob Agents Chemother 1978;13:570–576. O’Grady F, Pennington JH: Bacterial growth in an in vitro system simulating conditions in the urinary bladder. Br J Exp Pathol 1966;47:152–157. O’Grady F, Mackintosh IP, Greenwood D, Watson BW: Treatment of ‘bacterial cystitis’ in fully automatic mechanical models simulating conditions of bacterial growth in the urinary bladder. Br J Exp Pathol 1973;54:283–290. Musher DM, Griffiths DP: Generation of formaldehyde from methenamine: Effect of pH and concentration, and antibacterial effect. Antimicrob Agents Chemother 1974;6:708–711. Greenwood D, O’Grady F: An in vitro model of the urinary bladder. J Antimicrob Chemother 1978;4:113–120. Anderson JD, Eftekhar F, Aird MY, Hammond J: Role of bacterial growth rates in the epidemiology and pathogenesis of urinary infections in women. J Clin Microbiol 1979;10:766–771. Okada K, Ohkoshi M: In vitro activity of antibacterial agents in mixed infections in a urinary tract model; in Nelson JD, Grassi C (eds): Current Chemotherapy and Infectious Disease. Washington, American Society for Microbiology, 1980, pp 1290–1291. Greenwood D: An in vitro model simulating the hydrokinetic aspects of the treatment of bacterial cystitis. J Antimicrob Chemother 1985;15(suppl A):103–109. Nishimura M, Kumamoto Y, Hirose T, Ohya S: An experimental examination of the significance of the peak antimicrobial concentration value in urine and time above MIC on the effect of antimicrobial agents. An examination using a severely complicated in vitro bladder model with an autosimulation system for antimicrobial concentration in urine. Kansenshogaku-Zasshi 1994;68:353–365. Sano M, Kumamoto Y, Nishimura M, Tsukamoto T, Hirose T, Ohya S: Inhibition of biofilm formation by clarithromycin in an experimental model of complicated bladder infection. In vitro study using automated simulation of urinary antimicrobial concentration. Kansenshogaku-Zasshi 1994;68:894–904. Casellas JM, Tome G, Exeni R, Grimoldi I, Goldberg M, Farinati AE: Serum and urinary cefpodoxime levels and time-killing curves performed in the urine of children presenting urinary tract infections. Pathol Biol (Paris) 1993;41:385–391. Dubini F, Riviera L: Antibacterial activity in human urine of fosfomycin trometamol in an in vitro model of the urinary bladder. Chemioterapia 1988;7:15–19. Schlichter JG, MacLean H: A method of determining the effective therapeutic level in the treatment of subacute bacterial endocarditis with penicillin. Am Heart J 1947:34:209–211. Schlichter JG, MacLean H, Malzer A: Effective penicillin therapy in subacute bacterial endocarditis and other chronic infections. Am J Med Sci 1949;217:600–608. Klastersky J, Daneau D, Swings G, Weerts D: Antibacterial activity in serum and urine as a therapeutic guide in bacterial infections. J Infect Dis 1974;129:187–193. Guibert J, Kitzis MD, Brumpt I, Acar JF: Activité antibactérienne de la péfloxacine dans l’urine durant sept jours après prise oralle unique de 800 mg. Pathol Biol (Paris) 1989;37:406–410. Guibert J, Kitzis MD, Acar JF, Masquelier AM, Bellanger B: Activité antibactérienne de la loméfloxacine dans l’urine durant quatre jours après prise oralle unique de 400 mg. Pathol Biol (Paris) 1989;37:411–414. Guibert J, Kitzis MD, Acar JF: Activité antibactérienne de la ciprofloxacine dans l’urine durant quatre jours après prise oralle unique de 500 mg et 750 mg. Pathol Biol (Paris) 1994;42:587–592. Cruciani M, Monzillo V, Navarra A, Tinelli C, Concia E: Antibacterial activity after a single dose of norfloxacin, ofloxacin and pipemidic acid detected in urine of volunteers. Drugs Exp Clin Res 1988;14:533–537. Soussy CJ, Deforges L, Duval L: Activité antibactérienne de l’énoxacine in vitro et dans les urines. Pathol Biol (Paris) 1987;35:475–481. Bergan T: Degree of absorption, pharmacokinetics of fosfomycin trometamol and duration of urinary antibacterial activity. Infection 1990;18(suppl 2):65–69. Wiedemann B, Groos M: Antibacterial activity of fosfomycin trometamol in the urine after simulation of oral doses in a pharmacokinetic in vitro model. Eur Urol 1987;13(suppl 1):76–79.

Naber

82

BOIS:L01:BIF01015XA.77

42

43

44 45 46 47 48

FF: W1FP

E1: IN01

$$$

Hofbauer H, Kinzig M, Kresken M, Naber KG, Reiz A, Sörgel F, Rustige-Wiedemann C, Wiedemann B: Urine bactericidal titers of pefloxacin versus norfloxacin in volunteers receiving a single oral dose of 800 mg. 34th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), Orlando, Fla, Oct 1994, poster A61. Naber KG, Sörgel F, Kinzig M, Savov O, Boudnitzki G, Birner B, Hollauer K, Kirchbauer D: Urinary excretion and bactericidal titers following a single oral dose of 400 mg fleroxacin (F) versus 800 mg pefloxacin (P) in healthy volunteers. 19th International Congress of Chemotherapy (ICC), Montréal, July 1995, abstr 510. Acar JF: Dynamique de la bactériurie dans les infections urinaires a bacilles Gram-négatif traitées par les antibiotiques. Pathol Biol (Paris) 1969;17:859–864. Helm EB, Shah PM, Stille W: Kinetics of bacterial elimination in urine during antimicrobial treatment. J Antimicrob Chemother 1979;5(suppl B):191–191. Naber KG, Ahrens T: Cefroxadin und Cephalexin bei Harnwegsinfektionen. Kinetik der Keimelimination. Fortschr Med 1982;100:1827–1831. Naber KG, Timmler P: Keimelimination durch Fosfomycin bei komplizierten Harnwegsinfektionen. Therapiewoche 1983;33:3300–3306. Anderson JD: Relevance of urinary bladder models to clinical problems and to antibiotic evaluation. J Antimicrob Chemother 1985;15(suppl A):111–115.

Prof. Kurt G. Naber, Urologic Clinic, Elisabeth Hospital, Teaching Hospital of the Technical University of Munich, Schulgasse 20, D–94315 Straubing (Germany)

Antibacterial Activity in Urine

83

BOIS:L01:BIF01016XA.86

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 84–88

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Response of Escherichia coli to Fleroxacin in an in vitro Model of the Urinary Bladder Yukimichi Kawada, Yusuke Kanimoto Department of Urology, Gifu University School of Medicine, Tsukasa, Gifu, Japan

In most in vitro tests, antibiotic concentrations remain constant, whereas in vivo drug concentrations rise and fall depending on the drug and mode of administration. Conventional in vitro susceptibility tests expose the bacteria to different conditions than in urine during treatment of bacterial cystitis. In an attempt to minimize the differences between in vitro and in vivo conditions, O’Grady et al. [1, 2] have introduced an in vitro urinary bladder model in which the conditions of exposure of bacteria to antibacterial drugs in cystitis are simulated. We have modified a similar, in vitro model of the urinary bladder. In order to clarify whether once-daily treatment with fleroxacin, which has a long serum elimination half-life of approximately 10 h and urinary recovery rate of 60% within 24 h [3], is effective in the treatment of urinary tract infections (UTIs), the response of Escherichia coli to fleroxacin and ofloxacin was studied in an in vitro bladder model.

Materials and Methods Bacterial strain: E. coli ECSA-1, originally isolated from a urinary infection, was used. The MICs of fleroxacin and ofloxacin were determined by broth dilution to be 6.25 mg/l. Bladder model: The design and use of the bladder model have been described elsewhere [1, 2]. In the present study, overnight broth cultures of bacteria in 20 ml initially were diluted at a rate of 1 ml/min (simulating the ureteric urine flow rate), the ‘bladder’ voided at hourly intervals like micturition, levaing a residual ‘bladder’ volume of 20 ml. After four such cycles of dilution and micturition, fleroxacin or ofloxacin was introduced into the sys-

BOIS:L01:BIF01016XA.86

FF: W1FP

E1: IN01

$$$

Fig. 1. Suppression of E. coli by fleroxacin in the bladder model. E. coli was exposed to fleroxacin with concentration simulated to that achieved in urine after a single oral administration of 100 mg (single cycle exposure).

tem. Bacteria were exposed to each drug in one of two types of concentration profiles: either (1) those simulating urinary drug concentrations achieved after a single oral dose of 100 mg (single cycle exposure), or (2) simulating urinary drug concentrations achieved after a twice repeated oral administration of 50 mg every 12 h (double cycle exposure). Clinical study: A total of 104 patients with complicated UTIs were treated with fleroxacin either 100 mg b.i.d. or 200 mg once daily, for 7 days. All patients had pyuria of at least 5 white blood cells per high-power field, bacteriuria of at least 104 CFU/ml and no indwelling catheter. Clinical efficacy was assessed according to the criteria proposed by the Japanese UTI Committee as ‘excellent’, ‘moderate’ or ‘poor’ based on a combination of changes in pyuria and bacteriuria [4].

Results In the in vitro bladder model study, growth of E. coli was suppressed for 32 h when it was exposed to fleroxacin at concentrations simulating those achieved in urine after a single oral administration of 100 mg (fig. 1). When E. coli was exposed to fleroxacin at concentrations simulating those achieved in urine after 50 mg b.i.d. fleroxacin, growth of E. coli was suppressed for 26 h (fig. 2). The

Response of E. coli to Fleroxacin

85

BOIS:L01:BIF01016XA.86

FF: W1FP

E1: IN01

$$$

Fig. 2. Suppression of E. coli by fleroxacin in the bladder model. E. coli was exposed to fleroxacin with concentration simulated to that achieved in urine after oral administration of 50 mg b.i.d. (double cycle exposure).

latter was significantly shorter than that obtained after a single cycle exposure. Conversely, when E. coli was exposed to ofloxacin, which has a serum elimination half-life of approximately 5 h, a longer lasting duration of growth inhibition was observed with a double cycle exposure (table 1). Based on these results, clinical study on once daily treatment with fleroxacin was attempted, and a higher clinical efficacy was obtained with once daily treatment with fleroxacin (table 2).

Discussion Fleroxacin has a long serum elimination half-life of approximately 10 h and urinary recovery of 60% within 24 h after a single oral dose. These pharmacokinetic properties of fleroxacin prompted a clinical study on the efficacy and safety of fleroxacin once daily for complicated UTIs. However, because we have had no experiences with once daily treatment with fluoroquinolones for complicated UTIs, we investigated the response of E. coli to fleroxacin in an in vitro model of the urinary bladder before the clinical study.

Kawada/Kanimoto

86

BOIS:L01:BIF01016XA.86

FF: W1FP

E1: IN01

$$$

Table 1. Suppression of growth of E. coli by fleroxacin and ofloxacin in the bladder model Drug

Duration of growth inhibition, h single exposure double exposure

Fleroxacin Ofloxacin

32 25

26 32

Table 2. Clinical efficacy of fleroxacin for complicated UTI (7 days’ treatment) Dosage

100 mg ! 2 200 mg ! 1

Patients

36 68

Clinical efficacy

Excellent + moderate

excellent

moderate poor

n

%

12 27

11 25

23 52

63.9 76.5

13 16

The clinical experience showed that the results of the in vitro bladder model was useful in predicting the clinical efficacy of once daily treatment with fleroxacin. In another in vitro bladder model study, we have demonstrated that the concomitant presence of ß-lactamase producing Staphylococcus epidermidis substantially reduced the efficacy of ampicillin against ampicillin-susceptible E. coli [5]. These results explained why fully susceptible bacteria which would normally be eradicated when present alone, frequently persist when present in mixed infections. Furthermore, Nishimura et al. [6] with their kidney-bladder model have demonstrated that the combination of ciprofloxacin and clarithromycin was effective against biofilm-forming Pseudomonas aeruginosa because clarithromycin inhibited biofilm formation. They currently attempt to determine the clinical susceptibility break-point of MIC in UTI by the in vitro model. Consequently, an in vitro urinary bladder model, which simulates the dynamic conditions of bacterial exposure to antibacterial drugs during the treatment of cystitis, might render useful information in predicting the clinical efficacy of antibacterial drugs on UTIs.

Response of E. coli to Fleroxacin

87

BOIS:L01:BIF01016XA.86

FF: W1FP

E1: IN01

$$$

References 1 2 3

4

5

6

Greenwood D, O’Grady F: An in vitro model of the urinary bladder. J Antimicrob Chemother 1978;4:113–120. Greenwood D, O’Grady F: Is your dosage really necessary? Antibiotic dosage in urinary infection. Br Med J 1977;i:665–667. Nakashima M, Kanamaru M, Uematsu T, Takiguchi A, Mizuno A, Itaya T, Kawahara F, Ooe T, Sato S, Uchida H, Masuzawa K: Clinical pharmacokinetics and tolerance of fleroxacin in healthy male volunteers. J Antimicrob Chemother 1988;22:133–144. Kawada Y: An outline of Japanese criteria for evaluating the clinical efficacy of antimicrobial agents in complicated urinary tract infections; in Ohkoshi M, Kawada Y (eds): Clinical Evaluation of Drug Efficacy in UTI. Amsterdam, Excerpta Medica, 1990, pp 37–48. Kawada Y, Greenwood, O’Grady F: Response of Escherichia coli to beta-lactam antibiotics: Effect of concomitant presence of staphylococci exhibiting inducible beta-lactamase activity. Infection 1980;8:81–85. Nishimura M, Kumamoto Y, Sano M, Hirose T, Ohya S: Therapeutic study on biofilm of the urinary tract using severely complicated bladder model. J Jpn A Infect Dis 1994;1:386–398.

Dr. Yukimichi Kawada, Department of Urology, Gifu University School of Medicine, 40 Tsukasa-machi, Gifu-shi 500 (Japan)

Kawada/Kanimoto

88

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 89–97

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

The Role of the Animal Model in the Study of Prostatitis J. Curtis Nickel Department of Urology, Queen’s University, Kingston, Ont., Canada

The management of the various prostatitis syndromes has become less frustrating for both the clinician and patient than previously. Research over the last several decades has led to insights into the etiology and pathogenesis of prostatic inflammation, both bacterial and nonbacterial, improved diagnostic methods and better therapeutic strategies. These advances which are now only changing the clinical diagnosis and treatment of prostatitis were led by research in animal models. Prostatitis is not alone in employing animal modelling to develop an understanding of the etiology and pathogenesis of prostate diseases. In fact, many of the major advances in prostate research were made possible by animal modelling, i.e. etiology of benign prostatic hyperplasia, androgen and antiandrogens, oncogenes, the dihydrotestosterone and 5·-reductase story, etc. While it is always difficult to extrapolate from a somewhat artificial animal model system to human diseases, the importance of in vivo model research in prostatitis cannot be underestimated. Animal models have allowed us to study the pathogenesis of acute and chronic prostatic inflammation in a longitudinal fashion that is impossible in human disease. Newer diagnostic systems that will eventually be the future of prostatitis diagnosis will owe their origins to research in animal models. Therapeutic strategies such as appropriate antimicrobial therapy were based on antibiotic pharmacokinetics that were explored in animal models. This paper will examine a number of animal models of experimental prostatic inflammation that have led to insights into this complex inflammatory prostatic condition.

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

Etiology and Pathogenesis Experimental prostatic inflammation in animal models has given us insight into the pathogenesis of the various prostatitis syndromes. Combining the results of animal models, anatomical and clinical studies, we have developed insight into the actual etiopathogenesis of acute and chronic as well as bacterial inflammatory conditions of the prostate gland. It appears that prostatic inflammation results when two factors occur together. The first requirement is a relative high pressure dysfunctional voiding caused by anatomical or physiologic obstruction such as constriction of the bladder neck, external sphincter, urethra or urethral meatus. If this is combined with some form of intraprostatic ductal reflux, urine will reflux into the prostate gland. If the urine is sterile an immunologically mediated prostatic inflammation occurs. If pathogenic bacteria are present in the prostatic or bulbar urethra or even more distally, back eddies will draw these bacteria in a retrograde fashion, eventually into the prostate gland. If the animal (or patient) has not experienced a previous bacterial infection such as cystitis, urethritis or prostatitis, an acute bacterial prostatitis results. Recurrent episodes result in a much less severe and chronic bacterial inflammation. Experimental prostatitis by many researchers and in many animal models have allowed us to draw this hypothesis that can be further substantiated in clinical studies. Bacterial prostatitis (both acute and chronic) are caused by bacteria that are similar in type and incidence to those that cause simple urinary tract infection with aerobic Gram-negative enteric bacteria (Escherichia coli) predominating. The role of Gram-positive bacteria such as enterococci, coagulase-negative staphylococci and organisms such as Chlamydia trachomatis and Ureaplasma ureolyticum remain obscure. Experimental prostatitis with E. coli is well established in animal systems [1–6] and such models may be employed to test the hypothesis that other organisms, such as C. trachomatis [7], might be implicated in chronic prostatic inflammation. The route of bacterial inoculation in acute and chronic bacterial prostatitis is becoming clear. While Maglione et al. [8] noted that surgically induced E. coli seminal vesiculitis in rats was occasionally complicated by E. coli prostatitis, evidence is accumulating that the pathophysiology of bacterial prostatitis is associated with retrograde bacterial ascent from the urethra, likely secondary to some form of intraprostatic reflux. Jantos et al. [2] inoculated the bladder of male and female Mastomys natalensis with E. coli and produced severe prostatitis. The histologic and microbiologic course of the prostatic infection strongly resembled the human disease. Chronic bacterial and nonbacterial prostatitis developed by this route persisted for 6 months postinfection. Dilworth et al. [9] used a monkey model to study the ascending route infection in prostatitis and noted that the P pili were the principle mediators of adherence to urethral cells of the prostatic

Nickel

90

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

urethra. Neal et al. [3] employed this nonhuman primate model and found that after urethral inoculation of a wild-type clinical isolate of E. coli, the monkeys developed prostatitis similar to that reported in humans and concluded that infection in the nonhuman primate occurs by this ascending route. Again, histologic changes were similar to that in human disease. Goto et al. [5] induced experimental acute bacterial prostatitis in rats by four different routes of bacterial inoculation. The most simple and reproducible method of producing bacterial prostatitis was to instill the bacterial suspension in the prostatic urethra after the administration of appropriate antibiotics to prevent associated pyelonephritis. A reliable and consistent rat model of acute and chronic bacterial prostatitis has been developed by our research group [4, 6] and employed to undertake sophisticated microbial, histologic and immunologic studies that would not be possible in a clinical patient study to further elucidate the pathogenesis of this complicated disease. Research undertaken with this animal model has demonstrated that bacteria entering the ducts and ascini of the prostate gland multiply rapidly, inducing a host response with infiltration of acute inflammatory cells into the ducts. We have confirmed the observation of Dilworth et al. [9] that mannoseresistant (X and P) pili appear to be important in the pathogenesis of prostate infection and further noted that phase variation, to enhance virulence, can occur in the prostate (presented at Canadian Society for Microbiology Annual Meeting, Kingston, Canada, June 1995). In acute bacterial prostatitis the entire prostate gland or at least the major part of it is involved in the inflammatory process. The ducts become engorged with infiltrate of dead and live bacteria as well as living and dying acute inflammatory cells, desquamated epithelial cells and cellular debris. In acute bacterial prostatitis the rats become ill and approximately one third die of fulminating urosepsis. However, prior to the point of urosepsis it is relatively easy to eradicate all offending organisms with appropriate antibiotic therapy for complete resolution of the inflammatory process. We have demonstrated [10] that rats vaccinated with E. coli and then similarly inoculated did not go through the acute prostatitis stage. These rats developed a subacute and then chronic bacterial prostatitis. The rats did not become ill and it becomes much more difficult to eradicate the bacteria with antibiotic therapy. It appears the bacteria form small, sporadic bacterial microcolonies within the ductal system adherent to the epithelium. The bacteria produce an exopolysaccharide slime or ‘glycocalyx’ that envelops these microcolonies. The microorganisms become very quiescent, undergoing a sort of ‘hibernization’ when the environment becomes threatening. Surrounding these focal sites of bacterial persistence are areas of lymphatic invasion with variable infiltration of plasma cells and macrophages. Over time, fibrosis occurs with subsequent permanent scarring. The pathogenesis of nonbacterial prostatitis may be similar in that prostatic reflux of urine, urine products or even undetected organisms in prostatic ducts

Animal Models in Prostatitis

91

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

and ascini may occur. Keetch et al. [11] developed an animal model of nonbacterial prostatitis and characterized the immune parameters of this type of inflammation. Adoptive transfer studies demonstrated the prostatic inflammation to be at least in part immune mediated, and these researchers concluded that nonbacterial prostatitis may be an autoimmune process. This would explain the close association between nonspecific chronic prostatic inflammation and bacterial induced chronic inflammation. Even with complete resolution of the bacterial infection, the prostate may still demonstrate areas of chronic nonspecific prostatic inflammation. This would explain why patients with supposedly cured bacterial prostatitis continue to have low-grade symptoms. It also explains why chronic bacterial prostatitis recurs. Although eradication of bacteria is possible, the anatomical and voiding parameters that result in prostatitis in the first place remain and the process can recur. If the bacteria are not eradicated, the symptoms of chronic bacterial prostatitis relapse. Animal models have resulted in a further understanding of the pathogenic mechanisms involved in prostatic inflammation. Naslund et al. [12] suggested that genetic background, advancing age, and hormonal imbalance are important etiological factors for nonbacterial prostatitis in rats. Nonbacterial prostatitis is more common in Lewis rats than in Wistar rats and does not occur in SpragueDawley rats. The incidence of spontaneous prostatitis was significantly higher in older animals than in younger animals. The administration of exogenous 17ßestradial increased the incidence and severity of prostatitis in old Wistar rats. Castration had a similar effect. Similar prostatitis could even be induced in young adult Wistar rats by neonatal treatment with 17ß-estradial followed 7 months later in adulthood by testosterone administration. These and other studies of experimental prostatitis [2, 4, 13] also showed that nonbacterial prostatitis resulting from eradication of the bacterial agent in chronic bacterial prostatitis and chronic bacterial prostatitis with persistence of the bacterial agent are indistinguishable. This appears also to occur in human studies. Prostatic inflammation, which for the most part is asymptomatic, seen in association with benign prostatic hyperplasia in patients who have had prostatectomy, is really indistinguishable from the prostatitis seen in chronic bacterial and nonbacterial prostatitis. The symptoms of nonbacterial prostatitis in human patients appears to increase in severity with exogenous stress, anxiety and various diets [14]. Gatenbeck et al. [15] examined rat prostate glands after a 10-day period during which the rats had been submitted to standardized stimuli. They demonstrated similar inflammatory histopathologic changes in the glands of all rats submitted to longterm stress stimuli that were similar to those of human males with nonbacterial prostatitis. Aronsson et al. [13] confirmed this finding by subjecting rats to standardized experimental stress stimuli and again showed prostatic inflammatory changes compatible with nonbacterial prostatitis on histological examination.

Nickel

92

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

Diagnosis Meares and Stamey [16] standardized the technique to differentiate the various prostatitis syndromes. This technique is based on a rigid, quantitative segmented bacteriologic localization procedure that remains the gold standard for diagnosis today. However, numerous shortcomings with this technique have arisen and for the most part physicians have abandoned it in clinical practice. One of the main reasons is the perceived false-negative rate of this test. It has been demonstrated in a number of clinical studies that quantitative cultures of the expressed prostatic secretion or the urine after prostatic massage is sterile despite the bacterial presence in the prostate gland or a similar clinical response to antibiotics as that obtained when the culture was positive [14, 17]. Animal models explain some of this difficulty. Ling et al. [18] found similar difficulties in trying to correlate cultures of ejaculate, urine, urethral swab specimens and biopsy cultures of the prostate from dogs with suspected prostatitis. Our research group has demonstrated in experimental prostatitis that bacteria persist in the prostate gland of rats that have been suboptimally treated with antibiotics, even when they cannot be demonstrated in the prostatic secretion [19; Nickel, unpubl. data]. We have recently confirmed this generally held assumption that our accepted clinical diagnostic methods are not always adequate [20]. Patients with proven chronic bacterial prostatitis who became clinically resistant to antibiotic therapy were discovered to still have a pure culture of the initial bacterial agent in the prostatic cultures even when the prostatic secretion was sterile. The study of the immune response in both clinical and experimental prostatitis studies has led to a greater understanding of the pathogenesis of this disease and it appears that it may hold the key for more precise diagnosis of the etiologic agent. In experimental chronic bacterial prostatitis we have confirmed [4, 6] findings of increased levels of antigen-specific antibody in the prostatic secretion of animals with unresolved prostatitis as well as elevated serum antibody titers against their prostatic pathogens which return to normal with successful treatment of chronic bacterial prostatitis. Preliminary studies by Shortliffe and Wehner [21] have attempted to employ the defined immunological reaction seen in the prostate to identify the most common bacterial antigens; however, this work has never gone beyond the initial stages, primarily because it could not be validated in an adequate animal or clinical model. Our research group is exploiting our reliable and consistent animal model to determine whether simple immunologic diagnostic tests can differentiate nonbacterial from bacterial prostatitis, even after antibiotic therapy [Nickel, unpubl. data]. The results from these early studies have now been extrapolated to the clinical situation and a multicenter Canadian trial is now assessing the validity and reliability of using an immunologic test to predict the response of prostatic inflammation to antibiotic therapy.

Animal Models in Prostatitis

93

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

Treatment Treatment of acute bacterial prostatitis is consistently effective as predicted by our animal models [19, 22]. However, antibiotic therapy for chronic bacterial prostatitis remains dismal. These poor results are obtained even though the organisms which we occasionally culture are highly sensitive to the particular antimicrobial agents used and even remain so at the end of treatment [20]. Most human pharmacokinetic studies analyzed prostatic fluid obtained by prostatic massage or ejaculate for drug content following oral or parenteral administration of various antibiotics; however, these methods have since been disputed because of the contamination with seminal fluid or urine containing high concentrations of antibiotics. Homogenized human prostatectomy specimens have also been analyzed for drug content following preoperative antibiotic loading; however, it is apparent that this method does not measure the concentration in prostatic secretions and certainly is not an adequate measure of prostatic tissue levels in an inflamed prostate gland. Dog models were originally used to investigate prostatic secretion of various antibiotics. The vas deferens was divided and urine diverted by ureteral cannulation or suprapubic tube. Intravenous pilocarpine was used to stimulate copious secretion of pure prostatic fluid which could be collected uncontaminated by urine at the urethral meatus. The model allowed simultaneous evaluation of plasma, urine and prostatic secretion drug levels. Researchers employed such an animal model to systematically investigate the diffusion of antibiotics from the plasma into the prostatic secretion of normal dogs [23]. These and later studies led to the hypothesis that the distribution of a drug in the prostatic interstitium and prostatic secretion depends on absorption, plasma protein binding, lipid solubility, intercompartmental pH gradients, the individual antibiotic, the pKa of the antibiotic and biotransformation. Multiple modifications of this dog model were made by others [1] but unfortunately no real consistent methodology developed in the field. Sharer and Fair [24] have reviewed the various canine models used to quantitate antimicrobial drug diffusion, and from these particular types of studies a number of antibiotics (i.e. trimethoprim, erythromycin, quinolones) have been described as the most suitable drugs for the treatment of prostatitis. Animal models have consistently demonstrated that appropriate antibiotic therapy results in antibiotic concentration in the prostatic ducts which appear to be several times the minimal inhibitory concentration required for the eradication of offending bacteria. Why then does antibiotic therapy not cure prostatitis more consistently? Evidence from clinical studies and in animal models suggest incomplete sterilization of prostatic fluid, probably secondary to an inadequate concentration of antimicrobial agents. As noted before, with complete eradication of the offending bacterial organisms, the anatomical and physiologic factors

Nickel

94

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

remain that promote a recurrence of infection, perhaps with a different organism. It was believed [25] that the most important cause for therapeutic failure in chronic bacterial prostatitis was an inadequate concentration of antimicrobial agents. This was felt to be due to the simple fact that the prostatic intraductal compartment is entirely different in the inflamed human prostate compared to the uninfected dog. Also, prostatic inflammation may be a focal phenomenon and pharmacokinetics based on the entire gland may not be relevant to the sporadically infected areas. Ducts blocked with inflammatory debris, microabscesses, calculi and focal pH changes could potentially change the pharmacokinetics in the local microenvironment of the focal inflammatory lesion. To attempt to answer some of the questions, Baumueller and Madson [1] injected E. coli into the prostatic arterial system 1 week prior to antibiotic treatment to create a realistic treatment environment with an animal model with a bacterial prostatitis. Inflammatory changes were however mainly in the interstitial tissue and not in the acini presumably because of the iatrogenic hematogenous route of infection. Our group has recently reported on pharmacokinetic studies in our animal model of chronic prostatitis that develops after retrograde injection of pathogens and more closely resembles the focal and intraductal nature of human disease [19]. These studies, however, show that antibiotic concentrations in the inflamed prostatic duct are actually higher (but not significantly higher) than those obtained in the noninflamed prostate gland. We and others have shown in in vitro [26] and animal studies [19] that bacterial aggregates or microcolonies adherent to the ductal epithelium and covered with glycocalyx matrix are relatively resistant to both host defenses and normal concentrations of antibiotics. Animal studies in our rat model [4, 6] and further studies in human chronic bacterial prostatitis [20] consistently demonstrate these glycocalyx-enclosed bacterial biofilms, that occur in both treated and untreated conditions. Urinary and bacterial induced calculi in the genitourinary tract and prostate in animal models [27] also explains chronic bacterial persistence in some cases. Animal studies are presently helping us to determine whether novel forms of drug dosage, dosage intervals and drug delivery will improve bacterial eradication. We are presently treating our experimentally induced chronic prostatitis in the rat with high-dose pulse antibiotic therapy as well as high-pressure intravesical instillation of antibiotics and percutaneous antibiotic administration. Extrapolation of this research may guide us to improved treatment modalities for clinical prostatitis secondary to bacterial presence within the prostate gland.

Animal Models in Prostatitis

95

PIAM:L01:BIF01017XA.90

FF: W1FP

E1: IN01

$$$

The Importance of the Animal Model in Prostatitis This review should lead to a further understanding of the importance of two and one half decades of animal modelling research in the search for answers on the etiology, pathogenesis, diagnosis and treatment of prostatitis syndromes. Animal modelling of human disease, particularly human infectious diseases, remains open to skepticism, criticism and even misinterpretation. However, clinical studies have not answered important pathogenic and treatment questions despite decades of human research. Experimental prostatitis in appropriate animal models is providing the key that is unlocking some of the mysteries of prostatitis.

References 1 2 3 4 5 6 7 8 9 10 11 12 13

14 15 16 17

Baumueller A, Madsen PO: Experimental bacterial prostatitis in dogs. Urol Res 1977;5:211–213. Jantos C, Altmannsberger M, Weidner W, Shiefer HG: Acute and chronic bacterial prostatitis due to E. coli: Description of an animal model. Urol Res 1990;18:207–211. Neal DE, Dilworth P, Kaack MB, Didier P, Roberts JA: Experimental prostatitis in nonhuman primates. II. Ascending acute prostatitis. Prostate 1990;17:233–239. Nickel JC, Olson ME, Barabas A, Benediktsson H, Dasgupta MK, Costerton JW: Pathogenesis of chronic bacterial prostatitis in an animal model. Br J Urol 1990;66:47–54. Goto T, Kawahara M, Kawahara K, Mahinose S, Mizuma Y, Sakamoto N, Ohi Y: Experimental bacterial prostatitis in rats. Urol Res 1991;19:141–144. Nickel JC, Olson ME, Costerton JW: Rat model of experimental bacterial prostatitis. Infection 1991:19(suppl 3):126–130. Nielsen DS, Golubjatnikov R, Dodge R, Madsen PO: Chlamydial prostatitis in dogs: An experimental study. Urol Res 1982;10:45–49. Maglione W, Nardi A, Cranz C, Clavert A, Bollack C: Acute vesiculitis and its prostatic complications caused by E. coli. Urol Res 1986;14:265–266. Dilworth JP, Neal DE, Fussell EN, Roberts JA: Experimental prostatitis in nonhuman primates. 1. Bacterial adherence in the urethra. Prostate 1980;17:227–231. Nickel JC, Ceri H, Olson ME, Smidt S: The role of specific mucosal immunity in the pathophysiology of bacterial prostatitis. J Urol 1995;153:328A. Keetch DW, Humphrey P, Ratliff TL: Development of a mouse model for nonbacterial prostatitis. J Urol 1994;152:247–250. Naslund MJ, Strandberg JD, Coffey DS: The role of androgens and estrogens in the pathogenesis of experimental nonbacterial prostatitis. J Urol 1988;140:1049–1053. Aronsson A, Dahlgren S, Gatenbeck L, Stromberg L: Predictive sites of inflammatory manifestation in the prostate gland: An experimental study on nonbacterial prostatitis in the rat. Prostate 1988;13: 17–24. Nickel JC, Bruce AW, Reid G: The prostatitis syndromes; in Krane RJ, Siroky MB, Fitzpatrick JM (eds): Clinical Urology. Philadelphia, Lippincott, 1994, pp 925–938. Gatenbeck L, Aronsson A, Dahlgran S, Johansson B, Stromberg L: Stress stimuli-induced histopathological changes in the prostate: An experimental study in the rat. Prostate 1987;11:69–76. Meares EM Jr, Stamey TA: Bacteriologic localization patterns in bacterial prostatitis and urethritis. Invest Urol 1968;5:492–518. Nickel JC: New concepts in the pathogenesis and treatment of prostatitis. Curr Opin Urol 1992;2: 37–43.

Nickel

96

PIAM:L01:BIF01017XA.90

18

19 20 21 22 23 24 25 26 27

FF: W1FP

E1: IN01

$$$

Ling GV, Nyland TG, Kennedy PC, Hager DA, Johnson DL: Comparison of two sample collection methods for quantitative bacteriologic culture of canine prostatic fluid. J Am Vet Med Assoc 1990; 196:1479–1482. Nickel JC, Downey J, Clark J, Olson ME: Antibiotic pharmacokinetics in the inflamed prostate. J Urol 1995;153:527–529. Nickel JC, Costerton JW: Bacterial localization in antibiotic-refractory chronic bacterial prostatitis. Prostate 1993;23:107–114. Shortliffe LMD, Wehner N: The characterization of bacterial and non-bacterial prostatitis by prostatic immunoglobulins. Medicine 1986;65:399–414. Nickel JC, Olson ME, Ceri H: Experimental prostatitis; in Weidner W, Madsen PO, Schiefer HG (eds): Prostatitis: Etiopathology, Diagnosis and Treatment. Berlin, Springer, 1994, pp 123–130. Winningham DG, Nemoy NJ, Stamey TA: Diffusion of antibiotics from plasma into prostatic fluid. Nature 1968;219:139. Sharer WCV, Fair WR: The pharmacokinetics of antibiotic diffusion in chronic bacterial prostatitis. Prostate 1982;3:139–148. Fair WR, Crane DB, Schiller N, Heston WDW: Reappraisal of treatment in chronic bacterial prostatitis. J Urol 1979;121:437–441. Nickel JC, Costerton JW, McLean RJC: Bacterial biofilms: Influence on the pathogenesis, diagnosis and treatment of urinary tract infections. J Antimicrob Chemother 1994;33(suppl):31–41. Nickel JC, Olson M, McLean RJC, Grant SK, Costerton JW: An ecological study of infected urinary stone genesis in an animal model. Br J Urol 1987;59:21–31.

Dr. J. Curtis Nickel, Department of Urology, Queen’s University, Kingston General Hospital, Kingston, Ont. K7L 2V7 (Canada)

Animal Models in Prostatitis

97

BOIS:L01:BIF01018XA.88

FF: W1FP

E1: IN01

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

$$$

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 98–105

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Tropism in Bacterial Infections: Urinary Tract Infections James A. Roberts 1 Department of Urology, Tulane Regional Primate Research Center, Covington, La., USA

Tropism is the phenomenon by which both commensal and pathogenic bacteria are restricted to certain hosts, tissues and cell types [1]. The word is derived from the Greek ‘trope’ meaning to lean towards sustenance or food. It has been used for some time in botany and would be more familiar if we thought of phototropism, which we all understand to be the phenomenon by which plants lean toward the light. Tropism helps to explain the diversity of bacteria known to cause certain diseases. Examples include the involvement of group A streptococci with pharyngitis and Streptococcus mutans with dental caries [2], and the fact that only certain organisms are associated with meningitis, and these at only certain times of life. Thus, while Escherichia coli has been associated with meningitis in the newborn, it is not doing so at later life. Experimental studies have shown that endothelial receptors for E. coli in the brain are only present in the newborn [3]. Other examples of diseases of the lung, stomach, and colon also show the diversity. In the case of Helicobacter pylori, associated with both ulcers and perhaps gastric cancer, only certain cells in the stomach allow adherence of the bacteria [4]. In the case of urinary tract infection, it appears that the initiating event is adhesion to urothelial cells by means of the tip protein of P-fimbriae of E. coli (table 1). Bacterial adhesion is a necessary event for a bacterial infection of either the urinary, respiratory or gastrointestinal tract to occur, with adhesion to mucous membranes being the initial event in any infection, other than those associated with wounds, instrumentation or catheterization.

1

Supported by USPHS grants RR00164 and DK14681.

BOIS:L01:BIF01018XA.88

FF: W1FP

E1: IN01

$$$

Table 1. Examples of tropism Dental caries from S. mutans Skin: impetigo from group A streptococci Brain: meningitis from Haemophilus, Meningococcus, E. coli Pharynx: group A streptococci Lung: Pneumococcus Stomach: H. pylori Colon: Clostridium difficile Kidney: P-fimbriated E. coli (class II adhesin) Bladder: P-fimbriated E. coli (class II adhesin) Urethra: Neisseria gonorrhoeae

The surface energy theory of bacterial adhesion, devised by a group of biochemists, biophysicists and biologists, attempts to explain the mechanism of bacterial adhesion and the necessity of surface appendages for adhesion to occur in an energy-efficient manner [5] (fig. 1). The net negative surface charges of both the tissue cells and the bacteria, as well as diffuse ion clouds in the area, repulse adhesion. While at 15 nm there is very little repulsion, as bacteria approach 10 nm, maximum repulsion occurs. Since the magnitude of both attractive as well as repulsive forces increases with the diameter of the approaching body, bacterial fimbriae or polymers, being of a much smaller diameter, allow bacterial adherence that might not otherwise occur, the fimbriae reaching cell surface receptors for firm adherence to the cell surface. Thus, adherent, the normal flow of body fluids such as mucus or in the case of the urinary tract, urine, does not allow washing away of the bacteria. This leads to bacterial multiplication reaching a critical mass and leading to parenchymal invasion. The adhesins in urinary tract infections from E. coli include type 1 fimbriae which adhere to mucins containing mannose residues, P-fimbriae which adhere to glycolipids on cell surfaces, the afimbrial adhesins and, in the case of some organisms, such as Pseudomonas, polymers. The organisms that cause urinary tract infections can be divided into those of the endogenous flora, which cause infection by the hematogenous route, or by the ascending route, and nosocomial infections. Those causing hematogenous infections are said to do so because of their association with renal abscesses. Staphylococcus aureus kidney infections causing abscess are associated with skin lesions such as furuncles and carbuncles. Streptococcus renal infections occur in association with subacute bacterial endocarditis. E. coli more frequently is the cause of renal abscesses. While it has been assumed that this is due to a hematogenous route of infection, ascending infections can also cause renal abscsses. Ascending

Tropism in Bacterial Infections

99

BOIS:L01:BIF01018XA.88

FF: W1FP

E1: IN01

$$$

Fig. 1. The surface energy theory of bacterial adhesion [modified from 5]. From J Urol 1996;156:1552–1559; with permission.

infection is assumed to be the means by which both the bladder and kidney are infected. This has been well documented in the case of E. coli wherein organisms normally in the gut first colonize either the perineum of females [6] or the prepuce of uncircumcised males [7], prior to their causing an ascending infection. Nosocomial infections are associated with catheterization or instrumentation and both Pseudomonas or Serratia are commonly found in a catheterized urinary tract. How then, do these facts correlate with bacterial adhesion and tropism? Fimbriae of Enterobacteriaceae that cause urinary tract infections include type 1 fimbriae, which are very common in E. coli and Klebsiella, but are more frequently mannose-resistant fimbriae including P-fimbriated E. coli, X-fibriated E. coli, Sfimbriated E. coli [8], some strains of Proteus mirabilis [9] and some strains of Klebsiella [10]. Note that both bacteria such as nonfimbriated Pseudomonas and some species with fimbriae, such as Providencia stuartii, are most commonly associated with complicated infections, catheterization or instrumentation [10, 11]. To focus on the fimbriae of E. coli, type 1 are probably unimportant in urinary tract infections except in the catheterized patient [12], but do allow vaginal colonization. Korhonen’s type 1C, an afimbrial adhesin, has been associated with pyelonephritis, although the mechanism of binding is unknown [13]. P-fimbriae

Roberts

100

BOIS:L01:BIF01018XA.88

FF: W1FP

E1: IN01

$$$

bind to globoside, a glycolipid of urothelial cells and are found in acute pyelonephritis of the nonobstructive type [14–17]. S-fimbriae bind to urothelial cells but this binding appears to be inhibited by urine and thus S-fimbriated E. coli are more commonly associated with newborn sepsis and meningitis as at this time they also bind to endothelial cells in the brain [3]. Other forms of P-fimbriae, more commonly called X-fimbriae, include M and G which are both rare. Other afimbrial adhesins (such as the Dr adhesin) are known and others probably will be found because of the importance of bacterial adhesion in the pathogenesis of urinary tract infections. Several characteristics of P-fimbriae are important in understanding its role in bacterial adherence and the pathogenesis of urinary tract infections. This includes the fact that there are a multiplicity of fimbrial types and even a multiplicity of P-fimbriae with different epitopes of the tip protein [18]. In addition, phase variation of P-fimbriae to the fimbriated state as opposed to the nonfimbriated state does occur under certain conditions [19]. In vivo studies in monkeys of bacterial adherence by P-fimbriae of E. coli show that pyelonephritis occurs because nonhuman primates have the same urothelial receptor for the fimbriae as man [20]. When these bacteria were incubated with the putative receptor Gal-Gal and the bacteria introduced into the ureter of the monkey, the onset of pyelonephritis delayed. In addition, monkeys immunized with P-fimbriae did not develop pyelonephritis after bacterial ureteral challenge [21]. In the laboratories of Normark [22] the specific adhesin has been isolated by genetic studies and found to be the tip protein of P-fimbriae. In his laboratory a nephropathogenic strain with the class II tip protein was used to create a mutant without the tip protein. In our collaboration we have then shown that bacterial challenge with this mutant did not produce pyelonephritis, proving the importance of the class II G-tip protein in pyelonephritis [23]. There are multiple epitopes with different binding sites and thus association with specific infections, such as cystitis as opposed to pyelonephritis. In addition, while the tip protein is the most important adhesin of P-fimbriae for binding to the surface of cells, binding to a secondary binding site on normally unexposed cell membranes does occur by means of the fibrillum PapE which adheres to fibronectin [24]. There are three classes of the G-tip proteins. Class I has not been associated with disease in man, but the class II adhesin is associated with pyelonephritis in both man and monkey, the class III adhesin being associated with cystitis [1]. The difference in these classes has been shown by hemagglutination studies wherein P-fimbriated E. coli with the class I receptor globotriasylceramide agglutinate rabbit cells, the class II adhesin receptor is globoside and these bacteria agglutinate human pig cells, bacteria with the class III G-tip protein whose receptor is the Forssman glycolipid, agglutinate goat, sheep and probably human cells [25].

Tropism in Bacterial Infections

101

FF: W1FP

E1: IN01

$$$

➤

BOIS:L01:BIF01018XA.88

Fig. 2. The surface architecture of the class II and class III receptors for the tip protein of P-fimbriated E. coli. From J Urol 1996;156:1552–1559; with permission.

Table 2. Isoreceptor glycolipids for tip protein of P-fimbriae Class I · Gal 1/4 ß Gal 1/4 Glccer Class II GalNAcß 1/3 · Gal 1/4 ß Gal 1/4 Glccer Class III GalNAc· 1/3 GalNAcß 1/3 · Gal 1/4 ß Gal 1/4 Glccer

All of the isoreceptors for P-fimbriae contain the disaccharide ·Gal 1/4 ßGal but their position in the molecule differs (table 2). We are studying two organisms which contain either the class II tip protein, that being E. coli DS17, the other containing a class III tip protein, that being a mutant, now named DS17-1. The difference in adherence of the two tip proteins to different Gal–Gal disaccharidecontaining glycolipids can be shown by looking at the surface orientation of the saccharides. Note in the graphic illustration that the class II adhesin which is associated with P-fimbriated E. coli causing pyelonephritis can adhere to ·Gal 1/4 ßGal as its architecture places it at the surface even though its terminal sugar is GalNacß.

Roberts

102

BOIS:L01:BIF01018XA.88

FF: W1FP

E1: IN01

$$$

Initial events in acute pyelonephritis

➤ ➤ ➤ ➤ ➤

Initiating acute pyelonephritis

Fig. 3. Factors involved in ascending pyelonephritis from P-fimbriated E. coli. From J Urol 1996;156:1552–1559; with permission.

In the case of the class III adhesin which adheres to the Forssman antigen (GalNac·–GalNacß) its surface oritentation produces steric hindrance with adherence to the ·Gal 1/4 ßGal disaccharide (fig. 2). Thus while this glycoplipid contains the Gal-Gal disaccharide, it is not available for adhesion by the class II tip protein. The initial events in acute pyelonephritis are most often due to P-fimbriae E. coli. By means of the tip protein of P-fimbriae, the bacteria adhere and colonize the perineum or the prepuce, then ascend the urethra (fig. 3). They adhere and colonize the bladder and in the case of the class III tip adhesin, cystitis may occur. The E. coli may ascend the ureter if they contain the class II tip protein, adhere to

Tropism in Bacterial Infections

103

BOIS:L01:BIF01018XA.88

FF: W1FP

E1: IN01

$$$

the ureteral mucosa, and cause aperistalsis. This dilates the ureter and flattens the renal papilla allowing pyelotubular backflow of bacteria at a low pressure, thus allowing adhesion of bacteria to renal tubular cells followed by invasive disease initiating acute pyelonephritis [26]. Tropism should be important also with other bacteria causing urinary tract infections. Further studies of bacteria tropism in infections due to other Enterobacteriaceae which less commonly cause urinary tract infection will help us to further understand the pathogenesis of urinary tract infection.

References 1 2 3 4

5 6 7 8 9 10

11 12

13

14

15 16

Hultgren SJ, Abraham S, Caparon M, Falk P, St Geme JW III, Normark S: Pilus and nonpilus bacterial adhesins: Assembly and function in cell recognition. Cell 1993;73:887–901. Gibbons RJ, van Houte J: Bacterial adherence in oral microbial ecology. Annu Rev Microbiol 1975; 22:19–44. Parkkinen J, Korhonen TK, Pere A, Hacker J, Soinila S: Binding sites in the rat brain for Escherichia coli S-fimbriae associated with neonatal meningitis. Infect Immun 1988;56:2623–2630. Falk P, Roth KA, Boren T, Westblom TU, Gordon JI, Normark S: An in vitro adherence assay reveals that Helicobacter pylori exhibits cell lineage-specific tropism in the human gastric epithelium. Proc Natl Acad Sci USA 1993;90:2035–2039. Jones GW, Isaacson RE: Proteinaceous bacterial adhesins and their receptors. CRC Crit Rev Microbiol 1983;10:229–260. Kallenius G, Winberg J: Bacterial adherence to periurethral cells in girls prone to urinary tract infection. Lancet 1978;ii:540. Fussell EN, Kaack MB, Cherry R, Roberts JA: Adherence of bacteria to human foreskins. J Urol 1988;140:997–1001. Johnson JR: Virulence factors in Escherichia coli urinary tract infection. Clin Microbiol Rev 1991; 4:80–128. Mobley HLT, Island MD, Massad G: Virulence determinants of uropathogenic Escherichia coli and Proteus mirabilis. Kidney Int 1994;46(suppl 47):129–136. Mobley HLT, Chippendale GR, Tenney JH, Mayrer AR, Crisp LJ, Penner JL, Warren JW: MR/K hemagglutination of Providencia stuartii correlates with adherence to catheters and with persistence in catheter-associated bacteriuria. J Infect Dis 1988;157:264–271. Roberts JA, Kaack MB, Fussell EN: Adherence to urethral catheters by bacteria causing nosocomial infections. Urology 1993;41:338–342. Mobley HLT, Chippendale GR, Tenney JH, Hull RA, Warren JW: Expression of type 1 fimbriae may be required for persistence of Escherichia coli in the catheterized urinary tract. J Clin Microbiol 1987;25:2253–2257. Korhonen TK, Virkola R, Westurlund B, Holthöfer H, Parkkinen J: Tissue tropism of Escherichia coli adhesins in human extraintestinal infections. Curr Top Microbiol Immunol 1990;151:115– 127. Hagberg L, Jodal U, Korhonen TK, Lidin-Janson G, Lindberg U, Svanborg-Edén C: Adhesion, hemagglutination and virulence of Escherichia coli causing urinary tract infections. Infect Immun 1981;31:564–570. Källenius G, Möllby R, Svenson SB, Helin I, Hultberg H, Cederberg B, Winberg J: Incidence of P-fimbriated Escherichia coli in urinary tract infections. Lancet 1981;ii:1369–1371. Väisänen-Rhen V, Elo J, Väisänen E, Siitonen A, Orskov I, Orskov F, Svenson SB, Makela PH, Korhonen TK: P-fimbriated clones among uropathogenic Escherichia coli strains. Infect Immun 1984;43:149–155.

Roberts

104

BOIS:L01:BIF01018XA.88

17 18 19 20 21 22 23

24 25

26

FF: W1FP

E1: IN01

$$$

Johnson JR, Roberts PL, Stamm WE: P-fimbriae and other virulence factors in Escherichia coli urosepsis: Association with patient’s characteristics. J Infect Dis 1987;156:225–229. Stromberg N, Nyholm P-G, Pascher I, Normark S: Saccharide orientation at the cell surface affects glycolipid receptor function. Proc Natl Acad Sci USA 1991;88:9340–9433. Braaten BA, Nou X, Kallenbach LS, Low DA: Methylation patterns in pap regulatory DNA control pyelonephritis-associated pili phase variation in E. coli. Cell 1994;76:577–588. Roberts JA, Kaack B, Källenius G, Möllby R, Winberg J, Svenson SB: Receptors for pyelonephritogenic Escherichia coli in primates. J Urol 1984;131:163–168. Roberts JA, Hardaway K, Kaack B, Fussell EN, Baskin G: Prevention of pyelonephritis by immunization with P-fimbriae. J Urol 1984;131:602–607. Lund B, Lindberg F, Marklund B-I, Normark S: Tip proteins of pili associated with pyelonephritis: New candidates for vaccine development. Vaccine 1988;6:110–112. Roberts JA, Marklund B-I, Ilver D, Haslam D, Kaack MB, Baskin G, Louis M, Möllby R, Winberg J, Normark S: The Gal(·1-4)Gal-specific tip adhesin of Escherichia coli P-fimbriae is needed for pyelonephritis to occur in the normal urinary tract. Proc Natl Acad Sci USA 1994;91:11889– 11893. Westerlund B, Kuusela P, Vartio T, van Die I, Korhonen TK: A novel lectin-independent interaction of P-fimbriae of Escherichia coli with immobilized fibronectin. FEBS Lett 1989;243:199–204. Stromberg N, Marklund B-I, Lund B, Ilver D, Hamers A, Gaastra W, Karlsson K-A, Normark S: Host specificity of uropathogenic Escherichia coli depends on differences in binding specificity to Gal·1-4Gal-containing isoreceptors. EMBO J 1990;9:2001–2010. Roberts JA: Experimental pyelonephritis in the monkey. III. Pathophysiology of ureteral malfunction induced by bacteria. Invest Urol 1975;13:117–120.

Dr. James A. Roberts, Department of Urology, Tulane Regional Primate Research Center, 18703 Three Rivers Road, Covington, LA 70433 (USA)

Tropism in Bacterial Infections

105

PIAM:L01:BIF01019XA.89

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 106–108

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Experimental Prostatitis: Impact, Trends and Future Wolfgang Weidner Urologische Klinik, Justus-Liebig-Universität, Giessen, Deutschland

Recently, animal models of acute and chronic bacterial prostatitis caused by Echerichia coli have been developed in small laboratory animals [1]. These models offer an opportunity to study many unresolved questions concerning the etiology, pathogenesis and treatment of chronic bacterial prostatitis. One most important question to be clarified by an animal model concerns the prerequisites for the progression from acute to chronic bacterial prostatitis. In our opinion, only two experimental designs fulfill these criteria (table 1). In about 50% of all animals, chronic prostatitis with histological signs of chronic inflammation, evidence of bacteria in the tissue and serum antibody response [2] could be achieved. Nickel et al. [3] observed in their model the protection of bacteria in glycocalyx-embedded microcolonies, thus giving a hint for the mechanism of bacterial persistence. Furthermore, this protection hampers the immune system from recognizing the causative bacteria. New experiments [4] suggest that mucosal immunity developed from previous urinary tract infections or prostatitis does not confer protection against recurrence. An elevated number of leukocytes in expressed prostatic secretions is the main criterion for diagnosis of nonbacterial prostatitis in humans [5]. However, this measurement is not available in laboratory animals. Experimental studies were intended to induce an inflammatory response, mainly characterized by an interstitial infiltration by lymphocytes and macrophages without evidence of common pathogenic bacteria [1]. Animal models in nonbacterial prostatitis refer to the above-mentioned bacterial infections [2, 3] after eradication of the causative origin, spontaneous, nonacute age-dependent prostatitis, and the induction of nonbacterial inflammation due to an autoimmune process. There are, unfortu-

PIAM:L01:BIF01019XA.89

FF: W1FP

E1: IN01

$$$

Table 1. Animal models of chronic bacterial prostatitis Reference

Species/microorganisms

Route of infection

Acute Chronic prostatitis prostatitis

Jantos et al. [2]

Mastomys natalensis/ E. coli

Bladder (via incision)

+

Chronic prostatitis (50%) 3–6 months after infection, antibody response

Nickel et al. [3]

Sprague-Dawley rats/E. coli

Transurethral

+

Chronic prostatitis (50%) 50 days, antibody response, microcolonies

Table 2. Nonbacterial prostatitis and autoimmunity: experimental studies in mice [9] Type of reaction Autoimmune damage in the prostate after immunization with syngenic ventral prostate lobes, Freund’s adjuvant Lymphocytal infiltration Disease transfer by lymphocytes

nately, no convincing models available concerning prostatitis due to sexually transmitted microorganisms [1]. An interesting observation is the occurrence of spontaneous, nonacute, agedependent prostatitis in different rat species [1, 6, 7]. Incidence of prostatitis obviously increases with age, sexual activity decreases the degree of inflammation. An intraluminal accumulation of seminal vesicle protein was found in sexually inactive rats, thus suggesting a possible mechanical cause of prostatic inflammation by ductal obstruction [1]. In humans, a host response against prostatic invasion by spermatozoa has been claimed as an important stimulus for sterile inflammation [8]; data are in accordance with the experimental results that obstruction of the seminal pathways may sustain chronic inflammation of the prostate gland. Finally, recent experimental results concerning the induction of nonbacterial prostatitis in mice (table 2) [9] by autoimmunization with syngenic prostatic

Experimental Prostatitis: Impact, Trends and Future

107

PIAM:L01:BIF01019XA.89

FF: W1FP

E1: IN01

$$$

material provoke comments. The pattern of inflammatory response is similar to that demonstrated in humans and was shown to be of immune origin by adoptive transfer studies. These data fit in well with our understanding of host-parasite interactions in prostatitis and the opinion that a disturbed nonlaminar urine flow in the prostatic urethra, sustaining a reflux into the prostatic ducts, may perpetuate inflammation [10]. The intraprostatic reflux of bacteria, bacterial antigens, urine constituents, sperms, etc., may cause a tissue lesion and possibly induce an autoimmune reaction inside the gland – a hypothesis which may explain some aspects of idiopathic, nonbacterial prostatitis.

References 1 2 3 4 5 6 7 8 9 10

Weidner W, Jantos C, Schiefer H-G: Animal models of prostatitis. Infect Urol 1993;6:65–69. Jantos C, Altmannsberger M, Weidner W, Schiefer H-G: Acute and chronic bacterial prostatitis due to E. coli: Description of an animal model. Urol Res 1990;18:207–211. Nickel JC, Olson ME, Barabas A, Benediktsson H, Dagupta MK, Costerton JW: Pathogenesis of chronic bacterial prostatitis in an animal model. 1990;66:47–54. Nickel JC, Ceri H, Olson ME, Schmidt S: The role of specific mucosal immunity in the pathophysiology of bacterial prostatitis. J Urol 1995;153(suppl):328A. Weidner W: Prostatitis – Diagnostic criteria, classification of patients and recommendations for therapeutic trials. Infection 1991;19(suppl 3):227–231. Naslund MJ, Strandberg JD, Doffey DS: The role of androgens and estrogens in the pathogenesis of experimental nonbacterial prostatitis. J Urol 1988;140:1049–1053. Muntzing J, Sufrin G, Murphy GP: Prostatitis in the rat. Scand J Urol Nephrol 1979;13:17–22. McClinton S, Eremin O, Miller JD: Inflammatory infiltrate in prostatic hyperplasia – Evidence of a host response to intraprostatic spermatozoa? Br J Urol 1990;55:606–610. Keetch DW, Humphrey P, Ratliff T: Development of a mouse model for nonbacterial prostatitis. J Urol 1994;152:247–250. Blacklock NJ: The anatomy of the prostate: Relationship with prostatic infection. Infection 1991; 19(suppl 3):111–114.

Prof. Dr. Wolfgang Weidner, Urologische Universitätsklinik, Klinikstrasse 29, D–35392 Giessen (Germany)

Weidner

108

PIAM:L01:BIF01020XA.89

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 109–117

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Adherence and the Pathogenesis of Urinary Tract Infection Hugh Connell a, Catharina Svanborg a, Spencer Hedges a, William Agace a, Maria Hedlund a, Majlis Svensson a, Mikael Benson b, Ulf Jodal c a

b c

Department of Medical Microbiology, Section for Clinical Immunology, Lund University, Lund, and Departments of Pediatrics and Clinical Immunology, Göteborg University, Göteborg, Sweden

Urinary tract infections (UTIs) are among the most common bacterial infections in humans. The frequency varies with age, gender and socioeconomic background. Bacteriuria is found at screening in about 1% of girls from birth to puberty, in 2% or more of pregnant women, and in 15–20% of women at 70 years of age. In addition, symptomatic infections (acute pyelonephritis, acute cystitis) occur frequently in children and sexually active women. The susceptibility is generally higher in females than males, except for young boys and elderly men in whom the frequency of UTI is comparable to females of the same age group. Socioeconomic background variables include access to medical care (surgical correction of malformations, antenatal care of expectant mothers, estrogen treatment of women after menopause) and sexual and contraceptive practices [1]. Escherichia coli is the main cause of infections in the human urinary tract. The infecting strains fall into two groups: those that cause symptomatic UTI (pyelonephritis or cystitis), and those which colonize and persist as asymptomatic bacteriuria (ABU). The large intestine serves as a reservoir for uropathogenic E. coli [1–4]. Bacteria colonize the large intestine, spread to the vaginal and periurethral areas and ascend into the urinary tract. Bacterial Adherence The colonization of mucosal surfaces during the pathogenesis of UTI is influenced by bacterial adherence. Uropathogenic E. coli express an array of adhesins, including P, type 1, S, and Dr fimbriae [5]. P fimbriae have been shown

PIAM:L01:BIF01020XA.89

FF: W1FP

E1: IN01

$$$

to enhance bacterial virulence for the urinary tract [6]. Recent studies have suggested that P fimbriae are important also for the colonization of the large intestine. P-fimbriated E. coli adhere to human colonic epithelial cells in vitro and persisted longer in the large intestine of UTI-prone children than other E. coli strains [7, 8]. P-fimbriated E. coli occur more frequently in the urinary tract of patients with UTI than in the fecal flora of healthy individuals [6]. It is not clear if this accumulation of P-fimbriated E. coli in UTI is a result of increased intestinal persistence or if there is a difference in the ability to cause UTI between Pfimbriated and other E. coli in the large intestine of the UTI-prone patient. Certain individuals appear to be more susceptible to intestinal colonization by P-fimbriated E. coli [5, 9]. It has been shown that patients prone to UTI have an increased carriage of P-fimbriated E. coli in the large intestine, and of Enterobacteriaceae in the vagina and periurethral area. P fimbriae recognize as receptors Gal·1–4Galß and GalNAcß1–3Gal·1– 4Galß1 containing oligosaccharide sequences in the globoseries of glycolipids [10,11]. The fimbriae are encoded by the pap chromosomal gene cluster which contains 11 separate genes (papA to K) [12]. The pap gene clusters from different E. coli strains show extensive sequence homology except for papA, which encodes the antigenically variable fimbrial subunit and the sequences that specify the adhesin specificity (papG) [13]. While P fimbriae enhance the virulence of uropathogenic strains through specific adherence, the role of type 1 fimbriae in virulence remains undefined. Studies in animal models have suggested that type 1 fimbriation increases the survival of E. coli in the urinary tract [14–19]; however, epidemiological studies have failed to reveal a correlation between type 1 fimbriation and virulence [20]. Type 1 fimbriae are encoded by the chromosomally located fim gene cluster. The fimbriae consist of a major structural subunit (fimA) and several minor components including the adhesin (fimH). fimH recognizes terminally located D-mannose moieties on cell-bound and secreted glycoproteins [21, 22].

Mucosal Cytokine Responses What are the consequences of bacterial binding to uroepithelial cells during UTI? Bacteria infecting the urinary tract induce fever, elevate acute phase reactants (C-reactive protein (CRP) and erythrocyte sedimentation rate), lead to a leukocyte influx, reduce renal tubular function and elicit IgA and cytokine responses. Bacterial interactions with host cell receptors via P and type 1 fimbriae can result in the activation of the respiratory burst in granulocytes, degranulation of mast cells and cytokine release from epithelial cells in vitro [23–26]. These cytokines have a local and/or systemic effect. Two examples of cytokines released

Connell/Svanborg/Hedges/Agace/Hedlund/Svensson/Benson/Jodal

110

PIAM:L01:BIF01020XA.89

FF: W1FP

E1: IN01

$$$

➤

➤

Fig. 1. Urine cytokine responses from a patient deliberately colonized with an asymptomatic bacteriuria E. coli strain: (a) IL-6 pg/ml ([) and (b) IL-8 ng/ml (P) during the first 10 days of colonization. The arrows represent the times of bacterial instillation into the bladder.

by uroepithelial cells in response to bacteria are interleukin (IL)-6 and IL-8. Both of these cytokines can be found in the urines of patients infected with E. coli. IL-6 is an endogenous pyrogen which activates fever, is an activator of acute phase reactants (CRP), and a B-lymphocyte maturation factor (mucosal IgA B lymphocytes). IL-8 is a chemoattractant for neutrophils which migrate to the site of infection. We have studied the mucosal cytokine response to microbial challenge both in situ and in vivo. Deliberate colonization of the human urinary tract with E. coli stimulated the secretion of IL-6 and IL-8 into the urine without elevation of the serum levels (fig. 1) [27, 28]. Urinary levels of IL-6 were also elevated in patients with natural UTI, whereas circulating IL-6 levels were elevated only in those patients who developed febrile infections [29]. This separation of local and systemic IL-6 responses was also observed during experimental infection in mice [30]. The circulating IL-6 levels were elevated after intraperitoneal infection, but no urinary IL-6 response occurred unless the mice were challenged in the urinary tract. These observations suggested that mucosal cells were activated directly by bacteria to produce cytokines.

Adherence and Pathogenesis of UTI

111

PIAM:L01:BIF01020XA.89

▲

FF: W1FP

▲

E1: IN01

$$$

▲

▲

Fig. 2. The IL-6 response of kidney epithelial cells (A498) stimulated with E. coli isogenic strains carrying type 1 (X), P fimbriae ([) and vector controls (o, $), and with a wild-type uropathogenic strain ()).

Epithelial Cells as Cytokine Producers The rapidity of the IL-6 and IL-8 responses in vivo indicated that a local cell type such as the epithelial cell might be responsible for the secretion of the cytokines (IL-6 and IL-8) during UTI. We have shown that epithelial cell lines (kidney and bladder) are capable of secreting IL-6 after stimulation with E. coli, isolated fimbriae and lipopolysaccharide (LPS) [24]. We have also shown by indirect immunofluorescence and reverse transcriptase-polymerase chain reaction (RTPCR) that these same cells are capable of synthesizing a variety of cytokines [28, 31]. They synthesized IL-1·, IL-6 and IL-8 but not IL-1ß, TNF or GM-CSF after stimulation with E. coli; however, only IL-6 and IL-8 were secreted (not IL-1·). Fimbriation of the bacteria affects the magnitude of the cytokine response. We have stimulated uroepithelial cells (kidney) with P- and type 1-fimbriated E. coli. The fimbriated strains induced higher cytokine responses (IL-6 and IL-8) in epithelial cells than isogenic, nonfimbriated bacteria (fig. 2) [32]. Isolated P fimbriae have been shown to activate cytokine responses in an adhesin-dependent manner [24]. P fimbriae lacking the adhesin induced a lower cytokine response than fimbriae with intact adhesive capacity. An inhibitor (PDMP) of glycolipid receptor expression impaired attachment and cytokine responses to P-fimbriated

Connell/Svanborg/Hedges/Agace/Hedlund/Svensson/Benson/Jodal

112

PIAM:L01:BIF01020XA.89

FF: W1FP

E1: IN01

$$$

bacteria [33]. Soluble receptor analogues (globoside, ·-methyl-ß-mannoside) for P and type 1 fimbriae were also found to inhibit cytokine responses induced by Pand type 1-fimbriated bacteria [25, 34]. The oligosaccharide receptors recognized by P fimbriae are bound to ceramide in the outer leaflet of the lipid bilayer. Ceramide has recently been recognized as a second messenger in the sphingomyelin signal transduction pathway and is cleaved from sphingomyelin by sphingomyelinase. The hydrolysis of sphingomyelin can also occur upon exposure of cells to exogenous agonists that activate endogenous sphingomyelinases. Such agonists include TNF-· and IL-1ß. We have demonstrated that P-fimbriated E. coli activate the ceramide signal pathway and propose that this activation contributes to the epithelial cytokine response induced by P-fimbriated E. coli in the urinary tract [31].

Functions of Secreted Cytokines What are the consequences of cytokine expression by epithelial cells in the urinary tract? There is a correlation between the IL-6 response to bacteria in the urinary tract and the least response to infection measured as fever or CRP, in certain patient groups. One of the prime symptoms of acute pyelonephritis is pyuria – the efflux of polymorphonuclear cells (PMNs) into the urine. IL-8 as a chemoattractant for PMNs has been found in the urine of patients with UTI. In a study of patients deliberately colonized with fimbriated E. coli, it was shown IL-8 levels correlated directly to levels of PMNs present in the urine [25]. The strains used to colonize these patients were shown to be IL-8 in uroepithelial cell lines. This suggested that the production of IL-8 by epithelial cells may play an important role in the efflux of PMNs across the mucosal layer. To further investigate the role of IL-8 and neutrophils in the inflammatory response, in vitro experiments were carried out using the Transwell technology. A confluent epithelial (kidney or bladder) cell layer was grown on the underside of a Transwell insert. Media containing human neutrophil was placed in the upper compartment and media containing stimulant was placed in the lower compartment. Samples were removed from the system at various intervals to determine neutrophil migration. Neutrophils migrated across the cell layer in response to both E. coli and IL-1, but not to media alone (fig. 3) [35, 37]. How do the PMNs pass across the epithelial cell layer into the lumen of the urinary tract? PMN migration through endothelial cells involves cell adhesion molecules and ß2-integrins, a similar mechanism may be involved in transepithelial cell migration. Epithelial cells from the Transwells stimulated with E. coli or IL-1 were analyzed for the expression of cell adhesion molecules. Cells were stained with antibodies to various cell adhesion molecules and analyzed by flow

Adherence and Pathogenesis of UTI

113

PIAM:L01:BIF01020XA.89

FF: W1FP

E1: IN01

$$$

▲

▲

▲

Fig. 3. The kinetics of neutrophil migration across A498 kidney epithelial cell layers grown on Transwell membranes. Cells were prestimulated for 24 h with medium ()), E. coli (108 bacteria/ml) (o) or IL-1· (1 ng/ml) (P). Results are the mean (SE) of 5–6 separate experiments.

cytometry. Control cells (both bladder and kidney) stained positive for ICAM-1 and when stimulated with E. coli or IL-1 there was an up-regulation in the expression of this molecule [34]. To examine the involvement of IL-8, ICAM-1 and ß2-integrins in neutrophil migration, blocking antibodies to each molecule were added to the Transwell system. The effect of these antibodies was measured as an increase or decrease in neutrophil migration in response to E. coli. Antibodies to IL-8, ICAM-1, CD11b and CD18 reduced neutrophil migration by 60–70%, whereas antibodies to CD11a did not influence migration (table 1) [35]. It appears that neutrophil migration across epithelial cells is dependent upon the expression of ICAM-1 on the epithelial cell and the expression of the Mac-1 complex (CD11b/CD18) on the neutrophil surface and the presence of the neutrophil chemoattractant IL-8. Neutrophils and epithelial cells are not the only cells present in the mucosa of the urinary tract, and are not the only cell types capable of producing cytokines. Macrophages and lymphocytes are present in the urinary submucosa. These cells are capable of responding to exogenous cytokines and of producing and secreting a wide variety of cytokines. Questions arise as to the involvement of lymphocyte

Connell/Svanborg/Hedges/Agace/Hedlund/Svensson/Benson/Jodal

114

PIAM:L01:BIF01020XA.89

FF: W1FP

E1: IN01

$$$

Table 1. The role of IL-8, ICAM-1 and ß2-integrins in transepithelial neutrophil migration in the kidney epithelial cell line A498 Antibody

Percent neutrophil migration (SE)1 medium

E. coli

26 (4) 19 (1)

65 (5) 26 (3)

– Anti-ICAM-1 (10 Ìg/ml)

8 (4) 2 (1)

53 (6) 8 (3)

– Anti-CD11a (10 Ìg/ml) Anti-CD11b (10 Ìg/ml) Anti-CD18 (10 Ìg/ml)

27 (5) 27 (4) 9 (2) 3 (3)

79 (5) 73 (9) 23 (4) 21 (3)

– Anti-IL-8 (1 Ìg/ml)

1

Neutrophil migration (3 h) across A498 kidney epithelial cells prestimulated for 24 h with E. coli (108 bacteria/ml) or medium. Results are the mean (standard error) of four separate experiments.

derived cytokines and the role they play in the response of the epithelial cell. Epithelial cells have been shown to respond and release cytokines when stimulated with exogenous cytokines of lymphoid origin [36]. In summary, these findings are consistent with the following scenario. Bacteria attach to the mucosal lining of the urinary tract and trigger a cytokine response. The cytokines will in turn activate inflammation locally and spread to systemic sites where they cause fever and the release of acute phase reactants. This is why we see a direct association between bacterial virulence properties, inflammation and severity of infection.

Acknowledgments This study was supported by grants from the Swedish Medical Research Council, the Royal Physiographical Society of Lund, the Medical Faculties of the Lund and Göteborg Universities, and the Österlund and Crawford Foundations for Scientific Research.

Adherence and Pathogenesis of UTI

115

PIAM:L01:BIF01020XA.89

FF: W1FP

E1: IN01

$$$

References 1 2 3 4

5 6 7 8 9

10

11 12

13

14

15 16

17 18

19 20

21

Kunin C: Detection, Prevention and Management of Urinary Tract Infections. Lea & Febiger, Philadelphia, 1987. Bettelheim K, Taylor J: A study of Escherichia coli isolated from chronic urinary infections. J Med Microbiol 1969;2:225–236. Grüneberg R: Relationship of infecting urinary organisms to the faecal flora in patients with symptomatic urinary infections. Lancet 1969;ii:766–768. Lidin-Janson G, Hanson L, Kaijser B, Lincoln K, Lindberg U, Olling S, Wedel H: Comparison of Escherichia coli from bacteriuric patients with those from feces of healthy school children. J Infect Dis 1977;136:346–353. Johnson J: Virulence factors in Escherichia coli urinary tract infection. Clin Microbiol Rev 1991;4: 80–128. Svanborg C, Ørskov F, Ørskov I: Fimbriae and disease; in Klemm P (ed): Bacterial Fimbriae. Boca Raton, CRC Press, 1994, pp 253–266. Wold A, Thorssén M, Hull S, Svanborg C: Attachment of Escherichia coli via mannose of Gal·1– 4Galß containing receptors to human colonic epithelial cells. Infect Immun 1988;56:2531–2537. Wold A, Caugant D, Lidin-Janson G, de Man P, Svanborg C: Resident colonic Escherichia coli strains frequently display uropathogenic characteristics. J Infect Dis 1992;165:46–52. Plos K, Connell H, Jodal U, Marklund B-I, Mårild S, Wettergren B, Svanborg C: Intestinal carriage of P-fimbriated Escherichia coli and the susceptibility to urinary tract infection in young children. J Infect Dis 1995;171:625–631. Leffler H, Svanborg-Edén C: Chemical identification of a glycosphingolipid receptor for Escherichia coli attaching to human urinary tract epithelia cells and agglutinating human erythrocytes. FEMS Microbiol Lett 1980;8:127–134. Källenius G, Möllby R, Svensson S, Cedergren B: The Pk antigen as receptor for the hemagglutination of pyelonephritic Escherichia coli. FEMS Microbiol Lett 1980;7:297–302. Hull R, Gill R, Hsu P, Minshew B, Falkow S: Construction and expression of recombinant plasmids encoding type 1 or D-mannose-resistant pili from the urinary tract infection Escherichia coli isolate. Infect Immun 1981;33:933–938. Normark S, Båga M, Göransson M, Lindberg F, Lund B, Norgren M, Uhlin B: Genetics and biogenesis of Escherichia coli adhesins; in Mirelman D (ed): Microbial Lectins and Agglutinins: Properties and Biological Activity. New York, Wiley Interscience, 1994, pp 113–143. Aronson M, Medalia O, Schori L, Mirelman D, Sharon N, Ofek I: Prevention of colonisation of the urinary tract of mice with Escherichia coli by blocking of bacterial adherence with methyl-alphaD-mannopyranoside. J Infect Dis 1979;139:329–332. Hagberg L, Hull R, Hull S, Falkow S, Freter R, Svanborg-Edén C: Contribution of adhesion to bacterial persistence in the mouse urinary tract. Infect Immun 1983;40:265–272. Abraham S, Babu J, Giampapa C, Hasty D, Simpson W, Beachey E: Protection against Escherichia coli-induced urinary tract infections with hybridoma antibodies directed against type 1 fimbriae or complementary D-mannose receptors. Infect Immun 1985;48:625–628. Hultgren S, Porter T, Schaeffer A, Duncan J: Role of type 1 fimbriae and effects of phase variation on lower urinary tract infections produced by Escherichia coli. Infect Immun 1985;50:370–377. Iwahi T, Abe Y, Nakao A, Imada A, Tsuchiya K: Role of type 1 fimbriae in the pathogenesis of ascending urinary tract infection induced by Escherichia coli in mice. Infect Immun 1983;39:1307– 1315. Keith B, Maurer L, Spears P, Orndorff P: Receptor-binding function of type 1 pili effects bladder colonization by a clinical isolate of Escherichia coli. Infect Immun 1986;53:693–696. Hagberg L, Jodal U, Korhonen T, Lidin-Janson G, Lindberg U, Svanborg-Eden C: Adhesion, hemagglutination, and virulence of Escherichia coli causing urinary tract infections. Infect Immun 1981; 31:564–570. Giampapa C, Abraham S, Chiang T, Beachey E: Isolation and characterization of a receptor for type 1 fimbriae of Escherichia coli from guinea pig erythrocytes. J Biol Chem 1988;263:5362–5367.

Connell/Svanborg/Hedges/Agace/Hedlund/Svensson/Benson/Jodal

116

PIAM:L01:BIF01020XA.89

22

23

24 25 26 27 28 29

30 31 32 33 34 35 36 37

FF: W1FP

E1: IN01

$$$

Wold A, Mestecky J, Tomana M, Kobata A, Ohbayashi H, Endo T, Svanborg-Edén C: Secretory immunoglobulin A carries oligosaccharide receptors for Escherichia coli type 1 fimbrial lectin. Infect Immun 1990;58:3073–3077. Steadman R, Topley N, Jenner D, Davies M, Williams J: Type 1 fimbriate Escherichia coli stimulates a unique pattern of degranulation by polymorphonuclear leucocytes. Infect Immun 1988;56: 815–822. Hedges S, Svensson M, Svanborg C: Interleukin-6 response of epithelial cells to bacterial stimulation in vitro. Infect Immun 1992;60:1295–1301. Agace W, Hedges S, Ceska M, Svanborg C: Interleukin-8 and the neutrophil response to mucosal Gram-negative infection. J Clin Invest 1993;92:780–785. Malaviya R, Ross E, Jakschik B, Abraham S: Mast cell degranulation induced by type 1 fimbriated Escherichia coli in mice. J Clin Invest 1994;93:1645–1653. Hedges S, Anderson P, Lidin-Janson G, Svanborg C: Interleukin-6 response to deliberate Gramnegative colonization of the human urinary tract. Infect Immun 1991;59:421–427. Agace W, Hedges S, Andersson U, Andersson J, Ceska M, Svanborg C: Selective cytokine production by epithelial cells following exposure to Escherichia coli. Infect Immun 1993;61:602–609. Hedges S, Stenqvist K, Lidin-Janson G, Sandberg T, Svanborg C: Comparison of urine and serum concentrations of interleukin-6 in women with either acute pyelonephritis or asymptomatic bacteriuria. J Infect Dis 1992;166:653–656. de Man P, van Kooten C, Aarden L, Engberg I, Svanborg-Edén C: Interleukin-6 induced by Gramnegative bacterial infection at mucosal surfaces. Infect Immun 1989;57:3383–3388. Hedges S, Agace W, Svensson M, Sjögren A-C, Ceska M, Svanborg C: Uroepithelial cells are a part of a mucosal cytokine network. Infect Immun 1994;62:2315–2321. Hedlund M, Svensson M, Nilsson Å, Duang R, Svanborg C: Role of the ceramide signalling pathway in cytokine responses to P-fimbriated Escherichia coli. J Exp Med 1996;183:1–8. Svensson M, Lindstedt R, Radin N, Svanborg C: Epithelial glucosphingolipid expression as a determinant of bacterial adherence and cytokine production. Infect Immun 1994;62:4404–4410. Linder H, Engberg I, Hoschützky H, Mattsby Baltzer I, Svanborg-Edén C: Adhesion-dependent activation of mucosal IL-6 production. Infect Immun 1991;59:4357–4362. Agace W, Pattarroyo M, Svensson M, Svanborg C: Escherichia coli induce trans-epithelial neutrophil migration by an ICAM-1-dependent mechanism. Infect Immun 1995;63:4054–4062. Hedges S, Bjarnadottir M, Agace W, Svanborg C: Interleukin-4 and gamma interferon modify the epithelial cell cytokine response to Gram-negative bacteria. Cytokine, in press. Godaly G, Offord RE, Proudfoot AEI, Svanborg C, Agace WW: Escherichia coli induced transuroepithelial neutrophil migration: Role of epithelial interleukin-8 and neutrophil IL-8 receptor A (submitted).

Dr. Hugh Connell, Department of Medical Microbiology, Section for Clinical Immunology, Sölvegatan 23, Lund University, S–22 362 Lund (Sweden)

Adherence and Pathogenesis of UTI

117

PIAM:L01:BIF01021XA.88

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 118–124

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Identification of a Novel Antibacterial Factor from Human Urine Hugh Connell a, Hemant Sabharwal a, Lo Persson b, Michael Zasloff c, Catharina Svanborg a a b c

Departments of Medical Microbiology (Section for Clinical Immunology) and Physiology, Lund University, Lund, Sweden; Magainin Research Institute, Magainin Pharmaceuticals Inc., Plymouth Meeting, Pa., USA

Human urine has been regarded as a growth medium for bacteria [1]; however, not all Escherichia coli strains are capable of growth in this medium. Kaye [2] showed that isolates from patients with urinary tract infections (UTIs) grew well in urine while most fecal E. coli strains were killed. The ability of urinary tract E. coli strains to grow at faster rates in urine than non-urinary tract colonizers has been suggested as a means by which bacteria persist and avoid elimination by micturition [3]. Several studies have proposed models to aid in the understanding of bacterial colonization of the urinary tract [4–6]. O’Grady and Pennington [5] derived an artificial bladder model for bacterial colonization with theoretical values which must be met in order for a bacterium to colonize. In contrast, Cox and Hinman [3] showed a discrepancy between in vitro growth and bacterial colonization in human volunteers. Bacteria which successfully colonized the artificial bladder were eliminated from the human urinary tract. It was later shown that the bladder mucosa of guinea pigs was capable of killing the mucosal E. coli [5]. Schulte-Wisserman et al. [7] suggested that this killing effect was due to the production of bactericidal molecules by epithelial cells. These studies prompted a search for antibacterial molecules in urine and the identification of inhibiting factors like urea, organic acids, salts, pH and osmolarity [2, 8]. A zinc-containing protein isolated from male prostatic fluid was also found to be antibacterial [9, 10]. Asscher et al. [8] showed that urinary isolates of E. coli grew optimally at a pH of between 6.0 and 7.0 and that growth was inhibited in both alkaline and acidic urines outside this pH range. The physiological

PIAM:L01:BIF01021XA.88

FF: W1FP

E1: IN01

$$$

range of human urine pH is between 4.6 and 7.2. Kass and Zangwill [11] reduced bacterial counts in the urine of patients with chronic UTI by administering d,lmethionine that produces highly acidic urine. The inhibition of bacterial growth was due to undissociated organic acids which were in high concentration in acidic urine. Urine osmolarity has been shown to effect the ability of E. coli to grow in human urine [8]. At high osmolarity (1,200 mosm/kg), bacterial growth was inhibited by urea, sodium chloride, sodium sulfate, potassium chloride and potassium sulfate in the urine. The inhibition of growth was probably due to the hyperosmolarity of the urine rather than to one specific molecule. Urine of low osmolarity (^200 mosm/kg) was also shown to inhibit bacterial growth. This is probably due to the very low nutrient content of urine and lends support to the idea that high fluid intake assists in the clearance of bacteriuria. While urine appears to be inhibitory for bacterial growth, Chambers and Kunin [12] showed that urine confers an osmoprotective effect on E. coli. Urinary isolates of E. coli grown in a defined minimal medium were inhibited by high concentrations of electrolytes and sugars in direct relation to their osmotic strength. The addition of human urine and betaine to this medium increased the osmotic resistance of the E. coli isolates to these substances. It was proposed that urine contained low-molecular-weight (MW) osmoprotective substances which E. coli could use to protect against the hypertonic effect of urine. There appears to be a range of molecules in human urine other than electrolytes that are inhibitory for bacterial growth. For bacteria to persist in the urinary tract they must overcome these inhibitory and antibacterial factors in human urine. This study describes the isolation and partial purification of a fraction from human urine with antibacterial activity.

Results We found marked differences between E. coli strains in the ability to survive in human urine. E. coli strains from patients with different forms of UTI (acute pyelonephritis n = 5; acute cystitits n = 5; asymptomatic bacteriuria (ABU) n = 5) and from the intestinal flora of healthy individuals (n = 10) were tested for growth in a pool of male urine. All of the UTI isolates grew to reach a cell density of 6106 colony-forming units (CFU)/ml in 24 h. In contrast, most of the E. coli intestinal isolates (7/10) were killed. This suggested that urine contains antibacterial components to which the strains causing UTI are resistant. In order to identify the mechanism(s) explaining the observed antibacterial effect, two E. coli strains were selected as a model. E. coli 83972 was the clinical isolate that had been found to successfully establish bacteriuria following deliber-

Urine Antibacterial Factor

119

PIAM:L01:BIF01021XA.88

FF: W1FP

E1: IN01

$$$

Fig. 1. Colony-forming units (CFU) of E. coli in urine. ) = Strain No. 83972; [ = strain No. HB101.

ate colonization of 12 human hosts [13]. This strain was able to grow exponentially in pools of male and female urine, to reach a cell density of 108–109 CFU/ml of urine by 24 h (fig. 1). E. coli HB101 was a K-12 strain of the type that can be found in the fecal flora of healthy carriers. This strain was killed in male urine; at an inoculum of ! 105 CFU/ml, no viable bacteria remained at 24 h (fig. 1). At higher inoculum concentrations a reduction in CFU/ml of at least 2 logs10 occurred. In female urine, the concentration of E. coli HB101 decreased to 102 CFU/ml during the first 6 h, but by 24 h it had reached a value of 107 CFU/ml (fig. 1). This biphasic growth curve in female urine may be the result of a gene induction effect in a subpopulation of E. coli HB101, or be due to the consumption of a growthinhibiting substance in female urine. The possibility of gene induction was examined by taking E. coli HB101 from a 6-hour culture in urine and re-inoculating it into a fresh urine sample. The same decrease in bacterial counts was observed, suggesting that the bacteria that survived in the 6-hour culture retained their susceptibility to the growth-inhibiting factor. The possibility of consumption of an inhibitor was tested by filter-sterilizing urine in which E. coli HB101 had grown for 6 h, and re-inoculating the urine with broth grown E. coli HB101. The growth in this urine was exponential, suggesting that the decrease in E. coli HB101 in fresh urine was due to an inhibitory substance being consumed by the bacteria. Urine was subjected to gel filtration column chromatography in order to isolate the fraction causing the antibacterial activity (fig. 2). Fractions collected off a Sephadex G-25 column were pooled (I–VI) and tested for effects on bacterial growth in a minimal salts and glucose medium with amino acids and vitamins to

Connell/Sabharwal/Persson/Zasloff/Svanborg

120

PIAM:L01:BIF01021XA.88

FF: W1FP

E1: IN01

$$$

Fig. 2. Chromatogram (Sephadex G-25 column) and growth of E. coli, strains Nos. 83972 and HB101, in a defined minimal salts medium, supplemented with fractions I–VI.

supplement for the auxotrophic mutations in E. coli HB101. E. coli 83972 grew exponentially in medium supplemented with all of the urine fractions with the exception of fraction VI. E. coli HB101 grew exponentially in fractions I, II, IV and V but not in fractions III and VI. Fraction VI contains salts, inorganic acids, organic acids and urea which are known to be inhibitory for bacterial growth. Since this fraction inhibited both E. coli strains it was not considered further. The spectrum of activity against different E. coli strains (both urinary and fecal isolates) was tested with whole urine and fraction III. Whole male urine and fraction III inhibited the growth of intestinal E. coli isolates from healthy individuals (n = 10), but not UTI strains (n = 15). The UTI strains were able to grow in minimal media supplemented with fraction III whereas the 10 fecal isolates could not. The growth in fraction III was the same as that seen for growth in whole urine. The purification of the inhibitory component was continued from fraction III. Fraction III was enriched for the antibacterial component by organic solvent extraction, ion-exchange chromatography and ninhydrin-stained thin layer chromatography. This showed the antibacterial component to be weakly cationic (1 or 2 positively charged groups), ninhydrin reactive showing the presence of amine groups, was of low MW (between 500 and 1,000), and very hydrophilic. This type of profile shows compatibility with the chemical family of polyamines.

Urine Antibacterial Factor

121

PIAM:L01:BIF01021XA.88

FF: W1FP

E1: IN01

$$$

Determination of polyamine concentrations in human urine and fraction III was carried out using the method described by Holm et al. [14]. Spermine, spermidine and putrescine in the acetylated and nonacetylated forms were found. Testing of commercial polyamines (spermine, spermidine and putrescine) against our bacteria showed differential killing of our bacterial strains but at concentrations far higher than found in human urine. Further characterization is being carried out to determine the structure of this compound.

Discussion Bacterial adherence was initially thought to explain the establishment of bacteria in the urinary tract [15, 16]. By attaching to the mucosa the bacteria avoided removal during micturition. Recent evidence contradicts this hypothesis. E. coli strains isolated from patients with ABU fail to express the fimbrial determinants that characterize the virulent clones [17]. Human colonization studies have shown that an ABU strain which lacked the known adhesive determinants was able to establish bacteriuria (1 105 CFU/ml) in several different human hosts and to persist for periods of 30 days or more in their urinary tracts [13]. In contrast, isogenic derivatives of this ABU strain expressing P and type 1 fimbriae were rapidly eliminated from the urinary tract concomitant with the induction of an inflammatory response. This suggested that adherence was not required for bacterial persistence in the urinary tract and led us to re-examine the question about mechanisms which might control the initial establishment of bacteriuria. We found that urine contains a low MW, hydrophilic, amine-containing substance with bactericidal activity. It appears to be related to the family of polyamines but differs substantially in charge and size. The mechanism of the antibacterial activity has yet to be shown but polyamines are known to inhibit DNA replication in bacteria and to be taken up through an LPS-dependent transport system [18–20]. From purification analysis, the polyamine-like component we have isolated from urine is not a protein, or a peptide, ruling out the probability of it being a defensin, a magainin, a cecropin or a cryptdin. Defensins, magainins, cecropins and cryptdins are small peptide molecules which have been shown to have antibacterial activity [21–24]. They have been isolated from a variety of cell types and host species, and appear to be released in host secretions as part of a nonspecific defense mechanism. There was a difference in the antibacterial activity between male and female urine. While male urine effectively killed the sensitive strain, female urine had a moderate effect. The inhibitory component in female urine was consumed by the bacteria. There are two possible explanations for this, either (a) the compound we have isolated from male urine is different from that in female urine, or (b) there

Connell/Sabharwal/Persson/Zasloff/Svanborg

122

PIAM:L01:BIF01021XA.88

FF: W1FP

E1: IN01

$$$

are variations in the concentrations of this component in the urinary tract between males and females. Regardless of the mechanism, these findings indicate that differences in killing capacity exist between human hosts. This might contribute to the variation in susceptibility to UTI. Further studies should compare urine from patients prone to UTI and healthy controls for the bactericidal factor. The isolated component had a selective killing effect on bacteria. All UTI strains tested were resistant to the antibacterial effect whereas 70% of fecal strains were sensitive to it. The variation between strains in the sensitivity to this molecule suggested that the uropathogenic strains of E. coli are uniquely adapted to survival and persitence in the urinary tract. The large intestine serves as a reservoir for uropathogenic E. coli and for other Enterobacteriaceae that cause UTI [25–27]. The 30% of the fecal isolates that were resistant to this antibacterial urine factor may be the subset which can cause UTI.

Acknowledgments This study was supported by grants from the Swedish Medical Research Council, the Royal Physiographical Society of Lund, the Medical Faculty of the Lund University, and the Österlund and Crawford Foundations for Scientific Research.

References 1 2 3 4 5 6 7

8 9 10 11 12

Pasteur L: Exam du rôle attribué au gaz oxygène atmosphérique dans la destruction des matières animales et végétales après la mort. C R Acad Sci Paris 1863;56:734–740. Kaye D: Antibacterial activity of human urine. J Clin Invest 1968;47:2374–2390. Cox C, Hinman F: Experiments with induced bacteriuria, vesical emptying and bacterial growth on the mechanism of bladder defense to infection. J Urol 1961;86:739–748. O’Grady F, Pennington J: Bacterial growth in an in vitro system simulating conditions in the urinary bladder. Br J Exp Pathol 1966;47:283–290. Norden C, Green G, Kass E: Antibacterial mechanisms of the urinary bladder. J Clin Invest 1968; 47:2689–2700. Gordon D, Riley M: A theoretical and experimental analysis of bacterial growth in the bladder. Mol Microbiol 1992;6:555–562. Schulte-Wissermann H, Mannhardt W, Schwarz J, Zepp F, Bitter-Sauermann D: Comparison of the antibacterial effect of uroepithelial cells from healthy donors and children of asymptomatic bacteriuria. Eur J Pediatr 1985;144:230–233. Asscher A, Sussman M, Waters W, Harvard Davis R, Chick S: Urine as a medium for bacterial growth. Lancet 1966;i:1037–1041. Levy B, Fair W: The location of antibacterial activity in the rat prostatic secretions. Invest Urol 1973;11:173–177. Stamey T: Prostatitis. J R Soc Med 1980;74:22–40. Kass E, Zwangwill D: Biology of pyelonephritis. Boston, 1960, p 663. Chambers S, Kunin C: The osmoprotective properties of urine for bacteria: The protective effect of betaine and human urine against low pH and high concentrations of electrolytes, sugars, and urea. J Infect Dis 1985;152:1308–1316.

Urine Antibacterial Factor

123

PIAM:L01:BIF01021XA.88

13

14

15 16 17 18

19 20

21 22

23 24

25 26

FF: W1FP

E1: IN01

$$$

Anderson P, Engberg I, Lidin-Janson G, Lincoln K, Hull R, Hull S, Svanborg-Edén C: Persistence of Escherichia coli bacteriuria not determined by bacterial adherence. Infect Immun 1991;59:2915– 2921. Holm I, Persson L, Pegg A, Heby O: Effects of S-adenosyl-1,8-diamino-3-thio-octane and S-methyl5)-methylthioadenosine on polyamine synthesis in Ehrlich ascites-tumour cells. J Biochem 1989; 261:205–210. Svanborg-Edén C, Eriksson B, Hanson L: Adhesion of Escherichia coli to human uroepithelial cells in vitro. Infect Immun 1977;18:767–774. Hagberg L, Hull R, Hull S, Falkow S, Freter R, Svanborg-Edén C: Contribution of adhesion to bacterial persistence in the mouse urinary tract. Infect Immun 1983;40:265–272. Leffler H, Svanborg-Edén C: Glycolipid receptors for uropathogenic Escherichia coli on human erythrocytes and uroepithelial cells. Infect Immun 1981;34:920–929. Srivenugopal K, Ali-Osman F: Stimulation and inhibition of 1,3-bis(2-chloroethyl)-1-nitrosoureainduced strand breaks and interstrand cross-linking in ColE1 plasmid deoxyribonucleic acid by polyamines and inorganic cations. Biochem Pharmacol 1990;40:473–479. Tjandrawinata R, Hawel L, Byus C: Regulation of putrescine export in lipopolysaccharide or IFNÁ-activated murine monocytic-leukemic RAW 264 cells. J Immunol 1994;154:3039–3052. Wei T-F, Bujalowski W, Lohman T: Cooperative binding of polyamines induces the Escherichia coli single-strand binding protein-DNA binding mode transitions. Biochemistry 1992;31:6166– 6174. Eisenhauer P, Harwig S, Lehrer P: Cryptdins: Antibacterial defensins of the murine small intestine. Infect Immun 1992;60:3556–3565. Lee J-Y, Boman A, Chuanxin S, Andersson M, Jörnvall H, Mutt V, Boman H: Antibacterial peptides from pig intestine: Isolation of a mammalian cecropin. Proc Natl Acad Sci USA 1989;86: 9159–9162. Pattersson-Delafield J, Martinez R, Lehrer R: Microbicidal cationic proteins in rabbit alveolar macrophages: A potential host defense mechanism. Infect Immun 1980;30:180–192. Zasloff M: Maganins, a class of antimicrobial peptides from Xenopus skin: Isolation, characterization of two active forms, and partial cDNA sequence of a precursor. Proc Natl Acad Sci USA 1987; 84:5449–5453. Bettelheim K, Taylor J: A study of Escherichia coli isolated from chronic urinary infections. J Med Microbiol 1969;2:225–236. Grüneberg R: Relationship of infecting urinary organisms to the faecal flora in patients with symptomatic urinary infections. Lancet 1969;ii:766–768.

Dr. Hugh Connell, Department of Medical Microbiology, Section for Clinical Immunology, Sölvegatan 23, Lund University, S–22 362 Lund (Sweden)

Connell/Sabharwal/Persson/Zasloff/Svanborg

124

PIAM:L01:BIF01022XA.79

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 125–132

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Virulence Profile of Uropathogenic Escherichia coli in Patients with Nonobstructive Chronic Pyelonephritis Reinhard Fünfstück a, Niels Jacobsohn a, Helmut Tschäpe b, Günter Stein a a b

Klinik für Innere Medizin IV, Friedrich-Schiller-Universität Jena; Robert-Koch-Institut, Aussenstelle Wernigerode, Germany

Urinary tract infections are among the most common infectious diseases. For the development of an infection, pathogenicity and virulence of the infective agent are of importance as well as the efficacy of local and systemic immunological and nonimmunological defense mechanisms of the human organism. The interplay of these factors characterizes the clinical picture and determines the course of the disease. For colonization, invasion and infection, pathogenic germs must get adapted to host-specific defense reactions. Microorganisms are able to express virulence factors to different degrees. Investigations on the virulence properties of uropathogenic microorganisms seem particularly justified in the case of Escherichia coli. These bacteria are most frequently responsible for urinary tract infections [1, 2]. Several reports concerning the virulence of E. coli pathogenic to the urinary tract demonstrate their ability to produce hemolysin and hydroxamate as well as mannose-resistant or mannose-sensitive hemagglutination (fimbriae/adhesin capacity) and the occurrence of specific O- or K-antigens [3–6]. We investigated the four most important virulence properties in E. coli strains of patients with chronic nonobstructive pyelonephritis and evaluated the frequency of these virulence factors over a period of 3 years.

PIAM:L01:BIF01022XA.79

FF: W1FP

E1: IN01

$$$

Patients and Methods Patients In 53 women (mean age 42 B 12 years) with chronic nonobstructive pyelonephritis, E. coli strains were examined. The patients had been attending our outpatient department for more than 3 years. The diagnosis was established on the basis of clinical history as well as clinical, laboratory and radiological findings (renal scarring, caliceal clubbing and blunting). In all patients, a vesicoureteral reflux, an obstruction due to concrements, or a metabolic disorder (diabetes mellitus, hyperuricemia) were ruled out. In no case was urinary tract infection associated with glomerulonephritis or a gynecological disease involved. No immunocompromised host was included. All patients suffering from an acute urinay tract infection were examined microbiologically. Subsequent investigations of the patients showed no clinical or biochemical signs of an active infection. In all cases, significant bacteriuria (1 105 colonies/ml urine) was found in midstream urine samples. Methods Determination of Disease Activity. The inflammatory parameters erythrocyte sedimentation rate, leukocytes, ·2-globulin fraction, C-reactive protein, leukocyturia and erythrocyturia were determined. Microbiological Investigations. A total of 144 E. coli strains were studied over a period of 3 years. On average, 2.8 B 1.6 germ analyses were carried out per patient. Bacteria: The bacterial examination included bacterial count, species identification, and antibiotic susceptibility determination according to Edwards and Ewing [7]. Hemolysin: The hemolytic activity of the bacterial strains was determined as described by Springer and Goebel [8]. The amount of hemoglobin released was identified spectrophotometrically at 420 nm and served as a quantitative measure of the hemolytic activity of the strains. Fimbriae (mannose-resistant hemagglutination): Hemagglutination was tested using a slide agglutination test [9]. Washed red blood cells (group A) obtained from man, cattle and guinea pig were used in a 5% solution, and a drop of 0.1 M mannose was added at a ratio of 1:4. The strain to be studied was incubated for 24 h on CFA agar. Several colonies were suspended in a drop of a red blood cell suspension without adding D-mannose. The evidence of P-fimbriae and of type F 7 to 14-fimbriae was recorded in cases where agglutination of human group A erythrocytes without agglutination with erythrocytes from cattle or guinea pig occurred, often in the presence of D-mannose. Hydroxamate/aerobactin: The production of aerobactin was measured according to Stuart et al. [10] and Wittig et al. [11], E. coli K 12 strain LG 1622 serving as the indicator strain. This bioassay is 1,000 times more sensitive than chemical analysis. K 1-antigen: The detection of K 1-antigen was carried out by means of K 1-specific phages [12]. Confluent or semiconfluent lysis by K 1-phages was regarded as documenting K 1-antigen. Statistical Analysis. After calculation of frequency distribution for statistical testing, the exact two-tailed Fisher test was performed [13]. In order to test the dependences between qualitative features, Pearson’s ¯2 method and methods of logistic regression and discrimination were employed [14].

Fünfstück/Jacobsohn/Tschäpe/Stein

126

PIAM:L01:BIF01022XA.79

FF: W1FP

E1: IN01

$$$

Table 1. Characteristics of patient group (n = 53, females) Criteria

Acute episode of pyelonephritis

Duration of UTI, years

11.9B9.5

Symptoms Dysuria Pollakisuria Flank pain Rise in temperature (1 38 ° C)

+ + + +

Laboratory findings Blood sedimentation rate, mm Leukocytes, Gpt/l ·2-Globulin C-reactive protein (pos.) Creatinine, Ìmol/l Leukocyturia Erythrocyturia Bacteriuria

19.7B5.8 10.8B1.9 0.12B0.03 100% 104.7B20.1 53 (100%) 12 (22%) 53 (100%)

Asymptomatic stage

– – – – 10.4B9.4 6.4B1.9 0.08B0.02 8% 98.2B12.7 6 (11%) – 53 (100%)

Results Characteristics of the Patients In the first stage of investigation the patients were suffering an acute episode of pyelonephritis indicated by symptoms as dysuria, pollakisuria, flank pain and rise in temperature 1 38 ° C. The laboratory findings are presented in table 1. During the following microbiological investigations in all patients no signs of a clinically active disease were found. All subjects had been treated with chemotherapeutics for 7 (to 10) days. The following drugs were employed: ampicillin, gentamicin, sulfamerazine/trimethroprim, ciprofloxacin or ofloxacine. Microorganisms Incidence of virulence properties in the acute episode: The virulence factors of E. coli strains were demonstrable to varying degrees. Most frequently hemolysin formation (n = 26/49%), mannose-resistant hemagglutination/fimbriae (n = 26/ 49%) and the ability to produce hydroxamate/aerobactin (n = 25/47%) were detected. The expression of K 1-antigen (n = 9/17%) was seen less frequently. The frequency distribution is displayed in table 2.

Virulence Factor of E. coli in Patients with Pyelonephritis

127

PIAM:L01:BIF01022XA.79

FF: W1FP

E1: IN01

$$$

Fig. 1. Distribution of virulence properties of 144 E. coli strains in 53 patients with chronic nonobstructive pyelonephritis. Columns: A = hydroxamate (aerobactin); B = fimbriae (mannose-resistant hemagglutination); C = hemolysin; D = K 1-antigen.

Table 2. Incidence of virulence properties of E. coli in 53 patients with an acute episode of chronic nonobstructive pyelonephritis Virulence property

Incidence

Hemolysin formation Fimbriae (mannose-resistant hemagglutination) Hydroxamate/aerobactin K 1-antigen

26 (49%) 26 (49%) 25 (47%) 9 (17%)

Distribution of virulence properties over a period of 3 years. During the period of 3 years, in all patients, 144 E. coli strains were analyzed. The properties of hydroxamate/aerobactin (n = 47/33%) and mannose-resistant hemagglutination (n = 44/31%) were detected most frequently. The ability to form hemolysin (n = 38/27%) and the expression of K 1-antigen (n = 12/8%) were found in a minority of the bacterial strains. The distribution of all virulence markers in the 144 uropathogenic microorganisms is presented in figure 1. Between the ability of hemolysin formation and the property of mannoseresistant hemagglutination (evidence of fimbriae) on the one hand and the ability of hydroxamate production and K 1-antigen on the other hand, there is a statistically relationship (probability of error: 5%). The abilities of both hemolysin formation and mannose-resistant hemagglutination were found in isolates from 19 patients (36%). The combination of hydroxamate production and K 1-antigen occurred in microorganisms from 8 patients (15%). In long-term analysis throughout the observation period of 3 years, the decreasing virulence of bacterial

Fünfstück/Jacobsohn/Tschäpe/Stein

128

PIAM:L01:BIF01022XA.79

FF: W1FP

E1: IN01

$$$

Fig. 2. Incidence of virulence properties of 144 E. coli strains in 53 patient with nonobstructive pyelonephritis; results of a long-term follow-up over 3 years. Bands: A = mannoseresistant hemagglutination; B = hydroxamate; C = hemolysin; D = K 1-antigen.

strains in the course of their persistence in the urogenital tract was striking. The frequency of virulence properties of E. coli in these patients with nonobstructive pyelonephritis is summarized in figure 2. In particular, the loss of hemolysin production was statistically significant (p = 0.05). Upon increasing the observation period, strains with two or more virulence markers became rare.

Discussion A bacterial infection involves the ability of the microorganism to overcome the multiple host defense mechanisms. A pathogen must be able to survive in the host environment, to attach to, and to multiply on the body surface, to resist the defense mechanisms, and to produce a toxin or interfere with host physiology in several other ways [5, 6, 16]. But the development and the course of a disease certainly depends on complex immunological and nonimmunological host defense mechanisms. Primary events determine whether a bacterial strain is able to colonize the urinary tract. Microorganisms may damage host cells and maintain chronic infections without clinical symptoms, for instance, as asymptomatic bacteriuria [16, 17]. It is of clinical interest to determine what microbial property is responsible for urinary tract infection, i.e. occurs most frequently. By monitoring a group of

Virulence Factor of E. coli in Patients with Pyelonephritis

129

PIAM:L01:BIF01022XA.79

FF: W1FP

E1: IN01

$$$

patients over a period of 3 years, an attempt was made to harmonize host conditions. In all cases, vesicoureteral reflux, obstruction, metabolic disorders, gynecological diseases, glomerulonephritis, or an immunocompromised situation were ruled out. The first microbial investigation was performed during an acute episode of the infection, followed by analyses during asymptomatic bacteriuria. A total of 144 E. coli strains were studied. The possible uropathogenic virulence properties include O- and K-antigens, adherence to epithelial cells by different kinds of fimbriae, the ability of hemolysin formation, and hydroxamate (aerobactin) production [18–20]. It is insufficient to characterize uropathogenic bacterial strain by one virulence property alone; only the combination of several factors is helpful. In this study, the Oantigens of the E. coli strain were not determined. The O-antigen classification is necessary to characterize a germ and to differentiate between a relapse or a reinfection, especially in an investigation covering a long period of time. The E. coli strains emanating from an acute epidose of pyelonephritis most frequently showed mannose-resistant hemagglutination and hemolysin formation. For the development of a urinary tract infection, the ability of the bacteria to adhere to the epithelial cells is of great importance. Fimbrial antigens are responsible for adhesion of the microorganisms [16, 21]. In contrast to Löffler and Svanborg-Eden [22], and Elo [23], who described this property in 90% of the E. coli strains from patients with pyelonephritis, our investigations demonstrated the ability of mannose-resistant hemagglutination in 49% of the acute episodes and in 31% of cases with asymptomatic bacteriuria. Hemolysis was more frequent in strains isolated from acute urinary tract infection than in cases with asymptomatic disease (table 2). With a prolonged observation period, a loss of the ability to produce hemolysin was statistically significant (p = 0.05). This property appears to be a prerequisite for bacterial invasion into a host and for the initiation of an acute infection [24, 25]. Other virulence factors are more important for microorganisms to survive in the host. In the present analysis, this property was observed in 47% of the strains from acute disease and in 32% of all uropathogenic E. coli isolated during the 3-year follow-up period. The plasmids determining the production of hydroxamate/aerobactin additionally encode microbial resistance against serum complement factors [26]. On the whole, these properties enable the bacterial strains to survive in the host. K 1-antigen represents an important virulence factor enabling bacteria to become protected against the serum complement system [6, 16]. Our observation on the significance of the K 1-antigen confirms findings of Ørskov et al. [27]. The present results showed a statistically significant association with regard to the concomitant occurence of the properties of hydroxamate production and K 1-antigen (probability of error 5%). Based on our investigations, the virulence property mannose-resistant hemagglutination (fimbriae) appears to allow the microorganisms to adhere to uroepi-

Fünfstück/Jacobsohn/Tschäpe/Stein

130

PIAM:L01:BIF01022XA.79

FF: W1FP

E1: IN01

$$$

thelial cells and survive in a host. Hemolysin formation enables damage to the uroepithelium and initiation of acute infections. Fimbriae and hydroxamate (aerobactin) production enable E. coli to maintain asymptomatic urinary tract infections.

Acknowledgment The bacterial examinations (bacterial count, differentiation and resistance analysis) were performed at the Institute of Medical Microbiology of the University of Jena.

References 1 2

3 4 5 6 7 8 9 10 11

12

13 14 15

16 17

Ritzerfeld W: Zur Resistenzlage von pathogenen Mikroorganismen isoliert im 1. Quartal 1980 aus Urinproben. Urologie [B] 1980;20:222–224. Voigt T, Baier JF, Blechschmidt K: Erreger von Harnwegsinfektionen und deren Empfindlichkeit gegenüber Chemotherapeutika von ambulant und stationär betreuten Patienten des Gesundheitswesens WISMUT in Jahre 1994. Z Klin Med 1986;41:1711–1713. Brauner A, Jacobson SH, Kühn I: Urinary Escherichia coli causing recurrent infections. A prospective follow-up of biochemical phenotypes. Clin Nephrol 1992;38:318–325. Hacker J: Mechanism and methods for analysing pathogenicity. Swiss Biotech 1987;5:21–31. Stapleton A, Moseley S, Stamm WE: Urovirulence determinants in Escherichia coli isolates causing first-episode and recurrent in women. J Infect Dis 1991;163:773–779. Sloot N, Hacker J, Kreft B, Marre R: Uropathogenität von Escherichia coli. Chemother J 1992;1: 104–110. Edwards RP, Ewing WH: Identification of Enterobacteriaceae. Minneapolis, Progress Publishing Co, 1962, pp 1–269. Springer W, Goebel W: Synthesis and secretion of hemolysin by Escherichia coli. J Bacteriol 1980; 144:53–59. Evans DJ, Evans DG, Young LS, Pit I: Hemagglutination typing of E. coli: Definition of serum hemagglutination types. J Clin Microbiol 1980;12:235–242. Stuart SJ, Greenwood KR, Luke RKJ: Iron-suppressible production of hydroxamate by E. coli isolates. Infect Immun 1982;36:870–873. Wittig W, Prager R, Tietze E, Seltmann G, Tschäpe H: Aerobactin-positive Escherichia coli as causative agents of extraintestinal infections among animals. Arch Exp Vet Med 1988;42:211– 219. Nimmich W, Naumann G, Budde E, Straube E: K-Antigen, Adhärenzfaktor, Dulcitol-Abbau und Hämolysinbildung bei E. coli-R-Stämmen aus Urin. Zentralbl Bakteriol Mikrobiol Hyg [A] 1980; 247:35–42. Dixon WJ: BMDP Statistical Software Manual. Berkeley, University of California Press, 1990, vol 1. Brosius G: SPSS/PC+: Advanced Statistics and Tables. Hamburg, McGraw-Hill, 1989. Lomberg H, Hellström M, Jodal U, Leffler H, Lincoln K, Svanborg-Eden C: Virulence-associated traits in E. coli causing first and recurrent episodes of urinary tract infections in children with or without vesicourethral reflux. J Infect Dis 1984;150:561–569. Johnson JR: Virulence factors in Escherichia coli urinary tract infection. Clin Microbiol Rev 1991; 4:80–128. Trifillis AL, Donnenberg MS, Cui X, Russel GR, Utsalo SJ, Mobley HLT, Warren JW: Binding to and killing of human renal epithelial cells by hemolytic P-fimbriated E. coli. Kidney Int 1994;46: 1083–1091.

Virulence Factor of E. coli in Patients with Pyelonephritis

131

PIAM:L01:BIF01022XA.79

18

19 20 21 22

23 24

25

26 27

FF: W1FP

E1: IN01

$$$

Väisänen-Rhen V, Elo J, Väisänen E, Siitonen A, Orskov I, Orskov F, Svenson SB, Mäkela PH, Korhonen TK: P-fimbriated clones among uropathogenic Escherichia coli strains. Infect Immun 1984;43:149–155. Nimmich W, Zingler G, Falkenhagen U, Straube E, Raitner Y: Häufigkeit verschiedener E. coli O:K1:H-Stämme bei Patienten mit Harnwegsinfektionen. Z Klin Med 1986;41:1759–1761. Fünfstück R, Tschäpe H, Stein G, Kunath H, Bergner M, Wessel G: Virulence properties of Escherichia coli strains in patients with chronic pyelonephritis. Infection 1986;14:145–150. Mobley HLT, Island MD, Massad G: Virulence determinants of uropathogenic Escherichia coli and Proteus mirabilis. Kidney Int 1994;46(suppl 47):129–136. Löffler H, Svanborg-Eden C: Glycolipid receptors for uropathogenic E. coli binding to human erythrocytes and ureopithelial cells. Globoseries glycolipid versus other receptors. Infect Immun 1981;34:920–929. Elo J: Mikrobiologische Aspekte in der Pathogenese der Pyelonephritis; in Stein G, Fünfstück R (eds): Harnwegsinfektion. Frankfurt/Main, PMI-Verlag, 1991, pp 11–13. Donnenberg MS, Newman B, Utsalo SJ, Trifillis AL, Hebel JR, Warren JW: Internalization of Escherichia coli into human kidney epithelial cells: Comparison of fecal and pyelonephritis-associated strains. J Infect Dis 1994;169:831–838. O’Hanley P, Lalonde G, Ji G: Alpha-hemolysin contributes to the pathogenicity of piliated digalactoside Escherichia coli in the kidney: Efficacy of an alpha-hemolysin vaccine in preventing renal injury in the BALB/c mouse model of pyelonephritis. Infect Immun 1991;59:1153–1161. Tschäpe H, Prager R: Genetische Ursachen einiger möglicher Pathogenitätsfaktoren von Harnwegsstämmen (E. coli). Z Urol Nephrol 1984;78:407–413. Ørskov F, Ørskov I, Jann B, Jann K: Immunoelectrophoretic patterns of extracts from all Escherichia coli O- and K-antigen test strains: Correlation with pathogenicity. Acta Pathol Microbiol Immunol Scand [B] 1971;79:142–145.

Prof. Dr. Reinhard Fünfstück, Klinik für Innere Medizin IV, Friedrich-Schiller-Universität Jena, Erlanger Allee 101, D–07740 Jena (Germany)

Fünfstück/Jacobsohn/Tschäpe/Stein

132

PIAM:L01:BIF01023XA.90

FF: W1FP

E1: IN01

$$$

((W1: Bund + Kopf: 5 Cic.; Z1: Bund 5 Cic. / Kopf 6 Cic.))

7.4.1997

Bergan T (ed): Urinary Tract Infections. Infectiology. Basel, Karger, 1997, vol 1, pp 133–137

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Urinary Tract Infection: Some Research Priorities Allan R. Ronald, Stephen E. Sanche Section of Infectious Diseases, University of Manitoba, Winnipeg, Man., Canada

Well-executed studies over the past 30 years have made it possible for tens of millions of patients each year with urinary infection to be managed with reduced morbidity [1] and minimal mortality [2]. In addition, laboratory and imaging modalities have been evaluated for the diagnosis of urinary infection and their limitations are now reasonably well defined. As a result, clinical guidelines with appropriate algorithms for the investigation and treatment of patients with urinary infection have been developed and should be in widespread clinical use. Implementation of such guidelines results in lower direct costs of care and is thought to improve outcomes. However, successful clinical research for patients with acute urinary infection syndromes appears to have diverted scientific interest away from other aspects of urinary infection. As a result, we have documented a significant research waning with fewer individuals publishing fewer original contributions during the past decade [1]. At present, about half of the manuscripts published are concerned primarily with ‘comparative’ drug studies in patients with acute urinary tract infections. In almost every instance, the newer treatment regimens are only ‘equivalent’ to already established regimens and little, if any, scientific advance occurs as a result of most treatment trials. In this article, we identify several areas in which well-designed clinical trials and more basic fundamental research could possibly lead to cost-effective improvements in the management of urinary tract infections.

PIAM:L01:BIF01023XA.90

FF: W1FP

E1: IN01

$$$

Understanding Pathogenesis as a Strategy to Improve Care The brilliant studies of Svanborg and her colleagues [3] in Sweden have begun to unravel components of host response to bacterial colonization of the epithelial mucosa. The mucosal cytokine response appears to elicit pyuria and produce symptoms in the urinary tract. These investigators have also designed an organism which efficiently colonizes the urinary bladder without provoking an inflammatory response in most patients. Perhaps ‘designer organisms’ can be an alternate to antimicrobial prophylaxis and suppression in patients prone to recurrence. Chronic infection with these non-symptom-provoking organisms may be preferable to ongoing chronic treatment. In addition, well-defined urovirulence studies have characterized many of the microbial features of Escherichia coli that create a virulent pathogen [4]. Most other urinary tract pathogens have not been as well characterized and further studies are needed. Is a vaccine directed against fimbrial protein a possible strategy to prevent invasive upper tract infection [5]? How far can we proceed with animal studies including primates and when are human vaccine studies necessary? How do we proceed over the next five years to further understand the interventions that may make a much more significant difference in preventing urinary infections?

Preventing Urinary Infection During the past three years, this field of endeavor has seen several major advances which present opportunities for future research. The use of cranberry juice to prevent symptoms, pyuria and infection in elderly women deserves further studies for confirmation and operational research issues [6]. Should cranberry, or perhaps blueberry juice, be offered to all elderly women routinely? What is the role of this form of diet therapy in individuals with indwelling catheters or in young women with recurrent urinary infection? The observation that a biologic basis may exist for this long held ‘folk remedy’ is most intriguing. Studies by Raz and Stamm [7] demonstrating the value of intravaginal estrogens also need confirmation. What studies are needed before instituting this regimen? Are there any downsides to the use of this therapy in postmenopausal women with recurrent infection? Should they be investigated first to prove that they have estrogen deficiency? Will oral estrogens work as well? These and other issues are needed to expand our ability to prevent infections and reduce the need for long-term prophylaxis. Breakthrough infections continue to occur in patients on multi-year courses of trimethoprim/sulfamethoxazole or the quinolones. As a result, the efficacy of

Ronald/Sanche

134

PIAM:L01:BIF01023XA.90

FF: W1FP

E1: IN01

$$$

these regimens may gradually diminish. We also have limited data about the impact of these regimens on global intestinal microbial flora. Alternate strategies to prevent infections need to be a priority and appear to be possible. We also appear to have reached a plateau in our understanding of catheterrelated infections. We now can prevent most infections in patients with catheters in place for two weeks or less. However, in patients requiring long-term catheter drainage, strategies to prevent symptomatic infections, invasive pyelonephritis and sepsis are still not available. Can we do anything useful in this patient population? What are the alternatives to long-term catheter care? Will modification of the catheter or its care alter the risk of untoward events [8]? What is the significance of foul-smelling of turbid urine? How often should catheters be changed to prevent complications? Are organisms that split urea more likely to lead to complications? Again, it seems the questions are obvious, but no answers are available, despite the fact that we have numerous opportunities for study among the patients we care for within our clinics.

Complicated Urinary Infection Very few natural history and treatment studies have been carried out in patients with complicated urinary infection [9]. Despite the presence of large numbers of such patients in most of our clinical practices, we have generally classified these patients as one poorly defined entity and continue to use empiric, unproven treatment regimens. The International Reflux Study in Children, the largest surgical trial ever undertaken, has shown in children that primary vesicoureteric reflux does not require surgical treatment in order to prevent further renal damage and scarring [10]. Unfortunately, randomized trials appear to have limited impact on practice [11]. Winberg [11] states ‘It is sad to observe. In spite of the lessons learned from the study, new methods are introduced without controlled studies.’ As individuals committed to academic priorities within all of our areas of responsibility, we must insist that new tactics in the management of urinary infection, whether medical or surgical, are subject to careful study prior to widespread implementation.

Urinary Infection in Men The postulated incidence of urinary infection in men is 3/1,000 from the age of 20 to 50 years and it appears to increase thereafter. The health cost of managing these infections in the industrialized world probably exceeds USD 500 million annually. In spite of this major health impact, no population-based studies using

Urinary Tract Infection Research

135

PIAM:L01:BIF01023XA.90

FF: W1FP

E1: IN01

$$$

large well-characterized patient groups have determined the male incidence of urinary infection or identified the prevalence of underlying or concomitant disease or other risk factors. As a result, management strategies are based on subsets of men usually selected for treatment trials. A series of questions remain unanswered. We have taught that all urinary infections in men are categorized as ‘complicated’. What is the evidence of this classification? What routine investigations are cost-effective and necessary? When should imaging or cystoscopy be ordered? What regimens are most effective and what is an appropriate treatment duration? How do we define a ‘cure’ in men with urinary infection? Is a prostatic massage necessary following a treatment course in order to be confident of a cure? When should asymptomatic infection be treated in men? Carefully designed clinical trials can increase our knowledge and lead to improved care of men with urinary infection.

Delivery of Care Issues in Patients with Urinary Infection Few studies have addressed issues related to patient or provider knowledge, attitudes, or practice as they relate to the prevention or management of urinary tract infections. We have no information as to how physicians make antimicrobial choices, how they determine duration of therapy or how effectively they identify risk factors or prevent subsequent infections. Variation in the management of urinary tract infections is immense. Studies defining this variation, and its impact on outcome, are necessary. Planned interventions within communities could also be important strategies to alter practice, involve patients in decision-making, allow the patient choice among antibacterial regimens, and determine the value of additional investigations including urine culture, urinalysis, and imaging studies. These would all be of both academic and practical interest. Managed care organizations are collecting large amounts of data on patients with urinary tract infections. These studies may define less expensive strategies that can be used both within these organizations as well as elsewhere. If the annual global direct health cost of urinary infections is over USD 5 billion, substantial dollars could become available for other health care interventions if we could lower the ‘direct costs’ of managing urinary infection. Well-designed studies to address these questions require multidisciplinary investigators with expertise in epidemiology, economics and clinical medicine. In summary, urinary infection research is being carried out by only a few of the individuals who are involved in the care of these patients globally. We need to organize ourselves more effectively to identify the questions, encourage investigators to pursue improved experimental design, lobby for support, and find the answers. This is one of the goals of the Clinical Evaluation of Drug Efficacy in

Ronald/Sanche

136

PIAM:L01:BIF01023XA.90

FF: W1FP

E1: IN01

$$$

Urinary Tract Infection group. The meetings of this group in Montreal at the International Congress of Chemotherapy again addressed some of the opportunities and challenges of urinary infection. It is our intent that additional studies will emerge during the next two years that will further enlarge our areas of scientific competence as we manage patients with urinary infection.

References 1 2 3 4 5

6 7 8 9 10

11

Harding GKM, Ronald AR: The management of urinary infections: What have we learned in the past decade? Int J Antimicrob Agents 1994;4:83–88. Lipsky BA: Urinary tract infections in men: Epidemiology, pathophysiology, diagnosis, and treatment. Ann Intern Med 1989;110:138–150. Hedges S, Svanborg C: The mucosal cytokine response to urinary tract infections. Int J Antimicrob Agents 1994;4:89–93. Johnson JR: Virulence factors in Escherichia coli urinary tract infection. Clin Microbiol Rev 1991; 291:80–128. O’Hanley P, Lark D, Falkow S, Schoolnik G: Molecular basis of Escherichia coli colonization of the upper respiratory tract in BALB/c mice: Gal-gal pili immunization prevents Escherichia coli pyelonephritis in the BALB/c mouse model of human pyelonephritis. J Clin Invest 1985;75:347–360. Avorn J, Monane M, Gurwitz JH, Glynn RJ, Choodnovskiy I, Lipsitz LA: Reduction of bacteriuria and pyuria after ingestion of cranberry juice. JAMA 1994;271:751–754. Raz R, Stamm WE: A controlled trial of intravaginal estriol in postmenopausal women with recurrent urinary tract infections. N Engl J Med 1993;329:753–756. Kunin CM: Can we build a better urinary catheter? N Engl J Med 1988;319:365–366. Ronald AR, Patullo ALS: The natural history of urinary infection in adults. Med Clin North Am 1991;75:299–312. Smellie JM, Tamminen-Mobius T, Olbing H, Claesson T, Wikstad I, Jodal U, Seppanen U, on behalf of the International Reflux Study in Children (European Branch): Five-year study of medical or surgical treatment in children with severe reflux: Radiological renal findings. Pediatr Nephrol 1992;6:223–230. Olbing H, Tamminen-Mobius T, Jodal U, Smellie J: Correspondencee Re: J. Winberg: Management of primary vesico-ureteric reflux in children. Operation ineffective in preventing progress of renal damage (Infection 1994;22:S4–S7). Infection 1995;23:248–249.

Prof. Allan R. Ronald, Section of Infectious Diseases, C5124 St. Boniface General Hospital, 409 Tache Avenue, Winnipeg, Manitoba R2H 2A6 (Canada)

Urinary Tract Infection Research

137

1

PIAM:L01:ZIF01001AI.94

FF: W1PP

E1: 01X1

TY: B

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Author Index

Agace, W. 109 Bailey, R.R. 14 Benson, M. 109 Borleffs, J.C.C. 37 Connell, H. 109, 118 Dooyeweert, D.A. van 37 Fünfstück, R. 125 Hedges, S. 109 Hedlund, M. 109 Hoepelman, A.I.M. 37 Jacobsohn, N. 125 Jodal, U. 109 Jones, R.N. 67 Kanimoto, Y. 84 Kawada, Y. 84 Kotoh, S. 52 Krcˇméry, S. 48 Kumazawa, J. 19, 52 Ludwig, E. 57 Ludwig, M. 60 Matsumoto, T. 19, 52

Naber, K.G. 74 Nickel, J.C. 89 Nicolle, L.E. 8 Persson, L. 118 Roberts, J.A. 98 Ronald, A.R. 133 Rubin, R.H. 27, 34 Sabharwal, H. 118 Sakumoto, M. 52 Sanche, S.E. 133 Schiefer, H.-G. 60 Schneider, M.M.E. 37 Stamm, W.E. 1, 46 Stein, G. 125 Svanborg, C. 109, 118 Svensson, M. 109 Tolkoff-Rubin, N.E. 27, 34 Tschäpe, H. 125 Weidner, W. 60, 106 Zasloff, M. 118