Pancreatic Cancer

This page intentionally left blank

Pancreatic Cancer

Contemporary Issues in Cancer Imaging A Multidisciplinary Approach

Series Editor Rodney H. Reznek Cancer Imaging, St Bartholomew’s Hospital, London

Editorial Adviser Janet E. Husband Diagnostic Radiology, Royal Marsden Hospital, Surrey

Current titles in the series Cancer of the Ovary Lung Cancer Colorectal Cancer Carcinoma of the Kidney Carcinoma of the Esophagus Carcinoma of the Bladder Squamous Cell Cancer of the Neck Prostate Cancer Interventional Radiological Treatment of Liver Tumors Forthcoming titles in the series Gastric Cancer Primary Carcinomas of the Liver Breast Cancer

Pancreatic Cancer Edited by

Jay Heiken Series Editor

Rodney H. Reznek Editorial Adviser

Janet E. Husband

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521886925 © Cambridge University Press 2009 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2008

ISBN-13

978-0-511-46396-9

eBook (EBL)

ISBN-13

978-0-521-88692-5

hardback

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Every effort has been made in preparing this publication to provide accurate and up-to-date information which is in accord with accepted standards and practice at the time of publication. Although case histories are drawn from actual cases, every effort has been made to disguise the identities of the individuals involved. Nevertheless, the authors, editors and publishers can make no warranties that the information contained herein is totally free from error, not least because clinical standards are constantly changing through research and regulation. The authors, editors and publishers therefore disclaim all liability for direct or consequential damages resulting from the use of material contained in this publication. Readers are strongly advised to pay careful attention to information provided by the manufacturer of any drugs or equipment that they plan to use.

Contents

Series Foreword Preface to Pancreatic Cancer Contributors

page ix x xi

1. Epidemiology and genetics of pancreatic cancer Srinivasa K. R. Prasad and Rong Zeng

1

2. Pathology of pancreatic neoplasms Thomas C. Smyrk

10

3. Multi-detector row computed tomography (MDCT) techniques for imaging pancreatic neoplasms Alec J. Megibow

28

4. Magnetic resonance imaging (MRI) techniques for evaluating pancreatic neoplasms Shawyon Shadman and Vamsi R. Narra

46

5. Imaging evaluation of pancreatic ductal adenocarcinoma Steven S. Raman and David S. K. Lu

58

6. Imaging evaluation of cystic pancreatic neoplasms Dushyant Sahani and Onofrio A. Catalano

83

7. Imaging evaluation of pancreatic neuroendocrine neoplasms Ruedi F. Thoeni

104

8. Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms Rajesh N. Keswani and Riad R. Azar

130

vii

viii

Contents

9. Surgical staging and management of pancreatic adenocarcinoma Steven M. Strasberg

150

10. Treatment of locally advanced and metastatic pancreatic cancer Benjamin R. Tan, Jeffrey Carenza, and Joel Picus

166

11. Rare pancreatic neoplasms and mimics of pancreatic cancer Naoki Takahashi

Index The plate section can be found between pages 84 and 85.

175 193

Series Foreword

Imaging has become pivotal in all aspects of the management of patients with cancer. At the same time it is acknowledged that optimal patient care is best achieved by a multidisciplinary team approach. The explosion of technological developments in imaging over the past years has meant that all members of the multidisciplinary team should understand the potential applications, limitations and advantages of all the evolving and exciting imaging techniques. Equally, to understand the significance of the imaging findings and to contribute actively to management decisions and to the development of new clinical applications for imaging, it is critical that the radiologist should have sufficient background knowledge of different tumors. Thus the radiologist should understand the pathology, the clinical background, the therapeutic options and prognostic indicators of malignancy. Contemporary Issues in Cancer Imaging – A Multidisciplinary Approach aims to meet the growing requirement for radiologists to have detailed knowledge of the individual tumors in which they are involved in making management decisions. A series of single subject issues, each of which will be dedicated to a single tumor site, edited by recognized expert guest editors, will include contributions from basic scientists, pathologists, surgeons, oncologists, radiologists and others. While the series is written predominantly for the radiologist, it is hoped that individual issues will contain sufficient varied information so as to be of interest to all medical disciplines and to other health professionals managing patients with cancer. As with imaging, advances have occurred in all these disciplines related to cancer management and it is our fervent hope that this series, bringing together expertise from such a range of related specialties, will not only promote the understanding and rational application of modern imaging but will also help to achieve the ultimate goal of improving outcomes of patients with cancer. Rodney H. Reznek

London ix

Preface to Pancreatic Cancer

Our ability to diagnose and stage pancreatic adenocarcinoma, cystic pancreatic neoplasms and pancreatic neuroendocrine neoplasms continues to improve, owing to advances in ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI). Consequently, imaging has become an increasingly critical component in the clinical management of patients with pancreatic neoplasms. Accurate staging is important to determine appropriate treatment and to minimize the number of patients who undergo unnecessary laparotomy. In addition, much has been learned recently about the epidemiology and genetics of pancreatic neoplasms, which may lead to novel approaches to the prevention, diagnosis, and treatment of pancreatic cancer. As with other malignancies, a multidisciplinary team approach is essential to optimizing patient care and providing the most appropriate treatment. This issue of Contemporary Issues in Cancer Imaging provides a detailed review, not only of the imaging evaluation of pancreatic neoplasms, but also the epidemiology, genetics, pathology, and clinical management of these tumors. Separate chapters focus on the surgical staging and management of pancreatic cancers and the treatment of locally advanced and metastatic disease. Thus, although this volume is directed primarily at radiologists, it also will be of considerable value to other medical specialists involved in the care of patients with pancreatic cancer. Jay Heiken

x

Contributors

Riad R. Azar Department of Medicine Division of Gastroenterology Washington University School of Medicine Jeffrey Carenza Washington University School of Medicine St. Louis, Missouri USA Onofrio A. Catalano Department of Radiology Massachusetts General Hospital Boston, MA Rajesh N. Keswani Department of Medicine Division of Gastroenterology Washington University School of Medicine USA David S. K. Lu Division of Abdominal Imaging and Cross Sectional Interventional Radiology Department of Radiology David Geffen School of Medicine UCLA Medical Center Los Angeles, CA USA

Alec J. Megibow Department of Radiology New York University Medical Center New York, NY USA Vamsi R. Narra Mallinckrodt Institute of Radiology Washington University School of Medicine St. Louis Missouri, USA Joel Picus Washington University School of Medicine St. Louis, Missouri USA Srinivasa K. R. Prasad Department of Radiology University of Texas Health Science Center San Antonio, TX USA

xi

xii

Contributors

Steven S. Raman Division of Abdominal Imaging and Cross Sectional Interventional Radiology Department of Radiology David Geffen School of Medicine UCLA Medical Center Los Angeles, CA USA Dushyant Sahani Department of Radiology Massachusetts General Hospital Boston, MA USA Shawyon Shadman Mallinckrodt Institute of Radiology Washington University School of Medicine St. Louis Missouri, USA Thomas C. Smyrk Mayo Clinic Scottsdale, AZ USA Steven M. Strasberg Section of Hepatobiliary-Pancreatic and Gastrointestinal Surgery, Department of Surgery Washington University in Saint Louis Saint Louis Missouri USA

Naoki Takahashi Mayo Clinic Scottsdale, AZ USA Benjamin R. Tan Washington University School of Medicine St. Louis, Missouri USA Ruedi F. Thoeni Chief of Abdominal Imaging, SFGH University of California, San Francisco USA Rong Zeng Department of Radiology University of Texas Health Science Center San Antonio, TX USA

1 Epidemiology and genetics of pancreatic cancer Srinivasa K. R. Prasad and Rong Zeng

Introduction Pancreatic ductal adenocarcinoma (and its histological variants), also referred to as pancreatic cancer (PC) comprises 90% of exocrine pancreatic neoplasms [1]. This highly aggressive cancer is the fourth leading cause of cancer death in the USA [2]. More than 80% of patients with PC present with advanced disease that is incurable by surgery. Most tumors greater than 5 cm in size show disseminated metastases at presentation [3]. The 5-year survival rate of advanced PC is poor (< 5%) with a median survival of < 6 months. The 5-year survival rate improves to 20–30% in patients who harbor small, early invasive cancers (usually < 3 cm) and are candidates for surgical resection [2]. Thus, early diagnosis of PC before frank invasion occurs is critical to improve patient outcomes. Mucinous cystic neoplasms (MCNs) are mucin-secreting, cystic neoplasms of characteristic histopathology and variable clinico-biological profiles. They comprise 10–45% of cystic pancreatic neoplasms [4]. Intraductal papillary mucinous neoplasms (IPMNs) are characterized by cystic dilatation of ducts and intraductal papillary tumors with variable mucin production and tumor histobiology. Intraductal papillary mucinous neoplasms constitute 15–25% of cystic pancreatic neoplasms and typically show slow intra-luminal growth and low metastatic potential [4, 5]. Subsets of these two mucinous tumors progress to PC.

Pancreatic cancer Epidemiology

Pancreatic cancer is one of the most lethal cancers, characterized by invasive growth and rapid dissemination despite a relatively well-differentiated histomorphology [3]. In the USA, PC contributed to 32 300 estimated deaths in 2006 of an estimated Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

2

Pancreatic Cancer

33 730 new pancreatic cancer cases [6]. Eighty percent of PC manifest clinically in patients aged 60–80 years; only about 10% of patients are below the age of 50 years [7]. Pancreatic cancer is found more commonly in men. Predisposing risk factors include cigarette smoking, chronic pancreatitis, exposure to radiation and chemicals, diabetes mellitus and hereditary cancer syndromes [7]. Smoking is by far the most important risk factor accounting for approximately 25% of PCs [7]. Smokers show a two-fold increased risk for PC compared with nonsmokers. Chronic pancreatitis is a well-documented risk factor for PC [7, 8]. It is postulated that high cell turnover associated with chronic pancreatitis (either acquired or hereditary) predisposes to PC especially in the setting of defective DNA repair mechanisms [7]. It is interesting to note that KRAS and p16 mutations seen with PC are also observed in patients with chronic pancreatitis [9, 10]. In addition to the above risk factors, microscopic and macroscopic precursors of PC have been identified.

Microscopic precursors of pancreatic cancer It is now established that invasive PCs originate from microscopic, non-invasive neoplastic epithelial proliferations referred to as pancreatic intra-epithelial neoplasias (PanINs) [2, 11]. The PanINs develop in small-caliber pancreatic ducts (usually < 5 mm in diameter) and are histologically classified into three types, PanINs-1, 2 and 3 [12]. The histological spectrum of PanINs is summarized in Table 1.1. These microscopic ductal epithelial growths show histogenetic progression from low-grade PanIN (PanIN–1 to PanIN-2) through high-grade PanIN (PanIN-3) to PC analogous to the adenoma–carcinoma sequence in colorectal cancers [2]. The tumor progression is thought to occur through a series of sequential, polychromosomal genetic mutations [11, 13]. Early, intermediate and late genetic events of progression of PanINs to PC include KRAS activation, inactivation of the p16/CDKN2A tumor suppressor gene and the loss-of-function mutations of TP53 and DPC4/SMAD4 tumor suppressor genes respectively (Figure 1.1) [2].

Macroscopic precursors of pancreatic cancer Intraductal papillary mucinous neoplasms

Intraductal papillary mucinous neoplasms (IPMNs) are characterized by predominant intraductal growth, papillary epithelial configuration, mucin overproduction and variable histobiologic profiles [14]. Peak age of incidence is in the 7th and 8th decades. Men are more commonly affected. Most patients manifest abdominal pain,

Epidemiology and genetics of pancreatic cancer

Table 1.1. Histological spectrum of pancreatic intraepithelial neoplasias (PanINs) [2,12] Lesion type

Histological grade

Histological characteristics

Genetic pathways involved

Flat, micropapillary or papillary epithelial KRAS activation lesions composed of tall columnar cells. Minimal degree of atypia PanIN-2 Intermediate Mostly papillary epithelial lesions with Inactivation of the p16 grade some nuclear abnormalities and rare tumor suppressor gene mitoses. Moderate degree of atypia Inactivation of the *TP53 PanIN-3 High grade “Carcinoma-in-situ”; Papillary, and *DPC4 tumor micropapillary epithelial growths with suppressor genes cytonuclear abnormalities resembling non-invasive carcinoma (No basement membrane invasion) PanIN-1 Low grade

*TP53: Tumor Protein 53 *DPC4: Deleted in Pancreatic Cancer Locus 4

Normal Ductal Epithelium Telomere Shortening

KRAS (Activation) PanIN-1 p16 (Inactivation) PanIN-2

DPC4 (Inactivation)

TP53 (Inactivation) PanIN-3 HER-2/neu (Activation)

Invasive Pancreatic Cancer Figure 1.1. Step-wise cytogenetic progression of normal ductal epithelium through PanINs to invasive pancreatic cancer [Refs. 2, 25]

3

4

Pancreatic Cancer

weight loss, jaundice, chronic exocrine insufficiency and diabetes mellitus [4]. Patients with IPMN are at an increased risk of developing extra-pancreatic cancers, particularly colonic, gastric and lung malignancy [15, 16]. Intraductal papillary mucinous neoplasms may involve the main pancreatic duct, branch ducts or both. They are histologically classified into benign, borderline and malignant categories [1]. Branch duct tumors commonly show less aggressive pathologic changes compared with the main duct subtype [17]. In a large surgical series, 37% of main duct IPMNs developed invasive tumors, whereas only 15% of branch duct IPMNs harbored carcinoma in-situ changes [17]. A recent taxonomic schema of IPMNs identifies four different types of lining epithelium – gastric, pancreaticobiliary, intestinal and oncocytic [18]. It is postulated that different types of IPMN show distinct histogenetic pathways of progression. Most benign, branch-duct IPMNs show gastric type of epithelium, whereas IPMNs with intestinal and pancreaticobiliary epithelium show progression to ‘colloid’ carcinomas and PCs respectively [2, 19]. As with PCs, activating mutations in the KRAS oncogene constitutes an early event in pathogenesis of IPMNs [20]. However, there are some important differences between the cytogenetics of IPMNs and PCs. Loss of DPC4 protein is a characteristic finding associated with 55% of invasive PCs, whereas IPMNs are characterized by universal expression of DPC4 protein [21]. Also, inactivation of the Peutz–Jeghers gene STK11/LKB1 (a phenomenon that is rarely seen with PC) has recently been reported in about 30% of IPMNs [22, 23]. Mucinous cystic neoplasms

Mucinous cystic neoplasms (MCNs) occur almost exclusively in perimenopausal women in the 5th or 6th decades with a 9:1 female preponderance [1]. Mucinous cystic neoplasms comprise 10–45% of cystic pancreatic neoplasms. Most tumors are symptomatic on presentation and occur preferentially (> 90% of cases) in the body and tail of the pancreas [2, 5]. Mucinous cystic neoplasm histologically consists of tall mucin-producing epithelia that are supported by a characteristic ovarian-type stroma that expresses estrogen and progesterone receptors [1]. Microscopically, MCN can be categorized into three types: benign (cystadenoma), borderline (MCN with moderate dysplasia) and malignant (mucinous cystadenocarcinomas) [1]. Mucinous cystic neoplasms demonstrate a histogenetic progression model, akin to PanINs. Mutation of the KRAS oncogene is an early event being seen with 20%, 33% and 89% of adenomatous MCNs, borderline MCNs and MCNs with

Epidemiology and genetics of pancreatic cancer

Table 1.2. Summary of genetic alterations in sporadic pancreatic cancer [3, 25] Gene Oncogenes KRAS *HER-2/neu Tumor suppressor genes p16

TP53 DPC4

Mechanism of alteration

% of PC

Point mutation Overexpression

> 90 70 (invasive PC)

Homozygous deletion Loss of heterozygosity Promoter hypermethylation Loss of heterozygosity, mutation Homozygous deletion Loss of heterozygosity

40 40 15 50–80 35 20

*HER-2/neu: HER: Homolog of human epidermal growth factor receptor; neu: derived from murine neuroglioblastoma cell line.

carcinoma-in-situ respectively. Inactivation of p53 and DPC4 is seen in invasive MCNs only [24]. In summary, MCNs and IPMNs (with intestinal and pancreaticobiliary epithelium) are considered precursor lesions of PC (and its variants). The genetic changes that underlie the transformation of MCN and IPMN to PC are to some extent similar to the genetic changes responsible for PanIN to PC transformation.

Cytogenetics of pancreatic cancer Recent advances in cytogenetics and molecular biology have provided unique insights into the pathogenesis of PC. Pancreatic cancer is now thought to be a byproduct of genetic events that may be classified as promotion of proto-oncogenes (KRAS), suppression of tumor suppressor genes (p16, p53, DPC4, DUSP6), or alterations in growth factors (FGF, TGF-β and EGFR) [25]. Table 1.2 lists the known cytogenetics in sporadic PC. Hereditary pancreatic cancer syndromes

Up to 10% of PCs may show familial clustering [26]. Pancreatic cancer may be associated with several systemic hereditary cancer syndromes such as Peutz–Jeghers syndrome, hereditary breast–ovarian cancer syndrome, ataxia-telangiectasia,

5

6

Pancreatic Cancer

familial atypical multiple mole melanoma syndrome, familial adenomatous polyposis syndrome and Lynch syndrome [1]. Hereditary pancreatitis and cystic fibrosis are also associated with increased predisposition to PC [7]. Familial PC syndrome is a genetically and phenotypically heterogenous, autosomal dominant, hereditary disorder characterized by an increased risk of PC. BRCA2 germline mutations have been the most common, isolated genetic abnormality in these patients [27]. Molecular and genetic abnormalities involved in many hereditary syndromes have been characterized [1]. Different syndromes show characteristic genetic abnormality with variable penetrance and phenotypic risk of PC. Table 1.3 summarizes the cytogenetics and the risk of PC in genetic syndromes with predisposition to PC. Patients with hereditary predisposition to PC tend to develop multicentric cancers at an early age (40–50 years) [26]. They also are at increased risk of developing cutaneous and other systemic cancers. Although screening techniques have been suboptimal and have not been standardized, several approaches to early diagnosis and prophylactic treatment currently are being employed [26]. It is hoped that a thorough knowledge of hereditary PC syndromes may lead to routine screening of high-risk individuals, improved early diagnosis and better cancer control measures. Implications of cytogenetics on diagnosis and management

Pancreatic cancer is one of the most difficult cancers to treat due to late diagnosis, aggressive tumor biology and poor response to existing systemic therapy. Better understanding of the molecular mechanisms of the development and progression of PCs has allowed for the emergence of novel diagnostic and therapeutic paradigms. Genetic analysis of pancreatic fluid is a promising aid for accurate and early diagnosis of PC. Mesothelin, a protein that is over-expressed in 55% of PCs potentially can be used as a diagnostic marker of PC [2]. In addition, immunoliposomes may be used to selectively target mesothelin-bearing cancer cells [28]. Point mutations of the KRAS gene (found in over 90% of PCs) can be detected on analysis of pancreatic fluid [29]. Tissue-based proteomics may be used to define the “signature” of the PC proteome so that treatment may be tailored [30]. Specific, molecularly targeted drugs and novel gene therapy techniques are being developed to improve patient survival.

Conclusions Pancreatic cancer arises from precursor lesions (PanIN, MCN, IPMN), is associated with KRAS mutations (early changes), p53 mutations (late changes), and

Epidemiology and genetics of pancreatic cancer

Table 1.3. Genetic syndromes with inherited predisposition to pancreatic cancer [1, 26]

Hereditary pancreatic cancer Implicated syndromes gene Familial PC syndrome Hereditary pancreatitis Early-onset familial PC / Diabetes syndrome (Seattle family) Peutz–Jeghers syndrome

Gene location

BRCA-2? 13q12 Cationic 7q35 trypsinogen Unknown NA

STK11/LK B1 19p

Systemic syndrome Chronic pancreatitis Chronic pancreatitis

GI tract hamartomatous polyps and perioral pigmentation Hereditary breast– BRCA-2 13q12 Hereditary breast ovarian cancer syndrome and ovarian cancers MSH2, MLH1 11q22–23 Colorectal cancers Hereditary and other non-polyposis colorectal extracolonic cancer (HNPCC) cancers syndrome Familial adenomatous APC 5q21 Adenomatous polyposis syndrome polyps of the GI tract, colorectal cancers Familial atypical multiple p16 9p Melanomas mole melanoma (FAMMM) syndrome Ataxia-telangiectasia ATM, ATB 11q22 Ataxiatelangiectasia

Lifetime risk of ductal adenocarcinoma > 50% 30% 30%; high risk of pancreatitis and diabetes mellitus 36%

5–10%

< 5%

< 5%

10%

< 5%

shows an aggressive biological behavior with resultant poor prognosis. Cytogenetic abnormalities that are associated with step-wise, morphological progression of PC from precursor lesions have been elucidated. Several hereditary pancreatic syndromes and their underlying genetic and molecular abnormalities have been

7

8

Pancreatic Cancer

characterized. It is hoped that knowledge obtained from ongoing and future research into genetic and molecular mechanisms of PC will provide new perspectives that may lead to early diagnosis and better patient outcomes.

REFERENCES 1. DeLellis RA, Heitz PU, Eng C. World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of Endocrine Organs. Lyon: IARC Press, 2004. 2. Singh M, Maitra A. Precursor lesions of pancreatic cancer: molecular pathology and clinical implications. Pancreatology 2007; 7: 9–19. 3. Adsay NV, Basturk O, Cheng JD, Andea AA. Ductal neoplasia of the pancreas: nosologic, clinicopathologic, and biologic aspects. Semin Radiat Oncol 2005; 15: 254–264. 4. Brugge WR, Lauwers GY, Sahani D, Fernandez-del Castillo C, Warshaw AL. Cystic neoplasms of the pancreas. N Engl J Med 2004; 351: 1218–1226. 5. Sahani D, Prasad S, Saini S, Mueller P. Cystic pancreatic neoplasms evaluation by CT and magnetic resonance cholangiopancreatography. Gastrointest Endosc Clin N Am 2002; 12: 657–672. 6. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. Cancer J Clin 2006; 56: 106–130. 7. Lowenfels AB, Maisonneuve P. Epidemiology and risk factors for pancreatic cancer. Best Pract Res Clin Gastroenterol 2006; 20: 197–209. 8. Lowenfels AB, Maisonneuve P, Cavallini G, et al. Pancreatitis and the risk of pancreatic cancer. International Pancreatitis Study Group. N Engl J Med 1993; 328: 1433–1437. 9. Lohr M, Kloppel G, Maisonneuve P, Lowenfels AB, Luttges J. Frequency of K-ras mutations in pancreatic intraductal neoplasias associated with pancreatic ductal adenocarcinoma and chronic pancreatitis: a meta-analysis. Neoplasia 2005; 7: 17–23. 10. Rosty C, Geradts J, Sato N, et al. p16 Inactivation in pancreatic intraepithelial neoplasias (PanINs) arising in patients with chronic pancreatitis. Am J Surg Pathol 2003; 27: 1495–1501. 11. Hruban RH, Adsay NV, Albores-Saavedra J, et al. Pancreatic intraepithelial neoplasia: a new nomenclature and classification system for pancreatic duct lesions. Am J Surg Pathol 2001; 25: 579–586. 12. Hruban RH, Takaori K, Klimstra DS, et al. An illustrated consensus on the classification of pancreatic intraepithelial neoplasia and intraductal papillary mucinous neoplasms. Am J Surg Pathol 2004; 28: 977–987. 13. Furukawa T, Sunamura M, Horii A. Molecular mechanisms of pancreatic carcinogenesis. Cancer Sci 2006; 97: 1–7. 14. Prasad SR, Sahani D, Nasser S, et al. Intraductal papillary mucinous tumors of the pancreas. Abdom Imaging 2003; 28: 357–365.

Epidemiology and genetics of pancreatic cancer

15. Kamisawa T, Tu Y, Egawa N, et al. Malignancies associated with intraductal papillary mucinous neoplasm of the pancreas. World J Gastroenterol 2005; 11: 5688–5690. 16. Choi MG, Kim SW, Han SS, Jang JY, Park YH. High incidence of extrapancreatic neoplasms in patients with intraductal papillary mucinous neoplasms. Arch Surg 2006; 141: 51–56; discussion 56. 17. Terris B, Ponsot P, Paye F, et al. Intraductal papillary mucinous tumors of the pancreas confined to secondary ducts show less aggressive pathologic features as compared with those involving the main pancreatic duct. Am J Surg Pathol 2000; 24: 1372–1377. 18. Furukawa T, Kloppel G, Adsay NV, et al. Classification of types of intraductal papillarymucinous neoplasm of the pancreas: a consensus study. Virchows Arch 2005; 447: 794–799. 19. Adsay NV, Merati K, Basturk O, et al. Pathologically and biologically distinct types of epithelium in intraductal papillary mucinous neoplasms: delineation of an “intestinal” pathway of carcinogenesis in the pancreas. Am J Surg Pathol 2004; 28: 839–848. 20. Kondo H, Sugano K, Fukayama N, et al. Detection of K-ras gene mutations at codon 12 in the pancreatic juice of patients with intraductal papillary mucinous tumors of the pancreas. Cancer 1997; 79: 900–905. 21. Iacobuzio-Donahue CA, Klimstra DS, Adsay NV, et al. Dpc-4 protein is expressed in virtually all human intraductal papillary mucinous neoplasms of the pancreas: comparison with conventional ductal adenocarcinomas. Am J Pathol 2000; 157: 755–761. 22. Sato N, Rosty C, Jansen M, et al. STK11/LKB1 Peutz–Jeghers gene inactivation in intraductal papillary-mucinous neoplasms of the pancreas. Am J Pathol 2001; 159: 2017–2022. 23. Sahin F, Maitra A, Argani P, et al. Loss of Stk11/Lkb1 expression in pancreatic and biliary neoplasms. Mod Pathol 2003; 16: 686–691. 24. Jimenez RE, Warshaw AL, Z’Graggen K, et al. Sequential accumulation of K-ras mutations and p53 overexpression in the progression of pancreatic mucinous cystic neoplasms to malignancy. Ann Surg 1999; 230: 501–509; discussion 509–511. 25. Maitra A, Kern SE, Hruban RH. Molecular pathogenesis of pancreatic cancer. Best Pract Res Clin Gastroenterol 2006; 20: 211–226. 26. Hruban RH, Canto MI, Griffin C, et al. Treatment of familial pancreatic cancer and its precursors. Curr Treat Options Gastroenterol 2005; 8: 365–375. 27. Murphy KM, Brune KA, Griffin C, et al. Evaluation of candidate genes MAP2K4, MADH4, ACVR1B, and BRCA2 in familial pancreatic cancer: deleterious BRCA2 mutations in 17%. Cancer Res 2002; 62: 3789–3793. 28. Mamot C, Drummond DC, Noble CO, et al. Epidermal growth factor receptor-targeted immunoliposomes significantly enhance the efficacy of multiple anticancer drugs in vivo. Cancer Res 2005; 65: 11631–11638. 29. Wilentz RE, Chung CH, Sturm PD, et al. K-ras mutations in the duodenal fluid of patients with pancreatic carcinoma. Cancer 1998; 82: 96–103. 30. Chen R, Yi EC, Donohoe S, et al. Pancreatic cancer proteome: the proteins that underlie invasion, metastasis, and immunologic escape. Gastroenterology 2005; 129: 1187–1197.

9

2 Pathology of pancreatic neoplasms Thomas C. Smyrk

The term “pancreatic cancer” usually refers to ductal adenocarcinoma. While this entity accounts for 85% of primary pancreatic tumors, a variety of other neoplasms can arise from the range of cell types present in the normal pancreas (ducts, acini and islets) (Table 2.1) [1].

Ductal adenocarcinoma Approximately two-thirds of ductal adenocarcinomas are found in the head of the pancreas. The tumor tends to be very firm and ill-defined (Figure 2.1) Cystic change (usually due to tumor necrosis) can occur but is rare [2]. This is an aggressively infiltrating cancer with a propensity for direct invasion into distal common bile duct, ampulla of Vater, duodenum, blood vessels, nerves and extrapancreatic soft tissue, particularly posterior to the pancreas. Ductal adenocarcinoma is a gland-forming tumor. The glands tend to be round or only slightly angulated, giving the tumor a deceptively indolent appearance (Figure 2.2). There is almost always a dense stromal response to the tumor; this “desmoplastic stroma” of myofibroblasts and collagen can make it difficult to pick out the malignant cells. The glands may be lined by a single layer of cuboidal to columnar epithelium or may show complex papillary growth. Tumor nuclei usually vary in size and shape; in fact, the single most helpful diagnostic finding in a biopsy or cytologic preparation is a 4:1 variation in nuclear size within a single gland (Figure 2.3). Other helpful findings include an infiltrative, haphazard arrangement of the glands (as opposed to the lobular arrangement of normal pancreas), glands growing next to muscular arteries (a feature not part of normal pancreatic architecture) and the presence of foamy cytoplasm in the neoplastic cells [3]. Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

Pathology of pancreatic neoplasms

Table 2.1. Tumors of the pancreas Ductal adenocarcinoma Mucinous non-cystic carcinoma Signet ring cell carcinoma Adenosquamous carcinoma Undifferentiated (anaplastic or sarcomatoid) carcinoma Undifferentiated carcinoma with osteoclast-like giant cells Intraductal papillary mucinous neoplasm Non-invasive, epithelium graded as adenoma (low-grade dysplasia), borderline (moderate dysplasia) or carcinoma in-situ (high-grade dysplasia) Invasive carcinoma arising in intraductal papillary mucinous neoplasm Specify ductal carcinoma or mucinous non-cystic carcinoma Mucinous cystic neoplasm Non-invasive, epithelium graded as low-grade, moderate or high-grade dysplasia Invasive carcinoma arising in mucinous cystic neoplasm Acinar cell carcinoma Acinar cell cystadenocarcinoma Mixed acinar-endocrine carcinoma Solid pseudopapillary tumor Pancreatic endocrine neoplasm Serous cystadenoma Serous cystadenocarcinoma

Tumor grade correlates with patient outcome [4]. Grade is based on the extent and quality of gland formation. Tumors that form round, regular glands are well-differentiated, while those that form glands showing more variability in size and shape are moderately differentiated and those without appreciable gland formation are poorly differentiated. The tendency for ductal cancers to have heterogenous growth patterns makes this grading scheme difficult to apply at times; the general rule is to assign grade based on the worst growth pattern. In addition to the common form of ductal adenocarcinoma, there are several less common histologic subtypes. Mucinous non-cystic carcinoma (colloid carcinoma) accounts for 1–3% of pancreatic malignancies [5, 6]. Extracellular mucin accounting for at least 50% of tumor volume is the definitive diagnostic finding. These tumors have an intestinal phenotype, characterized by immunohistochemical positivity for MUC-2 and CDX2 (in contrast to usual ductal adenocarcinoma, which typically expresses MUC-1 and is negative for MUC-2 and CDX2). Colloid carcinoma is associated

11

12

Pancreatic Cancer

Figure 2.1. Ductal adenocarcinoma involving the body and tail of the pancreas. Tumor infiltrates diffusely into pancreatic parenchyma and extends into peripancreatic soft tissue (inferior). This figure is reproduced in the color plate section.

Figure 2.2. Ductal adenocarcinoma of the pancreas. Glands of irregular size and shape are lined by cells with variably sized nuclei. (Original magnification 66 ×). This figure is reproduced in the color plate section.

Pathology of pancreatic neoplasms

Figure 2.3. Fine needle aspiration of ductal adenocarcinoma. Three dimensional group of cells with marked variability in nuclear size and shape. (Pap stain, 132 ×). This figure is reproduced in the color plate section.

with intraductal papillary mucinous carcinoma (IPMN); at least half of colloid carcinomas arise in a background of IPMN. There is some evidence that colloid carcinoma has a better prognosis than ductal carcinoma; Adsay et al. found 55% five-year survival in a series of 17 patients with colloid carcinoma larger than 1 cm, despite the fact that eight of the patients had lymph node metastases at the time of diagnosis [5]. Adenosquamous carcinomas comprise 3–4% of pancreatic carcinomas [7]. At least one-third of the tumor should show squamous differentiation in order to qualify. This variant resembles usual ductal carcinoma in terms of demographics, gross appearance, site distribution, and outcome. Signet ring cell carcinoma of the pancreas is extremely rare. Signet ring cells in metastatic tumor are much more likely to be derived from colorectal, gastric or breast primaries than from the pancreas [8]. The prognosis has been uniformly poor for the handful of reported cases. The category of undifferentiated anaplastic carcinoma includes both giant cell carcinoma and sarcomatoid carcinoma [9, 10]. One percent of pancreatic cancers

13

14

Pancreatic Cancer

Figure 2.4. Anaplastic carcinoma of the pancreas. Discohesive cells with very large, bizarrely shaped nuclei. (132 ×). This figure is reproduced in the color plate section.

have anaplastic histology (Figure 2.4) [1]. The tumor cells are very large and have bizarrely shaped nuclei. The cells are uncohesive and are often accompanied by neutrophil inflammation. Sarcomatoid carcinoma is composed predominantly or exclusively of spindle cells, but is believed to arise from epithelial precursors. The spindle cell component may give rise to heterologous elements such as bone, cartilage or striated muscle. Undifferentiated anaplastic and sarcomatoid carcinomas appear to be even more aggressive than usual ductal carcinoma. Undifferentiated carcinoma with osteoclast-like giant cells features a mixture of epithelioid tumor cells and non-neoplastic giant cells (Figure 2.5) [11]. The giant cells are derived from histiocytes, just as in giant cell tumors of other organs. The neoplastic cells are thought to be ductally derived, despite the fact that they generally fail to show immunohistochemical evidence of epithelial differentiation. The prognosis for this rare subtype is not settled; early reports of favorable prognosis have not been borne out. Medullary carcinoma is a rare subtype showing solid growth, abundant cytoplasm and round nuclei with a prominent nucleus (Figure 2.6) [12]. Many small lymphocytes typically accompany the tumor cells. The appearance is similar to that seen in medullary carcinoma of the colon [13], and indeed the two tumors share a common molecular pathogenesis in the form of microsatellite instability. When patients with Hereditary Non-polyposis Colorectal Cancer develop pancreatic

Pathology of pancreatic neoplasms

Figure 2.5. Osteoclast-like giant cell tumor. Multinucleated giant cells mixed with epithelioid tumor cells. (132 ×). This figure is reproduced in the color plate section.

Figure 2.6. Medullary carcinoma. Solid sheets of tumor cells with many tumor-infiltrating lymphocytes. (66 ×). This figure is reproduced in the color plate section.

15

16

Pancreatic Cancer

cancer, it will sometimes have medullary phenotype. The prognosis for medullary carcinoma of the pancreas is probably better than that of ductal adenocarcinoma, but case numbers are small.

Precursors to ductal carcinoma: pancreatic intraepithelial neoplasia Morphologic and genetic evidence points to the existence of precursor lesions to ductal adenocarcinoma. The morphologic continuum is called “Pancreatic Intraepithelial Neoplasia (PanIN),” and is subdivided on the basis of architectural and cytologic features into PanIN-1A, 1B, 2 and 3 (Figure 2.7). PanIn-1A, descriptively termed mucinous hyperplasia, is a common incidental finding, while PanIN-3 is found almost exclusively in association with carcinoma. The genetic abnormalities found in ductal adenocarcinoma can be found in PanIN with increasing frequency as the grade of

(a)

(b)

Figure 2.7a. PanIN 1A. Tall mucinous cells with regular, basally-oriented nuclei. b: PanIN2. The nuclei are crowded and elongated. c: PanIN 3. Complex papillary architecture and large, dark, irregular nuclei. (66 ×). This figure is reproduced in the color plate section.

(c)

Pathology of pancreatic neoplasms

PanIN increases [14]. Rare examples of PanIN progression in individual patients have been reported, further supporting the precursor status of this lesion [15]. Unfortunately, the frequency and tempo of progression remain unknown, and clinical detection of PanIN remains difficult, limiting the practical value of PanIn.

Intraductal papillary mucinous neoplasm (IPMN) Papillary tumors arising in pancreatic ducts have been a recognized subset of pancreatic neoplasia for 20 years, described as mucinous duct ectasia, papillary carcinoma, mucin-producing tumor and villous adenoma of the pancreatic duct [16]. Intraductal papillary mucinous neoplasm usually involves the head of the pancreas (75%); up to 10% show involvement of the entire gland (Figure 2.8). Some tumors involve only the main duct, others affect both main and branch ducts, and some branch ducts only. Branch duct IPMN can be difficult to separate from other cystic lesions of the pancreas on clinical grounds, because connection with the duct system, a characteristic feature of IPMN, is hard to demonstrate in branch duct-type tumors. There is some evidence that branch-duct IPMN has better outcome than those involving main ducts [17]. As the name implies, IPMN often produces abundant mucin. In fact, mucin spillage from the ampulla of Vater is a classic finding in IPMN [18]. Ampullary biopsies taken in this setting will often show abnormal papillary epithelium, allowing morphologic confirmation of the diagnosis.

Figure 2.8. Intraductal mucinous papillary neoplasm involving most of the main pancreatic duct with extension in side branches. This figure is reproduced in the color plate section.

17

18

Pancreatic Cancer

The gross appearance of IMPN ranges from well-defined cystic lesion to diffuse, irregular dilation of the pancreatic duct system. The cyst lining may be smooth or thick and velvety, depending on the amount of papillary growth. Microscopically, the cyst lining ranges from bland mucinous epithelium to complex papillae with marked cytologic atypia. This range of morphologies is often seen in different areas of the same tumor, emphasizing the need for extensive sampling. The current WHO system subdivides the epithelial changes as “adenoma,” “borderline” and “carcinoma in-situ” [1]. These categories correspond roughly to the PanIN levels 1, 2 and 3, and can also be classified as low-grade, moderate or high-grade dysplasia. The epithelium of IPMN can also be subdivided according to its morphologic and immunohistochemical resemblance to intestinal, gastric or pancreatobiliary epithelium (Figure 2.9). Rarely, the epithelium is composed of cells with abundant pink

(a)

(c)

(b)

(d)

Figure 2.9. Histologic subtypes of IPMN. a: gastric-type mucosa with mucous glands. b: Intestinal type, lined by tall columnar epithelium. c: Pancreatobiliary type, with cuboidal cells. d: Oncocytic type, featuring abundant pink cytoplasm. This figure is reproduced in the color plate section.

Pathology of pancreatic neoplasms

cytoplasm, giving rise to the term intraductal oncocytic papillary neoplasm [19]. The histologic subtypes have specific mucin profiles, and there may be some differences in behavior: gastric-type IPMN is less likely to undergo malignant transformation than the intestinal or pancreatobiliary types, and when malignant transformation does occur, gastric and pancreaticobiliary epithelia give rise to ductal-type adenocarcinoma while intestinal epithelium gives rise to colloid carcinoma [20]. About one-third of IPMN described in the literature have had an invasive component [21]. Two histologic types of carcinoma arise in this setting: one resembling ductal adenocarcinoma and one producing abundant extracellular mucin (colloid carcinoma). Ductal adenocarcinoma arising in IPMN has a prognosis similar to that of usual ductal adenocarcinoma. Colloid carcinoma appears to have a somewhat better prognosis [5], although this is not the case in all studies [6].

Mucinous cystic neoplasm (MCN) Mucinous cystic neoplasm has many morphologic similarities to IPMN; in fact, the two entities were often jumbled together in papers published prior to 1995. The critical – indeed defining – histologic feature for MCN is a spindle cell stroma beneath the mucinous epithelial lining (Figure 2.10). The stroma is called “ovarian-like” because

Figure 2.10. Mucinous cystic neoplasm. A single layer of columnar epithelium above ovarian-like stroma. (66 ×). This figure is reproduced in the color plate section.

19

20

Pancreatic Cancer

it resembles the stroma of that organ, even to the point of expressing estrogen and progesterone receptors and containing scattered epithelioid cells that express inhibin [22]. When properly defined, MCN is almost exclusively a tumor of women, arising in the body or tail of the pancreas and not connected to the pancreatic duct system [23]. Mucinous cystic neoplasm usually has a round outer contour and is surrounded by a thick fibrous capsule. It may be unilocular or multilocular. Each locule is lined by columnar mucinous epithelium; the lining is usually flat, but may have papillary architecture. As with IPMN, the epithelium may display little or no cytologic atypia or the severe atypia of carcinoma in-situ. Again paralleling IPMN, the WHO recommends the terms “mucinous cystadenoma, mucinous cystic neoplasm, borderline, and mucinous cystadenocarcinoma” for the various levels of atypia [1], but low-grade, moderate and high-grade dysplasia are acceptable descriptors. Regardless of the grade of dysplasia, complete excision of MCN without an invasive component is curative [24]. Invasive carcinoma complicates 5–10% of MCN [23]. Because MCN can be quite large (mean diameter 10 cm) and the invasive component small, extensive sampling is the rule for all MCN. At least one section for each centimeter of cyst is the general guideline, but several authors have argued for more extensive or even complete sampling [24]. Since the invasive component of MCN typically looks different grossly from its benign background (solid, firm, discolored), generous sampling guided by judicious gross examination is probably sufficient. Carcinoma arising in MCN typically has ductal features, but colloid carcinoma, sarcomatoid carcinoma and undifferentiated carcinoma with osteoclastlike giant cells have all been described [25]. The prognosis for invasive carcinoma arising in MCN appears to be better than that of usual ductal adenocarcinoma; one study found 2- and 5-year survival rates of 67% and 31% [26].

Acinar cell carcinoma (ACC) This neoplasm is said to account for 1% to 2% of pancreatic malignancies. Acinar cell carcinoma resembles non-neoplastic pancreatic acinar cells and shows evidence of pancreatic enzyme production. The morphologic differential diagnosis includes endocrine tumor and solid pseudopapillary tumor. Electron micrographic demonstration of zymogen granules is one way to prove the diagnosis. Immunohistochemical markers can also be helpful, with trypsin being the most useful. Acinar cell carcinoma is a solid tumor that can arise in any part of the pancreas. It features nests of cells with uniform cells and round nuclei situated at the basal pole of the cell (Figure 2.11). The cytoplasm is granular and eosinophilic. Rare cystic

Pathology of pancreatic neoplasms

Figure 2.11. Acinar carcinoma. Gland-like spaces lined by cuboidal cells with round, regular nuclei. (132 ×). This figure is reproduced in the color plate section.

variants have been described (acinar cell cystadenocarcinoma) as well as mixtures of ACC with ductal or endocrine elements [27, 28]. The 5-year survival for ACC is given as 6% [29]. There are, however, welldocumented reports of long survival. The rarity of this tumor and difficulty in diagnosing it correctly make definitive pronouncements on prognosis difficult.

Solid pseudopapillary tumor (SPT) (also called solid pseudopapillary epithelial neoplasm – SPEN) Solid pseudopapillary tumor typically affects young females (mean age 27 years) but is well-documented in males [1]. Any part of the pancreas can be affected. As the name implies, these tumors produce solid or partly cystic masses, with the extent of cystic change often increasing with increasing size of tumor. The largest SPTs may become cysts filled with hemorrhage and degenerating tissue. In these situations, the diagnosis will depend on finding the rare islands of viable tissue that cluster along the inner wall of the cyst. Under the microscope, the solid areas show sheets of cells traversed by a rich vascular network (Figure 2.12). The cells are polygonal and often have a single, large eosinophilic globule rich in alpha-

21

22

Pancreatic Cancer

Figure 2.12. Solid pseudopapillary tumor. Sheets of cells traversed by slender capillaries. Many cells have intracytoplasmic globules. (132 ×). This figure is reproduced in the color plate section.

1-antitrypsin in the cytoplasm (as opposed to the smaller, multiple zymogen granules of ACC). This tumor does not form glands. When degenerative changes develop, the cells nearest the small vessels are preserved and the others drop out, producing the “pseudopapillary” pattern. The areas of cell dropout are often replaced by collections of foamy histiocytes. Hemorrhage and cholesterol clefts complete the classic histology of SPT. Clinical, gross and histologic findings usually are sufficient to make the diagnosis, but nuclear staining for beta-catenin is a useful confirmatory study [30]. The cell lineage of SPT is unknown. Immunohistochemical studies using markers of acinar, endocrine and ductal differentiation have produced inconsistent results. The behavior of SPT is similarly unpredictable; about 15% of patients will experience recurrence or metastasis [31]. When present, metastases are restricted to the peritoneum and liver. Histology does not seem to predict the potential for aggressive behavior; most SPT show some degree of infiltration into peripancreatic tissue, but this does not seem to correlate with outcome. The presence of vascular invasion has been linked to potential for aggressive behavior in some studies [32]. Even when metastases are present, affected patients typically have stable disease for many years.

Pathology of pancreatic neoplasms

Pancreatic endocrine neoplasm (PEN) Endocrine tumors account for 3–5% of primary pancreatic neoplasms. They are usually solid, but can undergo cystic degeneration. They lack the desmoplastic stroma of ductal adenocarcinoma, and are therefore much softer. As in endocrine tumors of other sites, there are several growth patterns, including trabecular, solid and pseudoacinar (Figure 2.13). Tumor nuclei are round and regular, with even chromatin. Cytoplasm is lightly eosinophilic, except in the rare clear cell variant. Clear cell change is thought to result from the accumulation of cytoplasmic lipid; this variant is strongly associated with von Hippel–Lindau syndrome [33]. The majority of PEN are low grade, meaning that they lack necrosis and have a low mitotic rate (fewer than 10 mitoses per 10 high power fields). Pancreatic endocrine neoplasm is typically positive for synaptophysin and chromogranin, allowing it to be separated from acinar cell carcinoma and solid pseudopapillary tumor. Antibodies are also available for the immunohistochemical evaluation of the peptides gastrin, insulin, somatostatin, glucagon, pancreatic polypeptide and vasoactive intestinal peptide. These markers are not usually necessary for the diagnosis of PEN, and the assessment of functionality is best based on clinical features and serum levels; still, immunohistochemical study is occasionally requested to document peptide production.

Figure 2.13. Pancreatic neuroendocrine tumor. Bland cells with trabecular and pseudacinar growth. This figure is reproduced in the color plate section.

23

24

Pancreatic Cancer

Predicting behavior for PEN remains a difficult issue. Many morphologically bland PEN will recur or metastasize after resection, whereas some with aggressivelooking histology will not. Features that predict adverse clinical outcome include tumor size greater than 2 cm, vascular invasion, elevated mitotic rate, relatively high labeling index (Ki-67) and DNA aneuploidy. Tumors with a greater number of adverse factors are at higher risk for recurrence, but predicting whether an individual tumor can be considered cured by complete resection is impossible at this time [34]. Rarely, PEN can have high-grade features resembling small cell carcinoma – scant cytoplasm, many mitoses and extensive necrosis. As with small cell carcinoma at any site, these lesions have a very poor outcome.

Serous cystic neoplasms Serous microcystic adenoma accounts for 1–2% of exocrine pancreatic tumors [35]. It affects women more often than men and occurs most frequently in the body or tail of the pancreas. It forms a well-circumscribed mass with a sponge-like cut surface. There are numerous tiny cysts filled with clear, watery fluid. There is often a central fibrous scar (Figure 2.14). The cysts are lined by cuboidal cells with clear cytoplasm and round, regular nuclei. The clear cytoplasm is due to the presence of abundant glycogen, which can be demonstrated by periodic acid-Schiff (PAS) stain

Figure 2.14. Serous microcystic adenoma. Sponge-like cut surface with central fibrosis. This figure is reproduced in the color plate section.

Pathology of pancreatic neoplasms

with and without diastase digestion. Less commonly, this neoplasm may have only a few cystic spaces (serous oligocystic adenoma). There are very few examples of serous cystadenocarcinoma (8 cases out of 2000) [35].

REFERENCES 1. Kloppel G, Hruban RH, Longnecker DS, et al. Ductal adenocarcinoma of the pancreas. In Hamilton SR, Aaltonen LA, eds. Pathology and Genetics of Tumours of the Digestive System, Lyon, France: IARC Press. 2000; 220–230. 2. Kosmahl M, Pauser U, Anlauf M, Kloppel G. Pancreatic ductal adenocarcinomas with cystic features: neither rare nor uniform. Mod Pathol 2005; 18: 1157–1164. 3. Adsay NV, Logani S, Sarkar F, Crissman J, Vaitkevicius V. Foamy gland pattern of pancreatic ductal adenocarcinoma: a deceptively benign-appearing variant. Am J Surg Pathol 2000; 24: 493–504. 4. Adsay NV, Basturk O, Bonnett M, et al. A proposal for a new and more practical grading scheme for pancreatic ductal adenocarcinoma. Am J Surg Pathol 2005; 29: 724–733. 5. Adsay NV, Pierson C, Sarkar F, et al. Colloid (mucinous noncystic) carcinoma of the pancreas. Am J Surg Pathol 2001; 25: 26–42. 6. Seidel G, Zahurak M, Iacobuzio-Donahue C, et al. Almost all infiltrating colloid carcinomas of the pancreas and periampullary region arise from in situ papillary neoplasms: a study of 39 cases. Am J Surg Pathol 2002; 26: 56–63. 7. Ishikawa O, Matsui Y, Aoki I, et al. Adenosquamous carcinoma of the pancreas: a clinicopathologic study and report of three cases. Cancer 1980; 146: 755–761. 8. Tracey KJ, O’Brien MJ, Williams LF, et al. Signet ring carcinoma of the pancreas; a rare variant with very high CEA values. Dig Dis Sci 1984; 29: 573–576. 9. Tschang TP, Garza-Garza R, Kissane JM. Pleomorphic carcinoma of the pancreas: a clinicopathologic study. Cancer 1977; 39: 2114–2126. 10. Hoorens A, Prenzel K, Lemoine NR, Kloppel G. Undifferentiated carcinoma of the pancreas: analysis of intermediate filament profile and K-ras mutations provides evidence of a ductal origin. J Pathol 1998; 185: 53–60. 11. Dworak O, Wittekind C, Koerfgen HP, Gall FP. Osteoclastic giant cell tumor of the pancreas: an immunohistological study and review of the literature. Pathol Res Pract 1993; 189: 228–231. 12. Wilentz RE, Goggins M, Redston M, et al. Genetic, immunohistochemical and clinical features of medullary carcinoma of the pancreas: a newly described and characterized entity. Am J Pathol 2000; 156: 1641–1651. 13. Jessurun J, Romero-Guadarrama M, Manivel JC. Medullary adenocarcinoma of the colon: clinicopathologic study of 11 cases. Hum Pathol 1999; 30: 843–848. 14. Hruban RH, Goggins M, Parsons J, Kern SE. Progression model for pancreatic cancer. Clin Cancer Res 2000; 6: 2969–2972.

25

26

Pancreatic Cancer

15. Brat DJ, Lillemoe KD, Yeo CJ, Warfield PB, Hruban RH. Progression of pancreatic intraductal neoplasias to infiltrating adenocarcinoma of the pancreas. Am J Surg Pathol 1998; 22: 163–169. 16. Adsay NV, Longnecker DS, Klimstra DS. Pancreatic tumors with cystic dilation of the ducts: intraductal papillary mucinous neoplasms and intraductal oncocytic papillary neoplasms. Semin Diagn Pathol 2000; 17: 16–30. 17. Terris B, Ponsot P, Paye F, et al. Intraductal papillary mucinous tumors of the pancreas confined to secondary ducts show less aggressive pathologic features as compared with those involving the main pancreatic duct. Am J Surg Pathol 2000; 24: 1372–1377. 18. Ohta T, Nagakawa T, Akiyama T, et al. The “duct-ectatic” variant of mucinous cystic neoplasm of the pancreas: clinical and radiologic studies of seven cases. Am J Gastroenterol 1992; 87: 300–304. 19. Adsay NV, Adair CF, Jeffess CS, Klimstra DS. Intraductal oncocytic papillary neoplasms of the pancreas. Am J Surg Pathol 1996; 20: 980–994. 20. Adsay NV, Merati K, Basturk O, et al. Pathologically and biologically distinct types of epithelium in intraductal papillary mucinous neoplasms: delineation of an “intestinal” pathway of carcinogenesis in the pancreas. Am J Surg Pathol 2004; 28: 839–848. 21. Chari ST, Yadav D, Smyrk TC, et al. Study of recurrence after surgical resection of intraductal papillary mucinous neoplasm of the pancreas. Gastroenterology 2002; 123: 1500–1507. 22. Zamboni G, Scarpa A, Bogina G, et al. Mucinous cystic tumors of the pancreas: clinicopathological features, prognosis and relationship to other mucinous cystic tumors. Am J Surg Pathol 1999; 23: 410–422. 23. Reddy RP, Smyrk TC, Zapiach M, et al. Pancreatic mucinous cystic neoplasm defined by ovarian stroma: demographics, clinical features and prevalence of cancer. Clin Gastroenterol Hepatol 2004; 2: 1026–1031. 24 Wilentz RE, Albores-Saavedra J, Zahurak M, et al. Pathologic examination accurately predicts prognosis in mucinous cystic neoplasms of the pancreas. Am J Surg Pathol 1999; 23: 1320–1327. 25. Wenig BM, Albores-Saavedra J, Buetow PC, Heffess CS. Pancreatic mucinous cystic neoplasm with sarcomatous stroma: a report of three cases. Am J Surg Pathol 1997; 21: 70–80. 26. Thompson LD, Becker RC, Przygodzki RM, Adair CF, Heffess CS. Mucinous cystic neoplasm of the pancreas: a clinicopathologic study of 130 cases. Am J Surg Pathol 1999; 23: 1–16. 27. Cantrell BB, Cubilla AL, Erlandson RA, Fortner J, Fitzgerald PJ. Acinar cell cystadenocarcinoma of human pancreas. Cancer 1981; 1547: 410–416. 28. Klimstra DS, Rosai J, Heffess CS. Mixed acinar-endocrine carcinomas of the pancreas. Am J Surg Pathol 1994; 18: 765–778. 29. Holen KD, Klimstra DS, Hummer A, et al. Clinical characteristics and outcomes from an institutional series of acinar cell carcinoma of the pancreas and related tumors. J Clin Oncol 2002; 1520: 4673–4678. 30. Abraham SC, Klimstra DS, Wilentz RE, et al. Solid-pseudopapillary tumors of the pancreas are genetically distinct from pancreatic ductal adenocarcinomas and almost always harbor beta-catenin mutations. Am J Pathol 2002; 160: 1361–1369.

Pathology of pancreatic neoplasms

31. Klimstra DS, Wenig BM, Heffess CS. Solid-pseudopapillary tumor of the pancreas: a typically cystic carcinoma of low malignant potential. Semin Diagn Pathol 2000; 17: 81–88. 32. Nishihara K, Nagoshi M, Tsuneyoshi M, Yamaguchi K, Hayashi Y. Papillary cystic tumors of the pancreas. Assessment of their malignant potential. Cancer 1993; 71: 82–92. 33. Hoang MP, Hruban HR, Albores-Saavedra J. Clear cell endocrine pancreatic tumor mimicking renal cell carcinoma: a distinctive neoplasm of von Hippel-Lindau disease. Am J Surg Pathol 2001; 25: 602–609. 34. Heitz PU, Komminoth P, Perren A, et al. Pancreatic endocrine tumours: introduction. In DeLellis RA, Lloyd RV, Heitz PU, Eng C. eds Pathology and Genetics of Tumours of Endocrine Organs. Lyon, France: IARC Press. 2004; 177–182. 35. Capella C, Solcia E, Kloppel G, Hruban RH. Serous cystic neoplasms of the pancreas. In Hamilton SR, Aaltonen LA, eds. Pathology and Genetics of Tumours of the Digestive System. Lyon, France: IARC Press. 2000; 231–233.

27

3 Multi-detector row computed tomography (MDCT) techniques for imaging pancreatic neoplasms Alec J. Megibow

Introduction From the earliest clinical experiences with single slice computed tomography (CT) to today’s multi-detector row CT (MDCT) era, it has been apparent that successful imaging detection of pancreatic neoplasms is significantly improved by the use of pancreas-specific examination protocols [1]. The primary goal in imaging patients suspected of a pancreatic neoplasm is to confidently detect or exclude its presence. If a pancreatic neoplasm is found, the imaging study should provide an accurate assessment of the parameters that will allow appropriate treatment selection for each patient. This chapter examines in detail the specific elements that comprise an MDCT pancreatic protocol. Radiologists reading this chapter will be able to adapt these recommended imaging parameters to their own site. Non-radiologist clinicians will gain a sense of the strengths and limitations of properly performed MDCT scanning for pancreatic disease and will be encouraged to make the radiologist aware of the clinical suspicion such that the patient can be examined with the best possible technique. The pancreatic protocol used at our institution will be reviewed with emphasis on how each element of the protocol improves imaging accuracy. Although the primary focus will be on the evaluation of pancreatic ductal adenocarcinoma, we will also explore specific utility for neuroendocrine tumors and cystic pancreatic neoplasms.

MDCT pancreatic protocol components A well-designed imaging protocol must specify the types of patients for whom the protocol is appropriate, provide the technologist with appropriate acquisition parameters, and take into consideration the radiation dose delivered to the patient. In addition it must specify the type and timing of oral contrast administration, the dose Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

MDCT techniques for imaging pancreatic neoplasms

and rate of intravenous contrast administration, the phases of imaging and appropriate scan timing, and the reconstruction parameters. Finally, the protocol should establish the distribution of images to be used for interpretation and further processing.

Clinical indications Pancreatic imaging protocols are utilized in all patients suspected of pancreatic neoplasm, acute or chronic pancreatitis, and for evaluation of jaundice. Many times, however, we use a pancreatic imaging protocol for patients who are referred to our institution with clinical symptoms such as “weight-loss” “recent onset of diabetes” or “severe epigastric pain.” The clinical radiologist is responding to a “clue” from the referring physician that there is a suspicion of pancreatic disease but the referrer does not wish to alarm the patient.

Acquisition parameters MDCT should be used whenever possible to evaluate suspected pancreatic neoplasms because it enables large volume coverage in short imaging times. MDCT also provides image data sets composed of near isotropic voxels [2], which results in high spatial resolution in all imaging planes. Typical scan duration for a 64-slice MDCT is 7–9 seconds to cover the abdomen and pelvis with sub-millimeter slice thickness. For most MDCT examinations, images are acquired with thin sections but displayed as thicker slices. The radiologist has a choice of acquisition slice thickness, which depends on the number of detector rows of the scanner. Sixteen-detector row scanners offer a detector configuration choice with elements of 0.625–0.75 mm or 1.25–1.5 mm. Images acquired with these detector settings can be reconstructed into voxels of 1 mm or 2 mm respectively. Forty- and 64-detector row scanners offer a choice of 0.6–0.625 mm or 1.2–1.25 mm detector configuration settings producing isotropic voxels of 0.75 mm or 1.5 mm respectively. Optimized pancreatic imaging protocols are based on the narrowest detector configuration, which results in high quality 3D images and frequently allows delineation of the pancreatic duct [3].

Radiation dose Concerns over radiation dose from MDCT have increased in recent years [4]. The effective radiation dose of an abdominal/pelvic MDCT study is approximately

29

30

Pancreatic Cancer

10 milliSieverts (mSv) per acquisition (http://www.radiologyinfo.org/en/safety/ index.cfm?pg=sfty_xray). Ten mSv is the equivalent of 3-years’ exposure to naturally occurring background radiation. With multiphase imaging, radiation dose is multiplied by the number of acquisitions. Several methods to minimize the delivered dose are available. Most scanners allow the user to choose an acceptable “noise factor.” Setting a low reference tube current in milliamperes (reference mA) assures the scanner will deliver the minimum dose necessary to maintain the noise level in the images that the user finds acceptable for diagnosis. Tube current modulation software can reduce delivered dose by a factor of 30%. At our institution we utilize reference mAs of 240, which result in a CT volume dose index (CTDIvol) of approximately 10–12 milliGray (mGy) for an average-sized patient. These parameters result in an effective dose of approximately 7–8 mSv. Patient radiation exposure decreases with increasing numbers of detector rows due to increased x-ray dose efficiency. With 40- and 64-row detector scanners, radiation dose is independent of detector configuration; using 4 or 16 channel scanners, radiation dose to the patient is increased by 50–100% when the narrowest detector configuration is used, and acquisition times increase by as much as 100% [5].

Oral contrast Properly delivered oral contrast material is critical to delineate the bowel loops adjacent to the pancreas and within the abdomen. For pancreatic imaging, we prefer to use a neutral (near water attenuation) contrast agent. Agents such as water [6], low Hounsfield unit barium suspension (VoLumen®, EZ-EM Co., Westbury Long Island) [7], and whole milk [8] have been employed. At our institution, we ask patients to consume approximately 900 ml of VoLumen in divided doses over a 30–40 minute period. In addition, 300 ml of water is given in the scan suite immediately before the patient is positioned for the localizing scan. Intravenous glucagon (0.1 mg) provides bowel hypotonia to facilitate visualization of the duodenal sweep, which aids in evaluating the duodenal papilla. The major benefit of neutral bowel contrast in pancreatic imaging is realized when 3-D processing is incorporated. Positive (high attenuation) contrast agents can interfere with evaluation of peripancreatic vessels. Improvements in visualization of the duodenal papilla improve the specificity of discriminating periampullary duodenal neoplasms from periampullary pancreatic neoplasms (Figure 3.1) [9, 10].

®

MDCT techniques for imaging pancreatic neoplasms

Figure 3.1a. Pancreatic phase MDCT image through the pancreatic head in a patient being evaluated for obstructive jaundice. The main pancreatic duct (PD) and common bile duct (CBD) are distended (double duct sign).

(a)

Figure 3.1b. The use of neutral oral contrast media allows identification of a polypoid ampullary adenocarcinoma (arrow) as the etiology of the ductal obstruction.

(b)

Intravenous contrast-dose Pancreatic imaging by MDCT requires intravenous contrast enhancement. Pancreatic adenocarcinoma elicits an intense desmoplastic response, encasing intra-pancreatic blood vessels explaining why 90% of these masses appear hypoattenuating on CT compared to background pancreas. When pancreatic carcinoma extends beyond the pancreas, it encases local arterial and venous structures. Noncontrast enhanced studies are insufficient for detection of neoplasms (Figure 3.2), evaluation of peripancreatic vasculature or detection of distant metastases. Low osmolality contrast (1.5–2.0 mgI/kg) provides sufficient pancreatic parenchymal and hepatic enhancement to maximize lesion to background attenuation differences

31

32

Pancreatic Cancer

Figure 3.2a. Importance of IV contrast. This non-IV contrast enhanced MDCT image was obtained in a patient suspected of pancreatic neoplasm. Although the head of the pancreas appears irregular, a definitive diagnosis of pancreatic neoplasm was not possible.

(a)

Figure 3.2b. Same patient as in “A” The patient underwent gadolinium enhanced MRI. The small tumor in the pancreatic head is evident. Non-visualization at MDCT is solely due to inability to administer iodinated contrast material.

(b)

[11, 12]. Peak hepatic enhancement and peak pancreatic parenchymal enhancement are directly related to the injection rate [13, 14]. We attempt to deliver the contrast at a minimum rate of 3 ml/s, but rates of 4–5 ml/s are preferable. Although improved pancreatic parenchymal enhancement can be achieved at rates of 8 ml/s, such rates may be impractical for routine pancreatic MDCT [15]. A power injector should always be used to deliver the intravenous contrast agent.

Acquisition timing and phases of imaging An examination protocol should provide maximal differentiation between normal and abnormal tissue. A challenge of pancreatic imaging is that the timing of peak pancreatic enhancement differs from that of other organs in the abdomen, most notably the liver. Therefore a pancreatic imaging protocol should specify the timing of acquisitions that coincide with the peak enhancement of the organs of interest. In the assessment of pancreatic tumors, there are four basic components: (a) detection

MDCT techniques for imaging pancreatic neoplasms

of the pancreatic tumor; (b) assessment of peripancreatic arteries; (c) assessment of peripancreatic veins; (d) detection of extrapancreatic metastases (most frequently liver). The pancreas is supplied by multiple end arteries arising from the celiac axis and the superior mesenteric artery, whereas the liver is supplied predominantly by the portal vein. Although arteries are homogeneously enhanced during the peak of pancreatic enhancement, the important peripancreatic venous structures may not be homogeneously enhanced until a time near the peak of hepatic enhancement. Because of this discrepancy in organ/vascular opacification, most radiologists employ a dual-phase protocol which incorporates a pancreatic parenchymal phase and a portal venous (hepatic parenchymal) phase [16, 17]. We begin to acquire data for the pancreatic phase at approximately 45–50 seconds after the initiation of the contrast injection. Because of the speed of current MDCT scanners, the pancreatic phase is usually completed in less than 5 seconds. A portal venous phase is acquired at 70–80 seconds after injection initiation. This dual phase acquisition guarantees that the pancreas will be imaged during peak parenchymal enhancement facilitating tumor detection (Figure 3.3), that the peripancreatic arteries and veins are homogeneously enhanced, and that the liver is imaged during the peak of hepatic parenchymal enhancement [18, 19]. A single-phase acquisition can be obtained with a 4-detector row scanner if careful scan timing is used. By acquiring images 50 seconds after initiation of the IV contrast bolus, equivalent sensitivity in assessing tumor resectability was achieved compared with dual-phase imaging [20]. If the images are acquired in a caudal to cephalad direction, the liver can be imaged closer to peak hepatic enhancement. With 16-detector row and higher scanners, this technique may not be appropriate as the liver likely will be imaged too early. The introduction of dual-energy MDCT scanners, which can simultaneously acquire data at 2 different killovoltage peaks (kVp) may facilitate a single-phase study because one tube acquires data at 140 kVp providing “traditional” MDCT contrast while the other tube simultaneously acquires data at 80 kVp leveraging the heightened sensitivity to iodine at low kVp, resulting in increased parenchymal and vascular attenuation [21].

Reconstruction parameters Image data are acquired utilizing the narrowest detector configuration available, and two sets of image data are reconstructed. The first reconstructions create images in familiar CT slice widths (3 or 4 mm) in both axial and coronal planes; these images are sent to the picture archiving and communication system (PACS). The second

33

34

Pancreatic Cancer

Figure 3.3a. Importance of dual-phase acquisition: a tumor visualized as a hypodense mass (arrow) is present in the pancreatic head.

(a)

Figure 3.3b. Importance of dual phase acquisition: same patient as in “A”; same examination, same date, preformed during the portal phase. The pancreatic mass cannot be seen (arrow).

(b)

Figure 3.3c. Importance of dual phase acquisition: pancreatic phase (left) and portal phase images from different patient. The small liver metastasis (arrow) is barely visible on the pancreatic phase image. It is maximally depicted during the portal phase of enhancement.

(c)

reconstructions create image data at the thinnest possible slice width as determined by the detector configuration, attempting to produce data sets composed of isotropic voxels (e.g. 1 mm sections from a 0.75 mm detector configuration on a 16-detector row scanner or 0.625–0.75 mm from a 64-detector row scanner). These “thin” slices are sent to a 3-D workstation which allows multiplanar reformatting, 3-D volume rendering, CT angiography and CT cholangiopancreatography (CTCP) (Figure 3.4). We do not archive the thin slice data sets. These multiple reconstructions are performed for both the pancreatic and portal venous phase images.

MDCT techniques for imaging pancreatic neoplasms

Figure 3.4a. Multipurpose capabilities of MDCT. An axial image from a pancreatic phase examination reveals a hypodense adenocarcinoma in the uncinate process (arrow).

(a)

Figure 3.4b. Multipurpose capabilities of MDCT: same patient as in “a”. Using MIP rendering and an interactive 3-D workstation, the superior mesenteric vein (arrow) can be displayed in its entire length. There is no evidence that the tumor invades the vessel.

(b)

Figure 3.4c. Multipurpose capabilities of MDCT: same patient as in “a”. Using MinIP, the biliary tree and pancreatic ducts can be displayed. The combination of these image acquisitions and 3D rendering techniques facilitates surgical planning.

(c)

35

36

Pancreatic Cancer

Additional MDCT imaging techniques Properly performed and timed MDCT acquisitions as outlined above yield a volume of data that can be viewed interactively. The data set can be imported into a workstation and subjected to a variety of manipulations and rendering techniques that facilitate visualization of key structures of clinical relevance to precisely define the status of a given patient (Figure 3.5). Because this volumetric data set is acquired to obtain near isotropic voxels, the data can be projected in any plane with no significant loss of resolution. The choice of 3-D processing techniques is individualized for the clinical question. We utilize 3-D volume rendering for surgical planning, maximum intensity projection (MIP) combined with volume rendering for CT angiography, and minimum intensity projection (MinIP) for evaluation of the biliary and pancreatic ducts. Ducts can be imaged with 3-D interactive methods or with curved multiplanar techniques [22–26]. Using 3-D techniques, images of the major peripancreatic arteries that rival those of conventional catheter angiography can be created, making recognition of the vascular invasion more apparent than on axial images alone [27–32]. Tumor

Figure 3.5. 3-D MDCT displays create clinically relevant displays of imaging data. This MinIP image reveals a low attenuation mass in the pancreatic head (straight arrow). The CBD (arrowhead) and main pancreatic duct (curved arrow) are both obstructed and can easily be localized in relation to the neoplasm.

MDCT techniques for imaging pancreatic neoplasms

Figure 3.6a. Arterial encasement. 3-D volume rendered image from pancreatic phase MDCT examination reveals an infiltrating tumor affecting the pancreatic tail presenting as a non-mass like region of decreased attenuation affecting the body and tail of the pancreas.

(a)

Figure 3.6b. Arterial encasement: same patient as in “A”. Maximum intensity projection (MIP) CT angiogram derived from the same data set that created the image in 3.6a. MIP utilizes only the brightest pixel along the projection ray, therefore contrast enhanced vascular structures are optimally visualized at the expense of background contrast resolution. Subtle narrowing of the splenic artery (arrowheads) reflects arterial encasement.

(b)

involvement is recognized by obliteration of the normal fat between the pancreatic margin and the adjacent vessel, > 180 degree contact between the tumor and the vessel, and morphologic changes in the artery including narrowing and contour abnormalities (Figure 3.6) [33, 34]. Criteria for venous invasion include >180 degree contact between the tumor and the vein. When the superior mesenteric vein is invaded by a tumor, it may display a “teardrop” configuration (Figure 3.7). Although soft tissue contact with the venous system is highly predictive of non-resectability, venous involvement may be found at surgery when the imaging study fails to reveal any direct contact with the vein. It is therefore critical to evaluate the presence and pattern of collateral venous channels surrounding the pancreatic head. In advanced cases, collateral venous channels are easily recognized. Common collateral channels include prominent short gastric varices, gastrohepatic ligament varices and gastroepiploic to gastrocolic trunk. It is important to look for presence of the small

37

38

Pancreatic Cancer

Figure 3.7a. Superior mesenteric venous (SMV) involvement: axial image from pancreas phase MDCT study reveals distortion of the usually circular SMV (arrow). This has been termed “teardrop” SMV and is highly predictive of tumor extension into the vein.

(a)

Figure 3.7b. Superior mesenteric venous involvement: different patient than in “a”. In this case, the tumor does not alter the shape of the vein; rather, it contacts >180˚ of the vessel circumference (arrow).

(b)

posterior pancreatico-duodenal veins; when these collaterals are present, there is a high likelihood that the tumor has involved the superior mesenteric vein to a degree that would preclude the ability to obtain a negative tumor margin (Figure 3.7) [35–37]. Detection of venous abnormalities is improved using CT angiographic displays [38]. Segmental dilation of the pancreatic duct is a finding that is highly suggestive of neoplasm. In some cases a mass may not be visible on even the most carefully performed studies (Figure 3.8). Differential diagnosis includes chronic pancreatitis, intraductal papillary mucinous tumor or pancreatic adenocarcinoma. Chronic pancreatitis almost always involves the entire pancreatic duct. If the involvement is segmental, one should carefully assess whether the entire duct is distended upstream from the point of caliber change (favoring adenocarcinoma). The importance of this finding cannot be overemphasized. In one study, segmental pancreatic

MDCT techniques for imaging pancreatic neoplasms

Figure 3.8a. Segmental pancreatic duct dilation. Scan in patient with colon cancer undergoing surveillance for recurrence.

(a)

Figure 3.8b. Segmental pancreatic duct dilation: same patient as in “a” 2 years later. During the interval, segmental dilatation of the pancreatic duct developed (arrow). No mass was seen on pancreas dedicated multi-detector row computed tomography (MDCT) or on magnetic resonance imaging (MRI). The downstream duct is normal in caliber making the possibility of chronic pancreatitis unlikely. Surgery was performed and an adenocarcinoma of the pancreatic body was resected at the site of change in caliber of the pancreatic duct. This finding is being increasingly recognized with improved imaging techniques. (b)

duct cutoff was the earliest visible feature that signaled pancreatic pathology and was seen in patients 18 months prior to diagnosis [39].

Accuracy of MDCT There is relatively uniform consensus at the time of writing that overall accuracy of MDCT in detecting tumor and predicting resectability ranges between 86 and 99% [19, 34, 40, 41]. Reported accuracy of MDCT in detecting arterial involvement is reported to be as high as 99% with negative predictive value for resectability of 100%

39

40

Pancreatic Cancer

[32, 34]. Confirmation of venous involvement is best assessed on hepatic phase images, the negative predictive value approaching 100% [19]. Detection of lymph node metastasis is limited. Using a short-axis diameter of greater than 10 mm as the criterion for nodal involvement, CT has reported sensitivity of 14%, a specificity of 85%, a negative predictive value of 82% and an overall accuracy of 73% [42]. The reported results for predicting resectability are higher when surgical resectability is used as a gold standard [34], than when pathology or resection margin positivity is used [43]. MDCT remains limited in demonstrating small extrapancreatic metastatic deposits in the liver and peritoneal cavity. Small peritoneal implants may be present without ascites or signs of peritoneal carcinomatosis making them extremely difficult to detect. Similarly, small hepatic metastases may be “too small to characterize.” A combination of CT scanning supplemented by laparoscopic ultrasound performed immediately prior to intended curative surgical resection in patients deemed resectable by pancreas MDCT is a strategy that has accuracies as high as those of more expensive tests, with no compromise of quality adjusted life-year survival [44].

MDCT and other forms of pancreatic neoplasms The acquisition techniques outlined above have general applicability to any patient suspected of having a pancreatic neoplasm. However, if a specific diagnosis is known, several of the techniques above can aid in further diagnosis. Neuroendocrine tumors, both primary pancreatic lesions and metastases, often present as small intensely enhancing masses. Therefore, if a patient is suspected to have a neuroendocrine tumor, a rapid intravenous contrast injection rate and an earlier, purely arterial phase, acquisition is obtained. At our institution, we will also add a non-contrast enhanced image acquisition through the liver in hopes of increasing our sensitivity for detecting hypervascular metastases. Recent studies have shown that MDCT has substantially improved the detection of neuroendocrine tumors compared with single slice CT (Figure 3.9) [45]. Cystic pancreatic neoplasms are recognized with increasing frequency. A critical differential diagnostic feature that can help to distinguish subtypes of cystic masses is establishing the presence or absence of communication with the pancreatic duct. This task is ideally served by utilizing MinIP CT pancreatographic imaging. Similar results can be achieved with interactive 3-D imaging or curved multiplanar reformatting (Figure 3.10) [46].

MDCT techniques for imaging pancreatic neoplasms

Figure 3.9. Neuroendocrine tumor: a hypervascular focus is present in the pancreatic body (arrow).

Figure 3.10a. Cystic lesion: value of three-dimensional (3-D) multi-detector row computed tomography (MDCT). Axial MDCT image reveals a lobulated cyst in the uncinate process of the pancreas (arrow).

(a)

41

42

Pancreatic Cancer

Figure 3.10b. Cystic lesion: value of 3-D MDCT. 3-D minimum intensity projection (MinIP) image derived from thin data set of pancreatic phase acquisition (“A”) clearly depicts communication with the main pancreatic duct establishing the diagnosis of branch duct intraductal papillary mucinous neoplasm (IPMN).

(b)

Figure 3.10c. Cystic lesion: value of 3-D MDCT. Axial MDCT image in a different patient reveals a unilocular cyst in the uncinate process of the pancreas (arrow). The lesion is similar in appearance and location to that in the previous patient.

(c)

Figure 3.10d. Cystic lesion: value of 3-D MDCT. 3-D MinIP image derived from thin data set of pancreatic phase acquisition (“C”) reveals that, as opposed to the patient in 3.10a and 3.10b the lesion does not communicate with the pancreatic duct. The lesion is an oligocystic serous cystadenoma.

(d)

MDCT techniques for imaging pancreatic neoplasms

REFERENCES 1. Megibow AJ. Pancreatic adenocarcinoma: designing the examination to evaluate the clinical questions. Radiology 1992; 183(2): 297–303. 2. Mahesh M. Search for isotropic resolution in CT from conventional through multiple-row detector. Radiographics 2002; 22(4): 949–962. 3. Itoh S, Ikeda M, Ota T, et al. Assessment of the pancreatic and intrapancreatic bile ducts using 0.5-mm collimation and multiplanar reformatted images in multislice CT. Eur Radiol 2003; 13(2): 277–285. 4. Brenner DJ, Hall EJ. Computed tomography – an increasing source of radiation exposure. N Engl J Med 2007; 357(22): 2277–2284. 5. Dalrymple NC, Prasad SR, El-Merhi FM, Chintapalli KN. Price of isotropy in multidetector CT. Radiographics 2007; 27(1): 49–62. 6. Winter TC, Ager JD, Nghiem HV, et al. Upper gastrointestinal tract and abdomen: water as an orally administered contrast agent for helical CT. Radiology 1996; 201(2): 365–370. 7. Megibow AJ, Babb JS, Hecht EM, et al. Evaluation of bowel distention and bowel wall appearance by using neutral oral contrast agent for multi-detector row CT. Radiology 2006; 238(1): 87–95. 8. Mitka M. Milk shows potential as CT contrast agent. JAMA 2007; 297(4): 353. 9. Horton KM, Fishman EK. MDCT of the duodenum: technique and clinical applications. Crit Rev Comput Tomogr 2004; 45(5–6): 309–334. 10. House MG, Yeo CJ, Cameron JL, et al. Predicting resectability of periampullary cancer with three-dimensional computed tomography. J Gastrointest Surg 2004; 8(3): 280–288. 11. Megibow AJ, Jacob G, Heiken JP, et al. Quantitative and qualitative evaluation of volume of low osmolality contrast medium needed for routine helical abdominal CT. Am J Roentgenol 2001; 176(3): 583–589. 12. Yanaga Y, Awai K, Nakayama Y, et al. Pancreas: patient body weight tailored contrast material injection protocol versus fixed dose protocol at dynamic CT. Radiology 2007; 245(2): 475–482. 13. Kim T, Murakami T, Takahashi S, et al. Pancreatic CT imaging: effects of different injection rates and doses of contrast material. Radiology 1999; 212(1): 219–225. 14. Tublin ME, Tessler FN, Cheng SL, Peters TL, McGovern PC. Effect of injection rate of contrast medium on pancreatic and hepatic helical CT. Radiology 1999; 210(1): 97–101. 15. Schueller G, Schima W, Schueller-Weidekamm C, et al. Multidetector CT of pancreas: effects of contrast material flow rate and individualized scan delay on enhancement of pancreas and tumor contrast. Radiology 2006; 241(2): 441–448. 16. Ichikawa T, Erturk SM, Sou H, et al. MDCT of pancreatic adenocarcinoma: optimal imaging phases and multiplanar reformatted imaging. Am J Roentgenol 2006; 187(6): 1513–1520. 17. Diehl SJ, Lehmann KJ, Sadick M, Lachmann R, Georgi M. Pancreatic cancer: value of dual-phase helical CT in assessing resectability. Radiology 1998; 206(2): 373–378.

43

44

Pancreatic Cancer

18. McNulty NJ, Francis IR, Platt JF, et al. Multi-detector row helical CT of the pancreas: effect of contrast-enhanced multiphasic imaging on enhancement of the pancreas, peripancreatic vasculature, and pancreatic adenocarcinoma. Radiology 2001; 220(1): 97–102. 19. Fletcher JG, Wiersema MJ, Farrell MA, et al. Pancreatic malignancy: value of arterial, pancreatic, and hepatic phase imaging with multi-detector row CT. Radiology 2003; 229(1): 81–90. 20. Imbriaco M, Megibow AJ, Camera L, et al. Dual-phase versus single-phase helical CT to detect and assess resectability of pancreatic carcinoma. Am J Roentgenol 2002; 178(6): 1473–1479. 21. Johnson TR, Krauss B, Sedlmair M, et al. Material differentiation by dual energy CT: initial experience. Eur Radiol 2007; 17(6): 1510–1517. 22. Francis IR. Pancreatic adenocarcinoma: diagnosis and staging using multidetector-row computed tomography (MDCT) and magnetic resonance imaging (MRI). Cancer Imaging 2007; 7 Spec No A: S160–S165. 23. Tunaci M. Multidetector row CT of the pancreas. Eur J Radiol 2004; 52(1): 18–30. 24. Prokesch RW, Schima W, Chow LC, Jeffrey RB. Multidetector CT of pancreatic adenocarcinoma: diagnostic advances and therapeutic relevance. Eur Radiol 2003; 13(9): 2147–2154. 25. Fishman EK, Horton KM. Imaging pancreatic cancer: the role of multidetector CT with three-dimensional CT angiography. Pancreatology 2001; 1(6): 610–624. 26. Kim HC, Yang DM, Jin W, et al. Multiplanar reformations and minimum intensity projections using multi-detector row CT for assessing anomalies and disorders of the pancreaticobiliary tree. World J Gastroenterol 2007; 13(31): 4177–4184. 27. Lu DS, Reber HA, Krasny RM, Kadell BM, Sayre J. Local staging of pancreatic cancer: criteria for unresectability of major vessels as revealed by pancreatic-phase, thin-section helical CT. Am J Roentgenol 1997; 168(6): 1439–1443. 28. Raptopoulos V, Steer ML, Sheiman RG, et al. The use of helical CT and CT angiography to predict vascular involvement from pancreatic cancer: correlation with findings at surgery. Am J Roentgenol 1997; 168(4): 971–977. 29. Fishman EK, Horton KM, Urban BA. Multidetector CT angiography in the evaluation of pancreatic carcinoma: preliminary observations. J Comput Assist Tomogr 2000; 24(6): 849–853. 30. Nakayama Y, Yamashita Y, et al. Vascular encasement by pancreatic cancer: correlation of CT findings with surgical and pathologic results. J Comput Assist Tomogr 2001; 25(3): 337–342. 31. Vargas R, Nino-Murcia M, Trueblood W, Jeffrey RB, Jr. MDCT in pancreatic adenocarcinoma: prediction of vascular invasion and resectability using a multiphasic technique with curved planar reformations. Am J Roentgenol 2004; 182(2): 419–425. 32. Li H, Zeng MS, Zhou KR, Jin da Y, Lou WH. Pancreatic adenocarcinoma: the different CT criteria for peripancreatic major arterial and venous invasion. J Comput Assist Tomogr 2005; 29(2): 170–175. 33. Li H, Zeng MS, Zhou KR, Jin da Y, Lou WH. Pancreatic adenocarcinoma: signs of vascular invasion determined by multi-detector row CT. Br J Radiol 2006; 79(947): 880–887. 34. Zamboni GA, Kruskal JB, Vollmer CM, et al. Pancreatic adenocarcinoma: value of multidetector CT angiography in preoperative evaluation. Radiology 2007; 245(3): 770–778.

MDCT techniques for imaging pancreatic neoplasms

35. Yamada Y, Mori H, et al. CT assessment of the inferior peripancreatic veins: clinical significance. Am J Roentgenol 2000; 174(3): 677–684. 36. Hough TJ, Raptopoulos V, Siewert B, Matthews JB. Teardrop superior mesenteric vein: CT sign for unresectable carcinoma of the pancreas. Am J Roentgenol 1999; 173(6): 1509–1512. 37. Hommeyer SC, Freeny PC, Crabo LG. Carcinoma of the head of the pancreas: evaluation of the pancreaticoduodenal veins with dynamic CT – potential for improved accuracy in staging. Radiology 1995; 196(1): 233–238. 38. Lepanto L, Arzoumanian Y, Gianfelice D, et al. Helical CT with CT angiography in assessing periampullary neoplasms: identification of vascular invasion. Radiology 2002; 222(2): 347–352. 39. Gangi S, Fletcher JG, Nathan MA, et al. Time interval between abnormalities seen on CT and the clinical diagnosis of pancreatic cancer: retrospective review of CT scans obtained before diagnosis. Am J Roentgenol 2004; 182(4): 897–903. 40. Procacci C, Biasiutti C, Carbognin G, et al. Spiral computed tomography assessment of resectability of pancreatic ductal adenocarcinoma: analysis of results. Dig Liver Dis 2002; 34(10): 739–747. 41. Nino-Murcia M, Tamm EP, Charnsangavej C, Jeffrey RB, Jr. Multidetector-row helical CT and advanced postprocessing techniques for the evaluation of pancreatic neoplasms. Abdom Imaging 2003; 28(3): 366–377. 42. Roche CJ, Hughes ML, Garvey CJ, et al. CT and pathologic assessment of prospective nodal staging in patients with ductal adenocarcinoma of the head of the pancreas. Am J Roentgenol 2003; 180(2): 475–480. 43. Smith SL, Basu A, Rae DM, Sinclair M. Preoperative staging accuracy of multidetector computed tomography in pancreatic head adenocarcinoma. Pancreas 2007; 34(2): 180–184. 44. McMahon PM, Halpern EF, Fernandez-del Castillo C, Clark JW, Gazelle GS. Pancreatic cancer: cost-effectiveness of imaging technologies for assessing resectability. Radiology 2001; 221(1): 93–106. 45. Rappeport ED, Hansen CP, Kjaer A, Knigge U. Multidetector computed tomography and neuroendocrine pancreaticoduodenal tumors. Acta Radiol 2006; 47(3): 248–256. 46. Sahani DV, Kadavigere R, Blake M, et al. Intraductal papillary mucinous neoplasm of pancreas: multi-detector row CT with 2D curved reformations – correlation with MRCP. Radiology 2006; 238(2): 560–569.

45

4 Magnetic resonance imaging (MRI) techniques for evaluating pancreatic neoplasms Shawyon Shadman and Vamsi R. Narra

Introduction The imaging of pancreatic neoplasms often presents a diagnostic challenge to the interpreting radiologist. The radiological questions that need to be answered include the differentiation between benign and malignant entities and the determination of extent of disease and resectability. The latter includes evaluation of the lesion’s relationship to the surrounding vasculature and assessment of the presence of local lymphadenopathy and metastatic disease to the liver and peritoneum. This thorough evaluation allows the referring clinician to determine the appropriateness of surgical resection of the lesion in question. The superior soft tissue contrast provided by magnetic resonance imaging (MRI) compared with other imaging tests has proven highly effective for imaging of the abdomen. Over the past decade, advancements in MRI techniques have resulted in a dramatic improvement in the ability of the radiologist to analyze pancreatic neoplasms. These advancements include the introduction of faster breath-hold sequences to limit motion artifact, the development of improved protocols for optimization of pancreatic contrast enhancement, the use of magnetic resonance cholangiopancreatography (MRCP) to evaluate the relationship of pancreatic lesions to the pancreaticobiliary system, and the use of magnetic resonance angiography (MRA) to evaluate the relationship of pancreatic masses to the adjacent vasculature. Use of these techniques results in accurate diagnosis of malignant pancreatic lesions with 95% sensitivity and specificity. Of similar importance, positive and negative predictive values of cancer non-resectability as high as 90% and 83%, respectively, can be achieved [1]. Emerging techniques including the use of diffusion-weighted imaging and proton magnetic resonance spectroscopy may prove to advance the power of MRI as a diagnostic tool in the imaging of pancreatic neoplasms. Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

MRI techniques for evaluating pancreatic neoplasms

Imaging techniques Patient preparation

In order to optimize visualization of the pancreaticobiliary tree with MRCP, the patient should take nothing by mouth for 4–6 hours preceding the examination. Alternatively, an oral contrast agent with decreased signal characteristics on T2-weighted imaging can be administered immediately prior to the examination in order to minimize the hyperintense signal caused by fluid normally found in the upper gastrointestinal tract. Such agents include commercially available products such as ferumoxsil (Gastromark, Mallinckrodt, Maryland Heights, MO) and dilute gadopentetate dimeglumine (Magnevist, Berlex Laboratories, Wayne, NJ) [2]. In addition, naturally occurring high manganese levels in blueberry juice [3] and pineapple juice [4] can be similarly effective. Oral barium also can be administered. The use of a phased array surface coil allows for a significantly higher signalto-noise ratio than can be achieved with a standard built-in body coil, allowing for improved image quality. This increased signal-to-noise ratio allows for decreased imaging sequence duration, permitting the use of breath-hold techniques which result in decreased artifact caused by patient motion [5]. Further image optimization is achieved with field strengths greater than 1.0 Tesla. This allows improved signal-to-noise ratio and improved chemically selective excitation-spoiling fat suppression [6]. Non-contrast MR sequences

The normal pancreatic parenchyma demonstrates moderately high signal intensity on T1-weighted images due to the high concentration of acinar proteinaceous fluid [7]. Non-contrast T1-weighted imaging of the pancreas should be performed using a spoiled gradient echo technique with both in phase and opposed phase sequences. The in phase sequence (100/4.6 [repetition time (TR) ms/echo time (TE) ms]; flip angle, 70–90°) allows for the summation of signal from lipid and water within the same voxel. The opposed phase sequence (100/ 2.3, flip angle 70–90°) subtracts the signal between intravoxel lipid and water. These sequences are most useful for evaluating areas of fatty change in the pancreas, which results in “drop out” of signal caused by intravoxel lipid. In addition, they also are useful in evaluating associated hepatic lesions by helping to distinguish primary hepatic neoplasms from metastases to the liver from a primary pancreatic neoplasm.

47

48

(a)

Pancreatic Cancer

(b)

Figure 4.1. Normal pancreatic parenchymal enhancement. Axial T1-weighted images of the pancreatic parenchyma using a three-dimensional, low flip angle spoiled gradient echo acquisition (a) prior to the administration of intravenous gadolinium and then in the (b) arterial and (c) portal venous phases of enhancement, demonstrate that the peak enhancement of normal pancreatic parenchyma occurs during the late arterial phase.

(c)

Additional non-contrast T1-weighted imaging using fat suppression also should be performed. Fat suppression enables improved visualization of the pancreatic parenchyma from the surrounding mesenteric fat. Fat suppression sequences are best performed using a traditional gradient echo breath-hold technique (212/2.38, flip angle, 70°) or a three-dimensional, low flip angle, spoiled gradient echo sequence (4.83/2.15; flip angle, 15°) (Figure 4.1). Non-contrast T2-weighted imaging is most useful for the evaluation of liver metastases, detection of pancreatic islet cell tumors, and evaluation of cystic lesions of the pancreas. Although many abdominal non-contrast T2-weighted imaging techniques have been described, the most successful have proven to be breath-hold inversion recovery sequences (4480/127.0; flip angle, 150°) and respiratorytriggered rapid acquisition with refocused echoes (RARE) sequences (3300/103, flip angle, 150°; echo train length, 29) [8]. In addition, excellent delineation of the common bile duct and pancreatic duct are provided by T2-weighted echo-train

MRI techniques for evaluating pancreatic neoplasms

(a)

(c)

(b)

Figure 4.2. Pancreatic adenocarcinoma. Axial T1-weighted images of the pancreatic head using a three-dimensional, low flip angle spoiled gradient echo acquisition (a) prior to the administration of intravenous gadolinium and then in the (b) arterial and (c) portal venous phases of enhancement, demonstrate a hypovascular mass occupying the head of the pancreas. A rim of surrounding normal, arterially enhancing pancreatic parenchyma is seen. The dynamic technique allows for excellent visualization of the relationship of the mass to the adjacent vasculature. In this case, note the intimate relationship of the adenocarcinoma to the superior mesenteric artery in (b) and superior mesenteric vein in (c).

sequences. Single-shot fast spin echo sequences such as the single-shot RARE sequence (1000/99, flip angle, 180°; echo train length, 256/134/256) can be performed rapidly in the transaxial, coronal and sagittal planes. Contrast-enhanced MR sequences

The administration of intravenous gadolinium chelate is of critical importance in the MR evaluation of pancreatic neoplasms. Gadolinium chelate enhanced imaging enables distinction of pancreatic masses from normal pancreatic parenchyma and allows differentiation between hypo-enhancing adenocarcinomas and hyperenhancing islet cell tumors. The gadolinium chelate contrast is administered in a dose of 0.1 mmol/kg, using a power injector at a rate of 2–3 ml per second.

49

50

Pancreatic Cancer

The most powerful technique for evaluating the pancreas and upper abdomen on contrast-enhanced MR imaging is a fat-suppressed, dynamic, T1-weighted, three-dimensional, low flip angle spoiled gradient echo acquisition [9]. This technique (4.83/2.15; flip angle, 15°) allows for acquisition of thin (2–3 mm) slices and provides the opportunity for near isotropic resolution. Consequently the acquired data can be reconstructed in multiple planes to help determine the relationship of a pancreatic neoplasm to the adjacent mesenteric vasculature [10]. Added benefits of the three-dimensional acquisition compared with two-dimensional gradient echo sequences include diminished motion-induced phase artifact and lack of aortic phase ghosting artifact overlying the body of the pancreas [11]. Pre-contrast imaging should be performed prior to intravenous gadolinium administration. Appropriate timing of post-contrast images is of the utmost importance when performing dynamic imaging. Image timing is best determined using a parallel signal processing unit that automatically determines imaging acquisition based upon the time required for arrival of contrast material into the abdominal aorta [12]. Alternatively, a small test bolus of 1 ml of gadolinium can also be used to determine the estimated aortic transit time. The peak of pancreatic parenchymal enhancement occurs approximately 15 seconds after contrast arrival in the aorta, whereas that of the liver and the peripancreatic veins occurs 25 seconds or more after aortic contrast arrival [13]. Timing is calculated in relation to the acquisition of the center of K-space. It is best to perform multiple post-contrast sequences for pancreatic neoplasms. The contrast-enhanced images should be compared directly with the non-contrast enhanced images. Additional post processing such as subtraction of the pre-contrast from the contrast-enhanced series can be of use in identifying areas of subtle enhancement. Magnetic resonance cholangiopancreatography (MRCP)

Magnetic resonance cholangiopancreatography (MRCP) has proven to be a highly reliable, non-invasive technique for evaluating the pancreaticobiliary tree [14]. The information provided by MRCP includes determination of the relationship of a pancreatic mass to the pancreatic and common bile ducts. The MRCP images assist the interpreting radiologist in distinguishing among pancreatic andenocarcinoma, islet cell neoplasms, focal pancreatitis and cystic pancreatic masses (Figure 4.3). Magnetic resonance cholangiopancreatography techniques are based on heavy T2-weighting, which renders as bright structures with high water content such as

MRI techniques for evaluating pancreatic neoplasms

(a)

(b)

Figure 4.3. Normal magnetic resonance cholangiopancreatography (MRCP). (a) Coronal and (b) oblique coronal heavily T2-weighted thick slab images after administration of a superparamagnetic oral MRI contrast agent allow for excellent delineation of the biliary tree and pancreatic duct. (c) A maximum intensity projection (MIP) rendering of a three-dimensional, multi-slab T2-weighted RARE sequence in the same patient is shown for comparison.

(c)

the fluid within the pancreaticobiliary tree. Very long TE sequences result in enhanced signal from fluid-containing structures and suppressed signal from the adjacent soft tissues of the upper abdomen. Traditionally, MRCP imaging has been performed using breath-hold, twodimensional sequences. Thick slab, heavily T2-weighted fast spin-echo sequences (4500/1050, flip angle, 150°; echo train length factor, 343) generally have a slice

51

52

Pancreatic Cancer

Figure 4.4. Pancreatic adenocarcinoma. Coronal heavily T2-weighted thick slab image demonstrates marked dilation of both the biliary tree and the pancreatic duct, converging on a region of signal void in the expected location of the pancreatic head. This is the “double duct” sign, caused by an adenocarcinoma within the head of the pancreas which obstructs the common bile duct and the pancreatic duct. Post-contrast dynamic imaging of this lesion is seen in Figure 4.2.

thickness of 40–80 mm and are obtained in the straight coronal plane and coronal-oblique planes in order to highlight the common bile duct and pancreatic duct (Figure 4.4). Alternatively, thin section, heavily T2-weighted single-shot fast spin-echo sequences, such as single-shot RARE, are performed with a slice thickness of 3–6 mm [15]. These techniques allow for rapid imaging of the pancreaticobiliary tree within a single breath-hold. However, a limitation of the thick section technique is that it often does not provide adequate anatomic detail. Limitations of the single-shot RARE technique are a relatively low signal-tonoise ratio, low resolution and the presence of significant gaps in the obtained slices [16].

MRI techniques for evaluating pancreatic neoplasms

An additional option for MRCP imaging is the use of a three-dimensional (3-D), multi-slab T2-weighted RARE technique (1600/654; flip angle, 170°; echo train length, 121). This can be performed using a breath-hold technique or a respiratory-triggered technique. Respiratory triggering nearly eliminates motion artifacts by synchronizing imaging with the movement of the diaphragm, allowing for data acquisition without requiring the patient to breath hold [17]. This 3-D technique allows for a slice thickness as low as 1 mm, resulting in improved spatial resolution [18]. This technique also has the advantage of postprocessing analysis including multiplanar reconstruction, volume rendering and maximum intensity projection. This 3-D technique has been reported to result in superior quality MRCP images and better delineation of pancreaticobiliary anatomy compared with two-dimensional (2-D) techniques [19]. Until recently the primary disadvantage of the 3-D sequences was their long duration. Imaging times of 10–15 minutes were required with free-breathing and respiratory triggering in order to achieve a slice thickness of 2 mm [20]. However, recent advances, including parallel acquisition and new pulse sequences have substantially decreased the required imaging time. Imaging times also can be reduced with a −90° radiofrequency restore pulse at the end of each echo train, which flips the transverse magnetization into the longitudinal direction, shortening the spin relaxation time without sacrificing contrast resolution [21]. Acquisition times of approximately 30 seconds with breath-hold and parallel acquisition and 3–4 minutes with free-breathing and respiratory-triggering now can be achieved [19]. Supplementary and emerging techniques

Although contrast-enhanced MR imaging using 3-D, low flip angle spoiled gradient echo acquisition is widely considered adequate for excellent visualization of the mesenteric vasculature adjacent to the pancreas, other techniques for visualizing these structures are occasionally used. Contrast-enhanced magnetic resonance angiography (CEMRA) is widely regarded as the premier MR method for evaluating the mesenteric vasculature. Advantages of CEMRA over noncontrast techniques such as time-of-flight and phase-contrast MRA include elimination of flow-related artifacts and loss of signal from slow flow or spin saturation [22], allowing higher spatial resolution. After intravenous gadolinium administration, the T1 relaxation time of blood is shortened, resulting in increased signal within the vasculature. CEMRA is performed using a T1-weighted, fast,

53

54

Pancreatic Cancer

spoiled gradient echo 3-D technique with fat saturation (3.6/1.2; flip angle, 25°). After an initial precontrast mask sequence is performed a test bolus or automated bolus detection is used to determine the aortic arrival time of the contrast bolus. The CEMRA sequence is then repeated during the arterial and portal venous phases of enhancement after administration of 0.1 mmol/kg gadolinium chelate contrast at a rate of 2 ml/s. The precontrast mask sequence is then subtracted from the contrast-enhanced sequences during post-processing to facilitate visualization of the mesenteric vasculature. The resulting 3-D angiographic data sets can then be reconstructed in multiple projections. However, CEMRA and dynamic 3-D gradient echo contrast-enhanced imaging cannot be performed during the same examination, thus limiting the use of CEMRA in the standard evaluation of pancreatic neoplasms. An additional rapid acquisition technique used to visualize the mesenteric vasculature without administration of an intravenous contrast agent is the balanced steady-state free precession gradient echo sequence (3.7/1.9; flip angle, 80°). Tissues with a high T2 to T1 ratio, such as bile and blood, demonstrate high signal intensity, which allows for rapid visualization of arterial and venous mesenteric vasculature as well as the larger caliber structures of the pancreaticobiliary system [23]. Although susceptible to significant artifacts, this imaging sequence can be of great use when evaluating patients who should not receive intravenous gadolinium or who are incapable of breath holding. Although widely utilized in MR imaging of the central nervous system, diffusion weighted imaging (DWI) currently is emerging as a technique for MR imaging of the upper abdomen. Diffusion-weighted imaging utilizes changes in Brownian motion of water protons to identify disease states such as infarction and neoplasia. Recently the use of a non-breath-hold, high-b value, spin-echo DWI technique (8000–10 000/73.2 – 73.4; 6 excitations) was reported to result in a sensitivity and specificity of 96% and 99% respectively for the detection of pancreatic adenocarcinoma [24] (Figure 4.5). The quantitative data obtained with DWI can be fused with conventional MR images to improve the anatomic resolution of the DWI findings. Diffusion-weighted imaging also has been used successfully to distinguish mucin-producing pancreatic neoplasms from benign cystic pancreatic tumors [25]. An additional MR imaging technique that is not routinely used clinically but has shown preliminary promise for pancreatic imaging is magnetic resonance spectroscopy (MRS). In one study, this technique, which is based upon the diminished lipid-peak found in regions of fibrosis associated with chronic pancreatitis, was

MRI techniques for evaluating pancreatic neoplasms

(a)

(b)

Figure 4.5. Pancreatic adenocarcinoma. (a) Nonbreath-hold, spin-echo diffusion-weighted image of an adenocarcinoma of the head of the pancreas with (b) corresponding apparent diffusion coefficient (ADC) map demonstrates that the adeoncarcinoma is characterized by increased signal on the diffusion-weighted image and decreased signal on the ADC, indicating internal restricted diffusion. (c) A corresponding T1-weighted three dimensional, low flip angle spoiled gradient echo acquisition image of the lesion in the arterial phase of contrast enhancement is provided. (c)

found to be highly sensitive, although not highly specific, in distinguishing pancreatic adenocarcinoma from changes of chronic pancreatitis [26].

Conclusion Magnetic resonance imaging is a powerful imaging technique for evaluating pancreatic neoplasms. Improvements in MR pulse sequences and MRCP have resulted in a highly reliable means of detecting and staging pancreatic neoplasms and distinguishing malignant from benign pancreatic disease processes. Emerging techniques such as DWI and MRS may help to improve the diagnostic capability of magnetic resonance imaging of the pancreas.

55

56

Pancreatic Cancer

REFERENCES 1. Hanninen EL, Amthauer H, Hosten N, et al. Prospective evaluation of pancreatic tumors: accuracy of MR imaging with MR cholangiopancreatography and MR angiography. Radiology 2002; 224: 34–41. 2. Heller SL, Lee VS. MR imaging of the gallbladder and biliary system. Magn Res Imag Clin N Am 2005; 13: 295–311. 3. Hiraishi K, Narabayashi I, Fujita O, et al. Blueberry juice: preliminary evaluation as an oral contrast agent in gastrointestinal MR imaging. Radiology 1995; 194: 119–123. 4. Heller SL, Lee VS. MR imaging of the gallbladder and biliary system. Magn Res Imag Clin N Am 2005; 13: 295–311. 5. Keogan MT, Edelman RR. Technologic advances in abdominal MR imaging. Radiology 2001; 220: 310–320. 6. Pamulkar E, Semelka R. MR imaging of the pancreas. Magn Res Imag Clin N Am 2005; 13: 313–330. 7. Miller F, Rini N, Keppke A. MRI of adenocarcinoma of the pancreas. Am J Roentgenol 2006; 187: W365–W374. 8. Keogan MT, Edelman RR. Technologic advances in abdominal MR imaging. Radiology 2001; 220: 310–320. 9. Rofsky NM, Lee VS, Laub G, et al. Abdominal MR imaging with a volumetric interpolated breath-hold examination. Radiology 1999; 212: 876–884. 10. Keogan MT, Edelman RR. Technologic advances in abdominal MR imaging. Radiology 2001; 220: 310–320. 11. Birchard KR, Semelka RC, Hyslop WB, et al. Suspected pancreatic cancer: evaluation by dynamic gadolinium-enhanced 3D gradient-echo MRI. Am J Roentgenol 2005; 185: 700–703. 12. Foo TK, Saranthan M, Prince MR, et al. Automated detection of bolus arrival and initiation of data acquisition in fast, three-dimensional, gadolinium-enhanced MR angiography. Radiology 1997; 203: 275–280. 13. Kanematsu M, Shiratori Y, Hoshi H, et al. Pancreas and peripancreatic vessels: effect of imaging delay on gadolinium enhancement at dynamic gradient-recalled-echo MR imaging. Radiology 2000; 215: 95–102. 14. Soto JA, Barish MA, Yucel EK, et al. Pancreatic duct: MR cholangiopancreatography with a three-dimensional fast spin-echo technique. Radiology 1995; 196: 459–464. 15. Reinhold C, Bret PM. Current status of MR cholangiopancreatography. Am J Roentgenol 1996; 166: 1285–1295. 16. Heller SL, Lee VS. MR imaging of the gallbladder and biliary system. Magn Res Imag Clin N Am 2005; 13: 295–311. 17. Zech CJ, Herrmann KA, Huber A, et al. High-resolution MR-imaging of the liver with T2-weighted sequences using integrated parallel imaging: comparison of prospective motion correction and respiratory triggering. J Magn Res Imag 2004; 20: 443–450.

MRI techniques for evaluating pancreatic neoplasms

18. Reinhold C, Bret PM. Current status of MR cholangiopancreatography. Am J Roentgenol 1996; 166: 1285–1295. 19. Zhang J, Israel GM, Hecht EM, et al. Isotropic #d T2-weighted MR cholangiopancreatography with parallel imaging: feasibility study. Am J Roentgenol 2006; 187: 1564–1570. 20. Soto JA, Barish MA, Yucel EK, et al. Pancreatic duct: MR cholangiopancreatography with a three-dimensional fast spin-echo technique. Radiology 1995; 196: 459–464. 21. Hoeffel C, Azizi L, Lewin M, et al. Normal and pathologic features of the postoperative biliary tract at 3D MR cholangiopancreatography and MR imaging. Radiographics 2006; 26: 1603–1620. 22. Nael K, Laub G, Finn JP. Three-dimensional contrast-enhanced MR angiography of the thoraco-abdominal vessels. Magn Res Imag Clin N Am 2005; 13: 359–380. 23. Keogan MT, Edelman RR. Technologic advances in abdominal MR imaging. Radiology 2001; 220: 310–320. 24. Ichikawa T, Erturk SM, Motosugi U, et al. High-b value diffusion-weighted MRI for detecting pancreatic adenocarcinoma: preliminary results. Am J Roentgenol 2007; 188: 409–414. 25. Yamashita Y, Namimoto T, Mitsuzaki K, et al. Mucin-producing tumor of the pancreas: diagnostic value of diffusion-weighted echo planar imaging. Radiology 1998; 208: 605–609. 26. Cho SG, Lee DH, Lee KY, et al. Differentiation of chronic focal pancreatitis from pancreatic carcinoma by in vivo proton magnetic resonance spectroscopy. J Comput Assist Tomogr 2005; 29: 163–169.

57

5 Imaging evaluation of pancreatic ductal adenocarcinoma Steven S. Raman and David S. K. Lu

Introduction In Western countries, pancreatic ductal adenocarcinoma (pancreatic cancer) is the 4th–5th leading cause of cancer-related death, with over 200 000 cases annually. In the USA, it is estimated that 33 370 deaths resulted from pancreatic cancer in 2007 [1]. Clinical presentation of pancreatic ductal adenocarcinoma depends on its location. Because there are no reliable clinical or laboratory parameters to detect lesions early in the course of the disease, approximately 80–90% of patients with pancreatic adenocarcinoma present at an advanced stage (III or IV). Approximately 60–70% of these malignancies arise in the pancreatic head, and most come to clinical attention when patients present with weight loss and painless jaundice. The remainder, which arise in the body, tail or uncinate process, usually are clinically occult until a late stage when patients often present with non-specific symptoms such as weight loss or back pain [2]. Surgical resection of ductal adenocarcinoma offers the only chance for cure in the 10–20% of patients who present at an early stage (I or II). Patients who undergo complete surgical resection of the primary mass have improved 5-year survival [3]. As with detection, no reliable clinical or laboratory tests are available to predict cancer stage and determine the potential for surgical resection in a given patient [2–5]. Surgical triage of patients is predicated on accurate preoperative imaging to maximize benefit for potentially resectable patients and to spare potentially morbid surgery (in up to 20% of attempted resections) for those with advanced stage disease that is unlikely to be resectable [6,7]. Pancreatic surgery has a mortality rate of 2.5% or less in high-volume centers [8]. Over the past few years, multi-detector row CT (MDCT) supplemented by magnetic resonance Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

Imaging evaluation of pancreatic ductal adenocarcinoma

imaging (MRI), endoscopic ultrasound (EUS) and positron emission tomography (PET) with computed tomography (PET-CT) has supplanted older techniques such as catheter angiography for detection and staging of pancreatic adenocarcinoma. The choice of imaging technique depends to some extent on available equipment and local expertise. Transabdominal ultrasound (US) also may be useful for establishing an initial diagnosis, but is not considered adequate for staging. The role of imaging with regard to screening average and high-risk patients has yet to be clearly defined. In this chapter, we will review the principles of CT and MR for detection and staging of pancreatic adenocarcinoma. In addition, we will provide an overview of the use of endoscopic ultrasound and PET-CT for evaluating pancreatic carcinoma.

Goals of preoperative imaging for pancreatic ductal adenocarcinoma The goals of preoperative imaging in the diagnosis of pancreatic ductal adenocarcinoma are 1. To detect and characterize a solid pancreatic mass as highly likely to be an adenocarcinoma. 2. To demonstrate the relationship of the mass to surrounding mesentery, bowel and major arteries and veins. 3. To detect surgery-precluding metastases primarily to liver and peritoneum. Detection of pancreatic ductal cancer

Detection of the vast majority of pancreatic ductal adenocarcinomas relies on high quality imaging, and the current standard is contrast-enhanced multi-detector row CT (MDCT). As stated previously, there are no reliable clinical or serum tumor markers to help stage patients. Although serum CA-19-9 is elevated to some degree in over 80% of patients with pancreatic adenocarcinoma, it is unreliable for detection of early stage lesions and may predict unresectable advanced stage pancreatic adenocarcinomas when highly elevated (> 1000 u/ml) [2–5]. Since contrast-enhanced MDCT and MRI are relied upon for preoperative diagnosis and staging, a high-quality technique becomes critically important. For both CT and MR, acquisition parameters, contrast dose, injection rate, post contrast image timing and multiplanar evaluation with a variety of post processing techniques are important to achieve consistently excellent results.

59

60

Pancreatic Cancer

Computed tomography

Over the past 15 years, advances in CT and ancillary technologies including intravenous contrast power injectors and post processing has enabled helical and multi-detector CT to become the dominant imaging modality for detection and staging of pancreatic adenocarcinoma [7]. Imaging for determination of preoperative resectablity was first described with non-helical CT scanners but was of limited reliability due to poor image quality from suboptimal slice thickness, scan duration and speed of gantry rotation [9]. Single-detector slip ring-based helical CT, introduced in the late 1980s, revolutionized pancreatic imaging by enabling acquisition of thinner collimation (2.5–3 mm) data sets of the entire pancreas during a single breath hold. The introduction of power injectors enabled optimized timing of an intravenous contrast bolus to improve contrast between the adenocarcinoma and uninvolved pancreas and its surrounding vessels. The combination enabled more optimized temporal and spatial resolution [10,11] improving accuracy of preoperative local staging [12,13]. The near exponential progression of multi-detector row CT from 4 to 64 detector rows with gantry rotation speeds as fast as 0.33 seconds has enabled high resolution, multi-planar imaging in short breath holds for the majority of patients. The planned introduction of 128- and 320-detector array systems will likely further improve temporal and spatial resolution with possibility of new applications such as whole organ functional imaging. Exponential advances in workstation hardware and software have enabled routine and widespread interactive post processing of volumetric MDCT data by radiologists and technologists. The optimal plane and post processing algorithm to demonstrate the primary pancreatic adenocarcinoma mass and its relationship to critical surrounding vessels can be easily selected [14]. Using internet-based client server commercial applications, data sets may routinely be manipulated by an end user such as the treating surgeon to demonstrate any feature of interest. Computed tomography technique

A high-quality preoperative CT to detect and stage suspected pancreatic ductal adenocarcinoma may be performed on both single- and multi-detector helical CT scanners with adherence to the following principles. Computed tomography data should be acquired with thin collimation (0.6–3 mm) and reconstructed to thin sections (1–3 mm). An unenhanced scan covering the liver, pancreas and kidneys should be acquired prior to contrast injection to allow detection of incidental hepatic steatosis, subtle calcifications within pancreatic lesions and subtle primary or metastatic lesions (by measuring relative enhancement). Iodinated intravenous

Imaging evaluation of pancreatic ductal adenocarcinoma

(a)

(b)

(c)

(d)

Figure 5.1. Post contrast axial CT images through pancreas in pancreatic phase: (a) body and tail, (b) head; and corresponding levels during portal venous phase: (c) body and tail, (d) head. Note optimal enhancement of pancreatic parenchyma along with peripancreatic arteries and veins in the pancreatic phase, with decreased enhancement in the hepatic phase (i.e. portal venous phase) but excellent hepatic enhancement.

contrast dosed to patient weight should then be injected using a power injector, and scans should be acquired during optimal pancreatic and hepatic parenchymal enhancement. Because the blood supply of the pancreas is entirely from the splanchnic arteries whereas that of the liver is mostly from the portal venous system, peak pancreas enhancement occurs earlier than peak hepatic enhancement. Therefore a dual phase protocol, one timed for the pancreas (pancreatic phase), and the other timed for the liver (hepatic phase or portal venous phase), is optimal for comprehensive evaluation (Figures 5.1–5.3) [10–13,15,16]. An optimal pancreatic phase is one in which there is simultaneous definition of tumor at its most conspicuous state along with opacification of adjacent

61

62

(a)

Pancreatic Cancer

(b)

Figure 5.2. Pancreatic adenocarcinoma: Post contrast axial CT image through pancreas in (a) pancreatic phase, and (b) hepatic phase. Note better delineation of hypoenhancing tumor margin (arrows) compared to enhancing normal pancreatic head tissue in the pancreatic phase as compared to the portal venous phase.

(a)

(b)

Figure 5.3. Liver metastasis from pancreatic cancer. Post contrast axial CT images from (a) pancreatic phase, and (b) hepatic phase, show small 8 mm metastatic lesion (arrow) apparent only in the hepatic phase. Also note incidental hemangioma (arrowhead).

Imaging evaluation of pancreatic ductal adenocarcinoma

peripancreatic arteries and veins, in order to optimize tumor detection and local vascular staging. Therefore, imaging too early in the early arterial phase would be suboptimal, as the major and minor surrounding veins important for determining surgical resectability are suboptimally enhanced. Rather, the pancreatic phase is better timed to what some may term as the “late arterial phase” or “early portal venous phase.” The name is less important than the concept – that of optimal pancreatic tumor conspicuity simultaneous with arterial and venous opacification. To this end, the pancreatic phase has been shown to provide superior pancreatic parenchymal and vascular opacification, with improved tumor–pancreas tissue contrast compared with the portal venous or hepatic phase (Figures 5.1, 5.2) [9, 10]. The image data set should be routinely reconstructed in axial and coronal planes and supplemented with other post processing techniques such as curved reformations, maximum intensity projections (MIP), or volume rendering when required [14]. We also recommend use of neutral oral contrast agents such as water [17] or a low Hounsfield unit barium based neutral agent (VoLumenTM; EZ EM). Normal or high-fat milk also has been used for this purpose. In Table 5.1 sample scan parameters for 16- and 64-detector scanners are given. Patient preparation

In our practice, the patient fasts for at least 4 hours and drinks up to 500 ml of water just before the scan, which helps distend the stomach and duodenum. The technologist inserts an 18- or 20-gauge intravenous catheter in an antecubital vein to facilitate power injection of low osmolality intravenous contrast at 3–4 ml/s. An unenhanced scan of the entire abdomen is then acquired, followed by contrastenhanced scans in the pancreatic and hepatic phases according to the protocol listed in Table 5.1. Contrast dose, injection rate and scan timing

The dose and rate of contrast administration must be optimized with a power injector. Between 40–45 grams of iodinated intravenous contrast (e.g. 125 ml of a 350 mg I/ml agent) should be administered to ensure adequate hepatic enhancement for detection of liver metastases (Figure 5.3). We recommend a contrast injection rate of 3–4 ml/s. A variety of timing methods can be used. The simplest and least reliable of these is a fixed timing protocol which is best reserved for single- or 4-row scanners. Given

63

64

Pancreatic Cancer

Table 5.1. Example of pancreatic multi-detector row computed tomography (MDCT) protocol MDCT generation

16 row

64 row

Contrast: 125 ml of 350 mg I/ml contrast agent infused at 3–4 cc/sec Scan timing Pancreatic phase (PP) Delay fixed 50 s 55 s Bolus tracking trigger150HU + 25 s trigger150HU + 30 s Hepatic phase (HP) Delay fixed 70 s 75 s Bolus tracking end of PP + 10 s end of PP + 10 s Scan volume diaph – iliac crest diaph – iliac crest Scan Direction cranio-caudal cranio-caudal Collimation (mm) PP: 0.75, HP:1.5 PP: 0.6, HP: 1.5 Table feed/rotation (mm) PP: 12, HP:24 PP: 38, HP: 76 Rotation time (s) 0.5 0.5 kVp 120 120 mAs 200–250 200–250 Slice reconstructions thickness/interval (mm) PP axial: 2/2 PP axial: 2/2 PP coronal: 2/1 PP coronal: 2/1 PP sagittal: 2/2 PP sagittal: 2/2 HP axial: 5/2.5 HP axial: 5/2.5 Optional for 3-D PP axial: 1/1 PP axial: 1/1

a fixed contrast dose and rate of injection, which at our institution is 125 ml at 3 ml/s, images are acquired for the pancreatic phase starting at 40–45 s on 1–4detector [10] or 50–55 s on 16–64-detector [16] scanners. The faster the scanner speed, the shorter the acquisition time for each phase, and therefore a greater delay should be programmed for the start time of each phase. For the hepatic phase, imaging should begin at 60–70 s on 1–4-detector scanners and delayed to 65–75 s on 16–64-detector scanners due to shorter acquisition times. Note these suggested timing delays only apply to the contrast volume and rate of injection mentioned, and should be modified according to general contrast principles depending on institutional preferences. The main limitation of the fixed timing method is that in up to 20% patients, the pancreatic phase may be mis-timed and imaged suboptimally, especially in patients with altered cardiac function.

Imaging evaluation of pancreatic ductal adenocarcinoma

Another timing option, which we prefer especially with the latest multi-detector scanners, is to time the contrast injection to each individual patient’s cardiac output. We image the pancreatic and portal venous phases using a bolus tracking program. Using a region of interest (ROI) of approximately 70% of aortic diameter, these programs serially measure the progressively increasing aortic attenuation within the ROI at the level of the celiac axis or superior mesenteric artery as the power-injected iodinated contrast bolus arrives. When the aortic attenuation reaches 150 HU, the scanner is triggered after a fixed delay to begin imaging in the pancreatic phase. The delay is programmed depending on scanner speed. On the 64-detector scanner, for example, a 30-s delay is programmed. The hepatic phase is then initiated at a fixed 10-s delay after completion of the pancreatic phase (Table 5.1). Post processing

Routine reconstructions, multiplanar reformations, and standardized 3-D renderings can be performed by a trained technologist. Currently at our institution the unenhanced images are automatically reconstructed at 5 mm thickness and interval in the axial plane, the pancreatic phase images are reconstructed at 2 mm thickness and interval in the axial plane and through the pancreas in both the coronal and sagittal planes. Images from the hepatic phase data set are reconstructed at 5 mm thickness and 2.5 mm interval in the axial plane. In addition, 3-D renderings can be performed either directly from the CT console or the image data can be sent to an independent 3-D workstation. A variety of display options can be generated according to the preference of the radiologist and surgeon to best highlight the desired feature of the individual tumor and its relationship to surrounding structures. These display options include maximum intensity projections (MIP), minimum intensity projections (MinIP), volume rendered (VR) or curved planar reformations (CPR) [Figures 5.4, 5.5] [14]. How much of the 3-D rendering is routinely performed by the technologist versus interactive 3-D by the radiologist is a matter of institutional preference. We currently include standardized MIP and VR renderings as part of routine protocol, and these images are available to the radiologist through the Picture Archiving and Communications System (PACS). With routine volumetric acquisition, multiplanar thin section displays, and 3-D post processing the image data sets are in essence CT angiograms (CTA). The main difference that must be noted is that unlike most CTAs, which are timed to the early arterial vascular phase to highlight only the arterial system, our

65

66

(a)

Pancreatic Cancer

(b)

Figure 5.4. Multiplanar and 3-D renderings from volumetric CT data sets. (a) Maximum intensity projection (MIP) image of pancreatic phase acquisition shows CT angiographic view of splanchnic arteries and veins in this patient with occluded superior mesenteric vein (SMV) (arrows indicate expected position) lateral to the superior mesenteric artery (SMA) (arrowhead) from pancreatic cancer. (b) Volume rendered (VR) image of pancreatic phase acquisition in another patient shows perspective view of SMV stenosis (arrows) from pancreatic uncinate cancer.

pancreatic CTA protocol is an acquisition in the pancreatic phase which captures both splanchnic arterial and venous systems simultaneously along with optimized tumor conspicuity, and it is supplemented by hepatic phase imaging for complete parenchymal evaluation of the abdomen. Computed tomography interpretation detection

The goals of the CT examination are detection, characterization and staging of pancreatic adenocarcinoma. Over 90% of pancreatic adenocarcinomas present as a focal, ill-defined, hypoattenuating, poorly enhancing mass as compared with arterially enhancing background pancreatic parenchyma. In the remainder, a variety of presentations may be possible including an isodense mass and diffuse disease. During the pancreatic phase of contrast enhancement, the difference in attenuation between the background pancreas and the hypodense mass averages 40 HU [10] [Figure 5.2] . Depending on location and size, the mass may present as a hypoattenuating, ill-defined convex deformity of the pancreas, although smaller lesions may not be contour deforming. Calcifications are rarely found within the mass, and their presence may raise a differential diagnosis of other pancreatic lesions. A small percentage may demonstrate cystic change or arise in a preexisting cystic lesion. A mass in the pancreatic head typically obstructs the common

Imaging evaluation of pancreatic ductal adenocarcinoma

(a)

(b)

(c)

(d)

Figure 5.5. Advantages of post processing for display of selected features are demonstrated. A 76-year-old man presented with painless jaundice. MDCT with (a) axial (b) coronal (c) curved reformatted and (d) coronal volume rendered images. There is a pancreatic head mass with occlusion of the superior mesenteric vein-portal confluence with enlarged peripancreatic veins (black arrow). A plastic biliary stent was placed prior to CT. Note how relationship of mass to the venous structures and ducts is optimally demonstrated on the reformatted and volume rendered images.

bile duct and the pancreatic duct with resulting upstream ductal dilation known as the “double duct sign.” If pancreatic duct obstruction is longstanding, the pancreatic body and tail atrophy. On curved planar reformations, the dilated ducts truncate abruptly at the mass.

67

68

Pancreatic Cancer

In approximately 10–11% of cases, pancreatic adenocarcinomas are isoattenuating to background pancreas and may be detected only by indirect signs such as mass effect, contour deformity or abrupt stricture and upstream dilation of the common bile duct and/or pancreatic duct [18]. Differentiating pancreatic adenocarcinoma from other malignancies or inflammatory processes usually is straightforward but occasionally can be challenging based on imaging criteria alone. Imaging diagnoses usually are suggested by clinical presentations of painless jaundice and weight loss with elevated levels of serum tumor markers such as CA-19-9. Hypoattenuating or isoattenuating lesions within the pancreas have a limited differential diagnosis. The most common mimicker of pancreatic adenocarcinoma is focal pancreatitis which may present as an ill-defined hypoattenuating area within pancreas with enlargement of only a portion of the pancreas, typically involving the pancreatic head. Unless recurrent or chronic, focal pancreatitis much less commonly results in upstream pancreatic duct dilation (in the neck, body or tail). Thus the “duct penetrating sign” in which a normal pancreatic duct courses through a hypoattenuating area within the pancreas should prompt consideration of benign conditions such as acute or mild chronic pancreatitis in the differential diagnosis with neoplastic conditions. Another increasingly recognized entity with imaging features mimicking pancreatic adenocarcinoma is autoimmune pancreatitis. This waxing and waning inflammatory disorder may result in focal or diffuse pancreatic parenchymal enlargement with strictures and dilation of the common bile duct and/or pancreatic duct. Primary tumors of the pancreas such as hyperfunctioning islet cell tumors typically are small and hypervascular whereas non-hyperfunctioning islet cell tumors typically are large and well defined. Metastases to pancreas, most commonly from renal cell or bronchogenic carcinoma, are rare and typically do not obstruct the common or pancreatic ducts. Occasionally, a tissue diagnosis by percutaneous or endoscopic fine needle aspiration (FNA) or percutaneous core biopsy may be necessary. Computed tomography interpretation: staging

The TNM staging of pancreatic cancer and its CT staging equivalent are given in Table 5.2. A critical determinant of surgical resectability is circumferential invasion of the mass around peripancreatic vessels including celiac artery (CeA), hepatic artery (HA), superior mesenteric artery (SMA), superior mesenteric vein (SMV) and portal vein (PV). Tumors involving the pancreatic body or tail may be resected (distal pancreatectomy) along with a splenectomy, and therefore splenic artery (SpA) and splenic vein (SpV) invasion does not preclude resection, unless there is

Imaging evaluation of pancreatic ductal adenocarcinoma

Table 5.2. TNM staging of pancreatic adenocarcinoma Primary tumor Tis: Carcinoma in-situ T1: Tumor within pancreas < 2 cm T2: Tumor within pancreas > 2 cm T3: Infiltration into peripancreatic tissue, duodenum or common duct T4: Infiltration into peripancreatic vessels, stomach, spleen, large bowel Regional lymph nodes: N0: No lymph node metastases N1: Metastases in peripancreatic lymph nodes Nx: Unknown Distant metastases: M0: No distant metastases M1: Distant metastases Mx: Unknown CT Staging of adenocarcinoma with TNM equivalent Stage 1: T1 or T2, N0, M0 (Typically resectable when < 50% circumferential vascular involvement Stage II: T1–T3, N0, M0; T3, N0 or N1, M0: (Typically resectable when < 50% circumferential vascular involvement Stage III: T4, N0 or N1, M0: (Typically unresectable as most present with > 50% circumferential vascular involvement or any vascular narrowing Stage IV: Any T or N stage; M1: (Unresectable due to liver, peritoneal or lung metastases)

involvement of these vessels within 1 cm of the celiac axis or spleno-portal venous confluence. Invasion of any of the critical arteries precludes resectability, but limited involvement of the SMV is not an absolute contraindication to surgical resection, if venous resection and grafting can be performed by an experienced surgeon [12,13]. The presence of dilated peripancreatic veins has been associated with a high likelihood of unresectability [19]. Other important landmarks include any potential invasion of the stomach or colon (via transverse mesocolon). Invasion of the anterior pancreatic capsule and invasion into nerve roots are all associated with worse outcomes. The presence of hepatic or peritoneal metastases also is a contraindication to surgical resection. The most durable criteria to predict the extent of vascular involvement in guiding surgical resection were developed by Lu et al. [12] and corroborated

69

70

Pancreatic Cancer

(a)

(b)

(c)

(d)

(e)

Figure 5.6. Schematic of local vascular staging as proposed by Lu et al. [12]. See text for abbreviations. Critical veins: MPV, SMV, SpV within 1 cm of the portal confluence. Critical arteries :Celiac, SMA, CHA, SpA within 1 cm of the celiac bifurcation. = cancer; = Portal vein/SMV; = SMA. Grade 0: No contact between mass and vessels. Grade 1: < 25% circumferential contact with tumor. Grade II: 25–50% degree circumferential contact with tumor. Grade III: 50–75% circumferential contact with tumor.

independently by O’Malley et al. [13]. This system is based on the extent of contact between the tumor and any critical vessel. Please note that any degree of vascular narrowing due to tumor should be graded as IV. The grading system is as follows (Figures 5.6–5.8): Grade 0: No contact between mass and surrounding major vessels. Grade 1: 0–24% circumferential contact (Almost no chance of invasion). Grade 2: 25–49% circumferential contact (approximately 50% chance of invasion). Grade 3: 50–74% circumferential contact (approximately 75% chance of invasion).

Imaging evaluation of pancreatic ductal adenocarcinoma

(a)

(b)

(c)

(d)

Figure 5.7. Grading of local vascular invasion by tumor. Examples of (a) grade 0, (b) grade 1, (c) grade 2, (d) grade 3, and (e,f) grade 4 vascular involvement by tumor. Note degree of circumferential contact of vessel with tumor (arrows) which is (a) none, (b) < 25% of SMV, (c) 25–50% of SMV, (d) 50–75% of SMA and (e,f) > 75% of and narrowing of SMV, and > 75% of SMA. See text for abbreviations.

Grade 4: 75–100% circumferential contact, or any vessel narrowing or occlusion (nearly 100% chance of invasion). A variety of imaging signs have been described to assess vascular invasion. One of these is the “tear drop sign” in which the tumor deforms the normally convex and round margin of the SMV or portal vein resulting in a teardrop-shaped vessel (Figure 5.8). Secondary signs correlated with vascular invasion include dilated peripancreatic veins anterior and posterior to the pancreatic head (Figure 5.5) [19].

71

72

Pancreatic Cancer

(e)

(f)

Figure 5.7. (cont.)

Figure 5.8. The “tear drop” sign. Note partial constriction or “puckering of the lateral wall of the SMV (arrow) on this axial pancreatic phase CT image, despite only approximately 25% wall contact with tumor. Such deformity indicates venous invasion.

Imaging evaluation of pancreatic ductal adenocarcinoma

In some cases, lymphatic obstruction by the mass produces perivascular soft tissue infiltration of the surrounding fat plane. Although this usually indicates unresectability, it is not reliable in all patients. Thus corroboration with serum CA 19–9 and further evaluation with EUS may be necessary. Invasion of a short segment of the SMV or PV does not always preclude resectability because at many institutions venous resection with graft placement is performed. Involvement of arteries, however, almost always predicts nonresectability. Therefore the terms “resectable” and “non-resectable” vary with institution and individual surgeon. These terms should be used carefully and only with full understanding between the radiologist and the treating surgeon. Aside from peripancreatic vascular invasion, other predictors of non-resectability include direct tumor invasion of the stomach or colon. Invasion of the duodenum does not preclude resectability as both the duodenum and the pancreatic head are removed during a Whipple procedure. With CT, enlarged peripancreatic lymph nodes (> 1 cm short axis) are poorly predictive of metastatic involvement as imaging features of inflammatory, reactive and malignant nodes overlap considerably. Based on a composite evaluation of these CT imaging signs, the diagnostic accuracy of determining non-resectability in pancreatic cancer is 95%, whereas the accuracy of determining resectability is 70–80%. Diagnostic accuracy of non-resectability is high because it is based on reproducible CT signs of local invasion and/or metastases. Diagnostic accuracy of determining resectability is more problematic as 5 mm or smaller metastases to liver and peritoneum are poorly detected on CT. A meta-analysis of 68 mostly older studies evaluating helical and conventional CT, MRI and US determined not surprisingly, that helical CT performed best for diagnosis and determination of resectability. Aggregate sensitivity and specificity for diagnosis was 91% and 85% and for resectability was 81% and 82% respectively [20]. In our own experience with 4-, 16- and 64-detector MDCT, sensitivity, specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) for resectability are 100%, 95%, 96%, 85% and 100% respectively [21]. In another recent study, performance similar to our own was achieved with 4-, 8-, 16- and 64-row scanners without significant improvement in PPV with the newest 64-row MDCT scanners [22]. The relatively decreased PPV in predicting resectability preoperatively by experienced readers is due to difficulty detecting 5 mm and smaller liver and peritoneal metastases by CT. Although studies have shown that scans with thinner collimation do help detect a greater percentage of liver metastases in general [23], the missed pancreatic adenocarcinoma metastases to both liver and peritoneum on CT are usually few in number. It is imperative

73

74

Pancreatic Cancer

therefore, that in pancreatic cancer evaluation, optimized hepatic phase images are obtained in addition to the pancreatic phase images, and peritoneal metastases are specifically looked for at interpretation. Magnetic resonance imaging

Advances in MR imaging hardware, pulse sequences, contrast injection precision, and post processing have enabled MRI to compete with single and early multi-detector helical CT in detection and staging of pancreatic adenocarcinoma [24]. However, the rapid advances in CT technology over the past few years with respect to improved temporal and spatial resolution likely have rendered outdated the performance comparisons of these earlier studies [25–27], which suggested that dynamic MRI performed better than available CT of that era. However, the previously mentioned meta-analysis did not demonstrate this [20]. With the high spatial and temporal resolution of MDCT, it is unlikely that dynamic contrastenhanced MRI would perform better than CT, although no current comparison studies have been performed. Currently MRI may be performed on both 1.5T and 3T scanners, although we perform the majority of our cases at 1.5T using the protocol in Table 5.2. Patient preparation: Patients usually fast for at least four hours prior to MR examination. Patients are given water to drink to help distend the duodenum. A variety of other agents may be used for duodenal distension including T2 dark contrast agents. Magnetic resonance imaging technique

Ability to image the pancreas well is highly dependent on the capabilities of the MR scanner. At 1.5 and 3.0 Tesla, with high gradient strengths and fast imaging capabilities, high-quality images should be routinely possible. Important pulse sequences for pancreatic imaging include precontrast and dynamic post contrast fat-saturated T1-weighted images, T2-weighted images, and MRCP images (Table 5.3). For T1-weighted acquisition, with 2-D gradient echo (GRE) sequences (e.g. FSPGR on GE scanners and FLASH on Siemens scanners) it is possible to obtain 3–4 mm images confined to the pancreas, or 1.5–3 mm images through the entire abdomen if 3-D GRE sequences are available (e.g. LAVA on GE scanners and VIBE on Siemens scanners). Dynamic multi-phasic post contrast imaging should be standard. However, because the various phases are shortened

Imaging evaluation of pancreatic ductal adenocarcinoma

Table 5.3. Example of multiphasic contrast-enhanced pancreatic magnetic resonance imaging protocol (on 1.5 T scanner) 1. Coronal T2 Half-Fourier single shot echo train spin echo 2. Axial T2 Half-Fourier single shot echo train spin echo 3. Axial dual echo (in and opposed phase) T1 2-D spoiled gradient echo 4. Coronal fat-saturated T1 2-D FLASH or 3-D spoiled gradient echo 5. Axial fat-saturated T1 2-D FLASH or 3-D spoiled gradient echo 6. Power Injection of 0.1 mmol/kg gadolinium-based contrast agent a. Bolus tracking with sagittal single shot spoiled gradient echo, first scan begins when contrast in abdominal aorta, or b. Fixed timing technique 7. Repeat #5 (if fixed, at 20 s and 50 s) 8. Repeat #4 (if fixed, at 80 s) 9. Repeat #5 at 120 s and 5 minutes 10. Thick slab rotational oblique T2-weighted long TE single shot echo train sequence (MRCP) 11. Navigator triggered thin section 3-D fast recovery echo train spin echo sequence with MIP of bile and pancreatic ducts (MRCP)

due to a small contrast bolus of only 15–20 ml, and each MR acquisition may take up to 20–25 seconds, it is at present not possible to image the pancreatic and hepatic phases of enhancement with the same precision as MDCT. In general two backto-back post contrast acquisitions are performed which correspond to an arterial dominant phase and a portal dominant phase. No definite parallel of the pancreatic phase has been described on MR. However, this is likely less important in MR due to its inherent high contrast. All the necessary detection, characterization and staging information can be obtained on MR with the arterial dominant and portal dominant phase images (Figure 5.9). Unlike CT, MRI allows acquisition of multiple data sets without the penalty of radiation exposure. Thus direct coronal and other multiplanar acquisitions can be performed in addition to the transaxial images, without the need for reformations as for CT data. Although mangafodipir trisodium (Teslascan, GE Healthcare, Oslo, Norway), provides improved pancreatic enhancement, lesion to background pancreatic tissue contrast and pancreatic mass and liver metastasis detection, in practice, this contrast agent has limited application for adenocarcinoma as it provides no vascular information for local staging [28], and it is not available in the USA.

75

76

(a)

(c)

Pancreatic Cancer

(b)

(d)

Figure 5.9. Magnetic resonance imaging of large uncinate mass. Axial fat-saturated T1 3-D GRE images (a) pre-contrast, post contrast in (b) arterial dominant phase and (c) portal dominant or hepatic phase. Note the mass (arrowheads) is hypointense against the bright intrinsic T1 signal of normal pancreatic tissue, but it is most conspicuous on the arterial dominant phase, and is already losing its contrast with surrounding pancreas by the portal dominant phase. Note that it completely encases a stenosed SMA (long arrow), and the SMV-spleno-portal confluence is narrowed and anteriorly displaced (short arrow). (d) The matching fat-saturated T2 RARE image shows the mass to be mildly hyperintense, but not as well delineated from adjacent structures. On (e) coronal 3D GRE venous phase image the mass (arrow) extends inferiorly to the duodenum (arrowheads) and on (f) thick slab T2 MRCP image, the second portion of the duodenum is dilated compatible with a duodenal obstruction. Note also that both the common bile duct and the pancreatic duct are spared. See text for abbreviations.

Normal pancreas: On T1-weighted images, the pancreas is of higher signal intensity than the liver and muscle presumably due to acinar proteins, whereas on T2-weighted images, the pancreas demonstrates uniformly high T2-signal intensity (similar to spleen). The normal pancreatic duct and bile ducts are hypointense to pancreas on T1-weighted sequences and hyperintense to pancreas on T2-weighted sequences. They are easily demonstrated on thin section imaging or on thin and thick slab MRCP sequences. On dynamic T1-weighted images after gadolinium injection, the

Imaging evaluation of pancreatic ductal adenocarcinoma

(e)

(f)

Figure 5.9. (cont.)

pancreas enhances early and is hyperintense to the liver in the arterial dominant phase, becoming isointense to liver on more delayed phases [24]. Pancreatic adenocarcinoma: detection. Pancreatic adenocarcinomas tend to be hypointense relative to normal pancreas on T1-weighted images and of variable intensity on T2-weighted images. As on CT, the margins of most adenocarcinomas tend to be ill defined [24]. Pancreatic ductal adenocarcinomas incite inflammatory and fibrotic reaction in surrounding tissue usually resulting in obstruction of the pancreatic duct and/or common bile duct with upstream ductal dilation. After intravenous administration of extracellular contrast agents such as gadolinium, the lesions show poor early enhancement relative to background pancreas, but slowly enhance over time on more delayed phases, like most adenocarcinomas and in keeping with the fibrotic nature of the mass. Magnetic resonance imaging may be useful when CT findings are non-specific in cases of suspected adenocarcinoma. In some series, up to 10% of ductal adenocarcinomas on CT have been reported to be non-contour deforming isoattenuating lesions [18]. Secondary signs such as a smoothly dilated or slightly beaded dilated upstream pancreatic duct and/or bile duct with abrupt tapering and distal pancreatic atrophy also may be present and easily demonstrated with MRCP. On MRI, a focal hypointense lesion may be seen on the arterial dominant phase images at the level of abrupt tapering. In other cases, the pancreatic head may be focally enlarged on CT without a hypoattenuating mass, upstream ductal dilation or pancreatic atrophy. In such

77

78

Pancreatic Cancer

cases, MRI may help confirm a normal variant or enable the diagnosis of other lesions such as islet cell tumors. Staging

In local staging of pancreatic ductal adenocarcinomas, morphological findings on MRI are similar to those on CT. The major vessels surrounding the pancreas, namely the SMV, PV, celiac axis and SMA are evaluated for evidence of circumferential tumor involvement. In the absence of a large body of validating literature, we adopt the CT staging system proposed by Lu et al. [12] and apply it to axial MR images. Thus the chances of successful resection decrease with increasing percentage circumferential involvement of the tumor surrounding the vessels. Signs of tumor-related compression of the major veins such as the teardrop sign may be seen on MRI. Distant staging

For detection of liver metastases, both CT and MR imaging are capable of demonstrating lesions larger than 1 cm in diameter with high sensitivity. However, for lesions less than 1 cm, especially serosal and omental lesions, sensitivity for both MR and CT remains unacceptably low. For nodal metastases, there are no reliable diagnostic MR or CT criteria, as enlarged nodes may be due to inflammatory or malignant causes. Endoscopic ultrasound

Endoscopic ultrasound (EUS) (see Chapter 8) is performed by an experienced gastroenterologist using a high frequency endoscopic transducer enabling examination of almost the entire pancreas. Endoscopic ultrasound is a valuable adjunct to MDCT and dynamic contrast-enhanced MR in diagnosis of small pancreatic cancers. In the majority of patients it is unnecessary as clinical symptoms, signs of obstructing jaundice, elevated serum Ca 19-9 and typical imaging findings suggest the diagnosis. However, in a minority of cases, especially those with a small isoattenuating or isointense lesion on MDCT or MRI, EUS may help demonstrate the obstructing lesion. Some studies have shown EUS to be better than MDCT in detecting small pancreatic tumors, although in these studies the ultrasonographer was not blinded to the MDCT findings [29,30]. DeWitt et al. detected 98% of lesions on EUS as compared with 86% on MDCT [30]. Another advantage of EUS is that it readily enables cytological confirmation of the primary lesion and of any suspicious nodal metastases with an FNA. Endoscopic ultrasound has a high negative

Imaging evaluation of pancreatic ductal adenocarcinoma

predictive value for excluding the diagnosis of primary pancreatic cancer [31] but not subtle peritoneal or liver metastases from a known pancreatic cancer. Thus a negative or normal EUS essentially excludes primary pancreatic cancer. Positron emission tomography with computed tomography

Positron emission tomography (PET) has become critical for detection of metastatic disease and intrinsic tumor activity in a variety of solid malignancies. As most malignancies rely on glycolysis for energy production, the relative consumption of glucose is up to 19 times higher than for most normal tissues that rely on the Krebs cycle. FDG ([18]Fluorodeoxyglucose), a glucose analog, is concentrated to a much greater extent in many glycolysis-dependent malignant tissues than in normal or inflammatory tissues. The normal pancreas in the fasting state has low background glucose usage compared to surrounding organs [32]. On PET scans, pancreatic adenocarcinomas up to 4 cm in diameter have been noted to have higher metabolic activity compared to lesions larger than 4 cm, which have less activity [33]. The development of hybrid PET-CT scanners has enabled combination of the anatomic multiplanar high-resolution imaging capability of CT with the metabolic imaging capability of positron emission tomography (PET). The role of PET-CT in the detection, staging and characterization of pancreatic ductal adenocarcinoma is currently being defined. Early studies evaluating PET alone generally have demonstrated a high level of performance for detecting pancreatic cancer [32–34]. However, PET alone cannot be used for local staging due to its inherently low spatial resolution. Its most practical role is to complement a high-quality multiphasic pancreatic CT and improve detection not only of small isodense pancreatic adenocarcinomas but also of small liver, lung, lymph node and bone metastases, thus precluding unnecessary surgery. In one study of 104 patients with pancreatic adenocarcinoma, who underwent separate PET and CT, blinded review of the fusion CT and PET images improved detection sensitivity from 76% (CT) and 84% (PET) to 89% (fusion). Ten of 31 proven lymph nodes with metastases were detected. On fusion images, an additional five patients with intra-abdominal metastases were detected as compared with CT alone [32]. Prior work has shown that PET activity is inversely correlated with prognosis – higher metabolic activity lesions having a poorer prognosis. PET may also be used to assess response to neoadjuvant chemotherapies and/or radiation [32]. Several different parameters of PET activity have been used to distinguish chronic pancreatitis, which tends to have lower PET activity, from ductal

79

80

Pancreatic Cancer

adenocarcinoma, which tends to have higher overall and progressively greater PET activity. However, there is enough overlap in the degree of PET activity between inflammatory conditions such as acute, chronic and autoimmune pancreatitis and pancreatic malignancies such as adenocarcinomas that it is not reliable enough to provide a diagnosis [32]. Positron emission tomography activity also can be affected by a number of factors, with false negative diagnoses reported in non-fasting patients and in those with hyperglycemia, which can decrease 18-FDG activity. False positive results have been described from a variety of causes resulting from chronic or autoimmune pancreatitis, nasobiliary tube insertion and other causes. Thus presently, there is no consensus on incorporating PET-CT into the routine imaging algorithm for detecting and staging pancreatic cancer. It is presently best reserved for detection of suspected distant metastases.

Conclusion Preoperative imaging for detection, staging and display of pancreatic ductal adenocarcinoma has developed considerably over the past ten years. Multidetector row computed tomography with a carefully chosen protocol remains the primary imaging method for this purpose, although dynamic contrastenhanced MRI is a competitive alternative and may be complementary to CT in detecting up to 10% of cancers that are isoattenuating on CT. Once acquired, image data sets may be displayed using a variety of methods to help surgeons understand the relationship of the cancer to surrounding structures. Endoscopic ultrasound may help detect and cytologically confirm small cancers. PET-CT may be useful for detecting primary pancreatic lesions as well as liver and lung metastases.

REFERENCES 1. American Cancer Society. Cancer Facts and Figures 2007: year 2007 Surveillance Research from the American Cancer Society. Bathesda, MD: American Cancer Society, 2007. 2. Warshaw A, Fernandez-Del Castillo C. Pancreatic carcinoma. N Engl J Med 1992; 326: 455–465. 3. Tsistos G, Farnell M, Sarr M. Are the results of pancreatectomy for pancreatic cancer improving? World J Surg 1999; 23: 913–919. 4. Warshaw A, Gu Z, Wittenberg J, Waltman A. Preoperative staging and assessment of resectability of pancreatic cancer. Arch Surg 1990; 125: 230–233.

Imaging evaluation of pancreatic ductal adenocarcinoma

5. Pasanen P, Eskelinen M, Partanen K, et al. A prospective study of the value of imaging, serum markers and their combination in the diagnosis of pancreatic carcinoma in symptomatic patients. Anticancer Res 1992; 12: 2309–2314. 6. Clarke D, Thomson S, Madiba T, et al. Preoperative imaging of pancreatic cancer: a management oriented approach. J Am Coll Surg 2003; 196: 119–129. 7. Horton K. Multidetector CT and three dimensional imaging of the pancreas; state of the art. J Gastrointest Surg 2002; 6: 126–128. 8. Gouma D, van Geenen R, van Gulik T, et al. Rates of complications and death after pancreaticoduodenectomy: risk factors and impact of hospital volume. Ann Surg 2000; 232: 786–795. 9. Freeny PC, Marks WM, Ryan JA, Traverso LW. Pancreatic ductal adenocarcinoma: Diagnosis and staging with dynamic CT. Radiology 1998; 166: 125–133. 10. Lu D, Vendantham S, Krasny R, et al. Two phase helical CT for pancreatic tumors: pancreatic versus hepatic phase enhancement of tumor, pancreas and vascular structures. Radiology 1996; 199: 697–701. 11. Boland G, O’Malley M, Saez M, et al. Pancreatic phase versus portal vein phase helical CT of the pancreas: optimal temporal window for evaluation of pancreatic adenocarcinoma. Am J Roentgenol 1999; 172; 605–608. 12. Lu D, Reber H, Krasny R, Kadell B, Sayre J. Local staging of pancreatic cancer: criteria for unresectability of major vessels as revealed by pancreatic–phase, thin section helical CT. Am J Radiol 1997: 168: 1439–1443. 13. O’Malley M, Boland G, Wood B, et al. Adenocarcinoma of the head of the pancreas: determination of unresectability with thin section pancreatic phase helical CT. Am J Roentgenol 1999; 173: 1513–1518. 14. Vargas R, Nino Murcia M, Trueblood W, Jeffrey RB. MDCT in pancreatic adenocarcinoma: prediction of vascular invasion and resectability using a multiphasic techinque with curved planar reformations. Am J Radiol 2004; 182: 419–425. 15. Fishman E, Horton K, Urban B. Multidetector CT angiography in the evaluation of pancreatic carcinoma: preliminary observations. J Comput Assist Tomogr 2000; 245: 849–853. 16. McNulty N, Francis I, Platt J, et al. Multi-detector row helical CT of the pancreas: effect of contrast enhanced multiphasic imaging on enhancement of the pancreas, peripancreatic vasculature, and pancreatic adenocarcinoma. Radiology 2001; 220: 97–102. 17. Richter G, Wunsch C, Schnider B, et al. Hydro-CT in detection and staging of pancreatic carcinoma. Radiology 1998; 38: 279–286. 18. Prokesch R, Chow L, Beaulieu C, Bammer R, Jeffrey R. Isoattenuating pancreatic adenocarcinoma at multidetector row CT: secondary signs. Radiology 2002; 224: 764–768. 19. Vedantham S, Lu D, Reber D, Kadell B. Small peripancreatic veins: improved assessment in pancreatic cancer patients using thin section pancreatic phase helical CT. Am J Radiol 1998; 170: 377–383.

81

82

Pancreatic Cancer

20. Bipat S, Phoa SS, van Delden OM, et al. Ultrasonography, computed tomography, and magnetic resonance imaging for diagnosis and determining resectability of pancreatic adenocarcinoma. J Comput Assist Tomogr 2005; 29: 438–445. 21. Lee D, Raman S, Kadell B, et al. Pancreatic phase multidetector CT angiography: performance in assessment of surgical resectability of pancreatic adenocarcinoma in 46 consecutive patients. [Abst.] RSNA 2007, Chicago, IL. 22. Zamboni, G, Kruskal J, Vollmer C, et al. Value of MDCT angiography in preoperative evaluation of pancreatic adenocarcinoma. Radiology 2007; 245: 770–778. 23. Weg N, Scheer M, Gabor M. Liver lesions: improved detection with dual-detector array CT and routine 2.5 m thin collimation. Radiology 1998; 209: 417–426. 24. Miller F, Rini N, Keppke A. MRI of adenocarcinoma of the pancreas. Am J Roentgenol 2006; 187: W365–372 (web exclusive article). 25. Ichikawa T, Haradome H, Hachiya J, et al. Pancreatic ductal adenocarcinoma: preoperative assessment with helical CT versus dynamic MR imaging. Radiology 1997; 202: 655–662. 26. Irie H, Honda H, Kaneko K, et al. Comparison of helical CT and MR imaging in detecting and staging small pancreatic adenocarcinoma. Abd Imaging 1997; 22: 429–433. 27. Obuz F, Dicle O, Coker A, Sagol O, Karademir S. Pancreatic adenocarcinoma: detection and staging with dynamic MR imaging. Eur J Radiol 2001; 38: 146–150. 28. Shima W et al. Diagnosis and staging of pancreatic cancer: comparison of mangofodipir enhanced MRI and contrast enhanced hydro CT. Am J Radiol 2002; 22: 429–433. 29. Leggmann P, Vignaux O, Dousset B, et al. Pancreatic tumors: comparison of dual phase helical CT and endoscopic sonography. Am J Roentgenol 1998; 170: 1315–1322. 30. DeWitt J, Devereaux B, Chriswell M, et al. Comparison of endoscopic ultrasonography and multidetector computed tomography for detecting and staging pancreatic cancer. Ann Int Med 2004; 141: 753–763. 31. Klapman J, Chang K, Lee J, Nguyen P. Negative predictive value of endoscopic ultrasound in a large series of patients with a clinical suspicion of pancreatic cancer. Am J Gastroenterol 2005; 100: 2658–2661. 32. Kalra M, Maher M, Boland G, Saini S, Fischman A. Correlation of positron emission tomography and CT in evaluating pancreatic tumors: technical and clinical implications. Am J Roentgenol 2003: 181: 387–393. 33. Friess H, Langhans J, Ebert M, et al. Diagnosis of pancreatic cancer by 2(18F)-fluoro-2-deoxy-Dglucose positron emission tomography. Gut 1995; 36: 771–777. 34. Ho C, Dehdashiti F, Griffeth L, et al. FDG PET evaluation of indeterminate pancreatic masses. J Comput Assist Tomogr 1996; 20 363–369. 35. Lytras D, Connor S, Bosonnet L, et al. Positron emission tomography does not add to computed tomography for the diagnosis and staging of pancreatic cancer. Dig Disease 2005; 22: 55–61. 36. Quon A, Chang ST, Chin F, et al. Initial evaluation of 18F-fluorothymidine (FLT) PET CT scanning for pancreatic adenocarcinoma. Eur J Nuc Med Mol Imaging 2007; 35: 527–531.

6 Imaging evaluation of cystic pancreatic neoplasms Dushyant Sahani and Onofrio A. Catalano

Introduction Cystic lesions of the pancreas have been increasingly diagnosed in recent years and often are discovered at a smaller size than in the past due to increased awareness and improved diagnostic imaging technology. Although all pancreatic tumors, including ductal adenocarcinomas, may undergo central necrosis and appear predominantly cystic, the term cystic neoplasms usually is applied to epithelial lined neoplastic cysts. Pancreatic cystic neoplasms are classified mainly on the basis of their epithelial lining and encompass a wide spectrum of different pathologic entities. Serous cystic neoplasm (SCN), mucinous cystic neoplasm (MCN) and intraductal papillary mucinous neoplasm (IPMN) account for more than 90% of these cases. The remaining 10% are represented by solid pseudopapillary epithelial neoplasm (SPEN), cystic pancreatic endocrine neoplasms (PEN), acinar cell cystadenocarcinomas, cystic metastases, and a few other even rarer tumors [1]. The clinical presentation of pancreatic cystic neoplasms is variable, and a substantial number of pancreatic cysts are discovered incidentally during imaging work-up for an unrelated medical problem. When symptomatic, patients often present with abdominal pain, early satiety, vomiting or jaundice. Occasionally they can present with recurrent pancreatitis when the lesion obstructs or communicates with the pancreatic duct. Advanced cystic malignancies may present with symptoms similar to pancreatic adenocarcinoma, i.e. pain, weight loss and jaundice [1–3]. Many cystic pancreatic masses are resected surgically because of the recognized malignant potential of some cystic neoplasms. However, cystic pancreatic neoplasms represent a spectrum from benign to malignant, and benign lesions are Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

84

Pancreatic Cancer

recognized with increasing frequency. Cyst morphology on imaging studies can offer clues to the histopathologic subtype and the potential risk of malignancy. Therefore diagnostic imaging is required to help discriminate between pseudocysts and cystic neoplasms and to distinguish benign from malignant lesions. Although cyst management often is dictated by the patient’s age and clinical presentation, the cyst morphology also is a critical determinant. For example a small cystic lesion discovered incidentally in an elderly patient, can be safely followed [1–4]. Although many diagnostic imaging methods can be used to evaluate cystic pancreatic lesions, multi-detector row computed tomography (MDCT) and magnetic resonance imaging (MRI) with magnetic resonance cholangiopancreatography (MRCP) are the most reliable and commonly used techniques. Both MDCT and MRI/MRCP provide high spatial and contrast resolution images that are acquired dynamically and can be reconstructed using 3-D post processing techniques. Therefore cyst morphology, the pancreatic duct and ancillary findings can be evaluated to enable diagnosis and preoperative planning. With MRCP and MDCT, the entire course of the main pancreatic duct (MPD) and its relationship with the cystic lesion can be displayed. In addition, the anatomic relationships of the cystic lesion to adjacent structures can be demonstrated. Because of the capabilities of MRCP and MDCT pancreatograms, invasive ERCP is reserved for selected cases. Transabdominal ultrasonography (US) although less expensive and widely available, is operator dependent, non-reproducible and limited by patient body habitus and abdominal gas overlying the pancreas. Endoscopic ultrasonography (EUS), on the other hand, offers high-resolution images of the pancreas and therefore provides detailed assessment of cyst morphology. Endoscopic ultrasonography also enables cyst fluid aspiration and sampling of the cyst wall or mural nodules. Cystic fluid analysis can provide diagnostic information about the nature of the cyst. High viscosity or extra-cellular mucin is indicative of a mucinous neoplasm, whereas a high amylase level indicates communication of the cyst with the pancreatic duct. Although amylase is present in both pseudocysts and intraductal papillary mucinous neoplasms, very high amylase levels are suggestive of a pseudocyst. Tumor markers also can support the diagnosis of malignancy in some cases. This chapter will focus primarily on the commonly encountered cystic lesions of the pancreas, including serous cystic neoplasms, mucinous cystic neoplasms and intraductal papillary neoplasms [1–5].

Figure 2.1 Ductal adenocarcinoma involving the body and tail of the pancreas. Tumor infiltrates diffusely into pancreatic parenchyma and extends into peripancreatic soft tissue (inferior).

Figure 2.2 Ductal adenocarcinoma of the pancreas. Glands of irregular size and shape are lined by cells with variably sized nuclei. (Original magnification 66×)

Figure 2.3 Fine needle aspiration of ductal adenocarcinoma. Three dimensional group of cells with marked variability in nuclear size and shape. (Pap stain, 132×)

Figure 2.4 Anaplastic carcinoma of the pancreas. Discohesive cells with very large, bizarrely shaped nuclei. (132×)

Figure 2.5 Osteoclast-like giant cell tumor. Multinucleated giant cells mixed with epithelioid tumor cells. (132×)

Figure 2.6 Medullary carcinoma. Solid sheets of tumor cells with many tumor-infiltrating lymphocytes. (66×)

Figure 2.7a PanIN 1 A. Tall mucinous cells with regular, basally-oriented nuclei.

Figure 2.7b PanIN2. The nuclei are crowded and elongated.

Figure 2.7c PanIN 3. Complex papillary architecture and large, dark, irregular nuclei. (66×)

Figure 2.8 Intraductal mucinous papillary neoplasm involving most of the main pancreatic duct with extension in side branches.

Figure 2.9a Histologic subtypes of IPMN. A: gastric-type mucosa with mucous glands.

Figure 2.9b Intestinal type, lined by tall columnar epithelium.

Figure 2.9c Pancreatobiliary type, with cuboidal cells.

Figure 2.9d Oncocytic type, featuring abundant pink cytoplasm.

Figure 2.10 Mucinous cystic neoplasm. A single layer of columnar epithelium above ovarian-like stroma. (66×)

Figure 2.11 Acinar carcinoma. Gland-like spaces lined by cuboidal cells with round, regular nuclei. (132×)

Figure 2.12 Solid pseudopapillary tumor. Sheets of cells traversed by slender capillaries. Many cells have intracytoplasmic globules. (132×)

Figure 2.13 Pancreatic neuroendocrine tumor. Bland cells with trabecular and pseudacinar growth.

below: Figure 2.14 Serous microcystic adenoma. Sponge-like cut surface with central fibrosis.

Figure 8.10 Cytology obtained from aspiration of a cyst. Fine needle aspiration during EUS can aid the clinician in determining the etiology of a cyst. This slide demonstrates undulating sheets of ductal epithelial cells with uniformly round to oval nuclei in an organized honeycomb pattern. These cytologic features can be seen in mucinous cystic lesions of the pancreas. (Courtesy of Lourdes Ylagan, MD)

Imaging evaluation of cystic pancreatic neoplasms

Serous cystic neoplasms Serous cystic neoplasms (SCNs) comprise 30–39% of all pancreatic cystic neoplasms and are regarded as slowly growing benign lesions with very low malignant potential [1, 4]. They occur predominantly in women (75%) and are diagnosed at a mean age of 62 years [1, 4]. Serous cystic neoplasms appear as well-circumscribed, lobulated, round, cystic masses, containing watery fluid that is devoid of mucin. They usually lack a capsule, contain no peripheral wall calcification, and do not communicate with the pancreatic duct. They typically have a microcystic “sponge-like” or “honeycomb-like” morphology, composed of innumerable small cysts, usually in the range of a few mm. A central stellate fibrous scar, with some calcifications, is found in up to 30% of cases and is considered specific for SCN. Larger cysts, if present, usually are less than 2 cm and peripheral. Uncommonly (10%) SCNs can be macrocystic (oligocystic), composed of a single cyst or a small number of large cysts. The macrocystic variety is encountered in younger individuals. The epithelial lining of SCNs is composed of cuboidal or flat, glycogen-rich cells which stain with periodic acid of Schiff (PAS positive) [1, 4, 6, 7]. On MDCT and MRI, the imaging appearance of SCNs closely parallels that of their macroscopic pathology. Fine, external lobulations and enhancement of septa and cyst wall are common features that are best appreciated in the portal venous phase of contrast enhancement. A central fibrous scar may have a characteristic stellate pattern of calcification and may enhance on delayed phase images (Figures 6.1 and 6.2). Dilation of the main pancreatic duct (MPD) is uncommon, unless the duct has been compressed by a large SCN. Serous cystic neoplasms most commonly are well-marginated lesions that have a “honeycomb” appearance, containing numerous small cysts surrounded by a matrix of enhancing septa. Thus the lesions may have soft tissue or mixed attenuation. Occasionally, SCNs can be difficult to differentiate from solid pancreatic masses on MDCT. Magnetic resonance imaging however can better display the microcystic nature of the lesions as numerous discrete high signal intensity foci on T2-weighted sequences. Macrocystic SCNs may be difficult to differentiate from mucinous cystic tumors. However, a lobulated contour or a central scar is supportive of SCN [1, 2, 4, 6, 8, 9]. Because SCNs are considered to be benign lesions, the decision to operate is based on size, clinical presentation and lesion location. Usually lesions more than 4 cm are resected in younger patients due to estimated growth of 4–12 mm/ year [1–4].

85

86

Pancreatic Cancer

Figure 6.1. Typical CT features of serous cystic neoplasm are shown. Axial contrast-enhanced CT image shows a sharply demarcated, lobulated, cystic mass in the body of the pancreas. The lesion contains a central scar (arrow) and internal septa, delimiting cystic components less than 2 cm in diameter.

Mucinous cystic neoplasms Mucinous cystic neoplasms comprise 10–45% of all pancreatic cystic neoplasms and encompass a wide range of biological behavior, from benign adenomas to frankly invasive adenocarcinomas [1, 10]. Mucinous cystic neoplasms occur almost exclusively in middle-age women (mean age 47 years), with only a few reported cases in men. They usually are located in the tail (72%) or body of the pancreas (13%), less frequently in the head (5%). Rarely MCN may occur as a diffuse lesion replacing the entire pancreas (9%) [11, 12]. Mucinous cystic neoplasms usually appear as large (average diameter 6–10 cm) round or oval, cystic masses, circumscribed by a fibrous pseudocapsule, which may contain calcifications. Typically they are multi-locular and contain thick mucin. Occasionally however, they are unilocular. They may also contain hemorrhage. The

Imaging evaluation of cystic pancreatic neoplasms

(a)

(b)

(c)

(d)

Figure 6.2. Serous cystic neoplasm in a 42-year-old man. (a) Axial contrast-enhanced CT image shows a microcystic lesion, with internal calcifications (arrow), in the tail of the pancreas. (b) Curved reformatted contrast-enhanced CT image displays the microcystic morphology of the SCN and its relationships to the pancreatic duct. Central calcification is indicated by an arrow. (c) T2-weighted fat saturated MR image shows a dark central scar (arrow) and radiating septa (arrowheads) against the hyperintense microcystic fluid. (d) Contrast-enhanced fat saturated T1-weighted image shows enhancement of the internal septa (arrowheads).

87

88

Pancreatic Cancer

cyst wall is composed of an ovarian-like stroma lined by mucin-producing epithelium that may display different grades of dysplasia, according to which lesions are classified as adenoma, borderline tumor or carcinoma [1, 7, 12]. Malignant MCNs contain soft tissue components such as mural nodules or papillary projections, whereas benign MCNs have a smooth internal surface. Invasive adenocarcinomas are found in 33% of the cases. Mucinous cystic neoplasms generally do not communicate with the pancreatic duct. On MDCT and MRI/MRCP, MCNs appear as multilocular macrocystic lesions, with individual compartments usually larger than 2 cm in diameter. The cyst fluid appears hyperintense on T2-weighted sequences and hypointense to minimally hyperintense on T1-weighted sequences. The internal architecture of the cysts including septations and internal wall nodularity is well appreciated on both MDCT and MRI. The infrequent finding of a peripheral eggshell calcification on CT is specific for mucinous cystic neoplasms and predictive of malignancy (Figures 6.3–6.5). In a minority of cases MCNs are unilocular. Patient age, cyst morphology and absence

(a)

(b)

Figure 6.3. Mucinous cystic neoplasm in a 47-year-old woman. (a) Fast Imaging Employing Steady State Acquisition (FIESTA) and (b) T1-weighted fat saturated gadolinium enhanced image display a well-demarcated, encapsulated lesion in the tail of the pancreas (arrow). Due to its fluid content it is hyperintense on FIESTA and hypointense on T1. An internal septum (arrowhead), visible in (a) demarcates a large cystic loculation (*), more than 2 cm in diameter, whose signal intensity on the T1-weighted image is slightly higher than that of the smaller loculation (thin arrow), probably due to the high mucin content.

Imaging evaluation of cystic pancreatic neoplasms

(a)

(b)

Figure 6.4. Mucinous cystic neoplasm. (a) Axial T1-weighted out of phase and (b) axial T2-weighted single shot fast spin echo images display an encapsulated, large, oval lesion (arrow) in the tail of the pancreas. Internal septa (arrowheads), dividing the lesion into large cystic components, are better appreciated in (b). Some of the loculations (double arrows) appear relatively hypointense in (b), due to differences in mucin composition of the cyst fluid.

of communication with the pancreatic duct are important features to differentiate MCNs from IPMNs [1, 9, 10, 12–15]. Although differentiation of benign from malignant lesions is not always possible based on imaging features, wall thickening or irregularity, mural nodules, papillary projections and peripheral calcification favor malignancy [1, 9, 10, 12, 14, 15]. Because of their malignant potential and the relatively young age at presentation, MCNs are treated by surgical resection, unless there is a contraindication to surgery. Even in patients with invasive malignant mucinous cystic neoplasm the postoperative 5-year survival rate is 38%, which far exceeds survival of patients with pancreatic adenocarcinoma. Patients with less aggressive lesions have excellent survival, and recurrence is rare (Figure 6.6) [1, 7, 11–13].

Intraductal papillary mucinous neoplasms Intraductal papillary mucinous neoplasms of the pancreas (IPMNs) comprise 21–33% of all pancreatic cystic neoplasms [15, 16]. They occur more often in men, and the mean age at presentation is 65.5 years. Most are located in the pancreas head

89

90

Pancreatic Cancer

(a)

(b)

Figure 6.5. Mucinous cystic neoplasm. Axial contrast-enhanced CT image (a), and corresponding T2-weighted fat saturated (b) and axial T1-weighted fat saturated contrast-enhanced (c) images display a large oval lesion in the tail of the pancreas. The cyst fluid is hypoattenuating on CT, hypointense on T1-weighted and hyperintense on T2-weighted images. A thick capsule, circumscribing the lesion, is marked by arrows.

(c)

(58%) or body (23%), less commonly in the tail (7%). In 12% of the cases the pancreas is diffusely involved [15, 16]. Intraductal papillary mucinous neoplasms are characterized by intraductal papillary growths of mucin-producing neoplastic cells, and they tend to progress through an adenoma-carcinoma sequence analogous to colorectal cancer. Consequently,

Imaging evaluation of cystic pancreatic neoplasms

(a)

(b)

Figure 6.6. Metastases from mucinous cystic neoplasm. Representative images of the upper (a) and lower (b) abdomen in a 54-year-old woman who underwent surgical resection of a malignant MCN show liver metastases (arrows in a) and peritoneal carcinomatosis (arrowheads in a and b).

IPMNs can have a wide range of histopathologies, ranging from hyperplasia to adenoma, carcinoma in situ and invasive carcinoma. Different stages of evolution of the tumor can coexist in the same patient. IPMNs are classified as main duct, branch duct or combined, according to the site of pancreatic duct involvement. In the case of main pancreatic duct IPMNs, the thick mucin secreted by the tumors usually results in dilation of the pancreatic duct, either focally or diffusely. Long-standing lesions also can result in low-grade obstruction with changes of chronic pancreatitis. Branch duct IPMNs occur most commonly in the head or uncinate process (about 60%) in the form of a single cystic structure or a grape-like cluster of communicating cysts, median size 20 mm (range 11–40 mm), separated by septa and containing fluid, mucin and tumor cells. They are connected to the pancreatic duct, which may be dilated (Figure 6.7) [13, 15–23]. Often both side-branch and MPD lesions coexist as independent or combined lesions. In the case of main duct IPMNs, MDCT and MRI reveal diffuse or segmental dilation of the MPD, without a transition point. Often the entire MPD is dilated to the papilla, and there is associated pancreatic parenchymal atrophy. In the case of branch duct IPMNs, one or more lobulated and septated cystic lesions are depicted (Figure 6.8). MRI/MRCP, MDCT with multiplanar reconstructions, EUS and ERCP all can demonstrate the morphologic details of IPMNs including communication with the pancreatic duct, extent of duct involvement, septa and mural nodules if present. Neoplastic mural nodules appear as areas of low signal intensity on T2-weighted

91

92

(a)

Pancreatic Cancer

(b)

Figure 6.7. Side branch intraductal papillary mucinous neoplasms. Coronal reformatted contrast-enhanced CT image (a), coronal T2-weighted single shot fast spin echo image (b), and coronal 3-D MRCP image (c) display a well-defined lobulated cystic lesion in the neck of the pancreas (arrowheads). A channel-like communication of the cystic lesion with the pancreatic duct, a diagnostic feature of IPMN, is indicated by the arrow.

(c)

MR images, exhibiting enhancement after the injection of gadolinium-based contrast agents. Lack of contrast enhancement is characteristic of mucin or calcification, which appear dark on T2-weighted images. MRCP is preferred to study IPMNs because ERCP may be unable to demonstrate a communication between branch duct IPMNs and the pancreatic duct, due to mucous plugging.

Imaging evaluation of cystic pancreatic neoplasms

Figure 6.8. Multiple side-branch intraductal papillary mucinous neoplasms. Coronal 3-D MRCP image displays multiple well-defined septated cystic lesions (arrowheads) scattered along the main pancreatic duct, with channel-like communications with the pancreatic duct (arrows).

With both the branch duct and MPD varieties, the dilated duodenal papilla may protrude into the duodenal lumen [2, 5, 13, 15–17, 21, 24]. Main duct IPMNs often present with low-grade pancreatitis symptoms due to thick mucin occluding the pancreatic duct. Malignant lesions can result in jaundice. The risk of malignancy in main duct IPMNs is high (57–92%), with invasive features in 50% (Figures 6.9 and 6.10). Surgery is usually advocated for MPD IPMNs, as the 5-year survival rate is 80%. Branch duct IPMNs on the other hand usually are detected incidentally and have a lower risk of malignancy, especially in lesions less than 3 cm in diameter (6–46%). Moreover, about 85% of branch duct IPMNs, if devoid of mural nodules, remain stable over time. Therefore, close follow-up can be undertaken for a small (< 3 cm) asymptomatic side branch IPMN that is devoid of signs of malignancy [1, 2, 15, 16].

93

94

Pancreatic Cancer

Figure 6.9. Malignant main duct intraductal papillary neoplasm. Coronal oblique 2-D MRCP image displays massive dilation of the entire main pancreatic duct (thin arrows). Some side branches are distended (arrowheads). A lobulated cystic lesion in the tail, due to associated branch duct IPMN, is indicated by the thick arrow. The extensive involvement and marked dilation of the pancreatic duct are suggestive of malignancy, a finding confirmed at surgery.

Strong predictors of malignancy are MPD diameter exceeding 9 mm, mural nodules, and signs of invasion. For side branch IPMN, moderate predictors of malignancy are diameter greater than 4 cm, thick septa, and irregular wall.

Solid pseudopapillary epithelial neoplasm (also called solid pseudopapillary tumor) Solid pseudopapillary epithelial neoplasms (SPEN) are rare pancreatic tumors (9% of cystic neoplasms) with low malignant potential. They generally are locally aggressive, but metastases have been reported. SPENs occur mostly in non-Caucasian, young (mean age 27 years) women (F:M, 9.5:1), usually in the body or tail of the pancreas. They usually present as large (mean diameter 5–9 cm), well-circumscribed masses,

Imaging evaluation of cystic pancreatic neoplasms

(a)

(b)

Figure 6.10. Combined malignant intraductal papillary mucinous neoplasm. Axial portal venous phase contrast-enhanced CT (a) and 2-D coronal oblique MRCP (b) images demonstrated massive dilation of the main pancreatic duct (arrowheads) and of side branches in the head and uncinate process (arrow).

containing an admixture of solid, cystic and papillary components along with hemorrhagic and necrotic areas. On histology, pseudopapillary architecture is observed [25, 26]. Calcifications, predominantly peripheral, are present on pathology in approximately one-third of lesions but are identified uncommonly on imaging studies. On cross-sectional imaging, the appearance of SPENs varies from completely solid masses to almost cystic structures (Figure 6.11), with most lesions demonstrating both solid and cystic components. Internal fluid-debris levels, due to hemorrhagic products that appear hyperintense at both T1- and T2-weighted sequences, can be observed. The lesions show a variable degree of enhancement after contrast medium administration [25–27]. Infiltrative lesions may appear poorly marginated [25–27]. Surgical resection is advocated in all cases of SPENs because they tend to infiltrate locally and metastasize distantly, mainly to the liver and lymph nodes. Even when

95

96

Pancreatic Cancer

Figure 6.11. Metastatic solid pseudopapillary epithelial neoplasm in a 24-year-old woman. Coronal reformatted contrast-enhanced CT image displays a large, well-defined lesion (thin arrows) arising from the body of the pancreas. The mass contains peripheral solid enhancing areas (arrowheads) and large non-enhancing fluid components. A metastasis in the liver is pointed out by the thick arrow.

aggressive behavior is encountered, resection of the primary tumor and liver metastases is recommended. The 5-year survival rate is 95% [25, 26].

Cystic pancreatic endocrine neoplasm Pancreatic endocrine neoplasms (PENs) are among the solid pancreatic neoplasms that can present as cystic lesions. Cystic PENs account for about 2% of all pancreatic cystic lesions, and are classified as either hyperfunctioning or nonhyperfunctioning, on the basis of hormone overproduction. Hyperfunctioning tumors may be associated with various syndromes, including multiple endocrine neoplasia type 1 (MEN 1), Von Hippel–Lindau and neurofibromatosis (NF). They show a slight female predominance and occur at a mean age of 55 years [28]. Cystic degeneration of a PEN likely is related to insufficient blood supply resulting in necrosis and hemorrhage. Such change occurs more often with the nonhyperfunctioning endocrine neoplasms (30%). Cystic PENs may be single or

Imaging evaluation of cystic pancreatic neoplasms

(a)

(b)

Figure 6.12. Cystic pancreatic endocrine neoplasm. Axial contrast-enhanced CT image (a), axial T1-weighted fat saturated contrast-enhanced (b) and axial T2-weighted fat saturated fast spin echo (c) images demonstrate a lobulated cystic (*) lesion in the body-tail of the pancreas, circumscribed by a thick highly enhancing peripheral rim (arrows).

(c)

multiple, and may be associated with one or more coexistent solid PENs. The lesions tend to be located in the body or tail of the pancreas and range in size from 2–10 cm (median 3.7 cm). On imaging, cystic PENs typically appear as well-defined, unilocular, thickwalled cystic masses or as mixed solid and cystic masses (Figure 6.12). Thick internal septations uncommonly may be seen. The most important imaging feature that differentiates cystic PENs from other cystic masses, is avid contrast enhancement of the cyst wall, solid component or septations when present. However, this feature is not universally present [28, 29].

97

98

Pancreatic Cancer

These tumors are locally invasive and also can metastasize to the liver and lymph nodes. Although most are non-malignant (78%) and non-hyperfunctional (79%), surgical resection is the treatment of choice for cystic PEN. The 5-year postoperative survival rate following surgery is excellent (96%) [28, 29].

Miscellaneous and rare lesions Every pancreatic tumor can undergo necrosis and cystic degeneration, mimicking the appearance of the primarily cystic neoplasms described above. In addition, non-neoplastic lesions such as pseudocysts (Figure 6.13) and lymphoepithelial cysts (LEC) can mimic cystic pancreatic neoplasms. Pancreatic adenocarcinomas that undergo cystic degeneration can appear as infiltrative masses with ill-defined margins, a thick irregular wall or a large solid component. They may invade into the retroperitoneum and encase peripancreatic vascular structures. Metastases to the liver, lymph nodes and peritoneum may be present. Lymphoepithelial cysts are extremely rare non-neoplastic cystic lesions that occur far more commonly in men than women (16:3), with a wide age range at

(a)

(b)

Figure 6.13. Pancreatic pseudocysts. Axial contrast-enhanced CT images from two different patients showing well-defined, encapsulated cystic lesions (arrow) arising from the pancreas. In (a) the pseudocyst compresses the posterior stomach wall, and in (b) peritoneal fat stranding, due to inflammation (arrowhead) is seen.

Imaging evaluation of cystic pancreatic neoplasms

presentation (32–73 years). They present as a single, lobulated, cystic mass, frequently containing internal septations. They are characterized by a lining of mature stratified squamous epithelium surrounded by abundant lymphocytic tissue and do not communicate with the pancreatic duct. Lymphoepithelial cysts appear hyperintense on T2-weighted and hypointense on T1-weighted MR images and may exhibit hypoattenuation or hyperattenuation at unenhanced CT. In some cases negative attenuation values have been reported. Preoperative diagnosis is difficult, even at cytopathology. Absence of microcysts helps to differentiate this lesion from SCNs. Lobulation and male predominance may be useful to differentiate them from MCN. Lymphoepithelial cysts do not infiltrate surrounding structures and do not recur after surgery.

Management Although the management of cystic pancreatic lesions is still evolving, surgical resection often is indicated for lesions having malignant potential or causing symptoms, provided that the surgical risk is acceptable for the individual patient [1, 3]. However, considering the high number of benign cystic lesions and the risks of morbidity and mortality from pancreatic surgery, resecting all cystic lesions would be inappropriate. When differentiation between a benign and malignant cyst is not possible based on imaging, EUS-guided cyst aspiration and/or surgical resection should be considered. Asymptomatic thin-walled unilocular cysts less than 3 cm in size commonly are benign and can be monitored safely with imaging. The accepted imaging time frame includes 6-month follow-up imaging for the first year, followed by annual imaging for a period of three years, and then every three years in patients less than 70 years of age. If cyst stability is established and the patient continues to be symptom free, no further evaluation may be needed [30]. On the other hand, an irregular thick-walled cyst often is indicative of an aggressive biology and should be treated more aggressively. Tumor location also influences patient management as cysts located in the head of the pancreas require a surgical procedure associated with greater risk of complications than those in the tail. Therefore, assessment of the risk–benefit ratio is imperative before deciding on treatment for an individual patient. Integrating imaging finding with demographics and clinical presentation is mandatory when evaluating cystic pancreatic lesions [1, 2]. A summary table and figure regarding the differential diagnosis of cystic pancreatic lesions are provided (Table 6.1, Figure 6.14).

99

Age *

M>F 6–7th

IPMN

F=M 5–6th

PEN

Head/tail/

Oval

Head > body/ Oval/Irregular

decade tail

M>F 5–7th

decade body

M>F 4–6th

Oval

Oval

(Combined)

type) Both

Wall

Tl High;

High

High; T2

No

pancreatitis) Obstructed.

thick

occasionally

enhancement

Syndromes

abdominal

back and

Weight loss,

Pancreatitis

(head)

High

Tl Low; T2 Thick, variable Yes

High

No

Endocrine

+/–

pain, jaundice

dilatation

Upstream

pancreatits)

Tl Low; T2 Thin/

Yes

contributory

Non-

contributory

Non-

contributory

Non-

uncommon

4–9 cm

(chronic

Dilated (chronic

Parenchymal,

pancreatitis)

Normal (acute

If malignant

If malignant

appear solid

(< 3 cm) may contributory

Non-

history

Clinical

Septa

component

solid

Cystic with

4–8 cm

High

component

enhancement

enhancement

Unilocular

components

Solid

Thick, variable Yes

Thick, strong

Cystic with

Sometimes

non-laminated T2 High

Peripheral,

calcification)

mucin plug

the case of

enhancement –

High Intraductal (in Tl Low/

Uniformly

thick

thick. Variable

Tl High/

High

low; T2

septal

Peripheral/

in 30%

Low; T2

Normal

Normal

intensity

Signal

Central stellate Tl Low; T2 –/ Occasionally No. Small

Calcification

Tl High/

2–10 cm

5–9 cm

type)

Dilated

compressed

Normal or rarely

compressed

Normal or rarely

duct (MPD)

Main pancreatic

solid

solid component

Cystic with

dilatation (Main component

> 5 mm (Main

solid

or diffuse MPD

1–4 cm

6–10 cm

5–11 cm

Size *

or cystic with (Branch Type)

Macrocystic

Macrocystic

stroma

Dense

Microcystic

appearance

Cystic

(Branch) Focal

Grape-like

Oval

Lobulated

Shape/borders

*Due to increased diagnostic imaging use and improved technology, lesions are currently discovered at a lower age and at a smaller size than reported in the literature. SCN, serous cystic neoplasm; MCN, mucinous cystic neoplasm; IPMN, intraductal papillary mucinous neoplasm; SPEN, solid pseudopapillary epithelial neoplasm; PEN, cystic pancreatic endocrine neoplasms.

oma

adenocarcin-

No

Body/tail

decade predilection

decade

2–3rd

F

SPEN

Cystic

Head/

Tail/body

decade uncinate

decade

4–5th

F

Pseudocysts

Head/body/

Location

decade tail

F>M 6–7th

Sex

MCN

SCN

Table 6.1.

Imaging evaluation of cystic pancreatic neoplasms

Microcystic Lobulation +++ Sharp Interface +++ Central Scar +–

Unilocular

–

IPMN

F/U– Intervention*

< 3 cm > 3 cm

F/U

Surgery > 4 cm Young age symptoms

EUS Aspriation and cystic fluid analaysis

High amylase No extracellular mucin

Cystic with solid components

Channel-like MPD communication + –

Clinical history/imaging Findings of Pancreatitis +

Macrocystic / septated

MCN

< 3 cm > 3 cm

F/U– Intervention* Surgery (unless contraindicated)

Low amylase Extracellular mucin CEA > 400 IU

* – Symptoms – Enlarging at F/U – Young Age – >4cm

Figure 6.14. Practical approach to pancreatic cystic lesions. Main duct IPMNs which are treated by surgery are not included in the diagram.

REFERENCES 1. Brugge WR, Lanwers GY, Sahani D, Fernandez-Castillo C, Warshaw AL. Cystic neoplasms of the pancreas. N Engl J Med 2004; 351(12): 1218–1226. 2. Sahani DV, Kadavigere R, Saokar A, et al. Cystic pancreatic lesions: a simple imaging-based classification system for guiding management. Radiographics 2005; 25(6): 1471–1484. 3. Sheehan MK, Beck K, Pickleman J, Aranha GV. Spectrum of cystic neoplasms of the pancreas and their surgical management. Arch Surg 2003; 138(6): 657–660; discussion 660–662. 4. Galanis C, et al. Resected serous cystic neoplasms of the pancreas: a review of 158 patients with recommendations for treatment. J Gastrointest Surg 2007; 11(7): 820–826. 5. Irie H, et al. MR cholangiopancreatographic differentiation of benign and malignant intraductal mucin-producing tumors of the pancreas. Am J Roentgenol 2000; 174(5): 1403–1408. 6. Goh BK, et al. Pancreatic serous oligocystic adenomas: clinicopathologic features and a comparison with serous microcystic adenomas and mucinous cystic neoplasms. World J Surg 2006; 30(8): 1553–1559.

101

102

Pancreatic Cancer

7. Zamboni GC, Pesci P, Brighenti A. Pathology of cystic tumors. In Procacci AJ, ed. Imaging of the Pancreas. Cystic and Rare Tumors. Berlin, Springer. 2003; 9–29. 8. Carbognin, GT, Petrella M, Fuini E, Procacci AJ. Serous cystic tumors. In Procacci AJ, ed. Imaging of the Pancreas. Cystic and Rare Tumors. Berlin, Springer. 2003; 31–55. 9. Sarr MG, et al. Primary cystic neoplasms of the pancreas. Neoplastic disorders of emerging importance – current state-of-the-art and unanswered questions. J Gastrointest Surg 2003; 7(3): 417–428. 10. Scott J, et al. Mucinous cystic neoplasms of the pancreas: imaging features and diagnostic difficulties. Clin Radiol 2000; 55(3): 187–192. 11. Goh BK, et al. A review of mucinous cystic neoplasms of the pancreas defined by ovarian-type stroma: clinicopathological features of 344 patients. World J Surg 2006; 30(12): 2236–2245. 12. Zamboni G, et al. Mucinous cystic tumors of the pancreas: clinicopathological features, prognosis, and relationship to other mucinous cystic tumors. Am J Surg Pathol 1999; 23(4): 410–422. 13. Suzuki Y, et al. Cystic neoplasm of the pancreas: a Japanese multiinstitutional study of intraductal papillary mucinous tumor and mucinous cystic tumor. Pancreas 2004; 28(3): 241–246. 14. Biasiutti CF, Venturini F, Pagnotta N, Schenal N, Procacci C. Mucinous cystic tumors, In Procacci AJ, ed. Imaging of the Pancreas. Cystic and Rare Tumors. Berlin, Springer. 2003; 57–74. 15. Tanaka M, et al. Clinicopathologic study of intraductal papillary-mucinous tumors and mucinous cystic tumors of the pancreas. Hepatogastroenterology 2006; 53(71): 783–787. 16. Tanno S, et al. Natural history of branch duct intraductal papillary-mucinous neoplasms of the pancreas without mural nodules: long-term follow-up results. Gut 2007. 17. Lim JH, Lee G, Oh YL. Radiologic spectrum of intraductal papillary mucinous tumor of the pancreas. Radiographics 2001; 21(2): 323–337; discussion 337–340. 18. Loftus, EV Jr. et al. Intraductal papillary-mucinous tumors of the pancreas: clinicopathologic features, outcome, and nomenclature. Members of the Pancreas Clinic, and Pancreatic Surgeons of Mayo Clinic. Gastroenterology, 1996; 110(6): 1909–1918. 19. Pais SA, Attasaranya S, Leblanc JK, et al. Role of endoscopic ultrasound in the diagnosis of intraductal papillary mucinous neoplasms: correlation with surgical histopathology. Clin Gastroenterol Hepatol 2007; 5(4): 489–495. 20. Tanaka M. Intraductal papillary mucinous neoplasm of the pancreas: diagnosis and treatment. Pancreas 2004; 28(3): 282–288. 21. Tanaka M, Kobayashi K, Mizumoto K, Yamaguchi K. Clinical aspects of intraductal papillary mucinous neoplasm of the pancreas. J Gastroenterol 2005; 40(7): 669–675. 22. Taouli B, et al. Intraductal papillary mucinous tumors of the pancreas: helical CT with histopathologic correlation. Radiology 2000; 217(3): 757–764. 23. Zamboni, GC, Bogina P, Pesci A, Brighent A. Pathology of intraductal cystic tumors. In Procacci AJ, ed. Imaging of the Pancreas. Cystic and Rare Tumors. 2003, Berlin, Springer. 2003; 85–95. 24. Procacci CS, Della Chiara G, Fuini E, Guarise A. Intraductal papillary mucinous tumors: imaging. In Procacci AJ, ed. Imaging of the Pancreas. Cystic and Rare Tumors. Berlin, Springer. 2003; 97–137.

Imaging evaluation of cystic pancreatic neoplasms

25. Geers C, et al. Solid and pseudopapillary tumor of the pancreas – review and new insights into pathogenesis. Am J Surg Pathol 2006; 30(10): 1243–1249. 26. Hernandez JM, Centeno BA, Kelley ST. Solid pseudopapillary tumors of the pancreas: case presentation and review of the literature. Am Surg 2007; 73(3): 290–293. 27. Casadei R, et al. Pancreatic solid-cystic papillary tumor: clinical features, imaging findings and operative management. J Pancreas 2006; 7(1): 137–144. 28. Goh BK, et al. Clinico-pathological features of cystic pancreatic endocrine neoplasms and a comparison with their solid counterparts. Eur J Surg Oncol 2006; 32(5): 553–556. 29. Ligneau B, et al. Cystic endocrine tumors of the pancreas: clinical, radiologic, and histopathologic features in 13 cases. Am J Surg Pathol 2001; 25(6): 752–760. 30. Fernandez-del Castillo C, Targarona J, Thayer SP, et al. Incidental pancreatic cysts: clinicopathologic characteristics and comparison with symptomatic patients. Arch Surg 2003; 138(4): 427–433; discussion 433–434.

103

7 Imaging evaluation of pancreatic neuroendocrine neoplasms Ruedi F. Thoeni

Introduction Neuroendocrine neoplasms (NEN) or islet cell tumors of the pancreas are rare neoplasms that produce and secrete variable amounts of hormones. These neoplasms can be intraparenchymal or exophytic and can be associated with genetic syndromes such as multiple endocrine neoplasia type I (MEN I), von Hippel–Lindau disease, neurofibromatosis type 1 and tuberous sclerosis [1]. Neuroendocrine neoplasms can be divided into hyperfunctioning or syndromic neuroendocrine tumors (H-NENs) that generate sufficient hormonal activity to create a clinically recognizable endocrine syndrome, and non-hyperfunctioning or non-syndromic neuroendocrine tumors (N-NENs) that are clinically silent because they produce insufficent levels of hormones. The H-NENs produce distinctive signs and symptoms that lead to an early clinical diagnosis. Insulinomas tend to be small whereas all other H-NENs (gastrinoma, glucagonoma, VIPoma, somatostinoma, corticotropinoma, GRFoma, parathyrinoma and carcinoid) usually are large [2]. The N-NENs present late, produce symptoms related to mass effect and may be large. Larger NENs exhibit more aggressive behavior, including local and vascular invasion and distant metastases. Imaging tests for detecting and characterizing NENs include multi-detector row CT, magnetic resonance imaging (MRI) with gadolinium enhancement, endoscopic ultrasound (EUS), somatostatin receptor scintigraphy (SRS) and positron emission tomography (PET) [3–10]. Under special circumstances, angiography also is used [11]. This chapter will review clinical signs, imaging features, diagnostic evaluation and treatment of pancreatic NENs. Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

Imaging evaluation of pancreatic neuroendocrine neoplasms

Clinical features of H-NENs Tumors producing islet- and gut-related hormones Insulinoma

Insulinoma is the most common NEN and presents with the Whipple triad consisting of hypoglycemia, low fasting glucose levels (< 50 mg/dl) and relief of symptoms with glucose administration (Table 7.1). Patients may experience additional symptoms such as ataxia, blurred vision, dizziness, palpitation, sweating and headaches. Most manifestations are related to hyperinsulinism and catecholamine release. Diagnosis is based on high plasma levels of insulin in association with low levels of serum glucose. Other causes of hypoglycemia must be ruled out, and radiographic localization should be attempted only when the endocrinologic diagnosis is certain. This holds true for all H-NENs of the pancreas. Table 7.1. Hyperfunctioning neuroendocrine tumors of the pancreas: hormones and endocrine syndromes* Hormone

Endocrine Syndrome

Cell Type

Islet-cell related hormones Insulin Glucagon Somatostatin

Insulinoma, Whipple triad Glucagonoma, “4 D syndrome” Somatostatinoma, non-specific symptoms

B A D

Gut-related and ectopic hormones Gastrin Gastrinoma, Zollinger–Ellison Vasoactive intestinal VIPoma, Verner–Morrison peptide (VIP) syndrome, WDHA syndrome ACTH Cushing syndrome, hypertension, osteoporosis, muscle atrophy GRF GRFoma, acromegaly Serotonin Carcinoid, syndrome without metastases PTH, PTH-like substance Parathyrinoma, hypercalcemia

G D1 G, atypical Atypical EC Atypical

* Adapted and modified from J. L. Chezmar Pancreatic neoplasms. In Textbook of Gastrointestinal Radiology, ed. R. M. Gore and M. S. Levine, vol. 2, 2nd edn. (Philadelphia, PA: W. B. Saunders Company, 2000), p. 1808. ACTH = adrenocorticotropic hormone; GRF = growth hormone-releasing factor; PTH = parathyroid hormone; VIPoma = vasoactive intestinal peptide-secreting tumor; WDHA = watery diarrhea, hypokalemia, achlorhydria

105

106

Pancreatic Cancer

Table 7.2. Features of hyperfunctioning and non-hyperfunctioning endocrine tumors of the pancreas Endocrine tumor Hyperfunctioning Insulinoma Gastrinoma

Glucagonoma VIPoma Somatostatinoma ACTHoma or Corticotropinoma GRFoma PTHoma Parathyrinoma Carcinoid

Features

Malignancy

Solitary, small, (< 2–3 cm), head, tail, occ. fibrous stroma, 10% often homogeneous, multiple (MEN I) = 5–10% Multiple in up to 60%, large (> 3 cm), in gastrinoma 60–65% triangle, often heterogeneous, MEN I = 20–25%, extrapancreatic 15% 70% Solitary, large, tail>body>neck, often heterogeneous, often invasive, metastases, occ. MEN Solitary, large, >2/3 body and tail, heterogeneous, rarely >60% MEN I, extrapancreatic 10% Solitary, large, head, aggressive; in men often ectopic to >70% pancreas (duodenum, small bowel), MEN IIb Solitary, rare, large, metastases at time of diagnosis, ~100% DDx to MEN I with islet cell and pituitary tumor Solitary, large (> 6 cm), heterogeneous, metastases in 60% >30% at time of diagnosis, MEN I in 33% Solitary, liver metastases, very rare, MEN I, but >90% symptoms more often from parathyroid adenoma in MEN I Solitary, usually large, cystic/necrotic degeneration, >90% metastases to liver at time of diagnosis

Non-hyperfunctioning Solitary, usually large, key feature: hypervascularity, no sex predilection, young to middle age, any location, often metastases, often invasive DDx: adenocarcinoma, metastasis, serous cystic neoplasm solid and papillary epithelial neoplasm, rarely sarcoma

60%

MEN = multiple endocrine neoplasia; occ = occasionally; DDx = differential diagnosis

Insulinomas tend to measure less than 2–3 cm in diameter, as even small tumors produce sufficient amounts of insulin to cause symptomatic hypoglycemia [12–13]. Most insulinomas are solitary lesions, and 90% are benign (Table 7.2). Larger lesions or calcifications in the tumor suggest malignancy. Sporadic insulinomas predominate among women (60%) and present in the fifth or sixth decade. The

Imaging evaluation of pancreatic neuroendocrine neoplasms

likelihood of multiple tumors in the pancreas is increased in patients with MEN I, and presentation at a young age should suggest this syndrome [14]. Insulinomas make up about 30% and gastrinomas about 10% of the NENs in patients with MEN I [1]. Gastrinoma

Gastrinoma, the second most common NEN, secretes high levels of gastrin and presents with diarrhea, acid hypersecretion and recurrent peptic ulcers (Zollinger–Ellison syndrome) (Table 7.1) [15]. A diagnosis of Zollinger–Elllison syndrome is made if fasting serum gastrin levels of greater than 1000 pg/ml (1000 ng/l) are detected or if the secretin stimulation test is positive [15]. Gastrinomas usually are larger than insulinomas, multiple in up to 60% (usually in MEN I), often extrapancreatic in location and malignant in 60–65% [16–17]. About 30% of patients with gastrinomas present with metastases upon diagnosis [16]. Most of these tumors are found in the “gastrinoma triangle” formed inferiorly by the second and third portion of the duodenum, medially by the junction of the pancreatic head and neck, and superiorly by the confluence of the cystic and common bile ducts. Gastrinomas grow slowly and have a better prognosis than pancreatic adenocarcinoma [18]. Glucagonoma

Patients with glucagonoma, a rare H-NEN, suffer from diabetes mellitus, diarrhea, dermatitis, painful glossitis, stomatitis, anemia, depression and deep vein thrombosis [19]. Often the symptoms are referred to as the “4D syndrome” consisting of dermatitis, diarrhea, depression and deep vein thrombosis (Table 7.1). The dermatitis consists of a skin rash, called “necrolytic migratory erythema,” which may be the first symptom and can be seen in over 70% of patients with glucagonoma [20–21]. Plasma glucagon levels are markedly elevated (> 1000 pg/ml). These tumors usually are intrapancreatic and most often occur in the pancreatic body or tail. They manifest themselves in middle-aged patients and are malignant in about 60% with a 5-year survival of 50% [22–23]. When the glucagonoma becomes symptomatic, it usually exceeds 5 cm in size, is locally invasive and has metastasized to regional lymph nodes. VIPoma

This tumor secretes large quantities of vasoactive intestinal polypeptides (VIP) that act directly on cyclic adenosine monophosphates in the epithelial cells of the bowel

107

108

Pancreatic Cancer

that cause massive watery diarrhea (at least 1 liter per day). The VIP and diarrhea also produce hypokalemia, achlorhydria (inhibition of gastrin production), acidosis and dehydration [24]. The symptom complex is called watery diarrhea, hypokalemia and achlorhydria (WDHA) syndrome or Verner–Morrison syndrome (Table 7.1). Zollinger–Ellison syndrome and laxative abuse can present in a similar fashion; the latter needs to be excluded. The tumors usually are larger than 3 cm in diameter and malignant in at least 60% of cases, with over two-thirds located in the body and tail of the pancreas [25–28]. Occasionally, tumors that produce similar symptomatology are located in the adrenal gland, the retroperitoneum, the sympathetic nerve chain, the lung and intestine [26, 28]. Rarely, these tumors are associated with MEN I [29]. Early diagnosis, extensive surgical resection combined with somatostatin analogues, and chemotherapy may result in prolonged survival, even in patients who present with poorly differentiated tumors or advanced disease [30]. Somatostinoma

The somatostatinoma syndrome comprises the triad of diabetes mellitus, gallbladder disease and diarrhea with or without steatorrhea [31]. Hypochlorhydria and weight loss also are common symptoms. The diagnosis is established with increased serum levels of somatostatin. Because the symptoms are non-specific, somatostatin levels often are not measured, and the tumor is rarely diagnosed early but is found only upon imaging or during laparotomy. Somatostatinomas may be associated with neurofibromatosis, MEN IIb, pheochromocytoma, paragangliomas or carcinoids [32–33]. Somatostatinomas are rare, usually solitary and large (2–10 cm in diameter) and tend to be aggressive (Table 7.2) [31]. Metastases are present at the time of diagnosis in over 70%, particularly if the primary tumor is larger than 2 cm. Over 75% occur in the head of the pancreas [34–35]. Although many are located within the pancreas, some arise from the small bowel, the duodenal ampulla or periampullary mucosa (52%) [32–33]. In women, the tumor occurs more often in the pancreas, and in men more often in the duodenum [36]. When the small bowel is involved, the neoplasm usually is a carcinoid that consists almost completely of somatostatincontaining cells but produces little somatostatin. If the neoplasm is outside the pancreas, it tends to be smaller (0.5–4 cm), which probably is related to the fact that it produces symptoms such as jaundice, bleeding and ulceration, which lead to its early discovery. The tumor occurs in the 4th–6th decades of life, and the prognosis is poor, with an average survival of 1–2 years [37].

Imaging evaluation of pancreatic neuroendocrine neoplasms

Tumors producing ectopic hormones Corticotropinoma

The rare corticotropinoma may secrete several different hormones. Adrenocorticotropic hormone (ACTH)-producing pancreatic H-NENs are found in 4–16% of patients with ectopic ACTH syndrome. Corticotropinoma may also secrete melanocyte-stimulating hormone (MSH), corticotropin-releasing hormone (CRH), gastrin or insulin [38–39]. Excessive ACTH increases cortisol levels in the serum and results in clinical symptoms including impaired glucose tolerance, Cushing syndrome, hypertension, osteoporosis and muscle atrophy (Table 7.1). Corticotropinomas must be distinguished from MEN I which consists of H-NEN of the pancreas, parathyroid adenoma and ACTH-secreting pituitary adenoma. Corticotropinomas are large, solitary, malignant and invariably present with metastases upon diagnosis (Table 7.2) [40]. GRFoma

GRFomas secrete a growth hormone-releasing factor, and patients with this tumor may develop acromegaly (Table 7.1) [41]. This tumor is large (> 6 cm) and has metastasized at the time of diagnosis in over 30% [38]. MEN I is present in 33% of these patients and 40% also have the Zollinger–Ellison syndrome (see gastrinoma above) [42]. Parathyrinoma

Very rarely, a parathyrinoma may be present in patients with NEN of the pancreas and hypercalcemia, but usually these patients have MEN I with a parathyroid adenoma. Parathyrinomas secrete a parathyroid hormone (PTH)-like protein [43–44]. This pancreatic tumor is almost always metastatic to the liver at the time of diagnosis. Only three cases have been reported [45–46]. Carcinoid

Some NENs of the pancreas can secrete serotonin and produce typical features of a carcinoid tumor. These tumors may be benign or have features of a carcinoma with poorly differentiated cells. Over 50% of patients with this tumor have the carcinoid syndrome but it is often atypical. Patients tend to suffer from more severe flushing, hypotension, periorbital edema and increased lacrimation [47]. Liver metastases need not be present for the syndrome to occur, but almost 90% of patients with the syndrome have liver metastases [38]. Small tumors are likely to be homogeneous

109

110

Pancreatic Cancer

and hypervascular, whereas larger tumors are heterogeneous and hypovascular with areas of cystic degeneration and necrosis (Table 7.2) [48].

Clinical features of N-NENs Non-hyperfunctioning neuroendocrine tumor is the third most common neuroendocrine neoplasm, and patients with this type of tumor do not demonstrate an endocrine syndrome. They usually present with metastatic disease and/or abdominal pain caused by the mass and its extension into surrounding tissue [49]. If the mass is located in the head of the pancreas, the patient may have jaundice. Some patients may present with variceal bleeding because of gastric varices caused by tumor invasion of the splenic vein. Upon presentation, these types of pancreatic NENs are larger than those that are accompanied by an endocrine syndrome [50–51]. More recently, probably owing to advances in imaging with MDCT and optimized bolus techniques, small N-NENs are discovered incidentally in asymptomatic patients [50]. The cumulative 5-year survival of N-NEN patients has been reported to be 52–58% [51–52].

Radiographic imaging features and results for H-NENs Computed tomography, MRI, ultrasound (transabdominal, endoscopic and intraoperative), somatostatin receptor scintigraphy (SRS), PET and selective arterial stimulation are employed today to localize and characterize NENs of the pancreas. To assess pancreatic neoplasms by MDCT, it is common practice to obtain images during more than one phase of enhancement (dual- or triple-phase pancreatic protocol in the arterial and/or pancreatic and the portal venous phase). All NENs demonstrate various degrees of hyperenhancement on MDCT. Computed tomography is the diagnostic imaging test of choice, whereas MRI is a problemsolving technique. In certain types of NENs, SRS is useful when combined with PET. Selective arterial stimulation is used only when other tests have failed to demonstrate a NEN but the clinical suspicion remains high. Insulinoma

In most cases, insulinomas are small (< 2–3 cm) hypervascular tumors that on MDCT appear as brightly enhancing round-to-oval masses in the arterial and pancreatic phases with rapid washout in the portal venous phase (Figure 7.1).

Imaging evaluation of pancreatic neuroendocrine neoplasms

(a)

(b)

Figure 7.1. A 65-year-old man who suffered from Whipple triad was found to have a fasting blood glucose of 30 mg/dl. At surgery a single insulinoma measuring 1.5 cm in diameter was found in the head of the pancreas. (a) An axial pancreatic phase image from a multislice helical CT outlines a small oval-shaped area of hyperenhancement (arrow) in the head of the pancreas that represents the insulinoma. (b) On a portal venous phase image from the same examination, the mass (arrow) is barely discernible from the remainder of the pancreatic parenchyma.

Over two-thirds are located to the left of the superior mesenteric artery [35]. In some cases, insulinomas appear as hyperattenuating lesions in the portal venous phase only or in all phases. Uncommonly, insulinomas present as cystic masses [5]. In the rare instance of malignant degeneration of an insulinoma, the appearance of liver metastases is similar to that of the primary tumor. This feature is shared by all hyperfunctioning islet tumors. In one retrospective study that assessed dual-phase thin-section MDCT and endoscopic ultrasonography for preoperative evaluation of insulinomas, 3.2 mm sections were obtained and reconstructed at 1.6 mm in the arterial-dominant phase [53]. The sensitivity of CT was 94.4% compared with 93.8% for endoscopic ultrasound. If the two tests were combined, the sensitivity rose to 100%. In another CT study using 2.5 mm collimation, the sensitivity was only 63% but 83% retrospectively [5]. This difference demonstrates that thin sections are essential for achieving high sensitivity with these small tumors. Some of the missed lesions were close to vessels and mistaken for a vascular structure [5]. Coronal and sagittal reconstruction might have eliminated such interpretative errors, as shown in a smaller series using multiplanar reformations in addition to transaxial slices [54]. On MR imaging, insulinomas of the pancreas are of low signal intensity on T1-weighted fat-suppressed images and of high signal intensity on T2-weighted

111

112

Pancreatic Cancer

Figure 7.2. A 55-year-old man who presented with profuse sweating and dizziness had a fasting glucose of 40 mg/dl. The endoscopic ultrasound shows a well-defined lesion (arrows). At surgery a single insulinoma was found in the body of the pancreas, which measured 1.4 cm × 1.2 cm in diameter. Following surgery, the patient was symptom-free.

images [55–56]. Occasionally, an insulinoma can be hypointense relative to pancreatic parenchyma on T2-weighted sequences due to the presence of fibrous stroma [57], as seen in 3 of 20 cases in one study [58]. Some insulinomas show marked enhancement with gadolinium throughout the lesion and some demonstrate ring-like peripheral enhancement [59]. The presence of fibrous tissue diminishes the degree of enhancement after gadolinium administration [58]. On transabdominal and endoscopic ultrasound, insulinomas appear as smoothly marginated hypoechogenic masses, often with posterior acoustic enhancement [60] (Figure 7.2). On transabdominal ultrasound, metastatic disease to the liver may appear hypoechoic or hyperechoic. Hyperechoic lesions are readily distinguished from pancreatic adenocarcinoma metastases, which appear hypoechoic. Intraoperative ultrasound is a reliable diagnostic method to confirm insulinomas seen preoperatively and to find undetected lesions [61]. For insulinomas, SRS generally is not as helpful as for other NENs [60,62]. The sensitivity of SRS is based on the number of somatostatin receptors and not the size of the lesion. Somatostatin receptor-rich lesions appear as focal areas of increased radiotracer activity. Insulinomas are not rich in somatostatin receptors and do not exhibit significant radiotracer activity. However, F-18 fluorodeoxyglucose PET has

Imaging evaluation of pancreatic neuroendocrine neoplasms

successfully demonstrated insulinomas in patients with confirmed hyperinsulinemic hypoglycemia and negative CT, MRI and ultrasound [63]. Similarly, arterial stimulation tests with calcium have revealed successful localization of insulinomas in patients with endogenous hyperinsulinism [64–65]. This test requires selective catheterization of the right hepatic vein and the arteries that supply the pancreas or the liver and is time consuming and invasive. Gastrinoma

Computed tomography readily demonstrates pancreatic gastrinomas because of their larger size (3–4 cm). The CT appearance of gastrinoma may be that of a mass that is lower, higher or equal in attenuation compared with the surrounding pancreas. Often lesions are better demonstrated in the arterial than in the portal venous phase. Large gastrinomas usually show hypervascularity in the periphery with areas of little or no enhancement in the center indicative of necrosis, fibrosis or cystic degeneration. Intra- and extrahepatic metastases tend to have an appearance similar to that of the primary tumor [66]. Extrapancreatic gastrinomas are frequent and more challenging to detect, but multiplanar reformations (MPRs) often are helpful [4] (Figure 7.3). In one study using non-helical CT, 80% of pancreatic gastrinomas but only 35% of extrapancreatic gastrinomas were detected [67]. More recent studies using MDCT with thin sections have shown high detection rates for extrapancreatic lesions [3, 68]. Gastrinomas are of low signal intensity on T1-weighted fat-suppressed MR images and high intensity on T2-weighted MR sequences, similar to all other NENs. On gadolinium-enhanced images, gastrinomas may enhance in a heterogeneous, uniform homogeneous or ring-like fashion, with the ring-like enhancement being more common than with any other NEN [7]. Transabdominal and endoscopic ultrasound are useful for imaging pancreatic but not extrapancreatic gastrinomas [69]. Somatostatin receptor scintigraphy demonstrates the primary gastrinoma and its metastases with a sensitivity of 86% due to the fact that gastrinomas contain a large number of somatostatin receptors [60, 70–71] (Figure 7.3). In one study SRS also was useful in distinguishing between H-NENs metastatic to the liver and hepatic hemangiomas with an accuracy of 96% [72]. In patients with clinical symptoms of Zollinger–Ellison syndrome but negative imaging data, the arterial secretin stimulation test with serotonin has helped detect pancreatic and ectopic gastrinomas [70, 73]. Multi-detector row computed tomography is the most effective imaging method for detecting and staging pancreatic gastrinomas but the overall sensitivity for all

113

114

Pancreatic Cancer

(a)

(b)

(c)

(d)

Figure 7.3. A 42-year-old woman presented with pain and diarrhea. Fasting gastrin levels were over 1200 pg/ml. Surgery revealed an ectopic gastrinoma. (a) An axial image from a multislice helical CT obtained in the pancreatic phase at the level of the portal vein demonstrates a bilobed hypervascular mass (arrows), which is adjacent to but separate from the pancreas (arrowhead). (b) The hyperenhancement of the ectopic gastrinoma (arrow) persists on the portal venous phase image. (c) The coronal reformation shows that the gastrinoma (arrows) is separate from the pancreas. (d) The sagittal reformation of the multislice helical CT reveals the bilobed nature (arrows) of the ectopic gastrinoma. (e) An octreotide scintigram demonstrates an area of hyperactivity near the porta hepatis (arrow) on the anterior (left side of image) and posterior projections (right side of image).

Imaging evaluation of pancreatic neuroendocrine neoplasms

Figure 7.3. (Cont.)

(e)

gastrinomas including ectopic locations is improved if SRS is added preoperatively. In cases of equivocal or negative cross-sectional imaging but high clinical suspicion, SRS may be the only technique that can reliably provide a preoperative diagnosis. Based on the improved results with cross-sectional imaging and SRS, invasive techniques currently are not recommended for diagnosis. Glucagonoma

On CT, glucagonomas usually appear as solidly enhancing lesions with occasional low attenuation areas (Figure 7.4). In one series, the size of the tumor ranged from 2.5–6 cm [74]. Glucagonomas rarely calcify and rarely are ectopic. Computed tomography, MRI, angiography, EUS, SRS and PET have been used with success, but all reports have included only anecdotal references to this rare tumor [52, 62, 75–78]. Somatostatin receptor scintigraphy serves as an adjunct to cross-sectional imaging

115

116

Pancreatic Cancer

(a)

(b)

Figure 7.4. This 67-year-old man presented with left upper quadrant pain, diarrhea, depression, swelling of both legs due to deep vein thrombosis and a skin rash. Plasma glucagon levels were elevated to 1200 pg/ml. (a) An axial image from a multi-slice helical CT in the portal venous phase shows an enhancing mass in the tail of the pancreas (large arrow) with direct invasion of the spleen, which caused a partial splenic infarct (small arrow). Several calcifications are present within the mass. Diffuse metastatic disease (arrowheads) is present in the enlarged liver. (b) An image at a slightly lower level depicts a large tumor thrombus (small arrows) in the splenic vein, collaterals (large arrows) and multiple liver metastases some of which have necrotic centers (arrowheads). Retroperitoneal lymphadenopathy (black arrow) also is seen.

[79]. Positron emission tomography has been used more recently as a reliable tool to localize the primary pancreatic mass and metastases to the liver [77–78]. VIPoma

VIPomas tend to be larger than 3 cm in diameter and most are located in the tail of the pancreas. Liver metastases usually are present at the time of diagnosis. Computed tomography demonstrates VIPomas as enhancing masses in the pancreas with similar appearing liver lesions [30] (Figure 7.5). On MR, metastatic lesions to the liver may show intense peripheral enhancement similar to the appearance of primary and metastatic lesions on CT. Due to the rarity of VIPomas, statistical data on the use of imaging are not available. Case reports or series with all types of H-NENs mixed together have been published. CT, MR, angiography and SRS have been used to localize the primary lesion and to identify metastases [80–82]. In one study of 11 cases, SRS showed the primary lesions in 10/11 with ectopic locations in 3 and liver lesions in 3/4 with metastases [30]. These results compared favorably to CT, MRI and ultrasound, which demonstrated only 4–6 of the 11 primary lesions. A combination of CT or MRI with SRS currently appears to be the most effective presurgical assessment.

Imaging evaluation of pancreatic neuroendocrine neoplasms

(a)

(b)

Figure 7.5. A 45-year-old woman suffered from massive watery diarrhea, hypokalemia and achlorhydria. Markedly elevated levels of vasoactive intestinal polypeptides were found. A VIPoma was resected in the body of the pancreas. (a) An axial image from a multislice helical CT scan obtained in the pancreatic phase reveals multiple peripherally enhancing lesions (arrows) in the liver. (b) A lower pancreatic phase image near the tip of the liver demonstrates several enhancing hepatic metastases (arrows) and a hyperenhancing mass in the head of the pancreas (arrowheads).

Somatostatinoma

Somatostatinomas are large enough to be seen easily on CT and generally occur in the pancreatic head or duodenum [83]. When they are located in the duodenum they often cause biliary or pancreatic duct obstruction as they tend to occur near the papilla of Vater [83]. Almost 50% of patients with somatostatinomas have metastases to the liver or lymph nodes at the time of diagnosis (Figure 7.6). Radiologic features of somatostatinomas resemble those of other H-NENs, and EUS, CT, MRI, angiography and SRS have been used for their diagnosis [7, 32, 76, 83–84]. All studies have reported only a few cases or have included somatostatinomas in a mixed series of NENs. Lesions in the pancreas and liver usually are well demonstrated. Radiological techniques often fail to demonstrate tumors in the duodenum but may show obstruction of the pancreatic or biliary ducts as indirect evidence. The diagnosis in cases of duodenal localization can be established by endoscopic techniques including biopsy [85]. Other hyperfunctioning tumors

Other pancreatic H-NENs such as corticotropinoma, parathryinoma and carcinoid are very rare, and only case reports or series of H-NENs including some of them

117

118

(a)

Pancreatic Cancer

(b)

Figure 7.6. A 63-year-old man presented with weight loss and diarrhea. Increased serum levels of somatostatin were found. At surgery, a 3-cm neuroendocrine tumor was found in the third portion of the duodenum. (a) An axial image from a multislice helical CT scan obtained in the pancreatic phase through the upper abdomen demonstrates multiple lesions (black arrows) in the liver with intense enhancement. (b) An image near the tip of the liver again demonstrates multiple enhancing liver lesions (arrowheads) and an enhancing lesion in the duodenum (arrows) near the uncinate process.

have been published. Like all the other H-NENs, they demonstrate the features of a hypervascular mass without or with metastases. Multi-detector row computed tomography initially is used. Endoscopic ultrasound can be employed for the primary tumor but cannot assess the frequent presence of liver metastases [86]. Magnetic resonance imaging is the problem-solving tool. Positron emission tomography and SRS may be useful [87]. Angiography and venous sampling are rarely employed today for these uncommon lesions.

Radiographic imaging features and results for N-NENs (Figures 7.7–7.8) Because N-NENs are detected incidentally, often in connection with uncharacteristic symptoms resulting from local mass, invasion or metastatic disease, they usually are diagnosed radiographically in contrast to H-NENs which are diagnosed clinically. Non-hyperfunctioning neuroendocrine neoplasms usually are larger than H-NENs at the time of presentation, the average size of the tumor ranging from 5.2–8.4 cm [50, 88]. On CT, the larger tumors often show calcifications, necrosis or cystic changes, signs of local or vascular invasion and metastatic disease. Similar to H-NENs, N-NENs demonstrate vascular enhancement in at least one phase, but the enhancement of N-NENs is

Imaging evaluation of pancreatic neuroendocrine neoplasms

(a)

(b)

Figure 7.7. A 45-year-old man presented with vague midabdominal pain. A multislice helical CT was obtained. An axial image from a multislice helical CT in the pancreatic phase outlines a heterogeneously enhancing mass in the uncinate process of the pancreas (arrows), which is well defined. Surgery and pathology demonstrated a low-grade non-hyperfunctioning neuroendocrine tumor in the uncinate process.

not as intense and the pattern is more heterogeneous, even in the absence of cystic or necrotic changes (Figure 7.7). Non-hyperfunctioning neuroendocrine neoplasms of the pancreas usually are located in the head of the pancreas but may involve the entire gland and thus be confused with chronic pancreatitis [89]. The distinguishing feature is hyperenhancement of the mass or the entire gland. In one study the sensitivity of CT and US was reported to be 94% [50]. More recently, N-NENs are being found incidentally owing largely to improved detection with MDCT [90]. Incidentally discovered tumors tend to be smaller and are less likely malignant.

Role of imaging for NENs In patients with NENs, cross-sectional imaging plays an important role in finding the site of the primary tumor, detecting metastases, and determining the response to therapy. Imaging of H-NENs is focused on lesion localization as these tumors are diagnosed clinically, and therefore there is no differential diagnosis. N-NENs are diagnosed radiographically and need to be distinguished from other pancreatic tumors (Table 7.2). Hypervascularity is the distinguishing feature of NENs. Multi-detector row computed tomography is now well established as the primary imaging method for NENs, but choice of imaging test often depends on the experience, expertise and personal preference of the individual radiologist. With recent technical

119

120

(a)

Pancreatic Cancer

(b)

Figure 7.8. A 58-year-old man suffered from weight loss and general malaise for several months. Biopsies revealed a non-functioning neuroendocrine neoplasm. (a) An axial image from a multislice helical CT in the portal venous phase demonstrates a very large enhancing mass (arrows) involving most of the pancreas. Cystic degeneration (black arrowhead) also is seen within the mass. (b) and (c) Coronal reformations show the large mass to better advantage. Tumor thrombus in the main portal vein (arrowheads) with its branches and metastases to the spleen (black arrows) also are visualized. (d) The indium111 octreotide scan outlines the large mass in the pancreas (large arrows), the tumor thrombus (arrowheads) and the metastases in the spleen (short arrows). (e) A gadolinium-enhanced T1-weighted MR image demonstrates the large pancreatic mass (large arrows), the tumor thrombus in the splenic vein (arrowheads) and the metastases to the spleen (small arrows) on a follow-up scan.

advances, MRI is challenging the results obtained with MDCT. Endoscopic ultrasound plays an important role in the preoperative evaluation of the pancreas when a small functioning tumor or multiple tumors are suspected [91]. Intraoperative ultrasound can further improve detection of these lesions during surgery, especially for insulinomas. High sensitivity for detecting small H-NENs can be achieved only if optimal technique is used, regardless of the imaging method employed. For MRI and CT optimized technique includes thin sections and appropriate scan timing. Magnetic resonance imaging currently remains a problem-solving tool. Somatostatin receptor scintigraphy is useful in detecting the primary tumor and metastases but suffers from low specificity, which can be improved by adding MDCT [3, 92–93]. Somatostatin receptor scintigraphy should not be used for detecting insulinomas [62]. Positron emission tomography/computed tomography with a new somatostatin analogue,

Imaging evaluation of pancreatic neuroendocrine neoplasms

(c)

(d)

Figure 7.8. (Cont.)

(e)

68Ga-DOTA-Tyr3-octreotide recently has been shown to improve on results from conventional SRS [94]. Somatostatin-receptor techniques also can be used to treat patients with H-NENs of the pancreas and to monitor the response to treatment [95]. In rare cases in which small tumors are not detected with the previously described imaging techniques, a selective arterial stimulation test employing secretin for gastrinomas and calcium for insulinomas can be used. Detection of NENs may continue to improve with advances in the various imaging techniques and by

121

122

Pancreatic Cancer

combining results from several different imaging tests [91]. In the future, molecular imaging also may have a role in these patients.

Treatment The primary goal of treatment is the relief of symptoms to enable an acceptable quality of life for the patient [96]. Effective disease management includes accurate diagnosis with hormonal testing, radiologic localization and surgical resection with histologic classification. Diagnosis needs to be combined with biotherapy, chemotherapy, or radionuclide therapy, depending on the behavior of the specific NEN [96]. Treatment of patients with NEN of the pancreas depends on lesion size and the presence or absence of endocrine symptoms and metastatic disease. For patients with an H-NEN and no metastases, complete surgical resection is performed. Small lesions can be enucleated, often during laparoscopy, particularly if they are exophytic [97]. For larger or deep-seated lesions, a distal pancreatectomy is performed if the mass is located in the pancreatic tail, but a Whipple procedure is needed for lesions in the head of the pancreas. Complete surgical resection may achieve long-term survival. Of all the H-NENs, insulinoma is most amenable to surgery, and operative removal usually is curative. Gastrinomas usually are multiple, often are difficult to localize and frequently are metastatic [60]. Therefore, surgery is curative in only 30% of patients with sporadic gastrinoma. Nevertheless medical therapy with octreotide usually controls the symptoms of gastrinomas [98]. Patients with metastatic H-NEN may still gain from resection of the primary tumor [15, 96], often experiencing relief from their clinical neuroendocrine symptoms for at least several months [99]. Patients with a limited number of hepatic metastases may gain longer survival from surgical resection, chemoembolization or ablation of the liver lesions [39, 100–101]. For multiple liver metastases that are refractory to chemotherapy, liver transplantation has been shown to improve survival [39]. Even with aggressive treatment of liver metastases, the recurrence rate of well-differentiated NENs remains high, but the symptoms can be controlled, and long-term survival is improved in many cases [30]. Patients with unresectable or residual disease may benefit from somatostatin analogues such as octreotide [39, 98]. Current therapeutic applications of octreotide for tumors expressing somatostatin receptors focus on stabilization of disease and tumor destruction using beta-emitting isotopes. In patients with poorly differentiated tumors or uncontrollable liver metastases and/or extrahepatic disease, chemotherapy often is used for palliation and leads to

Imaging evaluation of pancreatic neuroendocrine neoplasms

longer survival in some patients [30, 37]. Current clinical trials are exploring the role of newer antineoplastic drugs in combination with other agents. Patients with N-NENs of the pancreas who do not have distant metastases generally undergo radical resection (including multivisceral approach) [50, 102–103]. Some studies have demonstrated improved survival after radical resection of the primary tumors and metastases, but other studies have failed to show improved longterm survival [49, 52]. Adjuvant chemotherapy is recommended in all these cases.

Summary Neuroendocrine neoplasms of the pancreas manifest themselves with variable clinical appearances, biological behaviors and imaging findings. Based on their clinical presentation, they can be divided into hyperfunctioning or syndromic and non-hyperfunctioning (clinically silent) or non-syndromic tumors. Hyperfunctioning neuroendocrine tumors, especially insulinomas and gastrinomas, are detected early because of their typical endocrine symptoms and tend to be relatively small. Non-hyperfunctioning neuroendocrine tumors usually are detected late, are large and present as palpable or obstructing masses or as liver metastases. The key imaging feature of all these tumors is hyperenhancement. Overall, larger tumors tend to demonstrate cystic changes, calcification and necrosis. They frequently are associated with local and vascular invasion, and distant metastases. A variety of imaging techniques is available for localizing and characterizing these pancreatic neuroendocrine tumors. Multi-detector row computed tomography usually is the imaging test of choice followed by endoscopic ultrasound and MRI. In some cases the primary tumors and in many cases the extrapancreatic lesions are demonstrated only with SRS or PET. If all imaging tests are negative but the clinical suspicion remains high, an arterial stimulation test with calcium for insulinomas or secretin for gastrinomas may be diagnostic. The treatment and prognosis of these neoplasms is determined by the size, histopathology, hormonal activity and presence or absence of local invasion or metastases.

REFERENCES 1. Scarsbrook AF, Thakker RV, Wass JA, Gleeson FV, Phillips RR. Multiple endocrine neoplasia: spectrum of radiologic appearances and discussion of a multitechnique imaging approach. Radiographics 2006; 26: 433–451.

123

124

Pancreatic Cancer

2. Thoeni RF. Hyperfunctioning endocrine tumors. In Procacci CM, Megibow AJ, eds. Medical Radiology – Diagnostic Imaging and Radiation Oncology Volumes on Imaging of the Pancreas. Cystic and Rare Tumors. Heidelberg, Springer-Verlag. 2003; 219–240. 3. Rappeport ED, Hansen CP, Kjaer A, Knigge U. Multidetector computed tomography and neuroendocrine pancreaticoduodenal tumors. Acta Radiol 2006; 47: 248–256. 4. Horton KM, Hruban RC, Yeo C, Fishman EK. Multi-detector row CT of pancreatic islet cell tumors. Radiographics 2006; 26: 453–464. 5. Fidler JL, Fletcher JG, Reading CC, et al. Preoperative detection of pancreatic insulinomas on multiphasic helical CT. Am J Roentgenol 2003; 181: 775–780. 6. Ichikawa T, Peterson MS, Federle MP, et al. Islet cell tumor of the pancreas: biphasic CT versus MR imaging in tumor detection. Radiology 2000; 216: 163–171. 7. Semelka RC, Custodio CM, Cem Balci N, Woosley CT. Neuroendocrine tumors of the pancreas: spectrum of appearances on MRI. Magn Res Imaging 2000; 11: 141–148. 8. Bansal R, Tierney W, Carpenter S, Thompson N, Scheiman JM. Cost effectiveness of EUS for preoperative localization of pancreatic endocrine tumors. Gastrointest Endosc 1999; 49: 19–25. 9. Alexander HR, Fraker DL, Norton JA, et al. Prospective study of somatostatin receptor scintigraphy and its effect on operative outcome in patients with Zollinger–Ellison syndrome. Ann Surg 1998; 228: 228–238. 10. Ahlstrom H, Eriksson B, Bergstrom M, et al. Pancreatic neuroendocrine tumors: diagnosis with PET. Radiology 1995; 195: 333–337. 11. Morita S, Machida H, Kuwatsuru R, et al. Preoperative localization of pancreatic insulinoma by super selective arterial stimulation with venous sampling. Abdom Imaging 2007; 32: 126–128. 12. Chen X, Cai WY, Yang WP, Li HW. Pancreatic insulinomas: diagnosis and surgical treatment of 74 patients. Hepatobiliary Pancreat Dis Int 2002; 3: 458–461. 13. Quinkler M, Strelow F, Pirlich M, et al. Assessment of suspected insulinoma by 48-hour fasting test: a retrospective monocentric study of 23 cases. Horm Metab Res 2007; 7: 507–510. 14. Carney JA. Familial multiple endocrine neoplasia syndromes: components, classification and nomenclature. J Intern Med 1998; 243: 425–432. 15. Hung PD, Schubert ML, Mihas AA. Zollinger–Ellison Syndrome. Curr Treat Options Gastroenterol 2003; 6: 163–170. 16. Hoff A, Cote G, Gagel R. Management of neuroendocrine cancers of the gastrointestinal tract: islet cell carcinoma of the pancreas and other neuroendocrine carcinomas. In Abbruzzese J, Evans D, Willett C, Fenoglio-Preiser C, Gastrointestinal Oncology. New York, Oxford University Press. 2004; 780–800. 17. Turner HE, Wass JAH. Multiple endocrine neoplasia. In Turner HE, ed. Oxford Handbook of Endocrinology and Diabetes. Oxford, Oxford University Press. 2002; 718–730. 18. Campana D, Piscitelli L, Mazzotta E, et al. Zollinger–Ellison syndrome. Diagnosis and Therapy. Minerva Med 2005; 96: 187–206.

Imaging evaluation of pancreatic neuroendocrine neoplasms

19. Povoski SP, Zaman SA, Ducatman BS McFadden DW. Dermatitis, glossitis, stomatitis, cheilitis, anemia and weight loss: a classic presentation of pancreatic glucagonoma. W V Med J 2002; 98: 12–14. 20. Edney JA, Hofmann S, Thompson JS Kessinger A. Glucagonoma syndrome is an underdiagnosed clinical entity. Am J Surg 1990; 160: 625–628. 21. van Beek AP, de Haas ER, van Vloten WA, et al. The glucagonoma syndrome and necrolytic migratory erythema: a clinical review. Eur J Endocrinol 2004; 151: 531–537. 22. Bloom SR, Polak JM. Glucagonoma syndrome. Am J Med 1987; 82: 25–36. 23. Wynick D, Hammon PJ, Bloom SR. The glucagonoma syndrome. Clin Dermatol 1993; 11: 93–97. 24. Soga J, Yakuwa Y. ViPoma/diarrheogenic syndrome: a statistical evaluation of 241 reported cases. J Exp Clin Cancer Res 1998; 17: 389–400. 25. Remme CA, de Groot GH, Schrijver G. Diagnosis and treatment of VIPoma in a female patient. Eur J Gastroenterol Hepatol 2006; 18: 93–99. 26. Kloppel G, Heitz PU. Pancreatic endocrine tumors. Pathol Res Pract 1988; 183: 155–168. 27. Rothmund M, Stinner B, Arnold R. Endocrine pancreatic carcinoma. Eur J Surg Oncol 1991; 17: 191–199. 28. Jaffe BM. Surgery for gut hormone-producing tumors. Am J Med 1987; 82: 68–76. 29. Krejs GJ. VIPoma syndrome. Am J Med 1987; 82: 37–48. 30. Nikou GC, Toubanakis C, Nikolaou P, et al. VIPomas: an update in diagnosis and management in a series of 11 patients. Hepatogastroenterology 2005; 52: 1259–1265. 31. Nesi G, Marcucci T, Rubio CA, Brandi ML, Tonelli F. Somatostatinoma: clinicopathological features of three cases and literature reviewed. J Gastroenterol Hepatol 2008; 23: 521–526. 32. Hamissa S, Rahmouni A, Coffin C, Wolkenstein P. CT detection of an ampullary somatostatinoma in a patient with von Recklinghausen’s disease. Am J Roentgenol 1999; 173: 503–504. 33. Borobia FG, Fabregat J, Jorba R, et al. Exocrine pancreatic insufficiency caused by a somatostatinoma of the minor and major duodenal papilla in a patient with neurofibromatosis. Eur J Surg 2001; 167: 154–156. 34. Tanaka S, Yamasaki S, Matsushita H. Duodenal somatostatinoma: a case report and review of 31 cases with special reference to the relationship between tumor size and metastasis. Pathol Int 2000; 50: 146–152. 35. Howard TJ, Stabile BE, Zinner MJ, et al. Anatomic distribution of pancreatic endocrine tumors. Am J Surg 1990; 159: 258–264. 36. Patel YC, Ganda OP, Benoit R. Pancreatic somatostatinoma: abundance of somatostatin-28 (1–12)-like immunoreactivity in tumor and plasma. J Clin Endocrinol Metab 1983; 57: 1048–1053. 37. Soga J, Yakuwa Y. Somatostatinoma/inhibitory syndrome: a statistical evaluation of 173 reported cases as compared to other pancreatic endocrinomas. J Exp Clin Cancer Res 1999; 18: 13–22. 38. Solcia E, Sessa F, Rindi G, Bonato M, Capella C. Pancreatic endocrine tumors: nonfunctioning tumors and tumors with uncommon function. In Dayal Y, ed. Endocrine Pathology of the Gut and Pancreas. Boca Raton, FL, CRC. 1991; 105–131.

125

126

Pancreatic Cancer

39. Doi M, Imai T, Shichiri M, et al. Octreotide-sensitive ectopic ACTH production by islet cell carcinoma with multiple liver metastases. Endocr J 2003; 50: 135–143. 40. Doppman JL, Nieman LK, Cutle GB Jr, et al. Adrenocorticotropic hormone – secreting islet cell tumors: are they always malignant? Radiology 1994; 190: 59–64. 41. Berger G, Trouillas J, Bloch B, et al. Multihormonal carcinoid tumor of the pancreas. Secreting growth hormone-releasing factor as a cause of acromegaly. Cancer 1984; 54: 2097–2108. 42. Takahashi H, Nakano K, Adachi Y, et al. Multiple nonfunctional pancreatic islet cell tumor in multiple endocrine neoplasia type I. A case report. Acta Pathol Jpn 1988; 38: 667–668. 43. Kloppel G, Heitz PU. Pancreatic endocrine tumors. Pathol Res Pract 1988; 183: 155–168. 44. Bresler L, Boissel P, Conroy T, Grosdidier J. Pancreatic islet cell carcinoma with hypercalcemia: complete remission 5 years after surgical excision and chemotherapy. Am J Gastroenterol 1991; 86: 635–638. 45. Broughan TA, Leslie JD, Soto JM, Hermann RE. Pancreatic islet cell tumors. Surgery 1986; 99: 671–678. 46. Solcia E, Sessa F, Rindi G, et al. Classification and histogenesis of gastroenteropancreatic endocrine tumours. Eur J Clin Invest 1990; 20 Suppl. 1: S72–S81. 47. Friesen SR. Tumors of the endocrine pancreas. N Engl J Med 1982; 306: 580–590. 48. Dahan H, Soyer P, Cochand-Priollet B, et al. Imaging of primary carcinoid tumor of the pancreas. J Radiol 2001; 82: 987–990. 49. Eriguchi N, Aoyagi S, Hara M, et al. Nonfunctioning islet cell carcinoma of the pancreas: an evaluation of seven patients who underwent resection followed by long-term survival. Surg Today 1999; 29: 233–237. 50. Furukawa H, Mukai K, Kosuge T, et al. Nonfunctioning islet cell tumors of the pancreas: clinical, imaging and pathological aspects in 16 patients. Jpn J Clin Oncol 1998; 28: 255–261. 51. Liang H, Wang P, Wang XN, Wang JC, Hao XS. Management of nonfunctioning islet cell tumors. World J Gastroenterol 2004; 10: 1806–1809. 52. Phan GQ, Yeo CJ, Hruban RH, et al. Surgical experience with pancreatic and peripancreatic neuroendocrine tumors: review of 125 patients. J Gastrointest Surg 1998; 2: 472–482. 53. Gouya H, Vignaux O, Augui J, et al. CT, endoscopic sonography, and a combined protocol for preoperative evaluation of pancreatic insulinomas. Am J Roentgenol 2003; 181: 987–992. 54. Chung MJ, Choi BI, Han JK, et al. Functioning islet cell tumor of the pancreas: localization with dynamic spiral CT. Acta Radiol 1997; 38: 135–138. 55. Liessi G, Pasquali C, D’Andrea AA, Scandellari C, Pedrazzoli S. MRI in insulinomas: preliminary findings. Eur J Radiol 1992; 14: 46–51. 56. Moore, Rogers CE, Britton BJ. Magnetic resonance imaging of endocrine tumours of the pancreas. Br Med J 1995; 68: 341–347. 57. Mori M, Fukuda T, Nagayoshi K, Kohzaki S, et al. Insulinoma: correlation of short-TI inversion-recovery (STIR) imaging and histopathologic findings. Abdom Imaging 1996; 21: 336–341.

Imaging evaluation of pancreatic neuroendocrine neoplasms

58. Thoeni RF, Mueller-Lisse UG, Chan R, Do NK, Shyn PB. Detection of small, functional islet cell tumors in the pancreas: selection of MRI imaging sequences for optimal sensitivity. Radiology 2001; 214: 483–490. 59. Kraus BB, Ros PR. Insulinoma: diagnosis with fat-suppressed MR imaging. Am J Roentgenol 1994; 162: 69–70. 60. Zimmer T, Stolzel U, Bader M, et al. Endoscopic ultrasonography and somatostatin receptor scintigraphy in the preoperative localisation of insulinomas and gastrinomas. Gut 1996; 39: 562–568. 61. Kalafat H, Mihmanli I, Saribeyoglu K, Belli A. Intraoperative doppler ultrasound: a reliable diagnostic method in insulinoma. Hepatogastroenterology 2007; 54: 1256–1258. 62. Virgolini I, Traub-Weidinger T, Decristoforo C. Nuclear medicine in the detection and management of pancreatic islet-cell tumours. Best Pract Res Clin Endocrinol Metab 2005; 19: 213–227. 63. Kauhanen S, Seppanen M, Minn H, et al. Fluorine-18-L-dihydroxyphenylalanine (18F-DOPA) positron emission tomography as a tool to localize an insulinoma or beta-cell hyperplasia in adult patients. J Clin Endocrinol Metab 2007; 92: 1237–1244. 64. Pereira PL, Roche AJ, Maier GW, et al. Insulinoma and islet cell hyperplasia: value of the calcium intraarterial stimulation test when findings of other preoperative studies are negative. Radiology 1998; 206: 703–709. 65. Wiesli P, Brandle M, Schmid C, et al. Selective arterial calcium stimulation and hepatic venous sampling in the evaluation of hyperinsulinemic hypoglycemia: potential and limitations. J Vasc Interv Radiol 2004; 15: 1251–1256. 66. Debray MP, Geoffroy O, Laissy JP, et al. Imaging appearances of metastases from neuroendocrine tumours of the pancreas. Br J Radiol 2001; 74: 1065–1070. 67. Wank SA, Doppman JL, Miller DL, et al. Prospective study of the ability of computed axial tomography to localize gastrinomas in patients with Zollinger–Ellison syndrome. Gastroenterology 1987; 92: 905–912. 68. Pfannenberg AC, Burkart C, Krober SM, et al. Dual-phase multidetector thin-section CT in detecting duodenal gastrinoma. Abdom Imaging 2005; 30: 543–547. 69. Kann PH. Endoscopic ultrasound imaging in neuroendocrine pancreatic tumors. A critical analysis. Med Klin (Munich) 2006; 101: 546–551. 70. Gibril F, Doppman JL, Chang R, et al. Metastatic gastrinomas: localization with selective arterial injection of secretin. Radiology 1996; 198: 77–84. 71. Briganti V, Matteini M, Ferri P, et al. Octreoscan SPET evaluation in the diagnosis of pancreas neuroendocrine tumors. Cancer Biother Radiopharm 2001; 16: 515–524. 72. Termanini B, Gibril F, Doppman JL, et al. Distinguishing small hepatic hemangiomas from vascular liver metastases in gastrinoma: use of a somatostatin-receptor scintigraphic agent. Radiology 1997; 202: 151–158. 73. Imamura M, Takahashi K. Use of selective arterial secretin injection test to guide surgery in patients with Zollinger–Ellison syndrome. World J Surg 1993; 17: 433–438.

127

128

Pancreatic Cancer

74. Breatnach ES, Han SY, Rahatzad MT, Stanley RJ. CT evaluation of glucagonomas. J Comput Assist Tomogr 1985; 9: 25–29. 75. Solivetti FM, Giunta S, Caterino M, et al. CT findings in a case of glucagonoma with necrolytic migrating erythema. Radiol Med (Torino) 2001; 102: 410–412. 76. Anderson CJ, Dehdashti F, Cutler PD, et al. 64Cu-TETA-octreotide as a PET imaging agent for patients with neuroendocrine tumors. J Nucl Med 2001; 42: 213–221. 77. Fernandez-Represa JA, Fernandez Rodriguez D, Perez Contin MJ, et al. Pancreatic glucagonoma: detection by positron emission tomography. Eur J Surg 2000; 166: 175–176. 78. Nishiguchi S, Shiomi S, Ishizu H, et al. A case of glucagonoma with high uptake on F-18 fluorodeoxyglucose positron emission tomography. Ann Nucl Med 2001; 15: 259–262. 79. Johnson DS, Coel MN, Bornemann M. Current imaging and possible therapeutic management of glucagonoma tumors: a case report. Clin Nucl Med 2000; 25: 120–122. 80. Sofka CM, Semelka RC, Marcos HB, Woosley JT. MR imaging of metastatic pancreatic VIPoma. Magn Res Imaging 1997; 15: 1205–1208. 81. Thomason JW, Martin RS, Fincher ME. Somatostatin receptor scintigraphy: the definitive technique for characterizing vasoactive intestinal peptide-secreting tumors. Clin Nucl Med 2000; 25: 661–664. 82. Mortele KJ, Oei A, Bauters W, et al. Dynamic gadolinium-enhanced MR imaging of pancreatic VIPoma in a patient with Verner–Morrison syndrome. Eur Radiol 2001; 11: 1952–1955. 83. Tjon A, Tham RT, Jansen JB, Falke TH, Lamers CB. Imaging features of somatostatinoma: MR, CT, US, and angiography. J Comput Assist Tomogr 1994; 18: 427–431. 84. Jarufe NP, Coldham C, Orug T, et al. Neuroendocrine tumours of the pancreas: predictors of survival after surgical treatment. Dig Surg 2005; 22: 157–162. 85. Guo M, Lemos LB, Bigler S, Baliga M. Duodenal somatostatinoma of the ampulla of vater diagnosed by endoscopic fine needle aspiration biopsy: a case report. Acta Cytol 2001; 45: 622–626. 86. Varas Lorenzo MJ, Miquel Collell JM, Maluenda Colomer MD, Boix Valverde J, Armengol Miro JR. Preoperative detection of gastrointestinal neuroendocrine tumors using endoscopic ultrasonography. Rev Esp Enferm Dig 2006; 98: 828–836. 87. Hoegerle S, Altehoefer C, Ghanem N, et al. Whole-body 18F dopa PET for detection of gastrointestinal carcinoid tumors. Radiology 2001; 220: 373–380. 88. Buetow PC, Parrino TV, Buck JL, et al. Islet cell tumors of the pancreas: pathologicimaging correlation among size, necrosis and cysts, calcification, malignant behavior, and functional status. Am J Roentgenol 1995; 165: 1175–1179. 89. Simpson WF, Adams DB, Metcalf JS, Anderson MC. Nonfunctioning pancreatic neuroendocrine tumors presenting as pancreatitis: report of four cases. Pancreas 1988; 3(2): 223–231. 90. Vagefi PA, Razo O, Deshpande V, et al. Evolving patterns in the detection and outcomes of pancreatic neuroendocrine neoplasms: the Massachusetts General Hospital experience from 1977 to 2005. Arch Surg 2007; 142: 347–354.

Imaging evaluation of pancreatic neuroendocrine neoplasms

91. Rockall AG, Reznek RH. Imaging of neuroendocrine tumours (CT/MR/US). Best Pract Res Clin Endocrinol Metab 2007; 21: 43–68. 92. Saga T, Shimatsu A, Koizumi K. Morphological imaging in the localization of neuroendocrine gastroenteropancreatic tumors found by somatostatin receptor scintigraphy. Acta Radiol 2005; 46: 227–232. 93. Schillaci O, Spanu A, Scopinaro F, et al. Somatostatin receptor scintigraphy in liver metastasis detection from gastroenteropancreatic neuroendocrine tumors. J Nucl Med 2003; 44: 359–368. 94. Gabriel M, Decristoforo C, Kendler D, et al. 68Ga-DOTA-Tyr3-octreotide PET in neuroendocrine tumors: comparison with somatostatin receptor scintigraphy and CT. J Nucl Med 2007; 508: 508–518. 95. Ugur O, Kothari PJ, Finn RD. Ga-66 labeled somatostatin analogue DOTA-DPhe1-Tyr3octreotide as a potential agent for positron emission tomography imaging and receptor mediated internal radiotherapy of somatostatin receptor positive tumors. Nucl Med Biol 2002; 29: 147–157. 96. Simon P, Spilcke-Liss E, Wallaschofski H. Endocrine tumors of the pancreas. Endocrinol Metab Clin N Am 2006; 35: 431–447. 97. Toniato A, Meduri F, Foletto M, Avogaro A, Pelizzo M. Laparoscopic treatment of benign insulinomas localized in the body and tail of the pancreas: a single-center experience. World J Surg 2006; 30: 1916–1919. 98. Shojamanesh H, Gibril F, Louie A, et al. Prospective study of the antitumor efficacy of long-term octreotide treatment in patients with progressive metastatic gastrinoma. Cancer 2002; 94: 331–343. 99. Sarmiento JM, Que FG. Hepatic surgery for metastases from neuroendocrine tumors. Surg Oncol Clin N Am 2003; 12: 231–242. 100. Yao KA, Talamonti MS, Nemcek A, et al. Indications and results of liver resection and hepatic chemoembolization for metastatic gastrointestinal neuroendocrine tumors. Surgery 2001; 130: 677–682. 101. Moug SJ, Leen E, Horgan PG, Imrie CW. Radiofrequency ablation has a valuable therapeutic role in metastatic VIPoma. Pancreatology 2006; 6: 155–159. 102. Dralle H, Krohn SL, Karges W, et al. Surgery of resectable nonfunctioning neuroendocrine pancreatic tumors. World J Surg 2004; 28: 1248–1260. 103. Kazanjian KK, Reber HA, Hines OJ. Resection of pancreatic neuroendocrine tumors: results of 70 cases. Arch Surg 2006; 141: 765–769; discussion 769–770.

129

8 Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms Rajesh N. Keswani and Riad R. Azar

Introduction Endoscopic ultrasound (EUS) was developed in the 1980s to address, in part, the limitations of transabdominal ultrasound imaging of the abdomen. Adequate ultrasonography relies on the transmission of sound waves. Thus imaging of the abdomen by external ultrasonography is limited by interference of bowel gas. Endoscopic ultrasound reduces bowel gas interference by placing the ultrasound transducer directly within the gastrointestinal lumen via endoscopy. Thus, merging the technologies of endoscopy and ultrasonography allows the trained endosonographer to obtain detailed images of the pancreas, among many other uses. Endoscopic ultrasound is generally performed in an endoscopy facility with the availability of skilled nurses or technicians, sedation and appropriate endoscopic equipment (Figure 8.1) [1]. Upper EUS is routinely performed after an overnight fast and is most often completed in the outpatient setting. Patients are placed under intravenous conscious sedation or monitored anesthesia care, and the procedure is completed in 30–60 minutes, depending on the complexity of the examination. Endoscopic ultrasound accuracy is dependent on a number of factors including the quality of available equipment, operator experience in technique and interpretation, and patient anatomy and body habitus. There is a recognized learning curve for the performance of accurate EUS examinations and, as with standard ultrasonography, accuracy improves with increased experience [2]. Endoscopic ultrasound imaging of the pancreas can be performed via either radial or linear echoendoscopes (Figure 8.2). Radial echoendoscopes provide a complete, 360-degree image perpendicular to the axis of the endoscope. In contrast, linear array echoendoscopes produce images along the axis of the echoendoscope, similar to external ultrasound imaging. An advantage of linear array echoendoscopes Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

Figure 8.1. Endoscopic ultrasound equipment. (A) A radial echoendoscope provides a 360-degree image perpendicular to the axis of the endoscope. (B) A linear echoendoscope produces images along the axis of the endoscope, and also has an accessory channel. The accessory channel is useful in obtaining biopsies of the pancreas. (C) In addition to both a radial and linear echoendoscope, a dedicated processor for endoscopic ultrasonography is needed. (Photos provided courtesy of Olympus Corporation, Center Valley, PA).

Figure 8.2. Normal pancreas. (A) A normal head of the pancreas is pictured with the radial echoendoscope. The “stack sign” is demonstrated with the portal vein (lower white arrow), pancreatic duct (arrowhead), and common bile duct (upper white arrow) seen. (B) Normal head of the pancreas, as seen with a linear echoendoscope. The common bile duct (CBD), portal vein (PV), and pancreatic duct (PD) are seen.

is the ability to perform direct tissue sampling via an accessory channel of the endoscope and use of an elevator that maintains position of the aspiration needle in the ultrasound plane. This technology has transformed EUS from a purely diagnostic tool into a powerful tissue acquisition and interventional modality of particular use in gastrointestinal oncology [3].

Diagnosis of pancreatic adenocarcinoma A presumptive diagnosis of pancreatic adenocarcinoma is often made based on cross-sectional imaging. In these cases, although imaging is strongly suggestive

131

132

Pancreatic Cancer

Figure 8.3. Pancreatic adenocarcinoma in the head of the pancreas. Linear EUS is used to identify a suspicious lesion seen on computed tomography in an 83-year-old female. (A) Mass, identified by fine needle aspiration as adenocarcinoma, in the pancreatic head measuring 25 mm × 20 mm in maximal cross-sectional diameter. (B) Upstream of this mass, the main pancreatic duct (PD) is dilated to 10 mm.

of adenocarcinoma, the diagnosis must be confirmed with cytology or histopathology. Alternately, a strong clinical suspicion of adenocarcinoma may persist despite a lack of diagnostic findings on non-invasive imaging. In both cases, endosonography can aid the clinician in diagnosing pancreatic adenocarcinoma. Although the EUS appearance of pancreatic adenocarcinoma is variable, it most commonly appears as a hypoechoic mass with irregular contours (Figure 8.3). Depending on the location of the tumor, there may also be proximal pancreatic duct dilation [4]. The true sensitivity of EUS is difficult to ascertain. Patients undergoing EUS evaluation generally have preprocedure imaging studies suggestive of cancer that are available for review by the endosonographer, which may artificially increase the sensitivity of EUS. Furthermore, sensitivity is affected by the population being studied, due to a variable preprocedure probability of disease. Finally there may be publication bias. Most published studies compare the sensitivity of EUS with CT or MRI, either in patients with high clinical suspicion or a previous abnormal imaging study. Due to the rapid advancement of all three technologies, however, it is difficult to interpret the findings of early studies. More recent studies comparing the sensitivities of EUS and multi-detector row spiral CT scanners may be more representative of the capabilities of these technologies. A brief analysis of three studies, published in 2002, suggested that EUS was superior to helical CT in detecting adenocarcinoma [5]. This result was heavily weighted, however, by the poor performance of CT in a single study [6]. In a

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

more recent prospective study, 120 patients suspected of having pancreatic cancer based on an initial abnormal imaging study were enrolled. Subjects first underwent EUS followed by multi-detector CT (MDCT) [7]. Eighty subjects in this study ultimately were found to have pancreatic adenocarcinoma. The sensitivity of detecting a mass was greater with EUS compared with MDCT (98% vs. 86%, P = 0.012). There was also suggestion that EUS was superior at demonstrating lesions smaller than 25 mm (89% vs. 53%, P = 0.077) although the difference was not statistically significant. Detection of these small lesions is critical as they may be most responsive to therapy. Recently, a systematic review of all published studies comparing EUS and CT was reported [8]. This analysis acknowledges the limitations of prior published studies, including marked heterogeneity, methodological flaws and variability in technology. Despite these limitations, the authors concluded that EUS was superior to CT in detecting tumors, especially lesions smaller than 30 mm. In comparison to MRI, EUS has been found to be significantly more sensitive in the detection of pancreatic adenocarcinoma. In a recent retrospective study of 48 subjects who were ultimately diagnosed with pancreatic or ampullary adenocarcinoma, EUS detected more lesions than MRI, but the difference was not statistically significant (98% vs. 87.5%) [9]. Endoscopic ultrasound detected all 12 masses smaller than 25 mm whereas MRI detected 6/12 (100% vs. 50%, P = 0.04). In practice, EUS, CT and MRI are considered complementary for the detection of adenocarcinoma rather than competitive. In most institutions, patients with a clinical suspicion of pancreatic neoplasm are referred for an initial helical CT. Patients with a negative or equivocal CT are then referred for EUS to better characterize indeterminate lesions and detect small neoplasms not seen on CT (Figure 8.4) [10].

Endoscopic ultrasound with fine needle aspiration Endoscopic ultrasound with fine needle aspiration (FNA) plays an important role in the diagnosis of pancreatic adenocarcinoma. Although EUS-FNA has a specificity of nearly 100% for adenocarcinoma, it has a variable sensitivity reported between 82 and 91% [11, 12]. The sensitivity of EUS-FNA is improved with the use of on-site cytopathology, which is now standard in many tertiary care centers [13]. In centers where on-site cytopathology is not available, at least seven FNA passes are routinely performed, which in one study resulted in a sensitivity of 83.3% (Figure 8.5) [14].

133

134

Pancreatic Cancer

Figure 8.4. Small pancreatic adenocarcinoma in the tail of the pancreas. A 64-year-old female was referred for endoscopic ultrasound due to a dilated pancreatic duct seen on CT. No discrete mass was identified on CT. Linear EUS is used to locate a mass measuring 13 mm × 16 mm in the tail of the pancreas. A suspicious lymph node is seen adjacent to this mass.

Figure 8.5. Fine needle aspiration of a solid pancreatic mass. A pancreatic mass was seen on CT scan in the pancreatic head. Linear EUS with fine needle aspiration is performed for tissue diagnosis. The needle tip (arrow) is positioned in the center of the hypoechoic mass. Seven FNA passes were performed and demonstrated adenocarcinoma cells on cytology.

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

For patients with cross-sectional imaging findings that demonstrate unresectable disease, it is important to confirm the diagnosis of adenocarcinoma prior to the initiation of palliative chemoradiotherapy or placement of a metallic biliary stent. Endoscopic ultrasound with FNA of a pancreatic mass is a safe procedure, with a low risk of pancreatitis (0.85–0.91%) which is the most commonly reported complication [15, 16]. The role of FNA is less clear in patients with an apparently resectable mass that is highly suspicious for malignancy. In such patients, EUS-FNA adds the risk of complication without changing clinical management (surgical resection). Differentiating focal pancreatic inflammation from pancreatic neoplasm with cross-sectional and EUS imaging can be difficult as focal inflammation may mimic a tumor. For this reason, EUS-FNA is frequently used to evaluate focal pancreatic masses in patients with chronic pancreatitis. Although the diagnostic accuracy of EUS alone is lower in this setting, EUS-FNA was found to have a sensitivity of approximately 73% and a specificity of 100% in two recent studies [17, 18]. Thus, although EUS-FNA cannot definitively exclude malignancy, it is valuable when cytology is positive. When FNA of a lesion yields negative cytology, we generally recommend a follow-up imaging study. When a concerning lesion persists on follow-up imaging, surgery ultimately may be required to obtain a definitive diagnosis [19]. Newer methods, including DNA analysis of microsatellite loss and k-ras point mutation may ultimately improve the sensitivity of EUS-FNA in patients with chronic pancreatitis [20]. Endoscopic ultrasound with FNA is similarly helpful in patients with acute pancreatitis and a suspicious lesion on crosssectional imaging. Demonstration of malignant cytology allows for earlier surgical referral. When cytology is negative, the clinician may monitor the patient with repeat imaging after resolution of inflammation, rather than proceeding to early surgical resection. Computed tomography-guided percutaneous FNA is performed in many centers, but may be less sensitive than EUS-FNA with an increased risk of peritoneal and abdominal wall seeding, which are rare complications [21–24]. Only one randomized, prospective cross-over study comparing EUS and CT FNA has been reported [25]. In this study of 84 subjects with suspected pancreatic adenocarcinoma, EUS had a higher sensitivity compared with CT for positive adenocarcinoma cytology, although the difference was not statistically significant (84% vs. 62%, P = 0.074). In this study, EUS was particularly helpful in the six tumors measuring less than 30 mm in diameter. In a study of 185 patients with negative percutaneous CT-guided or US-guided biopsies, subsequent EUS-FNA yielded a sensitivity for

135

136

Pancreatic Cancer

malignancy of 90% [26]. Thus, EUS is the preferred method for FNA guidance due to its superior sensitivity and the decreased risk of tumor seeding.

Staging of pancreatic adenocarcinoma In addition to its role in the diagnosis of adenocarcinoma, EUS may also assist in local staging of adenocarcinoma and determining resectability. Prospectively validated EUS criteria for vascular invasion are shown in Table 8.1 [27, 28]. In clinical practice, the endosonographer generally has access to cross-sectional imaging that is reviewed prior to EUS. As with detecting pancreatic neoplasms, this prior knowledge likely results in overestimation of the true accuracy of EUS in determining resectability. Within these limitations, data suggest that EUS is accurate in identifying pancreatic adenocarcinoma invasion into the portal and splenic vein (Figure 8.6), but is limited in assessing invasion of the superior mesenteric artery and vein [29]. In an effort to eliminate the bias of knowledge of the CT findings prior to EUS, some authors have reviewed recordings of EUS procedures in a blinded fashion [30], which has resulted in much poorer accuracy for detecting vascular invasion. A meta-analysis published in 2007 found the sensitivity and specificity of EUS for diagnosing vascular invasion to be 73% and 90.2%, respectively [31]. Thus, EUS is better at identifying vascular invasion than excluding it. Further analysis did not identify any improvement in EUS sensitivity and specificity in recent studies compared with older studies, suggesting that improvements in EUS technology have not had a large effect on EUS accuracy. It is possible that differences in EUS technique and variable application of established criteria resulted in the reduced sensitivity and specificity seen in some studies [32]. The role of EUS for staging and determining resectability in the era of helical CT has been called into question. One study found EUS to be superior to CT in tumor staging for patients who underwent surgery (67% vs. 41%; P < 0.001) but equivalent in nodal staging and determining resectability [8]. A significant limitation of this study is that the investigators were not blinded to the radiology imaging. A further limitation is that the authors included as unresectable, tumors with a complete resection macroscopically, but positive microscopic margins (R1 resection), which may not be supported by recent evidence [33]. The authors concluded that the use of EUS for staging and assessing resectability should be determined by local expertise. In a small single-institution study comparing EUS and MRI, MRI was found to be superior to EUS for loco-regional staging and for prediction of resectability, but

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

Table 8.1. Endoscopic ultrasound criteria for vascular invasion 1. Peripancreatic venous collaterals in an area of a mass that obliterates the normal anatomic location of a major portal confluence vessel 2. Tumor within the vessel lumen 3. Abnormal vessel contour with loss of the vessel-parenchymal sonographic interference Adapted from Snady H, et al. Gastrointestinal Endoscopy 1994; 40: 326–333 [27].

Figure 8.6. Portal vein invasion from a pancreatic adenocarcinoma. (A) Linear EUS demonstrates a 23 mm × 17 mm mass (arrow) in the pancreatic head. (B) There is a loss of interface (arrow) between the mass and the adjacent portal vein confluence, suggesting invasion.

the difference was not statistically significant (78% vs. 33%, P = 0.13) [9]. In a larger study of 73 patients with pancreatic adenocarcinoma, EUS and MRI were individually found to have a poor positive predictive value for both resectability (69% and 77%, respectively) and unresectability (69% and 68%, respectively) [34]. The positive predictive value was somewhat improved when both MRI and EUS agreed on resectability (89%) and unresectability (76%). It is important to note, however, that despite both MRI and EUS agreeing on unresectability, 24% of patients (4/17) were found to be resectable at the time of surgery. The authors concluded that, though each individual test lacks sufficient sensitivity or specificity, the combination of both tests might be more clinically useful. In clinical practice, either multi-detector row CT or MRI and EUS are complementary in the management of pancreatic adenocarcinoma. We favor the management algorithm presented in Figure 8.7. Unless there is clearly unresectable disease, we refer patients for surgical evaluation given the limitations of the available imaging tests.

137

138

Pancreatic Cancer

Clinical suspicion of pancreatic adenocarcinoma

High-resolution CT scan Mass lesion

Clearly unresectable? Yes

CT or EUS-FNA to confirm diagnosis

No

No Mass lesion

EUS if suspicion persists No Mass lesion

Mass lesion

EUS (FNA and staging)

No further work-up

Clear vascular invasion Yes

Oncology referral (consider clinical trial in appropriate setting)

No

Surgical referral

Figure 8.7. Suggested algorithm for work-up of pancreatic adenocarcinoma. In patients in whom pancreatic adenocarcinoma is suspected based on clinical symptoms, we recommend initial work-up with a high-resolution CT scan. Due to its improved spatial resolution, endoscopic ultrasound is indicated in patients with a normal CT scan, but a high index of suspicion for a mass. In cases in which CT identifies a mass that is not clearly unresectable, EUS can be used to assess vascular invasion. Furthermore, EUS-FNA is used to obtain tissue diagnosis prior to surgery or palliative chemotherapy.

Endoscopic ultrasound evaluation of cystic lesions The widespread use of high-resolution cross-sectional imaging has resulted in the increasing diagnosis of incidental pancreatic cysts, many of which are asymptomatic and/or small at the time of diagnosis [35]. Although the management of pancreatic cysts varies depending on cyst type, cross-sectional imaging is limited in its ability to evaluate these lesions. The improved spatial resolution of EUS and the ability to perform FNA has triggered increased use of EUS, especially for cysts smaller than 2 cm in diameter [36]. Pseudocysts are the most common pancreatic cysts, with simple cysts and duplication cysts being less common benign cysts. Because these benign cysts occasionally appear similar to cystic neoplasms on CT or MRI, further imaging may be required in their management [37]. Cystic neoplasms have varying degrees

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

of malignant potential. Serous cystadenomas very rarely undergo malignant transformation, which is in contradistinction to mucinous cystic neoplasms (MCN) [38–40]. Thus, most physicians elect to observe asymptomatic patients with serous cystadenomas, especially in the setting of multiple medical comorbidities. In contrast, mucinous cystadenomas may progress to cystadenocarcinomas and thus generally are resected. Intraductal papillary mucinous neoplasms (IPMN), formerly referred to as mucinous ductal ectasia, are closely related to mucinous cystic lesions and arise from main or side branch pancreatic ducts. These lesions may progress to adenocarcinoma [41]. Rarely, neuroendocrine tumors may present as pancreatic cysts as well [42]. Other cystic lesions such as solid pseudopapillary neoplasms, lymphoepithelial cysts, and abscesses are significantly less common [43]. Most commonly, EUS is performed to differentiate mucinous from non-mucinous (including serous) cysts. Serous cystadenomas, previously called microcystic adenomas, generally appear as focal, well-demarcated lesions containing greater than six small fluid-filled cysts on EUS imaging [44–46]. They may have a central area of fibrosis or calcification (Figure 8.8) [43]. These cysts rarely contain echogenic material, in contrast to both

Figure 8.8. Serous cyst. A large cyst was identified on a CT scan performed for unrelated symptoms in this 71-year-old male. Linear EUS is performed for further evaluation. A cyst measuring 58 mm × 53 mm with many thinly septated compartments is seen in the pancreatic tail. A focus of central calcification (arrows) is seen, which is characteristic of a serous cyst.

139

140

Pancreatic Cancer

Figure 8.9. Side branch intraductal papillary mucinous neoplasm (IPMN). Linear EUS is performed for a history of recurrent pancreatitis. The main pancreatic duct is normal. A side branch of the pancreatic duct, however, is dilated. An associated 5 mm cyst is seen contiguous with this dilated side branch. This appearance is characteristic of a side branch IPMN, which may be the etiology of recurrent bouts of pancreatitis.

pseudocysts and mucinous cystadenomas. Mucinous cystadenomas are generally macrocystic with rare internal septa and may have a thick wall (> 3 mm) or solid component [44, 47]. Main duct or side branch IPMN appears as a macrocystic lesion with associated duct dilation (Figure 8.9). Mucous secretion via the papilla of Vater may also be seen on duodenoscopy with main duct IPMN. However, due to variation in endosonographic appearance, differentiating between serous and mucinous cysts can be challenging. An early report detailing the role of EUS in the evaluation of pancreatic cystic tumors was published in 1997 [48]. The investigators found EUS to be helpful in differentiating the internal architecture of cysts and determining whether cysts were mucinous, serous or simple cysts. This was possible even in cysts smaller than 2 cm, which had not previously been possible with CT [39]. However, a later report of 48 patients who had undergone surgery for cysts suggested the EUS features alone were not sufficient to differentiate benign and malignant cystic lesions [49]. In this study two experienced endosonographers reviewed static images from radial endosonography examinations, and the authors determined that there were no EUS

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

Figure 8.10. Cytology obtained from aspiration of a cyst. Fine needle aspiration during EUS can aid the clinician in determining the etiology of a cyst. This slide demonstrates undulating sheets of ductal epithelial cells with uniformly round to oval nuclei in an organized honeycomb pattern. These cytologic features can be seen in mucinous cystic lesions of the pancreas. (Courtesy of Lourdes Ylagan, MD). This figure is reproduced in the color plate section.

features significantly predictive of a malignant cyst. Acknowledging the limitations of this study, it demonstrated that morphology alone is not sufficient to distinguish between benign and malignant cysts. A similar study of eight endosonographers using videotaped EUS exams found that examiners also had a fairly poor overall accuracy rate (ranging from 40–93%) in determining whether a cyst was neoplastic or non-neoplastic [50]. Given the frequent misclassification by expert endosonographers, more recent work has focused on the yield of FNA in cyst evaluation. In a report of 63 patients, investigators found that EUS alone had an accuracy of 73% in determining cyst type [51]. The addition of cytology obtained via FNA improved the diagnostic accuracy to 97% (Figure 8.10). This degree of accuracy for cytologic analysis is significantly higher than that of other reports [52, 53]. In this study, tumor marker analysis did not significantly improve the diagnostic accuracy due to the high accuracy of cytology alone. In a large prospective multicenter study of 112 patients with confirmed cyst type, the accuracy of EUS morphology and cytology alone were 51% and 59%, respectively [52]. Carcinoembryonic antigen (CEA) content was evaluated due to early studies suggesting that the CEA

141

142

Pancreatic Cancer

concentration in mucinous cysts was elevated [54, 55]. Using a cut-off of 192 ng/ml, cyst fluid CEA was found to have an accuracy of 79% in distinguishing between mucinous and non-mucinous cysts. Combining morphology, cytology and CEA analysis resulted in a reduction in overall accuracy, due to decreased specificity. It is important to note that the measured CEA value varies depending on the assay system used, making it difficult to implement a strict cut-off value [56]. At present, cyst fluid analysis is used in combination with other clinical factors. In a pooled analysis of 12 studies, CEA, amylase and cytology were all found to have clinical utility in differentiating mucinous and non-mucinous cysts [57]. In practice, these tests are combined with clinical history to aid the physician in determining whether pancreatic resection is recommended. Cysts that are asymptomatic, microcystic in morphology and have a very low CEA (< 5 ng/ml) are observed if the cytology is benign or non-diagnostic. Conversely, surgery is recommended if cytology demonstrates malignancy or if the cyst fluid has a markedly elevated CEA. Management decisions must be individualized for cysts with indeterminate endosonographic findings or with borderline tumor marker analysis. In these cases, the clinician must take into account cyst location, size, associated symptoms and medical comorbidities. Recent trials have shown a possible role for DNA analysis and for testing extracellular mucin content and viscosity [58–60]. If these results are confirmed, use of such tests may enhance the diagnostic accuracy of EUS-FNA in the future.

Endoscopic ultrasound in the evaluation of pancreatic endocrine neoplasms Pancreatic endocrine neoplasms (PEN) are difficult to diagnose due to their relative infrequency (0.001% prevalence) and protean presentation [61]. However, even when a neuroendocrine tumor is suspected, confirmation of the diagnosis may still be difficult owing to the variable sensitivity of standard imaging tests. Although CT, MRI and somatostatin receptor scintigraphy (SRS) are useful non-invasive tests for evaluating PEN, they are unable to localize tumors in a minority of patients. Endoscopic ultrasound has been used to localize neuroendocrine tumors since its introduction in the 1980s [62–67]. Although endoscopic technique varies, initial standard upper gastrointestinal endoscopy is generally performed prior to EUS to evaluate for submucosal tumors of the upper gastrointestinal tract [68]. Following this, the radial echoendoscope is advanced into the duodenum distal to the ampulla. The echoendoscope is slowly withdrawn and careful imaging of the duodenal wall

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

and pancreas is performed. The exact location of any tumors is identified to help plan surgical excision. The complete examination takes approximately 30 minutes. The first large experience with EUS in detecting neuroendocrine tumors was published in 1992 [67]. Thirty-seven patients with tumors not seen on CT and ultrasound were studied. Insulinomas were the predominant tumor in this study. All of the patients were evaluated by EUS prior to surgical exploration. The sensitivity of EUS for tumor detection was 82%, and the specificity was 95% in 18 control patients without tumors. The sensitivity of angiography in this study was 27%. In a more recent single-center report of 82 patients, EUS was used as the first-line examination in patients with suspected PEN [62]. Endoscopic ultrasound was found to have an overall sensitivity of 93% and accuracy of 93%. The sensitivity of EUS in detecting gastrinomas was 100%. In early studies, EUS was clearly superior to non-spiral CT [65, 66]. However, with advances in CT technology, these differences may be less marked. In a study of 32 pancreatic insulinomas, French investigators found that thin-section dual-phase helical CT (n = 15) was significantly more sensitive than sequential CT (n = 7; 94.4% vs. 28.6%, P = 0.002) [64]. The overall sensitivity of EUS in this group of 32 patients was 93.8%. The single confirmed insulinoma not detected by biphasic helical CT measured 10 mm and was seen on EUS. In other studies, the improvement in sensitivity with recent CT technology and technique is less pronounced, with a significant proportion of tumors remaining undetectable on CT [69–71]. There have been no recent published studies comparing MRI and EUS [72]. Angiography is an invasive procedure with limited sensitivity compared with EUS [62, 67]. Selective arterial stimulation using intra-arterial calcium may have a role in selected patients but due to its invasive nature and cost, it is not an option to replace other imaging techniques such as EUS. Somatostatin receptor scintigraphy has a high sensitivity, except in the case of insulinomas, which may not express somatostatin receptors. In a study comparing EUS and somatostatin receptor scintigraphy, the sensitivity for detecting insulinomas was significantly greater with EUS (93% vs. 14%) [66]. Furthermore, scintigraphy may be unable to distinguish disease in a peripancreatic lymph node from pancreatic disease. The sensitivity of EUS also is likely superior to scintigraphy in smaller tumors [73]. An advantage of EUS over other imaging techniques is the ability to obtain tissue easily for diagnostic purposes (Figure 8.11). With experienced cytopathologists, a high accuracy can be achieved [74, 75]. However, EUS-FNA is of unproven benefit in evaluating neuroendocrine tumors. It is likely most useful in tumors that appear unresectable so that non-surgical therapy may be instituted or in patients who are

143

144

Pancreatic Cancer

Figure 8.11. Pancreatic endocrine tumor. A 67-year-old male is evaluated by linear EUS for jaundice. (A) A large hypoechoic mass with irregular margins is seen in the pancreatic head measuring 30 mm × 31 mm. (B) The FNA needle (arrow) is seen in the hypoechoic mass. Cells from cytology confirm a diagnosis of neuroendocrine tumor. (C) Evaluation of the left lobe of the liver demonstrates a small lesion (arrow) measuring 7 mm × 6 mm. Fine needle aspiration demonstrated metastatic neuroendocrine cells.

poor surgical candidates [68]. In these cases, immunohistochemical analysis may be beneficial [76]. Placement of an India ink tattoo under EUS guidance also can be performed for small tumors that may not be palpable at the time of surgery [77]. The choice of tests performed in a patient with a suspected PEN is dependent on local expertise. With EUS, pancreatic endocrine tumors can be detected costeffectively and with high sensitivity [78]. In addition, cytology can be obtained if necessary. Thus, if EUS expertise is available, it is an appropriate first-line examination. For patients with a tumor demonstrated on CT or MRI, EUS may be useful to evaluate for synchronous lesions as their presence may modify the surgical plan. The main limitation of EUS is its inability to evaluate for distant metastases. Thus we advocate the use of somatostatin receptor scintigraphy in concert with EUS to evaluate for metastatic disease.

REFERENCES 1. Liu J, Carpenter S, Chuttani R, et al. Endoscopic ultrasound probes. Gastrointest Endosc 2006; 63(6): 751–754. 2. Eloubeidi MA, Tamhane A. EUS-guided FNA of solid pancreatic masses: a learning curve with 300 consecutive procedures. Gastrointest Endosc 2005; 61(6): 700–708. 3. Oh YS, Early DS, Azar RR. Clinical applications of endoscopic ultrasound to oncology. Oncology 2005; 68(4–6): 526–537. 4. Yasuda K, Mukai H, Nakajima M. Endoscopic ultrasonography diagnosis of pancreatic cancer. Gastrointest Endosc Clin N Am 1995; 5(4): 699–712.

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

5. Hunt GC, Faigel DO. Assessment of EUS for diagnosing, staging, and determining resectability of pancreatic cancer: a review. Gastrointest Endosc 2002; 55(2): 232–237. 6. Mertz HR, Sechopoulos P, Delbeke D, Leach SD. EUS, PET, and CT scanning for evaluation of pancreatic adenocarcinoma. Gastrointest Endosc 2000; 52(3): 367–371. 7. DeWitt J, Devereaux B, Chriswell M, et al. Comparison of endoscopic ultrasonography and multidetector computed tomography for detecting and staging pancreatic cancer. Ann Intern Med 2004; 141(10): 753–763. 8. Dewitt J, Devereaux BM, Lehman GA, Sherman S, Imperiale TF. Comparison of endoscopic ultrasound and computed tomography for the preoperative evaluation of pancreatic cancer: a systematic review. Clin Gastroenterol Hepatol 2006; 4(6): 717–725; quiz 664. 9. Borbath I, Van Beers BE, Lonneux M, et al. Preoperative assessment of pancreatic tumors using magnetic resonance imaging, endoscopic ultrasonography, positron emission tomography and laparoscopy. Pancreatology 2005; 5(6): 553–561. 10. Tamm EP, Loyer EM, Faria SC, et al. Retrospective analysis of dual-phase MDCT and follow-up EUS/EUS-FNA in the diagnosis of pancreatic cancer. Abdom Imaging 2007; 32: 660–667. 11. Raut CP, Grau AM, Staerkel GA, et al. Diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration in patients with presumed pancreatic cancer. J Gastrointest Surg 2003; 7(1): 118–126; discussion 127–128. 12. Shin HJ, Lahoti S, Sneige N. Endoscopic ultrasound-guided fine-needle aspiration in 179 cases: the M. D. Anderson Cancer Center experience. Cancer 2002; 96(3): 174–180. 13. Klapman JB, Logrono R, Dye CE, Waxman I. Clinical impact of on-site cytopathology interpretation on endoscopic ultrasound-guided fine needle aspiration. Am J Gastroenterol 2003; 98(6): 1289–1294. 14. LeBlanc JK, Ciaccia D, Al-Assi MT, et al. Optimal number of EUS-guided fine needle passes needed to obtain a correct diagnosis. Gastrointest Endosc 2004; 59(4): 475–481. 15. Eloubeidi MA, Tamhane A, Varadarajulu S, Wilcox CM. Frequency of major complications after EUS-guided FNA of solid pancreatic masses: a prospective evaluation. Gastrointest Endosc 2006; 63(4): 622–629. 16. Eloubeidi MA, Varadarajulu S, Desai S, et al. A prospective evaluation of an algorithm incorporating routine preoperative endoscopic ultrasound-guided fine needle aspiration in suspected pancreatic cancer. J Gastrointest Surg 2007; 11: 813–819. 17. Varadarajulu S, Tamhane A, Eloubeidi MA. Yield of EUS-guided FNA of pancreatic masses in the presence or the absence of chronic pancreatitis. Gastrointest Endosc 2005; 62(5): 728–736; quiz 751, 753. 18. Ardengh JC, Lopes CV, Campos AD, et al. Endoscopic ultrasound and fine needle aspiration in chronic pancreatitis: differential diagnosis between pseudotumoral masses and pancreatic cancer. J Pancreas 2007; 8(4): 413–421. 19. Fritscher-Ravens A, Brand L, Knofel WT, et al. Comparison of endoscopic ultrasound-guided fine needle aspiration for focal pancreatic lesions in patients with normal parenchyma and chronic pancreatitis. Am J Gastroenterol 2002; 97(11): 2768–2775.

145

146

Pancreatic Cancer

20. Khalid A, Nodit L, Zahid M, et al. Endoscopic ultrasound fine needle aspirate DNA analysis to differentiate malignant and benign pancreatic masses. Am J Gastroenterol 2006; 101(11): 2493–2500. 21. Bergenfeldt M, Genell S, Lindholm K, Ekberg O, Aspelin P. Needle-tract seeding after percutaneous fine-needle biopsy of pancreatic carcinoma. Case report. Acta Chir Scand 1988; 154(1): 77–79. 22. Bret PM, Nicolet V, Labadie M. Percutaneous fine-needle aspiration biopsy of the pancreas. Diagn Cytopathol 1986; 2(3): 221–227. 23. Fornari F, Civardi G, Cavanna L, et al. Complications of ultrasonically guided fine-needle abdominal biopsy. Results of a multicenter Italian study and review of the literature. The Cooperative Italian Study Group. Scand J Gastroenterol 1989; 24(8): 949–955. 24. Micames C, Jowell PS, White R, et al. Lower frequency of peritoneal carcinomatosis in patients with pancreatic cancer diagnosed by EUS-guided FNA vs. percutaneous FNA. Gastrointest Endosc 2003; 58(5): 690–695. 25. Horwhat JD, Paulson EK, McGrath K, et al. A randomized comparison of EUS-guided FNA versus CT or US-guided FNA for the evaluation of pancreatic mass lesions. Gastrointest Endosc 2006; 63(7): 966–975. 26. Harewood GC, Wiersema MJ, Nelson H, et al. A prospective, blinded assessment of the impact of preoperative staging on the management of rectal cancer. Gastroenterology 2002; 123(1): 24–32. 27. Snady H, Bruckner H, Siegel J, et al. Endoscopic ultrasonographic criteria of vascular invasion by potentially resectable pancreatic tumors. Gastrointest Endosc 1994; 40(3): 326–333. 28. Buscail L, Pages P, Berthelemy P, et al. Role of EUS in the management of pancreatic and ampullary carcinoma: a prospective study assessing resectability and prognosis. Gastrointest Endosc 1999; 50(1): 34–40. 29. Brugge WR, Lee MJ, Kelsey PB, Schapiro RH, Warshaw AL. The use of EUS to diagnose malignant portal venous system invasion by pancreatic cancer. Gastrointest Endosc 1996; 43(6): 561–567. 30. Rosch T, Dittler HJ, Strobel K, et al. Endoscopic ultrasound criteria for vascular invasion in the staging of cancer of the head of the pancreas: a blind reevaluation of videotapes. Gastrointest Endosc 2000; 52(4): 469–477. 31. Puli SR, Singh S, Hagedorn CH, Reddy J, Olyaee M. Diagnostic accuracy of EUS for vascular invasion in pancreatic and periampullary cancers: a meta-analysis and systematic review. Gastrointest Endosc 2007; 65(6): 788–797. 32. Snady H. EUS criteria for vascular invasion: analyzing the meta-analysis. Gastrointest Endosc 2007; 65(6): 798–807. 33. Neoptolemos JP. Is endoscopic ultrasonography superior to multidetector CT for assessing pancreatic cancer? Nat Clin Pract Oncol 2005; 2(2): 78–79. 34. Ahmad NA, Lewis JD, Siegelman ES, et al. Role of endoscopic ultrasound and magnetic resonance imaging in the preoperative staging of pancreatic adenocarcinoma. Am J Gastroenterol 2000; 95(8): 1926–1931.

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

35. Fernandez-del Castillo C, Targarona J, Thayer SP, et al. Incidental pancreatic cysts: clinicopathologic characteristics and comparison with symptomatic patients. Arch Surg 2003; 138(4): 427–433; discussion 433–434. 36. Michael H, Gress F. Diagnosis of cystic neoplasms with endoscopic ultrasound. Gastrointest Endosc Clin N Am 2002; 12(4): 719–733. 37. Balcom IJ, Fernandez-Del Castillo C, Warshaw AL. Cystic lesions in the pancreas: when to watch, when to resect. Curr Gastroenterol Rep 2000; 2(2): 152–158. 38. Le Borgne J, de Calan L, Partensky C. Cystadenomas and cystadenocarcinomas of the pancreas: a multiinstitutional retrospective study of 398 cases. French Surgical Association. Ann Surg 1999; 230(2): 152–161. 39. Warshaw AL, Compton CC, Lewandrowski K, Cardenosa G, Mueller PR. Cystic tumors of the pancreas. New clinical, radiologic, and pathologic observations in 67 patients. Ann Surg 1990; 212(4): 432–443; discussion 444–445. 40. Compagno J, Oertel JE. Microcystic adenomas of the pancreas (glycogen-rich cystadenomas): a clinicopathologic study of 34 cases. Am J Clin Pathol 1978; 69(3): 289–298. 41. Longnecker DS, Adler G, Hruban RH, Kloppel G. World Health Organization Classification of Tumors. Lyon, IARC Press. 2000. 42. Spinelli KS, Fromwiller TE, Daniel RA, et al. Cystic pancreatic neoplasms: observe or operate. Ann Surg 2004; 239(5): 651–657; discussion 657–659. 43. Brugge WR. Evaluation of pancreatic cystic lesions with EUS. Gastrointest Endosc 2004; 59(6): 698–707. 44. Brugge WR. Role of endoscopic ultrasound in the diagnosis of cystic lesions of the pancreas. Pancreatology 2001; 1(6): 637–640. 45. Van Dam J. EUS in cystic lesions of the pancreas. Gastrointest Endosc 2002; 56(4 Suppl): S91–S93. 46. Ariyama J, Suyama M, Satoh K, Wakabayashi K. Endoscopic ultrasound and intraductal ultrasound in the diagnosis of small pancreatic tumors. Abdom Imaging 1998; 23(4): 380–386. 47. Gress F, Gottlieb K, Cummings O, Sherman S, Lehman G. Endoscopic ultrasound characteristics of mucinous cystic neoplasms of the pancreas. Am J Gastroenterol 2000; 95(4): 961–965. 48. Koito K, Namieno T, Nagakawa T, et al. Solitary cystic tumor of the pancreas: EUS-pathologic correlation. Gastrointest Endosc 1997; 45(3): 268–276. 49. Ahmad NA, Kochman ML, Lewis JD, Ginsberg GG. Can EUS alone differentiate between malignant and benign cystic lesions of the pancreas? Am J Gastroenterol 2001; 96(12): 3295–3300. 50. Ahmad NA, Kochman ML, Brensinger C, et al. Interobserver agreement among endosonographers for the diagnosis of neoplastic versus non-neoplastic pancreatic cystic lesions. Gastrointest Endosc 2003; 58(1): 59–64. 51. Frossard JL, Amouyal P, Amouyal G, et al. Performance of endosonography-guided fine needle aspiration and biopsy in the diagnosis of pancreatic cystic lesions. Am J Gastroenterol 2003; 98(7): 1516–1524.

147

148

Pancreatic Cancer

52. Brugge WR, Lewandrowski K, Lee-Lewandrowski E, et al. Diagnosis of pancreatic cystic neoplasms: a report of the cooperative pancreatic cyst study. Gastroenterology 2004; 126(5): 1330–1336. 53. Recine M, Kaw M, Evans DB, Krishnamurthy S. Fine-needle aspiration cytology of mucinous tumors of the pancreas. Cancer 2004; 102(2): 92–99. 54. Hammel P, Levy P, Voitot H, et al. Preoperative cyst fluid analysis is useful for the differential diagnosis of cystic lesions of the pancreas. Gastroenterology 1995; 108(4): 1230–1235. 55. Lewandrowski KB, Southern JF, Pins MR, Compton CC, Warshaw AL. Cyst fluid analysis in the differential diagnosis of pancreatic cysts. A comparison of pseudocysts, serous cystadenomas, mucinous cystic neoplasms, and mucinous cystadenocarcinoma. Ann Surg 1993; 217(1): 41–47. 56. Moparty B, Sluss PM, Melanson SE, Lewandrowski KB, Brugge WR. Pancreatic cyst fluid CEA: Caution! Diagnostic cutoff values are assay system dependent. Gastrointest Endosc 2007; 65(5): AB300. 57. van der Waaij LA, van Dullemen HM, Porte RJ. Cyst fluid analysis in the differential diagnosis of pancreatic cystic lesions: a pooled analysis. Gastrointest Endosc 2005; 62(3): 383–389. 58. Shami VM, Sundaram V, Stelow EB, et al. The level of carcinoembryonic antigen and the presence of mucin as predictors of cystic pancreatic mucinous neoplasia. Pancreas 2007; 34(4): 466–469. 59. Linder JD, Geenen JE, Catalano MF. Cyst fluid analysis obtained by EUS-guided FNA in the evaluation of discrete cystic neoplasms of the pancreas: a prospective single-center experience. Gastrointest Endosc 2006; 64(5): 697–702. 60. Khalid A, McGrath KM, Zahid M, et al. The role of pancreatic cyst fluid molecular analysis in predicting cyst pathology. Clin Gastroenterol Hepatol 2005; 3(10): 967–973. 61. Oberg K, Eriksson B. Endocrine tumours of the pancreas. Best Pract Res Clin Gastroenterol 2005; 19(5): 753–781. 62. Anderson MA, Carpenter S, Thompson NW, et al. Endoscopic ultrasound is highly accurate and directs management in patients with neuroendocrine tumors of the pancreas. Am J Gastroenterol 2000; 95(9): 2271–2277. 63. De Angelis C, Repici A, Arena V, Pellicano R, Rizzetto M. Preoperative endoscopic ultrasonography in decision making and management for pancreatic endocrine tumors: a 6-year experience. Endoscopy 1998; 30 Suppl 1: A182–A186. 64. Gouya H, Vignaux O, Augui J, et al. CT, endoscopic sonography, and a combined protocol for preoperative evaluation of pancreatic insulinomas. Am J Roentgenol 2003; 181(4): 987–992. 65. Palazzo L, Roseau G, Chaussade S, et al. Pancreatic endocrine tumors: contribution of ultrasound endoscopy in the diagnosis of localization. Ann Chir 1993; 47(5): 419–424. 66. Zimmer T, Stolzel U, Bader M, et al. Endoscopic ultrasonography and somatostatin receptor scintigraphy in the preoperative localisation of insulinomas and gastrinomas. Gut 1996; 39(4): 562–568. 67. Rosch T, Lightdale CJ, Botet JF, et al. Localization of pancreatic endocrine tumors by endoscopic ultrasonography. N Engl J Med 1992; 326(26): 1721–1726. 68. McLean AM, Fairclough PD. Endoscopic ultrasound in the localisation of pancreatic islet cell tumours. Best Pract Res Clin Endocrinol Metab 2005; 19(2): 177–193.

Role of endoscopic ultrasound in diagnosis and staging of pancreatic neoplasms

69. Fidler JL, Fletcher JG, Reading CC, et al. Preoperative detection of pancreatic insulinomas on multiphasic helical CT. Am J Roentgenol 2003; 181(3): 775–780. 70. Legmann P, Vignaux O, Dousset B, et al. Pancreatic tumors: comparison of dual-phase helical CT and endoscopic sonography. Am J Roentgenol 1998; 170(5): 1315–1322. 71. Rappeport ED, Hansen CP, Kjaer A, Knigge U. Multidetector computed tomography and neuroendocrine pancreaticoduodenal tumors. Acta Radiol 2006; 47(3): 248–256. 72. Rockall AG, Reznek RH. Imaging of neuroendocrine tumours (CT/MR/US). Best Pract Res Clin Endocrinol Metab 2007; 21(1): 43–68. 73. Lebtahi R, Cadiot G, Sarda L, et al. Clinical impact of somatostatin receptor scintigraphy in the management of patients with neuroendocrine gastroenteropancreatic tumors. J Nucl Med 1997; 38(6): 853–858. 74. Chang F, Vu C, Chandra A, Meenan J, Herbert A. Endoscopic ultrasound-guided fine needle aspiration cytology of pancreatic neuroendocrine tumours: cytomorphological and immunocytochemical evaluation. Cytopathology 2006; 17(1): 10–17. 75. Ardengh JC, de Paulo GA, Ferrari AP. EUS-guided FNA in the diagnosis of pancreatic neuroendocrine tumors before surgery. Gastrointest Endosc 2004; 60(3): 378–384. 76. Gines A, Vazquez-Sequeiros E, Soria MT, Clain JE, Wiersema MJ. Usefulness of EUS-guided fine needle aspiration (EUS-FNA) in the diagnosis of functioning neuroendocrine tumors. Gastrointest Endosc 2002; 56(2): 291–296. 77. Gress FG, Barawi M, Kim D, Grendell JH. Preoperative localization of a neuroendocrine tumor of the pancreas with EUS-guided fine needle tattooing. Gastrointest Endosc 2002; 55(4): 594–597. 78. Bansal R, Tierney W, Carpenter S, Thompson N, Scheiman JM. Cost effectiveness of EUS for preoperative localization of pancreatic endocrine tumors. Gastrointest Endosc 1999; 49(1): 19–25.

149

9 Surgical staging and management of pancreatic adenocarcinoma Steven M. Strasberg

Introduction Adenocarcinoma of the pancreas is one of the commonest gastrointestinal tumors, with 37 000 new cases predicted in the USA in 2007 [1]. It is also one of the deadliest with over 33 000 deaths predicted in 2007. In fact, although carcinoma of the pancreas is well down on the list of cancer sites in terms of incidence, it is the fourth most common cause of all cancer deaths in both males and females [1]. Surgery offers the only hope for cure, but in most cases the extent of the disease at presentation is such that resection is not possible. In the past 25 years some aspects of surgical treatment have improved dramatically. Surgical mortality rates have fallen from 15–20% to 1–2% in high volume centers [2], and morbidity rates, especially pancreatic fistula rates, have also fallen recently [3]. There is also evidence that long-term survival is improving, as the gap between the incidence of the disease and the mortality rate is widening annually [1]. Adenocarcinoma may arise throughout the gland but is more common in the head of the pancreas. The diagnosis and management issues facing the surgeon when dealing with cancer of the head differ from those associated with cancer of the body and tail of the gland, and we will deal with these sites separately.

Adenocarcinoma of the head of the pancreas Oncologic rationale for surgical resection of adenocarcinoma of the head of the pancreas

The operation used for resection of adenocarcinoma of the head of the pancreas is pancreatico-duodenectomy or Whipple procedure (Figure 9.1). The oncologic goals of the procedure are: (1) to remove the tumor with negative microscopic margins Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

Surgical staging and management of pancreatic adenocarcinoma

Figure 9.1. Extent of resection in a Whipple procedure is shown by shaded area. A pylorus-sparing Whipple is shown. In the Standard Whipple procedure an antrectomy is also done.

(R0 resection) and (2) to resect the regional lymph nodes, defined as the first set of nodes to which a tumor in the head of the pancreas might metastasize directly. These nodes lie in the pancreatoduodenal groove, on the anterior and posterior surfaces of the head of the pancreas, at the root of the mesentery along the right side of the superior mesenteric artery (SMA), along the common hepatic artery, in the region of the pylorus and in the hepatoduodenal ligament up to the level of the right hepatic artery. These nodes are referred to as N1 nodes. N2 nodes are nodes into which the N1 nodes drain. The goal of negative margins is based on the fact that resections that leave behind macroscopic tumor (R2 resection) uniformly fail to cure and because resections which leave microscopic tumor behind (R1 resection) have poorer outcomes than resections that do not (R0) [4]. N1 nodes are resected because resection of N1 nodes improves survival over patients who do not have nodal dissection, probably because in a small cohort of such patients the disease is limited to N1 nodes. Conversely, extending nodal dissections to N2 nodes does not improve survival as shown in several randomized trials [5–7]. Margin positivity is most likely to occur in four areas of the Whipple resection. These are the intrapancreatic margin where the pancreatic neck is divided, and three “tangential” margins – the posterior margin which lies on the inferior vena cava (IVC), the portal vein groove of the pancreatic head which lies against the superior

151

152

Pancreatic Cancer

Pancreatic neck margin SMV –PV groove Uncinate margin Posterior margin

Intrapancreatic resection line SMV –PV SMA

IVC

Figure 9.2. Likeliest sites of margin positivity in the Whipple procedure (see text). SMV, superior mesenteric vein; PV, portal vein; SMA, superior mesenteric artery; IVC, inferior vena cava.

mesenteric vein (SMV)-portal vein and the uncinate margin which lies against the SMA (Figure 9.2). To achieve the best chance of obtaining a negative intrapancreatic margin a biopsy is obtained of the cut surface of the neck once the specimen is removed. If the biopsy is positive the resection may be extended in an attempt to obtain a negative intrapancreatic margin. To maximize the chance of obtaining negative tangential margins the plane of dissection should be maintained directly on the relevant vessels (i.e. the IVC, SMV-portal vein and SMA). Stapling techniques are inadequate for this purpose. Occasionally tumors in the superior part of the head of the pancreas may approach the hepatic or celiac arteries, requiring that these arteries are also skeletonized on the surfaces abutted by the tumor. In about 25% of cases the tumor is adherent to or invades the portal vein and/or SMV [3]. Provided that the extent of invasion is limited, negative margins may still be obtained by resection and reconstruction of these veins (Figure 9.3) [8]. The role of more extensive vascular resections is debatable. Tumors with circumferential encasement of the SMA or hepatic/celiac arteries are inoperable. Lesser degrees of involvement, e.g. up to 180o of abutment of the tumor to these vessels, are considered “borderline” (Figure 9.4) [9]. Such patients are currently considered candidates for downsizing therapies in order to render them operable. The same is true for more extensive involvement of the SMV-portal vein. The bile duct margin rarely is involved if the duct is transected high in the hepatoduodenal ligament. Pancreatic head carcinomas may also extend into the small bowel mesentery or mesocolon, and portions of these structures may be resected to obtain negative margins. The pylorus-sparing variant of the Whipple procedure was introduced in an attempt to reduce postoperative morbidity of the procedure and not as an adjustment of the oncologic extent of the procedure. Its use does not affect survival [10]. Total pancreatectomy may be the procedure of choice in some cases of adenocarcinoma of the pancreas. Indications include diffuse or multifocal cancers,

Surgical staging and management of pancreatic adenocarcinoma

Figure 9.3. Example of vein reconstruction in surgery for adenocarcinoma of the pancreas. A–C: preoperative images showing the venous anatomy. A, patent uninvolved superior mesenteric vein (SMV) at level of uncinate process. B, involvement of SMV at confluence with the splenic vein. C, uninvolved portal vein above neck of pancreas. A resection of the SMV and portal vein were performed with resection of the tumor, and the vein was reconstituted with a tubular autograft taken from the superficial femoral vein. D, postoperative radiograph showing patency of graft.

153

154

Pancreatic Cancer

Figure 9.4. Borderline involvement of superior mesenteric artery. This case is “borderline” as a very thin tongue of tissue can be seen extending 180o around the vessel.

mucinous cancers in which the portion of the gland uninvolved by cancer shows high-grade dysplasia of ducts, and some cancers of the neck with involvement of the confluence of the splenic and superior mesenteric veins, especially when occlusion of the uninvolved portion of the splenic vein results in splenic congestion. Diagnosis and surgical staging Clinical presentation

The classic presentation of cancer of the head of the pancreas is unremitting jaundice and pruritus without pain or with minimal pain. Substantial pain,

Surgical staging and management of pancreatic adenocarcinoma

especially back pain, suggests invasion outside of the pancreas and unresectability. Other presentations include steatorrhea or diarrhea, weight loss, pain or a combination of these symptoms. Presentation with steatorrhea/diarrhea is usually due to a tumor in the uncinate process that obstructs the pancreatic duct, but not the bile duct. Migratory thrombophlebitis (Trousseau’s sign) is uncommon and when present usually signifies metastatic disease. A history of recent onset of diabetes is obtained in 5% of patients. Jaundice and a palpable gallbladder are the chief physical signs. Liver function tests are of limited value in diagnosis. The serum bilirubin level is elevated in jaundiced patients with the direct fraction being greater than 50%. The serum alkaline phosphatase is almost always elevated when the bile duct is obstructed, and levels 3–5 times normal are common. Serum Ca19–9 concentration is a useful tumor marker, which is often elevated in adenocarcinomas of pancreatic origin. The upper limit of normal is 37 IU. Concentrations over 100 IU are highly suggestive of malignancy, but elevations between 37–100 IU are less specific. Serum levels reflect the extent of tumor. Small tumors, less than 1 cm in size, are not usually associated with elevations over 100 IU. Very high levels in the thousands suggest metastatic disease. High levels may also accompany cholangitis. Diagnostic imaging – a surgical perspective

When a patient presents with painless jaundice the choice of initial imaging test is broad, including computed tomography (CT), magnetic resonance imaging (MRI), endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic ultrasound (EUS). Three types of CT findings may be obtained in such patients – typical mass, atypical mass and no mass. Subsequent management depends upon which of these is present. Typical mass

Patients who have typical hypoattenuating masses and whose tumors appear to be resectable need no other interventions prior to surgery unless surgery is to be delayed, for example to minimize the effect of comorbidities or to administer neoadjuvant therapies. Endoscopic retrograde cholangiopancreatography is not necessary for diagnosis, staging or for bile duct decompression in such patients. Indeed, preoperative stenting may be associated with negative outcomes [11]. Therefore CT should be selected over ERCP as the first imaging test in patients with new-onset jaundice of the type described above. Endoscopic ultrasound is more invasive and its advantage of providing biopsy material is negated by the fact that the typical mass on CT is almost always cancer, making biopsy unnecessary

155

156

Pancreatic Cancer

before resection. Magnetic resonance imaging theoretically is equivalent to CT, but in practice the quality of CT performed in most hospitals is higher than that of MRI. Computed tomography scans are also easier for most non-radiologists to interpret. Atypical mass

When a pancreatic head mass is atypical on CT the first issue to resolve is differentiation of a malignant from a benign process, most notably focal pancreatitis. Pancreatitis may occur without antecedent acute attacks, without a history of alcoholism, gallstones, or hyperlipidemia, without the presence of diabetes or steatorrhea and without calcifications in the gland. Diagnosis in such cases requires a biopsy of the mass in the pancreatic head. Endoscopic ultrasound has become invaluable in this role [12]. It is minimally invasive and because the biopsies are transduodenal the risk of needle tracking having an effect on outcome is eliminated. Other approaches to obtaining biopsies such as percutaneous CT or ultrasoundguided techniques or laparoscopic biopsy do not have this advantage and are less convenient or more invasive. No mass on CT scan

When no mass is seen on CT a pancreatic head malignancy may be suspected by the presence of a dilated bile duct or pancreatic duct terminating in the head. Endoscopic ultrasound may be slightly more sensitive than CT for small pancreatic head masses and is very useful in such circumstances for identifying and taking a biopsy of the mass. When no mass is seen on EUS, ERCP can be performed at the same sitting to obtain brush biopsies of the bile duct and delineate the radiologic characteristics of the bile duct stricture, i.e. whether benign or malignant appearing. Both pancreatic and bile ducts may be brushed for cytology. This test has a 45–50% sensitivity when cancers are present. Therefore, only a positive test is helpful. In some cases when an atypical mass or no mass is demonstrated on CT, laparoscopic or even open ultrasound-guided biopsies may be needed to diagnose or rule out cancer. Surgical staging (determination of resectability)

The term “surgical staging” (as opposed to pathologic staging) denotes the steps taken to determine whether a tumor is resectable. Surgical staging is started preoperatively and completed intraoperatively. Preoperative staging tests determine operability, i.e. whether the tumor appears resectable after preoperative testing. However, staging is continued intraoperatively and the final decision that the tumor is resectable is made only during the operation. Common reasons for unresectability are (1) vascular

Surgical staging and management of pancreatic adenocarcinoma

invasion with obstruction of the superior mesenteric and/or portal veins, or encasement of the superior mesenteric artery, and less commonly the hepatic or celiac artery; (2) lymph node metastases outside the scope of the pancreaticoduodenectomy such para-aortic and celiac lymph nodes; (3) hepatic metastases; (4) peritoneal metastases; (5) extra-abdominal metastases (usually pulmonary). As previously noted limited vascular invasion of the superior mesenteric and portal veins may be overcome by resection and reconstruction and therefore is only a relative contraindication to resection. This is especially true when the tumor is small and has arisen in the vicinity of the veins. In our experience about 20% of pancreatic cancer resections include resection of these veins [3]. Tests to establish diagnosis and to accomplish surgical staging go hand in hand. Abdominal CT or MRI, and thoracic CT or radiographs are used to detect hepatic metastases, vascular invasion and pulmonary metastases. To assess vascular invasion, thin section multiphase helical CT or MRI is required (for a complete discussion of appropriate CT and MRI technique, see Chapters 3 and 5). Enlarged lymph nodes are detectable by these tests but nodes may be enlarged for reasons other than cancer, and normal size nodes may contain metastatic disease. When ascites or peritoneal or omental nodules are identified, ascitic fluid aspirates may be sent for cytology, and omental nodules can be biopsied under ultrasound guidance or by laparoscopy. Invasion of the mesentery, mesocolon or retroperitoneal tissues also may be detected by CT. In the author’s experience the presence of such findings in the absence of concomitant vascular invasion, usually does not prevent resection with negative margins if the involved portion of the mesocolon or mesentery is resected. Endoscopic ultrasound may be used to biopsy suspicious lymph nodes that are outside the planned resection zone. Staging laparoscopy is particularly useful for identifying small hepatic and peritoneal nodules. About 20% of patients thought to have resectable pancreatic adenocarcinoma of the head of the pancreas prior to staging laparoscopy are found to have liver or peritoneal metastases at laparoscopy [13]. Staging is completed at surgery by careful inspection of the intra-abdominal contents, opening the lesser sac, mobilizing the head of the pancreas, performing biopsies of suspicious nodules or lymph nodes outside the zone of planned resection and attempting dissection of the tumor off the superior mesenteric/portal vein. All authorities agree that cross-sectional imaging of the abdomen and thorax (or chest radiographs) is standard practice for staging pancreatic cancer. However, there is controversy regarding the value of other staging tests. Many advocate omission of staging laparoscopy or EUS biopsy of lymph nodes on the grounds that patients are served better by palliative surgery than endoscopic stenting of the

157

158

Pancreatic Cancer

bile duct. There is no advantage to knowing if small liver metastases or celiac node metastases are present if laparotomy is to be undertaken anyway. Despite a number of trials it is not clear whether patients who are resectable after standard preoperative staging but not operable at the time of laparotomy or laparoscopy are better served by surgical double bypass or by endoscopic stenting. Finally FDG-PET may be useful in staging of pancreatic cancer, but its role is unclear. As inflammation can be confused with cancer the major role is in detection of distant metastases. Results of resection for adenocarcinoma of the head of the pancreas

Mortality rates after pancreatico-duodenectomy were 15–25% in the 1970s, but by 1990 had fallen to 1–2% in high volume centers [2]. Mortality rates remain high in low volume centers even today [14]. Postoperative morbidity is high, but the incidence of pancreatic fistula has fallen with the introduction of new techniques [3]. Delayed gastric empting remains a common problem [15] after both the standard and pylorus sparing variants, but is a self-limiting complication. Quality of life following pancreatico-duodenectomy is poor for 3–6 months but then returns to normal in most patients unless there is recurrence [16]. Although safety of pancreatic head resection has improved greatly in the past 25 years, the oncologic effectiveness of the procedure is still quite poor. Actuarial 5-year survival rates are about 20% [17,18] In one series of over 100 patients the actual 5-year survival rate was 15% [19], which is probably closer to the true outcome since actuarial survivals tend to overestimate survival rates [20]. Deaths due to recurrences presenting for the first time after 5 years are not unusual. Prognostic factors favoring long-term survival include R0 resection, absence of metastases in lymph nodes, low positive to total node ratio, absence of vascular or perineural invasion and moderateor well-differentiated grade of tumor [19,21–24]. The poor results of surgical resection are due to the aggressive local invasiveness of this tumor and its tendency to metastasize to the liver and other distant sites while the tumor is still quite small. Therefore at the time of resection positive microscopic margins are quite common (about 25% even when the procedure is performed at highly specialized centers). Progress in the treatment of this cancer will come from two sources – the introduction of techniques to detect cancers when they are quite small (i.e. 1 cm or less) at which point many tumors will still be entirely local, and the availability of effective chemotherapy. The latter would be particularly important both to eradicate minimal disease (either regional or distantly metastatic) which remains after resection and to downsize disease in an effective manner so that patients become operable.

Surgical staging and management of pancreatic adenocarcinoma

Adenocarcinoma of the body and tail of the pancreas Oncologic rationale for surgical resection of adenocarcinoma of the body and tail of the pancreas; radical antegrade modular pancreato-splenectomy (RAMPS)

This tumor is less common than adenocarcinoma of the head of the pancreas. The rationale for resection of body and tail adenocarcinomas logically should be the same as for those in the head of the pancreas, i.e. resection of the tumor and the N1 lymph nodes with negative tumor margins. In practice this result has not been achieved by the usual approach of retrograde (spleen first) distal pancreatectomy. With this traditional procedure, which is not based on the lymph node drainage of the pancreas, lymph node counts have been low, and positive tangential margin rates, especially posterior margin rates, have been high. A recently described technique called radical antegrade modular pancreato-splenectomy (RAMPS) accomplishes the desired goals through an antegrade resection from right-to-left which is based on the established lymph node drainage of the gland [25]. The procedure also allows early control of the vasculature. The extent of the node dissection is shown in Figure 9.5. The procedure is modular in that the posterior extent may be anterior or posterior to the adrenal gland depending upon whether preoperative CT has demonstrated tumor extension to or beyond the posterior border of the pancreas (Figure 9.6). When the tumor has not

gastrosplenic nodes celiac nodes

superior mesenteric nodes

gastroduodenal nodes

Splenic nodes infrapancreatic nodes

Figure 9.5. Extent of lymph node dissection in the radial antegrade modular pancreato-splenectomy (RAMPS) procedure.

159

160

Pancreatic Cancer

Figure 9.6. Left. Schematic showing the extent of the anterior radial antegrade modular pancreato-splenectomy (RAMPS) procedure. Right. Operative photograph showing the extent of the posterior RAMPS procedure. a, stump of pancreas; b, kidney; c, portal vein; d, celiac artery; e, superior mesenteric artery (SMA); f, aorta; g, renal vein. Note in the anterior RAMPS procedure the adrenal gland is not resected.

reached the posterior margin of the pancreas an anterior adrenal preserving RAMPS procedure is performed, whereas if the tumor has reached the posterior border of the pancreas a posterior adrenal resecting RAMPS operation is used (Figure 9.7). Carcinoma of the body and tail of the pancreas is resectable when it is confined to the gland or invades local structures such as the spleen or stomach, which also can be resected. Tumors of the mid body tend to invade posteriorly to involve the SMA, celiac axis, SMV or portal vein, even when the tumor is only 2–3 cm in size. Consequently tumors of the tail of the pancreas are more likely to be resectable than tumors of the mid-body (Figure 9.8). Considerations regarding limitations of resectability due to vascular invasion are the same as for cancers of the head of the pancreas. The commonest sites of positive margins are the posterior tangential margin followed by the intrapancreatic neck margin. Diagnosis and staging

Cancer of the body and tail tends to present at a later stage than cancer of the head of the pancreas because it does not produce jaundice. Therefore, presentation at an advanced stage of disease is common. Symptoms are non-specific and include weight loss, recent onset of diabetes, and abdominal and back pain, classically

Surgical staging and management of pancreatic adenocarcinoma

Figure 9.7. A, Abdominal computed tomography of patient with adenocarcinoma that has reached both the anterior and posterior surfaces of the pancreas (arrows). B, plane of dissection in the posterior radical antegrade modular pancreato-splenectomy (RAMPS) procedure passes behind the adrenal gland in order to avoid a positive posterior margin.

Figure 9.8. Abdominal computed tomography showing an extensive but resectable tumor of the tail of the pancreas. Such a large tumor would almost certainly be unresectable if it were in the body of the pancreas due to vascular invasion of the superior mesenteric artery or celiac artery.

161

162

Pancreatic Cancer

relieved by sitting up and leaning forward. The serum Ca 19–9 may be elevated. Computed tomography, which is the primary diagnostic test, typically shows a hypoattenuating mass (Figures 9.7 and 9.8), often with extension outside of the pancreas. The upstream pancreatic duct usually is dilated when the tumor is proximal to the tail of the gland. As with pancreatic head tumors, typical lesions do not require biopsy prior to surgical resection. Endoscopic ultrasound is very useful for diagnosis of indeterminate lesions. Surgical staging is performed as for cancers of the head of the pancreas and relies on CT of the abdomen and thorax. Unresectability due to local invasion usually is due to involvement of the superior mesenteric or celiac artery and less commonly the portal or superior mesenteric vein or the aorta. Another indicator of unresectability is enlarged para-aortic lymph nodes, which sometimes appear along the aorta below the pancreas. Invasion of the spleen, stomach, left adrenal gland, mesocolon, colon, retroperitoneum or even the left kidney are not contraindications to resection provided negative margins can be expected [26]. Staging laparoscopy is of great value in this disease as 20–50% of patients are found to be unresectable with this investigation [13]. Unlike cancer of the head of the pancreas there is no effective surgical palliation for cancer of the body or tail and therefore no rationale for laparotomy in unresectable patients. Celiac nerve block, which is very helpful in reducing use of narcotics for pain control, can also be performed laparoscopically or endoscopically. Results of resection for adenocarcinoma of the body and tail of the pancreas

Case series reporting long-term survival, especially those reporting margin status, are few and small in size. Shoup et al. reported on 57 patients[27], 28% of whom had positive margins. Overall survival, reported as “disease-specific survival” was 16 months, and the 5-year survival was about 15%. Christein et al. reported on 93 patients, 27 of whom had carcinomas arising in cystic diseases of the pancreas and 66 of whom had ductal adenocarcinoma [28]. Eighty-three percent had negative margins, but tangential (“radial”) margins were reported in only 62% of specimens. The negative margin rate fell to 73% in patients who had an adjacent organ resected. The median survival in the 66 patients with ductal adenocarcinoma was 16 months, and the 5-year survival about 5%. Shimada et al. (National Cancer Hospital, Tokyo) reported on 88 patients, 75% of whom had negative margins [29]. The median survival was 22 months, and the 5-year overall survival was 19%. We recently reported on 23 patients treated by the RAMPS procedure [26]. Negative tangential margins were attained in 91% of patients, the median survival was

Surgical staging and management of pancreatic adenocarcinoma

21 months, and the 5-year overall survival was 26%. It is difficult to compare the different series in a way that permits confident conclusions, as there is not a standard reporting schema. The modular approach of RAMPS based on high quality CT done in the immediate preoperative period seems to achieve a high rate of negative margins and acceptable survival in this very aggressive cancer, but further study is needed. Mucinous adenocarcinomas of the pancreas

Mucin-producing cancer is a special variant of pancreatic adenocarcinoma, which often arises in a pre-existing lesion – namely mucinous cystic neoplasm (MCN) or intraductal papillary mucinous neoplasm (IPMN). A discussion of the management of these cystic lesions in their premalignant state is beyond the purposes of this chapter. Once the tumors become malignant the issues regarding staging and surgical treatment are nearly the same as those discussed above. One difference is that when high-grade dysplasia or carcinoma in-situ is present in the part of the pancreas not affected by malignancy, a total pancreatectomy may be advisable. Another difference is that these tumors have a better prognosis than standard ductal adenocarcinomas of the pancreas, with a 5-year overall survival rate on the order of 45% [30].

REFERENCES 1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2007. CA Cancer J Clin 2007; 57: 43–66. 2. Strasberg SM, Drebin JA, Soper NJ. Evolution and current status of the Whipple procedure: an update for gastroenterologists. Gastroenterology 1997; 113(3): 983–994. 3. Strasberg SM, Drebin JA, Mokadam NA, et al. Prospective trial of a blood supply based technique of pancreaticojejunostomy: effect on anastomotic failure in the Whipple procedure. J Am Coll Surg 2002; 194: 748–758. 4. Neoptolemos JP, Stocken DD, Dunn JA, et al. Influence of resection margins on survival for patients with pancreatic cancer treated by adjuvant chemoradiation and/or chemotherapy in the ESPAC-1 randomized controlled trial. Ann Surg 2001; 234: 758–768. 5. Farnell MB, Pearson RK, Sarr MG, et al. A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma. Surgery 2005; 138: 618–628. 6. Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, part 2: randomized controlled trial evaluating survival, morbidity, and mortality. Ann Surg 2002; 236: 355–366.

163

164

Pancreatic Cancer

7. Pedrazzoli S, DiCarlo V, Dionigi R, et al. Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group. Ann Surg 1998; 228: 508–517. 8. Bold RJ, Charnsangavej C, Cleary KR, et al. Major vascular resection as part of pancreaticoduodenectomy for cancer: radiologic, intraoperative, and pathologic analysis. J Gastrointest Surg 1999; 3: 233–243. 9. Varadhachary GR, Tamm EP, Abbruzzese JL, et al. Borderline resectable pancreatic cancer: definitions, management, and role of preoperative therapy. Ann Surg Oncol 2006; 13: 1035–1046. 10. Seiler CA, Wagner M, Bachmann T, et al. Randomized clinical trial of pylorus-preserving duodenopancreatectomy versus classical Whipple resection – long term results. Br J Surg 2005; 92: 547–556. 11. Povoski SP, Karpeh MS Jr., Conlon KC, et al. Preoperative biliary drainage: impact on intraoperative bile cultures and infectious morbidity and mortality after pancreaticoduodenectomy. J Gastrointest Surg 1999; 3: 496–505. 12. Raut CP, Grau AM, Staerkel GA, et al. Diagnostic accuracy of endoscopic ultrasound-guided fine-needle aspiration in patients with presumed pancreatic cancer. J Gastrointest Surg 2003; 7: 118–126. 13. Vollmer CM, Drebin JA, Middleton WD, et al. Utility of staging laparoscopy in subsets of peripancreatic and biliary malignancies. Ann Surg 2002; 235: 1–7. 14. McPhee JT, Hill JS, Whalen GF, et al. Perioperative mortality for pancreatectomy: a national perspective. Ann Surg 2007; 246: 246–253. 15. Paraskevas KI, Avgerinos C, Manes C, et al. Delayed gastric emptying is associated with pylorus-preserving but not classical Whipple pancreaticoduodenectomy: a review of the literature and critical reappraisal of the implicated pathomechanism. World J Gastroent 2006; 12: 5951–5958. 16. Schniewind B, Bestmann B, Henne-Bruns D, et al. Quality of life after pancreaticoduodenectomy for ductal adenocarcinoma of the pancreatic head. Br J Surg 2006; 93: 1099–1107. 17. Zacharias T, Jaeck D, Oussoultzoglou E, et al. Impact of lymph node involvement on long-term survival after R0 pancreaticoduodenectomy for ductal adenocarcinoma of the pancreas. J Gastrointest Surg 2007; 11: 350–356. 18. Winter JM, Cameron JL, Campbell KA, et al. 1423 pancreaticoduodenectomies for pancreatic cancer: a single-institution experience. J Gastrointest Surg 2006; 10: 1199–1210. 19. Cleary SP, Gryfe R, Guindi M, et al. Prognostic factors in resected pancreatic adenocarcinoma: analysis of actual 5-year survivors. J Am Coll Surg 2004; 198: 722–731. 20. Smeenk HG, Tran TCK, Erdmann J, et al. Survival after surgical management of pancreatic adenocarcinoma: does curative and radical surgery truly exist? Langenbecks Arch Surg 2005; 390: 94–103. 21. Mitsunaga S, Hasebe T, Iwasaki M, et al. Important prognostic histological parameters for patients with invasive ductal carcinoma of the pancreas. Cancer Sci 2005; 96: 858–865.

Surgical staging and management of pancreatic adenocarcinoma

22. Jarufe NP, Coldham C, Mayer AD, et al. Favourable prognostic factors in a large UK experience of adenocarcinoma of the head of the pancreas and periampullary region. Dig Surg 2004; 21: 202–209. 23. Kuhlmann KF, de Castro SM, Wesseling JG, et al. Surgical treatment of pancreatic adenocarcinoma; actual survival and prognostic factors in 343 patients. Eur J Cancer 2004; 40: 549–558. 24. Pawlik TM, Gleisner AL, Cameron JL, et al. Prognostic relevance of lymph node ratio following pancreaticoduodenectomy for pancreatic cancer. Surgery 2007; 141: 610–618. 25. Strasberg SM, Drebin JA, Linehan D. Radical antegrade modular pancreatosplenectomy. Surgery 2003; 133(5): 521–527. 26. Strasberg SM, Linehan DC, Hawkins WG. Radical antegrade modular pancreatosplenectomy procedure for adenocarcinoma of the body and tall of the pancreas: ability to obtain negative tangential margins. J Am Coll Surg 2007; 204(2): 244–249. 27. Shoup M, Conlon KC, Klimstra D, Brennan MF. Is extended resection for adenocarcinoma of the body or tail of the pancreas justified? J Gastrointest Surg 2003; 7: 946–952. 28. Christein JD, Kendrick ML, Iqbal CW, et al. Distal pancreatectomy for resectable adenocarcinoma of the body and tail of the pancreas. J Gastrointest Surg 2005; 9: 922–927. 29. Shimada K, Sakamoto Y, Sano T, Kosuge T. Prognostic factors after distal pancreatectomy with extended lymphadenectomy for invasive pancreatic adenocarcinoma of the body and tail. Surgery 2006; 139: 288–295. 30. Sohn TA, Yeo CJ, Cameron JL, et al. Intraductal papillary mucinous neoplasms of the pancreas: an updated experience. Ann Surg 2004; 239: 788–797.

165

10 Treatment of locally advanced and metastatic pancreatic cancer Benjamin R. Tan, Jeffrey Carenza, and Joel Picus

Introduction Despite recent advances in multi-modality and targeted antineoplastic therapy, efforts to improve clinical outcomes for patients with pancreatic adenocarcinoma remain challenging. The potential for cure is applicable only to patients whose tumors are initially resectable or those with borderline, potentially resectable locally advanced tumors. However, these patients constitute < 10% of all newly diagnosed patients with pancreatic cancer [1]. For most patients with pancreatic cancer, the goal of therapy is palliation, and the prognosis is very poor. Patients with locally advanced pancreatic cancer have a median life expectancy of 6–10 months, and patients with metastatic disease have an even shorter median survival of approximately 3–6 months [2]. Current standard and experimental therapeutic options for patients with locally advanced and metastatic pancreatic cancer are evolving and are discussed below.

Locally advanced pancreatic adenocarcinoma A multi-disciplinary approach is required for the proper assessment and treatment of patients with locally advanced pancreatic cancer. Optimal imaging of the pancreas and its surrounding tissues may delineate tumors that truly are unresectable and those with “borderline” resectability. Whereas the median survival of patients with unresected locally advanced pancreatic cancer is 6–10 months [2], survival times of 21 months can potentially be achieved in patients who have had resection of their tumor after preoperative therapy [3]. This survival time is comparable to the median survival time achieved with upfront pancreatic cancer resection followed by adjuvant therapy. Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

Treatment of locally advanced and metastatic pancreatic cancer

(a)

(b)

Figure 10.1. Coronal (a) and sagittal (b) multi-planar reformatted computed tomography images of the abdomen demonstrate an infiltrating mass within the head of the pancreas (asterisk). The mass abuts the anterior undersurface of the portal-splenic vein confluence with loss of the fat plane between the mass and the vein (arrows).

High resolution imaging with computed tomography (CT) or magnetic resonance imaging (MRI) generally is used to establish anatomic relationships between the tumor and its surrounding tissues, with a positive predictive value of 91% and a negative predictive value of 79%, for unresectability [4]. The criteria for resectability of pancreatic adenocarcinoma are reviewed in Chapter 5. It is generally agreed that tumors with superior mesenteric artery (SMA), aortic, inferior vena cava (IVC) or celiac encasement and those with para-aortic and celiac lymph node metastases are unresectable (Figures 10.1a,b). Tumors abutting the SMA, or those that have portal vein or superior mesenteric vein impingement or with limited IVC involvement may be of borderline resectability (Figures 10.2a, b, c). However, direct visualization of the tumor via laparotomy or laparoscopic examination ultimately determines tumor resectability, as occult metastases may be found during these procedures, thereby precluding curative resection. For patients with borderline resectable pancreatic cancer wherein an incomplete R1 or R2 resection is anticipated based on imaging, neoadjuvant therapy has been proposed to increase rates of R0 resection, thereby possibly improving survival [5]. Preoperative chemotherapy or chemoradiation may also provide earlier treatment for micrometastatic disease, and in the setting of rapidly progressing disease, avoid unnecessary surgery for those patients. Currently, controversies exist regarding the practice of neoadjuvant therapy for pancreatic cancer, the use of radiation therapy, the choice of chemotherapy and the

167

168

Pancreatic Cancer

(a)

(b)

Figure 10.2. Transverse CT images of the abdomen in a patient with unresectable pancreatic cancer. A diffusely infiltrating mass (arrows) encases the celiac axis (a) and narrows the portal-splenic vein confluence (b). Soft tissue nodules along the course of middle colic vein and within the greater omentum (arrows) indicate peritoneal carcinomatosis (c).

(c)

duration of therapy. Most studies utilize 5 fluorouracil (5FU)-based chemoradiation, which appears to have a better therapeutic and safety index compared to gemcitabine-based chemoradiation [6]. Combination chemotherapy without radiation has also been evaluated in a small phase II study using gemcitabine plus docetaxel. Of the 61 patients treated, 48 (71%) had subsequent resection which resulted in a 1-year survival rate of 85% and a 3-year survival of 69% [7]. There is a general lack of consensus regarding the optimal management of patients with unresectable locally advanced pancreatic cancer. The conventional recommendation for the use of chemoradiation with 5FU stemmed from the Gastrointestinal Tumor Study Group (GITSG) study published in 1981 [8]. Split course radiation at 40–60 Gy with concurrent bolus and maintenance 5FU doubled

Treatment of locally advanced and metastatic pancreatic cancer

median survival times from 22.9 weeks to 42.2 weeks compared with radiation alone. Forty percent of patients treated with chemoradiation survived at 1 year compared with only 10% with radiation alone. Similar survival rates were observed among patients treated with chemoradiation using 40 Gy and 60 Gy. A subsequent GITSG trial confirmed a survival advantage for chemoradiation followed by chemotherapy using streptozotocin, mitomycin and 5FU (SMF) compared with SMF alone [9]. However, a larger study using bolus 5FU showed no difference in survival for patients treated with either 5FU-based chemoradiation or 5FU alone [10]. Radiation techniques have evolved since these initial publications, and split-course therapy is now infrequently utilized. In addition, infusional 5FU has been the favored administration with radiation and gemcitabine has replaced 5FU as standard therapy for patients with metastatic disease. Gemcitabine has been found to have significant radiosensitizing properties, although, with concurrent conventional radiation, doses needed to be markedly reduced to avoid excessive toxicities [11, 12]. In a very small 34-patient randomized study, chemoradiation using gemcitabine followed by maintenance gemcitabine conferred a significant survival advantage compared with 5FU-based chemoradiation followed by gemcitabine [13]. The high incidence of occult metastatic disease even in radiologically or laparoscopically assessed cases argues for the earlier use of adequate systemic therapy rather than locoregional therapy with radiation. Preliminary results from a French phase III study comparing chemoradiation with cisplatin and 5FU followed by gemcitabine versus systemic chemotherapy with gemcitabine alone showed a significantly inferior survival for patients treated with chemoradiation [14]. Median survival times and 1-year survival rates for chemoradiation and gemcitabine alone were 8.4 months and 24% versus 14.3 months and 51%, respectively. Sequential administration of chemotherapy followed by consolidation chemoradiotherapy is a rational, albeit investigational, approach for the treatment of locally advanced pancreatic cancer, although it has not been validated in randomized studies. A retrospective analysis of 181 patients with locally advanced pancreatic cancer enrolled in phase II and III European studies demonstrated that after 3 months of chemotherapy 30% of patients develop metastatic disease [15]. Among patients with no disease progression after initial chemotherapy, consolidation with chemoradiation resulted in significantly longer progression-free (10.8 vs. 7.4 months) and overall survival (15 vs. 11.7 months) compared to those patients who continued to receive chemotherapy alone. A systematic review of 323 patients treated at MD Anderson compared patients treated with chemoradiation as their initial therapy with patients treated with

169

170

Pancreatic Cancer

chemotherapy followed by consolidation chemoradiation [16]. Median progressionfree and overall survivals were superior with the induction chemotherapy followed by consolidation chemoradiation compared with initial chemoradiation (6.4 months and 11.9 months vs. 4.2 months and 8.5 months, respectively). Treatment of locally advanced pancreatic adenocarcinoma continues to be refined. The importance of identifying patients with truly unresectable, and thus incurable, tumors and those with borderline resectability, who are potentially curable with surgery, cannot be over-emphasized. Quality imaging is crucial, along with an experienced multi-disciplinary team consisting of radiologists, gastroenterologists, surgeons, radiation and medical oncologists, as well as their ancillary staff.

Metastatic pancreatic adenocarcinoma Prognosis is grim for patients with metastatic pancreatic cancer. Palliative chemotherapy with gemcitabine-based regimens is standard and confers a median survival of approximately 6 months. In the pivotal study by Burris et al., gemcitabine, given as a 30-minute intravenous infusion weekly × 7, followed by a 1-week break, then weekly × 3 of 4 weeks resulted in a significant improvement in overall survival and clinical benefit rate compared with 5FU in patients with advanced symptomatic pancreatic cancer. Median survival times and 1-year survival rates were 5.7 months and 18% with gemcitabine compared with 4.2 months and 2% with 5FU. However, response rates were low at 5.4% [17]. Multiple phase III studies comparing gemcitabine with combination chemotherapy in the treatment of pancreatic cancer have met with limited progress. Although response rates were increased with certain regimens, overall survival was not significantly improved. Only one preliminary report of gemcitabine plus capecitabine, an oral pro-drug of 5FU, showing a median survival and 1-year survival rate of 7.4 months and 26% compared with 6 months and 19% with gemcitabine alone, achieved statistical significance (P = 0.014) [18]. In a recent meta-analysis incorporating 51 randomized studies involving 9970 patients, gemcitabine combinations, specifically with platinum compounds and capecitabine resulted in improved survivals with hazard ratios of approximately 0.85 (confidence intervals 0.72–0.96) compared with single agent gemcitabine alone [19]. Various biologic and targeted agents also have been investigated either as a single agent or in combination with gemcitabine in the treatment of advanced pancreatic cancer. Initial studies using matrix metalloproteinases were disappointing [20–22]. Bevacizumab, an anti-vascular endothelial growth factor (VEGF) monoclonal

Treatment of locally advanced and metastatic pancreatic cancer

antibody, did not improve outcomes for pancreatic cancer patients compared to gemcitabine alone [23, 24]. Likewise, cetuximab, an anti-epidermal growth factor receptor (EGFR) monoclonal antibody did not result in a significant improvement in overall survival when combined with gemcitabine [25, 26]. Erlotinib, a small molecule anti-EGFR tyrosine kinase inhibitor, combined with gemcitabine did confer a significantly better survival compared with gemcitabine alone [27]. Although the median survival advantage appears modest with the addition of erlotinib to gemcitabine (6.24 months vs 5.91 months), the hazard ratio was 0.82 with a 1-year overall survival rate of 23% compared with 17% with gemcitabine alone. Interestingly, patients who develop grade 2 or greater acneiform rash secondary to erlotinib have a 10.5-month median survival and a 1-year survival of 43% compared with 5.29 months and 16%, respectively, for patient who did not develop a rash. Symptomatic management

Symptomatic management of patients with metastatic pancreatic cancer is paramount especially considering their short survival period. Patients with celiac involvement may develop excruciating pain and can be managed with narcotics, and possibly neurolytic celiac block [28]. Local progression of pancreatic tumors also may result in gastric outlet obstruction and require a surgical bypass procedure or alternative nutritional interventions. Malignant ascitis can be managed with diuretics and therapeutic paracentesis. Biliary stents and drains may relieve pruritus and cholangitis. Anticipation of impending complications resulting from disease progression or therapy through proper imaging may direct palliative management. Efforts to identify early prognostic and predictive factors are currently under way. Once pancreatic adenocarcinoma develops, outcomes are invariably poor, even for patients with resected pancreatic cancer. Strategies for endoscopic ultrasound and radiologic screening of high-risk patients with hereditary pancreatic cancer have been proposed [29]. The identification and management of pre-cancerous pancreatic lesions such as pancreatic intraepithelial neoplasia (panIN) and intraductal papillary mucinous neoplasia (IPMN) are being debated. Pancreatic juice analysis and proteomic and DNA array analyses may yield prognostic information. A committed effort to decrease pancreatic cancer incidence and mortality necessitates incorporation of behavioral and societal interventions, such as smoking cessation programs, which may have the greatest potential impact in the prevention of this deadly disease.

171

172

Pancreatic Cancer

REFERENCES 1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2007. CA Cancer J Clin 2007; 57(1): 43–66. 2. Evans D, Abbruzzesse J, Willett C. Cancer of the pancreas. In: DeVita V, Helmman S, Rosenberg S, eds. Cancer: Principles and Practice of Oncology, 6th Edn, 2001, Philadelphia, Lippincott Williams & Wilkins. 2001. 3. Breslin TM, et al. Neoadjuvant chemoradiotherapy for adenocarcinoma of the pancreas: treatment variables and survival duration. Ann Surg Oncol, 2001. 8(2): p. 123–132. 4. Diehl SJ, Lehmann KJ, Sadick M, Lachmann R, Georgi M. Pancreatic cancer: value of dual-phase helical CT in assessing resectability. Radiology 1998; 206(2): 373–378. 5. Lowy AM. Neoadjuvant therapy for pancreatic cancer. J Gastrointest Surg 2008; epublished Feb 8 2008. 6. Crane CH, et al. Is the therapeutic index better with gemcitabine-based chemoradiation than with 5-fluorouracil-based chemoradiation in locally advanced pancreatic cancer? Int J Radiat Oncol Biol Phys 2002; 52(5): 1293–1302. 7. Gnant M, Kuehrer I, Teleky B. Effect of neoadjuvant chemotherapy with gemcitabine and docetaxel on 3-year survival and resection rate in previously unresectable locally advanced pancreatic cancer. J Clin Oncol 2004; 22: 4234. 8. Moertel CG, et al. Therapy of locally unresectable pancreatic carcinoma: a randomized comparison of high dose (6000 rads) radiation alone, moderate dose radiation (4000 rads + 5-fluorouracil), and high dose radiation + 5-fluorouracil. The Gastrointestinal Tumor Study Group. Cancer 1981; 48(8): 1705–1710. 9. Gastrointestinal Tumor Study Group. Treatment of locally unresectable carcinoma of the pancreas: comparison of combined-modality therapy (chemotherapy plus radiotherapy) to chemotherapy alone. Gastrointestinal Tumor Study Group. J Natl Cancer Inst 1988; 80(10): 751–755. 10. Klaassen DJ, MacIntyre JM, Catton GE, Engstrom PF, Moertel CG. Treatment of locally unresectable cancer of the stomach and pancreas: a randomized comparison of 5-fluorouracil alone with radiation plus concurrent and maintenance 5-fluorouracil – an Eastern Cooperative Oncology Group study. J Clin Oncol 1985; 3(3): 373–378. 11. Ikeda M, Okada S, Tokuuye K, Ueno H, Okusaka T. A phase I trial of weekly gemcitabine and concurrent radiotherapy in patients with locally advanced pancreatic cancer. Br J Cancer 2002; 86(10): 1551–1554. 12. Blackstock AW, et al. Phase I trial of twice-weekly gemcitabine and concurrent radiation in patients with advanced pancreatic cancer. J Clin Oncol 1999; 17(7): 2208–2212. 13. Li CP, et al. Concurrent chemoradiotherapy treatment of locally advanced pancreatic cancer: gemcitabine versus 5-fluorouracil, a randomized controlled study. Int J Radiat Oncol Biol Phys 2003; 57(1): 98–104.

Treatment of locally advanced and metastatic pancreatic cancer

14. Chauffert B, Mornex F, Bonnetain F. Phase III trial comparing initial chemoradiotherapy (intermittent cisplatin and infusional 5FU) followed by gemcitabine versus gemcitabine alone in patients with locally advanced nonmetastatic pancreatic cancer. J Clin Oncol 2006; 24: 4008. 15. Huguet F, et al. Impact of chemoradiotherapy after disease control with chemotherapy in locally advanced pancreatic adenocarcinoma in GERCOR phase II and III studies. J Clin Oncol 2007; 25(3): 326–331. 16. Krishnan S, et al. Induction chemotherapy selects patients with locally advanced, unresectable pancreatic cancer for optimal benefit from consolidative chemoradiation therapy. Cancer 2007; 110(1): 47–55. 17. Burris HA, 3rd, et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol 1997; 15(6): 2403–2413. 18. Cunningham D, Chau I, Stockton. Phase III randomised comparison of gemcitabine versus gemcitabine plus capecitabine in patients with advanced pancreatic cancer. In: European Cancer Conference (ECCO 13). Paris, France. European Journal of Cancer Supplements 2005. 19. Sultana A, Smith CT, Cunningham D, et al. Meta-analyses of chemotherapy for locally advanced and metastatic pancreatic cancer. J Clin Oncol 2007; 25(18): 2607–2615. 20. Bramhall SR, et al. Marimastat as first-line therapy for patients with unresectable pancreatic cancer: a randomized trial. J Clin Oncol 2001; 19(15): 3447–3455. 21. Bramhall SR, Schulz J, Nemunaitis J, et al. A double-blind placebo-controlled, randomised study comparing gemcitabine and marimastat with gemcitabine and placebo as first line therapy in patients with advanced pancreatic cancer. Br J Cancer 2002; 87(2): 161–167. 22. Moore MJ, et al. Comparison of gemcitabine versus the matrix metalloproteinase inhibitor BAY 12–9566 in patients with advanced or metastatic adenocarcinoma of the pancreas: a phase III trial of the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 2003; 21(17): 3296–3302. 23. Kindler HL, et al. Phase II trial of bevacizumab plus gemcitabine in patients with advanced pancreatic cancer. J Clin Oncol 2005; 23(31): 8033–8040. 24. Kindler H, et al. A double-blind, placebo-controlled, randomized phase III trial of gemcitabine (G) plus bevacizumab (B) versus gemcitabine plus placebo (P) in patients (pts) with advanced pancreatic cancer (PC): A preliminary analysis of Cancer and Leukemia Group B 80303. J Clin Oncol 2007; 25(18S): a4508. 25. Xiong HQ, et al. Cetuximab, a monoclonal antibody targeting the epidermal growth factor receptor, in combination with gemcitabine for advanced pancreatic cancer: a multicenter phase II Trial. J Clin Oncol 2004; 22(13): 2610–2616. 26. Philip P, et al. Phase III study of gemcitabine [G] plus cetuximab [C] versus gemcitabine in patients [pts] with locally advanced or metastatic pancreatic adenocarcinoma [PC]: SWOG S0205 study. J Clin Oncol 2007; 25(18S): LBA4509.

173

174

Pancreatic Cancer

27. Moore MJ, et al. Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer: a phase III trial of the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 2007; 25(15): 1960–1966. 28. Wong GY, et al. Effect of neurolytic celiac plexus block on pain relief, quality of life, and survival in patients with unresectable pancreatic cancer: a randomized controlled trial. J Am Med Assoc. 2004; 291(9): 1092–1099. 29. Brentnall TA. Management strategies for patients with hereditary pancreatic cancer. Curr Treat Options Oncol 2005; 6(5): 437–445.

11 Rare pancreatic neoplasms and mimics of pancreatic cancer Naoki Takahashi

Introduction Certain histologic types of pancreatic neoplasms are encountered uncommonly. Although these tumors can have characteristic imaging features, they often are non-specific in appearance. In addition, non-neoplastic abnormalities that affect the pancreas can simulate the appearance of a pancreatic neoplasm. This chapter reviews the clinical and imaging features of rare pancreatic neoplasms and the non-neoplastic processes that can mimic pancreatic cancer.

Rare carcinomas Ductal adenocarcinoma of the pancreas is the most common type of cancer that originates from the pancreas. Other less common pancreatic cancers include acinar cell carcinoma, anaplastic carcinoma, osteoclast-like giant cell tumor, mucinous non-cystic carcinoma, adenosquamous carcinoma, signet-ring cell carcinoma and small cell carcinoma [1]. Acinar cell carcinoma is a rare pancreatic tumor comprising 1% of all exocrine neoplasms, despite the fact that acinar cells make up most of the pancreatic parenchyma [2]. Acinar cell carcinomas are less aggressive and have a better prognosis than the typical ductal carcinoma of the pancreas. This tumor may produce pancreatic enzymes and occasionally may cause polyarthritis and subcutaneous fat necrosis [2]. On computed tomography, acinar cell carcinoma typically is a well-circumscribed, exophytic, lowattenuation mass that may be partially cystic or contain calcifications (Figure 11.1) [3]. Anaplastic carcinoma comprises 2–7% of all pancreatic carcinomas. On CT, the tumor typically is large, partly necrotic, locally invasive, and invariably metastatic at the time of diagnosis [4]. Pancreatic Cancer, ed. Jay Heiken. Published by Cambridge University Press. © Cambridge University Press 2009.

176

Pancreatic Cancer

Figure 11.1. Acinar cell carcinoma. Contrast-enhanced CT shows a large exophytic mass involving the body and tail of the pancreas with a large non-enhancing area in the center of the mass. The mass encases splenic artery and vein.

Figure 11.2. Osteoclast-like giant cell tumor. Contrast-enhanced CT shows a large well-circumscribed multilocular cystic mass arising from the body and tail of the pancreas. There are solid components within the cystic mass. (Case courtesy of Dr. Jay P. Heiken.)

Rare pancreatic neoplasms and mimics of pancreatic cancer

Small cell carcinoma comprises 1% of all pancreatic carcinomas. Histologically, it is indistinguishable from small cell carcinoma of the lung. On CT, the tumor typically appears as a large, confluent, low-attenuation mass [4]. Biliary ductal obstruction is uncommon even though the tumor frequently is located in the head of the pancreas. Lymphadenopathy and liver metastases are common at the time of presentation. Radiographically, it resembles malignant lymphoma. Osteoclast-like giant cell tumor is extremely rare, comprising less than 1% of all pancreatic carcinomas. It typically appears as a large, well-defined, exophytic multilocular cystic mass containing intratumoral hemorrhage and fluid–fluid levels (Figure 11.2) [4, 5].

Solid pseudopapillary tumor (also called solid pseudopapillary epithelial neoplasm) Solid pseudopapillary tumor of the pancreas is a rare tumor with low malignant potential that occurs mainly in young women in the second to fourth decades of life. Synonyms for this tumor include solid and cystic tumor and solid and papillary epithelial neoplasm. Histologically, the tumor is usually a large, wellcircumscribed mass composed of a mixture of cystic and solid components. The cystic space often is filled with blood products. On CT, solid pseudopapillary tumor appears as a large, well-circumscribed mass with varying solid and cystic components caused by hemorrhagic degeneration (Figure 11.3). Solid components located along the wall of the mass show enhancement. The wall may calcify. On MR, the mass is well-defined and heterogeneous with a mixture of high and low signal intensity on T1- and T2-weighted images. Areas of high-signal intensity on T1-weighted images and low or heterogeneous signal intensity on T2-weighted images reflect blood products. Fluid–debris levels occasionally can be seen. A peripheral rim of low signal often is seen on T1- and T2-weighted images, representing a fibrous capsule. After administration of gadolinium, the mass shows enhancement at the periphery during the early phase, and progressively but incompletely fills in during the late phase. Atypical presentations of solid pseudopapillary tumor include presence of liver metastasis or extracapsular invasion, and completely solid appearance of the mass [6–8].

Lymphoma Pancreatic lymphoma may present as a large well-circumscribed, homogeneously hypo-attenuating mass or diffusely infiltrative mass on contrast-enhanced CT

177

178

Pancreatic Cancer

Figure 11.3. Solid pseudopapillary tumor. Contrast-enhanced CT shows a well-circumscribed cystic mass in the tail of the pancreas with heterogeneous solid component seen within the cyst.

Figure 11.4. Lymphoma. Contrast-enhanced CT shows a large, well-circumscribed, homogeneously low-attenuation mass involving the body and tail of the pancreas. A vessel running within the mass (arrowhead) is typical for lymphoma. Left para-aortic adenopathy also is present.

Rare pancreatic neoplasms and mimics of pancreatic cancer

(Figure 11.4) [9–11]. Calcification and necrosis are rare. The presence of lymphadenopathy, particularly below the level of renal vein and/or in the small bowel mesentery, is a feature that may help differentiate pancreatic lymphoma from ductal carcinoma. The lack of mass effect on the pancreatic duct or peripancreatic vessels despite the large size of the mass, is suggestive of the diagnosis [11].

Metastasis The most common primary tumors responsible for pancreatic metastases are renal cell carcinoma and lung carcinoma [12]. Other common primary tumors are melanoma, breast, ovarian and colon carcinomas [13]. In most cases, patients present with multifocal disease elsewhere in the body, but the pancreas may be the only site of metastatic involvement [13]. On imaging, pancreatic metastasis may appear as a solitary mass, multifocal masses or a diffusely enlarged pancreas (Figures 11.5a and 11.5b) [13]. Enhancement characteristics of pancreatic metastases resemble those of the primary tumors. Pancreatic metastases from renal cell carcinoma are often well defined and show greater enhancement than the normal pancreas [12, 14, 15]. Metastatic foci often are less conspicuous on delayed-phase imaging [14]. It is not uncommon for isolated metastases from renal cell carcinoma to develop in the pancreas more than 5 years after nephrectomy [15]. The imaging findings of pancreatic metastases may be indistinguishable from other pancreatic

(a)

(b)

Figure 11.5. Metastases from renal cell carcinoma. Contrast-enhanced CT obtained during the pancreatic phase (a) shows a well-circumscribed, markedly hypervascular mass in the head of the pancreas. Two additional small hypervascular lesions are seen in the tail of the pancreas (arrows). The lesions become less conspicuous during the portal phase (b). Note left nephrectomy change.

179

180

Pancreatic Cancer

neoplasms (adenocarcinoma in the case of hypo-attenuating metastasis or islet cell tumor in the case of hyper-attenuating metastasis), but a clinical history often allows the correct diagnosis.

Mesenchymal tumors Sarcoma

Primary sarcomas of the pancreas are extremely rare. Leiomyosarcoma, malignant fibrous histiocytoma, and Kaposi sarcoma have been reported. Leiomyosarcoma is the most common type, postulated to arise either from blood vessels within the pancreas or from the pancreatic duct [16]. It often appears as a homogeneous, large, solid, enhancing mass which may contain areas of necrosis on CT [16, 17].

Lymphangioma

Lymphangioma is a rare cystic lesion of the pancreas, seen predominantly in young female patients [18]. Lymphangioma is a well-circumscribed, multilocular cystic mass that occurs predominantly in the tail of the pancreas. The cyst is lined by endothelial cells and contains serous or chylous fluid [18]. On imaging, lymphangioma appears as a multilocular cystic mass with thin septations (Figure 11.6) [19–21]. Calcifications in the wall or septations may be present [22, 23]. Differentiation from other multiloculated cystic neoplasms, particularly mucinous cystic neoplasm, is difficult.

Lymphoepithelial cyst

Lymphoepithelial cyst is a rare cystic lesion of the pancreas seen predominantly in male patients. Histologically, the cyst can be unilocular or multilocular. The cyst is lined by squamous epithelium containing keratinizing material [24]. On CT, lymphoepithelial cyst appears as a well-circumscribed, non-enhancing cystic mass with or without fine septations (Figure 11.7.) [24, 25]. Calcification may be present. On MR, the cysts are hypointense on T1-weighted imaging and hyperintense on T2-weighted imaging [25, 26]. These imaging features are indistinguishable from pseudocysts or other common cystic neoplasms of the pancreas. Imaging features that may allow differentiation from other cystic lesions include presence of

Rare pancreatic neoplasms and mimics of pancreatic cancer

Figure 11.6. Lymphangioma. Contrast-enhanced CT shows a large well-circumscribed cystic mass in the anterior pararenal space. Note irregularity of the posterior aspect of the pancreatic head (arrow). Coarse calcification is present in the mass.

Figure 11.7. Lymphoepithelial cyst. Contrast-enhanced CT shows a well-circumscribed cystic lesion with a few thin septations in the head of the pancreas. This appearance is indistinguishable from mucinous cystic neoplasm.

181

182

Pancreatic Cancer

Figure 11.8. Lipoma. Contrast-enhanced computed tomography shows a well-circumscribed, lobulated, fat-attenuation lesion in the neck of the pancreas.

non-enhancing mural nodule [25, 27], which may have lipid attenuation or signal intensity on CT or MR, respectively [27], or granular hypointensity on T2-weighted MR images [28]. These features probably reflect keratinizing material within the cyst. Ultrasound appearance of the lymphoepithelial cyst varies from that of a cystic to a solid mass [24, 29]. Lipoma

Lipoma of the pancreas usually is found incidentally on CT. It appears as a well-circumscribed mass composed almost entirely of fat, with a few scattered vessels or septations (Figure 11.8.) [30]. Other mesenchymal tumors

Various other mesenchymal tumors of the pancreas have been reported. Schwannoma of the pancreas may have varying degrees of hypervascular solid areas and poorly enhancing low-attenuation areas on CT [31]. These CT features correspond to the dense cellular component (Antoni type A) and loose hypocellular component (Antoni type B). Teratoma of the pancreas is a very rare tumor

Rare pancreatic neoplasms and mimics of pancreatic cancer

Figure 11.9. Pancreatic cysts associated with autosomal dominant polycystic kidney disease. Contrast-enhanced computed tomography shows numerous cysts replacing almost the entire pancreatic parenchyma. Solid calcification is seen in the tail of the pancreas. The patient had bilateral nephrectomies and orthotopic renal transplant. Two cysts are also seen in the liver.

containing varying degrees of cystic and solid components. Computed tomography appearances depend on tissue component but may have fat, fat–fluid level or calcifications [32]. Angiomyolipoma of the pancreas is a rare tumor composed of a mixture of spindle cells, adipose tissue, epithelioid cells and blood vessels. It usually occurs in the kidney but can involve the liver and, rarely, other sites including the pancreas [33], and commonly is associated with tuberous sclerosis or lymphangiomyomatosis. Hepatoid carcinoma and hemangioma also have been reported. True cysts

Simple epithelium-lined cysts (true cysts) of the pancreas are rare, but may occur in association with genetic disorders such as von Hippel–Lindau disease, cystic fibrosis or polycystic kidney disease (Figure 11.9.) [34–37]. Pancreatic cysts in these disorders usually are multiple. Pancreatic involvement in von Hippel–Lindau disease is frequent, and cysts are seen in 56–70% of cases [35, 36]. Less common pancreatic manifestations of von Hippel–Lindau disease include islet cell tumor and serous cystadenoma. Pancreatic cysts in polycystic kidney disease are less frequent, occurring in 6% of patients with autosomal dominant polycystic kidney disease [34].

183

184

Pancreatic Cancer

Non-neoplastic process Chronic pancreatitis

Mass-forming chronic pancreatitis is a diagnostic challenge, often misdiagnosed as pancreatic adenocarcinoma clinically and radiographically. In approximately 5% of resections for suspected pancreatic cancer, a histologic diagnosis of chronic pancreatitis is made postoperatively [38, 39]. Alcohol-induced chronic pancreatitis and autoimmune pancreatitis (AIP) each account for approximately 30% of cases of mass-forming chronic pancreatitis [40]. On CT, mass-forming chronic pancreatitis is often iso-attenuating during both the pancreatic and hepatic parenchymal phases of enhancement. In some cases, the focal mass appears hypo-attenuating during the pancreatic phase but becomes iso-attenuating during the hepatic phase [41]. Pancreatic carcinomas, on the other hand, are typically hypo-attenuating during both phases of contrast enhancement. However, about 10% of pancreatic carcinomas are iso-attenuating during both the pancreatic and hepatic parenchymal phases of enhancement [42]. On MR, the enhancement characteristics of mass-forming pancreatitis are similar to those found on CT. On MR cholangiopancreatography (MRCP), the pancreatic duct within mass-forming pancreatitis may be visible but narrowed, whereas the duct within pancreatic carcinoma often is occluded. This sign is referred to as the duct penetrating sign, and in one study had an accuracy of 94% in differentiating the two entities [43]. Evidence of chronic pancreatitis elsewhere in or around the pancreas, including pancreatic calcifications or pseudocyst formation, is a helpful finding, but pancreatitis also can be caused by an obstructive tumor. Autoimmune pancreatitis

Autoimmune pancreatitis is an autoimmune systemic disease that involves the pancreas and several other organ systems including the bile ducts, kidneys, retroperitoneum and salivary glands. Clinical presentation is similar to that of pancreatic carcinoma. Patients with AIP are most commonly male, 50 years of age or older and present with jaundice, abdominal pain and weight loss. Elevation of serum IgG4 is the best serological marker, with a reported sensitivity of 73–75% and specificity of 93–95% [44, 45]. However, elevation of IgG4 is seen in approximately 10% of patients with pancreatic cancer [44]. Diffuse pancreatic enlargement is a typical imaging feature of AIP, seen in 40–60% of patients (Figure 11.10) [46–48]. A capsule-like rim can be seen as a halo of soft tissue attenuation around the enlarged

Rare pancreatic neoplasms and mimics of pancreatic cancer

Figure 11.10. Autoimmune pancreatitis with low-attenuation rim. Contrast-enhanced computed tomography shows diffuse enlargement of the pancreas. A low-attenuation rim of soft tissue is seen surrounding the tail of the pancreas, a sign characteristic of autoimmune pancreatitis.

pancreas in about 30% of patients [46–48]. On CT, the pancreas shows delayed enhancement [48, 49]. On MR, the pancreas appears diffusely hypointense on T1-weighted images, slightly hyperintense on T2-weighted images, and exhibits heterogeneously diminished enhancement during the early phase with delayed enhancement during the late phase [49]. The pancreatic border becomes featureless, with effacement of the lobular contour of the pancreas. Focal mass-like or segmental enlargement of the pancreas is seen in 30–40% of patients with AIP (Figures 11.11a and 11.11b) [46–48, 50]. On CT, the enlarged segments of the pancreas can appear iso-attenuating to the rest of the pancreas or demonstrate low-attenuation compared to the uninvolved pancreatic parenchyma, and may be indistinguishable from pancreatic cancer [50]. Rarely, a patient may present with an atrophic pancreas. Extrapancreatic involvement is relatively common, particularly in the bile ducts, kidneys, retroperitoneum and salivary glands, and often is a clue to the correct diagnosis when the pancreatic findings are non-specific [48, 51]. Groove pancreatitis

Groove pancreatitis is an uncommon type of focal chronic pancreatitis that affects the groove between the head of the pancreas and the duodenum. A sheet-like mass is seen between the pancreatic head and the duodenum, which is low-attenuation on CT, with delayed and progressive heterogeneous enhancement that reflects the

185

186

(a)

Pancreatic Cancer

(b)

Figure 11.11. Focal form of autoimmune pancreatitis with biliary involvement. Contrast-enhanced three-dimensional spoiled gradient-recalled (3-D-SPGR) MR image (a) shows a well-circumscribed, low signal intensity mass in the tail of the pancreas (arrow). The rest of the pancreas is normal in size. Note enhancement and thickening of the bile duct (arrowhead) (b). Coronal single-shot fast spain echo (FSE) image shows diffuse thickening of the extrahepatic bile duct (arrow). Note mild intrahepatic bile duct dilation.

(a)

(b)

Figure 11.12. Groove pancreatitis. Contrast-enhanced computed tomography obtained during the portal phase (a) shows heterogeneously enhancing soft tissue (arrows) between the head of the pancreas and the duodenum (d). The soft tissue between the head of pancreas and duodenum shows delayed enhancement (b). Cyst also is seen within the soft tissue (arrowhead).

fibrous nature of the tissue (Figures 11.12a and 11.12b) [52, 53]. Cysts commonly are seen in the duodenal wall. On MR, the mass appears hypointense to pancreatic parenchyma on T1-weighted images and iso- to slightly hyperintense on T2weighted images. After administration of gadolinium contrast, the mass shows delayed enhancement, as on CT. Duodenal wall thickening with duodenal stenosis is common. Distal bile duct stricture is common, but the bile duct typically shows smooth tapering rather than an abrupt cut-off.

Rare pancreatic neoplasms and mimics of pancreatic cancer

Intrapancreatic accessory spleen

An accessory spleen, often located in the splenic hilum, occurs in approximately 10% of the population. The tail of the pancreas is the second most common site, accounting for 17% of accessory spleens in one study [54]. On CT and MR, an intrapancreatic accessory spleen appears as a small well-circumscribed mass with enhancement and signal characteristics similar to those of the spleen in all phases and sequences of the study (Figures 11.13a and 11.13b). On contrast-enhanced CT, it appears as a hypervascular mass during the arterial phase, and a heterogeneous pattern of enhancement, similar to that of the splenic parenchyma, may be seen [55, 56]. On MR, the mass appears as hypointense and hyperintense compared with the pancreas, but iso-intense to the spleen, on unenhanced T1- and T2-weighted images, respectively [56]. Differential diagnosis of intrapancreatic accessory spleen includes hypervascular pancreatic tumors, such as islet cell tumors, and metastatic

(a)

(b)

Figure 11.13. Intrapancreatic accessory spleen. Contrast-enhanced three-dimensional spoiled gradient-recalled (3D-SPGR) MR image (a) shows a well-circumscribed, lobular, hypervascular mass in the tail of the pancreas (arrow). On T2 fast spin echo (FSE) MR image (b), the lesion is hyperintense compared with the pancreatic parenchyma, but iso-intense compared with the spleen (arrow). On T2 FSE MR image obtained after administration of superparamagnetic iron oxide (SPIO) (c), the signal of the lesion drops, and is similar to the spleen (arrow). (c)

187

188

Pancreatic Cancer

renal cell carcinomas. When an intrapancreatic accessory spleen is suspected, a sulfur-colloid nuclear medicine study or a superparamagnetic iron oxide (SPIO)enhanced T2-weighted MR study can be used to confirm the diagnosis. On SPIO-enhanced T2-weighted images, the signal of the mass decreases and is similar to that of the spleen compared with the pre-SPIO-enhanced T2-weighted images (Figure 11.13c) [56]. A cleft between the mass and the pancreas may be present in some cases, which is suggestive of the extrapancreatic origin of the mass [55]. Focal fatty replacement

Focal fatty replacement of the pancreas occasionally may mimic a pancreatic tumor appearing as a focal area of low attenuation on CT [57, 58]. It is located most commonly in the anterior portion of the head of the pancreas. Geographic or triangular shape of the low-attenuation area and absence of mass effect or ductal obstruction are clues to the correct diagnosis. However, in some cases it can be rounded in shape and difficult to differentiate from pancreatic carcinoma. The area of focal fatty replacement often has CT attenuation below −30 HU on non-contrast CT, but non-contrast CT often is not available. On MR, the presence

(a)

(b)

Figure 11.14. Focal fatty replacement. Contrastenhanced three-dimensional spoiled gradient-recalled (3D-SPGR) MR image (a) shows a well-circumscribed, low-signal intensity lesion in the head of the pancreas (arrow). Precontrast in- and opposed-phase gradient-recalled-echo (GRE) MR images (b, c) show presence of focal area of fat in the anterior aspect of the head of the pancreas (arrows).

(c)

Rare pancreatic neoplasms and mimics of pancreatic cancer

of fat can be confirmed by the reduction in signal intensity on the opposed phase of chemical shift MR image (Figures 11.14a–c) [58].

REFERENCES 1. Kloppel G, Solcia E, Longnecker DS, Capella C, Sobin LH. Histological Typing of Tumours of the Exocrine Pancreas. 2nd edn. Berlin, Springer-Verlag, 1998. 2. Klimstra DS, Heffess CS, Oertel JE, Rosai J. Acinar cell carcinoma of the pancreas. A clinicopathologic study of 28 cases. Am J Surg Pathol 1992; 16(9): 815–837. 3. Tatli S, Mortele KJ, Levy AD, et al. CT and MRI features of pure acinar cell carcinoma of the pancreas in adults. Am J Roentgenol 2005; 184(2): 511–519. 4. Ichikawa T, Federle MP, Ohba S, et al. Atypical exocrine and endocrine pancreatic tumors (anaplastic, small cell, and giant cell types): CT and pathologic features in 14 patients. Abdom Imaging 2000; 25(4): 409–419. 5. Molberg KH, Heffess C, Delgado R, Albores-Saavedra J. Undifferentiated carcinoma with osteoclast-like giant cells of the pancreas and periampullary region. Cancer 1998; 82(7): 1279–1287. 6. Ohtomo K, Furui S, Onoue M, et al. Solid and papillary epithelial neoplasm of the pancreas: MR imaging and pathologic correlation. Radiology 1992; 184(2): 567–570. 7. Buetow PC, Buck JL, Pantongrag-Brown L, et al. Solid and papillary epithelial neoplasm of the pancreas: imaging-pathologic correlation on 56 cases. Radiology 1996; 199(3): 707–711. 8. Cantisani V, Mortele KJ, Levy A, et al. MR imaging features of solid pseudopapillary tumor of the pancreas in adult and pediatric patients. Am J Roentgenol 2003; 181(2): 395–401. 9. Prayer L, Schurawitzki H, Mallek R, Mostbeck G. CT in pancreatic involvement of non-Hodgkin lymphoma. Acta Radiol 1992; 33(2): 123–127. 10. Van Beers B, Lalonde L, Soyer P, et al. Dynamic CT in pancreatic lymphoma. J Comput Assist Tomogr 1993; 17(1): 94–97. 11. Merkle EM, Bender GN, Brambs HJ. Imaging findings in pancreatic lymphoma: differential aspects. Am J Roentgenol 2000; 174(3): 671–675. 12. Klein KA, Stephens DH, Welch TJ. CT characteristics of metastatic disease of the pancreas. Radiographics 1998; 18(2): 369–378. 13. Ferrozzi F, Bova D, Campodonico F, et al. Pancreatic metastases: CT assessment. Eur Radiol 1997; 7(2): 241–245. 14. Ng CS, Loyer EM, Iyer RB, et al. Metastases to the pancreas from renal cell carcinoma: findings on three-phase contrast-enhanced helical CT. Am J Roentgenol 1999; 172(6): 1555–1559. 15. Ghavamian R, Klein KA, Stephens DH, et al. Renal cell carcinoma metastatic to the pancreas: clinical and radiological features. Mayo Clin Proc 2000; 75(6): 581–585. 16. Paciorek ML, Ross GJ. MR imaging of primary pancreatic leiomyosarcoma. Br J Radiol 1998; 71(845): 561–563.

189

190

Pancreatic Cancer

17. Srivastava DN, Batra A, Thulkar S, Julka PK. Leiomyosarcoma of pancreas: imaging features. Indian J Gastroenterol 2000; 19(4): 187–188. 18. Paal E, Thompson LD, Heffess CS. A clinicopathologic and immunohistochemical study of ten pancreatic lymphangiomas and a review of the literature. Cancer 1998; 82(11): 2150–2158. 19. Pandolfo I, Scribano E, Gaeta M, Fiumara F, Longo M. Cystic lymphangioma of the pancreas: CT demonstration. J Comput Assist Tomogr 1985; 9(1): 209–210. 20. Gray G, Fried K, Iraci J. Cystic lymphangioma of the pancreas: CT and pathologic findings. Abdom Imaging 1998; 23(1): 78–80. 21. Schneider G, Seidel R, Altmeyer K, et al. Lymphangioma of the pancreas and the duodenal wall: MR imaging findings. Eur Radiol 2001; 11(11): 2232–2235. 22. Hanelin LG, Schimmel DH. Lymphangioma of the pancreas exhibiting an unusual pattern of calcification. Radiology 1977; 122(3): 636. 23. Ferrozzi F, Zuccoli G, Bova D, Calculli L. Mesenchymal tumors of the pancreas: CT findings. J Comput Assist Tomogr 2000; 24(4): 622–627. 24. Truong LD, Stewart MG, Hao H, Yutani C, Jordan PH. A comprehensive characterization of lymphoepithelial cyst associated with the pancreas. Am J Surg 1995; 170(1): 27–32. 25. Koga H, Takayasu K, Mukai K, et al. CT of lymphoepithelial cysts of the pancreas. J Comput Assist Tomogr 1995; 19(2): 221–224. 26. Kazumori H, Sizuku T, Ueki T, Uchida Y, Yamamoto S. Lymphoepithelial cyst of the pancreas. J Gastroenterol 1997; 32(5): 700–703. 27. Fukukura Y, Inoue H, Miyazono N, et al. Lymphoepithelial cysts of the pancreas: demonstration of lipid component using CT and MRI. J Comput Assist Tomogr 1998; 22(2): 311–313. 28. Shinmura R, Gabata T, Matsui O. Lymphoepithelial cyst of the pancreas: case report with special reference to imaging – pathologic correlation. Abdom Imaging 2006; 31(1): 106–109. 29. Ueno S, Muranaka T, Maekawa S, et al. Radiographic features in lymphoepithelial cyst of the pancreas. Abdom Imaging 1994; 19(3): 232–234. 30. Katz DS, Nardi PM, Hines J, et al. Lipomas of the pancreas. Am J Roentgenol 1998; 170(6): 1485–1487. 31. Ferrozzi F, Bova D, Garlaschi G. Pancreatic schwannoma: report of three cases. Clin Radiol 1995; 50(7): 492–495. 32. Jacobs JE, Dinsmore BJ. Mature cystic teratoma of the pancreas: sonographic and CT findings. Am J Roentgenol 1993; 160(3): 523–524. 33. Heywood G, Smyrk TC, Donohue JH. Primary angiomyolipoma of the pancreas. Pancreas 2004; 28(4): 443–445. 34. Torra R, Nicolau C, Badenas C, et al. Ultrasonographic study of pancreatic cysts in autosomal dominant polycystic kidney disease. Clin Nephrol 1997; 47(1): 19–22. 35. Hough DM, Stephens DH, Johnson CD, Binkovitz LA. Pancreatic lesions in von Hippel-Lindau disease: prevalence, clinical significance, and CT findings. Am J Roentgenol 1994; 162(5): 1091–1094.

Rare pancreatic neoplasms and mimics of pancreatic cancer

36. Hammel PR, Vilgrain V, Terris B, et al. Pancreatic involvement in von Hippel-Lindau disease. The Groupe Francophone d’Etude de la Maladie de von Hippel-Lindau. Gastroenterology 2000; 119(4): 1087–1095. 37. Daneman A, Gaskin K, Martin DJ, Cutz E. Pancreatic changes in cystic fibrosis: CT and sonographic appearances. Am J Roentgenol 1983; 141(4): 653–655. 38. Smith CD, Behrns KE, van Heerden JA, Sarr MG. Radical pancreatoduodenectomy for misdiagnosed pancreatic mass. Br J Surg 1994; 81(4): 585–589. 39. van Gulik TM, Reeders JW, Bosma A, et al. Incidence and clinical findings of benign, inflammatory disease in patients resected for presumed pancreatic head cancer. Gastrointest Endosc 1997; 46(5): 417–423. 40. Yadav D, Notahara K, Smyrk TC, et al. Idiopathic tumefactive chronic pancreatitis: clinical profile, histology, and natural history after resection. Clin Gastroenterol Hepatol 2003; 1(2): 129–135. 41. Kim T, Murakami T, Takamura M, et al. Pancreatic mass due to chronic pancreatitis: correlation of CT and MR imaging features with pathologic findings. Am J Roentgenol 2001; 177(2): 367–371. 42. Prokesch RW, Chow LC, Beaulieu CF, Bammer R, Jeffrey RB Jr. Isoattenuating pancreatic adenocarcinoma at multi-detector row CT: secondary signs. Radiology 2002; 224(3): 764–768. 43. Ichikawa T, Sou H, Araki T, et al. Duct-penetrating sign at MRCP: usefulness for differentiating inflammatory pancreatic mass from pancreatic carcinomas. Radiology 2001; 221(1): 107–116. 44. Ghazale A, Chari ST, Smyrk TC, et al. Value of serum IgG4 in the diagnosis of autoimmune pancreatitis and in distinguishing it from pancreatic cancer. Am J Gastroenterol 2007; 102(8): 1646–1653. 45. Choi EK, Kim MH, Lee TY, et al. The sensitivity and specificity of serum immunoglobulin G and immunoglobulin G4 levels in the diagnosis of autoimmune chronic pancreatitis: Korean experience. Pancreas 2007; 35(2): 156–161. 46. Sahani DV, Kalva SP, Farrell J, et al. Autoimmune pancreatitis: imaging features. Radiology 2004; 233(2): 345–352. 47. Yang DH, Kim KW, Kim TK, et al. Autoimmune pancreatitis: radiologic findings in 20 patients. Abdom Imaging 2006; 31(1): 94–102. 48. Takahashi N, Fletcher JG, Fidler JL, et al. Dual-phase CT of autoimmune pancreatitis: multi-reader study. Am J Roentgenol 2008; 190(2): 280–286. 49. Irie H, Honda H, Baba S, et al. Autoimmune pancreatitis: CT and MR characteristics. Am J Roentgenol 1998; 170(5): 1323–1137. 50. Wakabayashi T, Kawaura Y, Satomura Y, et al. Clinical and imaging features of autoimmune pancreatitis with focal pancreatic swelling or mass formation: comparison with so-called tumorforming pancreatitis and pancreatic carcinoma. Am J Gastroenterol 2003; 98(12): 2679–2687. 51. Takahashi N, Kawashima A, Fletcher JG, Chari ST. Renal involvement in patients with autoimmune pancreatitis: CT and MR imaging findings. Radiology 2007; 242(3): 791–801. 52. Gabata T, Kadoya M, Terayama N, et al. Groove pancreatic carcinomas: radiological and pathological findings. Eur Radiol 2003; 13(7): 1679–1684.

191

192

Pancreatic Cancer

53. Itoh S, Yamakawa K, Shimamoto K, Endo T, Ishigaki T. CT findings in groove pancreatitis: correlation with histopathological findings. J Comput Assist Tomogr 1994; 18(6): 911–915. 54. Halpert B, Gyorkey F. Lesions observed in accessory spleens of 311 patients. Am J Clin Pathol 1959; 32(2): 165–168. 55. Miyayama S, Matsui O, Yamamoto T, et al. Intrapancreatic accessory spleen: evaluation by CT arteriography. Abdom Imaging 2003; 28(6): 862–865. 56. Kim SH, Lee JM, Han JK, et al. MDCT and superparamagnetic iron oxide (SPIO)-enhanced MR findings of intrapancreatic accessory spleen in seven patients. Eur Radiol 2006; 16(9): 1887–1897. 57. Matsumoto S, Mori H, Miyake H, et al. Uneven fatty replacement of the pancreas: evaluation with CT. Radiology 1995; 194(2): 453–458. 58. Kim HJ, Byun JH, Park SH, et al. Focal fatty replacement of the pancreas: usefulness of chemical shift MRI. Am J Roentgenol 2007; 188(2): 429–432.

Index

3-D processing techniques 36, 65–66 5 fluorouracil (5FU)-based chemoradiation 168–169 acinar cell carcinoma (ACC) 20–21, 175 adenosquamous carcinomas 13 anaplastic carcinoma 13–14, 175 angiomyolipoma 183 atypical mass, biopsies of 156 autoimmune pancreatitis (AIP) 68, 184–185 biologic agents for advanced pancreatic cancer 170–171 carcinoembryonic antigen (CEA), elevated level in mucinous cysts 141–142 carcinoid tumors 109–110 CEMRA (contrast enhanced magnetic resonance angiography) 53–54 chemoradiotherapy and chemotherapy 167–170 chronic pancreatitis 2, 38–39, 184 clinical presentation 29, 58 autoimmune pancreatitis 184–185 head of pancreas cancer 154–155 hyperfunctioning NENs 105–110 non-hyperfunctioning NENs 110 pancreatic cystic neoplasms 83 colloid carcinoma 11–13, 19 computed tomography (CT) 60 CT-guided FNA 135–136 interpretation detection 66–68 staging 68–74 patient preparation 63 procedure when no mass seen 156 staging technique 60–63 testing typical masses 155–156 see also multi-detector CT (MDCT) techniques contrast agents intravenous (IV) 31–32

gadolinimum-based 77, 92 injection rate 32, 63 mangafodipir trisodium 75 orally administered 30, 47, 63, 74 contrast-enhanced magnetic resonance angiography (CEMRA) 53–54 contrast-enhanced MR sequences 49–50 corticotropinoma 109 cystic lesions, EUS evaluation of 138–142 cystic pancreatic neoplasms 83–84 cystic pancreatic endocrine neoplasms (PENs) 96–98 intraductal papillary mucinous neoplasms (IPMNs) 90–94 management of 99 MDCT imaging 40 miscellaneous and rare lesions 98–99 mucinous cystic neoplasms (MCNs) 86–90 serous cystic neoplasms (SCNs) 85 solid pseudopapillary epithelial neoplasm (SPEN) 94–96 cytogenetics of pancreatic cancer 1, 5 differences from IPMN 4 implications for diagnosis/management 6 diagnosis of pancreatic cancer biopsies resulting from CT findings 155–156 cystic pancreatic lesions 99 cytogenetics 6 fine needle aspiration (FNA) 133–136 goals of preoperative imaging 59 role of endoscopic ultrasound (EUS) 131–133 see also imaging evaluation of PC diffusion weighted imaging (DWI) 54 ductal adenocarcinoma see pancreatic cancer (PC) ectopic hormones, tumors producing 109–110 endocrine tumors 23–24, 96–98 see also neuroendocrine neoplasms (NEN) endoscopic retrograde cholangiopancreatography (ERCP) 156

193

194

Index

endoscopic ultrasound (EUS) 78–79, 84, 130–131 in diagnosis of atypical mass 156 diagnosis of pancreatic adenocarcinoma 131–133 evaluation of cystic lesions 138–142 in the evaluation of PEN 142–144 staging of pancreatic adenocarcinoma 136–137 with fine needle aspiration (EUS-FNA) 133–136 epidemiology of pancreatic cancer 1–2 erlotinib treatment 171 familial PC syndrome 6 fine needle aspiration (FNA), EUS 133–136 focal fatty replacement, mimicking a tumor 188–189 focal pancreatitis 68 EUS-FNA for evaluation of 135 groove pancreatitis 185–186 gastrinomas 107, 113–115, 122 Gastrointestinal Tumor Study Group (GITSG) 168–169 gemcitabine treatment locally advanced cancer 167–169 metastatic cancer 170–171 genetics 1 cytogenetics of PC 5 implications for diagnosis and management 6 mucinous cystic neoplasms (MCNs) 4–5 syndromes with predisposition to PC 5–6 giant cell carcinomas 13–14, 177 glucagonoma 107, 115–116 GRFoma 109 groove pancreatitis 185–186 H-NENs see hyperfunctioning NENs hereditary PC syndromes 5–6 hybrid PET-CT scanners 79–80 hyperfunctioning NENs (H-NENs) radiographic imaging features and results 110–118 tumors producing ectopic hormones 109–110 tumors producing islet- and gut-related hormones 105–108 imaging evaluation of PC 58–59 computed tomography (CT) 60–63 contrast dose, injection rate and scan timing 63–65 detection of pancreatic ductal cancer 59 goals of preoperative imaging 59 magnetic resonance imaging (MRI) 74–78 patient preparation 63, 74 positron emission tomography (PET) 79–80 post processing 65–66 CT interpretation detection 66–68 staging 68–74 staging 78 see also endoscopic ultrasound; multi-detector CT (MDCT) techniques injection rate, contrast agent 32, 63 insulinomas 105–107, 110–113 detection using EUS 143

intraductal papillary mucinous neoplasms (IPMNs) 1, 2–4, 17–19 association with colloid carcinoma 11–13, 19 imaging of 90–94 intrapancreatic accessory spleen 187–188 invasive carcinoma, MCN 20 IPMNs see intraductal papillary mucinous neoplasms (IPMNs) islet- and gut-related hormones, tumors producing 105–108 KRAS oncogene, mutation of 4, 6 leiomyosarcoma 180 lipoma 182 liver metastases 116, 122 locally advanced pancreatic cancer, treatment of 166–170 lymphangioma 180 lymphoepithelial cyst (LEC) 98–99, 180–182 lymphoma 177–178 macroscopic precursors of PC intraductal papillary mucinous neoplasms (IPMNs) 2–4 mucinous cystic neoplasms (MCNs) 4–5 magnetic resonance cholangiopancreatography (MRCP) 50–53 magnetic resonance spectroscopy (MRS) 54–55 mangafodipir trisodium 75 mass-forming chronic pancreatitis 184 MCNs see mucinous cystic neoplasms (MCNs) MDCT see multi-detector CT (MDCT) techniques medullary carcinoma 14–16 mesenchymal tumors 182–183 lipoma 182 lymphangioma 180 lymphoepithelial cyst 180–182 sarcoma 180 true cysts 183 metastatic pancreatic cancer 178–180 H-HEN, liver metastases 122 treatment of 170–171 VIPomas, liver metastases 116 microscopic precursors of PC 2 mimics of pancreatic cancer 175 autoimmune pancreatitis 68, 184–185 focal fatty replacement 188–189 focal pancreatitis 68 mortality rates following surgery 58, 150, 158 MRI techniques 46, 74–75 advantages over CT 75 comparison to CT technology 74, 77–78 contrast-enhanced MR sequences 49–50 detection of pancreatic adenocarcinomas 77 MRCP (Magnetic Resonance Cholangiopancreatography) 50–53 non-contrast MR sequences 47–49 normal pancreas 76–77

Index

patient preparation 47, 74 staging 78 supplementary and emerging techniques 53–55 mucinous adenocarcinomas 163 mucinous cystic neoplasms (MCNs) 1 histologic features 19–20 imaging evaluation 86–90 invasive carcinoma 20 precursor of pancreatic cancer 4–5 mucinous non-cystic carcinoma 11–13 multi-detector CT (MDCT) techniques 28 accuracy of 39–40 acquisition parameters 29 additional techniques 36–39 clinical indications 29 IV contrast and dose 31–32 for NENs 119–120 oral contrast 30 and other forms of pancreatic neoplasms 40 pancreatic protocol components 28–29 phasing and timing of scans 32 dual phase acquisition 33 matching to cardiac output 65 single phase acquisition 33 protocol example 64 radiation dose 29–30 reconstruction parameters 33–34 see also computed tomography (CT) neoadjuvant chemotherapy 167–168 neuroendocrine neoplasms (NENs) 104, 123 H-NENs clinical features of 105–110 radiographic imaging features and results 110–118 MDCT imaging 40 N-NENs clinical features of 110 radiographic imaging features and results 118–119 role of imaging 119–122 treatment 122–123 see also pancreatic endocrine neoplasms (PENs) non-contrast MR sequences 47–49 non-neoplastic conditions autoimmune pancreatitis 68, 184–185 chronic pancreatitis 184 focal fatty replacement 188–189 groove pancreatitis 185–186 intrapancreatic accessory spleen 187–188 non-hyperfunctioning NENs (N-NENs) clinical features of 110 radiographic imaging features and results 118–119 non-syndromic tumors see non-hyperfunctioning NENs

pancreatic cancer (PC) 1 cytogenetics of 5 diagnosis of 131–133 epidemiology 1–2 hereditary syndromes 5–6 macroscopic precursors IPMNs 2–4 MCNs 4–5 microscopic precursors 2 pathology of 10–16 see also rare pancreatic neoplasms pancreatic ductal adenocarcinoma see pancreatic cancer (PC) pancreatic endocrine neoplasms (PENs) 23–24 cystic PEN 96–98 EUS in evaluation of 142–144 see also neuroendocrine neoplasms (NENs) pancreatic intraepithelial neoplasias (PanINs) 2, 16–17 pancreatic neuroendocrine neoplasms see neuroendocrine neoplasms (NENs) pancreatoduodenectomy 150–152 PanINs see pancreatic intraepithelial neoplasias parathyrinoma 109 pathology of pancreatic neoplasms 10 acinar cell carcinoma (ACC) 20–21 ductal adenocarcinoma 10–16 intraductal papillary mucinous neoplasm (IPMN) 17–19 mucinous cystic neoplasms (MCNs) 19–20 pancreatic endocrine neoplasm (PEN) 23–24 precursors to PC, PanIN 16–17 serous cystic neoplasms 24–25 solid pseudopapillary tumor (SPT) 21–22 patient preparation for CT and MRI 47, 63, 74 PC see pancreatic cancer PENs see pancreatic endocrine neoplasms positron emission tomography (PET) 79–80 preventive measures 171 pseudocysts 98, 138

oncogenes 4, 5, 6 osteoclast-like giant cell 14, 177

radiation doses, methods of minimizing 29–30 radical antegrade modular pancreato-splenectomy (RAMPS) 159–160 rare pancreatic neoplasms 175–177 lymphoma 177–178 mesenchymal tumors 180–183 metastasis 178–180 solid pseudopapillary tumor (SPT) 21–22, 177 renal cell carcinoma, pancreatic metastases from 178–180 resection of pancreatic adenocarcinoma accuracy in predicting using CT signs 73 body and tail, rationale for 159–161 determination of resectability 156–158 head of pancreas, rationale for 150–152 predictors of non-resectability 73

pancreatectomy distal (body or tail) 159–161 total 152–154

sarcoma 180 scan timing 32–33, 50, 63–65 schwannoma of the pancreas 182

195

196

Index

SCNs see serous cystic neoplasms screening 171 segmental pancreatic duct dilation 38–39 selective arterial stimulation 110, 113, 121, 123, 143 serous cystadenomas 139–140 serous cystic neoplasms (SCNs) 24–25, 85 signet ring cell carcinoma 13 small cell carcinoma 177 smoking, risk factor for PC 2 solid pseudopapillary epithelial neoplasm (SPEN) 21–22, 94–96, 177 solid pseudopapillary tumor (SPT) 21–22, 94–96, 177 somatostatin receptor scintigraphy (SRS) 112, 113–115, 116, 120–121, 143 somatostatinomas 108, 117 SPEN see solid pseudopapillary epithelial neoplasm SPT see solid pseudopapillary tumor (SPT) staging laparoscopy 157–158, 162 staging of pancreatic adenocarcinoma 68–74 distant staging 78 local staging 78 role of EUS 136–137 see also surgical staging steady-state free precession (SSFP) imaging 54 surgical staging 150 body and tail of pancreas diagnosis and staging 160–162 rationale for resection 159–161 results of resection 162–163 head of pancreas diagnosis clinical presentation 154–155 using imaging techniques 155–156 rationale for surgical resection 150–152 results of resection 158 surgical staging of tumor of 156–158 survival rates 1 acinar cell carcinoma 21 colloid carcinoma 13 glucagonoma 107 locally advanced cancer 166, 168–170

main pancreatic duct (MPD) IPMN following surgery 93 metastatic cancer 122, 170–171 mucinous cystic neoplasms (MCNs) 20, 90 non-hyperfunctioning NENs 110 pancreatic body and tail resection 162–163 pancreatic head resection 158 PEN following surgery 98 somatostatinomas 108 SPEN following surgery 96 symptoms 2–4, 29, 83 management of 171 syndromic tumors see hyperfunctioning NENs teratoma of the pancreas 182–183 timing of scans 32–33, 50, 63–65 total pancreatectomy 152–154 transabdominal ultrasound 84, 112, 113 treatment 166 cystic pancreatic neoplasms 99 implications of cytogenetics 6 locally advanced pancreatic adenocarcinoma 166–170 metastatic pancreatic adenocarcinoma 170–171 neuroendocrine neoplasms (NENs) 122–123 symptomatic management 171 true cysts 183 tumor suppressor genes 5 undifferentiated (anaplastic or sarcomatoid) carcinoma 13–14 undifferentiated carcinoma with osteoclast-like giant cells 14 vascular invasion 36–38 criteria predicting extent of 69–73 EUS criteria for 136 VIPomas 107–108, 116 von Hippel–Lindau disease 27, 96, 183 Whipple procedure 150–152 Zollinger–Elllison syndrome 107, 108, 109, 113