Frozen Section Library Series Editor Philip T. Cagle, MD Houston, Texas, USA
For further volumes: http://www.springer.com/series/7869
Frozen Section Library: Pancreas
Wendy L. Frankel, MD The Ohio State University Medical Center, Columbus, OH, USA
Daniela M. Proca, MD The Ohio State University Medical Center, Columbus, OH, USA
Including Chapter 2 co-authored by:
Mark Bloomston, MD The Ohio State University Medical Center Columbus, OH, USA
13
Wendy L. Frankel, MD The Ohio State University Medical Center Columbus, OH 43210, USA
[email protected] Daniela M. Proca, MD The Ohio State University Medical Center Columbus, OH 43210, USA
[email protected] ISSN 1868-4157 e-ISSN 1868-4165 ISBN 978-1-4419-7789-2 e-ISBN 978-1-4419-7790-8 DOI 10.1007/978-1-4419-7790-8 Springer New York Dordrecht Heidelberg London © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
To our fathers, Maurice A. Frankel, MD, and Valentin Cojocea, MD, who are both surgeons and serve as our role models Wendy L. Frankel, MD Daniela Proca, MD
Series Preface
For over 100 years, the frozen section has been utilized as a tool for the rapid diagnosis of specimens while a patient is undergoing surgery, usually under general anesthesia, as a basis for making immediate treatment decisions. Frozen section diagnosis is often a challenge for the pathologist who must render a diagnosis that has crucial import for the patient in a minimal amount of time. In addition to the need for rapid recall of differential diagnoses, there are many pitfalls and artifacts that add to the risk of frozen section diagnosis that are not present with permanent sections of fully processed tissues that can be examined in a more leisurely fashion. Despite the century-long utilization of frozen sections, most standard pathology textbooks, both general and subspecialty, largely ignore the topic of frozen sections. Few textbooks have ever focused exclusively on frozen section diagnosis, and those textbooks that have done so are now out-of-date and have limited illustrations. The Frozen Section Library series is meant to provide convenient, user-friendly handbooks for each organ system to expedite use in the rushed frozen section situation. These books are small and lightweight, copiously color illustrated with images of actual frozen sections, highlighting pitfalls, artifacts, and differential diagnosis. The advantages of a series of organ-specific handbooks, in addition to the ease of use and manageable size, are that (1) a series allows more comprehensive coverage of more diagnoses, both common and rare, than a single volume that tries to highlight a limited number of diagnoses for each organ and (2) a series allows more detailed insight by permitting experienced authorities to emphasize the peculiarities of frozen section for each organ system. As a handbook for practicing pathologists, these books will be indispensable aids to diagnosis and avoiding dangers in one of the most challenging situations that pathologists encounter. Rapid
vii
viii
SERIES PREFACE
consideration of differential diagnoses and how to avoid traps caused by frozen section artifacts are emphasized in these handbooks. A series of concise, easy-to-use, well-illustrated handbooks alleviates the often frustrating and time-consuming, sometimes futile, process of searching through bulky textbooks that are unlikely to illustrate or discuss pathologic diagnoses from the perspective of frozen sections in the first place. Tables and charts will provide guidance for differential diagnosis of various histologic patterns. Touch preparations, which are used for some organs such as central nervous system or thyroid more often than others, are appropriately emphasized and illustrated according to the need for each specific organ. This series is meant to benefit practicing surgical pathologists, both community and academic, and pathology residents and fellows and also to provide valuable perspectives to surgeons, surgery residents, and fellows who must rely on frozen section diagnosis by their pathologists. Most of all, we hope that this series contributes to the improved care of patients who rely on the frozen section to help guide their treatment. Houston, TX
Philip T. Cagle Series Editor
Preface
Intra-operative pathology consultation is one of the most difficult areas in diagnostic surgical pathology due to time constraints, tissue and freezing artifacts, and the lack of ancillary studies such as immunohistochemistry. Frozen sections of the pancreas tend to be one of the more challenging areas due to the relatively small number of these cases at most institutions. Additionally, adenocarcinoma and chronic pancreatitis coexist in many cases and can be difficult to distinguish. This monograph, Frozen Section Library: Pancreas, is a volume in the Frozen Section Library Series. The contents are a combination of personal experience and a review of the available literature. Frozen Section Library: Pancreas will provide a concise pictorial compendium to facilitate intra-operative consultations on pancreas specimens. This user-friendly handbook is divided into chapters that emphasize the common questions a pathologist must answer during frozen section examination and will provide guidance for the differential diagnosis of various histologic patterns. A summary of Key Points as well as Pitfalls will be discussed at the end of the chapters. Some of the pitfalls can be avoided by being aware they exist. When possible, ways to avoid pitfalls will be mentioned. The purpose of this book is to aid in the timely frozen section diagnosis of pancreatic lesions by using a broad array of illustrations, which would reinforce one s visual memory, and a condensed text, useful for rapid review of main diagnostic features. Currently, there is no other up-to-date, single-source reference specifically focused on frozen sections of the pancreas. This book, it is hoped, will be a useful tool to practicing surgical pathologists, as well as to pathology residents and fellows. We hope it will facilitate the accurate diagnosis of pancreatic lesions, by not only describing patterns of disease but also helping to understand
ix
x
PREFACE
the expectations of the surgeon and how the diagnosis will help guide intra-operative decisions. Columbus, OH Columbus, OH
Wendy L. Frankel Daniela M. Proca
Acknowledgments
The authors wish to thank Shawn Scully and Brian Rubin for their technical expertise in preparing the figures and Jacqui Lankford for assistance in preparing/editing the text.
xi
Contents
Series Preface . . . . . . . . . . . . . . . . . . . . . . . . . .
vii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ix
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . .
xi
1.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . .
1
2.
Clinical Indications and Surgeons’ Expectations . .
3
3.
The Distinction Between Chronic Pancreatitis and Adenocarcinoma . . . . . . . . . . . . . . . . . . . Criteria of Malignancy in Frozen Section . . . . . . . . Other Tumor-like Lesions . . . . . . . . . . . . . . . . . .
9 12 23
Variants of Pancreatic Adenocarcinoma and Pancreatic Intraepithelial Neoplasia . . . . . . . Variants of Adenocarcinoma . . . . . . . . . . . . . . . . Pancreatic Intraepithelial Neoplasia (PanIN) . . . . . .
29 29 35
5.
Cystic Lesions in the Pancreas . . . . . . . . . . . . . Non-Neoplastic Cysts . . . . . . . . . . . . . . . . . . . . Neoplastic Cysts . . . . . . . . . . . . . . . . . . . . . . .
43 45 46
6.
Other Pancreatic Tumors . . . . . . . . . . . . . . . . .
55
7.
Secondary Tumors and Miscellaneous Lesions Involving the Pancreas . . . . . . . . . . . . .
63
Metastases, Resectability, and Margins . . . . . . . . Metastasis and Resectability . . . . . . . . . . . . . . . . Margin Assessment . . . . . . . . . . . . . . . . . . . . .
73 73 80
4.
8.
xiii
xiv
CONTENTS
Suggested Reading . . . . . . . . . . . . . . . . . . . . . . .
89
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
93
Chapter 1
Introduction
Frozen sections of the pancreas are performed for tumor confirmation and to assess resectability and margin status. Accurate frozen section diagnoses are important to help guide the surgical procedure. Diagnostic difficulties in pancreatic frozen sections are related to technical and morphological factors, with the main problematic differential diagnosis being between chronic pancreatitis and duct adenocarcinoma. Most of the histologic features useful in frozen sections are similar to those useful in small biopsies, but the diagnosis is more challenging due to frozen artifacts and time constraints. A diagnosis that may be straightforward in large and well-fixed sections may be challenging in small samples with artifact. The majority of specimens sent for frozen section analysis consist of small portions of tissue, rather than whole resection specimens. Therefore, the focus of this monograph will be on the histologic features most useful in these specimens. There are instances when the surgeon takes a patient to the operating room based upon the clinicoradiographic diagnosis of pancreatic carcinoma without tissue confirmation. The use of endoscopic ultrasound-guided fine needle aspiration analysis has increased the number of cases that have definitive pre-operative diagnoses, but cases still remain unconfirmed prior to surgery. Since chronic pancreatitis frequently surrounds malignant lesions in the pancreas, a diagnosis of chronic pancreatitis should not preclude a pancreatic resection for a clinically suspicious mass in a patient with resectable disease. It is acceptable to surgeons to have up to 10% of pancreatic resections for suspected malignancy contain benign processes, many times forms of chronic pancreatitis.
1 W.L. Frankel, D.M. Proca, Frozen Section Library: Pancreas, Frozen Section Library 8, DOI 10.1007/978-1-4419-7790-8_1, © Springer Science+Business Media, LLC 2011
2
FROZEN SECTION LIBRARY: PANCREAS
Frozen sections for the intra-operative assessment of pancreatic lesions are helpful in many cases. Large centers have reported the accuracy of pancreatic frozen sections ranging from 80 to 90%. The false-negative rates range from 1.6 to 30% (where all deferrals were counted as false negatives) and the false-positive rate is very low, limited to a few cases in individual large studies. Therefore, the positive predictive value for a diagnosis of pancreatic cancer is nearly 100% and the negative predictive value is less but remains excellent.
Chapter 2
Clinical Indications and Surgeons’ Expectations Wendy L. Frankel1 and Mark Bloomston2 1
Department of Pathology, The Ohio State University Medical Center, Columbus, OH 43210, USA,
[email protected] 2 Department of Surgery, The Ohio State University Medical Center, Columbus, OH, USA
The primary goal of all frozen section analysis is to help determine the best treatment during the operative procedure. In order to accomplish this goal, it is vital to have an understanding of the surgeons’ expectations and his/her likely response to a stated diagnosis. This perspective may help determine what the more conservative response would be in a given situation. We may want to favor the diagnosis that would stop a major resection or defer to permanents, rather than making a diagnosis that would prompt the resection in a diagnostically difficult case. Depending on the question and site of the biopsy (pancreatic lesion or metastatic site), a “malignant” diagnosis may either prompt resection or stop the procedure. We typically assume the expected response for each case is either “malignant” or “benign,” but there are circumstances where that is not the most helpful answer. At times, a more appropriate response would be to describe what is present, i.e., the description of an epithelial lining in a cystic lesion is more helpful than merely stating “benign.” The finding of an epithelial lining will prompt resection of the cyst, rather than drainage for a pseudocyst. In contrast, the diagnosis of “benign” could be misunderstood by the surgeon as no tumor present; this may not be the
3 W.L. Frankel, D.M. Proca, Frozen Section Library: Pancreas, Frozen Section Library 8, DOI 10.1007/978-1-4419-7790-8_2, © Springer Science+Business Media, LLC 2011
4
FROZEN SECTION LIBRARY: PANCREAS
case if a low-grade mucinous epithelial lining was identified (suggesting the possibility of a mucinous cystic tumor). Mistakes can be made by not focusing on the most important question and not understanding how the response could alter treatment. An experienced surgeon does not typically take a patient to the operating room to establish a diagnosis, but to treat the presumed diagnosis. The surgeon is already relatively confident of the diagnosis from the clinical and radiologic features (Table 2.1) or pre-operative tissue or cytologic diagnosis. In some cases, there is a pre-operative diagnosis of chronic pancreatitis from a core biopsy or fine needle aspirate. Fibrosis and inflammation frequently surround tumors, so a “benign” biopsy or cytology diagnosis does not preclude a pancreatic resection in the proper clinical setting. This perspective is important and explains why, in the face of a benign frozen section diagnosis, a surgeon may proceed with a Whipple resection in a clinically resectable patient
TABLE 2.1 Clinical and demographic features useful in the differential diagnosis of pancreatic neoplasms.
Gender/age
Clinical presentation
Serous cystadenoma Ductal adenocarcinoma Intraductal papillary mucinous neoplasm Mucinous cystic neoplasm Neuroendocrine tumor Solid pseudopapillary Acinar cell carcinoma
F>M, older
Pain, mass
M>F, older
Jaundice, pain, weight loss Pain, exocrine loss Usually head, cystic/intraductal, mucin at ampulla at ERCP Pain, mass Usually tail, cystic
Pancreatoblastoma
M>F, children
M>F, older
Mostly F, middle age M=F, variable F>M, young adults M>F, older
Radiologic findings Usually body/tail, cystic Usually head, solid
Endocrine hypersecretion Pain, mass
Solid
Jaundice rare, some with fat necrosis and polyarthralgia Pain, mass
Solid
ERCP, endoscopic retrograde cholangiopancreatography.
Solid/cystic
Solid
CLINICAL INDICATIONS AND SURGEONS’ EXPECTATIONS
5
who can tolerate a major procedure. There is, however, some variation among surgeons’ comfort in proceeding with resection without a confirmed diagnosis. This is not only due to patient age and health, but the experience of the surgeon. Many experienced surgeons will not request frozen sections for confirmation of malignancy, but only use frozen sections for unusual findings at surgery, to evaluate close margins, or to diagnose a liver lesion or other possible metastatic lesions. Determining the resectability of a tumor may require correlation between the frozen section diagnosis and the surgeon’s clinical impression during laparotomy. Frozen sections may be used at different decision points during surgery. Fig. 2.1 shows commonly asked intra-operative questions and the typical responses of the surgeon to frozen section diagnoses. Prior to laparotomy, the initial determination of resectability and assessment of the patient’s ability to undergo a major resection have already been made. Since most surgeons take a resectable patient with a presumed pancreatic malignancy to the operating room planning on a pancreas resection, the most important initial surgical decision is the determination of metastasis and resectability. If metastatic carcinoma is present outside of the area typically included in the resection field, the resection will be aborted. Therefore, the diagnosis of a lesion in a distant lymph node, liver, or peritoneum is important for the decision to resect. If there are no obvious metastases, the following step is dependent upon local assessment of resectability. Usually this has been determined prior to surgery using radiologic studies, but sometimes palpation during surgery reveals previously unidentified arterial encasement or direct spread beyond the pancreas. If the tumor is technically resectable and the surgeon is confident in the diagnosis of carcinoma, he/she generally resects without frozen section. If the surgeon is less confident, a frozen section may be sent to pathology to confirm the malignant diagnosis, and confirmation will lead to a resection. A benign, deferred, or non-diagnostic response may prompt more tissue to be sent for additional frozen sections. In the case of a technically unresectable tumor, tissue confirmation is necessary for chemo/radiation therapy. A bypass of the pancreatic, biliary, and gastrointestinal tract may be performed depending on the clinical necessity. In most cases a diagnosis of “malignant” or “benign” is sufficient for the surgeon, but at times, a tumor does not show features of a typical adenocarcinoma. In these cases, the diagnosis of “malignant” or “benign” may not be enough. The tumor type suspected on frozen section may impact the surgical procedure. For
6
FROZEN SECTION LIBRARY: PANCREAS
FIGURE 2.1 The surgical response to frozen section diagnosis of pancreatic lesions and possible metastatic sites. After opening the abdomen, the surgeon determines if there are metastases (not seen on pre-operative evaluation) that would preclude the operation. Frozen sections may be requested at this point to help guide therapy. If there is no intra-operative evidence of metastasis, the decision to proceed with the resection will be based upon the assessment of technical resectability. If the tumor is not resectable, frozen section may be necessary to confirm the diagnosis of carcinoma prior to beginning chemo/radiation therapy. If the tumor is resectable, frozen section may be used to confirm carcinoma (if not previously confirmed on biopsy or fine needle aspiration) prior to resection.
example, neuroendocrine tumor suggested during intra-operative consultation allows surgical resection despite the presence of distant metastases; an unsuspected lymphoma in the pancreas likely precludes surgical resection, which would have been otherwise performed for the diagnosis of adenocarcinoma. If a pancreatic lesion is diagnosed on frozen as suspicious for metastasis from another site, it will likely still be resected, if possible. However, it may be helpful to suggest possible sites of origin so they can be investigated during the operative procedure.
CLINICAL INDICATIONS AND SURGEONS’ EXPECTATIONS
7
In summary, frozen sections of the pancreas are useful in diagnosing pancreatic mass lesions, determining the presence of metastases, and evaluating the margins of resection. Some less common types of pancreatic tumors can be suggested by intraoperative consultation, possibly affecting the surgical procedure.
Chapter 3
The Distinction Between Chronic Pancreatitis and Adenocarcinoma
Tumor-like lesions in the pancreas include any non-neoplastic pancreatic lesion that has the gross appearance of a mass. If diagnosed at frozen section, many of these lesions do not require a major procedure. However, some cases may still require a resection to exclude an adjacent malignancy. The final diagnosis is usually not made until the pancreas is widely sampled on the resection specimen. Examples of tumor-like lesions, include “pseudotumoral” forms of chronic pancreatitis, granulomatous inflammation, malakoplakia, other very rare infectious lesions, and noninflammatory lesions (hamartoma, heterotopic spleen). Some pancreatectomies performed with the pre-operative clinical diagnosis of carcinoma will prove to be non-neoplastic by pathologic examination. Chronic inflammatory lesions are the leading cause and among these, chronic pancreatitis (such as autoimmune or paraduodenal) is most important. Autoimmune pancreatitis (lymphoplasmacytic sclerosing pancreatitis) is characterized by periductal sclerosis and infiltration by lymphocytes and plasma cells and granulocytic epithelial lesions with secondary destruction of the ducts and venulitis. It usually involves the head of the pancreas and the distal bile duct. Paraduodenal (groove) pancreatitis is thought to be a form of alcoholic chronic pancreatitis that mimics periampullary carcinoma due to the selective involvement of the region of papilla minor. Histologic features of chronic pancreatitis that are helpful when seen in a frozen section include preservation of normal lobular architecture with irregular loss of acini, duct alterations,
9 W.L. Frankel, D.M. Proca, Frozen Section Library: Pancreas, Frozen Section Library 8, DOI 10.1007/978-1-4419-7790-8_3, © Springer Science+Business Media, LLC 2011
10
FROZEN SECTION LIBRARY: PANCREAS
chronic inflammation, and fibrosis. Duct alterations include dilatation, cyst formation, calcifications, and inspissated secretions (Fig. 3.1). Duct epithelial changes include mucinous or squamous metaplasia, papillary hyperplasia, atrophy, and slight nuclear enlargement and mild variation in nuclear size (Fig. 3.2). Pseudohyperplasia of the islets of Langerhans due to atrophy of the exocrine pancreas and aggregation of residual islets may give the erroneous impression of a neuroendocrine tumor. Clusters of islets in atrophic chronic pancreatitis can be seen in proximity to small nerve fascicles, and even pseudoperineural invasion by neuroendocrine cells has been described and should not be overcalled on a frozen section (Fig. 3.3).
FIGURE 3.1 Chronic pancreatitis. (a) The pancreas has fairly normal lobular architecture with some dilated ducts, acinar atrophy, and fibrosis. (b) Dilated ducts with fairly regular contours are shown.
CHRONIC PANCREATITIS AND ADENOCARCINOMA
11
FIGURE 3.2 Duct epithelial changes in chronic pancreatitis. (a) These ducts contain mucinous metaplasia (pancreatic intraepithelial neoplasia-1) (arrow) and are beneath an islet (arrowhead). (b) There is minimal nuclear enlargement without variation in size in this frozen section. (c) Mild nuclear variation in size is seen in this area.
12
FROZEN SECTION LIBRARY: PANCREAS
FIGURE 3.3 Islet pseudohyperplasia in chronic pancreatitis. (a) Nests of neuroendocrine cells are conglomerated in an area where there are no acini or ducts. (b) Higher power image of nests of neuroendocrine cells in a frozen section from a pancreas with chronic pancreatitis. (c) Islets (arrowhead) are prominent near a nerve (arrow) in this frozen section. (d) Permanent section from a resection specimen in a case of chronic pancreatitis showing pseudoperineural invasion by neuroendocrine cells.
One of the biggest challenges for a surgical pathologist is to distinguish chronic pancreatitis from ductal adenocarcinoma on a frozen section. There are many criteria that help with the differential diagnosis, but few are absolute. Some histologic features useful in permanent sections have not been found to be as useful in frozen sections, such as nuclear hyperchromasia. CRITERIA OF MALIGNANCY IN FROZEN SECTION Multiple histologic features can be assessed to distinguish malignant from benign lesions. In 1981, Hyland et al. proposed the histologic features most helpful in differentiating malignant versus benign lesions on frozen sections. They described major and minor criteria useful for the diagnosis of ductal adenocarcinoma derived from a review of 64 frozen sections. Major criteria were present in all malignant cases and not seen in benign cases and minor criteria were diagnostically useful, but present less often
CHRONIC PANCREATITIS AND ADENOCARCINOMA
13
and less reproducible between observers. The three “major criteria” for malignancy were nuclear size variation of at least 4:1, disorganized duct distribution, and partial ducts (Fig. 3.4). The “minor criteria” included large irregular nucleoli, necrotic debris, epithelial mitoses, and perineural invasion (Fig. 3.5). In addition, the presence of infiltrating, severely atypical single cells and desmoplasia represent other helpful features of malignancy on a frozen section. Glands adjacent to muscular arteries and within fat are other features useful in the diagnosis of malignancy (Fig. 3.6). Lipomatosis (secondary to aging, obesity, type II diabetes) may cause fat infiltration of the pancreas but ducts usually are accompanied by acini or islets.
FIGURE 3.4 Major criteria for malignancy. (a) This tumor shows nuclear atypia and size variation of greater than 4:1. (b) Haphazard and disorganized ducts are present in this frozen section. (c) Partial ducts are seen at high power. (d) Single cell infiltration and partial ducts are shown in this frozen section.
14
FROZEN SECTION LIBRARY: PANCREAS
FIGURE 3.4 Major criteria for malignancy (continued)
CHRONIC PANCREATITIS AND ADENOCARCINOMA
15
FIGURE 3.5 Minor criteria for malignancy. (a) There are large nucleoli (arrow) in some atypical cells in this image. (b) Necrotic debris is present within the malignant glands. (c) Occasional epithelial mitoses are present in these glands (arrow). (d) Perineural invasion is not commonly seen in frozen sections, but when present is helpful, as shown in this figure.
16
FROZEN SECTION LIBRARY: PANCREAS
FIGURE 3.5 Minor criteria for malignancy (continued)
CHRONIC PANCREATITIS AND ADENOCARCINOMA
17
FIGURE 3.6 Additional features of malignancy useful in a frozen section. (a) Desmoplasia and single cell infiltration are seen in this tumor. (b) Glands are seen adjacent to muscular arteries. (c) Glands are present in direct opposition to fat without acini or islets.
The approach we have found most useful when evaluating pancreas frozen sections is to initially scan the slide on low power to get a general idea of the architecture. In chronic pancreatitis, there is retention of the lobular distribution of the ducts rather than haphazard, irregular ducts that may be angulated and incomplete in adenocarcinoma (Fig. 3.7). At higher power, cytologic features can then be evaluated and include pleomorphism, nuclear crowding and stratification, high nuclear to cytoplasmic ratio (nuclei are up to three times the size of a lymphocyte), significant variation in nuclear size (at least 4:1), prominent nucleoli, and mitoses (Fig. 3.8). Figure 3.9 shows an example of this approach. Some cases remain difficult and an “atypical” diagnosis or deferral may be indicated (Fig. 3.10). The surgeon may submit additional frozen sections of the lesion and some are diagnostic. However, many surgeons will proceed with resection without a definitive diagnosis if they are clinically suspicious. Table 3.1 compares features most useful in distinguishing adenocarcinoma from chronic pancreatitis in frozen sections.
18
FROZEN SECTION LIBRARY: PANCREAS
FIGURE 3.7 Low power comparison between chronic pancreatitis and adenocarcinoma. (a and b) These frozen sections from chronic pancreatitis show retention of the lobular architecture. (c and d) These frozen sections from adenocarcinoma show haphazard and irregular glands with loss of lobular architecture.
CHRONIC PANCREATITIS AND ADENOCARCINOMA
19
FIGURE 3.7 Low power comparison between chronic pancreatitis and adenocarcinoma (continued)
20
FROZEN SECTION LIBRARY: PANCREAS
FIGURE 3.8 Higher power comparison between chronic pancreatitis and adenocarcinoma. (a) This case of chronic pancreatitis shows fairly rounded glands without nuclear atypia. (b) Rounded glands are present with slight nuclear variability in size. (c) This adenocarcinoma contains irregular glands with necrotic debris, nuclear atypia, and variation in size. (d) The malignant glands contain atypical cells with variation in nuclear size.
CHRONIC PANCREATITIS AND ADENOCARCINOMA
21
FIGURE 3.9 Evaluation of frozen section using the approach of scanning on low power for lobular architecture and then assessing cytologic features at higher power. (a) The lobular architecture appears mainly intact with one area that appears somewhat altered (arrow). The diagnosis is unclear at this power. (b) High-power image showing clear-cut malignant features including irregular glandular contours, nuclear atypia with fourfold variation in nuclear size.
22
FROZEN SECTION LIBRARY: PANCREAS
FIGURE 3.10 Challenging examples. (a) A few glands show slightly irregular contours in this frozen section. Nuclear atypia is present with variation in size but not fourfold (arrow). There were no other definitive features of malignancy, and this frozen section was called “atypical but favor benign.” The pancreatic resection was performed given the clinical suspicion for malignancy. Straightforward adenocarcinoma was found in the resection specimen that was much more pleomorphic than the area shown. (b) This frozen section shows rounded glandular contours with some nuclear enlargement (arrow) and mild variation in size. The frozen section was called “atypical but favor benign,” and the resection proceeded given the clinical suspicion for malignancy. Chronic pancreatitis was diagnosed in the resection specimen. (c) The lobular architecture is mainly intact but there are some dilated slightly irregular glands in this frozen section. (d) This high-power image of the previous figure shows the most cytologically atypical glands in the area. The glands show a rounded contour, but cells are atypical. No definitive criteria of malignancy were seen. The diagnosis was “atypical, defer,” and the resection proceeded based upon clinical suspicion. Adenocarcinoma was clearly seen in the area of the frozen section on permanent and deeper sections.
CHRONIC PANCREATITIS AND ADENOCARCINOMA
23
TABLE 3.1 Frozen section differential: duct adenocarcinoma versus chronic pancreatitis.
Glandular architecture Glands in fat/adjacent to arteries Nuclei Nucleoli Necrosis Mitoses Perineural invasion
Adenocarcinoma
Chronic pancreatitis
Haphazard, irregular, incomplete lumen ±
Lobular, round, smooth contours –
Variability ≥ 4:1 Prominent, may be multiple ± ± ±
