Endoscopic Ultrasound Guided Trucut Biopsy of the Cyst Wall for Diagnosing Cystic Pancreatic Tumors

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2005;3: Endoscopic Ultrasound Guided Trucut Biopsy of the Cyst Wall for Diagnosing Cystic Pancreatic Tumors MICHAEL J. LEVY,* THOMAS C. SMYRK, RAGHURAM P. REDDY,* JONATHAN E. CLAIN,* GAVIN C. HAREWOOD,* MICHAEL L. KENDRICK, RANDALL K. PEARSON,* BRET T. PETERSEN,* ELIZABETH RAJAN,* MARK D. TOPAZIAN,* KENNETH K. WANG,* MAURITS J. WIERSEMA, TONY E. YUSUF,* and SURESH T. CHARI* *Division of Gastroenterology and Hepatology, Department of Pathology, and Department of Surgery, Mayo Clinic School of Medicine, Rochester, Minnesota; and GI Consultants, Inc, Fort Wayne, Indiana Background & Aims: Nonoperative methods for diagnosis of pancreas cysts often lack sufficient accuracy. Accurate diagnosis is needed to determine prognosis and guide clinical management. The aim of this study was to determine whether the tissue obtained by endoscopic ultrasound guided trucut biopsy (EUS TCB) is sufficient for histologic diagnosis of cystic pancreatic tumors (CPTs). Methods: EUS TCB was performed in patients with a suspected CPT. A dedicated gastrointestinal pathologist reviewed the core biopsies. The final diagnosis was based on clinical, laboratory, imaging, and biopsy findings, and resected specimens when available. Results: EUS TCB was performed in 10 patients with a suspected CPT. Final diagnoses included serous cystadenoma (SCA, n 5), islet cell tumor (n 2), mixed seromucinous lesion (n 1), polycystic disease of the pancreas (n 1), and pseudocyst (n 1). EUS TCB was nondiagnostic in 3 of 10 patients. Among the other 7 patients, TCB diagnosed 4 SCAs, obviating the need for planned surgery in 3 patients. In the fourth patient with an SCA, the TCB result ruled out metastatic disease from locally recurrent lung cancer, allowing a narrowed radiation field. EUS TCB confirmed the need for surgery in 2 patients with an islet cell tumor. In 1 patient, EUS TCB findings were partially diagnostic, leading to previously unplanned surgery. Conclusions: This report establishes the capability and safety of EUS TCB to collect sufficient tissue for diagnosing CPTs. The results might help guide clinical management. Cystic lesions of the pancreas include simple cysts, pseudocysts, and various cystic pancreatic tumors (CPTs) including solid tumors that undergo cystic degeneration. 1,2 Traditionally, pseudocysts have accounted for most cystic lesions of the pancreas (80% 90%). However, improvement in cross-sectional imaging resolution has resulted in more frequent identification of cystic lesions of the pancreas. CPTs are broadly classified according to their malignant potential, which impacts prognosis and therapy. Clinically it is useful to classify them as low-risk lesions, which have little or no malignant potential, or as high-risk lesions, which are premalignant or frankly malignant. Low-risk lesions include nonmucinous lesions such as serous cystadenomas (SCAs), which have a very low malignant potential, and simple cysts and pseudocysts, which are always benign. 3,4 These lesions are generally only resected when causing symptoms or complications (eg, obstruction, pain) related to their location and size. High-risk lesions include mucinous lesions (mucinous cystic neoplasms [MCNs] and intraductal papillary mucinous neoplasms [IPMNs]), which are premalignant or malignant tumors at the time of diagnosis. 3,4 Other high-risk lesions include solid tumors that undergo cystic degeneration such as ductal adenocarcinoma, islet cell tumors, and pseudopapillary tumors. High-risk lesions are generally resected in patients who are good operative candidates in whom there is no evidence of distant metastasis. Noninvasive cross-sectional imaging offers limited sensitivity and specificity for classifying cystic lesions of the pancreas. 5,6 Although the performance characteristics of endoscopic ultrasound (EUS) are higher than conventional cross-sectional imaging, EUS morphologic features alone cannot accurately distinguish the tumor type or identify malignant transformation. 7 9 Cyst fluid aspiration can be performed at the time of EUS but seldom acquires a specimen adequate for cytologic diagnosis. 9,10 The addition of tumor marker levels and amylase levels Abbreviations used in this paper: CEA, carcinoembryonic antigen; CPT, cystic pancreatic tumor; EUS TCB, endoscopic ultrasound guided trucut biopsy; FNA, fine-needle aspiration; GI, gastrointestinal; IPMN, intraductal papillary mucinous neoplasm; MCN, mucinous cystic neoplasm; SCA, serous cystadenoma by the American Gastroenterological Association /05/$30.00 PII: /S (05)

2 October 2005 TRUCUT BIOPSY OF CYSTIC PANCREATIC TUMORS 975 improves the diagnostic accuracy, but much overlap is seen in these values among CPTs Unfortunately, despite an exhaustive evaluation, the diagnosis often remains uncertain. The result might be inappropriate management of CPTs, leading to resection of benign lesions or failure to resect a premalignant or malignant lesion falsely diagnosed as benign Definitive diagnosis of CPTs has until recently required surgical biopsy or resection. Development of an endoscopic ultrasound guided trucut biopsy (EUS TCB) needle allows acquisition of a core of tissue, permitting histologic examination. This offers the potential for definitive diagnosis without surgery and might altogether obviate the need for surgical intervention or further follow-up. The trucut biopsy device (Quick-Core; Wilson-Cook, Winston-Salem, NC) contains a 19-gauge needle with a tissue tray and sliding sheath that permits a histologic core to be obtained. 17,18 Our recently described experience with EUS TCB initially in swine 19 and later in human beings 17,18 demonstrated the safety of this device for acquiring histologic tissue representative of the target organs sampled. Recognizing that many types of CPTs have a unique cyst wall histology, we sought to determine whether the tissue obtained with EUS TCB of CPTs can permit a histologic diagnosis. Methods This study includes all patients (n 10) with a cystic pancreatic lesion that underwent EUS TCB from December 2003 November The decision to perform TCB was based on a particular endosonographer s experience with the TCB device, location of the lesion in the pancreatic body or tail allowing biopsy, and after discussion with the referring physician to determine the potential significance of the findings on patient management. Nine of 10 patients also underwent EUS FNA for cytologic analysis and determination of the carcinoembryonic antigen (CEA) level. Devices used included disposable 19- and 22-gauge needles (Quick-Core and Echotip; Wilson- Cook). The trucut needle was positioned so that the specimen tray overlapped a portion of the cyst wall as well as adjacent pancreatic parenchyma (Figure 1). The spring handle was then fired, and the needle was removed. The TCB tissue cores were fixed in formalin, embedded in paraffin, and stained with hematoxylin-eosin before histologic examination. The Institutional Review Board granted approval for retrospective review of the patient charts, and informed consent was obtained for all procedures described in this report. A dedicated gastrointestinal (GI) pathologist blinded to the clinical data reviewed the EUS FNA samples, when obtained, on the same day the EUS was performed. Because of the additional time required for processing of histologic samples (24 hours), a second dedicated GI pathologist, also blinded to the clinical data and EUS FNA interpretation, examined the histologic sample the day after Figure 1. Positioning of the trucut biopsy needle so that the specimen tray overlaps a portion of the cyst wall as well as adjacent pancreatic parenchyma. the EUS. Final diagnosis was determined by a combination of clinical presentation, patient outcome, laboratory findings, imaging studies, and pathology findings by using the World Health Organization guidelines. As per our routine clinical practice, patients were evaluated by a physician hours after biopsy to discuss results and assess for complications. Results EUS TCB was performed in 10 patients with a suspected CPT and followed for a mean of 6.7 months (range, 2 13 months). The mean age was 65 years (range, years), 6 of whom were female. A mean of 2.7 (range, 1 7) fine-needle aspirations (FNAs) was obtained in 9 of 10 patients, and all yielded nondiagnostic cytology. A mean of 2.4 trucut biopsies (range, 1 6) was obtained. No complications were identified after EUS TCB. Nine of 10 patients were given a final diagnosis of CPT: SCA (n 5) (Figure 2), islet cell tumor (n 2), polycystic disease of the pancreas, and mixed seromucinous lesion (n 1). One patient had a pseudocyst (Figure 3). The patient with polycystic disease appeared to have the isolated variant without evidence of associated von Hippel-Lindau disease or adult-type, autosomal dominant, polycystic kidney disease. EUS TCB was nondiagnostic in 3 of 10 patients, one patient each with an SCA, polycystic disease of the pancreas, and a pseudocyst. The TCB specimen from the patient with the pseudocyst was reported as disintegrated and not suitable for review. The final diagnosis of a pseudocyst was based on the patient history, clinical follow-up, and subsequent imaging that demonstrated complete resolution of the cyst. EUS FNA yielded negative cytology in each these 3 patients, and the CEA levels were 75, 16.4, and 54 ng/ml, respectively.

3 976 LEVY ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 10 the diagnosis was unclear after pancreatic protocol computed tomography, which led to EUS examination. The patients initially underwent EUS FNA, 5 and 3 biopsies, respectively. FNA yielded nondiagnostic, bloody specimens, despite the use of no negative pressure after the initial biopsy in each patient. In addition, fluid was sent for CEA testing, yielding values of 2.6 and 1.5 ng/ml. TCB and histologic review established the diagnosis of islet cell tumor and indicated the need for surgical intervention. The frequent overlap in CEA levels among mucinous and nonmucinous lesions did not allow firm exclusion of a mucinous lesion. In addition, although the CEA value for each of the 9 aforementioned patients suggested the presence of a nonmucinous lesion, the levels did not (and never do) allow firm distinction of the specific cyst histopathology. The lack of diagnostic distinction is important, given that certain types of nonmucinous lesions require surgical intervention versus others that are well managed by observation alone. For the last patient, EUS TCB findings were partially diagnostic in that histologic review revealed the mucinous component of a seromucinous lesion but failed to identify the serous component. FNA produced a negative cytology, but the CEA level of 1900 ng/ml strongly suggested a mucinous lesion. Seromucinous le- Figure 2. (A) A 43-year-old woman presenting with back pain underwent abdominal computed tomography scan and was found to have a cm pancreatic body cyst. It was unclear whether it represented a mucinous or nonmucinous lesion, and she was referred for resection. EUS FNA and cyst fluid analysis revealed a CEA level of 1.8 ng/ml and nondiagnostic cytology. (B) EUS TCB demonstrated classic histologic features of an SCA, obviating the need for surgery in favor of conservative management. For 6 of the remaining 7 patients, the diagnosis provided by histologic review of the TCB tissue core established the diagnosis. In 3 of these patients, the TCB results obviated the need for surgery, which was initially planned, after establishing the diagnosis of an SCA. In the fourth patient, the histologic finding of an SCA in the TCB specimen helped rule out metastatic disease from locally recurrent lung cancer, thereby allowing a more narrowed field of radiation. The CEA levels, which were obtained in 3 of these 4 patients, varied from ng/ml. EUS TCB findings confirmed the need for surgery in 2 patients as a result of histology diagnostic of an islet cell tumor. The islet cell tumors were predominantly cystic with a smaller solid component. For both patients Figure 3. A 50-year-old man with chronic pancreatitis caused by alcohol was found to have a dilated main pancreatic duct and a cyst in the pancreatic body/tail that was believed to represent a pseudocyst. Follow-up computed tomography scan demonstrated increased cyst size, wall thickening, and inflammation, raising concern for a cystic neoplasm. EUS identified focal wall thickening within what was believed to be a pseudocyst or less likely side branch IPMN. The cyst fluid CEA level was 54.2 ng/ml, and the amylase level was 7142 U/L. The EUS TCB specimen was reported as disintegrated and not suitable for review. With conservative care, the cyst has completely resolved. The clinical course and new radiographic findings support the diagnosis of a pseudocyst.

4 October 2005 TRUCUT BIOPSY OF CYSTIC PANCREATIC TUMORS 977 sions are rare and have an uncertain course and malignant potential. Identification of the mucinous component led to surgery, which is normally our practice for surgically fit patients with a mucinous lesion. Before EUS, the referring physician had planned conservative care and follow-up imaging for what was presumed to be an SCA. No complications were identified after EUS TCB. Discussion Management of CPTs remains a challenge as a result of limited understanding of their natural history and suboptimal diagnostic accuracy of noninvasive imaging. Incidentally discovered (asymptomatic) cystic pancreatic lesions place greater importance on identifying methods to more accurately characterize these abnormalities. Current clinical practice focuses on differentiation of mucinous and islet cell tumors from nonmucinous and inflammatory lesions by using clinical, radiologic, endosonographic, and biochemical criteria. These tests have suboptimal sensitivity and specificity, which might result in misdiagnosis and incorrect management of pseudocysts. 6,14,20 22 EUS FNA and cytologic analysis of aspirated fluid yield the correct diagnosis in a small number of patients. Core biopsies might be superior to FNA, because the larger tissue samples allow preservation of tissue architecture and histologic examination, which has been shown to be particularly useful for lymphomas, gastrointestinal stromal tumors, and autoimmune pancreatitis. 17,18 Specific to CPTs, TCB has the advantage of sampling a core of tissue consisting of the cyst wall and the supportive stroma. In addition to histopathologic examination, immunohistochemical staining can be performed. Furthermore, for FNA, the accuracy of cytologic review is hampered by the presence of blood, benign epithelial cells, desmoplasia, and well-differentiated tumors. These problems are often avoided with TCB. In addition, a cytopathologist does not need to be available during the procedure to assess adequacy of the specimen after TCB. Perhaps the greatest limitation with cytologic analysis of cyst fluid is the dependence on neoplasm shedding of viable cells. Well-differentiated and premalignant lesions might not provide sufficient cellularity to permit their early diagnosis on the basis of cytology alone. Certain types of CPTs have characteristic and even unique histology and immunophenotype that if acquired during biopsy can establish the diagnosis. For example, SCAs are lined by glycogen rich cuboidal epithelium, whereas MCNs and IPMNs are lined by mucinous epithelium. Furthermore, solid neoplasms that undergo cystic degeneration, such as ductal adenocarcinomas and islet cell tumors, contain characteristic histology. When demonstrated histologically in a trucut core biopsy, these characteristic histologic findings are diagnostic. However, lesions such as pseudocysts and simple cysts do not contain unique histologic findings and therefore might not be diagnosed by TCB. In addition, although MCN and IPMN both contain mucinous epithelium, this finding alone cannot distinguish these 2 entities. However, the presence of underlying ovarian-like stroma is diagnostic of MCN and considered a necessary criterion by many. 15,23 IPMNs are usually lined by mucin-producing epithelium; however, denudation of the epithelial lining is not uncommon. 24 Although mucin-producing epithelium without supportive ovarian stroma might suggest an IPMN, lack of such epithelium in the core biopsy sample is inconclusive. Distinction is important, given the difference in biologic behavior between MCN, main duct IPMN, and branch duct IPMN, including differences in the prevalence of malignancy at resection, recurrence after resection, and presence of synchronous lesions. 14,23,25 27 Although our findings support the capability and safety of EUS TCB to collect sufficient tissue for diagnosing certain CPTs, none of the patients had an MCN, and it is unclear whether TCB can verify the presence of ovarian stroma in vivo. However, our recent (unpublished) data regarding the use of TCB in postsurgical specimens demonstrate the ability of TCB to diagnose MCNs on the basis of the presence of ovarian stroma. This finding ex vivo offers promise that the same can be achieved in patients before surgery. Performance of this study provided important information concerning the method necessary for histologic review of core tissue samples. Core biopsy of a cystic lesion typically yields a fragmented sample containing scattered pieces of tissue interspersed among areas of the tray lacking tissue. Careful and complete histologic review is required because diagnostic tissue might be contained only within focal areas of the specimen. Failure to examine the entire tissue sample might lead to failed diagnosis. The decision to perform TCB on the basis of an endosonographer s TCB experience, location of the lesion in the pancreatic body or tail allowing biopsy, and after discussion with the referring physician to determine the potential significance of the findings on patient management might have favorably biased the TCB results. Histologic review of the TCB specimens firmly established the diagnosis in 6 patients, was partially diagnostic in 1 patient, and was nondiagnostic in 3 patients. This compares to cytologic review of the FNA specimens that was

5 978 LEVY ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 10 nondiagnostic in all 9 patients. The role of tumor markers is controversial, as is the threshold value that discriminates the lesion type with greatest accuracy. The considerable overlap in tumor marker levels between mucinous and nonmucinous lesions limits their diagnostic utility. 28,29 In addition, although the low CEA values strongly suggest nonmucinous lesions, the levels do not allow distinction of the various types of nonmucinous lesions that might lead to observation in a patient for whom surgery is indicated. In general, we view the CEA level as only one of several important pieces of information that must be considered when deciding patient care and alone cannot establish the specific tumor type. It is important to note that this study was not intended to directly compare the performance characteristics of EUS TCB with EUS FNA in patients with CPTs, and doing so would require different methodology. The intent instead was to determine whether the tissue obtained by EUS TCB is sufficient for histologic diagnosis of CPT, which our findings established. Summary Combined clinical, imaging, and laboratory evaluation of patients with CPTs is often insufficient to allow accurate diagnosis. The limited accuracy might negatively impact clinical management and prognosis. The addition of EUS TCB appears safe in patients with CPTs and might provide a sufficient tissue sample to allow adequate histologic examination and diagnosis. By doing so, the team of gastroenterologists and surgeons can more appropriately deliver care to their patients with CPTs. However, the use of EUS TCB might be of limited value in providing the diagnosis of lesions without characteristic histology. A clearer understanding of the utility and safety of EUS TCB for CPTs requires prospective controlled evaluation in a large number of patients. References 1. Fernandez-del Castillo C, Warshaw AL. Cystic tumors of the pancreas. Surg Clin North Am 1995;75: ReMine SG, Frey D, Rossi RL, et al. Cystic neoplasms of the pancreas. Arch Surg 1987;122: Siech M, Tripp K, Schmidt-Rohlfing B, et al. Cystic tumours of the pancreas: diagnostic accuracy, pathologic observations and surgical consequences. Langenbecks Arch Surg 1998;383: Sarr MG, Carpenter HA, Prabhakar LP, et al. Clinical and pathologic correlation of 84 mucinous cystic neoplasms of the pancreas: can one reliably differentiate benign from malignant (or premalignant) neoplasms? Ann Surg 2000;231: Zirinsky K, Abiri M, Baer JW. Computed tomography demonstration of pancreatic microcystic adenoma. Am J Gastroenterol 1984;79: Curry CA, Eng J, Horton KM, et al. CT of primary cystic pancreatic neoplasms: can CT be used for patient triage and treatment? AJR Am J Roentgenol 2000;175: Ahmad NA, Kochman ML, Lewis JD, et al. Can EUS alone differentiate between malignant and benign cystic lesions of the pancreas? Am J Gastroenterol 2001;96: 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: Sedlack R, Affi A, Vazquez-Sequeiros E, et al. Utility of EUS in the evaluation of cystic pancreatic lesions. Gastrointest Endosc 2002;56: Pinto MM, Meriano FV. Diagnosis of cystic pancreatic lesions by cytologic examination and carcinoembryonic antigen and amylase assays of cyst contents. Acta Cytol 1991;35: Alles AJ, Warshaw AL, Southern JF, et al. Expression of CA 72-4 (TAG-72) in the fluid contents of pancreatic cysts: a new marker to distinguish malignant pancreatic cystic tumors from benign neoplasms and pseudocysts. Ann Surg 1994;219: 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: Sperti C, Pasquali C, Guolo P, et al. Serum tumor markers and cyst fluid analysis are useful for the diagnosis of pancreatic cystic tumors. Cancer 1996;78: Warshaw AL, Compton CC, Lewandrowski K, et al. Cystic tumors of the pancreas: new clinical, radiologic, and pathologic observations in 67 patients. Ann Surg 1990;212: 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: Fernandez-del Castillo C, Warshaw AL. Current management of cystic neoplasms of the pancreas. Adv Surg 2000;34: Levy MJ, Wiersema MJ, Clain JE, et al. Comparison of endoscopic ultrasound-guided trucut biopsy (EUS-TCB) to endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) (abstr W1228). Gastrointest Endosc 2003;57:AB Levy MJ, Jondal ML, Clain JE, et al. Preliminary experience with an EUS-guided trucut biopsy needle compared with EUS-guided FNA. Gastrointest Endosc 2003;57: Wiersema MJ, Levy MJ, Harewood GC, et al. Initial experience with EUS-guided trucut needle biopsies of perigastric organs. Gastrointest Endosc 2002;56: Scott J, Martin I, Redhead D, et al. Mucinous cystic neoplasms of the pancreas: imaging features and diagnostic difficulties. Clin Radiol 2000;55: Procacci C, Biasiutti C, Carbognin G, et al. Characterization of cystic tumors of the pancreas: CT accuracy. J Comput Assist Tomogr 1999;23: Fukushima N, Mukai K, Kanai Y, et al. Intraductal papillary tumors and mucinous cystic tumors of the pancreas: clinicopathologic study of 38 cases. Hum Pathol 1997;28: Thompson LD, Becker RC, Przygodzki RM, et al. Mucinous cystic neoplasm (mucinous cystadenocarcinoma of low-grade malignant potential) of the pancreas: a clinicopathologic study of 130 cases. Am J Surg Pathol 1999;23: Levy MJ, Clain JE. Evaluation and management of cystic pancreatic tumors: emphasis on the role of EUS FNA. Clin Gastroenterol Hepatol 2004;2: Sperti C, Pasquali C, Pedrazzoli S, et al. Expression of mucin-like carcinoma-associated antigen in the cyst fluid differentiates mucinous from nonmucinous pancreatic cysts. Am J Gastroenterol 1997;92: 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: Chari ST, Yadav D, Smyrk TC, et al. Study of recurrence after

6 October 2005 TRUCUT BIOPSY OF CYSTIC PANCREATIC TUMORS 979 surgical resection of intraductal papillary mucinous neoplasm of the pancreas. Gastroenterology 2002;123: 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: Hammel P, Voitot H, Vilgrain V, et al. Diagnostic value of CA 72-4 and carcinoembryonic antigen determination in the fluid of pancreatic cystic lesions. Eur J Gastroenterol Hepatol 1998;10: Address requests for reprints to: Michael J. Levy, MD, Mayo Clinic School of Medicine, Division of Gastroenterology and Hepatology, 200 First Street SW, Rochester, Minnesota levy.michael@ mayo.edu; fax: (507)