Use of Samples From Endoscopic Ultrasound Guided 19-Gauge Fine- Needle Aspiration in Diagnosis of Autoimmune Pancreatitis

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2012;10: Use of Samples From Endoscopic Ultrasound Guided 19-Gauge Fine- Needle Aspiration in Diagnosis of Autoimmune Pancreatitis TAKUJI IWASHITA,* ICHIRO YASUDA,* SHINPEI DOI,* NOBUHIRO ANDO,* MASANORI NAKASHIMA,* SEIJI ADACHI,* YOSHINOBU HIROSE, TSUYOSHI MUKAI, KEISUKE IWATA, EIICHI TOMITA, TAKAO ITOI, and HISATAKA MORIWAKI* *First Department of Internal Medicine, Department of Clinical Pathology, Gifu University Hospital, Gifu; Department of Gastroenterology, Gifu Municipal Hospital, Gifu; Department of Gastroenterology and Hepatology, Tokyo Medical University, Tokyo, Japan BACKGROUND & AIMS: Histologic techniques are used to distinguish autoimmune pancreatitis (AIP) from pancreatic malignancies and to confirm the etiology of pancreatitis. Endoscopic ultrasound guided fine-needle aspiration (EUS-FNA) is a well-established technique used in the diagnosis of pancreatic cancer. However, it is unclear whether specimens obtained from pancreatic lesions by EUS- FNA are adequate for the histologic diagnosis of AIP, because the evaluation of tissue architecture and immunostaining assays usually require larger samples. METHODS: We evaluated samples collected by EUS-FNA with a conventional 19-gauge needle by histologic analysis, looking for features of AIP. We analyzed data from 44 patients who were diagnosed with AIP and underwent EUS-FNA with a 19-gauge needle from January 2004 to September The FNA specimens were reviewed by histologic analysis; AIP was diagnosed based on the presence of lymphoplasmacytic sclerosing pancreatitis or immunoglobulin (Ig)G4-positive plasma cells in the infiltrate. RE- SULTS: The specimen amount was inadequate from 3 patients. Among the remaining 41 patients, histopathologic analysis revealed lymphoplasmacytic sclerosing pancreatitis in 17 samples and IgG4- positive plasma cells in 5 (3 samples were positive for both); no samples had granulocytic epithelial lesions. Therefore, 19 patients (43%) were diagnosed with AIP based on histologic analysis. One patient had temporary abdominal pain. CONCLUSIONS: EUS- FNA, with a 19-gauge needle, is a safe and reliable procedure for obtaining pancreatic samples for the histologic analysis of AIP. Although it does not have a high diagnostic yield, it might be useful in patients without typical features of AIP because it would allow patients to avoid surgery. Keywords: Diagnostics; Noninvasive; Pancreas; Marker. Watch this article s video abstract and others at tiny.cc/bz9jv. Scan the quick response (QR) code to the left with your mobile device to watch this article s video abstract and others. Don t have a QR code reader? Get one at mobiletag.com/en/download.php. main pancreatic duct (MPD), an increase in the levels of autoimmune-related serologic markers, fibrotic changes with lymphocytic infiltration, and improved response to steroid therapy. 1 A diagnosis of AIP usually is made on the basis of a combination of these clinical characteristics. 2 7 In the diagnosis of AIP, pathologic analysis of the pancreatic lesion is important for 2 reasons. The first is to distinguish AIP from other pancreatic malignancies because AIP can mimic pancreatic cancer and can be misdiagnosed as malignancy. 8 The second is to confirm the etiology of pancreatitis on the basis of the histologic hallmark of AIP, that is, lymphoplasmacytic sclerosing pancreatitis (LPSP) 9 11 or infiltration of immunoglobulin (Ig)G4-positive plasma cells, 12 particularly in patients lacking typical findings of AIP. Endoscopic ultrasound guided fine-needle aspiration (EUS- FNA) is a well-established technique and has been used widely to obtain pathologic specimens from the pancreas. The diagnostic accuracy of EUS-FNA for pancreatic malignancies has been reported as 75% to 92% based only on cytologic assessment Although the possibility of false-negative results still remains, EUS-FNA appears to be effective in distinguishing AIP from pancreatic cancer. However, it is still unclear whether specimens obtained by EUS-FNA from pancreatic lesions of AIP patients are adequate for the histologic diagnosis of AIP because the histologic characteristics of AIP have been investigated mostly in surgically resected specimens. Moreover, the evaluation of tissue structure and immunohistochemical staining usually requires large amounts of specimens. We previously reported that EUS-FNA with a larger caliber (19 gauge) needle, rather than a 22-gauge or 25-gauge needle, provided better yields of specimens for histologic analysis Subsequently, a conventional 19-gauge needle has been used routinely when technically possible. The aim of this study was to evaluate the yield of EUS-FNA with a conventional 19-gauge needle for the histologic diagnosis of AIP in patients with suspected clinical diagnoses of AIP. Autoimmune pancreatitis (AIP) is defined as chronic inflammation of the pancreas owing to an autoimmune mechanism. The presence of pancreatitis associated with an autoimmune mechanism had been recognized, but the term autoimmune pancreatitis was first proposed in Thereafter, AIP has attracted increasing interest and has been investigated by many researchers. The clinical characteristics of AIP include a diffusely enlarged pancreas, diffuse irregular narrowing of the Abbreviations used in this paper: AIP, autoimmune pancreatitis; CT, computed tomography; EUS-FNA, endoscopic ultrasound guided fine-needle aspiration; GEL, granulocytic epithelial lesion; HPF, high-power field; ICP, idiopathic chronic pancreatitis; IDCP, idiopathic duct centric chronic pancreatitis; Ig, immunoglobulin; LPSP, lymphoplasmacytic sclerosing pancreatitis; MPD, main pancreatic duct; TCB, Tru-Cut biopsy needle by the AGA Institute /$36.00 doi: /j.cgh

2 March 2012 EUS-FNA FOR HISTOLOGIC DIAGNOSIS OF AIP 317 Figure 1. Storiform fibrosis and lymphoplasmacytic infiltration around a small pancreatic duct (H&E staining; original magnification, 20). Methods Study Population Patients who received clinical diagnoses of AIP were identified by a retrospective review of the EUS database, which included all EUS cases performed at the First Department of Internal Medicine at Gifu University Hospital between January 2004 and September The original clinical diagnosis of AIP had been made on the basis of satisfying any of the diagnostic criteria for AIP of the Japan Pancreas Society. 2,6 Patients were selected and narrowed down to 46 patients from the EUS database analysis. The diagnosis of AIP was confirmed in these patients by applying the Asian diagnostic criteria 7 on the basis of their imaging and serologic findings and response to steroid therapy, but excluding histologic findings, which would enable the identification of true AIP patients. The diagnoses of AIP were reconfirmed in all 46 patients. In these 46 patients, FNA was performed with a 22-gauge needle in 2 patients and a 19-gauge needle in the remaining 44 patients. These 44 patients finally were reviewed in this study. All of the patients provided written informed consent to undergo EUS- FNA. The study protocol was approved by the review board for human research at Gifu University and was conducted in accordance with the human and ethical principles of research set forth by the Declaration of Helsinki. Endoscopic Ultrasound and Endoscopic Ultrasound Guided Fine-Needle Aspiration EUS was performed using an oblique forward-viewing electronic linear scanning video echoendoscope equipped with an elevator and a 2.8-mm diameter working channel (GF- UC240P-AL5; Olympus, Co, Ltd, Tokyo, Japan). The echoendoscope was connected to a processor with color Doppler function (SSD-5000; Aloka Co, Ltd, Tokyo, Japan). The patients underwent EUS-FNA under conscious sedation with midazolam and pentazocine, and vital signs were monitored. A single experienced endosonographer (I.Y.) performed EUS-FNA in all patients. The puncture was achieved with a 19-gauge needle (EchoTip; Wilson-Cook, Winston-Salem, NC) and was guided by real-time EUS imaging via the stomach or duodenum after careful examination including color Doppler imaging to ascertain that no major vessels or the MPD was interposed on the puncture path. The stylet was removed completely, and a 10-mL syringe was attached to the needle. This was followed by the application of 10 ml of negative pressure. Two to 5 multiple movements were made within the lesion and the suction slowly was released after the completion of this step. The needle then was withdrawn into the sheath, and the entire system was withdrawn from the biopsy channel. The aspirated material was expelled onto glass slides by carefully re-inserting the stylet into the needle. The macroscopic characteristics of this material were evaluated, and the whitish portions were collected and placed on a small piece of filter paper. The sample then was placed in formalin solution for histologic examination. The remaining materials were smeared on glass slides for cytologic examination. Because neither pathologists nor cytologists were present on-site at our institutions, the puncture was repeated until a whitish material could be observed macroscopically. However, if no such specimen was obtained after 3 consecutive punctures, only a cytologic examination on smeared glass slides usually was performed. The procedure was performed on an outpatient basis unless the patient already was hospitalized for other medical conditions. Outpatients were observed for immediate complications in the recovery room for 2 hours and followed up by a telephone conversation the day after the procedure to monitor for possible complications. Histologic Examinations Tissue samples were fixed in formalin and embedded in paraffin. A paraffin block then was cut into serial sections and stained with hematoxylin-eosin (H&E), elastica van Gieson, and an anti-igg4 antibody (Binding Site, Birmingham, UK). If the size of the sampled tissue was smaller than 5 high-power fields (HPFs, 400) on H&E staining, the samples were excluded from the histologic analysis. Histologic analysis included the evaluation of the following: (1) storiform fibrosis with lymphoplasmacytic infiltration on H&E staining, (2) lymphoplasmacytic infiltration around pancreatic ducts on H&E staining (Figure 1), (3) obliterative phlebitis on elastica van Gieson staining (Figure 2), (4) abundant IgG4-positive plasma cells on IgG4 immunostaining (Figure 3), and (5) granulocytic epithelial Figure 2. Obstruction of a vein by the infiltration of inflammatory cells (elastica van Gieson stain; original magnification, 20).

3 318 IWASHITA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 3 Data Analysis The primary end point was the diagnostic yield of EUS-FNA using a 19-gauge needle for the histologic diagnosis of AIP. The secondary end points were to evaluate the rate of complications encountered in this procedure and to assess the predictive factors for positive histologic diagnosis of AIP through a comparison between patients with a histologic diagnosis of AIP and those with a histologic diagnosis of ICP. Continuous variables are presented as the median and range. We calculated sensitivity and the 95% confidence interval. Intergroup comparisons were performed with the Fisher exact test for categoric variables and the Wilcoxon rank-sum test for continuous variables. All tests were 2-tailed and a P value of less than.05 was considered to indicate a statistically significant difference. JMP software version (SAS Institute, Inc, Cary, NC) was used for all statistical analyses. Figure 3. Abundant IgG4-positive plasma cells ( 10 cells/hpf) with fibrosis, lymphocytes, and negatively stained plasma cells (IgG4 immunostaining; original magnification, 40). lesions (GELs) on H&E staining. Obliterative phlebitis was defined as an area enclosed by elastic fibers covering more than half of the circumference and obliterated with inflammatory cells on elastica van Gieson staining. The number of infiltrated IgG4-positive plasma cells was evaluated by IgG4 immunostaining in 1 HPF. The presence of more than 10 positive cells per HPF was considered as extensive IgG4-positive plasma cell infiltration. All slides were reviewed by a single pathologist (Y.H.) who was completely blinded to any clinical information and was asked to fill out our pathologic data sheet on the basis of the histologic findings. The pathologist was given no indication of the kinds of studies that we were conducting. The histologic diagnosis of AIP was made on the basis of the presence of LPSP. In this study, LPSP was diagnosed when storiform fibrosis with lymphoplasmacytic infiltrates surrounding the pancreatic ducts and obliterative phlebitis were observed, which constituted histologic findings A patient with storiform fibrosis with lymphoplasmacytic infiltration and extensive IgG4-positive plasma cell infiltration also was considered as having AIP, which constituted histologic findings 1 4. Idiopathic duct-centric chronic pancreatitis (IDCP) was defined as the presence of storiform fibrosis with lymphoplasmacytic infiltration and GEL, which constituted histologic findings 1 5. Finally, a patient showing fibrosis only or storiform fibrosis with lymphoplasmacytic infiltration without sufficient histologic findings for a diagnosis of AIP or IDCP was considered to be idiopathic chronic pancreatitis (ICP) histologically. Results Baseline Characteristics We enrolled 7 women and 37 men with a median age of 64 years (range, y). The initial symptoms were jaundice in 18 patients, abdominal pain in 15 patients, cervical lymphadenopathy in 2 patients, and body weight loss in 2 patients. Although the remaining 7 patients had no clinical symptoms, pancreatic lesions were detected as incidental findings in 3 patients on abdominal ultrasound or computed tomography (CT) during a periodic health check-up and further work-up for bilateral hilar lymphadenopathy, hyperamylasemia, abnormal results on a zinc sulfate turbidity test, and cirrhosis in each of the 4 remaining patients. Abdominal CT scans showed swelling of the whole pancreas in 19 patients, and enlargement of the head of the pancreas in 5 patients, the body in 1 patient, and the tail in 2 patients. Furthermore, the head and body of the pancreas were swollen in 9 patients, the body and tail in 6 patients, and the head and tail in 2 patients. On endoscopic retrograde pancreatography or magnetic resonance pancreatography, narrowing of the entire MPD was observed in 15 patients, the head and body in 10 patients, the body and tail in 6 patients, the head and tail in 1 patient, the head in 8 patients, the body in 2 patients, and the tail in 2 patients. The serum antinuclear antibody test was positive (1:40 or more) in 18 patients (41%). The serum IgG level increased in 26 patients (57%; cut-off level, 1800 mg/dl; median, 1910 mg/dl; range, mg/dl), and the serum IgG4 level was increased in 39 patients (89%; cut-off level, 135 mg/dl; median, 322 mg/dl; range, mg/dl). None of the patients had either ulcerative colitis or Crohn s disease (Table 1). Findings of Endoscopic Ultrasound and Endoscopic Ultrasound Guided Fine-Needle Aspiration EUS clearly visualized pancreatic lesions as a hypoechoic and enlarged pancreatic parenchyma with scattered hyperechoic foci in all patients. A peripancreatic hypoechoic margin was detected in 28 patients (63%) (Figure 4). EUS-FNA was performed successfully in all patients using a 19-gauge needle for the pancreatic lesions. FNA of the head lesion was performed using the transduodenal approach in 2 patients, and FNA of the neck, body, and tail was performed using the transgastric approach in 3, 35, and 4 patients, respectively. In 3 of 5 patients whose CT scans showed head lesions, FNA of the neck was performed via the stomach because the lesion involved the neck of the pancreas. The median number of FNA passes was 2 (range, 1 3). A macroscopically evaluable material was obtained in all patients. Pathologic Assessments Cytologic assessment showed adequate specimens and no evidence of malignant or atypical cells in all 44 patients.

4 March 2012 EUS-FNA FOR HISTOLOGIC DIAGNOSIS OF AIP 319 Table 1. Baseline Characteristics of the Patients Sex, n Female:male 7:37 Age, y Median 64 Range Symptoms, n (%) Jaundice 18 (41) Abdominal pain or discomfort 15 (34) Cervical lymphadenopathy 2 (5) Body weight loss 2 (5) No symptoms 7 (16) Swollen part of the pancreas on CT, n (%) Entire pancreas 19 (43) Head and body 9 (20) Body and tail 6 (14) Head and tail 2 (5) Head 5 (11) Body 1 (2) Tail 2 (5) Narrowing part of the MPD on ERP or MRP, n (%) Entire pancreas 15 (34) Head and body 10 (23) Body and tail 6 (14) Head and tail 1 (1) Head 8 (18) Body 2 (5) Tail 2 (5) Antinuclear antibody, n Positive (1:40 or more) 18 (41%) Serum IgG level, mg/dl Median 1910 Range Increase (cut-off level, 1800), n (%) 26 (57) Serum IgG4 level, mg/dl Median 322 Range Increase (cut-off level, 135), n (%) 39 (89) ERP, endoscopic retrograde pancreatography; MRP, magnetic resonance pancreatography. Three patients were excluded from the histologic analyses because the amount of FNA specimen was microscopically smaller than 5 HPFs. Among the remaining 41 patients (93%), histopathologic examination revealed storiform fibrosis with infiltration of lymphocytes and plasma cells in 38 patients, but only dense fibrotic changes in 3 patients. Among those with storiform fibrosis with lymphoplasmacytic infiltration, a diagnosis of LPSP was made in 17 patients (39%) because the presence of lymphoplasmacytic infiltration around pancreatic ducts and obliterative phlebitis was detected in 23 and 21 patients, respectively, and positive findings for both were observed in the 17 patients. Extensive IgG4-positive plasma cell infiltration was observed in 5 patients (11%), including positive findings for LPSP in 3 patients. However, there were no findings indicative of neutrophilic granulocytes infiltration consistent with GEL. Therefore, 19 patients (43%; 95% confidence interval, 30% 58%) were given histologic diagnoses of AIP, and the remaining 22 patients (50%) were given histologic diagnoses of ICP (Figure 5). Patients with a histologic diagnosis of AIP did not differ from those with a histologic diagnosis of ICP with respect to age, sex, imaging, or serologic findings (Table 2). Complications There was one minor complication. One patient had epigastric pain immediately after EUS-FNA. However, blood tests and abdominal CT revealed no specific findings. The patient was monitored carefully and the symptoms resolved within a few hours. There were no other reported complications in any of the other patients. Discussion The histopathologic hallmark of AIP is known as LPSP, which is characterized by lymphoplasmacytic infiltration surrounding pancreatic ducts, swirling fibrosis around pancreatic ducts and veins (storiform fibrosis), and obliterative phlebitis Suda et al 11 compared the histologic findings of AIP with those of chronic alcoholic pancreatitis and chronic obstructive pancreatitis caused by pancreatic cancer. They reported the specific histologic characteristics of AIP to be a combination of lymphoplasmacytic infiltration, interlobular and intralobular fibrosis, acinar atrophy, obliterative phlebitis, and pancreatic duct involvement. Zamboni et al 22 also studied the resected specimens of 53 patients with AIP and reported periductal lymphoplasmacytic infiltration, fibrosis, and venulitis as characteristic findings of AIP. These specific histologic findings of AIP were thought to be indicative of LPSP. Another supporting pathologic finding of AIP is the extensive infiltration of IgG4-positive plasma cells. Hamano et al first reported that the serum concentrations of IgG4 were significantly and specifically increased in patients with AIP, 23 and extensive infiltration of IgG4-positive plasma cells in the pancreas of AIP patients subsequently was reported. 12 Surgically resected pancreatic specimens of patients with AIP showed significantly more IgG4-positive plasma cells than those of patients with pancreatic cancer or chronic pancreatitis. 24,25 Considering these results, AIP can be distinguished from other types of chronic pancreatitis on the basis of LPSP and extensive infiltration of IgG4-positive plasma cells. However, these histologic findings have been examined mainly in surgically resected pancreatic specimens. Whether a small amount of specimen Figure 4. A pancreatic lesion on EUS showing a diffuse hypoechoic and enlarged pancreatic parenchyma with scattered hyperechoic foci and a peripancreatic hypoechoic margin (arrowheads) (left, B-mode image; right, color Doppler mode image). EUS-FNA was performed successfully using a 19-gauge needle via the stomach.

5 320 IWASHITA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 3 Figure 5. Study profile of histologic assessment. obtained by a needle biopsy allows histologic diagnosis of AIP has not yet been elucidated. In this study, EUS-FNA with a 19-gauge conventional needle was performed in 44 AIP patients and adequate amounts of specimens for histologic analysis were obtained in 93% of the patients (41 of 44). However, a histologic diagnosis of AIP was made only in 43% of the patients (19 of 44). This relatively low yield of histologic diagnosis in the current study could be attributed to the limited amount of FNA specimens. Zamboni et al 22 reported that pancreatic biopsy specimens, including those of intraoperative wedge biopsy in 5 patients and fineneedle core biopsy in 4 patients, revealed adequate histologic findings for AIP diagnosis in only 2 of 9 patients, even though most of the surgically resected specimens showed sufficient findings to make a diagnosis. This fact suggests that the amount of specimen could affect the yield of histologic diagnosis in AIP patients. Another factor that could be associated with the low yield of histologic diagnosis is the patchy distribution of specific histologic findings of AIP. LPSP does not feature diffuse findings, but, rather, findings in specific areas such as around the pancreatic ducts and veins Moreover, with regard to the infiltration of IgG4-positive plasma cells, the density and distribution of positive cells were patchy. 26 Furthermore, different severities of the specific histologic findings of AIP, which may represent the different disease stages of AIP, were recognized in surgically resected specimens of AIP patients. 22 These factors could affect the detection of LPSP and the extensive infiltration ( 10/HPF) of IgG4-positive plasma cells in a limited amount of needle biopsy specimens and cause sampling errors during EUS-FNA. A comparison of the basic characteristics of the patients with a histologic diagnosis of AIP and those with a histologic diagnosis of ICP did not reveal any statistically significant differences in the present study. However, these comparisons also could be affected by possible sampling errors during EUS-FNA. A previously reported method for obtaining greater amounts of pancreatic specimens from AIP patients is to use a 19-gauge Table 2. Comparison Between Those With Histologic Diagnosis of AIP and ICP Histologic diagnosis of AIP (n 19) Histologic diagnosis of ICP (n 22) P value Age, y, median (range) 64 (49 79) 63.5 (36 79).56 Male sex, n (%) 17 (89) 17 (77).42 Symptoms Jaundice, n (%) 9 (47) 7 (32).35 Abdominal pain, n (%) 5 (26) 10 (45).33 Entire pancreas swollen on CT, n (%) 11 (58) 6 (27).06 Hypoechoic rim on EUS, n (%) 15 (79) 13 (59).20 Antinuclear antibody positive, n (%) 10 (53) 7 (32).22 Serum IgG level, mg/dl, median (range) 1910 ( ) ( ).75 Increased IgG level (cut-off 1800 mg/dl), n (%) 12 (63) 12 (55).75 Serum IgG4 level, mg/dl, median (range) 338 ( ) ( ).81 Increased IgG4 level (cut-off 135 mg/dl), n (%) 19 (100) 18 (82).11

6 March 2012 EUS-FNA FOR HISTOLOGIC DIAGNOSIS OF AIP 321 Tru-Cut biopsy needle (TCB) under EUS guidance. Levy et al 27,28 described the results of their retrospective study using EUS-TCB in patients with AIP. In one of their reports, the specimens obtained were adequate for histologic analysis in all 14 AIP patients and revealed diagnostic findings, that is, LPSP or abundant IgG4-positive plasma cells in 57% of the patients (8 of 14). 28 Mizuno et al 29 performed both EUS-TCB and EUS- FNA with a 22-gauge needle in 11 patients who were given a final diagnosis of AIP. They reported that pancreatic specimens obtained by EUS-TCB showed full-spectrum LPSP in 45% (5 of 11) of the patients. Although a TCB needle obtains the specimen on the tissue tray after cutting the tissue with only one sliding movement of the outer sheath, the FNA needle suctions the specimen inside the needle after multiple movements of the needle inside the lesion. Theoretically, these differences in the specimen collection method might cause more distortion in the FNA specimen than that obtained by a TCB needle. However, the diagnostic sensitivity (43%) of our technique was similar to that of previous studies using a TCB needle. In addition, the EUS-TCB has a technical limitation with regard to needle manipulation in the duodenum and in the approach to the pancreatic head because a puncture from the duodenum to the head of the pancreas usually requires an angulated scope position, endoscopic tip angulation, and the use of an elevator function, which make the passage of the TCB needle difficult. 30 In the earlier-described studies, a puncture was never made for the head lesion. 28,29 This is possibly the main drawback of using the TCB needle. On the other hand, in our study, EUS-FNA from the duodenum with a 19-gauge needle for the lesion in the pancreatic head was performed successfully in 2 patients. Histologic analyses of pancreatic specimens resected from patients with mass-forming chronic pancreatitis from Europe and the United States identified another unique histologic pattern called IDCP 10 or AIP with GEL. 22 The clinical profile of patients with IDCP was reported to be different from that of patients with LPSP in the following respects: patients with IDCP were younger, 22 showed an almost equal ratio among men and women, 22 more frequently showed complications of ulcerative colitis and Crohn s disease, 22 and showed a lower rate of positive results in serum IgG4 tests and other organ involvement. 31 In the current study, FNA specimens did not reveal any findings indicative of GEL or neutrophilic infiltration and the clinical profile of the cohort appeared to be different from that previously reported for IDCP. Furthermore, the prevalence of IDCP is rare in Japan. 32 We believe that this study did not include any IDCP patients. However, again, there may have been a sampling error because of the limited amount of needle biopsy specimens and the patchy distribution of GEL. Taking into consideration the results of previous studies on EUS-TCB and those of our study, the yield of EUS-guided needle biopsy for the histologic diagnosis of AIP was not high. It may be difficult to further increase the amount of specimens obtained because of the limited caliber of the needle. The histopathologic features of LPSP and IDCP have been studied comprehensively on the basis of surgically resected pancreatic specimens. We believe that further studies are required to evaluate the efficacy of EUS-guided needle biopsy of pancreatic lesions in the diagnosis of AIP and to determine how to interpret the histologic findings obtained from a limited amount of pancreatic specimen in a diagnostic algorithm for AIP. In our facility, when possible, a 19-gauge needle is used routinely as the primary needle and FNA usually is repeated for up to 3 passes until a whitish material is obtained. Only a single pass was performed in a few patients in which the FNA specimen immediately showed an adequate amount of whitish material; we subsequently considered such an FNA specimen to be adequate for evaluation. Although we do not have specific data, we previously found that the adequacy of the FNA specimen for histologic analysis can be confirmed by the presence of a macroscopically visible whitish material Therefore, the presence of a whitish material has been used as an indicator to decide the number of FNA passes necessary because on-site cytotechnicians are not available in our facility. We also observed that additional FNA passes usually did not yield a whitish material if none was obtained within the first 3 sufficient FNA passes. Considering our past experience, the specific conditions at our facility, and the risk-to-benefit ratio for patients, a default setting for the FNA needle and restrictions in the number of FNA passes were established as our practice norm. This study had several limitations. All EUS procedures were performed by a single endosonography specialist at a single center, which might affect the reproducibility of the procedures. Furthermore, the study might contain potential biases in patient selection and data, despite careful selection of true AIP patients and blinded histologic analysis, because of its retrospective nature. Further prospective study is needed to evaluate the accuracy of needle biopsy for the histologic diagnosis of AIP. Conclusions EUS-FNA with a 19-gauge needle was shown to be a safe and reliable procedure for obtaining pancreatic specimens. Although the yield of this technique for the histologic diagnosis of AIP was not high, the method still may be useful, particularly in suspected AIP patients lacking typical serologic or imaging findings, to avoid unnecessary surgical exploration. Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at and at doi: /j.cgh References 1. Yoshida K, Toki F, Takeuchi T, et al. Chronic pancreatitis caused by an autoimmune abnormality. Proposal of the concept of autoimmune pancreatitis. Dig Dis Sci 1995;40: Japan Pancreas Society. Diagnostic criteria for autoimmune pancreatitis by the Japan Pancreas Society. J Jpn Pancreas (Suizo) 2002;17: Pearson RK, Longnecker DS, Chari ST, et al. Controversies in clinical pancreatology: autoimmune pancreatitis: does it exist? Pancreas 2003;27: Chari ST, Smyrk TC, Levy MJ, et al. Diagnosis of autoimmune pancreatitis: the Mayo Clinic experience. Clin Gastroenterol Hepatol 2006;4: Kim KP, Kim MH, Kim JC, et al. Diagnostic criteria for autoimmune chronic pancreatitis revisited. World J Gastroenterol 2006; 12: Okazaki K, Kawa S, Kamisawa T, et al. Clinical diagnostic criteria

7 322 IWASHITA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 3 of autoimmune pancreatitis: revised proposal. J Gastroenterol 2006;41: Otsuki M, Chung JB, Okazaki K, et al. Asian diagnostic criteria for autoimmune pancreatitis: consensus of the Japan-Korea Symposium on Autoimmune Pancreatitis. J Gastroenterol 2008;43: Abraham SC, Wilentz RE, Yeo CJ, et al. Pancreaticoduodenectomy (Whipple resections) in patients without malignancy: are they all chronic pancreatitis? Am J Surg Pathol 2003;27: Kawaguchi K, Koike M, Tsuruta K, et al. Lymphoplasmacytic sclerosing pancreatitis with cholangitis: a variant of primary sclerosing cholangitis extensively involving pancreas. Hum Pathol 1991;22: Notohara K, Burgart LJ, Yadav D, et al. Idiopathic chronic pancreatitis with periductal lymphoplasmacytic infiltration: clinicopathologic features of 35 cases. Am J Surg Pathol 2003;27: Suda K, Takase M, Fukumura Y, et al. Histopathologic characteristics of autoimmune pancreatitis based on comparison with chronic pancreatitis. Pancreas 2005;30: Hamano H, Kawa S, Ochi Y, et al. Hydronephrosis associated with retroperitoneal fibrosis and sclerosing pancreatitis. Lancet 2002;359: Williams DB, Sahai AV, Aabakken L, et al. Endoscopic ultrasound guided fine needle aspiration biopsy: a large single centre experience. Gut 1999;44: 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: Volmar KE, Vollmer RT, Jowell PS, et al. Pancreatic FNA in 1000 cases: a comparison of imaging modalities. Gastrointest Endosc 2005;61: Imazu H, Uchiyama Y, Kakutani H, et al. A prospective comparison of EUS-guided FNA using 25-gauge and 22-gauge needles. Gastroenterol Res Pract 2009;2009: Wilson JL, Kalade A, Prasad S, et al. Diagnosis of solid pancreatic masses by endoscopic ultrasound-guided fine-needle aspiration. Int Med J 2009;39: Turner BG, Cizginer S, Agarwal D, et al. Diagnosis of pancreatic neoplasia with EUS and FNA: a report of accuracy. Gastrointest Endosc 2010;71: Yasuda I, Tsurumi H, Omar S, et al. Endoscopic ultrasoundguided fine-needle aspiration biopsy for lymphadenopathy of unknown origin. Endoscopy 2006;38: Iwashita T, Yasuda I, Doi S, et al. The yield of endoscopic ultrasound-guided fine needle aspiration for histological diagnosis in patients suspected of stage I sarcoidosis. Endoscopy 2008;40: Iwashita T, Yasuda I, Tsurumi H, et al. Endoscopic ultrasoundguided fine needle aspiration biopsy for splenic tumor: a case series. Endoscopy 2009;41: Zamboni G, Lüttges J, Capelli P, et al. Histopathological features of diagnostic and clinical relevance in autoimmune pancreatitis: a study on 53 resection specimens and 9 biopsy specimens. Virchows Arch 2004;445: Hamano H, Kawa S, Horiuchi A, et al. High serum IgG4 concentrations in patients with sclerosing pancreatitis. N Engl J Med 2001;344: Deshpande V, Chicano S, Finkelberg D, et al. Autoimmune pancreatitis: a systemic immune complex mediated disease. Am J Surg Pathol 2006;30: Zhang L, Notohara K, Levy MJ, et al. IgG4-positive plasma cell infiltration in the diagnosis of autoimmune pancreatitis. Mod Pathol 2007;20: Chu KE, Papouchado BG, Lane Z, et al. The role of Movat pentachrome stain and immunoglobulin G4 immunostaining in the diagnosis of autoimmune pancreatitis. Mod Pathol 2009;22: Levy MJ, Reddy RP, Wiersema MJ, et al. EUS-guided trucut biopsy in establishing autoimmune pancreatitis as the cause of obstructive jaundice. Gastrointest Endosc 2005;61: Levy MJ, Wiersema MJ, Chari ST. Chronic pancreatitis: focal pancreatitis or cancer? Is there a role for FNA/biopsy? Autoimmune pancreatitis. Endoscopy 2006;38(Suppl 1):S30 S Mizuno N, Bhatia V, Hosoda W, et al. Histological diagnosis of autoimmune pancreatitis using EUS-guided trucut biopsy: a comparison study with EUS-FNA. J Gastroenterol 2009;44: Itoi T, Itokawa F, Kurihara T, et al. Experimental endoscopy: objective evaluation of EUS needles. Gastrointest Endosc 2009; 69: Sah RP, Chari ST, Pannala R, et al. Differences in clinical profile and relapse rate of type 1 versus type 2 autoimmune pancreatitis. Gastroenterology 2010;139: ; quiz, e12 e Kamisawa T, Notohara K, Shimosegawa T. Two clinicopathologic subtypes of autoimmune pancreatitis: LPSP and IDCP. Gastroenterology 2010;139: Reprint requests Address requests for reprints to: Ichiro Yasuda, MD, PhD, First Department of Internal Medicine, Gifu University Hospital, 1-1 Yanagido, Gifu , Japan. yasudaic@aol.com; fax: (81) (0) Acknowledgments The authors are indebted to Mr Roderick J. Turner, Assistant Professor Edward F. Barroga, and Professor J. Patrick Barron, Chairman of the Department of International Medical Communications at Tokyo Medical University, for their editorial review of the English manuscript. Conflicts of interest This author discloses the following: Takao Itoi is a consultant and gives lectures for Olympus Medical Systems, Ltd, Japan. The remaining authors disclose no conflicts.