SYSTEMATIC REVIEWS AND META-ANALYSES

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2013;11:913 921 SYSTEMATIC REVIEWS AND META-ANALYSES Fasiha Kanwal, Section Editor Cyst Features and Risk of Malignancy in Intraductal Papillary Mucinous Neoplasms of the Pancreas: A Meta-Analysis NEERAJ ANAND,*,, KARTIK SAMPATH, and BECHIEN U. WU,*,, *Center for Pancreatic Care, Southern California Permanente Medical Group, Division of Gastroenterology, Department of Internal Medicine, Kaiser Permanente, Los Angeles Medical Center, Los Angeles, California This article has an accompanying continuing medical education activity on page e60. Learning Objectives At the end of this activity, the successful learner will be able to understand intraductal papillary mucinous neoplasms and the characteristics that represent a risk for malignancy. BACKGROUND & AIMS: International guidelines for the management of intraductal papillary mucinous neoplasms (IPMNs) of the pancreas recommend surgical resection of those with specific characteristics. We performed a meta-analysis to evaluate the risk of malignancy associated with each of these features of IPMNs. METHODS: We performed a comprehensive search of MEDLINE from January 1, 1996, to November 11, 2011, for studies that included any of the features mentioned in the consensus guidelines for surgical resection of main duct and branch duct IPMNs. Data were analyzed from 41 studies for the following features: cyst size greater than 3 cm, the presence of mural nodules, dilated main pancreatic duct, symptoms, and main duct vs branch duct IPMNs. Malignant IPMNs were defined as those with carcinoma in situ or more advanced histology. A separate meta-analysis was performed for each risk factor to calculate pooled odds ratios (ORs). A random-effects model was used, based on the assumption of variation among study populations. RESULTS: CONCLUSIONS: The risks of malignancy associated with individual cyst features were as follows: cyst size greater than 3 cm (OR, 62.4; 95% confidence interval [CI], 30.8 126.3), presence of a mural nodule (OR, 9.3; 95% CI, 5.3 16.1), dilatation of the main pancreatic duct (OR, 7.27; 95% CI, 3.0 17.4), and main vs branch duct IPMN (OR, 4.7; 95% CI, 3.3 6.9). There was a moderate level of heterogeneity among studies (I 2 range, 34 67). Based on a meta-analysis, cyst features proposed by the international guidelines for resection of IPMN were highly associated with malignancy. However, based on our findings, not all cyst features should be weighted equally when considering risk of malignancy; cyst size greater than 3 cm was associated most strongly with malignant IPMN. Keywords: Pancreatic Cancer; Pancreatic Tumor; Surgery; Treatment. Pancreatic cysts increasingly have been detected over the past several years through increased use of enhanced imaging modalities. 1 A primary concern related to these cystic lesions is the potential for malignant transformation. In particular, mucinous lesions have an increased potential for malignant transformation compared with serous lesions. The World Health Organization (WHO) classifies cystic mucin-producing pancreatic neoplasms into 2 distinct entities: intraductal papillary mucinous neoplasm (IPMN) and mucinous cystic neoplasm. 2,3 IPMNs are intraductal mucin-producing neoplasms of the pancreas that cause cystic dilation of the pancreatic duct. 2 They are divided further into main duct and branch duct subtypes. These lesions are thought to undergo transformation from adenoma to borderline neoplasms, and finally to carcinoma. 1 Management for IPMN, in particular the branch duct subtype, remains controversial. In 2006, a consensus meeting of the working group of the International Association of Pancreatology published guidelines (Sendai criteria) for the diagnosis and treatment of IPMN. 3 The group recommended Abbreviations used in this paper: CI, confidence interval; CT, computerized tomography; ERCP, endoscopic retrograde cholangiopancreatography; EUS, endoscopic ultrasound; IPMNs, intraductal papillary mucinous neoplasms; MPD, main pancreatic duct; MRI, magnetic resonance imaging; OR, odds ratio; WHO, World Health Organization. 2013 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2013.02.010

914 ANAND ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 11, No. 8 that surgical resection of IPMN should be performed if any of the following 4 factors are present: (1) a cyst greater than 3 cm in diameter, (2) mural nodules, (3) dilatation of the main pancreatic duct, and (4) symptoms attributable to the tumor. Recently, these recommendations were updated to reflect changes in expert opinion but still incorporate many of the same cyst features mentioned in the previous guidelines. 4 The aim of this study was to perform a quantitative metaanalysis to evaluate the risk associated with each of the cyst features highlighted by the International Consensus Guidelines for IPMN. Methods Search Strategy A literature search was conducted in PubMed including studies published between January 1, 1996 (publication of the WHO classification of mucinous cysts), and November 11, 2011, to identify potentially relevant articles. The following terms were used for the initial search: intraductal papillary mucinous neoplasms, mucinous cysts, pancreatic cyst, cystic neoplasm of the pancreas and IPMNs. In addition, a recursive search of the reference sections of selected studies, review articles, and practice guidelines was performed manually to identify other potentially relevant articles. Study Selection Criteria Abstracts of articles from the literature search were evaluated individually for possible inclusion. Complete texts were obtained for articles that were potentially relevant. Studies meeting the following criteria were included: (1) English language articles; (2) full-article publication; (3) publication years from 1996 to 2011; (4) study design of prospective cohorts, retrospective cohorts, and case series; and study population of patients with IPMNs confirmed by histology; (5) preoperative imaging with either computed tomography, magnetic resonance imaging, or endoscopic ultrasound; (6) histopathology of cysts or long-term follow-up evaluation (at least 1 year clinical follow-up period); and (7) sufficient data available to calculate an odds ratio (study metameter) for malignancy for at least one of the prespecified cyst features. Exclusion criteria. Editorials, review articles, duplicate publications, and case reports were excluded. Data Abstraction and Analysis Eligible articles were reviewed independently by the investigators. Agreement between the investigators was greater than 95% for selection of articles to include in this review. Disagreement was resolved by consensus opinion among the study investigators. Data were extracted based on an a priori abstraction sheet to include the following: (1) patient demographics, (2) frequency of IPMN, (3) preoperative imaging, (4) risk factors associated with malignancy as outlined in the Sendai criteria, (5) histopathology of surgically resected cysts, and (6) clinical follow-up data. The WHO histologic classification divides IPMNs into adenoma, borderline IPMNs, carcinoma in situ, and invasive IPMN. 6 In this study, malignant IPMN was defined as carcinoma in situ and invasive carcinomas. Adenomas and borderline dysplastic IPMNs were classified as nonmalignant lesions. A separate meta-analysis was performed for each risk factor to calculate a pooled odds ratio with 95% confidence limits. A random-effects model was chosen for study analysis based on the assumption of variation among study populations. Study heterogeneity was assessed using the I 2 method. Subgroup analysis was performed by study region. Analysis was performed using Comprehensive Meta-Analysis version 2.2 (Biostat, Englewood, NJ). All reported P values were 2-sided with an value of.05. Evaluation of cyst features. Because surgical resection currently is recommended for all main duct IPMNs, analyses of cyst features such as cyst size, dilated main pancreatic duct, and the presence of mural nodules included only branch duct IPMNs. Comparison of main duct with branch duct IPMNs then was performed as a separate analysis (see the Main Duct Versus Branch Duct Intraductal Papillary Mucinous Neoplasms section later). Cyst size. We analyzed the relationship between cyst size and risk of malignancy. We evaluated studies that included cyst size characterized by cross-sectional imaging (computerized tomography [CT] and magnetic resonance imaging [MRI]), abdominal ultrasound, and endoscopic ultrasound. For each study, the method for determining cyst size was evaluated. For standardization, only studies that used a cut-off value of 3 cm or greater (in accordance with the Sendai guidelines) were included. Those studies that used a separate size threshold were excluded. Mural nodule. We analyzed the relationship between the presence of mural nodules and malignancy. We also sought to determine which imaging modalities were used commonly to assess for the presence of mural nodules. We sought to determine how studies defined mural nodules and what imaging modalities or combination of modalities appeared to be best at identifying mural nodules. Main duct versus branch duct intraductal papillary mucinous neoplasms. We evaluated the risk of malignancy associated with main duct vs branch duct or mixed-type IPMNs. We also analyzed studies to determine criteria for preoperative classification of cysts as main duct or branch duct type. Cysts were determined to be main duct, branch duct, and combined-type or mixed-type IPMNs based on preoperative imaging, endoscopic ultrasound (EUS), and endoscopic retrograde cholangiopancreatography (ERCP). We also sought to determine criteria for mixed- or combined-type IPMNs. We searched for studies that assessed pancreatic duct dilatation based on multiple imaging modalities including EUS and ERCP and provided pathologic and/or long-term follow-up evaluation to identify malignant lesions. In studies of branch duct IPMN, we included studies that used a threshold of 6 mm in duct size and also searched for the radiologic method used to assess duct size. Studies using a different duct size cut-off value were excluded. Symptoms. We analyzed studies to determine the relationship between the presence of symptoms and risk of malignancy. We incorporated studies that provided a description of which patients were classified as symptomatic and asymptomatic. We sought to further distinguish which symptoms were attributed to cysts in the published literature. All studies that provided data on individual symptoms and the associated risk of malignancy were included in the pooled analysis. Regional differences (subgroup analysis). We analyzed studies to determine whether there were regional differ-

August 2013 IPMNS OF THE PANCREAS 915 study by Kang et al, 38 cyst size was defined as an average of the major and minor axis diameters on axial images. In the study by Takeshita et al, 22 2 radiologists reviewed all imaging and decided on all cyst characteristics including cyst size, but further details were not provided. Figure 1. Flow chart of literature search and study selection. ences in the strength of association for various cyst features and risk of malignancy. We performed a subgroup analysis in which studies were grouped according to study origin. Specifically, we compared pooled odds ratios between Asian studies and studies conducted in Western populations. Results Our initial search yielded 1047 studies, of which 526 abstracts were reviewed based on relevance. Of these, 188 articles were fully reviewed and 41 met the prespecified inclusion criteria as outlined in Figure 1. Overall, there were 5788 patients and 3304 branch duct IPMNS included in the final analysis (Table 1). There were 19 (47.5%) studies from Japan, 5 22 10 (25.0%) studies from the United States, 23 32 7 (20.0%) studies from Korea, 33 39 2 (5.0%) studies from France, 40,41 and 1 study from both Italy 42 and China. 43 Twenty-nine (72.5%) of the studies included were published after the release of the Sendai consensus guidelines. 5 7,9,11 17,19,22 26,28,29,31 33,35,37 39,42,43 Twentyfive (62.5%) of the studies used in this analysis were published between 2006 and 2010. 7,11 14,16,17,19,22 26,28,29,31 33,37,39,42,43 In addition, all studies used cross-sectional imaging in the evaluation of cysts. There were 29 (72.5%) studies that used EUS and ERCP to evaluate cysts; in all of the studies these modalities were used only for selected patients. 8,17,18,21,30,33,34,36,41,42 Cyst Size Sixteen studies 5,8,9,11,15,17,20,22,28,31,32,38,39,41,42,44 including 1058 patients were incorporated in the analysis of cyst size. Cyst size greater than 3 cm substantially increased the risk of malignancy, the pooled odds ratios (ORs) was 62.4 (95% confidence interval [CI], 30.8 126.3) (Figure 2). The I 2 test for heterogeneity was 67. A detailed method of determining cyst size was included in the methods section of only 4 (10.0%) studies. 9,22,23,38 In studies that described the method of cyst measurement, various methods were used. In the study by Maguchi et al 9 and Ferrone et al 23 cyst size was determined by greatest diameter on cross-sectional imaging studies. In the Mural Nodule Nineteen studies 5,6,9 11,15,18 20,22,28,31,33,35,37 39,42,44 including 1452 patients were incorporated in the analysis of mural nodules. The presence of a mural nodule within the cyst was also a strong risk factor for malignancy (OR, 9.3; 95% CI, 5.3 16.1) (Figure 3). The I 2 was 55.7. The presence of mural nodules was determined by cross-sectional imaging and/or endoscopic ultrasound. Two of the 19 studies (11%) included additional information on how mural nodules were characterized: Wakabayashi et al 44 defined any protuberance seen on imaging as a mural nodule but Jang et al 37 defined mural nodules as papillary projections greater than 2 mm. Pancreatic Duct Dilatation Eight studies 5,9,11,17,18,22,42,44 including 358 patients were incorporated in the analysis of main pancreatic duct dilation. A dilated main pancreatic duct greater than 6 mm was associated with increased risk, with a pooled OR of 7.27 (95% CI, 3.0 17.4) for malignancy (Figure 4). Pancreatic main duct dilation was determined by a combination of CT scan, MRI, ERCP, and EUS. Wakabayashi et al 44 and Maguchi et al 9 used the maximal caliber of the main pancreatic duct for measurement. The remainder of the studies did not explicitly describe their method for ascertainment of main duct measurement. Main Duct Versus Branch Duct Intraductal Papillary Mucinous Neoplasms Twenty-nine studies 7,8,10,12 16,19 25,27 31,33,34,36,40,41,43,44 were included in the analysis of main duct IPMNs vs side branch IPMNs and the risk of malignancy. In a pooled analysis, main duct IPMN was associated with significantly increased risk of malignancy compared with side branch lesions (pooled OR, 4.7; 95% CI, 3.3 6.9) (Figure 5). All studies used the WHO pathologic classification to define IPMNs postoperatively. The following studies further classified IPMNs into mixed-type or combined-type IPMNs, which included IPMNs with both main duct and branch duct involvement: Nagai et al 11 defined mixed-type lesions as having a combination of main duct and branch duct involvement with predominantly branch duct involvement; Choi et al 34 identified mixed-type IPMNs when the pancreas contained more than one cyst and the diameter of the dilated MPD was more than 5 mm; Sainani et al 29 stated that cysts were classified as branch duct vs main duct IPMNs based on published criteria and independent review by 2 radiologists but without further explanation; Ohno et al 14 classified cysts as main duct IPMNs when a mural nodule was detected by EUS in the dilated MPD. Branch duct IPMNs were defined as an IPMN with a dilated branch and mixed IPMNs were defined as a cystic lesion in the branch duct in combination with a mural nodule continuous with a dilated MPD. Symptoms Thirteen studies 5,8,11,15,17,23,26,28,31,35,38 were incorporated in the analysis of symptoms as a risk for malignancy. There was

916 ANAND ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 11, No. 8 Table 1. Studies Included in the Meta-Analysis Study name Study country Study year Age, y Females Patients, n BD IPMN, n Patients resected, n Fukushima et al 8 Japan 1997 64.9 9 38 11 38 Sugiyama and Atomi 20 Japan 1998 64 14 41 16 41 Paye et al 40 France 2000 60 20 41 12 41 Terris et al 41 France 2000 55 20 43 13 43 Wakabayashi et al 44 Japan 2001 69 21 58 26 35 Choi et al 34 Korea 2003 61 14 46 12 46 Matsumo et al 10 Japan 2003 66 11 54 16 43 Sai et al 18 Japan 2003 68 9 33 33 5 Salvia et al 30 United States 2004 65 69 140 0 140 Suzuki et al 21 Japan 2004 67 469 1558 509 1024 Carbognin et al 42 Italy 2005 63 12 65 65 29 Jang et al 36 Korea 2005 61 61 208 79 208 Chiu et al 33 Korea 2006 60 16 56 22 56 Okabayashi et al 16 Japan 2006 67 7 23 8 23 Serikawa et al 19 Japan 2006 67 30 120 68 103 Fujino et al 7 Japan 2007 65.5 25 64 29 64 Lee et al 24 United States 2007 63 99 166 10 166 Nakagohri et al 13 Japan 2007 65 27 82 61 82 Pais et al 25 United States 2007 65 36 74 19 74 Pelaez-Luna et al 26 United States 2007 65 84 147 147 77 Rodriguez et al 28 United States 2007 67 83 145 145 145 Schmidt et al 31 United States 2007 64 79 150 103 150 Jang et al 37 Korea 2008 60 51 138 138 138 Nagai et al 12 Japan 2008 63 28 72 49 72 Takeshita et al 22 Japan 2008 65 25 53 46 53 Tang et al 32 United States 2008 66 32 204 31 204 Ferrone et al 23 United States 2009 62 244.6 401 185 256 Nagai et al 11 Japan 2009 63 36 105 84 84 Ohno et al 14 Japan 2009. 34 87 48 87 Sainani et al 29 United States 2009 68 17 30 14 30 Tan et al 43 China 2009 62 9 20 5 20 Woo et al 39 Korea 2009 63 35 85 85 85 Akita et al 5 Japan 2010 62.6 18 38 38 46 Arikawa et al 6 Japan 2010 65.2 25 25 25 Pitman et al 27 United States 2010. 28 81 55 112 Sadakari et al 17 Japan 2010 66 25 73 73 73 Hwang et al 35 Korea 2011 63.1 100 237 237 237 Kang et al 38 Korea 2011 62.6 90 201 201 35 Maguchi et al 9 Japan 2011 66 170 349 349 29 Moriya et al 49 United States 2011 66 210 210 Ohtsuka et al 15 Japan 2011. 39 99 99 99 NOTE. Forty-one studies were included from the following countries: 19 studies were from Japan, 11 studies were from the United States, 7 studies were from Korea, 2 studies were from France, and 1 study each from both Italy and China. a weak association between patient symptoms and malignancy (OR, 1.6; 95% CI, 1.0 2.6). The I 2 was 34 (Figure 6). Only 2 studies specified what symptoms were attributed to cysts. 11,23 Ferrone et al 23 classified cysts as asymptomatic if they were detected on imaging for nonspecific abdominal pain or symptoms of nonpancreatic origin. Nagai et al 11 attributed patients symptoms to cysts if they presented with jaundice, abdominal pain, back pain, fever, weight loss, pancreatitis, or recent-onset diabetes. The remainder of the studies included in this analysis did not specify which symptoms were attributed to the IPMN. International/Regional Differences For evaluation of regional differences the studies were divided into Asian or non-asian origin. There were 27 (67.5%) studies 5 22,33 39,43,44 from Asian countries and 13 (32.5%) studies 23 32,40 42 from non-asian countries; 11 (25.0%) studies in the non-asian group were from the United States. Overall, there was no difference in cyst features and risk of malignancy in studies originating from Asia and studies originating from Western countries (data not shown). Discussion In this meta-analysis we confirmed that the cyst features proposed by the International Sendai consensus guidelines for resection of IPMN initially in 2006 and updated in 2012 are associated with an increased risk of malignancy. However, the present study findings suggest that not all cyst features carry a similar risk of malignancy. In particular, cyst size was associated most strongly with malignant IPMN in the present meta-analysis. In addition, the presence of mural nodules had a

August 2013 IPMNS OF THE PANCREAS 917 Figure 2. Forest plot for cyst size greater than 3 cm and risk of malignancy. Sixteen studies were included (pooled OR, 62.4; 95% CI, 30.8 126.3). significant impact on risk of malignancy. This is important because objective findings appeared to be more reliable predictors of malignancy compared with subjective features such as the presence of symptoms. Pancreatic cystic neoplasms likely are being discovered more frequently with the use of cross-sectional imaging. Lee et al 45 found the prevalence of incidental pancreatic cysts to be 13.5% in 616 consecutive patients who had undergone MRIs. This was higher than in a study published in 2008 that examined 2832 consecutive CT scans performed in adults without a history of pancreatic lesions or factors predisposing to pancreatic disease and found 73 (2.6%) patients with pancreatic cysts. 46 The disparity between these 2 studies may be related to differences in imaging techniques used as well as the study populations. Figure 3. Forest plot for mural nodules and risk of malignancy. Nineteen studies were included (pooled OR, 9.3; 95% CI, 5.3 16.1).

918 ANAND ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 11, No. 8 Figure 4. Forest plot for pancreatic ductal dilatation and risk of malignancy. Eight studies were included (pooled OR, 7.27; 95% CI, 3.0 17.4). A subset of pancreatic cystic neoplasms may undergo malignant transformation. As a result, guidelines for the management of suspected mucinous lesions were proposed in 2006 based on the literature available at that time and expert opinion. 3 Surgical resection was recommended for all main duct IPMNs and for branch duct IPMNs that were greater than 3 cm, symptomatic, had the presence of a mural nodule, or were associated with main pancreatic duct dilatation. These guidelines were updated recently. 4 Resection still is recommended for all main duct IPMNs if the patient is a surgical candidate but the most recent guidelines included an algorithm for the treatment of branch duct IPMNs. 4 In its most recent iteration the consensus guideline indications for resection of branch duct IPMNs are based on the presence of worrisome features and high-risk stigmata. The group considered high-risk stigmata to include main pancreatic duct (MPD) Figure 5. Forest plot for risk of malignancy in main duct vs branch duct IPMN. Twenty-nine studies were included (pooled OR, 4.7; 95% CI, 3.3 6.9).

August 2013 IPMNS OF THE PANCREAS 919 Figure 6. Forest plot for symptoms and risk of malignancy. Thirteen studies were included (pooled OR, 1.6; 95% CI, 1.0 2.6). greater than 10 mm and an enhanced solid component. Worrisome features were considered cysts 3 cm or greater, thickened enhanced cyst walls, nonenhanced mural nodules, MPD size of 5 to 9 mm, or an abrupt change in MPD caliber with distal pancreatic atrophy and lymphadenopathy. In patients with worrisome features, endoscopic ultrasound with fine-needle aspiration has been suggested by other groups as well. 47 Overall, the revised Sendai guidelines de-emphasize the role of cyst size. However, findings from the present study suggest that cyst size greater than 3 cm might be considered a high-risk feature based on the significant increased risk of malignant transformation noted. It is possible that this discrepancy is because the recent consensus guidelines were based to a large extent on expert opinion. Systemic reviews such as the present study allow for objective synthesis of the available literature and are based on a systematic careful identification, appraisal, and integration of the published literature to make conclusions. 48 In this analysis of risk factors for malignant transformation, symptoms appeared to be the least significant marker of malignancy. This is likely because it is difficult to ascertain which symptoms can be attributed to a cyst. It is possible that certain symptoms such as jaundice or pancreatitis would be a better indicator of malignancy risk when compared with symptoms such as nausea or abdominal pain (as suggested by the updated consensus guidelines). Determining the predictive value of certain symptoms in comparison with others was beyond the scope of this analysis and unable to be performed based on the available literature. Future studies examining the relationship between particular symptoms and malignant IPMNs would be useful. There were several limitations to the present study. The present meta-analysis was subject to significant selection bias because the majority of data were obtained from surgical case series. We also were limited in our ability to determine which methods were used to determine specific cyst features. Many of the studies included in this meta-analysis did not detail the method by which cyst size, pancreatic duct dilatation, and presence of mural nodules were determined. This may have been because the studies were largely retrospective and measurements were based on radiographic imaging reports. In addition, several studies were excluded based on lack of consistent data presentation. In particular, several studies used size thresholds other than those suggested by the Sendai criteria. An additional limitation was the inability to evaluate the relationship between IPMN histologic subtype (gastric, intestinal, pancreatobiliary, oncocytic) and risk of malignancy based on lack of sufficient studies for meta-analysis. Finally, it is important to note that there was moderate heterogeneity among the studies in each of the analyses potentially related to variation in study population or lack of standardization in methods used to determine cyst features. Based on findings from the present meta-analysis, we have several recommendations for future research. First, adoption of standardized reporting guidelines for evaluation of cyst features in patients with cystic neoplasms would help evaluate the quality of future studies. The methods of determining cyst features should be standardized including the methods for measurement of cyst size and determining the presence of mural nodules. Second, the analysis of specific symptoms and their association with malignancy risk should be ascertained to aid in the surveillance of worrisome cysts. Third, because the vast majority of studies included in this meta-analysis were from surgical case series, longitudinal data from patients with IPMN who have not undergone resection would provide a more accurate estimate of the true overall rate of malignant transformation in this patient population. Finally, additional studies are needed to evaluate the additive role that multiple risk factors may contribute to the risk of malignant transformation. In summary, in this meta-analysis we determined that cyst size, presence of mural nodule, and main duct involvement each were associated with increased risk of malignancy in IPMN of the pancreas. However, our findings suggest that not all fea-

920 ANAND ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 11, No. 8 tures carry the same level of risk. In particular, cyst size may need to play a larger role in the decision of when to pursue surgical resection of side-branch IPMN. References 1. Kimura W, Nagai H, Kuroda A, et al. Analysis of small cystic lesions of the pancreas. Int J Pancreatol 1995;18:197 206. 2. Klöppel G. Histological typing of tumours of the exocrine pancreas. Heidelberg, Germany: Springer Verlag, 1996. 3. Tanaka M, Chari S, Adsay V, et al. International consensus guidelines for management of intraductal papillary mucinous neoplasms and mucinous cystic neoplasms of the pancreas. Pancreatology 2006;6:17 32. 4. Tanaka M, Fernández-del Castillo C, Adsay V, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012;12:183 197. 5. Akita H, Takeda Y, Hoshino H, et al. Mural nodule in branch duct-type intraductal papillary mucinous neoplasms of the pancreas is a marker of malignant transformation and indication for surgery. Am J Surg 2011;202:214 219. 6. Arikawa S, Uchida M, Uozumi J, et al. Utility of multidetector row CT in diagnosing branch duct IPMNS of the pancreas compared with MR cholangiopancreatography and endoscopic ultrasonography. Kurume Med J 2011;57:91 100. 7. Fujino Y, Matsumoto I, Ueda T, et al. Proposed new score predicting malignancy of intraductal papillary mucinous neoplasms of the pancreas. Am J Surg 2007;194:304 307. 8. 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;28:1010 1017. 9. Maguchi H, Tanno S, Mizuno N, et al. Natural history of branch duct intraductal papillary mucinous neoplasms of the pancreas: a multicenter study in Japan. Pancreas 2011;40:364 370. 10. Matsumoto T, Aramaki M, Yada K, et al. Optimal management of the branch duct type intraductal papillary mucinous neoplasms of the pancreas. J Clin Gastroenterol 2003;36:261 265. 11. Nagai K, Doi R, Ito T, et al. Single-institution validation of the international consensus guidelines for treatment of branch duct intraductal papillary mucinous neoplasms of the pancreas. J Hepatobiliary Pancreat Surg 2009;16:353 358. 12. Nagai K, Doi R, Kida A, et al. Intraductal papillary mucinous neoplasms of the pancreas: clinicopathologic characteristics and long-term follow-up after resection. World J Surg 2008;32:271 278; discussion 279 280. 13. Nakagohri T, Kinoshita T, Konishi M, et al. Surgical outcome of intraductal papillary mucinous neoplasms of the pancreas. Ann Surg Oncol 2007;14:3174 3180. 14. Ohno E, Hirooka Y, Itoh A, et al. Intraductal papillary mucinous neoplasms of the pancreas: differentiation of malignant and benign tumors by endoscopic ultrasound findings of mural nodules. Ann Surg 2009;249:628 634. 15. Ohtsuka T, Kono H, Nagayoshi Y, et al. An increase in the number of predictive factors augments the likelihood of malignancy in branch duct intraductal papillary mucinous neoplasm of the pancreas. Surgery 2012;151:76 83. 16. Okabayashi T, Kobayashi M, Nishimori I, et al. Clinicopathological features and medical management of intraductal papillary mucinous neoplasms. J Gastroenterol Hepatol 2006;21: 462 467. 17. Sadakari Y, Ienaga J, Kobayashi K, et al. Cyst size indicates malignant transformation in branch duct intraductal papillary mucinous neoplasm of the pancreas without mural nodules. Pancreas 2010;39:232 236. 18. Sai JK, Suyama M, Kubokawa Y, et al. Management of branch duct-type intraductal papillary mucinous tumor of the pancreas based on magnetic resonance imaging. Abdom Imaging 2003;28:694 699. 19. Serikawa M, Sasaki T, Fujimoto Y, et al. Management of intraductal papillary-mucinous neoplasm of the pancreas: treatment strategy based on morphologic classification. J Clin Gastroenterol 2006; 40:856 862. 20. Sugiyama M, Atomi Y. Intraductal papillary mucinous tumors of the pancreas: imaging studies and treatment strategies. Ann Surg 1998;228:685 691. 21. Suzuki Y, Atomi Y, Sugiyama M, et al. Cystic neoplasm of the pancreas: a Japanese multiinstitutional study of intraductal papillary mucinous tumor and mucinous cystic tumor. Pancreas 2004;28:241 246. 22. Takeshita K, Kutomi K, Takada K, et al. Differential diagnosis of benign or malignant intraductal papillary mucinous neoplasm of the pancreas by multidetector row helical computed tomography: evaluation of predictive factors by logistic regression analysis. J Comput Assist Tomogr 2008;32:191 197. 23. Ferrone CR, Correa-Gallego C, Warshaw AL, et al. Current trends in pancreatic cystic neoplasms. Arch Surg 2009;144:448 454. 24. Lee CJ, Scheiman J, Anderson MA, et al. Risk of malignancy in resected cystic tumors of the pancreas or 3 cm in size: is it safe to observe asymptomatic patients? A multi-institutional report. J Gastrointest Surg 2008;12:234 242. 25. Pais SA, Attasaranya S, Leblanc JK, et al. Role of endoscopic ultrasound in the diagnosis of intraductal papillary mucinous neoplasms: correlation with surgical histopathology. Clin Gastroenterol Hepatol 2007;5:489 495. 26. Pelaez-Luna M, Chari ST, Smyrk TC, et al. Do consensus indications for resection in branch duct intraductal papillary mucinous neoplasm predict malignancy? A study of 147 patients. Am J Gastroenterol 2007;102:1759 1764. 27. Pitman MB, Genevay M, Yaeger K, et al. High-grade atypical epithelial cells in pancreatic mucinous cysts are a more accurate predictor of malignancy than positive cytology. Cancer Cytopathol 2010;118:434 440. 28. Rodriguez JR, Salvia R, Crippa S, et al. Branch-duct intraductal papillary mucinous neoplasms: observations in 145 patients who underwent resection. Gastroenterology 2007;133:72 79; quiz 309 310. 29. Sainani NI, Saokar A, Deshpande V, et al. Comparative performance of MDCT and MRI with MR cholangiopancreatography in characterizing small pancreatic cysts. AJR Am J Roentgenol 2009;193:722 731. 30. Salvia R, Fernández-del Castillo C, Bassi C, et al. Main-duct intraductal papillary mucinous neoplasms of the pancreas: clinical predictors of malignancy and long-term survival following resection. Ann Surg 2004;239:678 685; discussion 685 687. 31. Schmidt CM, Yip-Schneider MT, Ralstin MC, et al. PGE(2) in pancreatic cyst fluid helps differentiate IPMN from MCN and predict IPMN dysplasia. J Gastrointest Surg 2008;12:243 249. 32. Tang RS, Weinberg B, Dawson DW, et al. Evaluation of the guidelines for management of pancreatic branch-duct intraductal papillary mucinous neoplasm. Clin Gastroenterol Hepatol 2008; 6:815 819; quiz 719. 33. Chiu SS, Lim JH, Lee WJ, et al. Intraductal papillary mucinous tumour of the pancreas: differentiation of malignancy and benignancy by CT. Clin Radiol 2006;61:776 783. 34. Choi BS, Kim TK, Kim AY, et al. Differential diagnosis of benign and malignant intraductal papillary mucinous tumors of the pancreas: MR cholangiopancreatography and MR angiography. Korean J Radiol 2003;4:157 162. 35. Hwang DW, Jang JY, Lim CS, et al. Determination of malignant and invasive predictors in branch duct type intraductal papillary mucinous neoplasms of the pancreas: a suggested scoring formula. J Korean Med Sci 2011;26:740 746. 36. Jang JY, Kim SW, Ahn YJ, et al. Multicenter analysis of clinico-

August 2013 IPMNS OF THE PANCREAS 921 pathologic features of intraductal papillary mucinous tumor of the pancreas: is it possible to predict the malignancy before surgery? Ann Surg Oncol 2005;12:124 132. 37. Jang JY, Kim SW, Lee SE, et al. Treatment guidelines for branch duct type intraductal papillary mucinous neoplasms of the pancreas: when can we operate or observe? Ann Surg Oncol 2008; 15:199 205. 38. Kang MJ, Jang JY, Kim SJ, et al. Cyst growth rate predicts malignancy in patients with branch duct intraductal papillary mucinous neoplasms. Clin Gastroenterol Hepatol 2011;9:87 93. 39. Woo SM, Ryu JK, Lee SH, et al. Branch duct intraductal papillary mucinous neoplasms in a retrospective series of 190 patients. Br J Surg 2009;96:405 411. 40. Paye F, Sauvanet A, Terris B, et al. Intraductal papillary mucinous tumors of the pancreas: pancreatic resections guided by preoperative morphological assessment and intraoperative frozen section examination. Surgery 2000;127:536 544. 41. Terris B, Ponsot P, Paye F, et al. Intraductal papillary mucinous tumors of the pancreas confined to secondary ducts show less aggressive pathologic features as compared with those involving the main pancreatic duct. Am J Surg Pathol 2000;24:1372 1377. 42. Carbognin G, Zamboni G, Pinali L, et al. Branch duct IPMTs: value of cross-sectional imaging in the assessment of biological behavior and follow-up. Abdom Imaging 2006;31:320 325. 43. Tan L, Zhao YE, Wang DB, et al. Imaging features of intraductal papillary mucinous neoplasms of the pancreas in multi-detector row computed tomography. World J Gastroenterol 2009; 15:4037 4043. 44. Wakabayashi T, Kawaura Y, Morimoto H, et al. Clinical management of intraductal papillary mucinous tumors of the pancreas based on imaging findings. Pancreas 2001;22:370 377. 45. Lee KS, Sekhar A, Rofsky NM, et al. Prevalence of incidental pancreatic cysts in the adult population on MR imaging. Am J Gastroenterol 2010;105:2079 2084. 46. Laffan TA, Horton KM, Klein AP, et al. Prevalence of unsuspected pancreatic cysts on MDCT. AJR Am J Roentgenol 2008;191:802 807. 47. Varadarajulu S, Fockens P, Hawes RH. Best practices in endoscopic ultrasound-guided fine-needle aspiration. Clin Gastroenterol Hepatol 2012;10:697 703. 48. Kanwal F, White D. Systematic Reviews and Meta-analyses in clinical gastroenterology and hepatology. Clin Gastroenterol Hepatol 2012;10:1184 1186. 49. Moriya T, Hashimoto Y, Traverso LW. The duration of symptoms predicts the presence of malignancy in 210 resected cases of pancreatic intraductal papillary mucinous neoplasms. J Gastrointest Surg 2011;15:762 770; discussion 770 771. Reprint requests Address requests for reprints to: Bechien U. Wu, MD, 1526 North Edgemont Street, 7th Floor, Los Angeles, California 90027. e-mail: bechien.u.wu@kp.org; fax: (323) 783-7056. Conflicts of interest The authors disclose no conflicts.