High Prevalence of Pancreatic Cysts Detected by Screening Magnetic Resonance Imaging Examinations

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2010;8:806 811 High Prevalence of Pancreatic Cysts Detected by Screening Magnetic Resonance Imaging Examinations KOEN DE JONG,* C. YUNG NIO, JOHN J. HERMANS, MARCEL G. DIJKGRAAF, DIRK J. GOUMA, CASPER H. J. VAN EIJCK, # EDDY VAN HEEL,** GUNTER KLASS, PAUL FOCKENS,* and MARCO J. BRUNO *Department of Gastroenterology and Hepatology, Department of Radiology, Department of Biostatistics and Clinical Epidemiology, and Department of Surgery, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands; Department of Radiology, # Department of Surgery, and Department of Gastroenterology and Hepatology, Erasmus Medical Center, Rotterdam, The Netherlands; **Prescan, Hengelo, The Netherlands; and Department of Radiology, Mathias Spital Hospital, Rheine, Germany BACKGROUND & AIMS: The prevalence of pancreatic cysts is not known, but asymptomatic pancreatic cysts are diagnosed with increasing frequency. We investigated the prevalence of pancreatic cysts in individuals who were screened by magnetic resonance imaging (MRI) as part of a preventive medical examination. METHODS: Data from consecutive persons who underwent abdominal MRI (n 2803; 1821 men; mean age, 51.1 10.8 y) at an institute of preventive medical care were included from a prospective database. All individuals had completed an application form including questions about possible abdominal complaints and prior surgery. MRI reports were reviewed for the presence of pancreatic cysts. Original image sets of all positive MRI reports and a representative sample of the negative series were re-assessed by a blinded, independent radiologist. RESULTS: Pancreatic cysts were reported in 66 persons (2.4%; 95% confidence interval, 1.9 3.0); prevalence correlated with increasing age (P.001). There was no difference in prevalence between sexes (P.769). There was no correlation between abdominal complaints and the presence of pancreatic cysts (P.542). Four cysts (6%) were larger than 2 cm and 3 (5%) were larger than 3 cm. Review of the original image sets by the independent radiologist did not significantly change these findings. CONCLUSIONS: The prevalence of pancreatic cysts in a large consecutive series of individuals who underwent an MRI at a preventive medical examination was 2.4%. Prevalence increased with age, but did not differ between sexes. Only a minority of cysts were larger than 2 cm. Keywords: Epidemiology; Radiologic Imaging; Pancreatic Cancer. Pancreatic cysts are diagnosed with increasing frequency because of the increased use of advanced cross-sectional imaging such as computed tomography (CT) and magnetic resonance imaging (MRI) for different indications. Pancreatic cysts may be discovered by coincidence. Consequently, physicians are confronted with all inherent diagnostic and therapeutic dilemmas. At the same time there is a trend for healthy asymptomatic individuals to undergo preventive health checkups, which may include a full-body MRI. Pancreatic cysts are a heterogeneous group of diagnostic entities that constitute a wide spectrum of congenital, inflammatory, and neoplastic lesions. 1 The former two are truly benign and do not require any further medical attention unless symptomatic, although some neoplastic cysts have malignant potential or may harbor frank malignancy. A majority of asymptomatic pancreatic cysts that are discovered incidentally are neoplasms, including intraductal papillary mucinous neoplasms and mucinous cystadenomas. 2 5 Making a diagnosis of pancreatic cysts is not straightforward and often is a clinical challenge. One would like to prevent patients with truly benign lesions from being exposed to unwarranted investigations and surgery. Simultaneously, patients should not be withheld a justifiable resection in the case of a (pre)malignant lesion. In light of these considerations it would be helpful to have a better understanding about the true prevalence of pancreatic cysts. Little is known about the prevalence of pancreatic cysts in the general population. Reports on the subject mostly have been conducted in selected groups of patients for a wide range of medical indications using various types of imaging modalities. These include transabdominal ultrasonography, CT, and MRI, which all have a different sensitivity for cyst detection. Hence, the reported prevalence spans a wide range: from 0.21% to 24.3%. 6 10 Ideally, to estimate the prevalence of pancreatic cysts one would like to perform the most sensitive type of imaging investigation for cyst detection (MRI) 11 in an unselected sample of the general population. These data are not likely to become available owing to the high costs of such a study. The aim of our study was to approximate the ideal situation by investigating the prevalence of pancreatic cysts in a group of individuals who underwent MRI as part of a preventive medical examination without medical indication, at their own initiative and costs. Materials and Methods Patients For this study we included consecutive individuals who had undergone an abdominal MRI in the Mathias Spital Hospital in Rheine, Germany, as part of a preventive medical examination. All examinations were arranged through Prescan (Hengelo, The Netherlands), a commercial agency offering preventive health check-ups, including MRI (www.prescan.nl). All individuals were Dutch, but the investigations were performed in a German hospital just across the Dutch-German border Abbreviations used in this paper: CI, confidence interval; CT, computed tomography; HASTE, half-fourier acquired single turbo spinecho; MRI, magnetic resonance imaging; US, ultrasound. 2010 by the AGA Institute 1542-3565/$36.00 doi:10.1016/j.cgh.2010.05.017

September 2010 PANCREATIC CYSTS DETECTED BY MRI 807 because in The Netherlands it is not legal to undergo preventive MRI investigation without a referral from a medical doctor. We included all examinations that were performed between December 2006 and September 2008. All individuals decided to undergo screening MRI at their own initiative and costs, either without having any complaint, or because of (nonspecific) complaints that did not prompt the individual to seek regular medical advice or for which further medical work-up was not indicated according to their general physician. All individuals were older than age 18. If a person had undergone multiple MRI scans of the abdomen during the designated study period, only the first scan was included in this study. All individuals had completed an application form before undergoing abdominal MRI that included items such as medical history, physical complaints, family medical history, prior surgery, and medication use. No specific questions about the family history of pancreatic cancer were asked. This application form was discussed with a medical doctor before the MRI investigation. After the investigation the results were discussed with the individuals and, when deemed indicated by the medical doctor, the advice was given to seek further medical assistance in The Netherlands. The study was approved by the institutional review board of the Mathias Spital Hospital and by Prescan. The requirement for informed consent was waived for this retrospective review of MRI reports. Because no informed consent was obtained we did not have the possibility to contact subjects to determine subsequent medical history after a pancreatic cyst was identified. Magnetic Resonance Imaging Technique All MRI examinations were performed in the Mathias Spital Hospital in Rheine, Germany, using a 1.5-T system (Avanto; Siemens, Erlangen, Germany). All patients underwent MRI before and after contrast medium administration with 0.1 mmol/kg gadopentetate dimeglumine (Magnevist; Bayer-Schering, Berlin, Germany). The following imaging sequences were acquired both coronal and axial: half-fourier acquired single turbo spin-echo (HASTE) (echo time, 90 ms; repetition time, 1000 ms; slice thickness, 8 mm) without fat saturation; HASTE (echo time, 383 ms; repetition time, 1040 ms; slice thickness, 8 mm) without fat saturation; T1 fast low angle shot MRI (Flash- 2D) before and after intravenous contrast, portal phase (echo time, 2,64 ms; repetition time, 124 ms; slice thickness, 8 mm) with fat saturation. End Points The primary end point was the prevalence of cystic lesions of the pancreas. Secondary end points were age distribution, sex distribution, pancreatic cyst characteristics, presence of liver or kidney cysts, presence of abdominal complaints, and medical history. Data Collection MRI data were originally read by radiologists at the Mathias Spital Hospital. The written reports of every MR investigation as well as the application forms were scrutinized by the investigators. All complete MRI reports were read by the investigators. Data regarding specific features of each individual (presence of abdominal complaints, age, sex, previous medical history, family history) and cyst features (size, location, multiplicity, simple vs complex, communication with pancreatic duct, simultaneous presence of kidney and/or liver cysts) were entered in a dedicated study database. For quality control, the original image sets of all positive MRI reports and a random sample of the negative series were re-assessed by an independent abdominal radiologist (Y.N.). MRI images were reviewed on site with the use of a picture archiving and communication system. Statistical Analysis Before the start of the study a power analysis was performed. With an estimated proportion of 0.018, a distance from proportion to limit of 0.005, and a confidence interval (CI) of 95%, the needed number of MRIs was 2717. Data were recorded in a Microsoft Office Access 2007 database (Microsoft Corporation, Redmond, WA). Statistical analysis was performed using SPSS 16.0 (SPSS Inc, Chicago, IL). Descriptive statistics were used to prepare overviews of relevant data regarding demographics as well as cyst characteristics. Quantitative data are presented by median values with range or mean values with standard deviations, depending on distributional properties. For the comparison of continuous variables, appropriate t tests or nonparametric tests were used. When appropriate, the 95% CI for the difference was calculated. For categoric data, the chi-square test for trend was used. P values less than.05 were considered statistically significant. A multivariable logistic regression model was used to analyze associations between the presence of pancreatic cysts (ie, dependent variable) and sex, age, abdominal complaints, abdominal surgery, liver cysts, and kidney cysts (ie, independent variables). Variables that were associated with the presence of pancreatic cysts using univariable analysis (P.1) were analyzed in the multivariable model. Odds ratios plus 95% CI were used to express the strength of the association. Results Retrospective Evaluation Between December 2006 and September 2008 a total of 2803 first-time MRI scans had been performed and these were all included in this study. Mean age at the time of MRI investigation was 51.1 years (standard deviation, 10.8 y) and 65% of individuals were men. Pancreatic cysts were identified in 66 individuals, representing a prevalence of 2.4% (95% CI, 1.8 3.0). Age was associated positively (odds ratio, 1.087; 95% CI, 1.060 1.113) with the presence of pancreatic cysts. In 443 individuals between ages 18 and 39 only one cyst (0.23%) was found (95% CI, 0.04% 01.27%). In the group between ages 40 and 49 cysts were found in 12 of 893 (1.3%) individuals (95% CI, 0.8% 2.3%), and 21 individuals had cysts in the group between ages 50 and 59 (823 individuals), representing a prevalence of 2.6% (95% CI, 1.7% 3.9%). In the group between ages 60 and 69, cysts were found in 20 of 551 (3.6%) persons (95% CI, 2.4% 5.6%), and 9 cysts (10.6%) were found in the group between ages 70 and 79 (95% CI, 5.7% 18.9%). Only 8 persons were older than 80 and the oldest was 87 years. In this group 3 individuals (37.5%) had pancreatic cysts (95% CI, 13.7% 69.4%). Findings are summarized in Figure 1. The mean age of persons with pancreatic cysts was 60.1 years (standard deviation, 10.9 y), whereas the mean age of persons without cysts was 50.9 years (standard deviation, 10.9 y) (P

808 DE JONG ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 9 Confirmatory Prospective Evaluation Original image sets of 200 MRI reports (66 positive reports and 134 negative reports) were re-assessed independently by an abdominal radiologist in random order. The radiologist was blinded to the results of the first evaluation. Reassessment of the 66 positive reports revealed no discrepancies between the first and the second reviewer. Three doubtful lesions were identified after re-assessment of the 134 negative reports by the second reviewer, and these findings were discussed with the first observer. In one case a dilated side branch was reported by the first observer, which was scored as a small cyst (4 mm) by the second observer. In 2 cases (4 and 6 mm) no agreement could be reached because it was considered that the presumed cysts also could represent small blood vessels. Figure 1. Prevalence of pancreatic cysts per age group. The black line shows the prevalence and the gray boxes represent the CI. For patients older than age 80, see the text for more details..001). In the group with a pancreatic cyst, 33.3% were women compared with 35.1% in the group without a cyst (P.769). In the group with pancreatic cysts 9% reported abdominal complaints on their pre-mri questionnaire compared with 7% in the group without cysts (P.524). None of the subjects in the study reported a history of acute or chronic pancreatitis. Of all individuals with pancreatic cysts 26% had abdominal surgery in their medical history, compared with 21% in the group without pancreatic cysts (P.354). Significantly more liver cysts were described in the group with pancreatic cysts (39.4% vs 25.0%; P.008). In the group with pancreatic cysts also more kidney cysts were identified (50.0% vs 35.1%; P.012). A detailed comparison between the 2 groups is provided in Table 1. Univariable and multivariable logistic regression was performed to analyze the association between the presence of pancreatic cysts and sex, age, abdominal complaints, abdominal surgery, liver cysts, and kidney cysts. In the univariable analysis, age, liver cysts, and kidney cysts were associated with the presence of pancreatic cysts. In the multivariable logistic regression analysis, age and the presence of liver cysts were shown to be associated with the presence of pancreatic cysts. Findings are provided in Table 2. Pancreatic cyst size varied from 2 to 54 mm (median, 8.0 mm). Thirteen cysts had a size between 1 and 2 cm, 4 had a size between 2 and 3 cm, and 3 cysts were larger than 3 cm. Cysts were distributed equally throughout the pancreas: head (23%), corpus (30%), tail (27%), and multifocal (12%), and in 8% of all cases location was not mentioned in the report. Six cysts (9%) were multilocular. These multilocular cysts were larger compared with the unilocular cysts with a median of 15 mm (range, 11 30 mm). Two examples of pancreatic cysts are shown in Figures 2 and 3. A connection with the pancreatic duct was reported in 8% of all cysts (Table 3). No features were reported that suggested malignant degeneration of a cyst such as the presence of an intracystic nodule or a thickened irregular cyst wall. Discussion Pancreatic cysts are discovered with increasing frequency because of a more liberal application of cross-sectional imaging techniques, not only for medical indications but also because apparently healthy individuals seek preventive medical consultation including whole-body MRI scanning. The apparent question is how to proceed after the detection of a pancreatic cyst in an asymptomatic individual. In clinical practice, most cysts are analyzed extensively, including the application of endoscopic ultrasonography with fine-needle aspiration, because a certain percentage is potentially neoplastic. 2 5 However, relatively little is known about the natural prevalence and progression of pancreatic cysts. The observation that clinicians are not frequently confronted with cystic malignancies gives some food for thought. It prompts the physician to construct a tailored application of differential diagnostic utilities, follow-up evaluation, and surgery with a trade-off of emotional distress, costs, and risks of medical intervention against preventing people from dying of a malignant degenerated cyst. To get a grip on this difficult consideration, a methodical assessment of the natural existence, occurrence, and development of pancreatic cysts is required. In this particular study we sought to determine the natural prevalence of pancreatic cysts in the general population. Table 1. Characteristics of Patients With and Without a Pancreatic Cyst Characteristics PC (n 66) PC (n 2737) P value Sex, male 44 (67) 1777 (65).769 Mean age, y 60 (10.9) 51 (10.9).001 (standard deviation) Abdominal complaints 6 (9) 193 (7).524 Abdominal surgery 17 (26) 576 (21).354 Smoking 16 (24) 709 (26).447 Pancreatic atrophy 6 (9) 75 (3).002 Liver cyst 26 (39) 685 (25).008 Kidney cyst 33 (50) 960 (35).012 Solid liver lesions 10 (15) 199 (7).016 Lymphadenopathy 1 (2) 33 (1).820 NOTE. Values are given as the number of cases with percentages in parentheses, except where indicated. PC, pancreatic cyst.

September 2010 PANCREATIC CYSTS DETECTED BY MRI 809 Table 2. Univariable and Multivariable Logistic Regression Analysis of Pancreatic Cyst Prevalence Characteristics Univariable analysis Multivariable analysis OR (95% CI) P OR (95% CI) P Sex 1.080 (0.644 1.813).770 Age, y 1.087 (1.060 1.113).001 1.083 (1.055 1.111).001 Abdominal complaints 1.358 (0.562 3.090).525 Abdominal surgery 1.302 (0.744 2.277).356 Liver cysts 1.947 (1.180 3.214).009 1.561 (0.938 2.599).087 Kidney cysts 1.851 (1.135 3.018).014 1.152 (0.692 1.919).587 OR, odds ratio. In our study in a group of 2803 individuals who decided to undergo abdominal MRI without referral from a medical doctor, pancreatic cysts were identified in 66 individuals, amounting to a prevalence of 2.4%. To date, only a few studies have been conducted that investigated the prevalence of pancreatic cysts. A recently published study by Laffan et al 8 reported a prevalence of 2.6%. In retrospect, 2832 consecutive CT scans were reviewed. Patients with known pancreatic disease or symptoms related to the pancreas were excluded. Other radiologic studies performed to date report prevalence rates ranging from 0.21% to 14.4%. 6 9 Ikeda et al 6 performed an ultrasound (US) study and reported that in 0.21% of 30,951 examinations a cystic lesion in the pancreas was recognized. This is the lowest reported prevalence rate to date and is probably easily explained because these US investigations were not performed with an explicit intent to investigate the pancreas and US has a lower sensitivity for (smaller) pancreatic cysts than MR, especially in the body and tail of the pancreas. 11 Moreover, US investigations were not videotaped, which precluded an independent review by a second observer with a focus on the pancreas. In an Italian study by Spinelli et al 7 reports of 24,039 MRI and CT scans were reviewed retrospectively with a computerized search. Pancreatic cysts were reported in 1.2%, of which 58% (0.7% of the total study population) did not have a history of pancreatitis. This study was hampered because a computerized search was performed with only a few specific search terms. We decided to review every single MRI report so as not to miss lesions because of an atypical description. Furthermore, no second independent review of the original image sets was performed in contrast to our study. The highest prevalence of pancreatic cysts using a radiologic imaging technique was found in a study by Zhang et al. 9 Spin-echo MR images of 1444 patients were reviewed for pancreatic cysts by 2 radiologists and pancreatic cysts were described in 19.6% of patients. MRIs were performed for various indications including pancreatic disorders. Of the 283 patients with pancreatic cysts 47 (16.6%) had a history of pancreatitis and 16 (5.7%) had a malignant pancreatic neoplasm, which provides some explanation for the high prevalence found in this cohort. In an autopsy study of 300 cases a stunning 24.3% were found to have pancreatic cysts. 10 It is of note that this study was performed in elderly patients ( 80% were 65 y), and no information was provided of a possible history of pancreatic disease. Similar to the autopsy study and the studies by Zhang et al 9 and Laffan et al, 8 increasing age strongly correlated with cyst presence in our study. 10 Only one cyst (0.23%) was identified in persons younger than 40 years of age, whereas 10.6% of all individuals between 70 and 80 years had pancreatic cysts. No difference in prevalence by sex was found, similar to the results of Zhang et al 9 and Laffan et al. 8 Neither the presence of abdominal complaints nor prior abdominal surgery correlated with the presence of pancreatic cysts in our study. In the other studies on the prevalence of pancreatic cysts no data were available on the presence of abdominal complaints or prior abdominal surgery. Figure 2. A 65-year-old woman with a small (3-mm) cyst in the tail of the pancreas. Figure 3. A 64-year-old woman with a multilocular cyst of 32 mm in the head of the pancreas compatible with a side-branch intraductal papillary mucinous neoplasm.

810 DE JONG ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 9 Table 3. Cyst Characteristics Characteristic Number (N 66) Median size, mm (range) 8 (2 54) Location Head 15 (23) Corpus 20 (30) Tail 18 (27) Multiple 8 (12) Not reported 5 (8) Multilocular 4 (6) Communication PD 5 (8) NOTE. Values are given as the number of cases with percentages in parentheses. PD, pancreatic duct. The median size of the cysts was 8 mm, and 6% of cysts were greater than 2 cm. These numbers are comparable with the studies using CT, MRI, and US. 6,8,9 None of the cysts larger than 2 cm showed additional morphologic features suggestive of malignancy. Multiple cysts were found in 12% of all individuals, comparable with the study by Laffan et al 8 in which 15% of all patients had more than one cyst. Other studies reported much higher rates of multiple cysts. In the study by Zhang et al, 9 almost half of the patients (44%) were reported to have multiple cysts whereas more than half of the patients in the autopsy study had more than one cyst. 10 In the remainder of individuals, cysts were distributed equally throughout the pancreas, comparable with the results of Laffan et al. 8 In the study by Zhang et al 9 most cysts were identified in the body of the pancreas, and in the autopsy study a majority of the cysts were found in the tail of the pancreas. 10 Cysts were reported to communicate with the pancreatic duct in 8% of cases, suggestive of branch-type intraductal papillary mucinous neoplasms. Zhang et al 9 found a possible communication with the pancreatic duct in 35% of all cysts. However, this study was performed in a hospital setting, and the Zhang et al speculate that their study population could have consisted of many patients with pancreatic disease, although they did not verify this in their study. Re-assessment of MRIs by an independent radiologist for quality control did not reveal relevant discrepancies from the original reports. Only 3 additional small lesions were detected during this second evaluation that had not been discovered during the first evaluation. The images were once again evaluated by the first observer and 1 of the 3 lesions, which was identified as a cyst by the second reviewer, was reported as a dilated side-branch during the first evaluation. In 2 other cases it proved impossible to distinguish between a true cyst and a blood vessel after the second evaluation and no consensus agreement was reached. In the absence of discrepancies regarding prevalence estimates between the first and confirmatory evaluations and considering the representative selection of negative reports from the first evaluation for re-assessment, no further re-assessment of the remaining negative reports from the first evaluation was deemed necessary. The ideal scenario would have been to review all MRIs, but this would have been a significant task and time investment. For practical reasons we decided to re-assess a random selection of the negative MRIs, which gives enough reassurance about a low to almost zero percent false-negative rate. All the studies described earlier were performed in patients within a hospital or out-patient setting, some series not explicitly excluding patients with potential pancreatic disease, using a range of different imaging techniques with various sensitivities for detecting cysts. As such, the cyst prevalence found in these patient cohorts may not be representative for the general population. Our study was strengthened because of the unique population of relatively unselected individuals who decided to undergo MRI of the abdomen at their own cost without a referral from their medical doctor. Examinations were performed in individuals who did not have any complaints (the vast majority of this series), or because of (nonspecific) complaints that did not prompt the individual to seek regular medical advice or in whom further medical examinations by their physicians were denied. We have to acknowledge that individuals in our study had financial means to drive to a hospital in Germany and pay for a medical health check-up. This constitutes a potential bias in the representativeness of our study population. However, we believe that this probably comes closest as one can possibly get to a general population sample. Moreover, in previous series cysts may have been missed either because a less-sensitive imaging technique was used or simply because cysts were overlooked because the pancreas was not the primary focus of attention of the radiologist. In our series MRIs also were not performed with a focus on the pancreas, but as part of a total body scan. Because this may have caused potential bias, a second assessment by an independent expert abdominal radiologist was performed of all the original image sets of individuals in whom a cyst was reported and a representative sample of individuals without cysts. In conclusion, the prevalence of pancreatic cysts in a relatively unselected group of individuals undergoing abdominal MRI at their own initiative and costs was 2.4%. Prevalence strongly increased with age and there was no difference in prevalence by sex. Only a small minority of cysts were larger than 2 cm, and none of the cysts showed features suggestive of malignant degeneration. However, several studies have shown that a majority of asymptomatic cysts are neoplasms including side-branch intraductal papillary mucinous neoplasms and that these lesions require follow-up evaluation. 2 5 Given the observation that pancreatic cysts are so common in the general population and detected with increased frequency because of the application of cross-sectional imaging techniques, there is no doubt that gastroenterologists will be consulted more often regarding how to manage these cysts. The next challenge is to define a tailored management approach deciding which cyst to remove, discard, or follow up with which technique at which interval, taking into consideration cyst characteristics but also the age and comorbidity of the patients. References 1. Brugge WR, Lauwers GY, Sahani D, et al. Cystic neoplasms of the pancreas. N Engl J Med 2004;351:1218 1226. 2. Fernandez-del CC, Targarona J, Thayer SP, et al. Incidental pancreatic cysts: clinicopathologic characteristics and comparison with symptomatic patients. Arch Surg 2003;138:427423. 3. 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;212:432 443.

September 2010 PANCREATIC CYSTS DETECTED BY MRI 811 4. Kiely JM, Nakeeb A, Komorowski RA, et al. Cystic pancreatic neoplasms: enucleate or resect? J Gastrointest Surg 2003;7:890 897. 5. Talamini MA, Pitt HA, Hruban RH, et al. Spectrum of cystic tumors of the pancreas. Am J Surg 1992;163:117 123. 6. Ikeda M, Sato T, Morozumi A, et al. Morphologic changes in the pancreas detected by screening ultrasonography in a mass survey, with special reference to main duct dilatation, cyst formation, and calcification. Pancreas 1994;9:508 512. 7. Spinelli KS, Fromwiller TE, Daniel RA, et al. Cystic pancreatic neoplasms: observe or operate. Ann Surg 2004;239:651 657. 8. Laffan TA, Horton KM, Klein AP, et al. Prevalence of unsuspected pancreatic cysts on MDCT. AJR Am J Roentgenol 2008;191:802 807. 9. Zhang XM, Mitchell DG, Dohke M, et al. Pancreatic cysts: depiction on single-shot fast spin-echo MR images. Radiology 2002; 223:547 553. 10. Kimura W, Nagai H, Kuroda A, et al. Analysis of small cystic lesions of the pancreas. Int J Pancreatol 1995;18:197 206. 11. Bassi C, Salvia R, Molinari E, et al. Management of 100 consecutive cases of pancreatic serous cystadenoma: wait for symptoms and see at imaging or vice versa? World J Surg 2003;27: 319 323. Reprint requests Address requests for reprints to: Marco J. Bruno, MD, PhD, Department of Gastroenterology and Hepatology, Erasmus Medical Center, 3015 CE Rotterdam, The Netherlands. e-mail: m.bruno@erasmusmc.nl; fax: (31) 107034682. Conflicts of interest The authors disclose no conflicts.