• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of annsurgLink to Publisher's site
Ann Surg. May 2004; 239(5): 651–659.
PMCID: PMC1356273

Cystic Pancreatic Neoplasms

Observe or Operate



The objectives of this analysis were to define the incidence, natural history, and predictors of neoplasia in pancreatic cysts to determine which patients can safely be observed and which should undergo an operation.

Summary Background Data:

With advancements in imaging technology, cystic lesions of the pancreas are being detected with increased frequency. Many of these lesions are small and asymptomatic, but they may be associated with pancreatitis or have malignant potential. Therefore, the management of these patients is complex, and knowledge of pancreatic cyst natural history and predictors of neoplasia are important.


From January 1995 through December 2002, all radiologic, surgical, and pathology records were reviewed for the presence of pancreatic cysts. In determining natural history, only patients with 2 scans more than 1 month apart at our institution were included. Patients with a clinical history and laboratory evidence of pancreatitis and/or pathologic confirmation of a pseudocyst were excluded. Factors analyzed as potential predictors of neoplasia included age, gender, cyst size, and symptoms. Serous cystadenomas, solid and cystic papillary (Hamoudi) tumors, lymphoepithelial cysts and simple cysts were all benign, whereas mucinous cystic neoplasms, intraductal papillary mucinous neoplasm, cystic neuroendocrine tumors, and cystadenocarcinomas were considered to be premalignant or malignant.


Among 24,039 CT or MR scans, 290 patients (1.2%) had pancreatic cysts, and 168 of these patients (0.7%) had no documentation of pancreatitis. Seventy-nine of these patients with 103 cysts had more than 1 scan with an average interval of 16 months. These cysts increased in size in 19%, did not change in 59% and decreased in 22% of patients. Forty-nine patients underwent surgery for 14 benign (serous cystadenomas = 10, Hamoudi = 2, lymphoepithelial = 1, simple = 1) 25 premalignant (mucinous cystic neoplasm =16, intraductal papillary mucinous neoplasm = 5, neuroendocrine tumors = 4), or 10 malignant (intraductal papillary mucinous neoplasm = 7, cystadenocarcinomas = 3) lesions. Gender and cyst size did not predict neoplasia. However, presence of symptoms predicted premalignant or malignant pathology (60% vs. 23%, P < 0.05), and age over 70 years was associated with malignancy (60% vs. 21%, P < 0.02).


These data suggest that cystic pancreatic neoplasms 1) occur in 0.7% of patients, 2) increase in 19% over 16 months, and 3) are likely (60%) to be malignant in patients older than 70 years. Therefore, we recommend surgical excision for pancreatic cysts that are increasing under observation, symptomatic, or detected radiologically in fit older patients.

Cystic lesions of the pancreas are being detected more often as sensitive abdominal imaging tests are being used for multiple indications.1–3 Many of these patients in whom a small cystic lesion is detected are elderly and asymptomatic. If resection is undertaken when the pancreas is otherwise normal, however, the risk of a pancreatic fistula is increased.4–6 On the other hand, the majority of these lesions are either premalignant or malignant.1,2,4,5,7,8 Thus, the decision of whether to observe or to operate on these small cystic pancreatic lesions is difficult.

Little information is presently available on the incidence and natural history of pancreatic cysts. The decision with respect to observation or surgery is further complicated by the fact that a significant percentage of patients with pancreatic cysts have pseudocysts secondary to pancreatitis rather than a pancreatic neoplasm. In addition, patients with mucinous cystic neoplasms (MCNs) and intraductal papillary mucinous neoplasms may present with pancreatitis as viscous secretions occlude the pancreatic duct. Moreover, neither imaging studies nor cyst aspiration has sufficient sensitivity and specificity to accurately differentiate among the multiple benign, premalignant and malignant pathologies. Thus, the aims of this analysis are to define the incidence, natural history, and predictors of neoplasia in pancreatic cysts to determine which patients can safely be observed and which should undergo an operation.



A retrospective study was carried out of all patients seen over an 8-year period from 1995 to 2002 at the Froedtert Memorial Hospital and the Medical College of Wisconsin. The Institutional Review Board granted approval for the protocol. All radiologic, surgical, and pathology records were reviewed for the presence of pancreatic cysts. Radiology reports of abdominal computerized tomography (CT) and magnetic resonance images (MRI) were searched using the radiology information system for the words “pancreas” or “pancreatic” and “cysts” or “cystic.” The hospital medical records were searched to determine the patient's age at time of first imaging study, gender, presenting symptoms, and any occurrence of pancreatitis or surgical resection. Additional information was gathered from the medical record for all patients undergoing surgery and from the pathology records of all surgical specimens.

In determining natural history, only patients with 2 scans more than 1 month apart at our institution were included. Patients with a clinical history and laboratory evidence of pancreatitis and/or pathologic confirmation of a pseudocyst were excluded. The various patient populations are depicted in Figure 1. A total of 290 patients with CT or MR scans were identified. During this time period, 49 patients underwent surgery for pancreatic cystic neoplasms, 24 of whom (49%) had their scans done at Froedtert Memorial Hospital. The 168 patients without a history of pancreatitis detected radiologically, the subset of 79 of these patients with follow-up scans, and the 49 patients undergoing surgery were compared with respect to age, gender, symptoms, cyst size, and location. The mean age of the radiologically detected patients (64.3 years) was slightly older than the surgical patients (59.3 years). On the other hand, the surgical patients were slightly more likely to be female (71% vs. 63%) and to be symptomatic (49% vs. 42%). In addition, the surgical patients had slightly larger cysts as measured radiologically (4.0 cm vs. 2.5 cm), and their cysts were slightly more likely to be in the pancreatic head (55% vs. 44%).

figure 9FF1
FIGURE 1. The various patient populations, including 290 patients discovered by CT or MR scan and the 49 patients undergoing surgery subclassified by pancreatitis, scans, and cyst type.


A wide variety of imaging modalities, protocols, and equipment were used over the 7-year interval. Abdominal CT was the most common imaging procedure employing both single detector helical CT in the initial phase of the study and subsequently 4-, 8-, and 16-channel multidetector row CT. Standard injection/acquisition protocols using 5-mm image thickness and parenchymal phase acquisition (beginning 60–70 seconds after intravenous contrast medium injection of a 42-g iodine load) was used most commonly. In patients with acute pancreatitis or suspected to have a pancreatic neoplasm, a double pass acquisition employing “pancreatic phase” and conventional parenchymal phase timing following a bolus intravenous contrast medium injection (5 mL/s) and 2.5-mm image thickness was used. All studies were performed using General Electric CT scanners (Milwaukee, WI): CTi, single detector helical CT, or LightSpeed multidetector row CT (4-, 8-, 16-channel). MRIs were performed with a variety of imaging protocols, including T2-weighted images, gadolinium-enhanced images, and magnetic resonance cholangiopancreatography on 1.5 Tesla GE scanners, with Signa Advantage, Echo Speed Horizon, or Horizon CVi.

Natural History

Seventy-nine patients without a history of pancreatitis had follow-up imaging, mostly enhanced CT of the abdomen (Fig. 1). The average length of follow-up was 16.5 months. These 79 patients were similar in age and gender to the entire radiologic group, and 58% were asymptomatic. These 79 patients had 103 cysts. The mean cyst size was 2.3 cm. Forty-six patients (58%) had cysts less than 2 cm, 22 (28%) had cysts between 2.0 cm and 4.9 cm, and 11 (14%) had cysts more than 5 cm in greatest diameter. This subset of 79 radiologic patients was slightly more likely to have a cyst in the body or tail of the pancreas (73% vs. 56%). An example of one of the patients with a follow-up CT scan is demonstrated in Figure 2. For a cyst to be considered as having changed in size, an increase or decrease of at least 0.2 cm was required.

figure 9FF2


Only patients undergoing surgery for cystic pancreatic lesions were included in this analysis. The mean age was 59.3 years, and 71% were female. All 49 patients had preoperative CT scans, and 2 also had MRI scans. Eighteen of these patients (37%) also had preoperative endoscopic retrograde cholangiopancreatography, and 9 (18%) had endoscopic ultrasound with fine needle aspiration. Twenty-five (51%) were asymptomatic while 19 (39%) presented with pain, 4 (8%) had nausea and fullness, and 1 (2%) presented with hypoglycemic coma. Twenty patients (41%) underwent pancreatoduodenectomy, 18 (90%) with pylorus preservation. Fifteen patients (31%) underwent distal pancreatectomy, 10 (67%) with splenic preservation. Eleven patients (22%) underwent cyst enucleation. Three patients (6%) with liver or peritoneal metastasis only underwent exploratory laparotomy (2) or palliative bypass procedures (1).


All pathologic specimens were reviewed by 1 surgical pathologist (R.A.K.), and tumors were classified as benign, premalignant, or malignant. Serous cystadenoma (SCA, n = 10), solid and cystic papillary (Hamoudi, n = 2) tumors, lymphoepithelial (LE, n = 1) cysts, and simple (n = 1) cysts were all considered to be benign tumors. Premalignant cysts included MCNs (n = 16), intraductal papillary mucinous neoplasms (IPMNs, n = 5) without invasive carcinoma, and cystic neuroendocrine tumors (NETs, n = 4). Malignant cystic tumors included IPMNs (n = 7) with invasive cancer and cystadenocarcinomas (CACs, n = 3), each of whom had the diagnosis established by biopsy of liver or peritoneal metastases. Thus, 14 patients (29%) had benign cysts, 25 patients (51%) had premalignant cysts, and 10 patients (20%) had MCNs (Fig. 1).


Data are presented as mean ± standard error of the mean (SEM). Statistical analyses were performed by analysis of variance, Student's t test, χ2 analysis, or the Mann-Whitney rank sum test where appropriate with statistical significance achieved at P < 0.05.


Natural History

Among 24,039 CT or MR scans, 290 patients (1.2%) had pancreatic cysts, and 168 of these patients (0.7%) had no documentation of pancreatitis (Fig. 1). Thus, 59% of the cysts were detected in patients with no history of pancreatitis. Of these 168 patients, 79 had imaging follow-up at least 1 month after the initial scan. The natural history of the cysts with demographics of the subgroup of these 79 patients is presented in Table 1. In 15 patients (19%) cysts increased in size; in 47 patients (59%) cysts did not change; and in 17 patients (22%) cysts decreased in size. These 3 subgroups of patients did not differ with respect to mean age, gender, length of follow-up (21.8 months for increase, 15.1 month for no change, and 16.3 months for decrease), or location within the pancreas. Patients with cysts that did not change were more likely (67%, P < 0.04) to be asymptomatic compared with those that increased (47%) or decreased (41%). Patients with cysts that increased had larger cysts (3.5 cm, P < 0.04) compared with the other groups. In addition, cysts larger than 2.0 cm were more likely to be associated with pain (33% vs. 16%, P < 0.01). Further analysis of the 15 patients whose “cysts” increased revealed that 11 had malignancies, including 3 unresectable pancreatic cancers and 3 resectable solid pancreatic neoplasms. Four patients had advanced cirrhosis, 1 had von Hippel-Lindau syndrome, and 1 refused surgery.

Table thumbnail
TABLE 1. Cyst Natural History


Data on the 49 surgical patients grouped by their final pathology as benign, premalignant, or malignant are presented in Table 2 and Figure 3. Patients with malignant pathology were significantly older (P < 0.02) than patients with benign or premalignant cysts (Table 2). In addition, the percentage of patients over 70 years of age was greatest in patients with malignant cystic tumors (60% vs. 21%, P < 0.02, Fig. 3A). Moreover, the median age of patients with benign cysts was 50 years, with premalignant cysts was 60 years, and with malignant cysts was 75 years. Thus, age clearly correlated with final pathology.

figure 9FF3
Table thumbnail
TABLE 2. Surgical Patients by Pathology

Similarly, patients with premalignant or malignant pathology were less likely (P < 0.05) to be asymptomatic (Table 2) and more likely to be symptomatic (60% vs. 23%, P < 0.05, Fig. 3B). Neither gender, nor cyst size, nor location within the pancreas predicted final pathology (Table 2). Patients with malignant cystic neoplasms were more likely (P < 0.05) to undergo a pancreatoduodenectomy and less likely (P < 0.05) to undergo a distal pancreatectomy. Diagnostic or palliative surgery was only performed in patients with metastatic cystic tumors, whereas enucleation was only done in patients with benign or premalignant cysts.

Morbidity, Mortality, and Follow-up

The most frequent postoperative complication was pancreatic fistula, which occurred in 10 patients (20%). A pancreatic fistula occurred in 5 of 14 patients (36%) with benign pathology, in 4 of 25 patients (16%) with premalignant cysts, and in 1 of 7 patients (14%) undergoing resection for a malignant cystic neoplasm. A pancreatic fistula occurred in 2 of 20 patients (10%) undergoing pancreatoduodenectomy, in 5 of 15 patients (33%) after distal pancreatoduodenectomy, and in 3 of 11 patients (27%) whose cysts were enucleated. A variety of nonpancreatic complications occurred in 10 patients (20%), including 4 (29%) in the benign group, 4 (16%) in the premalignant group, and 2 (20%) in the malignant group. One hospital death (2%) from multiple organ dysfunction occurred in a 79-year-old woman 7 weeks after a pancreatoduodenectomy for a malignant IPMN. Three late complications (6%), an intra-abdominal abscess, a pseudocyst, and a small bowel obstruction, occurred all in patients with premalignant pathology.

The median postoperative length of stay was 15 days for the benign group, 11 days for the premalignant group, and 13 days for the malignant group as well as for all 49 surgical patients. The mean and median survival for the benign group was 60.5 and 37.2 months, respectively, with all patients alive. The mean and median survival for the premalignant group was 42.7 and 34.7 months, respectively, with all patients alive. The mean and median survival for the malignant group was 37.7 and 30.4 months, respectively, with 4 patients alive and 6 dead, including all 3 with metastatic disease and 3 of 7 with localized malignancies. The median survival for the 3 patients with metastatic CACs was 14.5 months but was 47.4 months for the 7 patients with resectable malignant IPMNs.


This analysis of more than 24,000 CT and MR scans over an 8-year period demonstrates that pancreatic cysts are seen in 1.2% of these patients. A total of 168 of these 290 patients (59%) had no history of pancreatitis, suggesting that 0.7% of patients undergoing abdominal imaging may have a variety of pancreatic cystic neoplasms. More than 1 scan was available for comparison in 79 of the nonpancreatitis patients with a mean interval between scans of 16.5 months. In 15 of these patients (19%) the cysts increased in size, in 47 (59%) the cysts did not change, and in 17 (22%) the cysts decreased under observation. Of 49 patients who underwent surgery, 35 (71%) had premalignant (51%) or malignant (20%) lesions. Neither gender, cyst size nor cyst location predicted neoplasia. However, symptomatic patients were more likely to have premalignant or malignant pathology (60% vs. 23%, P < 0.05), and age over 70 years was a predictor of malignancy (60% vs. 21%, P < 0.02).

Very few data are available with respect to the incidence of pancreatic cysts in patients without a history of pancreatitis. In an ultrasound (US) screening study from Japan, Yoda et al found cystic lesions of the pancreas in 0.2% of the study population.9 This incidence is lower than the 0.7% found in the present study. This difference may be due to multiple factors including differences in screening modalities (US vs. CT/MR), ethnic background (Asian vs. white), and study population (general population vs. hospital patients). At the other end of the spectrum, Kimura et al found cysts at least 0.4 cm in size in 73 of 300 autopsy cases (24%).10 This very high incidence is probably due to the inclusion of very small cysts not routinely imaged by CT or MR.

Also, of note is the fact that only 41% of the patients with cysts detected in the present study had a history of pancreatitis. This percentage is significantly lower than the classic teaching that 90% of pancreatic cysts are pseudocysts.11 However, the results of this study are similar to those reported by Megibow et al who found a history of pancreatitis in 89 of their 192 patients (46%).3 At the other end of the spectrum, Zhang et al12 documented pancreatitis in 75 of 283 patients (27%); Ikeda et al13 found evidence of pancreatitis in 19 of 120 patients (16%); and Fernandez-del Castillo et al2 diagnosed pseudocysts in only 29 of 212 patients (14%). These dramatic differences in the percentage of patients with pancreatitis are probably due to increased detection of small asymptomatic cysts in some series as well as differences in referral patterns.

Data on the natural history of pancreatic cysts in patients without pancreatitis also are limited. Ikeda et al from Japan followed 31 patients with small pancreatic cysts by ultrasound and found that only 1 (3%) increased, whereas 4 (13%) decreased over 3 years.13 Similarly, Allen et al14 from New York found that only 6 of 138 patients (4%) they initially observed grew sufficiently over 31 months to warrant surgery. However, Megibow et al3 found that 2 of 21 cysts (10%) increased over 30 months. Thus, the finding in this study that 15 of 79 patients (19%) had cysts that increased over 16 months is a higher incidence than in these other reports. However, Handrich et al from the Mayo Clinic found that cysts increased in size in 10 of 23 patients (43%) over 8 years of follow-up.15

Most authorities agree that no imaging modality is sufficiently accurate to differentiate among the multiple benign, premalignant, and malignant lesions that are visualized by US, CT, or MR scans.1–5,12,16,17 In the 1990s some authorities recommended cyst aspiration with fluid analysis for cytology, amylase, viscosity and various tumor markers.18,19 More recently, endoscopic ultrasound (EUS) with fine-needle aspiration has been suggested as a method to differentiate among benign, premalignant, and malignant lesions.20-22 However, percutaneous or endoscopic aspiration has the potential to spill malignant cells with the possibility of reducing survival. For this reason, as well as ongoing concerns about the accuracy of fluid analysis, we1 and others2,5 do not routinely use these expensive, invasive diagnostic modalities. In the absence of a good radiologic or pathologic test to preoperatively determine the diagnosis, clinical characteristics such as age, gender, the presence of symptoms, cyst size, or location might be helpful.

In the present analysis, age greater than 70 years was a predictor of malignant pathology. This observation is consistent with multiple other reports. Both serous cystadenomas and benign MCNs occur in patients with a median age in their 50s.4,5,7,8,23,24 Similarly, solid and cystic papillary (Hamoudi) tumors as well as lymphoepithelial cysts and simple cysts have generally been reported in younger patients.4,7,25,26 Among the premalignant tumors, cystic NETs also have a mean age of approximately 50 years.27 Similarly, benign MCN have a median age around 50 to 55 years, whereas mucinous CACs generally have a median age of 60 to 65 years.4,7,11,23 Moreover, patients with benign IPMNs have a median age of 60 to 65 years, whereas malignant IPMNs usually have a median age of 65 to 70 years.28,29 Thus, older patients are more likely to have premalignant or malignant cystic pancreatic neoplasms.

In the present analysis, the presence of symptoms also predicted premalignant or malignant pathology. This observation is also supported by the recent literature. Moesinger et al found that less than half of their patients with serous cystadenomas and benign mucinous cystic tumors had symptoms, whereas 80% to 85% of their patients with CACs were symptomatic.7 Similarly, Fernandez-del Castillo et al found that 134 of their 212 patients (63%) with pancreatic cysts were symptomatic.2 Malignant lesions were found in 40% of their symptomatic patients compared with only 17% of their asymptomatic patients (P < 0.001). These same authors found that 66% of their surgical patients had premalignant or malignant pathology, which is similar to the 71% in the present report. Interestingly, Fernandez-del Castillo et al found premalignant or malignant pathology in 65% of cysts less than 2 cm and in 66% of cysts greater than 2 cm in diameter.2 Thus, cyst size did not predict pathology.

In the present analysis, cyst size also did not correlate with final pathology. This observation is consistent with earlier reports as many serous cystadenomas, benign MCNs, and solid and cystic papillary (Hamoudi) tumors may become quite large before they become symptomatic.4,7,23 Similarly, the mean size of reported cystic NETs is 6.4 cm.27 In addition, Sarr et al found no difference in the mean size of benign MCN and CACs.23 On the other hand, Sohn et al found that the mean size of malignant IPMNs was only 4.0 cm, which is similar to the 4.7 cm in this report.28 In comparison, Chari et al found that their 73 noninvasize IPMNs averaged 5.2 cm, whereas their 40 invasive IPMNs averaged 6.6 cm.29 Thus, size alone does not predict pathology or malignancy.

In the present report, cyst location also did not predict premalignant or malignant pathology. This lack of correlation may be due to the relatively small number of patients but also may be the result of the multiple potential pathologies. In some reports, serous cystadenomas and benign MCNs are relatively more likely to be found in the body or tail of the pancreas.4,11,23 Conversely, CACs and malignant IPMNs have a tendency to occur more often in the head of the pancreas.23,28 On the other hand, some of the more rare tumors, such as solid and cystic papillary (Hamoudi) tumors and cystic NETs, tend to be evenly distributed throughout the pancreas.7,27

In this analysis, gender did not predict pathology. Most cystic lesions of the pancreas, including serous cystadenomas, MCNs, solid and cystic papillary tumors and cystic NETs, occur more often in women.2,4,7,8,11,23,24,27 In comparison, in 2 recent reports of IPMNs in 60 patients from Johns Hopkins and 113 patients from the Mayo Clinic, 56% occurred in men.28,29 Thus, male gender may suggest an IPMN, and analyses with larger numbers of patients might indicate that gender, like advanced age, is a clinical predictor of a premalignant or malignant cystic pancreatic neoplasm.

Considerable controversy presently exists with respect to the management of small, asymptomatic pancreatic cysts. Ikeda et al13 have recommended observation with endoscopic ultrasonography in 1 year and then every 2 years if the cyst has not changed. Similarly, Allen et al have recommended high-quality CT or magnetic resonance cholangiopancreatography on an annual or biannual basis for cysts smaller than 2.5 cm.14 Fernandez-del Castillo el al have also recommended periodic imaging for cysts less than 2 cm in size.2 They recommend resection for larger cysts in young and middle-aged patients and further workup including EUS with aspiration for older patients. Disadvantages of these strategies include the fact that cyst size does not predict pathology, that 65% to 70% of these lesions are premalignant or malignant, and that biannual CT, magnetic resonance cholangiopancreatography, or EUS is quite costly. For a 50-year-old woman with a 1.5-cm cyst, this strategy would cost $25,000 to $50,000 in present dollars for observation over the remainder of her expected life.

The majority of benign lesions are serous cystadenomas as well as solid and cystic papillary (Hamoudi) tumors. These benign lesions frequently can become quite large and symptomatic, eventually requiring a major pancreatic resection, whereas early intervention might allow enucleation or limited resection. Similarly, early surgery in patients with MCNs, cystic NETs, and benign IPMNs will prevent malignant degeneration and is likely to be more cost-effective than observation. Moreover, even small lesions in older patients, especially if they are symptomatic or present with pancreatitis, are likely to be malignant and deserve exploration in fit patients. Thus, we recommend surgical excision for pancreatic cysts that increase under observation, are symptomatic, and are discovered in healthy older patients.


Dr. William H. Nealon (Galveston, Texas): I would like to thank the authors for the opportunity to discuss their paper and providing their excellent manuscript in a timely fashion for my review. Dr. Spinelli, Dr. Pitt and co-workers have provided some important insights into the management of patients with cystic neoplasms of the pancreas.

The authors provide a retrospective search of all CT scans performed at their institution over 8 years. Among more than 24,000 scans, 290 had pancreatic cysts, 168 of these had no history of pancreatitis, and 79 of these patients had follow-up CT scans, permitting the authors to focus on the subset of patients who over time had serial changes or did not.

Among the numerous risk factors that the authors evaluate as predictors of malignancy, their only significant findings were in those patients who were symptomatic and those who were elderly. Although there were some trends, they specifically did not find changes in size either increasing or decreasing as predictors of malignancy. I have 2 questions and a comment.

You have excluded all patients with pancreatitis. As you know, several of the neoplastic mucinous cysts may present with pancreatitis, and thus there is a likelihood that you may have excluded neoplasms by excluding pancreatitis in your evaluation. Have you evaluated the patients who had pancreatitis?

I know that 14% of your operated patients had a history of pancreatitis. Can you tell us whether those specific patients had pre-malignant or malignant lesions? It certainly is vital to emphasize that pancreatitis may actually be a clue to the presence of these neoplasms.

My second question relates to the fact that your survey included a wide variety of cystic lesions. Since the most intense scrutiny is now focused upon IPMN, do you feel that your conclusions made on the basis of all these cysts can be applied individually to IPMN?

I think this is a very valuable paper for what you did not show. As you mentioned repeatedly in your manuscript, there remains at this time no accepted system to select the patients who require resection. There are numerous papers which advocate monitoring, all based upon what they admit to be an inadequate system of follow-up, potentially permitting a pre-malignant lesion to progress to malignancy.

You have shown convincingly that changes in the size of the cyst over time does not correlate with the presence of benign or malignant neoplasms. This is a strong argument for uniformly advocating resectional therapy until more precision can be established for selecting patients whose cysts are safe to follow.

Dr. Andrew L. Warshaw (Boston, Massachusetts): I guess the world is made up of lumpers and splitters. This is a lumper paper, and I rise as a splitter. While Dr. Nealon has raised the question about IPMN, there are 2 categories of cystic tumors that have not really been distinguished here.

One, perhaps the most prevalent individual one, is the serous cystadenoma. There are 6 reported cases in the world of that being or becoming malignant; therefore, the only justification for operating on the serous cystadenoma would be symptoms or growth to the point of visceral or biliary obstruction.

Second, there are reports that small mucinous cystic tumors, perhaps less than 3 cm, have only about a 3% prevalence of malignancy in them; so those very small tumors can safely be observed, especially in an elderly patient with comorbidities.

Finally, there are data suggesting that, when in doubt, image or endoscopic ultrasound-guided fine needle aspiration for biochemical parameters such as mucin or CEA and cytology can be useful in making a clinical decision.

Dr. John M. Daly (Philadelphia, Pennsylvania): Dr. Spinelli, I was going to follow up on what Dr. Warshaw just said. I thought you showed remarkable restraint as a radiologist in not putting a needle into these cystic lesions, and I wondered if you just didn't do that as policy at your institution.

Second, the cystic lesions that occur in the head of the pancreas often are the most problematic, because it is a much larger operation to deal with them surgically. And I wondered if your endoscopists, either in surgery or in GI, were using transduodenal ultrasound and aspiration of these lesions in the pancreatic head to do the fluid and cytology analyses that Dr. Warshaw was discussing?

Dr. J. Patrick O'Leary (New Orleans, Louisiana): In my experience, I have always thought that if septa were demonstrated that suggested a neoplasm. Were there any radiologic characteristics that might separate these cystic tumors from a pseudocyst?

Dr. Henry A. Pitt (Milwaukee, Wisconsin): Dr. Nealon, with respect to the patients with pancreatitis that were excluded from this analysis, obviously they are an interesting group as well. You are correct to note that patients with cystic pancreatic malignancies can present with pancreatitis. We have tried to stress this point with our medical colleagues, and we agree with you.

We are in the process of further analyzing the 41% of our 290 patients who had cysts and pancreatitis. As you might imagine, this group had a higher percentage whose cysts decreased in size. However, we need to look a little more closely to be sure that we are not missing some malignancies in that group.

In response to your question and Dr. Warshaw's comment about lumping the various patients together, we believe that you have to group these patients when you look at them preoperatively because you don't have a tissue diagnosis. If you think of these patients in these broad categories of benign, pre-malignant, or malignant and realize that two-thirds or more of these cystic tumors are pre-malignant or malignant, then that helps in your analysis.

What if we then apply these data to the subset of patients with IPMN? We will hear 2 papers, one from Hopkins and one from MGH tomorrow about IPMNs. Both papers will completely confirm what we have shown in our analysis. Patients with benign IPMNs are in their early sixties whereas the malignant IPMN patients are in the late sixties and are more likely to be symptomatic. Therefore, an age difference and a symptom difference exists between benign and malignant IPMNs, confirming exactly what we have shown in our presentation.

Another important negative finding from our study, which is confirmed by a paper published by the MGH group earlier this year, is that cyst size does not predict pathology. In the MGH data, two-thirds of their patients with cysts smaller than 2 centimeters had pre-malignant or malignant lesions, as did two-thirds of their patients with cysts larger than 2 centimeters. The message is that you should not be fooled by a small cyst, these can be pre-malignant or malignant lesions.

Dr. Warshaw, with respect to serous cystadenomas, you are right, there have been a few isolated reported cases of malignant degeneration. However, S. T. Fan and his colleagues in Hong Kong summarized the world literature a couple years ago and showed that there is a 97% 5-year survival in these patients. Although an occasional patient may demonstrate malignant degeneration, the vast majority of these patients have a benign lesion.

In addition, if you operate on patients with serous cystadenomas at a relatively young age when the lesion is relatively small and can do a more minor operation, for example, an enucleation or a central pancreatectomy, you may avoid a very big operation when the patient is older and less fit.

With respect to both Dr. Warshaw's and Dr. Daly's comments about fine needle aspiration and EUS, we do these studies selectively. Only 18% of our patients who came to surgery had EUS and FNA. Unfortunately, these studies are not 100% accurate, and you can get a false sense of security by a false negative FNA and follow a cyst that is pre-malignant or malignant. Moreover, these studies are expensive, and there are some risks to doing these aspirations. As a result, we have not encouraged our endoscopists to do too many EUSs on these patients but have asked them to refer these patients directly to surgery.

Finally, with respect to Dr. O'Leary's comments, unfortunately, there is no way to accurately differentiate among these lesions radiologically. For those surgeons and radiologists who are interested in this area, we might be accurate 80% or 90% of the time. However, even at surgery many of these lesions have a similar gross appearance, and only the pathologist can give us the final answer.


Reprints: Henry A. Pitt, MD, Department of Surgery, Medical College of Wisconsin, 9200 W. Wisconsin Ave., Milwaukee, WI 53226. E-mail: ude.wcm@ttipah.


1. Kiely JM, Nakeeb A, Komorowski RA, et al. Cystic pancreatic neoplasms: enucleate or resect? J Gastrointest Surg. 2003;7:890–897. [PubMed]
2. Fernández-del Castillo C, Targarona J, Thayer SP, et al. Incidental pancreatic cysts: clinicopathologic characteristics and comparison with symptomatic patients. Arch Surg. 2003;138:427–434. [PubMed]
3. Megibow AJ, Lombardo FP, Guarise A, et al. Cystic pancreatic masses: cross-sectional imaging observations and serial follow-up. Abdom Imaging. 2001;26:640–647. [PubMed]
4. Talamini MA, Pitt HA, Hruban RH, et al. Spectrum of cystic tumors of the pancreas. Am J Surg. 1992;163:117–124. [PubMed]
5. Talamini MA, Moesinger R, Yeo CJ, et al. Cystadenomas of the pancreas: is enucleation an adequate operation? Ann Surg. 1998;227:896–903. [PMC free article] [PubMed]
6. Yeo CJ, Cameron JL, Maher MM, et al. A prospective randomized trial of pancreatico-gastrostomy versus pancreaticojejunostomy after pancreatico-duodenectomy. Ann Surg. 1995;222:580–592. [PMC free article] [PubMed]
7. Moesinger RC, Talamini MA, Hruban RH, et al. Large cystic pancreatic neoplasms: pathology, resectability, and outcome. Ann Surg Oncol. 1999;6:682–691. [PubMed]
8. Warshaw AL, Compton CC, Lewandowski K, et al. Cystic tumors of the pancreas: new clinical, radiologic, and pathologic observations in 67 patients. Ann Surg. 1990;212:432–443. [PMC free article] [PubMed]
9. Yoda Y, Suzuki Y, Yamada K, et al. A study of mass survey for pancreatic tumor using ultrasonography [in Japanese]. J Gastroenterol Mass Surv. 1992;96:50–55.
10. Kimura W, Nagai H, Kurodu A, et al. Analysis of small cystic lesions of the pancreas. Int J Pancreatol. 1995;18:197–206. [PubMed]
11. Meyer W, Köhler J, Gebhardt C. Cystic neoplasms of the pancreas: cystadenomas and cystadenocarcinomas. Langenbeck's Arch Surg. 1999;384:44–49. [PubMed]
12. Zhang X-M, Mitchell DG, Dohke M, et al. Pancreatic cysts: depiction on single-shot fast spin-echo MR images. Radiology. 2002;223:547–553. [PubMed]
13. Ikeda M, Sato T, Ochiai M, et al. Ultrasonographic follow-up study of small pancreatic cysts of unknown etiology. Bildgebung. 1993;60:209–214. [PubMed]
14. Allen PJ, Jaques DP, D'Angelica M, et al. Cystic lesions of the pancreas: selection criteria for operative and non-operative management in 209 patients. J Gastrointest Surg. 2003; in press. [PubMed]
15. Handrich SJ, Hough DM, Fletcher JG, et al. The clinical significance of the incidentally discovered small simple pancreatic cyst: follow-up and outcome. Proceedings of the Pancreas Club, Inc. 2003:40.
16. Curry CA, Eng J, Horton KM, et al. CT of primary cystic pancreatic neoplasms: can CT be used for patient triage and treatment. AJR Am J Roentgenol. 2000;175:99–103. [PubMed]
17. Koito K, Namieno T, Ichimura T, et al. Mucin-producing pancreatic tumors: comparison of MR cholangiopancreatography with endoscopic retrograde cholangiopancreatography. Radiology. 1998;208:231–237. [PubMed]
18. Alles AJ, Warshaw AL, Southern JF, et al. Expression of CA 7204 (TAG-72) in fluid contents of pancreatic cysts: a new marker to distinguish malignant pancreatic cystic tumors from benign neoplasms and pseudocysts. Ann Surg. 1994;219:131–134. [PMC free article] [PubMed]
19. Sperti C, Pasquali C, Guolo P, et al. Serum tumor markers and cyst fluid analysis are useful for the diagnosis of pancreatic cystic tumors. Cancer. 1996;78:237–243. [PubMed]
20. Walsh RM, Henderson JM, Vogt DP, et al. Prospective preoperative determination of mucinous pancreatic cystic neoplasms. Surgery. 2002;132:628–634. [PubMed]
21. Maguchi J, Osani M, Yanagawa N, et al. Endoscopic ultrasonography diagnosis of pancreatic cystic disease. Endoscopy. 1998;30:108–110. [PubMed]
22. Sand JA, Hyoty MK, Muttila J, et al. Clinical assessment compared with cyst fluid analysis in the differential diagnosis of cystic lesions of the pancreas. Surgery. 1996;119:275–280. [PubMed]
23. Sarr MG, Carpenter HA, Prabhakar LP, et al. Clinical and pathologic correlation of 84 mucinous cystic neoplasms of the pancreas. Ann Surg. 2003;138:427–434. [PMC free article] [PubMed]
24. Pyke CM, vanHeerden JA, Colby TV, et al. The spectrum of serous cystadenoma of the pancreas: clinical, pathologic and surgical aspects. Ann Surg. 1992;215:132–139. [PMC free article] [PubMed]
25. Mandavilli SR, Port J, Ali SZ. Lymphoepithelial cyst (LEC) of the pancreas: cytomorphology and differential diagnosis on fine-needle aspiration (FNA). Diagn Cytopathol. 1999;20:371–374. [PubMed]
26. Bergin D, Ho LM, Jowell PS, et al. Simple pancreatic cysts: CT and endosonographic appearances. AJR Am J Roentgenol. 2002;178:837–840. [PubMed]
27. Ahrendt SA, Komorowski RA, Demeure MJ, et al. Cystic pancreatic neuroendocrine tumors: is preoperative diagnosis possible? J Gastrointest Surg. 2002;6:66–74. [PubMed]
28. Sohn TA, Yeo CJ, Cameron JL, et al. Intraductal papillary mucinous neoplasms of the pancreas: an increasingly recognized clinicopathologic entity. Ann Surg. 2001;234:313–322. [PMC free article] [PubMed]
29. Chari ST, Yadav D, Smyrk TC, et al. Study of recurrence after surgical resection of intraductal papillary mucinous neoplasm of the pancreas. Gastroenterology. 2002;123:1500–1507. [PubMed]

Articles from Annals of Surgery are provided here courtesy of Lippincott, Williams, and Wilkins
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...