• 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. Feb 2000; 231(2): 205–212.
PMCID: PMC1420988

Clinical and Pathologic Correlation of 84 Mucinous Cystic Neoplasms of the Pancreas

Can One Reliably Differentiate Benign From Malignant (or Premalignant) Neoplasms?

Abstract

Objective

To determine whether the long-term behavior of cystic mucinous neoplasms of the pancreas could be predicted using a novel, precisely defined classification of benign mucinous cystadenomas, noninvasive proliferative cystic mucinous neoplasms, and invasive mucinous cystadenocarcinomas. The primary interest was to obtain long-term follow-up after complete resection to determine the recurrence rates based on this objective classification.

Background

Current understanding is that all cystic mucinous neoplasms of the pancreas are potentially malignant and that mucinous cystadenomas, when completely removed, are biologically benign. Cystadenocarcinomas are thought to be less aggressively malignant than ordinary ductal adenocarcinoma, but reported recurrence rates vary widely and are unpredictable.

Methods

All patients who underwent “curative” resection for cystic mucinous neoplasms at Mayo Clinic Rochester from 1940 to 1997 were identified. All available pathology slides, gross specimens, and clinical records were reviewed, eliminating patients with inadequate documentation. Neoplasms were reclassified as mucinous cystadenomas, noninvasive proliferative mucinous cystic neoplasms, or invasive cystadenocarcinomas based on specific histologic criteria.

Results

Of 84 patients (70 women, 14 men) with cystic mucinous neoplasms of the pancreas, 54 were classified as cystadenomas, 23 as noninvasive proliferative cystic mucinous neoplasms, and only 7 as cystadenocarcinomas. Recurrent disease developed in none of the 77 patients without invasion, but 5 of the 6 patients surviving resection for cystadenocarcinomas died of recurrent cystadenocarcinoma within 5 years.

Conclusions

When the neoplasm is completely resected and subjected to adequate histopathologic examination based on these objective criteria, absence of tissue invasion predicts a curative operation and detailed follow-up may be unnecessary. In contrast, a histologic diagnosis of invasive cystadenocarcinoma portends a dismal prognosis, similar to that of typical ductal adenocarcinoma of the pancreas.

Cystic neoplasms of the pancreas are rare primary neoplasms of the pancreas. There are continuing problems in differentiating cystic neoplasms before surgery from other nonmalignant cystic lesions of the pancreas and in understanding the natural history and malignant potential of the different cystic neoplasms. Commonly, the diagnosis of cystadenocarcinoma is loosely based on a combination of epithelial proliferation and nuclear dysplasia, irrespective of the demonstration of actual tissue invasion. Biologically, when so categorized, these “cystadenocarcinomas” as a group are thought to be less aggressively malignant than ordinary ductal adenocarcinoma, although the reported rate of recurrence is unpredictable and the incidence of metastases varies widely. 1–5 Moreover, classifying cystic neoplasms as benign or malignant has been said to be difficult due to the discontinuous epithelial lining of the cystic wall. 6 In addition, some investigators suggest that identification of benign and malignant epithelia within the same specimen of patients with mucinous cystic neoplasms indicates that there is a spectrum of disease rather than true benign or malignant entities. 7

Compagno and Oertel 8,9 first characterized and differentiated the universally benign, serous cystic neoplasms from the premalignant or overtly malignant mucinous cystic neoplasms. Because of the reputed difficulty of classifying mucinous cystic neoplasms as “benign” cystadenomas, some pathologists classify mucinous cystic neoplasms of the pancreas with only minor areas of epithelial hyperplasia or dysplasia as grade 1 cystadenocarcinoma, 2,5,7,10 thereby implying a more aggressive natural history and predetermining their future management and follow-up as a malignancy despite successful resection.

Our overall objective was to identify histologic features of individual neoplasms that would predict long-term outcome after complete “curative” surgical resection. Our hypotheses were that after curative resection, mucinous cystadenomas do not recur or metastasize, proliferative dysplastic neoplasms without tissue invasion do not metastasize, and invasive cystadenocarcinomas are likely to recur or metastasize, provided the examination was sufficiently thorough to detect invasion when present. Our aims were to determine whether pathologic examination can reliably differentiate benign from malignant mucinous cystic neoplasms of the pancreas and to determine whether completely resected neoplasms with no evidence of tissue invasion would follow a benign or malignant clinical course. This study was thus designed to provide useful data to support a more or less aggressive resection for subsets of these patients and to support the opinion that an expensive postoperative surveillance program (repeated imaging, screening with tumor markers, regular follow-up examinations) is unnecessary in subsets of patients in whom a benign course could be demonstrated.

METHODS

Patient Selection/Study Groups

We reviewed the medical records and pathology slides and reports of all consecutive patients with cystic neoplasms of the pancreas and those with a pathologic diagnosis of mucinous adenocarcinoma of the pancreas managed at our institution from 1940 to 1997 (n > 200 patients). These patients were identified by our Medical Records Department and then confirmed by the original pathology report. We eliminated patients with serous cystadenoma, ordinary ductal adenocarcinoma, and low-grade mucinous or papillary adenocarcinoma of the pancreatic ductal system. We also eliminated patients if there was insufficient tissue to confirm the diagnosis of a cystic mucinous neoplasm of the pancreas objectively and if gross pathologic, clinical, or radiologic information was incompatible with a diagnosis of cystic mucinous neoplasm of the pancreas. Ultrasonography and computed tomography became available routinely in our institution about 1975; before this, no objective imaging procedures were available, but pathology reports of patients with mucinous cystadenocarcinoma from 1940 to 1975 were reviewed to ensure we included only patients with bona fide cystic mucinous neoplasms.

Next, we reviewed all operative reports and excluded patients who did not have a complete “curative” resection. The remaining 84 patients were divided using in large part the criteria set forth by Compagno and Oertel 9 as modified (see below) into two groups to segregate patients according to the thoroughness of the pathologic evaluation. Group 1 (n = 54) was defined as having gross surgical specimens stored in formalin in our tissue registry that were adequate for complete pathologic study and recutting of slides if necessary. Group 2 (n = 30) did not have the gross specimen available because they had deteriorated or had been discarded; only the original pathologic slides were available but were adequate for histopathologic diagnosis.

Histologic Evaluation

The surgical specimens of group 1 (Fig. 1) and the original slides of groups 1 and 2 were reviewed by a gastrointestinal pathologist (HAC). If the slides were deemed insufficient to evaluate the lesion fully, more slides were prepared by recutting specific regions of the specimen when available (group 1). If the number and location of the original slides were judged to be sufficient to evaluate the reexamined specimen fully, no further slides were made. Using the modified criteria of Compagno and Oertel, 9 the neoplasms were then classified as benign (mucinous cystadenomas); a borderline group we termed “noninvasive proliferative cystic mucinous neoplasms” with or without dysplasia but without invasion of the stroma by neoplastic cells; or malignant (invasive cystadenocarcinomas). Neoplasms classified as noninvasive proliferative mucinous cystic neoplasms displayed a spectrum of histologic changes ranging from epithelial proliferation without dysplasia (hyperplasia) to low-grade or high-grade dysplasia; carcinoma in situ was considered high-grade dysplasia. However, these noninvasive proliferative mucinous cystic neoplasms had no areas of tissue invasion. A diagnosis of cystadenocarcinoma required unequivocal tissue invasion.

figure 9FF1
Figure 1. Cystic mucinous neoplasms of pancreas, resection specimens. (A) Cystadenoma: multiple, large (>2 cm) cysts with thin walls and no obvious or prominent solid component. (B) Proliferative cystic mucinous neoplasm: more granular appearance ...

The following histologic features were assessed: number of layers of cells in the mucinous epithelium; papillary, glandular, or solid growth patterns; gross intracystic polyp formation; low- or high-grade epithelial dysplasia; frequency of mitotic features; presence of invasion; differentiation of carcinoma; presence of mesenchymal stroma and intestinal metaplasia; and evaluation of the pancreatic duct for hyperplasia, dysplasia, or papillary changes when possible.

Low-grade dysplasia was defined as uniform, slightly enlarged, and mildly hyperchromatic nuclei with evenly distributed chromatin and absent or small nuclei. High-grade dysplasia was recognized by large, hyperchromatic, variably sized nucleoli with either vesicular or irregularly clumped chromatin, prominent macronuclei, and large numbers of mitoses. Adenocarcinoma in situ was classified as high-grade dysplasia. Intracystic polypoid masses were grossly visible, intracystic papillary growths. Mesenchymal stroma consisted of primitive, uniform spindled cells resembling ovarian cortex. Intestinal metaplasia consisted of goblet cells, paneth cells, or both cell types.

Specimens were designated as mucinous cystadenomas if they contained a single layer of uniform cuboidal or columnar, mucin-containing cells with no dysplasia (Fig. 2). Neoplasms classified as noninvasive proliferative mucinous cystic neoplasms (Figs. 2 and 3) had epithelium that in areas was composed of multiple cell layers (three or more) and/or microscopic or gross papillary/polypoid formations with and without low- or high-grade dysplasia, but without objective evidence of invasion. Microscopic papillary formations may or may not have demonstrated dysplasia, but gross lesions were always dysplastic. Neoplasms with gross intracystic polyp formation (Fig. 1B) but without tissue invasion would be classified in this group. Invasion was documented by one or more of the following features: infiltrating irregular dysplastic glands with stromal desmoplasia, irregular cell nests, single cell invasion, or vascular space infiltration. Lesions with invasion were classified as invasive mucinous cystadenocarcinomas (Fig. 4).

figure 9FF2
Figure 2. Mucinous cystadenoma of the pancreas, histologic features. Monotonously regular, single layer of benign-appearing mucinous epithelial cells.
figure 9FF3
Figure 3. Proliferative cystic mucinous neoplasms of the pancreas, spectrum of histologic changes. (A) Irregular papillary epithelial growth. Epithelial cells heaped up; nuclei mildly atypical. (B) Papillary epithelial fronds with low-grade nuclear dysplasia. ...
figure 9FF4
Figure 4. Mucinous cystadenocarcinoma, histologic features. Small irregular glands and single cells with high-grade nuclear dysplasia invading the stroma and eliciting a desmoplastic stromal response (i.e., proliferating fibroblast surrounding infiltrating ...

Data Collection and Analysis

After pathologic review, follow-up of the 84 patients with confirmed mucinous cystic neoplasms of the pancreas resected “for cure” was carried out over the 57-year period. Follow-up was 100% complete. Information was obtained through review of the medical record, autopsy reports, and telephone interviews with the patient, the patient’s family, or the local physician in conjunction with the Mayo Tumor Registry. Patient demographics such as age, gender, clinical presentation, past medical history, laboratory values, cause of death, characteristics of the primary neoplasm such as size and location, and the type of operation were examined to determine differences between the three groups.

Because this was a noncontrolled, retrospective study, the only formal statistical analyses were descriptive comparisons. Data are summarized as mean values ± standard error of the mean unless stated differently.

RESULTS

The 84 patients with mucinous cystic neoplasms of the pancreas who had a complete resection were classified histopathologically: 54 were mucinous cystadenoma, 23 were noninvasive proliferative mucinous cystic neoplasms, and only 7 were cystadenocarcinomas. Initially in each of the classification subsets, group 1 (slides and gross specimens reviewed) and group 2 (slides alone reviewed) were analyzed separately because data from group 1 was considered a priori potentially more reliable. Group 1 contained 30 of the 54 patients with mucinous cystadenomas, 19 of 23 with noninvasive proliferative mucinous cystic neoplasms, and 5 of the 7 with invasive cystadenocarcinomas. Because no significant differences were noted between these groups, we combined groups 1 and 2 for this presentation. Overall clinical and pathologic characteristics are outlined in Table 1.

Table thumbnail
Table 1. CLINICAL AND PATHOLOGIC CHARACTERISTICS OF MUCINOUS CYSTIC NEOPLASMS OF THE PANCREAS*

Mucinous Cystadenoma

All but 4 of these 54 patients were women; the mean age was 48 years (range 19–82). The most common presenting symptom was often vague, nonspecific abdominal pain (59%), followed by gastrointestinal symptoms such as nausea, vomiting, or other nonspecific gastrointestinal complaints (26%), weight loss (20%), and back pain (7%). Only one patient had jaundice. A history of “pancreatitis” was obtained in 17% (most patients offered this history by subjective review without objective documentation by imaging test or laboratory confirmation, and thus this high prevalence may be less reliable). Two patients had a history of alcohol abuse.

The neoplasms were typically large: 29 (54%) were larger than 5 cm (up to 27 cm) and the rest were smaller than 5 cm. Most (55%) were in the distal body and tail of the pancreas, 28% were in the body of the gland, and only 16% arose in the head. Distal pancreatectomy (40 patients) was the most commonly performed operation, followed by enucleation (13 patients) and total pancreatectomy (1 patient). There was one surgical death (4%) after enucleation related to postoperative fistula formation and hemorrhage 2 months after surgery. Significant complications occurred in two patients after resection (both required reoperation for drainage of an intraabdominal abscess) and in four patients after enucleation (two of whom required reoperation for drainage of a pancreatic pseudocyst and for a pancreaticocutaneous fistula).

Follow-up was a mean of 11 ± 1 years (range 2–31 years). There were no recurrences in this group, and of the 43 patients with at least a 5-year follow-up, 89% survived the entire 5 years. Of the six deaths, one was an operative death and the other five were due to unrelated causes. Of the 26 patients followed up for more than 10 years, all deaths were due to unrelated causes.

Noninvasive Proliferative Mucinous Cystic Neoplasm

Fifteen of the 23 patients were women. The overall mean age was 53 ± 3 years (range 26–74). Abdominal pain (57%) was the most common complaint, followed by nonspecific upper gastrointestinal symptoms (26%), weight loss (25%), and back pain (18%). Jaundice was present in only one patient. No patients in this group had a history of alcohol abuse, and 20% had a history of pancreatitis.

Eighteen of the 23 neoplasms were larger than 5 cm, ranging up to 30 cm. The neoplasm arose in the distal body or tail in 11 (48%), in the body of the gland in 5 (22%), in the head in 6 (26%), and in the body and head region in 1. Complete resection included distal pancreatectomy in 16, pancreatoduodenectomy in 5, a completion proximal pancreatectomy in 1 patient who had undergone a previous distal pancreatectomy/cystojejunostomy, and a total pancreatectomy in 1. There was one operative death after completion proximal pancreatectomy, related to multiple organ failure. Significant complications occurred in one patient (intraabdominal abscess).

Follow-up was 8 ± 1 years (2–25 years), excluding an early unrelated cardiac death. There were no recurrences. In those followed up for more than 10 years (nine patients), all but one remained alive and well; that patient died of diabetic coma 12 years after total pancreatectomy. Eleven of these 23 patients were considered to have “cystadenocarcinoma” based on the original pathologic report. Ten of these 11 neoplasms were originally described as well-differentiated grade 1 cystadenocarcinoma.

Mucinous Cystadenocarcinoma

The presence of unequivocal tissue invasion defined this group as cystadenocarcinoma. Five of these seven patients were women. The mean age was 64 ± 5 years. Four had abdominal pain and weight loss and two had jaundice. Three were in the pancreatic tail and were treated by distal pancreatectomy; four were in the head of the gland (three treated with pancreaticoduodenectomy, one with total pancreatectomy). All were 5 to 10 cm, but only one had lymph node involvement within the specimen. There was one operative death (after pancreatoduodenectomy), but there were no other serious postoperative complications. Only one patient lived more than 5 years without evidence of recurrent carcinoma; the other five died 0.5 to 5 years after surgery from recurrent disease. The lone 5-year survivor had only a single, microscopic focus of invasion (individual cells isolated in sclerotic hyaline stroma within one high-power field from a region of intracystic polypoid growth).

DISCUSSION

Primary cystic neoplasms of the pancreas are rare. Serous cystic neoplasms of the pancreas are benign, 8,11,12 except for the rare malignant variant (only six cases have been reported to date). 13 However, mucinous cystic neoplasms are at best potentially malignant and present a spectrum of neoplasia ranging from benign neoplasms (mucinous cystadenomas) to presumably premalignant (noninvasive proliferative cystic mucinous neoplasms) to frankly invasive carcinoma (mucinous cystadenocarcinoma). The behavior of the cystic mucinous neoplasms is only partially understood. Most investigators have presumed that histologically benign neoplasms, if not removed, may become malignant. Likewise, the reported 5-year survival rate of patients with mucinous “cystadenocarcinomas” resected for cure appears to be significantly better than that of ordinary ductal adenocarcinomas, ranging from 50% to 70%. 1–5 It remains unclear, however, based on the literature which patients with cystic mucinous neoplasms will do well with complete resection, which require close follow-up for potential recurrence, and which (if any) will benefit from adjuvant therapy. 14

To predict behavior reliably and design rational therapy, we need to improve the diagnosis and understanding of the malignant potential of these neoplasms, but this is difficult at best. Adequate sampling is crucial to obtain an accurate diagnosis, but this is impossible, short of complete surgical excision. 6 These tumors are believed to begin as uni- or multilocular cysts separated from the pancreatic ductal system and lined by a single layer of benign-appearing, mucin-secreting columnar epithelium resembling pancreatic duct epithelium (or the mucinous cells of an ovarian mucinous cystadenoma). Sometimes mucin accumulating within tumor lobules causes pressure necrosis of the lining epithelium, making the epithelial lining discontinuous, 6 and thus complicating histologic differentiation from an inflammatory pseudocyst. Characteristically, when these neoplasms do become malignant, only portions of the tumor dedifferentiate (i.e., become proliferative, develop nuclear dysplasia, and eventually invade the stroma). These areas may not be sampled unless the entire neoplasm is examined. In addition, separation of cystic mucinous neoplasms from low-grade intraductal, papillary, and mucinous adenocarcinomas and high-grade ductal adenocarcinomas that incidentally have cystic areas (related to necrosis or ductal obstruction) requires both adequate sampling and clinicoradiologic and pathologic correlation. Sampling error and the lack of standard histologic criteria for the diagnosis of invasive malignancy have confused our ability to predict tumor behavior in the individual patient. Based on poorly described histologic criteria and a presumed aggressive natural history, some pathologists have diagnosed all mucinous cystic neoplasms with any epithelial proliferation as carcinoma because of their presumed malignant potential and the obvious problems of obtaining a representative sample. Not infrequently, pathologists have diagnosed carcinoma based solely on cellular dysplasia or proliferation in the absence of demonstrated invasion. 2 Such a diagnosis of “cystadenocarcinoma” then necessitates a different prognosis and follow-up management plan, to say nothing of the subjective outlook of the patient who has been told he or she had a malignant pancreatic “cancer.”

The careful and novel histologic classification presented in our study shows that when fully excised, benign cystadenomas and more importantly the noninvasive proliferative cystic mucinous neoplasms do not recur. Whether benign cystadenomas have a premalignant potential remains unknown, but this is likely. In contrast, the noninvasive proliferative cystic mucinous neoplasms manifest cellular and nuclear changes of a premalignant character, but in the absence of documented tissue invasion, they should not be considered an invasive cancer. Indeed, our study suggests that complete excision is curative. Although we have no evidence to prove that these noninvasive proliferative cystic mucinous neoplasms, if left unresected, would eventually become invasive cystadenocarcinomas, the cellular and nuclear characteristics noted are characteristic of malignant transformation of the cells. Similarly, the invasive cystadenocarcinomas in this series had areas of these proliferative changes within the neoplasm distant from the areas of tissue invasion. Nevertheless, because the preoperative differentiation of these three variants of mucinous cystic neoplasms of the pancreas is unreliable, complete excision is the suggested treatment for all cystic mucinous pancreas neoplasms.

The incidence and natural history of resected cystadenocarcinomas of the pancreas is confusing. Fairly large series of resected “cystadenocarcinomas,” including one from our institution 23 years ago of 16 patients with resected “mucinous cystadenocarcinoma of the pancreas,” have been reported. 1–5 Cure rates (5-year survival rates) have ranged from 50% to 70%, with all but one group 1 reporting 5-year survival rates of more than 60%. As a result of our study, in which we used a novel histopathologic classification, we challenge the validity of these diagnoses of “mucinous cystadenocarcinoma,” including our own previous report, 2 and thus the 5-year survival rates after resection. In our 57-year experience at a large tertiary referral center of 84 fully documented, histologically confirmed cystic mucinous neoplasms of the pancreas that were surgically resected for cure, only 7 fulfilled the criteria of objective tissue invasion (i.e., cystadenocarcinoma). Moreover, only one of six patients (one died in the postoperative period) survived 5 years without recurrent cystadenocarcinoma, and that lone patient had but a single, minute microscopic focus of invasion. This dismal survival rate is consistent with a report from the Lahey Clinic in 1989 1 but differs radically from other reports addressing curative resections of mucinous cystadenocarcinomas. 2–5 Thus, we suggest that cystic mucinous neoplasms of the pancreas that contain documented evidence of tissue invasion do not have a favorable natural history or cure rate after complete resection; rather, they behave like ductal adenocarcinoma of the pancreas, with a dismal prognosis and a disappointing 5-year survival rate (<20%).

The results of our study and the histologic classification we used offer potentially important therapeutic data to manage most patients with mucinous cystic neoplasms of the pancreas. First, if the lesion is completely resected and no area of tissue invasion is seen on careful, complete histologic review, our long-term follow-up of 77 such patients showed no recurrence, either in the 54 patients with benign cystadenoma or in the 23 patients with noninvasive proliferative cystic mucinous neoplasms. These results strongly suggest that a regular, oncologic-type follow-up program with surveillance using imaging tests or serum tumor markers may not be necessary to search for local recurrence or for a new second primary neoplasm. Such an approach would save money and eliminate worry on the patient’s part.

This approach disagrees with the results of a recent report by Thompson et al 10 from the Armed Forces Institute of Pathology. This group carefully reviewed the clinicopathologic correlation of 130 mucinous cystic neoplasms referred to their institute. They believed that all should be considered as mucinous cystadenocarcinomas of low-grade malignant potential because recurrence developed after resection in 7% of the 60 patients with a mucinous cystic neoplasm without atypia. In addition, they found “stromal invasion” in 60% of the specimens considered as harboring some element of atypia, a much higher incidence than in our series. This experience is obviously somewhat selected and may not represent the more usual spectrum of disease that would occur at a single institution, as in our study. Nevertheless, their findings differ markedly from ours.

Second, if tissue invasion is present, some form of adjuvant therapy should be strongly considered despite a “curative” resection, even if there are no nodal metastases. Although the efficacy of neoadjuvant or postoperative adjuvant chemotherapy or radiation therapy for mucinous cystadenocarcinomas of the pancreas is unknown, there are two reports suggesting possible benefit of chemoradiation therapy 14,15; however, these reports are anecdotal at best.

Third, our study does not necessarily provide support for the use of a nonanatomical resection or “enucleation” of cystic mucinous neoplasms of the pancreas. Many of these lesions have peripherally based small daughter cysts contiguous with the primary lesion that might be left in situ with an attempt at enucleation. In addition, there is evidence to suggest that the pancreatic ducts just outside the periphery of the neoplasm harbor K-ras mutations, a proposed precursor to malignant transformation (personal communication, Andrew Warshaw, MD, Massachusetts General Hospital, 1999). Also, we treated 13 patients with benign cystadenomas with enucleation; however, complications related directly to the enucleation (pancreatic fistula, pseudocyst) developed in 4 of these 13 patients. In contrast, a spleen-preserving distal pancreatectomy might be considered in selected patients.

In summary, our long-term follow-up of 84 patients with completely resected cystic mucinous neoplasms of the pancreas has shown that only a minority (7/84) had a true cystadenocarcinoma (tissue invasion), and as a group those with invasion have a poor prognosis (only one of six eligible patients lived >5 years). In contrast, most of the patients amenable to complete resection, according to our proposed histopathologic classification, have either a benign cystadenoma or a noninvasive proliferative mucinous cystic neoplasm that if completely resected does not recur locally or distantly. Thus, we do not believe this latter group harbors a malignancy or a neoplasm at risk for recurrence, and a formal follow-up with surveillance imaging tests and serum markers is not necessary.

Acknowledgment

The authors thank Deborah I. Frank for her assistance in manuscript preparation.

Footnotes

Correspondence: Michael G. Sarr, MD, Division of Gastroenterologic & General Surgery, Gastroenterology Research Unit, Mayo Clinic, 200 First St. SW, Rochester, MN 55905.

Reprints will not be available.

Accepted for publication April 5, 1999.

References

1. Katoh H, Russi RL, Braasch JW, Munson JL, Shimozawa E, Tanabe T. Cystadenoma and cystadenocarcinoma of the pancreas. Hepatogastroenterology 1989; 36:424–430. [PubMed]
2. Hodgkinson DJ, ReMine WH, Weiland LH. A clinicopathologic study of 21 cases of pancreatic cystadenocarcinoma. Ann Surg 1978; 188:679–684. [PMC free article] [PubMed]
3. Talamini MA, Pitt HA, Hruban RH, Boitnott JK, Coleman JA, Cameron JL. Spectrum of cystic tumors of the pancreas. Am J Surg 1992; 163:117–124. [PubMed]
4. Delcore R, Thomas JH, Forster J, Hermreck AS. Characteristics of cystic neoplasms of the pancreas and results of aggressive surgical treatment. Am J Surg 1992; 164:437–442. [PubMed]
5. Siech M, Tripp K, Schmidt-Rohlfiing B, et al. Cystic tumors of the pancreas: diagnostic accuracy, pathologic observations, and surgical consequences. Langenbecks Arch Surg 1998; 383:56–61. [PubMed]
6. Warshaw AL, Compton CC, Lewandrowski K, Cardenosa G, Muller PR. Cystic tumors of the pancreas: new clinical, radiologic, and pathologic observations in 67 patients. Ann Surg 1990; 212:432–445. [PMC free article] [PubMed]
7. Albores-Saavedra J, Angeles-Angeles A, Nadji M, Henson DE, Alvarez L. Mucinous cystadenocarcinoma of the pancreas: morphologic and immunocytotechnical observations. Am J Surg Pathol 1987; 11:11–29. [PubMed]
8. Compagno J, Oertel JE. Microcystic adenomas of the pancreas. Glycogen-rich cystadenomas. Am J Clin Pathol 1978; 69:281–298. [PubMed]
9. Compagno J, Oertel JE. Mucinous cystic neoplasms of the pancreas with overt and latent malignancy (cystadenocarcinoma and cystadenoma). A clinicopathologic study of 41 cases. Am J Clin Pathol 1978; 69:573–580. [PubMed]
10. Thompson LDR, Becker RC, Przygodzki RM, Adair CF, Heffess CS. Mucinous cystic neoplasm (mucinous) cystadenocarcinoma of low-grade malignant potential) of the pancreas: a clinicopathologic study of 130 cases. Am J Surg Pathol 1999; 23:1–16. [PubMed]
11. Pyke CM, van Heerden JA, Colby TV, Sarr MG, Weaver AL. The spectrum of serous cystadenoma of the pancreas. Ann Surg 1992; 215:132–139. [PMC free article] [PubMed]
12. Alpert LC, Truong LD, Bossart MI, Spjut HJ. Microcystic adenoma (serous cystadenoma) of the pancreas: a study of 14 cases with immunohistochemical and electron-microscopic correlation. Am J Surg Pathol 1988; 12:251–263. [PubMed]
13. Wood D, Silberman AW, Heifetz L, Memsic L, Shabot MM. Cystadenocarcinoma of the pancreas: neo-adjuvant therapy and CEA monitoring. J Surg Oncol 1990; 43:56–60. [PubMed]
14. George DH, Murphy F, Michalski R, Ulmer BG. Serous cystadenocarcinoma of the pancreas: a new entity? Am J Surg Pathol 1989; 13:61–66. [PubMed]
15. Doberstein C, Kirchner R, Gordon L, Silberman AW, Morgenstern L, Shapiro S. Cystic neoplasms of the pancreas. Mt Sinai J Med 1990; 57:102–105. [PubMed]

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

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...