Search

Purpose: Effective new markers of pancreatic carcinoma are urgently needed. In a previous analysis of gene expression in pancreatic adenocarcinoma using serial analysis of gene expression (SAGE), we found that the tag for the mesothelin mRNA transcript was present in seven of eight SAGE libraries derived from pancreatic carcinomas but not in the two SAGE libraries derived from normal pancreatic duct epithelial cells. In this study, we evaluate the potential utility of mesothelin as a tumor marker for pancreatic adenocarcinoma., Experimental Design: Mesothelin mRNA expression was evaluated in pancreatic adenocarcinomas using reverse-transcription PCR (RT-PCR) and in situ hybridization, whereas mesothelin protein expression was evaluated by immunohistochemistry., Results: Using an online SAGE database (http://www.ncbi.nlm.gov/SAGE), we found the tag for mesothelin to be consistently present in the mesothelioma, ovarian cancer, and pancreatic cancer libraries but not in normal pancreas libraries. Mesothelin mRNA expression was confirmed by in situ hybridization in 4 of 4 resected primary pancreatic adenocarcinomas and by RT-PCR in 18 of 20 pancreatic cancer cell lines, whereas mesothelin protein expression was confirmed by immunohistochemistry in all 60 resected primary pancreatic adenocarcinomas studied. The adjacent normal pancreas in these 60 cases did not label, or at most only rare benign pancreatic ducts showed weak labeling for mesothelin., Conclusions: Mesothelin is a new marker for pancreatic adenocarcinoma identified by gene expression analysis. Mesothelin overexpression in pancreatic adenocarcinoma has potential diagnostic, imaging, and therapeutic implications.

Purpose: SMAD4 (also called Dpc4) is a tumor suppressor in the TGF-beta signaling pathway that is genetically inactivated in approximately 55% of all pancreatic adenocarcinomas. We investigated whether prognosis after surgical resection for invasive pancreatic adenocarcinoma is influenced by SMAD4 status., Experimental Design: Using immunohistochemistry, we characterized the SMAD4 protein status of 249 pancreatic adenocarcinomas resected from patients who underwent pancreaticoduodenectomy (Whipple resection) at The Johns Hopkins Hospital, Baltimore, MD, between 1990 and 1997. The SMAD4 gene status of 56 of 249 (22%) pancreatic carcinomas was also determined. A multivariate Cox proportional hazards model assessed the relative risk of mortality associated with SMAD4 status, adjusting for known prognostic variables., Results: Patients with pancreatic adenocarcinomas with SMAD4 protein expression had significantly longer survival (unadjusted median survival was 19.2 months as compared with 14.7 months in patients with pancreatic cancers lacking SMAD4 protein expression; P = 0.03). This SMAD4 survival benefit persisted after adjustment for prognostic factors including tumor size, margins, lymph node status, pathological stage, blood loss, and use of adjuvant chemoradiotherapy. The relative hazard of mortality for cancers lacking SMAD4 after adjusting for other prognostic factors was 1.36 (95% confidence interval, 1.01-1.83; P = 0.04)., Conclusion: Patients undergoing Whipple resection for pancreatic adenocarcinoma survive longer if their cancers express SMAD4.

The INK4a locus encodes two structurally unrelated tumor suppressor proteins, p16(INK4a) and p14(ARF). Although the former is one of the most common targets for inactivation in human neoplasia, the frequency of p14(ARF) abrogation is not established. We have developed an immunohistochemical assay that allows the evaluation of p14(ARF) expression in formalin-fixed, paraffin-embedded tissues, using commercially available antibodies. p14(ARF) positive cells showed nuclear/nucleolar staining, which was absent in all cell lines and tumors with homozygous deletions of the INK4a gene. The assay was applied to 34 paraffin-embedded cell buttons, 30 non-small cell lung cancers and 28 pancreatic carcinomas, and the staining results were correlated with p16(INK4a) expression. Loss of p14(ARF) expression was common but less frequent than down-regulation of p16(INK4a) (53% versus 76% of all specimens). The p14(ARF) and p16(INK4a) expression pattern was concordant in 65 of 92 cases (71%). Significantly, 24 cases were p16(INK4a)-/p14(ARF)+, while the opposite staining pattern was observed in three cases, consistent with the notion that the two proteins have nonredundant functions. The immunohistochemical assay described here may facilitate studies on the prevalence and significance of aberrant p14(ARF) expression in human tumors.

Until recently, pancreatic cancer was a poorly understood disease. Research in the past decade has shown conclusively, however, that pancreatic cancer is primarily genetic in nature. Inactivation with a variety of tumor-suppressor genes such as p16, DPC4, and p53, coupled with activation of oncogenes such as K-ras, are a few of the mutations that trigger the growth of cancerous cells. Understanding these mutations is critical to a better understanding of familial pancreatic cancer and to the development of gene-based screening tests and therapies. In this article, we review the genetic alterations identified in pancreatic cancer and provide examples of how this information can be applied to patient care.

Pancreatic intraepithelial neoplasia (PanIN) is thought to be a precursor lesion of infiltrating pancreatic ductal adenocarcinoma (IPA). DPC4 is a tumor-suppressor gene on chromosome 18q21.1 and is inactivated in approximately 55% of IPAs. Recently, immunohistochemical labeling using a monoclonal antibody to the Dpc4 protein has been shown to mirror DPC4 genetic status in invasive adenocarcinomas of the pancreas. In the present study, we examined the role of Dpc4 loss in neoplastic progression and recurrence. Two cases in which a PanIN clinically progressed to an invasive adenocarcinoma and a third of a patient with IPA of the head of the pancreas who later developed invasive adenocarcinoma in the tail of the pancreas were studied using Dpc4 immunolabeling. The first patient underwent pancreatic resection, which revealed PanIN-3 that lacked Dpc4 expression, and the patient developed an invasive pancreatic ductal carcinoma 10 years later that shared this loss of expression. The second patient had a pancreaticoduodenectomy for recurrent pancreatitis, and the resected pancreas contained PanIN-3 with intact Dpc4 expression. Seventeen months later, the patient developed an invasive adenocarcinoma of the distal pancreas that also had intact Dpc4 expression. In the third case, the patient underwent pancreaticoduodenectomy for an invasive ductal adenocarcinoma with negative margins. This carcinoma lacked Dpc4 expression. Three years later, resection of the pancreatic tail showed a second invasive adenocarcinoma. The cancer in the tail of the gland showed intact Dpc4 expression, suggesting it represented a second primary tumor, not a recurrence. We conclude that Dpc4 expression in PanIN can be predictive of Dpc4 expression in the subsequent invasive ductal adenocarcinoma. Additionally, Dpc4 expression can be used to differentiate recurrent or persistent adenocarcinoma from a second primary adenocarcinoma., (Copyright 2001 by W.B. Saunders Company.)

Serial analysis of gene expression (SAGE) can be used to quantify gene expression in human tissues. Comparison of gene expression levels in neoplastic tissues with those seen in nonneoplastic tissues can, in turn, identify novel tumor markers. Such markers are urgently needed for highly lethal cancers like pancreatic adenocarcinoma, which typically presents at an incurable, advanced stage. The results of SAGE analyses of a large number of neoplastic and nonneoplastic tissues are now available online, facilitating the rapid identification of novel tumor markers. We searched an online SAGE database to identify genes preferentially expressed in pancreatic cancers as compared with normal tissues. SAGE libraries derived from pancreatic adenocarcinomas were compared with SAGE libraries derived from nonneoplastic tissues. Three promising tags were identified. Two of these tags corresponded to genes (lipocalin and trefoil factor 2) previously shown to be overexpressed in pancreatic carcinoma, whereas the third tag corresponded to prostate stem cell antigen (PSCA), a recently discovered gene thought to be largely restricted to prostatic basal cells and prostatic adenocarcinomas. PSCA was expressed in four of the six pancreatic cancer SAGE libraries, but not in the libraries derived from normal pancreatic ductal cells. We confirmed the overexpression of the PSCA mRNA transcript in 14 of 19 pancreatic cancer cell lines by reverse transcription-PCR, and using immunohistochemistry, we demonstrated PSCA protein overexpression in 36 of 60 (60%) primary pancreatic adenocarcinomas. In 59 of 60 cases, the adjacent nonneoplastic pancreas did not label for PSCA. PSCA is a novel tumor marker for pancreatic carcinoma that has potential diagnostic and therapeutic implications. These results establish the validity of analyses of SAGE databases to identify novel tumor markers.

Background: Biliary tract carcinomas are clinically heterogeneous. It is not known if molecular heterogeneity underlies the clinical differences., Methods: The authors evaluated 128 bile duct carcinomas, 88 of the distal common bile duct and 40 of more proximal origin (28 perihilar carcinomas, 12 intrahepatic carcinomas), immunohistochemically for abnormalities in the expression of the products of the DPC4 and p53 tumor-suppressor genes. Prognostic factors were evaluated in the series of distal bile duct carcinomas for which follow-up information was available., Results: The authors found that a significantly higher percentage of distal bile duct carcinomas (55%) demonstrated loss of DPC4 expression than did the proximal bile duct carcinomas (15%; P < 0.001). They also found that a significantly higher percentage of the distal tumors abnormally expressed the p53 gene product (51% vs. 26%; P < 0.001). Among the distal common bile duct carcinomas, the presence of poorly differentiated histology correlated with decreased survival in multivariate analysis, while labeling for p53 or Dpc4, margin status, lymph node status, and tumor dimension did not correlate significantly with survival., Conclusions: These results demonstrate that abnormalities in DPC4 and p53 gene expression are frequent in distal common bile duct carcinomas, just as they are in pancreatic ductal adenocarcinoma, suggesting that these two tumor types might share a similar molecular pathogenesis. They also show that proximal and distal bile duct carcinomas have different patterns of inactivation of tumor-suppressor genes, indicating that they often arise through different molecular mechanisms likely reflecting their differing etiologies., (Copyright 2001 American Cancer Society.)

It has been estimated that familial aggregation and genetic susceptibility play a role in as many as 10% of patients with pancreatic cancer (PC). The quantified prospective risk of PC among first-degree relatives of PC patients has not been investigated. Families enrolled in the National Familial Pancreas Tumor Registry (NFPTR) prior to September 1, 1998 were followed to estimate the risk and incidence of PC among first-degree relatives of patients with PC. Analyses were performed separately on kindreds with at least two first-degree relatives with PC (familial pancreatic carcinoma (PC); n = 150) at the time the kindred was enrolled in the NFPTR and on kindreds without a pair of affected first-degree relatives (sporadic PC; n = 191). A subanalysis was performed on familial PC kindreds containing three or more affected members at the time of enrollment in the NFPTR (n = 52). Risk was estimated by comparing observed new cases of PC during the observation period with expected numbers based on the United States population-based Surveillance, Epidemiology and End Results program data. Incidence was estimated using person-years risk analyses. During the observational period, six incident PCs developed in the first-degree relatives: two in the sporadic PC kindreds, and four in the familial PC kindreds. The PC risk in the sporadic PC kindreds was not significantly greater than expected [observed/expected = 6.5 (95% CI = 0.78-23.3)] with an incidence rate of 24.5/10(5)/ year. There was a significantly increased 18-fold risk (95% CI = 4.74-44.5) of PC among first-degree relatives in familial PC kindreds, with an incidence of 76.0/10(5)/year. In the subset of familial PC kindreds with three or more affected family members at the time of enrollment, there was a 57-fold (95% CI = 12.4-175) increased risk of PC and an incidence of 301.4/10(5)/year compared with the Surveillance, Epidemiology and End Result age-adjusted incidence of PC in the U.S. (8.8/10(5)/year). When stratified by age, the risk was largely confined to relatives over the age of 60. This study is the first analysis of incident PC occurring in familial PC kindreds. The risk and incidence of PC is exceptionally high among at-risk first-degree relatives in familial PC kindreds in which at least three first-degree relatives have already been diagnosed with PC. Familial PC kindreds are a reasonable high-risk group for PC screening and chemoprevention research.

Endogenous fatty acid synthesis has been observed in some rapidly proliferating cells and tissues, both normal and neoplastic, and probably supports membrane synthesis. Sterol regulatory element binding proteins (SREBPs) are transcription factors that regulate the expression of genes for both cholesterol and fatty acid synthesis. The inactive precursor form resides in cytoplasmic membranes. Intracellular lipid depletion triggers proteolytic cleavage of SREBP, allowing the amino terminus to enter the nucleus and activate the expression of enzymes, including acetyl-CoA carboxylase (ACC) and fatty acid synthase (FAS), major biosynthetic enzymes for fatty acid synthesis. The expression patterns of ACC, FAS, SREBP, and Ki-67 in fetal tissues were compared to determine whether SREBP is likely to participate in the regulation of proliferation-associated fatty acid synthesis during fetal growth. Tissues from 22 fetuses, 12 first-trimester and 10 second-trimester (range 7.0 to 21.6 weeks), were studied. Serial 5-microm sections were stained with antibodies to ACC, FAS, SREBP, and Ki-67 and were compared. ACC, FAS, SREBP, and Ki-67 were coexpressed in the proliferative compartments of the intestines, skin, and kidney. ACC, FAS, and Ki-67 were coexpressed with little SREBP in lung and cytotrophoblast. SREBP, ACC, and FAS were coexpressed without Ki-67 in hepatocytes, ganglion cells, and intermediate trophoblast. The close linkage of SREBP, ACC, FAS, and Ki-67 in some proliferating fetal tissues suggests that in these tissues SREBP participates in the transcriptional regulation of lipogenic genes during proliferation. SREBP, ACC, and FAS coexpression without Ki-67 occurs in differentiated tissues that may synthesize fatty acids for other functions.

DPC4 (MADH4, SMAD4) is a nuclear transcription factor shown to be genetically inactivated in over half of infiltrating ductal adenocarcinomas of the pancreas. Immunohistochemical labeling for the DPC4 gene product using a monoclonal antibody has recently been shown to be an extremely sensitive and specific marker for DPC4 gene alterations in pancreatic adenocarcinomas. Mucinous cystic neoplasms (MCNs) are a biologically less aggressive subtype of pancreatic neoplasm that may show benign, borderline, or overtly malignant features. However, the role of DPC4 inactivation in the development of MCNs has not been examined. The immunohistochemical expression of Dpc4 protein was therefore examined in 36 mucinous cystic neoplasms using this previously characterized monoclonal antibody. The 36 mucinous cystic neoplasms studied included 23 adenomas, 1 tumor with borderline potential, 5 tumors with carcinoma in situ, and 7 invasive carcinomas. Twenty-nine (100%) of the 29 noninvasive mucinous cystic neoplasms strongly expressed Dpc4 in the neoplastic epithelium. In striking contrast, only one (14%) of seven infiltrating carcinomas expressed Dpc4 in the neoplastic epithelium (p = 0.0001). The adjacent stroma retained expression of this protein in all 36 cases. In invasive MCNs with loss of Dpc4 expression, areas of carcinoma in situ were identified in the same paraffin sections, and these areas of carcinoma in situ retained expression of Dpc4. The frequent loss of Dpc4 expression in invasive MCNs indicates that genetic inactivation of Dpc4 occurs late in the neoplastic progression of these tumors and suggests a relationship to the development of invasion.

DPC4 (MADH4, SMAD4) encodes a nuclear transcription factor shown to be genetically inactivated in over one-half of conventional infiltrating ductal adenocarcinomas of the pancreas. Intraductal papillary mucinous neoplasms (IPMNs) of the pancreas have been suggested to be distinct neoplasms with a significantly less aggressive course than conventional ductal adenocarcinomas of the pancreas, but molecular comparisons of these tumor types have previously been impaired by technical difficulties. Recently, immunohistochemical labeling for the DPC4 gene product has been shown to be an extremely sensitive and specific marker for DPC4 gene alterations in pancreatic adenocarcinomas. Therefore, we analyzed the immunohistochemical expression of Dpc4 protein in 79 IPMNs using a previously characterized monoclonal antibody. Twenty-nine of the IPMNs also had an associated infiltrating adenocarcinoma available for analysis. The labeling patterns observed were compared to those we have previously reported for conventional ductal carcinomas. All 79 of the intraductal components of the IPMNs strongly expressed Dpc4 protein. In 77 of the 79 cases (97%), the labeling was diffusely positive, and in 2 of the 79 (3%) the labeling was focally positive. Dpc4 expression was seen in 28 (97%) of the associated 29 invasive cancers. The one infiltrating carcinoma that showed loss of Dpc4 expression was associated with an intraductal component which showed focal loss of Dpc4 expression. The strong and almost universal expression of Dpc4 in IPMNs contrasts sharply with the loss of Dpc4 expression seen in approximately 30% of in situ adenocarcinomas of the pancreas (so-called pancreatic intraepithelial neoplasms, grade 3; P: < 0.001) and in 55% of pancreatic duct carcinomas (P: < 0.0001). Differences in Dpc4 expression between IPMNs and ductal carcinomas suggest a fundamental genetic difference in tumorigenesis, which may relate to the significantly better clinical outcomes observed for IPMNs.

Medullary carcinomas of the pancreas are a recently described, histologically distinct subset of poorly differentiated adenocarcinomas that may have a unique pathogenesis and clinical course. To further evaluate these neoplasms, we studied genetic, pathological, and clinical features of 13 newly identified medullary carcinomas of the pancreas. Nine (69%) of these had wild-type K-ras genes, and one had microsatellite instability (MSI). This MSI medullary carcinoma, along with three previously reported MSI medullary carcinomas, were examined immunohistochemically for Mlh1 and Msh2 expression, and all four expressed Msh2 but did not express Mlh1. In contrast, all of the medullary carcinomas without MSI expressed both Msh2 and Mlh1. Remarkably, the MSI medullary carcinoma of the pancreas in the present series arose in a patient with a synchronous but histologically distinct cecal carcinoma that also had MSI and did not express Mlh1. The synchronous occurrence of two MSI carcinomas suggests an inherited basis for the development of these carcinomas. Indeed, the medullary phenotype, irrespective of MSI, was highly associated with a family history of cancer in first-degree relatives (P < 0.001). Finally, one medullary carcinoma with lymphoepithelioma-like features contained Epstein-Barr virus-encoded RNA-1 by in situ hybridization. Therefore, because of medullary carcinoma's special genetic, immunohistochemical, and clinical features, recognition of the medullary variant of pancreatic adenocarcinoma is important. Only by classifying medullary carcinoma as special subset of adenocarcinoma can we hope to further elucidate its unique pathogenesis.

Infiltrating adenocarcinomas of the pancreas are believed to arise from histologically identifiable intraductal precursors [pancreatic intraepithelial neoplasias (PanINs)] that undergo a series of architectural, cytological, and genetic changes. The role of DPC4 tumor suppressor gene inactivation in this progression has not been defined. Immunohistochemistry for the Dpc4 protein in formalin-fixed, paraffin-embedded tissue is a sensitive and specific marker for DPC4 gene status, providing a tool to examine DPC4 status in these putative precursor lesions. A total of 188 PanINs were identified in 40 pancreata, 38 (95%) of which also contained an infiltrating adenocarcinoma. Sections containing these 188 duct lesions were labeled with a monoclonal antibody to Dpc4. All 82 flat (PanIN-1A), all 54 papillary (PanIN-1B), and all 23 atypical papillary (PanIN-2) intraductal lesions expressed Dpc4. In contrast, 9 of 29 (31%) severely atypical lesions (PanIN-3 lesions, carcinomas in situ) did not. The difference in Dpc4 expression between histologically low-grade (PanIN-1 and -2) and histologically high-grade (PanIN-3) duct lesions was statistically significant (P < 0.0001). In three cases, the pattern of Dpc4 expression in the PanIN-3 lesions did not match the pattern of expression in the associated infiltrating carcinomas, indicating that these high-grade lesions did not simply represent infiltrating carcinoma growing along benign ducts. Loss of Dpc4 expression occurs biologically late in the neoplastic progression that leads to the development of infiltrating pancreatic cancer, at the stage of histologically recognizable carcinoma.

Since their initial description, mucinous cystic neoplasms have been difficult to classify. This article attempts to clarify histological, clinical, and genetic criteria so that the pathologist can categorize each mucinous cystic neoplasm into 1 of 4 possible categories. Mucinous cystadenomas contain a single layer of mucin-producing, columnar epithelium lacking significant atypia. Borderline mucinous cystic neoplasms contain cells with moderate atypia. Mucinous cystic neoplasms with in situ carcinoma show significant architectural and cytological atypia. When invasive carcinoma is present in association with a mucinous cystic neoplasm, then the diagnosis of invasive mucinous cystadenocarcinoma should be made. The categorization of mucinous cystic neoplasms into these groups is essential because it accurately predicts outcome, provided that the tumor has been sampled and examined thoroughly. Completely removed mucinous cystadenomas, borderline mucinous cystic neoplasms, and mucinous cystic neoplasms with in situ carcinoma follow benign courses. Partial resection should be avoided as evidence suggests that mucinous cystic neoplasms can progress from adenomas to borderline lesions to carcinomas in situ to invasive carcinomas over time; partial resection should be avoided if possible. Modern molecular genetic techniques are helping to unravel the origins of rare variants of mucinous cystic tumors, such as the mucinous cystic tumor with an associated osteoclast-like giant cell tumor and the mucinous cystic tumor with sarcomatous stroma.

DPC4 (MADH4, SMAD4) is a tumor suppressor gene inactivated by allelic loss in approximately 55% of pancreatic adenocarcinomas. Unfortunately, it can be technically very difficult to detect the inactivation of DPC4 at the genetic level because genetic analyses require the microdissection of relatively pure samples of neoplastic and normal tissues. This is especially true for pancreatic adenocarcinomas, which elicit vigorous, non-neoplastic, stromal responses. Immunohistochemical labeling can overcome this hurdle because it preserves morphological information. We therefore studied the expression of the DPC4 gene product in 46 cancers, including 5 cancer cell lines by Western blot analysis and 41 primary periampullary adenocarcinomas by immunohistochemistry. The status of exons 1-11 of the DPC4 gene in all 46 of the cancers had been previously characterized at the molecular level, allowing us to correlate Dpc4 expression directly with gene status. Three cell lines had wild-type DPC4 genes, and Dpc4 expression was detected in all three by Western blot. The two cell lines with homozygously deleted DPC4 genes did not show Dpc4 protein by Western blot analysis. Immunohistochemical labeling revealed that 17 (94%) of the 18 primary adenocarcinomas with wild-type DPC4 genes expressed the DPC4 gene product, whereas 21 (91%) of 23 primary adenocarcinomas with inactivated DPC4 genes did not. Cases in which there was discordance between the immunohistochemical labeling and the genetic analyses were reanalyzed genetically, and we identified a deletion in exon 0 of DPC4 in one of these cases. This is the first report of a mutation in exon 0 of DPC4 in a pancreatic cancer. The contrast between the strong expression of Dpc4 by normal tissues and the loss of expression in the carcinomas was highlighted in several cases in which an infiltrating cancer was identified growing into a benign duct. These observations suggest that immunohistochemical labeling for the DPC4 gene product is an extremely sensitive and specific marker for DPC4 gene alterations in pancreatic carcinomas. The sensitivity and specificity of immunohistochemical labeling for Dpc4 in other periampullary carcinomas has yet to be determined.

The behavior of pancreatic mucinous cystic neoplasms has long been debated. Some authors contend that histologically benign neoplasms can recur and metastasize. We reviewed the gross and microscopic findings and outcomes of 61 mucinous cystic neoplasms diagnosed at The Johns Hopkins Hospital from March 20, 1984 to July 8, 1998. Each neoplasm was placed into one of four categories based on complete histologic examination: invasive mucinous cystadenocarcinoma, mucinous cystic neoplasm with in situ carcinoma, borderline mucinous cystic neoplasm, and mucinous cystadenoma. Neoplasms in the latter three categories were included only if they were entirely resected and completely examined. Patient outcomes were obtained from hospital records and patient and physician follow-up. Twenty (33%) of the patients had invasive mucinous cystadenocarcinomas, and they had 2- and 5-year disease-specific survival rates of 67% and 33% (mean follow-up of survivors, 4.2 years), respectively. Nine (15%) patients had mucinous cystic neoplasms with in situ carcinoma (mean follow-up of survivors, 4.1 years). Five (8.2%) patients had borderline mucinous cystic neoplasms (mean follow-up of survivors, 5.6 years). Twenty-seven (44%) patients had mucinous cystadenomas (mean follow-up of survivors, 5.1 years). No mucinous cystadenoma, borderline mucinous cystic neoplasm, or mucinous cystic neoplasm with in situ carcinoma recurred or metastasized. No patient with the diagnosis of mucinous cystadenoma, borderline mucinous cystic neoplasm, or mucinous cystic neoplasm with in situ carcinoma died of disease. The difference in disease-specific survival rates between patients with invasive mucinous cystadenocarcinomas and those with noninvasive tumors was significant (p < 0.0001, log-rank test). One case, originally showing only benign histology on incisional biopsy, contained foci of invasive carcinoma on complete resection. Completely resected and entirely examined mucinous cystadenomas, borderline mucinous cystic neoplasms, and mucinous cystic neoplasms with in situ carcinoma follow benign courses. Because invasive carcinoma can be focal, failure to study an entire mucinous cystic neoplasm may result in the miscategorization of a malignant neoplasm as benign.

DPC4/SMAD4 is a candidate tumor suppressor gene with a strikingly high frequency of gene alterations in pancreatic cancer that suggests a discrete role for DPC4 in these tumors. DPC4 tumor-suppressive function has been implicated to mediate the transforming growth factor-beta (TGFbeta)-suppressive pathway; however, in a DPC4-null pancreatic cancer cell line, TGFbeta growth-inhibitory and transcriptional responses were found to be DPC4-independent. This was observed within native cells having a natural homozygous deletion and in clones engineered for stable expression of wild-type DPC4 integrated into the genome. This observation contrasted with the absolute DPC4 dependence of TGFbeta responses in a breast cancer cell line studied in parallel. This growth-inhibitory response to TGFbeta in DPC4-null cells relied on an intact ras effector pathway. These data further suggest a major categorization of TGFbeta responses into DPC4-dependent and -independent signaling pathways and specifically suggest that disruption of the TGFbeta-independent signal might be a basis of selection for the emergence of DPC4 alterations during tumorigenesis in the pancreas and other sites.

Background: An understanding of incipient pancreatic neoplasia is an essential foundation for the future development of effective screening tests for pancreatic cancer. Only when we understand early pancreatic neoplasms will we be able to detect tumors that are curable with surgical resection., Method: Two approaches have helped define incipient pancreatic neoplasia. First, the histologic examination of pancreata has helped identify the most common histologic lesions in pancreatic tissues adjacent to infiltrating carcinomas. The assumption is that some of these lesions represent the precursors to the infiltrating cancers. Second, advances in molecular genetics now make it possible to define the genetic alterations present in small lesions. The demonstration that a suspected precursor lesion and an infiltrating pancreatic carcinoma share the same genetic alterations would help establish that the lesion is indeed a precursor to infiltrating pancreatic carcinoma., Results: A spectrum of intraductal proliferations in the pancreas has been found associated with infiltrating adenocarcinoma of the pancreas. These lesions, called "duct lesions," have even been identified in pancreata years before patients develop infiltrating carcinoma. Molecular genetic analysis of these lesions has revealed that they frequently harbor many of the same alterations present in infiltrating carcinoma of the pancreas. These alterations include activation of K-ras and inactivation of p16, p53 and, rarely BRCA2., Conclusions: From these morphologic and molecular observations, we can now develop a progression model for the development of infiltrating carcinoma of the pancreas. Infiltrating carcinomas of the pancreas may arise from pancreatic duct lesions, and the progression of duct lesions to infiltrating cancer is associated with the accumulation of generalized and specific genetic alterations.

Pancreatic adenocarcinoma develops from histologically identifiable intraductal lesions that undergo a series of architectural, cytological, and genetic changes. Limited genetic evidence recently suggested that the p16 gene plays a role in the progression of these "duct lesions." Duct lesions were identified in pancreata from 33 pancreaticoduodenectomies performed for infiltrating adenocarcinoma. All of these infiltrating adenocarcinomas were previously shown to contain alterations in the p16 gene or its promoter. Monoclonal and polyclonal anti-p16 antibodies were used for histological immunodetection. One hundred twenty-six duct lesions were identified. Nine (30%) of 30 flat, 4 (27%) of 15 papillary, 37 (55%) of 67 papillary with atypia, and 10 (71%) of 14 carcinoma in situ duct lesions showed loss of p16 expression. These included 30% of the flat lesions versus 53% of the nonflat lesions and 29% of the nonatypical lesions versus 58% of the atypical lesions. For both comparisons, the differences were statistically significant (P = 0.036 and P = 0.003, respectively). Loss of p16 expression occurs more frequently, but not exclusively, in higher-grade duct lesions. These data support the hypothesis that pancreatic duct lesions are neoplastic and that they represent the precursors of infiltrating adenocarcinoma. Immunohistochemical detection of p16 provides a new technology to study the genetic alterations in and stages of progression of large numbers of morphologically defined pancreatic duct lesions.

Although many have lumped nearly 20 different neoplasms under the umbrella term "cancer of the pancreas," each of these neoplasms is pathologically and clinically distinct. In addition, each may require a specific treatment and result in a different outcome. Understanding the pathology of pancreas cancer, therefore, forms the cornerstone for rational treatment and prognostication. This article describes the pathology of a number of primary, metastatic, and systemic cancers that can involve the pancreas. The clinical relevance of each gross and histologic tumor feature is emphasized.

It can be difficult to distinguish benign bile duct proliferations (BDPs) from well-differentiated metastatic peripancreatic adenocarcinomas on histological grounds alone. Most peripancreatic carcinomas harbor activating point mutations in codon 12 of the K-ras oncogene, suggesting that K-ras mutational status may provide a molecular basis for distinguishing BDPs from liver metastases. The ability of tests for mutations in codon 12 of K-ras to make this distinction was examined in a two-part study. In the first part we determined the K-ras mutational status of 56 liver lesions and 48 primary peripancreatic adenocarcinomas obtained from 48 patients. In the second part of this study an additional 45 liver lesions were studied. In the first 48 patients, activating point mutations in codon 12 of K-ras were detected in 28 (61%) of the 46 primary carcinomas, in 8 (100%) of 8 liver metastases, in 2 (6.5%) of 31 BDPs, and in none (0%) of 14 liver granulomas. Three BDPs and two primary carcinomas did not amplify. To further estimate the prevalence of K-ras mutations in BDPs we analyzed an additional series of 45 mostly incidental BDPs for K-ras mutations. Three (6.7%) of these 45 harbored K-ras mutations. These results suggest that K-ras mutations may be useful in distinguishing BDPs from metastases in the liver; however, there is some overlap in the mutational spectra of BDPs and pancreatic carcinomas.