Periampullary adenocarcinoma: analysis of 5-year survivors

OBJECTIVE

This single-institution experience retrospectively reviews the outcomes in a group of patients treated 5 or more years ago by pancreaticoduodenectomy for periampullary adenocarcinoma.

SUMMARY BACKGROUND DATA

Controversy exists regarding the benefit of resection for periampullary adenocarcinoma, particularly for pancreatic tumors. Many series report only Kaplan-Meier actuarial 5-year survival rates. There are believed to be discrepancies between the actuarial 5-year survival data and the actual 5-year survival rates.

METHODS

From April 1970 through May 1992, 242 patients underwent pancreaticoduodenal resection for periampullary adenocarcinoma at The Johns Hopkins Hospital. Follow-up was complete through May 1997. All pathology specimens were reviewed and categorized. Actual 5-year survival rates were calculated. The demographic, intraoperative, pathologic, and postoperative features of patients surviving > or =5 years were compared with those of patients who survived <5 years.

RESULTS

Of the 242 patients with resected periampullary adenocarcinoma, 149 (62%) were pancreatic primaries, 46 (19%) arose in the ampulla, 30 (12%) were distal bile duct cancers, and 17 (7%) were duodenal cancers. There was a 5.3% operative mortality rate during the 22 years of the review, with a 2% operative mortality rate in the last 100 patients. There were 58 5-year survivors, 28 7-year survivors, and 7 10-year survivors. The tumor-specific 5-year actual survival rates were pancreatic 15%, ampullary 39%, distal bile duct 27%, and duodenal 59%. When compared with patients who did not survive 5 years, the 5-year survivors had a significantly higher percentage of well-differentiated tumors (14% vs. 4%; p = 0.02) and higher incidences of negative resection margins (98% vs. 73%, p < 0.0001) and negative nodal status (62% vs. 31%, p < 0.0001). The tumor-specific 10-year actuarial survival rates were pancreatic 5%, ampullary 25%, distal bile duct 21%, and duodenal 59%.

CONCLUSIONS

Among patients with periampullary adenocarcinoma treated by pancreaticoduodenectomy, those with duodenal adenocarcinoma are most likely to survive long term. Five-year survival is less likely for patients with ampullary, distal bile duct, and pancreatic primaries, in declining order. Resection margin status, resected lymph node status, and degree of tumor differentiation also significantly influence long-term outcome. Particularly for patients with pancreatic adenocarcinoma, 5-year survival is not equated with cure, because many patients die of recurrent disease >5 years after resection.

Pancreatic adenocarcinomas with DNA replication errors (RER+) are associated with wild-type K-ras and characteristic histopathology. Poor differentiation, a syncytial growth pattern, and pushing borders suggest RER+

The clinical and pathological features of carcinomas of the pancreas with DNA replication errors (RER+) have not been characterized. Eighty-two xenografted carcinomas of the pancreas were screened for DNA replication errors using polymerase chain reaction amplification of microsatellite markers. Cases with microsatellite instability in at least two markers of a minimum of five tested were considered RER+. RER status was correlated with histological appearance, karyotype of the carcinomas when available, K-ras mutational status, and patient outcome. Three (3.7%) of the eighty-two carcinomas were RER+. In contrast to typical gland-forming adenocarcinomas of the pancreas, all three RER+ carcinomas were poorly differentiated and had expanding borders and a prominent syncytial growth pattern. Neither a Crohn's-like lymphoid infiltrate nor extracellular mucin production were prominent. Ductal adenocarcinomas of the pancreas typically contain a mutant K-ras gene, yet all three RER+ carcinomas had wild-type K-ras. One of the three RER+ carcinomas was karyotyped and showed a near diploid pattern. All three of the RER+ tumors were removed via Whipple resection. One of the three patients is free of disease 16 months after pancreaticoduodenectomy, one is alive and free of tumor at 52 months but developed two colon carcinomas during this period, and the third died of pancreatic cancer at 4 months. None of the three patients had a family history of colorectal carcinoma. A review of the K-ras wild-type carcinomas in a previously characterized series of pancreatic carcinomas with known K-ras mutational status identified two additional cancers with poor differentiation, a syncytial growth pattern, and pushing borders. Both of the cancers were diploid and both patients were longterm survivors (over 5 years). The inclusion of such patients in previous prognostic studies of pancreas cancer may explain the failure of histological grade to be a predictor of prognosis. These data suggest that DNA replication errors occur in a small percentage of resected carcinomas of the pancreas and that wild-type K-ras gene status and a medullary phenotype characterized by poor differentiation, and expanding pattern of invasion, and syncytial growth should suggest the possibility of DNA replication errors in carcinomas of the pancreas.

Should pancreaticoduodenectomy be performed in octogenarians?

As the population in the United States ages, an increasing number of elderly patients may be considered for pancreaticoduodenal resection. This high-volume, single-institution experience examines the morbidity, mortality, and long-term survival of 727 patients undergoing pancreaticoduodenectomy between December 1986 and June 1996. Outcomes of patients 80 years of age and older (n = 46) were compared to those of patients younger than 80 years. In these older patients, pancreaticoduodenectomy was performed for pancreatic adenocarcinoma (n = 25; 54%), ampullary adenocarcinoma (n = 9; 20%) distal bile duct adenocarcinoma (n = 5; 11%), duodenal adenocarcinoma (n = 2; 4%), cystadenocarcinoma; (n = 2; 4%), cystadenoma (n = 1; 2%), and chronic pancreatitis (n = 2; 4%). When compared to the 681 concurrent patients younger than 80 years who were undergoing pancreaticoduodenectomy, the two groups were statistically similar with respect to sex, race, intraoperative blood loss, and type of pancreaticoduodenectomy performed. Patients 80 years of age or older had a shorter median operative time (6.4 hours vs. 7.0 hours; P = 0.02) but a longer postoperative length of stay (median = 15 days vs. 13 days; P = 0.01) and a higher complication rate (57% vs. 41%; P = 0.05) when compared to their younger counterparts. Pancreaticoduodenectomy in the older group resulted in a 4.3% perioperative mortality rate compared to 1.6% in the younger group (P = NS). In the subset of patients undergoing pancreaticoduodenectomy for periampullary adenocarcinoma (n = 495), patients 80 years of age or older (n = 41) had a median survival of 32 months and a 5-year survival rate of 19%, compared to 20 months and 27%, respectively, in patients younger than 80 years (n = 454; P = 0.77). These data demonstrate that pancreaticoduodenectomy can be performed safely in selected patients 80 years of age or older, with morbidity and mortality rates approaching those observed in younger patients. Based on these data, age alone should not be a contraindication to pancreaticoduodenectomy.

Development and characterization of a cytokine-secreting pancreatic adenocarcinoma vaccine from primary tumors for use in clinical trials

Preclinical studies with murine tumor models have demonstrated that tumor cell vaccines engineered to secrete certain cytokines in a paracrine fashion elicit systemic immune responses capable of eliminating small amounts of established tumor. In particular, tumors that express the cytokine GM-CSF produce potent systemic antitumor immune responses against poorly immunogenic murine tumors. These results have encouraged the development of paracrine-cytokine secreting tumor vaccines for gene therapy of human cancer. GM-CSF recruits professional antigen-presenting cells, which in turn activate effector T cells. These findings suggest that allogeneic as well as autologous tumor cells can be used as the tumor source for developing cancer vaccines. A major obstacle to creating genetically modified human allogeneic tumor vaccines is the absence of stable cell lines required for efficient gene transfer, because most human tumors isolated from primary surgical specimens fail to proliferate in long-term culture. We have developed a method for the routine generation of in vitro cell lines from primary tumors of the pancreas. This method overcomes the common problem of stromal and fibroblast overgrowth that can inhibit the in vitro expansion of many histologic types of tumors. In addition, we have analyzed 12 of these cell lines for cytokeritin and mutated K-ras expression to demonstrate that they derive from the original epithelial tumor tissue. The lines can be genetically modified to stably express the cytokine GM-CSF. These methods should be helpful to investigators attempting to establish cell lines from other histologic tumor types for the development of allogeneic genetically modified tumor vaccines.

DNA content and other factors associated with ten-year survival after resection of pancreatic carcinoma

BACKGROUND AND OBJECTIVES

The 5-year survival rates after resection of pancreatic carcinoma have recently increased and are predicted by tumor size, DNA content, and lymph node metastases at the time of resection. However, whether the 10-year survival rates have also increased and are similarly predicted by these factors is not known.

METHODS

The influence of preoperative imaging tests, alcohol consumption, cigarette smoking, K-ras mutations, anatomic location, details of surgical resection, pathologic findings, and tumor DNA content on survival was tested for 96 patients after a successful resection of a pancreatic carcinoma with 17 patients being followed for more than 5 years.

RESULTS

The 5- and 10-year patient survival rates were 18% and 3%, respectively. Univariate and multivariable analyses showed that tumor DNA content, pathologic tumor size, and lymph node metastases were the strongest prognostic indicators for long-term patient survival, although the importance of tumor size may diminish 2 or more years after resection. Surprisingly, the 11 patients with diploid carcinomas > or = 4 cm had an estimated 10-year survival rate of 36%.

CONCLUSION

These results show that the 10-year survival rate for pancreatic carcinoma remains very low, although the subset of patients with biologically favorable tumors has a prolonged survival and possible cure after resection.

DNA content and other factors associated with ten-year survival after resection of pancreatic carcinoma

BACKGROUND AND OBJECTIVES

The 5-year survival rates after resection of pancreatic carcinoma have recently increased and are predicted by tumor size, DNA content, and lymph node metastases at the time of resection. However, whether the 10-year survival rates have also increased and are similarly predicted by these factors is not known.

METHODS

The influence of preoperative imaging tests, alcohol consumption, cigarette smoking, K-ras mutations, anatomic location, details of surgical resection, pathologic findings, and tumor DNA content on survival was tested for 96 patients after a successful resection of a pancreatic carcinoma with 17 patients being followed for more than 5 years.

RESULTS

The 5- and 10-year patient survival rates were 18% and 3%, respectively. Univariate and multivariable analyses showed that tumor DNA content, pathologic tumor size, and lymph node metastases were the strongest prognostic indicators for long-term patient survival, although the importance of tumor size may diminish 2 or more years after resection. Surprisingly, the 11 patients with diploid carcinomas > or = 4 cm had an estimated 10-year survival rate of 36%.

CONCLUSION

These results show that the 10-year survival rate for pancreatic carcinoma remains very low, although the subset of patients with biologically favorable tumors has a prolonged survival and possible cure after resection.

The Quilty lesion enigma: focal apoptosis/necrosis and lymphocyte subsets in human cardiac allografts

Quilty lesions, as first described by Billingham in 1981, or 'Quilty Effect' (QE) are distinct endomyocardial mononuclear cell infiltrates that have been observed in human heart transplant recipients, as well as in experimental models of heart transplantation. In the present investigations, the pattern and extent of apoptosis (programmed cell death) and myocyte necrosis, as well as specific lymphocyte subsets in Quilty lesions was assessed. Endomyocardial biopsies obtained from 13 patients at 10-3362 days post-transplant were examined. Apoptosis, as identified by DNA nick end-labeling, was found in myocytes at the periphery of Quilty lesions in 11 of 13 cases (85%), and 'early' stages of myocyte necrosis, as demonstrated by specific staining with alpha light chain myosin monoclonal antibodies (mAb), was observed at the same sites in 10 of 13 cases (77%) of both Quilty type A and type B lesions. Apoptosis was not identified in the lymphocyte infiltrates of any of the lesions examined. Lymphocyte subsets were characterized using mAb for T cell receptor (CD3), for helper/inducer T cells (CD4), for cytotoxic/suppressor T cells (CD8) and for mature B cells (CD20). Immunostaining revealed separate clusters of T lymphocytes with less prevalent B cells within the Quilty lesions. CD4+ cells were found in larger numbers than CD8+ cells in all cases. Non-B, non-T large lymphocytes were occasionally present. Except for the extent of the cellular infiltrate, no major cytochemical lymphocyte distribution differences were found between Quilty type A and B lesions. Myocyte apoptosis and early necrosis at the periphery of Quilty lesions suggest that early myocyte injury occurring in B lesions may represent initial or 'abortive stages' of cardiac allograft rejection. Why these lesions do not progress to overt rejection indeed warrant further detailed studies.

K-ras mutations in the duodenal fluid of patients with pancreatic carcinoma

BACKGROUND

Many patients with carcinoma of the pancreas die because their disease is not detected until late in its course. Methods that detect these cancers earlier will improve patient outcome. Over 80% of pancreatic carcinomas contain mutations in codon 12 of the K-ras gene. Screening duodenal fluid for these mutations may lead to early detection of these cancers and assist in establishing a diagnosis of pancreatic carcinoma.

METHODS

Polymerase chain reaction (PCR), with and without restriction enzyme-mediated mutant enrichment, was performed on DNA isolated from duodenal fluid specimens from 61 patients who underwent pancreaticoduodenectomy (Whipple's operation) for either periampullary cancer or a benign condition of the pancreas. Representative sections of pancreas pathology (primary carcinoma, benign tumor, or chronic pancreatitis) from the patients with duodenal fluid specimens containing amplifiable DNA were also analyzed for K-ras mutations. Wild-type and mutant K-ras were detected by hybridization of the PCR products with K-ras codon 12 mutant and wild-type specific probes.

RESULTS

Seven of the 61 duodenal fluid specimens contained DNA that did not amplify. Thirteen (24% of the 54 duodenal fluid specimens with amplifiable DNA and 21% of the total of 61 specimens) contained activating point mutations at codon 12 of the K-ras gene. Mutations were detected in 13 of the 51 duodenal fluid specimens from patients with cancer (sensitivity, 25%), whereas mutations were not detected in any of the 9 amplifiable duodenal fluid specimens from patients with benign conditions of the pancreas (specificity, 100%). One duodenal fluid specimen from a patient with adenocarcinoma of the pancreas had two different K-ras mutations. DNA from three of the primary carcinomas did not amplify or was not available. Twenty-nine (69%) of the 42 primary tumors with amplifiable DNA contained K-ras mutations, whereas 3 (30%) of the 10 pancreata with benign conditions harbored mutations. Twenty-two (65%) of 34 ductal adenocarcinomas of the pancreas with amplifiable DNA had K-ras mutations. It is noteworthy that the same mutation was present in both the duodenal fluid and the primary carcinomas of 11 (92%) of the 12 patients who had primary tumors with amplifiable DNA as well as K-ras mutations in their duodenal fluid specimens.

CONCLUSIONS

The identification of genetic alterations in cancer-causing genes in duodenal fluid may form the basis for the development of new approaches to the detection of carcinoma of the pancreas. Some pancreata without cancer, however, may also harbor K-ras mutations, potentially limiting the specificity of K-ras-based tests.

Identifying markers for pancreatic cancer by gene expression analysis

To begin to identify new tumor markers, we recently performed a systematic study of gene expression in cancers of the colon and pancreas. Of the 45,000 genes identified, 183 were found to be expressed at significantly elevated levels in pancreatic cancer. One of the genes was tissue inhibitor of metalloproteinase type I (TIMP-1), which encodes a secreted protein. Analysis of TIMP-1 serum levels revealed significant increases in pancreatic cancer patients, but TIMP-1 by itself was inadequate as a serum marker for cancer. However, a combination of individually suboptimal markers (TIMP-1, CA19-9, and carcinoembryonic antigen) detected 60% of 85 patients with pancreatic cancers in a highly specific manner. These results suggest that a systematic analysis of gene expression can reveal novel serum markers and that individually suboptimal markers can be combined to yield higher sensitivity and specificity.

Progression of pancreatic intraductal neoplasias to infiltrating adenocarcinoma of the pancreas

Pancreata with cancer also frequently have intraductal proliferative lesions, suggesting an association between pancreatic cancer and these lesions. We present three cases in which atypical papillary hyperplasia of the pancreas was documented 17 months to 10 years before the development of an infiltrating adenocarcinoma of the pancreas. The first patient was a 70-year-old woman who underwent pancreaticoduodenectomy for adenocarcinoma of the pancreas. Atypical papillary duct hyperplasia extended to the pancreatic neck margin of resection, but the margin was negative for infiltrating carcinoma. Nine years later, an infiltrating adenocarcinoma developed in the remaining pancreas. The second patient was a 58-year-old man who underwent distal pancreatectomy for chronic pancreatitis with pseudocyst. Histologic examination showed chronic pancreatitis and multiple foci of atypical papillary duct hyperplasia. Ten years later, the patient underwent a Whipple procedure for infiltrating adenocarcinoma of the pancreas. The third patient was a 46-year-old woman with recurrent pancreatitis who underwent a Whipple procedure. Histologic examination showed atypical papillary duct hyperplasia and chronic pancreatitis but no infiltrating carcinoma. At the time of surgery, the tail of the pancreas was grossly and radiographically normal. Seventeen months later, a malignant pleural effusion developed, and postmortem examination showed infiltrating adenocarcinoma in the tail of the pancreas. In the cases presented, atypical papillary hyperplasia was documented 17 months, 9 years, and 10 years before the development of infiltrating adenocarcinoma of the pancreas, supporting the concept that there is a progression from intraductal hyperplasia to infiltrating carcinoma of the pancreas, just as there is a progression from adenoma to infiltrating carcinoma in the colorectum. Based on evidence that these intraductal lesions are precursor lesions to infiltrating adenocarcinoma of the pancreas, we suggest that the term "hyperplasia" be replaced by the more specific term "pancreatic intraepithelial neoplasia."

Pathology of cancer of the pancreas

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.

Adenocarcinoma of the duodenum: factors influencing long-term survival

This single-institution retrospective analysis reviews the management and outcome of patients with surgically treated adenocarcinoma of the duodenum. Between February 1984 and August 1996, fifty-five patients with adenocarcinoma of the duodenum underwent surgery at The Johns Hopkins Hospital. Univariate analysis was performed to identify possible prognostic indicators. Curative resection was performed in 48 patients (87%): 35 of these patients (73%) underwent a pancreaticoduodenectomy (PD), whereas 27% (n = 13) underwent a pancreas-sparing duodenectomy (PSD). Patients undergoing PD were comparable to those undergoing PSD with respect to demographic factors, presenting symptoms, and tumor pathology. The remaining 13% of patients (n = 7) were deemed unresectable at the time of surgery and underwent biopsy and/or palliative bypass. PD was associated with an increase in postoperative complications when compared to PSD (57% vs. 30%), but this difference was not statistically significant. One perioperative death occurred following PD (mortality 2.9%). The overall 5-year survival rate for the 48 patients undergoing potentially curative resection was 53%. Negative resection margins (P <0.001), PD (P <0.005), and tumors in the first and second portions of the duodenum (P <0.05) were favorable predictors of long-term survival by univariate analysis. Nodal status, tumor diameter, degree of differentiation, and the use of adjuvant chemoradiation therapy did not influence survival. These data support an aggressive role for resection in patients with adenocarcinoma of the dueodenum

Genetics of pancreatic cancer. From genes to families

Cancer of the pancreas is a genetic disease. Sporadic cancers of the pancreas are frequently associated with the activation of an oncogene, K-ras, and the inactivation of multiple tumor suppressor genes, including p53, DPC4, p16, and BRCA2. An improved understanding of the genetics of pancreas cancer should lead to new tests to screen for this disease and novel rational gene-based therapies.

K-ras mutations in the duodenal fluid of patients with pancreatic carcinoma

BACKGROUND

Many patients with carcinoma of the pancreas die because their disease is not detected until late in its course. Methods that detect these cancers earlier will improve patient outcome. Over 80% of pancreatic carcinomas contain mutations in codon 12 of the K-ras gene. Screening duodenal fluid for these mutations may lead to early detection of these cancers and assist in establishing a diagnosis of pancreatic carcinoma.

METHODS

Polymerase chain reaction (PCR), with and without restriction enzyme-mediated mutant enrichment, was performed on DNA isolated from duodenal fluid specimens from 61 patients who underwent pancreaticoduodenectomy (Whipple's operation) for either periampullary cancer or a benign condition of the pancreas. Representative sections of pancreas pathology (primary carcinoma, benign tumor, or chronic pancreatitis) from the patients with duodenal fluid specimens containing amplifiable DNA were also analyzed for K-ras mutations. Wild-type and mutant K-ras were detected by hybridization of the PCR products with K-ras codon 12 mutant and wild-type specific probes.

RESULTS

Seven of the 61 duodenal fluid specimens contained DNA that did not amplify. Thirteen (24% of the 54 duodenal fluid specimens with amplifiable DNA and 21% of the total of 61 specimens) contained activating point mutations at codon 12 of the K-ras gene. Mutations were detected in 13 of the 51 duodenal fluid specimens from patients with cancer (sensitivity, 25%), whereas mutations were not detected in any of the 9 amplifiable duodenal fluid specimens from patients with benign conditions of the pancreas (specificity, 100%). One duodenal fluid specimen from a patient with adenocarcinoma of the pancreas had two different K-ras mutations. DNA from three of the primary carcinomas did not amplify or was not available. Twenty-nine (69%) of the 42 primary tumors with amplifiable DNA contained K-ras mutations, whereas 3 (30%) of the 10 pancreata with benign conditions harbored mutations. Twenty-two (65%) of 34 ductal adenocarcinomas of the pancreas with amplifiable DNA had K-ras mutations. It is noteworthy that the same mutation was present in both the duodenal fluid and the primary carcinomas of 11 (92%) of the 12 patients who had primary tumors with amplifiable DNA as well as K-ras mutations in their duodenal fluid specimens.

CONCLUSIONS

The identification of genetic alterations in cancer-causing genes in duodenal fluid may form the basis for the development of new approaches to the detection of carcinoma of the pancreas. Some pancreata without cancer, however, may also harbor K-ras mutations, potentially limiting the specificity of K-ras-based tests.

Pancreaticoduodenectomy for selected periampullary neuroendocrine tumors: fifty patients

BACKGROUND

Most resectable pancreatic or peripancreatic neuroendocrine tumors are treated by enucleation or distal pancreatectomy. A minority of tumors may require pancreaticoduodenectomy for complete tumor excision because of their large size, location, or lymph node involvement.

METHODS

This study reviews the management of 50 patients treated by pancreaticoduodenectomy for periampullary neuroendocrine tumors between 1962 and 1996 at a single institution.

RESULTS

There were 30 men and 20 women with a mean age of 52 +/- 2 years. Functional tumors were resected in 17 patients: insulinoma, seven tumors; gastrinoma, eight tumors; vipoma, one tumor; and glucagonoma, one tumor. Tumors were classified as malignant in 29 patients and benign in 21. The median intraoperative blood loss was 800 ml, and the median number of units of blood transfused was zero. The postoperative length of stay was 20 +/- 2 days. Postoperative morbidity included 11 patients (24%) with a pancreatic fistula and four patients (8%) with a biliary fistula. There was one in-hospital death (2%), in 1967. The actuarial survival rates at 2, 5, and 7 years are 81%, 73%, and 65%, respectively. Patients with benign tumors had a significantly improved 5-year survival rate (94%) compared with those with malignant tumors (61%; p = 0.03).

CONCLUSIONS

Selected patients with periampullary neuroendocrine tumors can be managed successfully by pancreaticoduodenectomy, with low mortality and acceptable morbidity rates.

Genetic progression and heterogeneity in intraductal papillary-mucinous neoplasms of the pancreas

Intraductal papillary-mucinous neoplasms (IPMNs) of the pancreas are ideal neoplasms to study clonal progression and genetic diversity because of their large size and prominent intraductal component. We microdissected 55 histologically defined areas from 13 IPMNs, extracted the DNA from each, and performed polymerase chain reaction (PCR)-based microsatellite analysis to detect loss of heterozygosity on chromosome arms 1p, 3p, 6q, 8p, 9p, 17p, 18q, and 22q. LOH was identified at 1p in two cases, at 3p in four cases, at 6q in seven cases, at 8p in four cases, at 9p in eight cases, at 17p in five cases, at 18q in five cases, and at 22q in one of the IPMNs examined. In one of the IPMNs, the allelic losses were uniform throughout multiple microdissected areas, and in four of the IPMNs, there was evidence of clonal progression. In contrast, in three of the IPMNs, substantial allelic heterogeneity was seen. This remarkable heterogeneity may, in part, be due to the slow growth rate of these neoplasms.

Can K-ras codon 12 mutations be used to distinguish benign bile duct proliferations from metastases in the liver? A molecular analysis of 101 liver lesions from 93 patients

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.

Inducible nitric oxide synthase expression in coronary arteries of transplanted human hearts with accelerated graft arteriosclerosis

Inducible nitric oxide synthase (iNOS) is a high-output isoform of NOS that produces nitric oxide (NO), a nonspecific immune effector molecule. In some animal models of autoimmunity, the induction of iNOS has been shown to lead to inflammation and tissue damage, and it has been suggested that iNOS is an immune mediator in humans as well. Using in situ hybridization and immunohistochemical techniques, we demonstrate that iNOS mRNA and protein are present in the coronary arteries of transplanted human hearts with accelerated graft arteriosclerosis (AGA). iNOS is expressed in cells morphologically consistent with macrophages in the neointima of 7 of 10 of the transplanted vessels with AGA that were examined. In serial sections, these same cells express the macrophage marker CD68. In contrast, iNOS is absent from five native coronary arteries with atherosclerosis and absent from two normal coronary arteries. Although iNOS is expressed in macrophages in AGA, its role in the pathogenesis of AGA is unknown.

Six hundred fifty consecutive pancreaticoduodenectomies in the 1990s: pathology, complications, and outcomes

OBJECTIVE

The authors reviewed the pathology, complications, and outcomes in a consecutive group of 650 patients undergoing pancreaticoduodenectomy in the 1990s.

SUMMARY BACKGROUND DATA

Pancreaticoduodenectomy has been used increasingly in recent years to resect a variety of malignant and benign diseases of the pancreas and periampullary region.

METHODS

Between January 1990 and July 1996, inclusive, 650 patients underwent pancreaticoduodenal resection at The Johns Hopkins Hospital. Data were recorded prospectively on all patients. All pathology specimens were reviewed and categorized. Statistical analyses were performed using both univariate and multivariate models.

RESULTS

The patients had a mean age of 63 +/- 12.8 years, with 54% male and 91% white. The number of resections per year rose from 60 in 1990 to 161 in 1995. Pathologic examination results showed pancreatic cancer (n = 282; 43%), ampullary cancer (n = 70; 11%), distal common bile duct cancer (n = 65; 10%), duodenal cancer (n = 26; 4%), chronic pancreatitis (n = 71; 11%), neuroendocrine tumor (n = 31; 5%), periampullary adenoma (n = 21; 3%), cystadenocarcinoma (n = 14; 2%), cystadenoma (n = 25; 4%), and other (n = 45; 7%). The surgical procedure involved pylorus preservation in 82%, partial pancreatectomy in 95%, and portal or superior mesenteric venous resection in 4%. Pancreatic-enteric reconstruction, when appropriate, was via pancreaticojejunostomy in 71% and pancreaticogastrostomy in 29%. The median intraoperative blood loss was 625 mL, median units of red cells transfused was zero, and the median operative time was 7 hours. During this period, 190 consecutive pancreaticoduodenectomies were performed without a mortality. Nine deaths occurred in-hospital or within 30 days of operation (1.4% operative mortality). The postoperative complication rate was 41%, with the most common complications being early delayed gastric emptying (19%), pancreatic fistula (14%), and wound infection (10%). Twenty-three patients required reoperation in the immediate postoperative period (3.5%), most commonly for bleeding, abscess, or dehiscence. The median postoperative length of stay was 13 days. A multivariate analysis of the 443 patients with periampullary adenocarcinoma indicated that the most powerful independent predictors favoring long-term survival included a pathologic diagnosis of duodenal adenocarcinoma, tumor diameter <3 cm, negative resection margins, absence of lymph node metastases, well-differentiated histology, and no reoperation.

CONCLUSIONS

This single institution, high-volume experience indicates that pancreaticoduodenectomy can be performed safely for a variety of malignant and benign disorders of the pancreas and periampullary region. Overall survival is determined largely by the pathology within the resection specimen.

17beta-estradiol inhibits apoptosis of endothelial cells

Endothelial cells provide an antithrombotic and anti-inflammatory barrier for the normal vessel wall. Dysfunction of endothelial cells has been shown to promote atherosclerosis, and normalization of previously dysfunctional endothelial cells can inhibit the genesis of atheroma. In normal arteries, endothelial cells are remarkably quiescent. Acceleration of the turnover rate of endothelial cells can lead to their dysfunction. Apoptosis is a physiological process that contributes to vessel homeostasis, by eliminating damaged cells from the vessel wall. However, increased endothelial cell turnover mediated through accelerated apoptosis may alter the function of the endothelium and therefore, promote atherosclerosis. Apoptotic endothelial cells can be detected on the luminal surface of atherosclerotic coronary vessels, but not in normal vessels. This finding links endothelial cell apoptosis and the process of atherosclerosis, although a causative role for apoptosis in this process remains hypothetical. Estrogen metabolites have been shown to be among the most potent anti-atherogenic agents available to date for post-menopausal women. The mechanism of estrogen's protective effect is currently incompletely characterized. Here we show that 17beta-estradiol, a key estrogen metabolite, inhibits apoptosis in cultured endothelial cells. Our data support the hypothesis that 17beta-estradiol's anti-apoptotic effect may be mediated via improved endothelial cell interaction with the substratum, increased tyrosine phosphorylation of pp125 focal adhesion kinase, and a subsequent reduction in programmed cell death of endothelial cells. Inhibition of apoptosis by estrogens may account for some of the anti-atherogenic properties of these compounds.

Abrogation of the Rb/p16 tumor-suppressive pathway in virtually all pancreatic carcinomas

The Rb/p16 tumor-suppressive pathway is abrogated frequently in human tumors, either through inactivation of the Rb or p16INK4a/CDKN2/MTS1 tumor-suppressor proteins, or through alteration or overexpression of the cyclin D1 or cyclin-dependent kinase 4 oncoproteins. We reported previously that the p16 gene was genetically inactivated in 82% of pancreatic carcinomas. Nearly half of these inactivations were by intragenic mutation of p16, and the remainder were by homozygous deletion of the gene. Here, we analyzed pancreatic carcinomas for additional mechanisms by which the Rb/p16 pathway might be inactivated. Transcriptional silencing of the p16 gene in association with methylation of its 5'-CpG island was examined by methylation-specific PCR in 18 pancreatic carcinomas. Nine of these were known to harbor an intragenic mutation in p16, and nine had a wild-type p16 coding sequence. Seven of the 18 tumors were hypermethylated, and all 7 were p16 wild-type (P = 0.001). Complete silencing of transcription from methylated wild-type gene sequences was demonstrated. Immunohistochemical analysis revealed normal expression levels of the Rb protein in all carcinomas studied. None of the carcinomas had genomic amplification of the cyclin D1 or CDK4 genes, and none had mutation of the p16-binding domain of CDK4. An additional p16 mutation was identified. In total, the Rb/p16 pathway was abrogated in 49 of the 50 carcinomas (98%) studied, all through inactivation of the p16 gene. Similar results were obtained in an independently analyzed series of 19 pancreatic carcinomas. These data demonstrate the central role of the Rb/p16 pathway in the development of pancreatic carcinoma.

The autopsy: still important in cardiac surgery

BACKGROUND

This study examined the ability of autopsy to confirm or dispute presumptive cause of death among cardiac surgery patients.

METHODS

Autopsy reports were compared with mortality conference notes that were dictated prospectively before autopsy results were available. Between January 1985 and December 1995, there were 600 hospital deaths among 13,029 adult cardiac surgery patients (4.6% mortality). Of these 600 deaths, 147 (24.5%) had postmortem examination.

RESULTS

Annual autopsy rate remained constant over the course of the study. Autopsied patients were younger (60.4 +/- 15 versus 66.7 +/- 13 years [mean +/- standard error of the mean]; p < 0.0001), but their race and sex distributions were similar to deceased patients not having autopsy. Autopsy confirmed clinical presumptive cause of death in 52% (76), disputed clinical diagnosis in 9.5% (14), provided definitive diagnosis in the absence of clinical diagnosis in 13.6% (20), and failed to provide definitive diagnosis in 25% (37). One third of autopsies (39%; 57) provided information that was clinically unrecognized and might have altered therapy and outcome if known premortem. As determined by autopsy, common causes of death were cardiac (27%; 39), unknown (25%; 37), sepsis (14%; 21), stroke (8.8%; 13), cholesterol embolism (4.1%; 6), pulmonary embolism (4.1%; 6), and adult respiratory distress syndrome (4.1%; 6).

CONCLUSIONS

Autopsy reveals or confirms cause of death in nearly three quarters of cardiac surgical deaths and provides information that differs significantly from premortem clinical impression more than 20% of the time. As such, the autopsy remains important to quality assurance in cardiac surgical care.