Search

OBJECTIVES: Pancreatic cystic lesions (PCLs) may be precancerous. Those likely to harbor high-grade dysplasia (HGD) or pancreatic cancer (PC) are targets for surgical resection. Current algorithms to predict advanced neoplasia (HGD/PC) in PCLs lack diagnostic accuracy. In pancreatic tissue and cystfluid (CF) from PCLs, we sought to identify and validate novel methylated DNA markers (MDMs) that discriminate HGD/PC from low-grade dysplasia (LGD) or no dysplasia (ND). METHODS: From an unbiased whole-methylome discovery approach using predefined selection criteria followed by multistep validation on case (HGD or PC) and control (ND or LGD) tissues, we identified discriminant MDMs. Top candidate MDMs were then assayed by quantitative methylation-specific polymerase chain reaction on archival CF from surgically resected PCLs. RESULTS: Of 25 discriminant MDMs identified in tissue, 13 were selected for validation in 134 CF samples (21 cases [8 HGD, 13 PC], 113 controls [45 ND, 68 LGD]). A tree-based algorithm using 2 CF-MDMs (TBX15, BMP3) achieved sensitivity and specificity above 90%. Discrimination was significantly better by this CF-MDM panel than by mutant KRAS or carcinoembryonic antigen, with areas under the receiver operating characteristic curve of 0.93 (95% confidence interval: 0.86–0.99), 0.71 (0.57–0.85), and 0.72(0.60–0.84), respectively. Cutoffs for the MDM panel applied to an independent CF validation set (31 cases, 56 controls) yielded similarly high discrimination, areas under the receiver operating characteristic curve 5 0.86 (95% confidence interval: 0.77–0.94, P = 0.2). DISCUSSION: Novel MDMs discovered and validated in tissue accurately identify PCLs harboring HGD/PC. A panel of 2 MDMs assayed in CF yielded results with potential to enhance current risk prediction algorithms. Prospective studies are indicated to optimize and further evaluate CF-MDMs for clinical use.

There is lack of consensus on post-operative surveillance for resected non-invasive intraductal papillary neoplasms (IPMNs). In this study we explored risk factors for subsequent PC in patients with MD-IPMN undergoing partial pancreatectomy. We searched the Mayo Clinic surgical pathology database for all cases of resected MD-IPMN between 1997 and 2014. Cases with histologically confirmed main pancreatic duct involvement either isolated or in a mixed pattern with branch-duct involvement were included. Outcomes of PC in the remnant pancreas, and death related to MD-IPMN were assessed with survival analyses (Kaplan–Meier and Cox regression). Among the 179 patients with resected MD-IPMN the incidence of concomitant PC and high-grade dysplasia (HGD) in the resected specimen was 23 and 14%, respectively. The mean duration of follow-up was 4.31 years (range 0.12–13.5 years). Excluding 28 subjects who either underwent initial total pancreatectomy or partial pancreatectomy with surgical margins positive for PC/HGD, the 5-year incidence of subsequent PC was 12%, including 60.6% and15.6% in those with initial PC and HGD, respectively. The 10-year incidence of PC was 21.2% overall, 60.6% for PC, 38.3% for HGD, and 3.0% for LGD. Risk of subsequent PC was significantly higher for those with initial PC compared with HGD (HR = 4.95, 95% CI: 1.63–15.03, p = 0.005 and for HGD compared with LGD (HR = 11.30, 95% CI: 1.55–82.26, p = 0.017). Patients with MD-IPMN with PC or HGD undergoing segmental pancreatectomy are at higher risk of subsequent PC and may benefit from post-operative surveillance. The post-operative surveillance intervals in resected MD- IPMNs need to be tailored based on dysplasia grade.

Minimally invasive methods have been described to detect Barrett’s esophagus (BE), but are limited by subjectivity and suboptimal accuracy. We identified methylated DNA markers (MDMs) for BE in tissue and assessed their accuracy on whole esophagus brushings and capsule sponge samples. Step 1: Unbiased whole methylome sequencing was performed on DNA from BE and normal squamous esophagus (SE) tissue. Discriminant MDM candidates were validated on an independent patient cohort (62 BE cases, 30 controls) by quantitative methylation specific PCR (qMSP). Step 2: Selected MDMs were further evaluated on whole esophageal brushings (49 BE cases, 36 controls). 35 previously sequenced esophageal adenocarcinoma (EAC) MDMs were also evaluated. Step 3: 20 BE cases and 20 controls were randomized to swallow capsules sponges (25 mm, 10 pores or 20 pores per inch (ppi)) followed endoscopy. DNA yield, tolerability, and mucosal injury were compared. Best MDM assays were performed on this cohort. Step 1: 19 MDMs with areas under the ROC curve (AUCs) >0.85 were carried forward. Step 2: On whole esophageal brushings, 80% of individual MDM candidates showed high accuracy for BE (AUCs 0.84–0.94). Step 3: The capsule sponge was swallowed and withdrawn in 98% of subjects. Tolerability was superior with the 10 ppi sponge with minimal mucosal injury and abundant DNA yield. A 2-marker panel (VAV3 + ZNF682) yielded excellent BE discrimination (AUC = 1). Identified MDMs discriminate BE with high accuracy. BE detection appears safe and feasible with a capsule sponge. Corroboration in larger studies is warranted. ClinicalTrials.gov number NCT02560623.

Objectives The objective of this study was to determine whether constipation-predominant irritable bowel syndrome (IBS-C) is associated with changes in intestinal barrier and secretory function. Methods A total of 19 IBS-C patients and 18 healthy volunteers (all females) underwent saccharide excretion assay (0.1 g 13C mannitol and 1 g lactulose), measurements of duodenal and colonic mucosal barrier (transmucosal resistance (TMR), macromolecular and Escherichia coli Bio-Particle translocation), mucosal secretion (basal and acetylcholine (Ach)-evoked short-circuit current (Isc)), in vivo duodenal mucosal impedance, circulating endotoxins, and colonic tight junction gene expression. Results There were no differences in the in vivo measurements of barrier function between IBS-C patients and healthy controls: cumulative excretion of 13C mannitol (0-2 h mean (s.e.m.); IBS-C: 12.1 (0.9) mg vs. healthy: 13.2 (0.8) mg) and lactulose (8-24 h; IBS-C: 0.9 (0.5) mg vs. healthy: 0.5 (0.2) mg); duodenal impedance IBS-C: 729 (65) Ω vs. healthy: 706 (43) Ω; plasma mean endotoxin activity level IBS-C: 0.36 (0.03) vs. healthy: 0.35 (0.02); and in colonic mRNA expression of occludin, zonula occludens (ZO) 1-3, and claudins 1-12 and 14-19. The ex vivo findings were consistent, with no group differences: duodenal TMR (IBS-C: 28.2 (1.9) Ω cm2 vs. healthy: 29.8 (1.9) Ω cm2) and colonic TMR (IBS-C: 19.1 (1.1) Ω cm2 vs. healthy: 17.6 (1.7) Ω cm2); fluorescein isothiocyanate (FITC)-dextran (4 kDa) and E. coli Bio-Particle flux. Colonic basal Isc was similar, but duodenal basal Isc was lower in IBS-C (43.5 (4.5) μA cm-2) vs. healthy (56.9 (4.9) μA cm-2), P=0.05. Ach-evoked ΔIsc was similar. Conclusions Females with IBS-C have normal colonic barrier and secretory function. Basal duodenal secretion is decreased in IBS-C.

GOALS Our aim was to develop a scoring system to predict risk of microscopic colitis (MC), to identify patients at low risk, potentially avoiding unnecessary biopsies. BACKGROUND Patients with chronic diarrhea often undergo colonoscopy with biopsy, but few have histologic abnormalities. STUDY We conducted a retrospective study of patients with chronic diarrhea and a macroscopically normal colonoscopy at our institution over a 9-month period. Multivariable logistic regression assessed the association between predictors and the presence of biopsy-proven MC. RESULTS The derivation cohort included 617 patients. Median age was 55.1 (39.6 to 68.1) years; 397 (64.3%) were female and 81 (13.1%) had MC. Age ≥55 years, duration of diarrhea ≤6 months, ≥5 bowel movements per day, body mass index

Mismatch-repair (MMR) immunohistochemistry is used to detect tumor MMR deficiency associated with high-level microsatellite instability (MSI). Rare tumors show heterogenous loss of mutS homolog 6 (MSH6) with immunohistochemistry, defined by areas of retained staining and separate areas of complete loss of staining. To investigate the clinical interpretation of this phenomenon, we identified 22 cases of heterogenous MSH6 loss interpreted at Mayo Clinic from January 2001 through December 2012 and reviewed histologic features, MSH6 and other MMR immunohistochemistry, and accompanying MSI testing results (n=20). Heterogenous MSH6 loss was seen in colorectal carcinoma (n=18), endometrial carcinoma (n=3), and sebaceous neoplasm (n=1). In the 18 colorectal carcinoma cases, it accompanied complete loss of mutL homolog 1 (MLH1) or PMS2, or both. Heterogenous MSH6 loss was characterized by MSI and MSH6 C8 tract instability in treatment-naive cases and showed mucinous or signet-ring zones in one quarter of cases. Two cases status post neoadjuvant chemoradiation showed heterogenous MSH6 loss but were microsatellite and C8 tract stable. C8 tracts were unstable in 2 of 4 MSH6-associated Lynch syndrome (LS) tumors, but all 4 showed complete MSH6 loss on immunohistochemistry. Further, 12 such MSH6-associated LS cases showed complete MSH6 loss. In conclusion, heterogenous MSH6 loss is uncommon, usually caused by instability in MSH6 exon 5 polycytosine tract, and not associated with germline MSH6 mutation. Although heterogenous MSH6 loss provides evidence against germline MSH6 mutation, patients whose tumors exhibit this immunolabeling pattern may have LS due to a defect in a different MMR gene.

To further elucidate the anticancer mechanisms of metformin against pancreatic cancer, we evaluated the inhibitory effects of metformin on pancreatic tumorigenesis in a genetically engineered mouse model and investigated its possible anti-inflammatory and antiangiogenesis effects.Six-week-old LSL-Kras;Trp53 mice (10 per group) were administered once daily intraperitoneally with saline (control) for 1 week or metformin (125 mg/kg) for 1 week (Met_1wk) or 3 weeks (Met_3wk) before tumor initiation. All mice continued with their respective injections for 6 weeks after tumor initiation. Molecular changes were evaluated through quantitative polymerase chain reaction, immunohistochemistry, and Western blotting.At euthanasia, pancreatic tumor volume in the Met_1wk (median, 181.8 mm) and Met_3wk (median, 137.9 mm) groups was significantly lower than those in the control group (median, 481.1 mm; P = 0.001 and 0.0009, respectively). No significant difference was observed between the Met_1wk and Met_3wk groups (P = 0.51). These results were further confirmed using tumor weight and tumor burden measurements. Furthermore, metformin treatment decreased the phosphorylation of nuclear factor κB and signal transducer and activator of transcription 3 as well as the expression of specificity protein 1 transcription factor and several nuclear factor κB-regulated genes.Metformin may inhibit pancreatic tumorigenesis by modulating multiple molecular targets in inflammatory pathways.

The risk of pancreatic and extrapancreatic cancer in autoimmune pancreatitis (AIP) has not been systematically evaluated. We compared the risk of malignancy in AIP patients and matched control subjects.We identified 116 subjects with AIP from a prospectively maintained database. From patients evaluated in primary care clinics, we selected 3 control subjects for each AIP patient matched on age, registration date, and sex (n = 344) . Risk for developing cancer after the index date was compared using a stratified Cox model.The proportion of patients diagnosed with cancer before AIP diagnosis (10.3%) was lower than that in the matched control subjects (17.4%). After a median follow-up of over 3 years, the risk of developing cancer after the index date was similar in AIP and control subjects (hazard ratio, 0.64; 95% confidence interval, 0.27-1.51). The 3 most commonly diagnosed malignancies in the AIP group were prostate cancer, lymphoma, and bladder cancer.Cancer risk before and after diagnosis of AIP is similar to that of control subjects. Specifically, there is no increased risk of cancer immediately preceding or following AIP diagnosis. Additional follow-up is needed to determine if there is a cumulative increase in cancer risk in AIP.

Microscopic colitis (MC) is prevalent in adults investigated for chronic watery diarrhea, yet characterization of pediatric MC is limited.Our pathology database was searched from 1995 to 2011 for pediatric cases of lymphocytic colitis (LC) or collagenous colitis (CC). Those with diarrhea persisting for2 weeks and visually normal colonoscopy were accepted as cases. Demographics, laboratory results, medication use within 3 months of presentation, medical and family history of autoimmune disease, and response to treatment were abstracted.A total of 27 cases were histologically consistent with MC on biopsy; 5 with concomitant enteric infection or isolated abdominal pain were excluded. Twenty-two cases of MC (female patients, 59%; median age at diagnosis, 15.3 years) were included (19 LC and 3 CC). Two had type 1 diabetes mellitus, 2 were anti-nuclear antibody positive, and 2 had common variable immunodeficiency. Of 20 patients who underwent an esophagogastroduodenoscopy, 1 had collagenous sprue and 4 had celiac disease. One presented after the clearance of recurrent Clostridium difficile infection. Previous drug exposures included nonsteroidal anti-inflammatory drugs (n = 7), proton pump inhibitors (n = 6), and selective serotonin reuptake inhibitors (n = 3). Common symptoms in addition to diarrhea included abdominal pain (77.3%) and weight loss (27.3%). Of 17 patients with follow-up, all of the 8 treated with steroids had some response: 57.1% (4/7) responded to mesalamine and 42.9% (3/7) responded to bismuth subsalicylate.In this cohort of pediatric patients, LC was much more common than CC. As described in adults, we observed associations with celiac disease, type 1 diabetes mellitus, and medications; we additionally saw an association with immunodeficiency. Our patients showed greater response to steroids than mesalamine or bismuth.

To evaluate the outcomes of corticosteroid-treated microscopic colitis (MC) in a population-based cohort, and to compare these outcomes in patients treated with prednisone or budesonide.A historical cohort study of Olmsted County, Minnesota residents diagnosed with collagenous or lymphocytic colitis (LC) between 1986 and 2010 was performed using the Rochester Epidemiology Project.Of 315 patients with MC, 80 (25.4%) were treated with corticosteroids. The median age at colitis diagnosis was 66.5 years (range: 16-95) and 78.7% were female. Forty patients (50%) had LC and 40 (50%) had collagenous colitis. Prednisone was used in 17 patients (21.2%) and budesonide in 63 (78.8%); 56 (75.6%) had complete response and 15 (20.3%) had partial response. Patients treated with budesonide had a higher rate of complete response than those treated with prednisone (82.5 vs. 52.9%; odds ratio, 4.18; 95% CI, 1.3-13.5). Six patients were lost to follow-up. The remaining 74 had a median follow-up of 4 years (range 0.2-14). Fifty patients out of the 71 who responded (70.4%) had a recurrence after corticosteroid discontinuation. Patients treated with budesonide were less likely to recur than those treated with prednisone (hazard ratio, 0.38; 95% CI, 0.18-0.85; P=0.02). After 397 person years of follow-up in the 73 patients with long-term data, 47 (64.4%) required maintenance with corticosteroids.Patients with MC often respond to corticosteroid therapy, but with a high relapse rate. Budesonide had a higher response rate and a lower risk of recurrence than prednisone.

Small bowel intraepithelial lymphocytosis (IELs) with normal villous architecture is a relatively common finding, often of uncertain significance. The aims of our study were to determine the prevalence of this finding, its clinical associations, its specificity for celiac disease (CD), and whether histologic clues exist that increase the specificity for CD in this setting, all in the pediatric population.The Mayo Clinic electronic pathology database was searched between January 1, 2000 and December 31, 2009 for patients younger than 18 years who had the terms "normal villi" and "increased intraepithelial lymphocytes" in their small bowel biopsy reports. All of the slides were reviewed to confirm the histologic findings. Demographic, serologic, pathologic, and clinical informations were obtained.Among 1290 duodenal biopsies obtained from children during the years 2000 and 2009, 56 (4.3%) were noted to have "normal villous architecture with increased intraepithelial lymphocytes." In the 54 patients not known to have CD before biopsy, 48 (89%) had serologic testing for CD. Of these 48 patients, 9 were labeled with CD, although only 5 of 9 met the definite criteria for the diagnosis, based on a combination of serologic markers, human leukocyte antigen haplotyping, and response to gluten-free diet. No clinical features pointed to a diagnosis of CD. There was no correlation between CD and number of IELs, but patients with newly diagnosed CD were more likely to have a tip-heavy lymphocyte distribution. Other diagnoses made during the study period and in follow-up were inflammatory bowel disease (5), Helicobacter pylori infection (3), medication-related injury (10), and systemic autoimmune disorders (2).Increased IELs with normal villous architecture in small bowel biopsies are clinically important in children, and are associated with a new definite diagnosis of CD in 9% of pediatric patients. Even at this low sensitivity, clinical work-up for CD is mandated in all of the patients with this finding.

Background Biliary intraductal papillary mucinous neoplasm (B-IPMN) has been proposed as a unique clinicopathologic disease with distinct histopathologic features, although wide acceptance remains controversial. A recent consensus conference classified pancreatic IPMN (P-IPMN) into 4 subtypes (ie, gastric, intestinal, pancreatobiliary, oncocytic) based on morphologic appearance and mucin (MUC) staining properties. The aim of this study was to determine whether B-IPMN has similar histopathologic and immunologic subtypes to P-IPMN. Study Design Specific immunostaining for MUC1, MUC2, and deleted for pancreas cancer, locus 4 were performed on specimens from 19 patients with a histopathologic diagnosis of B-IPMN. Immunostaining patterns of B-IPMN were correlated with histopathology. Results Based on histopathology, the following subtypes of B-IPMN were identified: pancreatobiliary n=9 (47%), intestinal n=8 (42%), oncocytic n=2 (11%), and gastric n=0 (0%). Pancreatobiliary and oncocytic subtypes of B-IPMN were positive for MUC1 and negative for MUC2, and intestinal subtypes were positive for MUC2 and negative for MUC1. Thirteen of the 19 B-IPMN were associated with invasive carcinoma; loss of deleted for pancreas cancer, locus 4 was found in 6 of 13 invasive components and in 3 of 19 noninvasive components of B-IPMN. Five-year survival for patients with resected B-IPMN and invasive carcinoma was 38%, which is similar to that for resected P-IPMN with invasive carcinoma. Conclusions Histopathologic subtypes and type-specific MUC expression patterns of B-IPMN resemble those of P-IPMN. MUC1 expression and/or absence of MUC2 expression, which correlate with aggressive features of P-IPMN, were found in B-IPMN and correlate with invasive B-IPMN. Loss of deleted for pancreas cancer, locus 4 parallels the findings observed in P-IPMN. These findings provide additional support that B-IPMN is a unique entity with similarities to main duct P-IPMN.

IgG4-related systemic disease is an autoimmune disease that was first recognized in the pancreas but also affects other organs. This disease may manifest as tubulointerstitial nephritis (IgG4-TIN), but its clinicopathologic features in the kidney are not well described. Of the 35 patients with IgG4-TIN whose renal tissue specimens we examined, 27 (77%) had acute or progressive chronic renal failure, 29 (83%) had involvement of other organ systems, and 18 of 23 (78%) had radiographic abnormalities. Elevated total IgG or IgG4 serum levels were present in 79%. All pathologic specimens featured plasma cell–rich TIN, with most showing diffuse, expansile interstitial fibrosis. Immune complexes along the tubular basement membranes were present in 25 of 30 (83%). All specimens had a moderate to marked increase in IgG4+ plasma cells by immunohistochemistry. We used a control group of 175 pathologic specimens with plasma cell–rich interstitial infiltrates that can mimic IgG4-TIN to examine the diagnostic utility of IgG4 immunostaining. Excluding pauci-immune necrotizing and crescentic glomerulonephritis, IgG4 immunohistochemistry had a sensitivity of 100% (95% CI 90–100%) and a specificity of 92% (95% CI 86–95%) for IgG4-TIN. Of the 19 patients with renal failure for whom treatment and follow-up data were available, 17 (89%) responded to prednisone. In summary, because no single test definitively diagnoses IgG4-related systemic disease, we rely on a combination of histologic, immunophenotypic, clinical, radiographic, and laboratory features. When the disease manifests in the kidney, our data support diagnostic criteria that can distinguish IgG4-TIN from other types of TIN.

Intestinal endometriosis (IE) can present with varied symptoms, making the diagnosis difficult. Modalities have been described to evaluate IE, but few can provide a confirmatory diagnosis. A preoperative diagnosis of IE may help guide management. We sought to describe the presentation, diagnostic evaluation, histology and operative management of 89 patients with tissue-confirmed symptomatic IE.The records of 89 patients from a single institution with histologically confirmed, symptomatic IE from 1 January 1994 to 30 September 2009 were reviewed.Abdominal pain was the most common symptom in patients with IE; however, rectal bleeding was significantly associated with IE of the distal colon (P=0.02), while dysfunctional uterine bleeding was seen more in patients with proximal IE (P=0.01). Preoperative confirmation of IE was uncommon; colonoscopy with biopsy confirmed the diagnosis in 29.6% of patients tested and only 15% of patients with IE had histologic lesions involving mucosa. In the five patients who underwent endoscopic ultrasound (EUS), the diagnosis of IE was established in all cases (n=4) where histology or cytology was obtained. Malignancy was considered nearly as frequently as IE preoperatively, and 90.4% of patients underwent laparotomy as the initial surgical approach.IE can present with a variety of manifestations, which may provide clues to location of bowel affected. Patients with known pelvic endometriosis and rectal bleeding are more likely to have distal bowel affected; EUS with tissue sampling may play a role if routine endoscopy fails to reveal the diagnosis. Making a diagnosis of IE preoperatively may allow for less invasive surgical approaches and better patient outcomes.

The spectrum of IgG4-related systemic disease (IgG4-RSD) continues to widen. At most of the sites involved by this condition, the clinical presentation can mimic neoplasm. Pathologic assessment of small biopsies can be critical to proper management. This review summarizes the histologic features of IgG4-RSD and the role of immunohistochemistry of IgG4 in the diagnosis.The review period saw further expansion of the list of sites putatively involved by IgG4-RSD, with new, or more detailed, entries related to lung, lymph nodes, stomach, and thyroid. A tentative consensus was reached on the issue of subtypes of autoimmune pancreatitis. The role of immunohistochemistry for IgG4 as an adjunct to the diagnosis of IgG4-RSD was further clarified.Sclerosing lymphoplasmacytic inflammation at almost any site can represent a manifestation of IgG4-RSD. There are several histologic features that can suggest the diagnosis. Immunohistochemistry for IgG4 is a useful diagnostic test to further support the diagnosis.

The objective of this study was to evaluate and validate cyst fluid carcinoembyronic antigen (CEA) and amylase in differentiating (1) nonmucinous from mucinous pancreatic cystic lesions (PCLs), (2) benign mucinous from malignant mucinous PCLs, and (3) pseudocysts from nonpseudocysts (amylase only).A retrospective analysis of patients with histologically confirmed PCLs from February 1996 to April 2007 was performed. Cyst fluid CEA (n=124) and/or amylase (n=91) were measured and correlated to cyst type.Carcinoembyronic antigen levels (P=0.0001), but not amylase, were higher in mucinous versus nonmucinous cysts. The sensitivity, specificity, and diagnostic accuracy of CEA 200 ng/mL or greater for the diagnosis of mucinous PCLs were 60%, 93%, and 72%, respectively. Carcinoembyronic antigen levels did not differentiate benign from malignant mucinous cysts. Whereas amylase levels were higher in pseudocysts than nonpseudocysts (P=0.009), 54% of noninflammatory PCLs had a level greater than 250 IU/L, including mucinous cystic neoplasms (median, 6800 IU/L; interquartile range, 70-25,295 IU/L). Malignant mucinous cysts had lower amylase levels than benign mucinous cysts (P=0.0008).Cyst fluid CEA and amylase levels are suggestive but not diagnostic in differentiating PCLs. Unlike CEA, amylase may help differentiate benign from malignant mucinous cysts. Novel biomarkers are needed.

Patients with ulcerative colitis (UC) are at increased risk for developing colorectal cancer (CRC). Surveillance in this at-risk population remains challenging. We assessed the methylation status of genes in the non-neoplastic mucosa of UC-CRC patients and controls to identify potential biomarkers of CRC.We evaluated the methylation status of 10 genes (p16, p14, runt-related transcript factor-3 (RUNX3), cyclooxygenase-2 (COX-2), E-cadherin, methylated-in-tumor-1 (MINT1), MINT31, HPP1, estrogen receptor, SLC5A8) in UC-CRC tumors and non-neoplastic sections from both UC-CRC cases and UC controls (n=114 for each) using methylation-specific PCR.Amplification was successful for 96 UC controls, 83 tumors, and 66 non-adjacent, non-neoplastic samples. The prevalence of methylation was significantly greater in UC-CRC tumors for p16, RUNX3, MINT1, MINT31, and HPP1. Methylation of COX-2 and E-cadherin was greater in UC controls than in tumors. Univariate testing of these genes using non-adjacent, non-neoplastic sections from UC-CRC cases indicated that associations between p16, RUNX3, MINT1, MINT31, E-cadherin, and COX-2 and UC-CRC remained significant. In multivariable analysis of the six genes, only RUNX3, MINT1, and COX-2 remained significantly associated with the UC-CRC cases (odds ratio=12.6, 9.0, and 0.2, respectively). The results remained unaffected by the presence of PSC or severity of inflammation. Logistic regression modeling with the three genes showed interactions that increased the odds ratio for each gene.RUNX3, MINT1, and COX-2 are potential biomarkers for detecting the presence of CRC in patients with UC. These genes should be evaluated as biomarkers for colorectal dysplasia.

Carney triad, as originally described in 1977, was the association of 3 tumors: gastric epithelioid leiomyosarcoma [later renamed gastrointestinal stromal tumor (GIST)], extra-adrenal paraganglioma, and pulmonary chondroma. The disorder affected mostly young women and was not familial. We studied the clinical and pathologic features of the gastric neoplasm in 104 patients with the syndrome. Most (88%) were young women (mean age, 22 y), and the usual presentation was gastric bleeding. The tumors, commonly antral-based (61%), were multifocal, and ranged from 0.2 to 18.0 cm in dimension. Most (86%) featured round and polygonal (epithelioid) cells. Metastasis occurred in 49 patients (47%): to gastric lymph nodes (29%), liver (25%), and peritoneum (13%). Immunopositivity was detected in the tumors tested as follows: KIT, 100%; CD34, 75%; PKCtheta, 21%; PDGFRA, 90%; and smooth muscle actin, 6%. Fourteen patients (13%) died of metastatic GIST at a mean age of 45 years (range, 30 to 69 y). Estimated 10 and 40-year survivals were 100% and 73%, respectively. Median survival time was 26.5 years (range, 16 to 60 y). There was no correlation between the National Institutes of Health tumor risk classification and the tumor behavior. Compared with sporadic gastric GISTs, the gastric stromal tumor in Carney triad showed distinctive features: female predilection, young patient age, epithelioid cell predominance, multifocality, frequent lymph node metastasis, serial tumor occurrence, and unpredictable behavior. Thus, the Carney triad gastric stromal tumor is different clinically, pathologically, and behaviorally from sporadic gastric GIST.

Patients with primary sclerosing cholangitis (PSC) are at increased risk for developing cholangiocarcinoma (CCA). However, the morphology of precursor lesions and the prevalence of biliary dysplasia among patients undergoing liver transplantation for PSC are incompletely defined, and the earlier studies using relatively small number of cases have yielded conflicting results. We retrospectively evaluated 100 consecutive formalin-fixed PSC liver explants (including 30 with CCA) by randomly sampling the hilar and large intrahepatic bile ducts (10 additional tissue cassettes submitted per case). The following histologic features were evaluated and quantitated according to the number of ducts involved: mucinous metaplasia, pyloric metaplasia, intestinal metaplasia, pancreatic acinar metaplasia, and biliary dysplasia [low-grade vs. high-grade (biliary intraepithelial neoplasia-2 or neoplasia-3), papillary vs. flat]. Using Fisher exact test and t test, these features were correlated with the presence or absence of CCA and with the following clinical parameters: sex, age, PSC duration, cirrhotic-stage liver disease, and inflammatory bowel disease at the time of transplant. We found high frequencies of mucinous metaplasia (77%), pyloric metaplasia (73%), and pancreatic acinar metaplasia (10%), which did not differ between CCA and non-CCA livers. However, livers with CCA were more likely to harbor intestinal metaplasia (43% vs. 19%, P=0.013), dysplasia (of any grade) (83% vs. 36%, P

Prediction of the biologic behavior of pancreatic endocrine tumor (PET) without local invasion or metastasis is often difficult. The 2004 World Health Organization (WHO) classification uses size, angioinvasion, mitotic activity, and Ki-67 index as prognostic criteria. Recently, cytokeratin 19 (CK19) was shown to be another prognostic marker, but the mechanism by which CK19 predicts prognosis is unknown. As CK19 is the first cytokeratin expressed in all epithelial cells in fetal pancreas, we sought to test expression of other markers of islet cell differentiation including KIT, Pdx-1, Pax4, and Pax6 in PET and correlation of these markers with clinical behavior. Clinical information and histology was reviewed in 97 PETs. All tumors were classified according to WHO criteria and a tumor, node, and metastases stage system. Immunohistochemistry was performed using antibodies to Ki-67, KIT, CK19, Pdx-1, Pax4, and Pax6. Associations of clinicopathologic and immunohistochemical features with prognosis were evaluated using Cox proportional hazards regression models. WHO and tumor, node, and metastases classifications, mitotic counts and Ki-67 labeling, infiltrative border, necrosis, perineural invasion, extrapancreatic extension, tumor size, and positive CK19 and KIT expression were significantly associated with death from disease in a univariate setting. In multivariate analysis, only WHO criteria and KIT expression were shown to be independent. An immunohistochemical classification system was derived from a combination of KIT and CK19 expression: low risk (KIT-/CK19-), intermediate risk (KIT-/CK19+), and high risk (KIT+/CK19+). Survival, metastases, and recurrence of PET were significantly different among the 3 groups. These results indicate that KIT is a new and independent prognostic marker for PETs. The classification system derived from KIT and CK19 was able to predict clinical behavior of PET.

Autoimmune pancreatitis (AIP) is a chronic inflammatory disease of the pancreas that is increasingly encountered worldwide. It has generated considerable interest, in part because the inflammatory process usually responds dramatically to corticosteroid therapy. The most common presentation mimics that of pancreatic cancer; thus, a correct diagnosis of AIP can avoid major surgery. However, the diagnosis is challenging, because its incidence is far lower than that of the diseases it mimics and there is no single diagnostic clinical feature or test that can identify the full spectrum of AIP. Therefore, we are increasingly encountering patients misdiagnosed as having AIP. The misdiagnosis typically occurs in three scenarios: (i) treatment of pancreatic or biliary malignancy with corticosteroids and/or immunomodulators, (ii) treatment of chronic abdominal pain with corticosteroids and/or immunomodulators, or (iii) performance of operative resection for autoimmune disease. Our purpose is to highlight this growing clinical problem and to reinforce the use of published guidelines for the diagnosis and management of AIP.

IgG4-associated systemic disease (ISD) is a multiorgan fibroinflammatory disorder whose pancreatic manifestation is called autoimmune pancreatitis (AIP). We describe 3 patients who developed non-Hodgkin lymphoma during the follow-up of ISD.At our institution's pancreas clinic, we have prospectively and retrospectively examined patients with ISD with (n = 101) or without (n = 10) AIP (mean age, 59 years; 90 males and 21 females). We reviewed the medical records of all 111 patients to identify patients who developed non-Hodgkin lymphoma during the follow-up since their first presentation of ISD. Standardized incidence rate was calculated.The 111 patients with ISD with or without AIP had 331 patient-years of observation during which 3 patients had a diagnosis of non-Hodgkin lymphoma 3 to 5 years after the diagnosis of ISD. In these patients who later developed lymphoma, ISD involved the pancreas (AIP) in 2 and salivary gland in 1. Non-Hodgkin lymphoma had extranodal involvement in all patients (liver [n = 2], adrenal glands [n=1], kidney [n= 1], and lung [n = 1]). Standardized incidence rate was 16.0 (95% confidence interval, 3.3-45.5).We report 3 cases of non-Hodgkin lymphoma that developed during the follow-up of ISD suggesting that patients with ISD may be at an increased risk of developing non-Hodgkin lymphoma.

In immunosuppressed patients with branch duct intraductal papillary mucinous neoplasm (IPMN-Br) associated with solid organ transplantation, the risk of major pancreatic surgery has to be weighed against the risk of progression to malignancy. Recent studies show that IPMN-Br without consensus indications for resection (CIR) can be followed conservatively. We analyzed the course of IPMN-Br in patients with and without solid organ transplant.We compared clinical and imaging data at diagnosis and follow-up of 33 IPMN-Br patients with solid organ transplant (T-IPMN-Br) with those of 57 IPMN-Br patients who did not undergo transplantation (NT-IPMN-Br). In T-IPMN-Br, we noted pre- and post-transplant imaging and cyst characteristics. This case-control study was conducted in a tertiary-care hospital for patients with IPMN-Br.T-IPMN-Br patients were younger than the NT-IPMN-Br patients (63 vs. 68 years, P = 0.01). The median duration of follow-up for the groups was similar (29 vs. 28 months, P = NS). CIR were present in 24% (8/33) of T-IPMN-Br patients and 32% (18/57) of NT-IPMN-Br. New CIR were noted in 6% (2/33) of patients in the T-IPMN-Br group during a median follow-up of 17 months (range, 3-100 months) compared with 4% (2/57) of patients in the NT-IPMN-Br group (P = NS). Eleven patients (10 NT-IPMN-Br, 1 T-IPMN-Br) underwent surgery during follow-up. Only one NT-IPMN-Br patient was diagnosed with malignancy; all others had benign IPMN-Br.In participants with IPMN-Br, short-term follow-up after solid organ transplant was not associated with any significant change in cyst characteristics suggesting that incidental IPMN-Br, even in the setting of immunosuppression post-transplant, can be followed conservatively.

Morphologic, clinical, and genetic evidence suggests that pancreatic intraepithelial neoplasia (PanIN) is a precursor to ductal adenocarcinoma. But understanding precursor lesions in a pancreas with existing tumor is hampered by the fact that chronic pancreatitis often accompanies carcinoma, and the possible interactions between tumor, chronic pancreatitis, and PanIN are complex. Furthermore, cancerization of ducts can mimic high-grade PanIN. Heterotopic pancreas has a genetic make-up, physiologic function, and local environmental exposure similar to that of the pancreas. It offers an opportunity to study putative precursor lesions in a setting with fewer confounding factors. We identified 6 pancreatic cancer patients who had heterotopic pancreas removed at the time of surgery. All 6 cases were immunostained for p53, cyclin D1, and p16. Molecular analysis of K-ras mutation was also done. All 6 cancer-associated heterotopias had PanIN-1A or 1B; 5 had PanIN-2 and 1 had PanIN-3. Three of 6 cases harbored the same K-ras codon 12 mutation as the PanINs in orthotopic pancreas, and a similar pattern of p53, cyclin D1, and p16 expression was observed between heterotopic and orthotopic pancreas in all 6 cases. There was no chronic pancreatitis in the cancer-associated heterotopias, but chronic pancreatitis was seen adjacent to carcinoma in 5 of 6 cases. The presence of PanIN in heterotopic pancreas from patients with ductal adenocarcinoma supports the progression model.

During the last decade, patients with steroid responsive sclerosing cholangitis have been described in a number of case reports and a few cases series. These patients meet diagnostic criteria for primary sclerosing cholangitis (PSC) but seem to have a better prognosis and are often, but not always, associated with autoimmune pancreatitis (AIP). Unlike biliary strictures in PSC, most PSC-like biliary strictures seen in association with AIP respond well to steroids and some have been shown to resolve after immunosuppressive treatment during prolonged follow-up. In addition, steroid-responsive dominant bile duct lesions, sometimes mimicking infiltrating hilar cholangiocarcinoma, without pancreatic abnormalities have also been reported. Here we review the literature on this emerging clinical entity and suggest that this condition of steroid responsive biliary strictures be named Immunoglobulin G4 (IgG4) Associated Cholangitis (IAC). We performed an electronic database search of MEDLINE for case reports and case series of sclerosing cholangitis with pancreatitis and/or pancreatic mass (1960–July 2006). Articles were limited to English and these case reports and cases series were reviewed and all references on the combination of sclerosing cholangitis and pancreatitis as well as IgG4 positive cholangitis without pancreatic involvement not identified by our original search were obtained. Terminology and Definition

OBJECTIVES: Biliary strictures, similar to primary sclerosing cholangitis (PSC), have been reported in patients with autoimmune pancreatitis, which is characterized by elevated serum lgG4 levels and responsiveness to corticosteroids. We sought to determine the frequency of elevated lgG4 in patients with PSC and to clinically compare PSC patients with elevated and normal lgG4 levels. METHODS: We measured serum lgG4 in 127 patients with PSC and 87 patients with primary biliary cirrhosis, as disease controls. Demographic, clinical, and laboratory characteristics were compared between the PSC groups with normal and elevated lgG4 (>140 mg/dL). RESULTS: Elevated lgG4 was found in 12 PSC patients (9%) versus one PBC patient (1.1%) (p = 0.017). Patients with elevated lgG4 had higher total bilirubin (p = 0.009), alkaline phosphatase (p = 0.01), and PSC Mayo risk score (p = 0.038), and lower frequency of IBD (p < 0.0001). Importantly, the time to liver transplantation was shorter in patients with elevated lgG4 (1.7 vs 6.5 yr, p = 0.0009). The type of biliary involvement (intrahepatic, extrahepatic, or both) and pancreatic involvement were similar in both groups. CONCLUSIONS: A small proportion of PSC patients had elevated serum lgG4. In these patients parameters of liver disease severity were more pronounced and time to liver transplantation was shorter, suggesting a more severe disease course. It is possible that this subset of patients behaves similarly to autoimmune pancreatitis patients with biliary strictures, and could potentially respond to corticosteroids. Testing PSC patients for lgG4 and treating those with elevated levels with corticosteroids in clinical trials should be considered.

This study was designed to review the clinical characteristics of surgically treated gastrointestinal stromal tumors of the colon and rectum, evaluate their immunohistochemical and pathologic features based on the current National Institutes of Health criteria, and correlate clinicopathologic findings with the subsequent clinical course. Patient and disease characteristics at presentation, pathologic features, surgical management, and clinical outcomes of 18 patients with gastrointestinal stromal tumors (4 colon and 14 rectum) diagnosed and primarily treated at our institution between 1979 and 2004 were evaluated. Tumors were classified on basis of size and mitotic rate according to current National Institutes of Health recommendations: 67 percent (n = 12) were high-risk, 5 percent (n = 1) were intermediate-risk, 17 percent (n = 3) were low-risk, and 11 percent (n = 2) were very low-risk gastrointestinal stromal tumors. Fifteen of 18 tumors were KIT-positive. The three KIT-negative tumors were platelet-derived growth factor receptor alpha positive. All patients with colonic gastrointestinal stromal tumors (n = 4) underwent segmental resection, whereas patients with rectal gastrointestinal stromal tumors had local excision (n = 5) or radical resection (n = 9). Sixty-six percent (8/12) of patients with high-risk colorectal gastrointestinal stromal tumors developed metastases. None of the patients (n = 6) with intermediate-risk, low-risk, or very low-risk gastrointestinal stromal tumors died of their disease after a median follow-up of 65 (range, 15–266) months. The majority of gastrointestinal stromal tumors of the colon and rectum are high-risk. Patients with high-risk colorectal gastrointestinal stromal tumors have a significant likelihood of developing metastases that is associated with poor prognosis. These patients need to be closely followed for an extended period and should be considered for adjuvant therapy with tyrosine kinase inhibitors.

Enterocolic lymphocytic phlebitis (ELP) is a rare cause of gastrointestinal ischemia. Unlike most vasculitic diseases affecting the gastrointestinal tract, ELP involves only the mural and mesenteric veins, which are surrounded by a lymphocytic and sometimes granulomatous infiltrate. The mesenteric arterial system and the systemic vasculature are characteristically spared. Most patients with ELP present with an acute abdomen that resolves following surgical resection of the involved bowel. ELP has been reported to involve the small bowel, colon, or both, but involvement of the upper gastrointestinal tract has not been previously described. Here we report a case of lymphocytic phlebitis that affected only the stomach and duodenum. The patient, a 68-year-old man, had a nonhealing gastric antral ulcer and underwent hemigastrectomy with vagotomy and Billroth II reconstruction. Both the resected stomach and duodenum showed characteristic lymphocytic and granulomatous infiltrates that involved the submucosal and mural veins, with associated obliteration of vascular lumina; the adjacent arteries were completely spared. The patient developed late postoperative complications including bile reflux gastritis and erosive esophagitis, but he had no recurrence of gastrointestinal ulceration or ischemia over a 2-year follow-up. We hypothesize that there may be more cases of upper gastrointestinal ELP than are diagnosed as such, in part because the diagnosis can be made only on surgical resections specimens.

New-Onset Microscopic Colitis in an Ulcerative Colitis Patient After Fecal Microbiota Transplantation

Microscopic colitis is a relatively common cause of chronic watery diarrhea, often accompanied by abdominal pain and weight loss. The colonic mucosa appears normal grossly, and the diagnosis is made when there is an intraepithelial lymphocytosis and a mixed inflammatory infiltrate in the lamina propria. The two main subtypes, collagenous and lymphocytic colitis, are similar clinically and histologically, distinguished by the presence or absence of a thickened subepithelial collagen band. Many potential pathophysiological mechanisms have been described, although none have been conclusively proved. There is a paucity of randomized treatment trials in these patients, although a rational approach to therapy often leads to satisfactory control of symptoms.

PURPOSE: We updated an Uruguayan family with hereditary nonpolyposis colorectal cancer first described in 1977, incorporating knowledge of how thehMLH1 germline mutation has been established and shown to segregate in accord with the expected autosomal dominant mode of genetic transmission. METHODS: DNA-based molecular genetic testing was performed in conjunction with genetic counseling. Individuals were provided with their genetic test results, so that at-risk family members would be able to benefit from targeted management programs. RESULTS: We counseled 19 members of this kindred, 13 of whom were positive for thehMLH1 germline mutation. Specific recommendations for surveillance and management were provided. We were able to describe follow-up, including anecdotal cancer survival and pathology findings extending from the initial 1977 report of this family to the present. A remarkable sibship within this kindred was comprised of eight siblings, six of whom underwent resections for colorectal carcinoma between 1963 and 1971. Colon carcinomas before 1977 in this sibship were treated with classic hemicolectomies. Of those who had hemicolectomies for their first primary colorectal cancers, two had a second colon cancer primary, and two had a third colon cancer primary. CONCLUSIONS: Attention given to this extended family with hereditary nonpolyposis colorectal cancer has had a positive impact on the physician community in Uruguay, leading to the identification of additional families with hereditary nonpolyposis colorectal cancer.

INTRODUCTION: This study sought to determine whether dietary arginine influences colonic anastomotic healing in the rat model. METHODS: Three groups of 42 Sprague-Dawley rats were fed 0, 1, and 3 percent arginine diets for three preoperative and three postoperative days. Animals underwent transection of the transverse colon with handsewn anastomosis. Subgroups of 14 animals in each dietary group were killed on postoperative Days 6, 10, or 14, and bursting pressures, histologic inflammation, and collagen content were compared. RESULTS: Mean anastomotic bursting pressures on postoperative Day 6 were lower for the 0 percent arginine group than the 1 and 3 percent arginine groups (mean ± standard error of the mean =134±6 mmHg, 164±7 mmHg, and 166±7 mmHg, respectively;P

PURPOSE: Hereditary nonpolyposis colorectal cancer is reported to have special histologic features. This study compares the histologic features of hereditary nonpolyposis colorectal cancer to colorectal cancers from the general population when hereditary nonpolyposis colorectal cancer cases are restricted to families with known MSH2 and MLH1 mutations. METHODS: Thirty-seven cancers from kindreds carrying MSH2 mutations, 27 cancers from kindreds carrying MLH1 mutations, and 37 colorectal cancers from the general population were reviewed by a pathologist blinded to hereditary nonpolyposis colorectal cancer gene status. Tumor grade, growth pattern, Crohn's-like lymphoid reaction, mucin production, extent of disease in the bowel wall, and lymph node status were evaluated. RESULTS: Poor differentiation and Crohn's-like reaction were a feature of 44 and 49 percent of hereditary nonpolyposis colorectal cancer compared with 14 percent (P=0.002) and 27 percent (P=0.049) of colorectal cancers from the general population, respectively. There was no difference in growth pattern, mucin production, lymph node involvement, or local extent of disease between hereditary nonpolyposis colorectal cancer and colorectal cancers from the general population. Poor differentiation and lymph node metastases were found in 57 and 49 percent of MSH2 compared with 26 percent (P=0.002) and 10 percent (P=0.03) of MLH1-associated cancers, respectively. There was no difference in growth pattern, mucin production, Crohn's-like lymphoid reaction, or local extent of disease between subgroups of hereditary nonpolyposis colorectal cancer. CONCLUSIONS: Poor differentiation and Crohn's-like reaction are more common in hereditary nonpolyposis colorectal cancer than colorectal cancers from general population. Poor differentiation and lymph node metastases are more commonly seen in MSH2-associated cancers than MLH1. Evaluation of the natural history, pathogenesis, and prognosis of colorectal cancer in hereditary nonpolyposis colorectal cancer should include consideration of which mismatch repair genes are mutated and what the specific mutations are.

Recent advances in hereditary nonpolyposis colorectal cancer (HNPCC) have been made based on the discovery early in this decade that germline mutations in genes responsible for repair of DNA mismatches formed the molecular basis for the syndrome. Several studies during the past year described the prevalence of germline mutations in those deemed at risk for HNPCC and helped define who should be tested for such mutations. Investigators are also beginning to make connections between genotype and phenotype; it appears that certain mutations are more likely than others to generate a broad spectrum of extracolonic tumors. Carcinogenetic mechanisms in HNPCC also received attention; evidence continues to accumulate that the critical somatic mutations driving malignant transformation in HNPCC (and in sporadic colorectal cancer with microsatellite instability) are different from the critical mutations seen in most colon cancers. Finally, several contributions dealt with the complicated question of how to manage germline carriers and affected individuals.

Objective The authors investigate the effects of gastric juice on tumorigenesis in a rat model of esophageal adenocarcinoma. Summary Background Data In rats treated with the carcinogen methyl-n-amyl nitrosamine, squamous cancer of the esophagus develops in a time- and dose-dependent manner. When methyl-n-amyl nitrosamine treatment is preceded by an operation to induce reflux of duodenal and gastric juice into the esophagus, there is an increased yield of esophageal tumors, many of which are adenocarcinomas. When only gastric juice refluxes into the esophagus, the tumor yield is less and adenocarcinomas are not found. Methods Two hundred seventy 8-week old Sprague-Dawley rats were studied. Twenty unoperated rats served as controls. The remaining rats underwent the following operations : esophagoduodenostomy with gastric and vagal preservation to induce duodenogastroesophageal reflux (n = 48) ; esophagoduodenostomy with antrectomy and Billroth 1 reconstruction to produce reflux of duodenogastric juice with the exclusion of the antrum (n = 53) ; esophagoduodenostomy with proximal gastrectomy to induce hypergastrinemia and reflux of duodenogastric juice with exclusion of the body and forestomach (n = 51) ; esophagoduodenostomy plus total gastrectomy to produce reflux of duodenal juice alone (n = 50) ; and esophagoduodenostomy with vagal and gastric preservation but with division of the duodenum just beyond the pylorus and reimplantation into the jejunum, 13 cm distal to the esophagoduodenostomy. This produced reflux of duodenal juice with gastric juice diverted downstream, (n = 48). At 10 weeks of age, all rats were given 4 weekly doses of carcinogen (methyl-n-amyl nitrosamine, 25 mg/kg intraperitoneally), and survivors were killed at 36 weeks of age. Results The prevalence rate of esophageal adenocarcinoma was 30% in rats with duodenogastroesophageal reflux and 87% in rats with reflux of duodenal juice alone. Fifty-six percent of rats with reflux of duodenogastric juice with exclusion of the antrum and 72% of rats with reflux of duodenogastric juice with the exclusion of the body and forestomach developed adenocarcinoma, showing a progressive increase in the prevalence of adenocarcinoma as less gastric juice was permitted to reflux with duodenal juice into the esophagus. Conclusion In this rat model, the presence of gastric juice in refluxed duodenal juice protects against the development of esophageal adenocarcinoma. The protective effect appears to be due to acid secretion from the stomach. Continuous profound acid suppression therapy may be detrimental by encouraging esophageal metaplasia and tumorigenesis in patients with duodenogastroesophageal reflux.

Angiomyolipoma (AML) is a rare tumor characterized histologically by a mixture of spindle cells, adipose tissue, epithelioid cells, and blood vessels. AML usually occurs in the kidney but can involve the liver and, rarely, other sites. We describe a 74-year-old woman without tuberous sclerosis who presented with spontaneous hemorrhage into a primary AML of the pancreas and underwent curative surgical resection. We are aware of only one previous report describing a clear cell "sugar" tumor of the pancreas, which probably represents a monotypic epithelioid variant of AML. The pancreas should be included among sites that can give rise to AML.

PURPOSE: The effect of prostaglandin E1 (PGE1) and corticosteroids alone and in combination were studied in the healing rat colon to determine whether PGE1 could not only improve healing but reverse the negative effect of steroids on colonic wound healing. METHODS: Colonic anastomoses were performed in 144 male Sprague-Dawley rats divided into four groups. The control group (I) received no further treatment. The steroid group (II) received cortisone acetate (5 mg/kg/day) beginning six days preoperatively and continuing until sacrifice. The PGE1 group (III) received 2 μg of PGE1 intra-aortically at surgery and for three days postoperatively. The combination PGE1/steroid group (IV) received both drugs in the same doses as those in Groups II and III. Animals were sacrificed on postoperative days 6, 10, and 14. Wound healing was evaluated by hydroxyproline content, bursting pressures, and histology. RESULTS: The hydroxyproline assay at day 10 revealed that steroid-treated rats have significantly lower levels than any other group. The PGE1 group (III) had the highest level of significance in comparison to the steroid group (II) (P=0.001). The addition of PGE1 to steroid (Group IV) appeared to abolish the negative effect of the steroid as measured by hydroxyproline content on day 10 (P=0.038). When measuring bursting pressures, the PGE1 group (III) had significantly higher pressures than any other group at day 10. However, no amelioration of the steroid effect on bursting pressures was seen. Histologic evaluation of the anastomosis did not reveal any significant differences among the four groups. CONCLUSIONS: PGE1 reverses the negative effect of the steroid on hydroxyproline levels at day 10. Furthermore, using bursting pressure as a parameter of wound healing, administration of PGE1 results in significantly improved anastomotic healing at day 10.

We describe the clinical and pathologic features in four extended kindreds that are consistent with the hereditary flat adenoma syndrome (HFAS). This colon cancer susceptibility disorder is believed to be inherited as an autosomal dominant. The principal phenotypic marker is multiple colonic adenomas (usually less than 100), with a tendency for proximal location. The majority of these adenomas are flat or slightly raised and plaquelike, as opposed to polypoid. Colon cancers have typically developed in middle age and show no unusual histologic features. There are a variety of extracolonic manifestations, including adenomas and carcinomas of the small bowel and fundic gland polyps. The HFAS is contrasted with hereditary nonpolyposis colorectal cancer and familial adenomatous polyposis (FAP) and shown to be distinct from both in the numbers and distribution of colonic adenomas and the typical age of cancer diagnosis. The clinical implications of these findings are discussed. Given its linkage to the FAP locus on 5q and the phenotypic parallels between HFAS and FAP, we conclude that HFAS is a variant of FAP.