Use of microsatellite instability and immunohistochemistry testing for the identification of individuals at risk for Lynch syndrome.

It is now generally recognized that a specific subset of those patients clinically defined as having hereditary non polyposis colon cancer (HNPCC) have germline mutations in any one of several genes involved in DNA mismatch repair (MMR). This important subset of HNPCC families is now defined as having Lynch syndrome. A considerable amount of data has shown that tumors from patients with Lynch syndrome have characteristic features resulting from the underlying molecular involvement of defective MMR, that is, the presence of microsatellite instability (MSI) and the absence of MMR protein expression by immunohistochemistry (IHC). As a result, identifying patients with Lynch syndrome can now be accomplished by testing tumors for these tumor-related changes. Together, MSI and IHC are powerful tools that help identify individuals at risk for having Lynch syndrome and to distinguish these cases from HNPCC cases with other hereditary gene defects. Furthermore, IHC analysis provides valuable clues as to which MMR gene is mutated, allowing for comprehensive mutational analyses of that gene. Here, we discuss the current and historical perspectives regarding MSI and IHC analyses in tumors from sporadic colon cancer and from patients with Lynch syndrome. Given this background, we also provide a testing strategy for the identification of patients at risk for Lynch syndrome and subsequent gene testing.

Isolated loss of PMS2 expression in colorectal cancers: frequency, patient age, and familial aggregation.

PURPOSE: Most colorectal cancers that have high levels of microsatellite instability (MSI-H) show loss of immunohistochemical expression of proteins that participate in the DNA mismatch repair process, most often involving MLH1 and MSH2. Less commonly, a third DNA mismatch repair protein, MSH6, may also be lost as the primary event. Rarely, tumors with MSI-H show normal expression of these three proteins. The genetic deficiency leading to the MSI-H phenotype in such cases is unknown. PMS2 is another member of the DNA mismatch repair complex. Its expression is generally lost in tumors with MLH1 loss of expression. Rarely, there is selective loss of PMS2 expression. We sought to describe the frequency and clinical correlates of selective loss of expression of PMS2 with the MSI-H tumor phenotype. EXPERIMENTAL DESIGN: Two thousand seven hundred nineteen colorectal cancers from both clinic- and research-based ascertainment were studied. Tumor MSI testing and immunohistochemistry for MLH1, MSH2, MSH6, and PMS2 were conducted. Medical records were abstracted for age at diagnosis, gender, colorectal cancer site, and family history. RESULTS: Five hundred thirty-five of the 2,719 tumors were MSI-H. Of these, 93% showed loss of expression of MLH1, MSH2, and/or MSH6. Thirty-eight showed normal expression for these proteins. PMS2 immunohistochemical staining was successful in 32 of 38 of these tumors. Of the 32, 23 showed selective loss of expression of PMS2. This was associated with young age of diagnosis and right-sided location but not with a striking family history of cancer. CONCLUSIONS: Overall, 97% of the MSI-H tumors showed loss of expression for one or more of these four mismatch repair proteins. Selective loss of expression of PMS2 was present in 72% of cases in which colorectal cancers had an MSI-H phenotype but no alteration of expression of MLH1, MSH2, and MSH6. The underlying mechanism involved cannot be determined from this study but could involve point mutations in other DNA mismatch repair genes with retention of immunohistochemical expression, somatic inactivation of PMS2, or germ line mutation of PMS2.

Widespread EMR: a new technique for removal of large areas of mucosa.

BACKGROUND: Widespread EMR is a novel technique for resection of large areas of mucosa as a single bloc. Large lesion size (>2 cm) is a recognized limitation of current EMR techniques. This study assessed the technical feasibility, efficacy, and safety of widespread EMR in a porcine model. METHODS: Widespread EMR was performed in 6 pigs. A submucosal fluid cushion was created in the distal esophagus by using a 0.83% solution of hydroxypropyl methylcellulose. A mucosal strip 5 cm in length and including at least 50% of the luminal circumference was marginated by using a prototype cutting device. The proximal end of the mucosa was stripped from the submucosa by using a grasping forceps. The distal end was snare resected. Resected tissue was assessed histologically. Endoscopy was repeated at weeks 1 and 6. At week 6, a second widespread EMR of the mucosa on the wall opposite the initial resection was attempted to create a full circumferential resection. RESULTS: Widespread EMR was completed in all animals. The esophagus was denuded of mucosa in 5-cm lengths that included 50% of the circumference of the lumen. No procedure-related complication was observed. Histologic assessment demonstrated that the resection specimen included mucosa and submucosa but not muscularis propria. Endoscopy at 1 week revealed mucosal regrowth in two animals and ulceration in 4. At week 6, regrowth of mucosa was noted in all animals. The second wide EMR proved to be technically difficult and was associated with perforation, peri-esophageal abscess, and stricture formation. CONCLUSIONS: Wide EMR appears to be technically feasible for removal of large areas of mucosa. Mucosal strips 5 cm long that included over 50% of the luminal circumference were resected safely. Resection was followed by complete regrowth of the mucosa. However, a second wide EMR to create a circumferential resection proved to be technically difficult and unsafe.

Colorectal tumour microsatellite instability test results: perspectives from patients.

PURPOSE: To determine which individuals with colorectal cancer (CRC) were interested in knowing the results of their tumour microsatellite instability (MSI) and immunohistochemistry (IHC) testing. We were also interested in the patients' reasons for choosing to learn their results and in the impact of those results on overall self-assessed quality of life. PATIENTS AND METHODS: CRCs from 414 individuals were assayed for MSI and IHC for DNA mismatch repair gene products (MLH1, MSH2, MSH6). Individuals were invited to learn their MSI/IHC results. They randomly received either brief or extended educational materials about the testing and a pretest survey to learn reasons for their interest and to assess their pretest quality of life. RESULTS: Of the 414 individuals, 307 (74%) chose to learn their results. There was no significant difference in interest in knowing test results according to gender, age, educational level, or family history of colon cancer. The level of detail in the information piece received by the patients did not influence their desire to know their test results. Self-assessed quality of life was not altered by receiving results and was not correlated with the test outcome. CONCLUSIONS: Individuals with colorectal cancer had a high level of interest in learning their individual MSI/IHC test results and did not seem deterred by the inherent complexity or ambiguity of this information. Regardless of test outcome, results did not significantly affect self-assessed quality of life. Further studies are needed to assess comprehension of results and behavioural changes resulting from the learning of MSI/IHC results.

First endoluminal system for transmural resection of colorectal tissue with a prototype full-thickness resection device in a porcine model.

BACKGROUND: Endoluminal transmural resection of colorectal lesions is a pivotal advance in endoscopic technology. A full-thickness resection device has been developed that functions through a combination of tissue grasping, stapling, and cutting under endoscopic guidance. This preclinical study evaluated the performance, safety, and effectiveness of the full-thickness resection device in a porcine model. METHODS: The full-thickness resection device consists of an operative handle, flexible shaft, and a resection chamber head. Eight pigs were randomized to 2 survival groups (4 each) of 14 and 28 days. The mucosa was marked electrosurgically to delineate target areas in the distal colon. A gastroscope inserted through the channel of the full-thickness resection device was advanced together with the device to the target. Targeted tissue was pulled into the resection chamber with a forceps, staples were deployed, and the isolated specimen was resected with a cutting blade. Histologic and radiographic evaluations were performed. RESULTS: All resections were transmural. Mean resected tissue diameter was 3.6 cm (1.4-5.2 cm). Mean procedure time was 30.2 (15) minutes. Minor mechanical problems required the use of replacement devices to complete 3 procedures. Resection sites were identified endoscopically and radiographically at sacrifice. In one animal, local adhesions were present. Histology evaluation disclosed resection line mucosal ulceration in 4 animals with fibroinflammatory changes consistent with healing. CONCLUSIONS: The full-thickness resection device can be used for endoluminal transmural localized resection of colorectal tissue in an animal model.

Chemoprevention of esophageal adenocarcinoma by COX-2 inhibitors in an animal model of Barrett's esophagus.

BACKGROUND & AIMS: Carcinogenesis in Barrett's esophagus (BE) is associated with an increased expression of cyclooxygenase (COX) 2. However, there has been no direct evidence that inhibition of COX-2 prevents cancer in BE. We studied the effect of MF-Tricyclic, a selective COX-2 inhibitor, on the development of BE and adenocarcinoma in a rat model. METHODS: Four weeks after esophagojejunostomy, 105 Sprague-Dawley rats were randomized to a chow containing MF-Tricyclic or Sulindac, or a placebo. Ninety-six (92%) rats completed the study and were sacrificed at 28 +/- 2 weeks. The animals were assessed for the presence of cancer, tumor volume, BE, degree of inflammation, and COX-2 expression and activity. RESULTS: MF-Tricyclic and Sulindac reduced the relative risk of development of esophageal cancer by 55% (95% confidence interval [CI] = 43%-66%, P < 0.008) and by 79% (95% CI = 68%-87%, P < 0.001), respectively, compared with controls. No significant differences were noted in the risk of esophageal cancer between the MF-Tricyclic and the Sulindac group (P = 0.34). The median tumor volume was not significantly different among the 3 groups (P = 0.081). Moderate to severe degree of inflammation was significantly more common (P = 0.005) in the control compared with the MF-Tricyclic and the Sulindac group; however, the prevalence of BE was not significantly different between groups (P = 0.98). Rats in the control group had higher tissue PGE2 level compared with the MF-Tricyclic and Sulindac groups (P = 0.038). CONCLUSIONS: Selective and nonselective COX-2 inhibitors can inhibit inflammation, COX-2 activity, and development of adenocarcinoma induced by reflux. This provides direct evidence that COX-2 inhibitors may have chemopreventive potential in BE.

2-Octyl-cyanoacrylate (Dermabond), a new glue for variceal injection therapy: results of a preliminary animal study.

BACKGROUND: Endotherapy of bleeding gastric varices is problematic. The aim of this descriptive study in an animal model was to compare 2-octyl-cyanoacrylate (Dermabond), a Food and Drug Administration-approved agent for superficial wound closure, to N-butyl-2-cyanoacrylate (Histoacryl), an agent available outside of the United States for the endoscopic treatment of bleeding gastric varices. METHODS: Eight New Zealand white rabbits were randomly assigned to either the study agent Dermabond or the control agent, Histoacryl. Both active agents were equally mixed with ethiodized poppy seed oil (Ethiodol) and injected into the auricular vein. The animals were euthanized at various time intervals after injection. A 2 cm strip of tissue on either side of the injection site along with the adjacent perivascular tissues were resected for histologic evaluation. RESULTS: The use of 0.5 mL of Dermabond effectively induced vascular occlusion compared with 0.2 mL of Histoacryl. The histologic changes acutely and at 1 week were similar with each mixture. CONCLUSION: Dermabond may be useful in the treatment of gastric variceal bleeding, but further studies are necessary to determine dose response rates in animals and humans.

Immunohistochemistry versus microsatellite instability testing in phenotyping colorectal tumors.

PURPOSE: To compare microsatellite instability (MSI) testing with immunohistochemical (IHC) detection of hMLH1 and hMSH2 in colorectal cancer. PATIENTS AND METHODS: Colorectal cancers from 1,144 patients were assessed for DNA mismatch repair deficiency by two methods: MSI testing and IHC detection of hMLH1 and hMSH2 gene products. High-frequency MSI (MSI-H) was defined as more than 30% instability of at least five markers; low-level MSI (MSI-L) was defined as 1% to 29% of loci unstable. RESULTS: Of 1,144 tumors tested, 818 showed intact expression of hMLH1 and hMSH2. Of these, 680 were microsatellite stable (MSS), 27 were MSI-H, and 111 were MSI-L. In all, 228 tumors showed absence of hMLH1 expression and 98 showed absence of hMSH2 expression: all were MSI-H. CONCLUSION: IHC in colorectal tumors for protein products hMLH1 and hMSH2 provides a rapid, cost-effective, sensitive (92.3%), and extremely specific (100%) method for screening for DNA mismatch repair defects. The predictive value of normal IHC for an MSS/MSI-L phenotype was 96.7%, and the predictive value of abnormal IHC was 100% for an MSI-H phenotype. Testing strategies must take into account acceptability of missing some cases of MSI-H tumors if only IHC is performed.