Allelic imbalance in gastric cancer: an affected site on chromosome arm 3p.

In order to detect regions of DNA containing tumor suppressor genes involved in the development of gastric cancer, we performed an allelotype study on 78 gastric adenocarcinomas from a population composed largely of Texan Hispanics and Anglos, two ethnic groups that have a ratio of incidence rates of gastric cancer of approximately 2:1. In total, 42 microsatellite markers were employed, which detected at least one site per arm of each autosome in the human genome. These included several markers linked to known tumor suppressor genes (TP53, APC, DCC, RB1, and BRCA1). Sites showing quantitative allelic imbalance (AI) greater than 30% were located on 3p (36%), 11q (31%), 12q (38%), 13q (33%), 17p near TP53 (74%), and 17q near BRCAI (32%). Among the 22% of cases showing microsatellite instability (MI), a subset (4 of 17) showed instability at 59% or more of sites tested. No ethnic bias was detected in cases showing MI or in cases with AI at sites with rates of AI above 30%. Tumors of the intestinal subtype were significantly more likely than diffuse tumors to show AI at DI3S170 (P = 0.01). A deletion map of chromosome arm 3p was prepared for tumors with AI at D3S1478. These data indicate that a tumor suppressor gene on chromosome arm 3p is involved in the development of a subset of gastric cancers.

p53 mutations in gastric and colorectal cancers in Texas Hispanics versus Anglos.

Gastric cancer is more than twice as common in Hispanics as in Anglos in Texas, while colorectal cancer is almost twice as common in Anglos as Hispanics. To test the hypothesis that mutations in the p53 tumour suppressor gene are involved in these differences, we examined 131 gastric and 138 colorectal cancers from Hispanic and Anglo patients from South Texas and Mexico using immunohistochemistry (IHC) as a screening assay for p53 mutations. The fraction of p53 positive cases was not significantly different in gastric cancers from Hispanics compared to Anglos (43% versus 61%, respectively, p = 0.13) or in colorectal cancer (57% versus 58%, respectively, p = 1.0), suggesting that p53 mutations are not involved in causing the different incidences of these cancers in these populations. In addition, the types of p53 mutations arising in gastric tumours from Hispanic patients were consistent with those reported in gastric tumours in other populations. Sequencing of mutations in five gastric cancers revealed two G:C to A:T transitions, two A:T to G:C transitions and one complex deletion. In contrast with findings in studies in other tumour types, neither stage nor survival was associated with p53 positive staining by IHC in either gastric or colorectal tumours in this study. Positive p53 immunostaining was associated with the diffuse histological subtype in gastric carcinoma (p = 0.05) and high histological grade in colorectal carcinoma (p = 0.04).