COLORECTAL-CANCER AFTER GASTRIC-SURGERY - A METAANALYSIS

Background: Until now, the question of whether previous gastric surger y results in an increased risk of colorectal cancer has remained contr oversial. A meta-analysis was performed to resolve this issue. Methods : The numbers of colorectal tumors in the two populations with and wit hout gastric surgery were retrieved from all relevant articles found t hrough a MEDLINE, search. If available, the most detailed breakdown of the data by patient gender, type of peptic ulcer, type of surgery, ti me since gastric surgery, and by tumor location was chosen. The Mantel -Haenszel procedure was used to calculate a weighted odds ratio from t he individual studies. The method of DerSimonian and Laird was used to calculate a weighted average of the difference between tumor rates in patients with and without gastric surgery. Results: The summary odds ratio was 1.09 with a 95% confidence interval of 0.96-1.24. The length of time after surgery, gender, tumor location, initial ulcer type, an d different types of surgery were not associated with any significant influence. The summary rate difference was not significantly different from zero. The test for heterogeneity yielded a chi(2) = 26.87, df = 13, p = 0.013, After ignoring one obvious outlier study, the chi(2) te st dropped to a nonsignificant level. A L'Abbe plot failed to show any prevailing influence of rate size on the overall heterogeneity among the various studies. Conclusions: Surgery for peptic ulcer does not re sult in an increased risk for colorectal tumors. This lack of influenc e was shown consistently by all tests of the present meta-analysis.