Amsterdam criteria for the hereditary nonpolyposis colorectal cancer (HNPCC) exclude most suspect cases of possible hereditary colorectal cancer (CRC). By contrast, revised Bethesda guidelines excessively broaden the disease spectrum. The aim of this study is to retrospectively evaluate the cliniciopathilogical characteristics of patients fulfilling the revised Bethesda guidelines and to review the efficacy and limitations of the revised guidelines.
This retrospective study enrolled 3609 patients who underwent curative surgery for primary CRC. Patients were classified into the Bethesda group or the control group according to whether they fulfilled the revised Bethesda guidelines. Patients were further categorized when they fulfilled a minimum of 2 items of the revised guidelines. Individual items were analyzed for deficient mismatch repair (d-MMR).
The median follow-up was 82.9 (interquartile range, 72–101) months. Patients in the Bethesda group were younger and had a higher rate of reduced mismatch repair (MMR) protein expression, microsatellite instability, and right colonic involvement (all P < 0.001) than the control group. As a predictor of d-MMR, the revised Bethesda guidelines showed a sensitivity of 63.0% and a specificity of 72.6%. Items 1 and 2, respectively, or the item pair 1 and 2, were independent predictors of d-MMR (all P < 0.001). Patients fulfilling the Bethesda guidelines showed clinicopathological features of HNPCC.
The revised Bethesda guidelines appear to be a competent predictor of d-MMR. Specifically, items 1 and 2 are significant predictors of d-MMR and may be relevant to the application of the revised Bethesda guidelines.