Prognostic implications of BAX protein expression and microsatellite instability in all non-metastatic stages of primary colon cancer treated by surgery alone.

PURPOSE: This study examined whether the apoptosis-related protein, BAX, or the microsatellite-instability phenotype provide prognostic information in patients with resected colon cancer. METHODS: A total of 371 stage I-III patients that previously underwent radical surgery were included (mean follow-up 51.8 months). BAX expression was examined by immunohistochemical staining; high-frequency microsatellite instability (MSI+) was determined by assessing the specific marker, BAT26, using single-strand conformation polymorphism (SSCP)-based analysis. RESULTS: High BAX expression was found in 66.4% of patients. MSI+ tumors were observed in 14.8% of 344 patients. Univariate analysis showed that unlike MSI, low BAX expression was significantly correlated with poor disease-specific overall survival (OS) in stages I-III (p = 0.04). Multivariate subgroup analyses revealed that unlike MSI, low BAX was an independent predictor for OS in stage II (p = 0.009); however, in stages I or III, BAX or MSI were not independent predictors of OS. CONCLUSIONS: In stage II colon cancer treated with surgery alone, BAX protein expression may be a predictor for prognosis.

Low BAX protein expression correlates with disease recurrence in preoperatively irradiated rectal carcinoma.

PURPOSE: To determine the prognostic impact of BAX in correlation to its upstream effector p53 as well as clinicopathologic variables and patient outcome in preoperatively irradiated rectal carcinoma. METHODS AND MATERIALS: We investigated 92 rectal carcinoma patients treated by preoperative radiotherapy to a total dose of 30 Gy followed by surgery. Median follow-up was 71 months. Immunohistochemistry was performed on paraffin sections of pretreatment biopsy samples for BAX protein. Also, we considered the previously determined p53 expression data from this cohort. RESULTS: BAX protein expression was classified as high and low in 63 (68.5%) and 29 (31.5%) tumors, respectively. Unlike clinicopathologic variables, high BAX expression was significantly associated with improved disease-free survival by univariate analysis (p = 0.048). Moreover, in multivariate analyses, high BAX expression was an independent prognostic indicator for both improved local recurrence-free interval and improved disease-free survival (p = 0.03 and 0.047, respectively). Concerning the p53/BAX pathway, subgroup analysis yielded no association between p53 immunonegative/BAX high vs. p53 immunopositive/BAX low expressing tumors with regard to overall, disease-free, or local recurrence-free survival in either univariate (p = 0.88, 0.54, and 0.16, respectively) or multivariate analysis. CONCLUSIONS: This study demonstrates that BAX protein expression might help to predict disease recurrence in preoperatively irradiated rectal carcinoma, whereas determination of p53, the proposed upstream regulator of BAX-induced apoptosis, did not provide additional prognostic information.