Apoptosis in rectal cancer: prognostic significance in comparison with clinical histopathologic, and immunohistochemical variables.

Purpose: The aim of this study was to evaluate the prognostic value of the apoptotic index for recurrence and disease-free survival after curative surgery for rectal cancer, particularly in relation to clinicopathologic variables, p53- and bcl-2 expression.

Methods: Formalin-fixed, paraffin-embedded tissue samples of rectal carcinomas resected curatively within a five-year period were used (N = 160). Apoptotic cells with fragmented DNA were detected by the terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphatase-biotin nick-end-labeling method. The ratio of apoptotic tumor cells (in percent) was classified into low apoptotic index (less than 10 percent) and high apoptotic index (10 percent or more). Immunohistochemical analysis was performed using monoclonal antibodies (DO-1 for p53 and clone 124 for bcl-2). Statistics included univariate and multivariate analysis, and survival was calculated using the Kaplan-Meier method.

Results: Seventy-five percent of tumors showed a low apoptotic index, and 25 percent had a high apoptotic index. No correlation was found between apoptotic index and International Union Against Cancer stage (P > 0.05). However, significant correlations were documented with histologic differentiation (mean apoptotic index, 5.74 percent in moderately vs. 3.98 percent in poorly differentiated carcinomas; P = 0.0173), lymph node involvement (mean apoptotic index, 6.11 percent in pN1 vs. 3.72 percent in pN2; P = 0.0074), p53 status (mean apoptotic index, 6.26 percent in p53- vs. 4.42 percent in p53+; P = 0.0085), and bcl-2 expression (mean apoptotic index, 5.13 percent in bcl-2- vs. 6.51 percent in bcl-2+; P = 0.0418). Tumors of the lower rectum had a lower apoptotic index than those of the upper rectum (P = 0.0277). Neither univariate nor multivariate analysis assessed apoptotic index as predictor of prognosis: Recurrence rates did not differ between tumors related to apoptotic index (22 percent with low apoptotic index vs. 15 percent with high apoptotic index; P > 0.05), and no significant differences were found regarding survival (P > 0.05). On multivariate analysis, International Union Against Cancer stage (P = 0.0002), p53 (P = 0.0002), gender (P = 0.0136), and bcl-2 (P = 0.0243) were independent predictors of recurrence. These variables, except for bcl-2, were also independently related to disease-free survival.

Conclusions: Reflecting tumor biology, apoptotic index as single variable showed no prognostic significance, whereas p53 was an independent predictor for both recurrence and survival, and bcl-2 was independently related to recurrence, but not to survival. Clinically, International Union Against Cancer stage and gender were independent prognostic factors after curative surgery for rectal cancer.