Background: Recently, we have analyzed new prognostic markers in colorectal cancer including neuroendocrine differentiation, overexpression of the sialyl-Lex antigen, overexpression of the peripheral benzodiazepine receptor (PBR), BAX protein expression and p53 mutational status. The predictive power of all markers in combination has not yet been evaluated.
Patients and methods: Between 1989 and 1991, 48 consecutive patients underwent surgery for stage III colorectal cancer at our hospital. All patients received a complete 5-year follow-up. Paraffin-embedded tumor samples were analyzed for all 5 markers. Multivariate discriminant analysis was performed to determine the prognostic value of all markers in combination.
Results: Based on these prognostic markers a mathematical discriminant function was obtained. This function allowed to correctly predict the further course of disease in 77% of the patients (specificity: 83.3%, sensitivity: 70.8%). The discriminant function was confirmed in another group of 19 patients. Single marker analysis allowed the prediction of the further course of disease only in 58-70%.
Conclusion: Our study shows that in colorectal cancer, multimarker analysis is superior to unimarker analysis in predicting prognosis. The derived discriminant function allows patient stratification according to risk. Therefore, a multimarker analysis provides a rationale for future individualized risk-adapted therapies in stage III colorectal cancer.