The relationship between different staging methods and survival in colorectal carcinoma.

A routine clinicopathologic (CP) staging system for patients who have had surgical resection for colorectal carcinoma was established at Concord Hospital in 1971. Research on this prospective series of resections has evaluated the CP staging system as a guide to prognosis. The aim of this study was to compare the CP system with the classic Dukes' staging system and its modified form introduced by Astler and Coller to determine which method provided the most accurate basis for prognosis. Life table survival analysis was used to examine the survival of 709 patients according to each staging system. Relative mortality rates for groups of patients cross-classified by each possible pair of staging systems were examined and the Cox regression model was used to determine the independent effects of staging by each system on survival. The CP system was found to have a stronger association with survival than either the classic Dukes' system or the Astler-Coller modification of the Dukes' system. The importance of supplementing data on the operative specimen with data about the spread of tumor beyond the limits of surgical resection is emphasized.