Purpose: The aim of this study was to obtain additional biologic determinants that may be of use in segregating into subgroups with different prognosis patients with similarly staged colorectal cancers.
Methods: Between 1989 and 1991, a prospective study of prognostic factors has been performed in a group of 98 consecutive, unselected patients who underwent curative resections for primary untreated large bowel carcinoma. The fate of all patients is known at three years after operation. Clinical and pathologic data were recorded at the time of presentation and operation, and patients have been the subjects of regular follow-up. Tumor DNA content was determined by flow cytometry, and cell proliferative activity was determined by autoradiography with tritiated thymidine labeling index (LI).
Results: Univariate analysis revealed that the most important predictors of survival (P < 0.001) were the presence of positive lymph nodes, the presence of preoperative complications, Dukes stage, and LI. The multivariate analysis showed that Dukes stage (P < 0.002) and LI (P < 0.0001) were the only factors significantly related to survival. Disease-free survival was influenced significantly by Dukes stage (P < 0.001), LI, according to the classification in the two groups of high and low proliferative activity, respectively, (P < 0.0001), LI, calculated as a continuous variable (P < 0.0002), and the presence of lymph node metastases (P < 0.003). Outcome (favorable/unfavorable) was influenced significantly by Dukes stage (P < 0.0001) and LI (P < 0.0001). Concordance for each patient between Dukes stage and outcome was 73.1 percent and between LI, calculated as a continuous variable, and outcome was 74.1 percent. If, on the other hand, Dukes stage and LI are used together, concordance with outcome reaches 89.2 percent.
Conclusion: We can conclude that, from a practical point of view, LI is an essential factor that must be combined with pathologic variables for a better prediction of patient outcome.