Improvement of staging by combining tumor and treatment parameters: the value for prognostication in rectal cancer.

Abstract

BACKGROUND & AIMS: Staging of cancer is based on the TNM system. This valuable system takes only tumor-related parameters into account, but in the era of refined surgery and preoperative therapy treatment-related factors are of equal importance. By using rectal cancer as a model we explored the hypothesis that a combination of tumor- and treatment-related parameters will result in improved prognostication. METHODS: Standardized clinicopathologic and histologic factors considered predictive for survival were studied in eligible patients treated in a trial for rectal cancer (n = 1324). These factors were analyzed in relation to survival using log-rank tests, Kaplan-Meier curves, and Cox regression both individually and in combination, the latter including TNM staging. A second data set from an independent trial (n = 316) was used for data validation. RESULTS: Multivariate analysis identified nodal status (P = .001) and circumferential margin (P = .001) involvement as the most important prognostic factors for survival. The combination of these factors formed an improved staging system (node status and circumferential margin [NCRM]) compared with the present TNM staging with respect to 5-year cancer-specific survival. The results were confirmed in our independent patient population. CONCLUSIONS: NCRM staging of rectal cancer results in a broad range of survival rates and favorable patient grouping. Our data give strong evidence that a staging system combing tumor- and treatment-related factors provides better prognostic information than the classic TNM system, which is based solely on tumor-related factors. Similar results might be obtained in other types of cancer in which quality of treatment is important for outcome.