Prognostic value of various staging and grading systems in proximal colon cancer.

OBJECTIVE: To compare the ability of various methods of staging and grading to predict survival in proximal colon cancer. DESIGN: Retrospective study. SETTING: University hospital, Finland. SUBJECTS: 153 patients with primary proximal colon cancer. MAIN OUTCOME MEASURES: Staging by four classification systems, grading by two grading systems, and survival analysis based on Kaplan-Meier survival curves. RESULTS: In all staging systems the survival curves of different stages differed significantly from each other. The modified Dukes' classification was still the best predictor of survival. Grade of tumours had no significant effect on long term survival, but short term survival was affected adversely by the presence of anaplastic tissue. Tumours with no mucin had a worse prognosis than those that produced mucin. All staging methods were superior to either of the histological grading systems tested as prognosticators. Tumour depth correlated with the operator's clinical impression of radicality of operation, and also predicted survival. CONCLUSION: The clinicopathological modified Dukes' staging system was the most powerful prognosticator in proximal colonic cancer and its use in clinical practice should continue. Even a small amount of anaplastic tissue (> or = 5%) had an adverse effect on short term survival.