A predictive model of survival after radical cystectomy for carcinoma of the bladder.

Abstract

To evaluate the effect of patient and tumour characteristics on the disease-free survival after radical cystectomy for infiltrating bladder cancer, and to use these to help in constructing a meaningful prognostic index. The disease-free survival was initially evaluated in 1026 patients (the reference series, 1969-1990). A multivariate analysis showed that the tumour P stage, grade and nodal involvement were the only factors which had an independent and significant association with survival. The computed regression coefficients were then used to classify patients into one of four risk categories and the results then validated by applying the model to a prospective test series (1991-1995). The 5-year disease-free survival of both groups was similar. When the results for the risk categories of the reference series were compared with those of the test series, there was no significant difference. This comprehensive prognostic model for the results of radical cystectomy was validated and verified in a prospective group of patients. Adjuvant therapies are indicated for patients with a high risk score.