Abstract

A number of studies have suggested that serum CA 125 levels may be an important prognostic factor for survival of patients with ovarian carcinoma. We investigated, in a large group of patients from 11 UK centers, which combination of CA 125 measurements provided the best prognostic index, and whether the predictive power could be improved by the addition of other factors. Analysis of the data from 248 patients showed that the absolute value of the third CA 125 sample was the single most important factor for predicting progression at 12 months, with the addition of residual bulk only slightly improving the predictive power. Seventy-four patients had CA 125> 70, and of these 57% were correctly predicted to progress or die within 12 months, but 43% remained alive and progression free. The best predictor for progression produced a false positive rate of 19%. We therefore conclude that prognostic information based upon CA 125 measurements up to the start of the third course of initial chemotherapy is not accurate enough to be used to manage individual patients.