Prediction of the survival of patients with advanced ovarian cancer according to a risk model based on a scoring system

Purpose of investigation: The aim of the study was to devise a scoring syst em based on significantly prognostic parameters which might predict the sur vival time for each individual patient. Methods: The study group included 30 patients with stage III or IV invasive ovarian cancer, operated on between 1987 and 1994. Different preoperative and intraoperative parameters were examined and their impact on survival wa s assessed using the Kaplan-Meier method. Survival rates were compared by t he log rank test. The: parameters, which have been found to be significant prognostic factors, were incorporated into the final risk model, based on a scoring system. The scores were given according to the cutoff point of eac h prognostic parameter. A correlation coefficient was calculated from the s urvival curve, representing the combined scores of individual patients vers us their survival. Results. From ail studied parameters, only gravidity, the stage of ovarian cancer and the preoperative levels of LDH and CA-125 were demonstrated to h ave a significant impact on survival. On the basis of these four parameters a scoring system was devised. The scores were given considering the cut-of f level of each parameter: gravidity less than or equal to 2 versus > 2; st age of the cancer - IIIA, IIIB and stage IIIC versus stage IV; LDH level le ss than or equal to 240 versus > 240 Un; CA-125 less than or equal to 100 v ersus > 100 U/ml. The average of the combined scores representing the four parameters was the final risk factor, which was plotted against the surviva l of each individual patient. A correlation coefficient of -0.99 was found. Conclusions: Our data suggest that the survival of each individual patient with advanced ovarian cancer could be predicted quite accurately considerin g our risk model based on a scoring system which incorporates four preopera tive and intraoperative parameters. This scoring system needs to be validat ed by larger prospective trials.