Prognosis-predicting system based on factors related to survival of cervical carcinoma

Objective. To establish a different category system based on grouping of th e risk factors in patients with cervical smear. Methods: Univariate and mul tivariate analyses of factors associated with survival time were performed in 443 patients with stage Ib-IIa squamous cell carcinoma (SCC) from 1985 t o 1989. Results: Univariate analysis showed that parametrial extension, bul ky tumor size (greater than or equal to 4 cm), uterine body involvement, po or differentiation and pelvic lymph node (LN) metastasis were prognosticall y significant. However, by multiple regression method, only LN metastasis, deep stromal invasion (DSI), and poor differentiation (PD) were significant ly related to the patient's survival time with risk ratios of 2.78, 2.61 an d 1.52, respectively. A prognosis-predicting system was established on the basis of these simplified factors: groups of high, intermediate, law and ve ry low-risks, obtained survival rates of 93-96%, 77%, 51% and 25-37%, respe ctively. Conclusions: Using this simple model, cervical carcinomas can be c lassified prognostically for predicting 5-year-survival rates, and for risk -guided therapy in the future. (C) 1998 International Federation of Gynecol ogy and Obstetrics.