SURGICALLY DEFINED PROGNOSTIC PARAMETERS IN PATIENTS WITH EARLY CERVICAL-CARCINOMA - A MULTIVARIATE SURVIVAL TREE ANALYSIS

BACKGROUND. This study was performed to identify a statistical combina tion of independent pathologic and clinical features that best predict 5-year disease free survival (DFS) in patients with early stage cervi cal carcinoma treated by radical hysterectomy. The main goal of the st udy was to identify subsets of patients based on risk factors with max imal differences in DFS. METHODS. Three hundred and seventy patients w ere found for whom complete clinical and pathologic material, includin g cone and cervical biopsies, were available for analysis. Variables s tudied included age, weight, race, marital status, economic status, tu mor size (TS), depth of invasion (DI), lymph-vascular space involvemen t (LVSI), cell type, tumor grade, lymph node metastasis (LNM), and num ber of lymph nodes removed. Patients with LNM, parametrial involvement , and positive or close surgical margins were offered postoperative ra diation. After excluding patients with microinvasive and small cell ca rcinoma, data from the remaining 301 patients were submitted to univar iate and multivariate analyses to define those variables that best pre dict DFS. RESULTS. Univariate analysis showed that, ranked by degree o f significance, DI, TS, LVSI, LNM, tumor volume (TV) and clinical stag e were significant in predicting survival. Significant (P < 0.05) sing le parameters and other variables considered important were chosen for multivariate analysis, including the creation of a survival tree. Wit h this method, DI (less than or equal to 6 mm and >2 cm), LVSI, age (g reater than or equal to 40 yrs), and LNM were found to be the best com bination of risk factors to define prognosis. CONCLUSIONS. The multiva riate survival tree analysis maximally separates patients with early s tage invasive carcinoma of the cervix into 3 subgroups with 5-year DFS of 91%, 68%, and 43%, respectively. The authors excluded patients wit h microinvasive carcinoma (SGO, Society of Gynecologic Oncologists), w ho have an excellent DFS of 100%, and patients with small carcinoma, w ho have a poor DFS of 36.4% based on cell type alone, to define indepe ndent risk factors that maximally separate the remaining patients by D SF. The survival tree prognostic scoring system is easy to apply, and only requires DI (mm), LVSI (+, -), LNM, and age to assign an individu al patient to one of three risk groups. (C) 1996 American Cancer Socie ty.