PROGNOSTIC FACTORS OF EARLY-STAGE CERVICAL-CANCER TREATED BY RADICAL HYSTERECTOMY

Background. This study was performed to identify pathologic and clinical features that best correlate with lymph node metastasis and disease free survival among patients with Stage I and II cervical cancer treated by radical hysterectomy. Methods. Three hundred-seventy patients with complete clinical information and pathologic material, including cone and cervical biopsies, were selected for analysis. Of these patients, 301 with clinical stages I and II disease were the subject of this paper, The results of patients with microinvasive carcinoma of the cervix, as defined by the Society of Gynecologic Oncologists (depth of invasion less than or equal to 3 mm and no lymph node vascular space invasion), were reported previously(1) and excluded from this analysis. Patients with small cell carcinoma of the cervix were found to have a very poor prognosis (disease free 5-year survival of 36%) and were also excluded from this analysis (Sevin BU, Nadji M, Metkoch MW, Lu Y, Averette HE. Unpublished data, 1995). Variables studied were patient age, weight, race, marital status, and economic status; tumor size; depth of invasion; lymph node-vascular space involvement; cell type; tumor grade; lymph node metastasis; and number of lymph nodes removed. The influence of these variables on survival was examined by univariate analysis with use of Cox's regression model and the log rank test for comparison of survival curves. Results. Factors that predict disease free survival, ranked by degree of significance, were depth of invasion, tumor size, lymph node-vascular space invasion, number of positive nodes, tumor volume, clinical stage, and tumor extension to the vagina or surgical margins. Conclusions. Radical hysterectomy and bilateral lymphadenectomy is standard therapy for patients with Stage IB and IIA carcinoma of the cervix. A variety of surgically defined risk factors predict ti-year disease free survival, and many of these factors are related. Identification of independent risk factors requires a multivariate analysis of data.