Impact of tumor size on survival in cancer of the cervix and validation of stage IIA1 and IIA2 subdivisions

Objective A change has recently been made to the Federation of Gynecology and Obstetrics (FIGO) staging system for cervical cancer to account for size within stage IIA cancers. This study was designed to investigate the impact of size within stage I-IIIB cervical carcinoma, and to validate these changes. Methods The Surveillance, Epidemiology, and End Results Program database was used to extract data on patients from 1988 to 2008. Patients were included who had information recorded regarding stage, size, and type of treatment received. They were then stratified by stage and size, and analyzed for cause-specific survival (CSS) using Kaplan Meier estimates, as well as hazard ratios using Cox proportional hazards regression modeling. Results A total of 18,649 cases were evaluated. All stages evaluated demonstrated improved CSS on Kaplan Meier estimates for smaller tumor sizes (largest p = 0.0003). Hazard ratios were significantly worse for larger tumor sizes on both univariate and multivariate modeling. Specifically, stage IIA cancers demonstrated a hazard ratio of 2.0 on univariate, and 1.69 on multivariate analysis (C.I. 1.46-2.75, p < 0.0001 and C.I. 1.20-2.38, p = 0.0025, respectively). Further size subdivisions of 2 and 4 cm for stage I, 4 cm for stage IIB, and 4 and 6 cm for stage IIIB also maintained prognostic significance. On multivariate analysis within each stage, size was the only variable to maintain independent significance in all stages evaluated. Conclusions Size is independently prognostic within each stage in cervical cancer, validating the recent changes to the FIGO staging system.