Predictive value of FIGO and AJCC staging systems in patients with uterine leiomyosarcoma

Abstract

Background Cancer staging systems aim to identify patient cohorts with different outcomes based on clinically relevant prognostic factors. Uterine leiomyosarcoma (ULMS) is classified using the FIGO staging system developed for epithelial malignancies; other sarcomas use the AJCC staging system. Neither has been validated in ULMS. We critically evaluated both systems to determine if either identified patient groups with distinct outcomes. Methods We staged 230 ULMS patients by the FIGO and AJCC systems. Progression-free survival (PFS) and overall survival (OS) rates were calculated; statistical pairwise comparisons were performed. Results The number of stages I and III patients varied by staging system. There were few stage II patients by either system. Using the FIGO system, PFS was better in stage I patients versus stage III or IV patients, and OS was better in stage I patients versus stage IV patients. Using the AJCC system, PFS and OS were better in stage I patients (low grade) versus stage II, III or IV patients, and OS was better in stage III patients versus stage IV patients. Prognosis of patients with serosal involvement (FIGO III) was similar to that of patients with metastases (FIGO IV). Conclusion Neither system classifies ULMS patients into four clinically meaningful, non-overlapping stages predictive of PFS and OS. This analysis highlights the relevance of certain factors (low grade, serosal involvement) and rarity of others (FIGO stage II, cervical invasion). Once identified, prognostic factors relevant to this malignancy should be incorporated into a new staging system in an effort to identify appropriate cohorts for different treatments.