Survival benefit of postoperative adjuvant chemotherapy for patients in poor-differentiated early-stage cervical cancer: A multicenter, cohort study.

e17509 Background: Postoperative adjuvant therapy decisions for early-stage cervical cancer (ESCC) without high risk factor, including parametrial extension, positive nodes and positive margins, now depend on Sedlis criteria, which do not mention the degree of differentiation. In this study, we aimed to explore the benefit of adjuvant chemotherapy for patients of ESCC with poor differentiation but without positive Sedlis criteria after radical hysterectomy, and develop a score to predict the recurrence risk in these patients, so as to further improve the indications of postoperative adjuvant therapy for patients of ESCC. Methods: Patients of ESCC with pathologically proven poor differentiation after radical hysterectomy were prospectively included in a multicenter study. Univariate Cox proportional hazards regression were developed to evaluate the independent risk factors for recurrence. The log-rank test was performed to compare the progression-free survival (PFS) and overall survival (OS) of adjuvant chemotherapy (ACT) and no further treatment (NFT) groups. Selected variables were analyzed in a Multivariate Cox Regression Model and used to build a nomogram for ESCC to predict the survival probabilities. Results: In total, 352 patients were randomly assigned to two groups, a training group and a validation group, at a ratio of 1:1 (176 patients: 176 patients). In both of training and validation cohort, the ACT groups had longer PFS than the NFT groups ( P= 0.007 and 0.016, respectively), while the differences of OS were not statistically significant ( P= 0.922 and 0.107, respectively). In the multivariate analysis, ACT was proved an independent prognostic factor for the PFS (P = 0.048), and the strongest predictor of PFS even analyzing with other potential independent prognostic factors (HR 0.116, 95%CI 0.014-0.981). Five factors, including postoperative treatment, histologic type, isthmus uteri involvement, FIGO stage, SCCA level, were included in the nomogram to predict survival probabilities in the training cohort and performed well in the validation cohorts. Conclusions: For the patients of ESCC who don't meet to Sedlis criteria but with poor differentiation, ACT after radical hysterectomy can reduce the recurrence rate.