Validation of Nomograms for Survival and Metastases after Hysterectomy and Adjuvant Therapy in Uterine Cervical Cancer with Risk Factors

Abstract

Background. Three nomogram models for early stage uterine cervical cancer have been developed (KROG 13-03 for overall survival [OS], SNUH/AMC for disease-free survival [DFS], and KROG 12-08 for distant metastases-free survival [DMFS]) after radical hysterectomy (RH) and pelvic lymph node dissection (PLND). This study aimed to validate these models using our cohort with adjuvant radiotherapy. Methods. According to the eligibility criteria of nomogram studies, patients were enrolled in Group A (N=109) for the two KROG models (RH with PLND and whole pelvic irradiation) and Group B (N=101) for the SNUH/AMC model (RH with PLND and squamous histology). Using Cox-regression hazard models, the prognostic factors of our cohorts were evaluated. The risk probabilities induced from published nomogram scores were calculated and the concordance indices were evaluated. Results. Group A had 88.1% 5-year OS and 86.0% 5-year DMFS. Group B had 83.0% 5-year DFS. In multivariate analyses, large tumor size for OS (HR 8.62, P<0.001) and DMFS (HR 5.13, P=0.003), young age (≤40 versus 41-64 years) for OS (HR 4.63, P=0.097) and DFS (HR 3.44, P=0.051), and multiple lymph node metastases (0 versus ≥3) for DMFS (HR 4.03, P=0.031) and DFS (HR 3.90, P=0.038) were significantly correlated. The concordance indices for OS, DMFS, and DFS were 0.612 (P=0.002), 0.597 (P=0.014), and 0.587 (P=0.020), respectively. Conclusion. The developed nomogram models after RH and PLND are clinically useful in predicting various types of survival with significance.