Objective: The aim of this study was to determine whether knowledge of lymph node status improves survival prediction in clinically early-stage endometrial cancer.
Methods: The records of 661 patients with apparently uterine-confined disease were reviewed. The performance in predicting overall survival and cause-specific survival was compared between a multivariate prognostic model with nodal status and a model without nodal status by calculating Harrell concordance index.
Results: Among 661 patients with clinically early-stage endometrial cancer, the lymph node metastasis rate was 8.3% (55/661). Lymph node metastasis independently associated with cause-specific survival only when no stratification according to adjuvant treatment was applied (P = 0.035). After stratification according to adjuvant radiotherapy, lymph node status marginally associated with cause-specific survival (P = 0.073), whereas myometrial invasion retained its strong association with cause-specific survival (P < 0.001). However, there was no significant difference in the performance of the survival model using only uterine factors and the model using lymph node status and uterine factors (concordance index, 0.77 vs 0.77, respectively; P = 0.798).
Conclusions: Knowledge of lymph node status did not significantly improve the performance of survival prediction in apparently uterine-confined endometrial cancer, although it was independently associated with survival. In the patients with clinically early-stage endometrial cancer, the accuracy of the prediction of survival was comparable between risk grouping without lymph node status and that including lymph node status.