Stage IIIC endometrial carcinoma (EC) represents pathologically heterogenous patients with single/multiple pelvic (stage IIIC1) or paraaortic (stage IIIC2) lymph nodes (LNs). There is an increasing trend to offer adjuvant chemotherapy (CT) +/− radiation (RT) uniformly to these patients, regardless of substage. We investigate the prognostic significance of positive LN (pLN) number, ratio (%pLN), location (IIC1 vs. IIC2), and adjuvant treatment on patterns of failure and survival in a large collaborative multi-institutional series.Materials and Methods:
Clinical data for stage III EC patients such as patient characteristics, surgery/pathologic details, adjuvant therapies (including CT, RT, and chemotherapy and radiation), and outcomes (including pelvic control [PC], disease-free survival [DFS], distant DFS, and overall survival [OS]) were collected from 3 academic institutions. Log-rank analyses, Cox regression univariate and multivariate analyses were performed.Results:
Of the 264 patients queried for stage III disease, 237 (73%) had pLN, and complete LN sampling for analysis. The mean number of pLN in the combined data were 3.9, with 26.1% of all LN sampled positive; 121 patients (51%) staged IIIC1, and 116 patients (49%) staged IIIC2. There was a significant difference in number of pLN (P=0.0006) and total LN sampled by institution (range, 13 to 35; P=0.0004), without a difference in %pLN (P=0.35). Ninety-seven of 220 (44.1%) have ≥20% pLN. While controlling for substage and institution, a decrease in DFS (hazard ratio [HR], 1.1; P=0.007), and OS (HR, 1.1; P=0.01) was observed with every increase of 10% in the pLN ratio. There was a significant difference in DFS (HR, 1.8; P=0.003), PC (HR, 1.9; P=0.004), and distant DFS (HR, 1.6; P=0.03), as well as a trend for decreased OS (HR, 1.6; P=0.08) for substage IIIC2 versus IIIC1 disease; 5 years DFS 40% versus 45%, OS 50% versus 57%. Patients received no adjuvant therapy (10%), CT alone (27%), RT alone (16%), or chemotherapy and radiation (47%). There was no significant difference in PC, DFS, or OS between the various treatment regimens. On univariate analysis, while pLN was significant, treatment type did not impact DFS or OS. On multivariate analysis for DFS, patient age, race, and IIIC1 versus IIIC2 substage retained significance (HR, 0.56; P=0.01).Conclusions:
Stage III EC patients with substage IIIC2 disease have a significantly increased risk of local and distant disease recurrence and death from EC. A decrease in DFS and OS was observed with every increase of 10% in the pLN ratio. Stage IIIC2 patients represent a high-risk subpopulation for whom clinical trials, or targeted regimens should be explored to achieve improved oncologic outcomes.