Predictors of non-sentinel lymph node metastasis in patients with positive sentinel lymph node (SLN) in endometrial cancer

Abstract

histology (Grade 3, Type I vs. Type II), depth of invasion, and lymphovascular space invasion. Group B cases were older (65.5 ± 9.2 vs. 63.0 ± 10.9 years, P= 0.02) and had more pelvic LNs harvested (26.4 ± 10.5 vs. 18.8 ± 8.5, P b 0.0001). Aortic LN yields were identical (9.0 ± 5.6 vs. 9.0 ± 6.0). More pelvic LN metastases were detected in Group B (30.3% vs. 13.6%, P b 0.0001). Group B had more Stage IIIC1 cases (19.3% vs. 6.0%, P b 0.0001), fewer Stage IIIA cases (0.8% vs. 6.6%, P b 0.01), more GOG high-risk cases (32.8% vs. 19.8%, P b 0.01), fewer low-risk cases (21.8% vs. 36.3%, P b 0.01), and more patients receiving combination chemotherapy + radiation (28.6% vs. 16.3%, P b 0.003). A total of 18/36 (50%) Group B cases with LNmetastases were isolated to the SLN, and 12/18 (66.7%) of these were micrometastases or isolated tumor cells that were missed on hematoxylin-and-eosin pathology. These 12 upstaged cases would have been considered GOG low-risk (n= 4), low-intermediate risk (n= 2), and high-intermediate risk (n= 6) without SLN mapping. Conclusions: This study suggests that compared to traditional pelvic lymphadenectomy, the addition of SLN mapping identifies more pelvic lymph node metastases and shifts GOG lowand intermediate-risk cases to high-risk cases, possibly resulting in administration of more postoperative chemotherapy and radiation therapy.