Purpose: This study aimed to assess the optimal tumor diameter for predicting lymphatic metastasis and to determine intraoperatively the need for lymph node dissection in patients with endometrioid endometrial cancer.

Methods: Military beneficiaries diagnosed with stage I–III endometrioid endometrial cancer during 2003–2016 who had at least 7 pelvic and/or paraaortic lymph nodes removed during the time of hysterectomy were studied. Tumor diameter was compared against the presence of positive nodes, using the prior models of 20 mm (ie, Mayo model) and 50 mm (ie, Milwaukee model), to determine the false-negative rate of each threshold. A separate analysis was completed to determine the optimal diameter for our population. Receiver operating characteristic curve analysis models of tumor diameter were evaluated for model fit and predictive power of lymph node involvement.

Results: Of the 1224 patients with endometrioid endometrial cancer included, 13% (n = 160) had positive lymph node involvement. Tumor sizes ranged from 1 mm to 100 mm. In contrast to Mayo and Milwaukee models (ie, Mayo, Milwaukee), the optimal tumor diameter independent of myometrial invasion and grade of tumor to predict lymph node metastasis was found to be 35 mm.

Conclusions: Endometrioid endometrial cancer tumor diameter of 35 mm was found to be the optimal threshold for lymphadenectomy when the operating surgeon has no knowledge of tumor invasion.