Determinants of Lymph Node Count in Endometrial Cancer Surgical Staging

Abstract

Lymphadenectomy is a fundamental procedure in gynecologic oncology, but there is an ongoing debate concerning its indication in endometrial cancer. Lymph node (LN) count has been used as a surrogate marker for quality of staging in endometrial cancer. Because of variability in reported LN counts in the literature and within our practice, we aimed to better understand the factors that influence the final LN count in endometrial cancer staging. We conducted a retrospective case study of patients with endometrial cancer who underwent surgical staging at our institution between April 1, 2005, and February 3, 2007. Linear regression was used to determine the association between LN count and a series of predictor variables. Of 131 patients, 100 patients (76%) had stage I disease and 9 patients (7%) had LN metastasis. The mean (SD) LN count was 9.5 (7.8). We found no significant difference in LN count according to age, tumor histology, stage, or surgeon. Lymph node count decreased by 1 for each 5-unit (kg/m(2)) increase in body mass index (coefficient, -0.2; P = 0.038). The strongest predictor associated with LN count was the pathologist, with 2 groups of pathologists counting an average 7.7 (P < 0.001) and 6.42 (P = 0.001) fewer LNs per case compared to the referent group. Our study confirms that LN count varies markedly. Although not the only contributor, the pathologist, we found, was the most significant determining factor in LN count variation. This highlights the need to exercise caution when drawing conclusions from published LN counts in endometrial cancer research.