Patients with endometrial intraepithelial neoplasia do not need routine lymph node sampling

Abstract

<h3>Objectives:</h3> The use of sentinel lymph node (SLN) protocols in endometrial cancer has emerged as a standard of care for staging invasive endometrial cancers, while in endometrial intraepithelial neoplasia (EIN), their role is not well defined. We sought to identify preoperative characteristics that may predict the presence of concurrent endometrial cancer at the time of hysterectomy in order to identify who may benefit from nodal assessment at the time of hysterectomy. <h3>Methods:</h3> Using billing codes for EIN or complex endometrial hyperplasia, we identified all patients with a pre-operative diagnosis of EIN from 2010 through 2020. Those who subsequently underwent surgical management (i.e. minimally invasive or open hysterectomy±adnexal surgery and±lymph node dissection) were included in this analysis. Patients were excluded if they opted for medical management or if subsequent pathology review indicated an alternative preoperative diagnosis. Data was abstracted from medical records. Data are presented as n (%) and median (interquartile range); we used modified Poisson regression to calculate risk ratios (RR) and 95% confidence intervals (CI). <h3>Results:</h3> Of the 492 patients with a diagnosis of EIN, 378 patients underwent hysterectomy. Surgical pathology revealed 275 (73%) had EIN or no residual disease, and 103 (27%) had endometrial cancer. Age (p=0.003), race (p=0.02), and concurrent diagnosis of hypertension (p=0.02) were significantly associated with the presence of endometrial cancer on final pathology. Notably, the median preoperative endometrial thickness was significantly greater in the endometrial cancer group [14mm (10-19)] than in the EIN group [11mm (8-16); p=0.002]. When dichotomizing endometrial thickness, patients with a preoperative endometrial stripe ≥15mm were 1.77 times more likely to have endometrial cancer than those with an endometrial stripe <15mm (95% CI: 1.24-2.54). The corresponding RR for those with a preoperative endometrial stripe ≥20mm was 1.97 (95% CI: 1.34-2.90). A total of 5 patients (1.3%) were diagnosed with stage IB (>50% myometrial invasion) disease, 30 (8%) patients had tumors > 2cm, 1 (0.3%) had grade 3 histology, and 3 (0.8%) had lymphovascular space invasion (LVSI). In this cohort, only 10 (3%) of patients underwent lymph node evaluation at time of surgery based on intra-operative factors, including frozen section. <h3>Conclusions:</h3> In a large cohort of patients with a preoperative diagnosis of EIN, 27% were found to have invasive carcinoma, but very few patients had pathologic features portending increased risk of nodal metastasis. Even fewer met intraoperative criteria for lymph node evaluation. Increasing endometrial thickness may be a useful pre-operative marker to identify who is at higher risk for concurrent endometrial cancer and could be considered a criterion for use of a sentinel lymph node algorithm in patients with EIN. Prospective studies are warranted.