GSOR20 Presentation Time: 12:05 PM: Vaginal Cuff Brachytherapy versus External Beam in High Risk (T1b Grade 3) Endometrioid Endometrial Cancer: An NCDB Analysis

Abstract

<h3>Purpose</h3> The optimal adjuvant management of localized, resected high risk (HR) Stage I endometrioid endometrial cancer (i.e. FIGO grade 3 with > 50% myometrial invasion) remains unclear. A Cochrane meta-analysis showed a 10% overall survival (OS) benefit when EBRT (external beam radiation therapy) was given versus no radiation treatment. However, no randomized trials have compared adjuvant EBRT to VCB (vaginal cuff brachytherapy) without chemotherapy. A SEER analysis from 2012 showed reported that in women with lymph node dissection, there was no difference in OS between EBRT and VCB in patients with HR Stage I endometrial cancer. We hypothesized that adjuvant EBRT or VCB would both be equally associated with improved OS in HR Stage I endometrial cancer in a large, modern database. <h3>Materials and Methods</h3> The National Cancer Database (NCDB) was queried for women diagnosed with endometrioid endometrial cancer from 2004-2017, who had undergone total hysterectomy or greater, and lymph node dissection revealing a pathologic stage of IB (AJCC Edition 6) or IC (AJCC Edition 7), and FIGO Grade of 3. Women were grouped by receipt of multiagent chemotherapy (MCT), VCB (without EBRT), EBRT (without VCB), or EBRT+VCB. OS was measured by Kaplan-Meier estimator, and compared using log rank test. Cox proportional hazard modeling was done, using as additional covariates tumor size, lymphovascular space invasion (LVSI), margin status, age, extent of comorbidities, treatment year, and several demographic factors (race, insurance status, educational and income level of zip code, urban versus rural location). A propensity score weighted analysis was performed, using the above covariates. <h3>Results</h3> 10,357 women were analyzed. As adjuvant treatment, 22% received EBRT alone, 26% received VCB alone, 11% received EBRT+VCB, and 42% received no radiation. 30% received adjuvant chemotherapy. Median OS for the entire group was 11.0 years. For the whole group, median OS times were 12.8 years with VCB alone, 11.3 years for EBRT alone, and 11.5 years with EBRT+brachytherapy, which were not significantly different (p > 0.05). However, median OS was only 8.6 years in patients not receiving any radiation, (p <0.0001 compared to receipt of each of the radiation modalities). For patients receiving no adjuvant chemotherapy, median OS was 12.3 years for EBRT alone, 12.8 years with VCB alone, 11.9 years with EBRT+brachytherapy (p > 0.05), but was only 9.6 years in patients not receiving any radiation (p < 0.00001). On multivariate analysis, receipt of radiation (but not receipt of multiagent chemotherapy), smaller tumor size, absence of LVSI, less comorbidities, younger age, and white race were associated with improved overall survival (p <0.003). Propensity score weighted analysis confirmed the OS benefit for women who received any RT. <h3>Conclusions</h3> In this large database, VCB and EBRT had comparable effectiveness in improving OS for HR Stage I endometrial cancer. With equal effect, VCB is favored given the more convenient schedule and decreased side effects as shown in previous randomized trials. The results of this study give impetus for a randomized trial comparing EBRT and VCB without chemotherapy in women with HR endometrioid endometrial cancer. In addition, a large percentage of women with HR disease who did not receive any radiation, despite this being the standard of care with known survival advantage, which reveals likely deficiencies in health care delivery.