BACKGROUND : Endometrial cancer (EC) is the most common gynecologic cancer in the United States. The objective of this cohort study was characterize the clinical and pathologic features that are associated with endometrial cancer-specific death for women cared for at a single NCI-designated comprehensive cancer center.

PATIENTS AND METHODS : This is a retrospective cohort from 2014-2017 including all women who had a hysterectomy for EC. Charts were reviewed for clinical and pathologic data, focusing on survival outcomes.

RESULTS : 771 EC patients underwent hysterectomy with 760 informative for outcomes. Seventy-six (10%) deaths were related to their EC; 62 women died from recurrent EC. Non-endometrioid histology and advanced stage were predictors of recurrence and EC death. Among patients with endometrioid ECs, mismatch repair status was significantly associated with EC specific survival (RR=4.8, 95% CI 2.3-10.3; p<0.0001). Most EC patients who recurred died of their disease 62/83 (74.7%). Nearly half of the patients that recurred (27/62) had no additional therapy at the time of recurrence. Overall survival was significantly longer for those women who had additional treatment at the time of recurrence; however the improvement in overall survival with therapy at recurrence was largely attributable to effects in those women who were adjuvant therapy naïve.

CONCLUSIONS : While there is benefit of treatment at the time of recurrence for treatment-naïve women; only about half of patients were able to receive therapy. There is an urgent need for continued efforts for more effective EC therapy in both the front-line and recurrent setting as well as early identification of cancer diagnosis and recurrence.

IMPLICATIONS FOR PRACTICE : Approximately 10% of patients die of their endometrial cancer. Most deaths were from recurrent disease however almost 20% of endometrial cancer deaths were within 120 days of surgery. While treatment at the time of recurrence improves overall survival only about half of patients will receive therapy at the time of recurrence. Traditional prognostic features like histology and stage remain important to predict risk of recurrence and newer biomarkers like mismatch repair status may improve risk stratification and targeted therapy. There remains an urgent need for improved therapy and early detection of diagnosis and recurrence.