Abstract
Background: The incidence of complex atypical hyperplasia (CAH) and early stage endometrioid endometrial cancer (EEC) is increasing. While standard upfront treatment for CAH and EEC is hysterectomy, non-surgical options are needed for the increasing population unable to undergo hysterectomy due to medical comorbidities precluding surgery or desired future fertility. Our recently published phase II trial of levonorgestrel intrauterine device (LIUD) showed substantial response for patients with CAH or grade 1 (g1) EEC, with 83% overall response rate at 12 months of LIUD treatment. Yet, relapse rates have not been well-defined in a large prospective clinical trial population. Further, we propose that estrogen-regulated gene expression biomarkers in the endometrium could reflect differences in estrogen signaling that drive later outcomes. Methods: This study was conducted under an approved IRB protocol at MD Anderson Cancer Center. Follow-up data for complete responders (those without evidence of hyperplasia or cancer following 12 months of LIUD) were abstracted from the medical record to determine long-term outcomes and relapse status (presence of CAH or grade 1 EEC). Estrogen-regulated gene expression biomarkers (PR, EIG121, IGF1, IGF2, RALDH2, survivin) were measured by real-time PCR from baseline endometrial biopsies and compared by relapse status. Results: Follow-up data were available for 30 of 39 participants that experienced complete response with 12 months of treatment with LIUD (median follow-up was 40 months). Eleven of 30 patients (36.7%) with complete response at the end of the 12-month LIUD treatment went on to experience relapse (6/23 with initial diagnosis of CAH and 5/7 g1 EEC). Median time to relapse after complete response was 14.3 months. Relapse occurred in patients with LIUD still in place (n=6) and LIUD removed (n=5). Of 11 cases with relapse, 5 subsequently responded to LIUD (including 3 cases with LIUD previously removed) and 6 went on to have hysterectomy. Two out of 30 (6.7%) complete responders showed progression of their initial g1 EEC lesions following the study: one patient progressed to grade 2 EEC and one patient progressed to grade 2 mixed EEC/clear cell carcinoma. Ten of 30 later had hysterectomy (equally from CAH and g1 EEC cohorts). Finally, although 13/30 patients indicated desired future fertility at baseline, only 3 attempted to conceive, and 1 achieved pregnancy and live birth. When comparing estrogen-regulated gene expression, participants that exhibited relapse had lower pre-treatment IGF1 than those that did not relapse (p=0.028). However, other biomarkers were not significantly different by relapse status. Conclusions: These findings show that although many patients with CAH or grade 1 EEC are initially responsive to LIUD, the response may not be durable. Additional studies are needed to evaluate biomarkers to predict risk of relapse and identify candidate targets for improved therapeutic strategies.
Citation Format: Mikayla Waters, Navdeep Pal, Misty Woodall, Jessica Gallegos, Shannon Westin, Melinda Yates. Long-term follow-up for a prospective phase II trial of levonorgestrel intrauterine device as non-surgical treatment for complex atypical hyperplasia and early endometrial cancer [abstract]. In: Proceedings of the AACR Virtual Special Conference: Endometrial Cancer: New Biology Driving Research and Treatment; 2020 Nov 9-10. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(3_Suppl):Abstract nr PO034.