Long-term survival rates and cure modeling with dostarlimab plus chemotherapy in mismatch repair deficient/microsatellite instability-high (dMMR/MSI-H) primary advanced or recurrent endometrial cancer in the ENGOT-EN6-NSGO/GOG-3031/RUBY trial.

Abstract

5501 Background: Dostarlimab + carboplatin-paclitaxel (CP) demonstrated significant progression-free survival (PFS) and clinically meaningful overall survival benefits in patients with dMMR/MSI-H primary advanced or recurrent endometrial cancer (EC) in Part 1 of the phase 3 RUBY trial (NCT03981796). Median PFS was not reached for dostarlimab + CP vs 7.7 months for CP alone in this population. Moreover, plateauing of the PFS Kaplan-Meier (KM) curve suggested that a subset of patients with dMMR/MSI-H EC achieved durable disease control and potential for curative intent with dostarlimab + CP. Mixture cure models (MCMs) complement conventional outcome metrics by quantifying the proportion of patients considered "cured" and by modeling PFS among patients who remain uncured. Here we present an MCM applied to the PFS data from the dMMR/MSI-H population with up to 4.5 years of follow-up treated with dostarlimab + CP. Methods: At a median follow-up of 55.6 months (range 49.9–67.7 months), descriptive PFS analyses were conducted in the dMMR/MSI-H population from RUBY Part 1. MCMs were fitted to PFS data to estimate the proportion of patients treated with dostarlimab + CP who had a curative potential. In the MCM, patients with curative potential were assumed to be free of recurrence- and disease-related mortality risks, implying that any additional PFS events would come from death. Mortality risks were assumed to be similar to the background mortality expected in the general population (with matched demographic characteristics). Conversely, the uncured population was considered to be at risk for disease recurrence and death from all causes. Results: There was a low rate of progression since PFS was first analyzed, with only 4 new events reported with an additional 2.5 years of follow-up. The 4-year PFS rate was 57.9% (95% CI, 42.3%–70.6%) in the dostarlimab + CP arm vs 15.7% (95% CI, 7.2%–27.0%) in the control arm. The MCM analysis demonstrated that the majority of patients with dMMR/MSI-H EC (54%; 95% CI, 35%–72%) were considered “cured” with dostarlimab + CP. The PFS survival curves estimated from MCMs closely followed the KM curves reported from the RUBY trial data. The model appropriately captured the long-term survival plateau, instilling confidence in the robustness of the model predictions. Conclusions: With a minimum of 4 years of follow-up, updated PFS analyses demonstrated that the majority of patients remained alive and progression-free, illustrating durable disease control. Model-based predictions indicated the potential for cure in the dMMR/MSI-H primary advanced or recurrent EC population receiving dostarlimab + CP. These analyses may assist oncologists in understanding long-term survival outcomes and the proportion of patients with a curative potential. Clinical trial information: NCT03981796 .