39MO Phase III ENGOT-En9/LEAP-001 study: Lenvatinib + pembrolizumab (LEN/PEMBRO) vs chemotherapy (chemo) as first-line (1L) therapy for advanced or recurrent endometrial cancer

LEN/PEMBRO following prior systemic therapy in any setting, including neo/adjuvant, is a standard of care for advanced endometrial cancer (EC). The phase 3 ENGOT-en9/LEAP-001 trial (NCT03884101) compared 1L LEN/PEMBRO vs chemo in patients (pts) with advanced/recurrent EC. Eligible pts had stage III–IV or recurrent, measurable/non-measurable, radiographically apparent EC, with no prior chemo or PD ≥6 mo after neo/adjuvant platinum-based chemo. Pts were randomized 1:1 (stratified by proficient vs deficient mismatch repair status [pMMR/dMMR] and, in the pMMR stratum, by ECOG PS [0/1], measurable disease [yes/no], and chemo/chemoradiation [yes/no]) to lenvatinib 20 mg QD plus pembrolizumab 200 mg Q3W or paclitaxel 175 mg/m2 Q3W plus carboplatin AUC 6 Q3W. Dual primary endpoints were PFS (RECIST v1.1, blinded independent central review) and OS in the pMMR population and among all-comers. Secondary endpoints included ORR and safety; duration of response (DOR) was an exploratory endpoint; and efficacy outcomes assessed by tumor histology was a prespecified exploratory analysis. 842 pts were randomized. At final analysis (data cutoff, 2 Oct 2023), after median follow-up of 38.4 (range, 30.3–52.9) mo, statistical significance for noninferiority (NI) OS endpoint was not achieved for LEN/PEMBRO vs chemo in the pMMR population (HR, 1.02 [95% CI, 0.83–1.26]; NI P = 0.2459875; Table). PFS and OS results for LEN/PEMBRO vs chemo by histological subtype will be presented for the pMMR population and all-comers. Treatment-related AEs occurred in 411/420 (97.9%) vs 398/411 (96.8%) treated pts in the LEN/PEMBRO vs chemo groups.Table: 39MOpMMRAll-comersLEN/PEMBRO n = 320Chemo n = 322LEN/PEMBRO n = 420Chemo n = 422OS HR (95% CI)1.02 (0.83–1.26)a0.93 (0.77–1.12)PFS HR (95% CI)0.99 (0.82–1.21)0.91 (0.76–1.09)ORR (95% CI), %50.6 (45.0–56.2)54.7 (49.0–60.2)55.7 (50.8–60.5)55.5 (50.6–60.3)Median DOR (range), mo16.1 (1.0+ to 48.7+)10.6 (1.1+ to 43.8+)23.2 (1.0+ to 49.0+)10.9 (1.1+ to 46.9+)a1-sided NI P = 0.2459875 (nonsignificant), not crossing prespecified OS NI boundary, P = 0.0158890, so no further statistical testing of efficacy endpoints was performed per prespecified multiplicity strategy for type 1 error control. Open table in a new tab a1-sided NI P = 0.2459875 (nonsignificant), not crossing prespecified OS NI boundary, P = 0.0158890, so no further statistical testing of efficacy endpoints was performed per prespecified multiplicity strategy for type 1 error control. The prespecified statistical criteria for OS and PFS in pts with pMMR IL advanced/recurrent EC were not met. Subgroup analyses identifying pts who may benefit most from LEN/PEMBRO will be presented. The safety profile was manageable and consistent with prior experience.