Lenvatinib plus pembrolizumab and chemotherapy versus chemotherapy in advanced, metastatic gastroesophageal adenocarcinoma: The phase 3, randomized LEAP-015 study.

4001 Background: LEAP-015 (NCT04662710), is a randomized, open-label, phase 3 study of pembrolizumab plus lenvatinib and chemotherapy as first-line treatment for advanced/ metastatic gastroesophageal adenocarcinoma. We report results from the interim and final analyses of LEAP-015. Methods: Eligible participants (pts) had untreated HER-2 negative locally advanced unresectable or metastatic gastroesophageal adenocarcinoma, measurable disease and ECOG PS 0-1. All pts were randomly assigned 1:1 to induction with pembrolizumab 400 mg IV Q6W (x2) plus oral lenvatinib 8 mg QD and investigators choice chemotherapy (CAPOX Q3W x4 or mFOLFOX6 Q2W x6) then consolidation with pembrolizumab 400 mg Q6W for ≤16 doses plus lenvatinib 20 mg QD (only if 8 mg tolerated for at least 3 weeks), or chemotherapy alone (CAPOX or FOLFOX). Randomization was stratified by region, ECOG PS, and chemotherapy choice. Dual primary endpoints were PFS (RECIST v1.1, BICR) and OS in pts with PD-L1 combined positive score (CPS) ≥1 and in all pts; secondary endpoints included ORR and DOR (RECIST v1.1, BICR) in pts with PD-L1 CPS ≥1 and in all pts, and safety and tolerability in all pts. The data cut-off date was Oct 29, 2024. Results: A total of 880 pts (78% PD-L1 CPS ≥1; 75% gastric primary) were randomized (443 pembrolizumab plus lenvatinib and chemotherapy; 437 chemotherapy alone).Median follow-up was 32.2 mo (range 19.0 – 41.7) in pts with PD-L1 CPS ≥1 and 31.8 mo (range, 19.0 – 41.7) in all pts. At interim analysis, PFS difference was statistically significant with pembrolizumab plus lenvatinib and chemotherapy vs chemotherapy in pts with PD-L1 CPS ≥1 (median 7.3 vs 6.9 mo; HR 0.75; 95% CI, 0.62-0.9; P = 0.0012), with 24-mo PFS of 20% vs 7%, and in all pts (median 7.2 vs 7.0 mo; HR 0.78; 95% CI, 0.66-0.92; P = 0.0019), with 24-mo PFS of 21% vs 8%. ORR was 59.5% vs 45.4% in pts with PD-L1 CPS ≥1 and 58.0% vs 43.9% in all pts; P < 0.0001 for both. At final analysis, OS in pts with PD-L1 CPS ≥1 was not statistically significant (median 12.6 vs 12.9 mo; HR 0.84; 95% CI, 0.71-1.00; P = 0.0244 (P-value boundary for significance of 0.0204), with 24-mo OS of 31% vs 23%. OS in all pts was not tested per multiplicity strategy (median 13.1 vs 13.0; HR 0.87; 95% CI 0.75-1.01). Drug-related adverse event (AE) rates were 98% vs 92% in pts receiving pembrolizumab plus lenvatinib and chemotherapy vs chemotherapy. Grade ≥3 drug-related AE rates were 65% vs 49% (grade 5 AEs 5% vs < 1%). Conclusions: Pembrolizumab plus lenvatinib and chemotherapy vs chemotherapy provided statistically significant improvement in PFS and ORR in pts with advanced unresectable or metastatic gastroesophageal carcinoma at interim analysis. However, there was no significant improvement in OS in pts with PD-L1 CPS ≥1 at final analysis. Safety profiles were consistent with known regimens, with higher AE rates seen in pts receiving the experimental treatment. Clinical trial information: NCT04662710 .