Background
Phase III double-blind RCT (KEYNOTE-006), N=834 untreated or ≤1 prior line unresectable stage III/IV melanoma. Stratified by ECOG PS, prior therapies, PD-L1 (CPS). 1:1:1 randomization; pembrolizumab arms pooled in final analyses. 87 sites, enrolled 2013–2014. 10-yr follow-up via KEYNOTE-587 extension.
Interventions and follow up
Arm A: Pembrolizumab 10 mg/kg q2wk × up to 2 yr (n=279)
Arm B: Pembrolizumab 10 mg/kg q3wk × up to 2 yr (n=277)
Arm C: Ipilimumab 3 mg/kg q3wk × 4 (n=278)
Primary endpoint: OS and PFS
mFollow up: 10 yr (KEYNOTE-587)
Arm B: Pembrolizumab 10 mg/kg q3wk × up to 2 yr (n=277)
Arm C: Ipilimumab 3 mg/kg q3wk × 4 (n=278)
Primary endpoint: OS and PFS
mFollow up: 10 yr (KEYNOTE-587)
Results
mOS (pembro pooled vs ipi): 32.7 vs 15.9 mo, HR 0.71 (95% CI 0.60–0.85)
10-yr OS: 34.0% (pembro) vs 23.6% (ipi)
mPFS (pembro pooled vs ipi): 9.4 vs 3.8 mo, HR 0.64 (95% CI 0.54–0.75)
8-yr OS among ≥94-wk completers: 80.8% (median OS not reached)
10-yr OS: 34.0% (pembro) vs 23.6% (ipi)
mPFS (pembro pooled vs ipi): 9.4 vs 3.8 mo, HR 0.64 (95% CI 0.54–0.75)
8-yr OS among ≥94-wk completers: 80.8% (median OS not reached)
Adverse events
Grade ≥3 AEs: pembrolizumab ~17% vs ipilimumab ~20%
Immune-mediated AEs any grade: pembrolizumab ~25% vs ipilimumab ~18% (more frequent but less severe)
Grade ≥3 colitis: 1.4% (pembro) vs 7.5% (ipi)
Grade ≥3 hypophysitis: 0.4% (pembro) vs 1.8% (ipi)
Discontinuation due to AE: 7.4% (pembro) vs 9.4% (ipi)
Immune-mediated AEs any grade: pembrolizumab ~25% vs ipilimumab ~18% (more frequent but less severe)
Grade ≥3 colitis: 1.4% (pembro) vs 7.5% (ipi)
Grade ≥3 hypophysitis: 0.4% (pembro) vs 1.8% (ipi)
Discontinuation due to AE: 7.4% (pembro) vs 9.4% (ipi)
Conclusions
Pembrolizumab significantly improved OS (HR 0.71) and PFS (HR 0.64) vs ipilimumab in untreated advanced melanoma. 10-yr data confirm durable benefit: 34% alive at 10 yr; among ≥2-yr completers, 8-yr OS 80.8%. With CheckMate 067, established PD-1 inhibition as foundational in advanced melanoma.
Key Limitations
Comparator (ipilimumab 3 mg/kg monotherapy) is no longer a standard 1L option since CheckMate 067, making benefit vs current standard (ipi+nivo) indirect. Pembrolizumab and ipi+nivo never compared in a randomized phase III. Pembrolizumab dose (10 mg/kg) higher than approved flat dose (200 mg q3wk); approved on PK equivalence. PD-L1 was stratification only, not a validated required biomarker (benefit regardless of PD-L1).
Clinical Context
Established pembrolizumab superior to ipilimumab in 1L advanced melanoma (FDA 2015) — first PD-1 inhibitor approval in this indication. 10-yr data confirm a survival-curve tail: ~34% alive at 10 yr. Current 1L standards: nivolumab+ipilimumab ± relatlimab (high-burden) or pembrolizumab/nivolumab monotherapy (lower-burden). For BRAF V600+, DREAMseq supports IO-first sequencing. ESMO-MCBS: 5 (highest).