Background
Phase III RCT (DREAMseq; EA6134; ECOG-ACRIN), N=265 untreated BRAF V600-mutant unresectable/metastatic melanoma. Determined optimal 1L sequence: IO first (ipi+nivo → BRAF/MEK on progression) vs targeted first (dab+tram → ipi+nivo on progression). Stopped early at planned interim for efficacy (IO-first superiority). Enrolled 2017–2019, US cooperative groups.
Interventions and follow up
Arm A (IO first): Nivolumab 3 mg/kg + ipilimumab 1 mg/kg q3wk × 4 → nivolumab 3 mg/kg q2wk maintenance; on progression → dabrafenib 150 mg BID + trametinib 2 mg daily
Arm B (targeted first): Dabrafenib 150 mg BID + trametinib 2 mg daily; on progression → nivolumab 3 mg/kg + ipilimumab 1 mg/kg q3wk × 4 → nivolumab maintenance
Primary endpoint: 2-yr OS
mFollow up: 27.7 mo
Arm B (targeted first): Dabrafenib 150 mg BID + trametinib 2 mg daily; on progression → nivolumab 3 mg/kg + ipilimumab 1 mg/kg q3wk × 4 → nivolumab maintenance
Primary endpoint: 2-yr OS
mFollow up: 27.7 mo
Results
2-yr OS: 71.8% (Arm A, IO first) vs 51.5% (Arm B, targeted first), P=.010
OS HR: 0.57 (95% CI 0.37–0.87)
PFS on first-line therapy (Arm A): 11.2 mo (ipi+nivo) vs 8.5 mo (dab+tram)
Subsequent therapy receipt: 55% of Arm B received 2L ipi+nivo; 59% of Arm A received 2L dab+tram
OS HR: 0.57 (95% CI 0.37–0.87)
PFS on first-line therapy (Arm A): 11.2 mo (ipi+nivo) vs 8.5 mo (dab+tram)
Subsequent therapy receipt: 55% of Arm B received 2L ipi+nivo; 59% of Arm A received 2L dab+tram
Adverse events
Grade ≥3 AEs: 59% (Arm A) vs 54% (Arm B)
Grade ≥3 immune-mediated AEs (Arm A): colitis 9%, hepatitis 7%, endocrinopathy 5%
Grade ≥3 BRAF/MEK AEs (Arm B): pyrexia 5%, elevated liver enzymes 4%, rash 3%
AE-related discontinuation of first-line: 15% (Arm A) vs 9% (Arm B); no unexpected safety signals
Grade ≥3 immune-mediated AEs (Arm A): colitis 9%, hepatitis 7%, endocrinopathy 5%
Grade ≥3 BRAF/MEK AEs (Arm B): pyrexia 5%, elevated liver enzymes 4%, rash 3%
AE-related discontinuation of first-line: 15% (Arm A) vs 9% (Arm B); no unexpected safety signals
Conclusions
IO-first sequencing (ipi+nivo then BRAF/MEK on progression) was significantly superior to targeted-first in BRAF V600-mutant advanced melanoma, with 2-yr OS 71.8% vs 51.5% (P=.010). Driven by durable IO responses and the difficulty of re-establishing IO efficacy after prior BRAF/MEK-induced immunosuppression.
Key Limitations
Stopped early (N=265 vs planned 500) after interim showed IO-first superiority — smaller sample may amplify effect size and limit subgroup precision. Ipi+nivo dosing (nivo 3mg/kg + ipi 1mg/kg) is the approved melanoma combination; outcomes may not apply to 3mg+3mg dosing. Only 55% of targeted-first patients received 2L ipi+nivo, partly explaining the OS gap. Does not address pembrolizumab monotherapy sequencing. Patients with urgent need for rapid response (high-volume visceral disease, very high LDH) may still benefit from BRAF/MEK first.
Clinical Context
DREAMseq resolved a major clinical question: IO-first sequencing is preferred for BRAF V600-mutant metastatic melanoma when rapid cytoreduction is not urgently required. With SECOMBIT (phase II) favoring IO-first on total OS, this is now reflected in ESMO guidance. For most BRAF V600+ patients, ipi+nivo (or pembrolizumab monotherapy for lower-burden disease) is recommended first-line, reserving BRAF/MEK for progression or rapid symptomatic control. DREAMseq did not include newer regimens (relatlimab+nivolumab, BRAF/MEK + IO triplets).