Background
CheckMate 204 was a phase II single-arm trial (n=119) evaluating nivolumab + ipilimumab combination immunotherapy in patients with active untreated melanoma brain metastases. Two cohorts: Cohort A (asymptomatic, no steroids, n=101) and Cohort B (symptomatic or leptomeningeal, n=18).
Interventions and follow up
Regimen: Nivolumab 1 mg/kg + ipilimumab 3 mg/kg IV q3w × 4 doses → nivolumab 3 mg/kg q2w (or 480 mg q4w) until progression or unacceptable toxicity
Primary endpoint: Intracranial clinical benefit rate (CR + PR + SD ≥6 months) per investigator; secondary: intracranial ORR, extracranial ORR, OS, safety
mFollow up: 34 months
Primary endpoint: Intracranial clinical benefit rate (CR + PR + SD ≥6 months) per investigator; secondary: intracranial ORR, extracranial ORR, OS, safety
mFollow up: 34 months
Results
Cohort A iORR: 54.2% (CR 26.7%, PR 27.5%)
Cohort A intracranial clinical benefit: 57.4%
Cohort A 3-yr OS: 51.9%
Cohort A 3-yr intracranial PFS: 54%
Cohort A extracranial ORR: 49.5%
Cohort B iORR: 16.7%
Cohort B 3-yr OS: 26.7%
Cohort A intracranial clinical benefit: 57.4%
Cohort A 3-yr OS: 51.9%
Cohort A 3-yr intracranial PFS: 54%
Cohort A extracranial ORR: 49.5%
Cohort B iORR: 16.7%
Cohort B 3-yr OS: 26.7%
Adverse events
Grade ≥3 AEs: 54.4% (Cohort A)
Immune-mediated Grade ≥3: 19.8% (hepatitis 9.9%, diarrhea/colitis 6.9%, CNS immune events 2.0%)
Discontinuation due to AEs: 28.7%
Corticosteroid use for immune-mediated AEs: 51%
Grade ≥3 neurologic AEs: 4% Cohort A vs 17% Cohort B
Immune-mediated Grade ≥3: 19.8% (hepatitis 9.9%, diarrhea/colitis 6.9%, CNS immune events 2.0%)
Discontinuation due to AEs: 28.7%
Corticosteroid use for immune-mediated AEs: 51%
Grade ≥3 neurologic AEs: 4% Cohort A vs 17% Cohort B
Conclusions
Nivolumab + ipilimumab achieved a 54% intracranial ORR and 52% 3-year OS in asymptomatic untreated melanoma brain metastases, establishing dual checkpoint blockade as a highly effective CNS-active systemic regimen.
Key Limitations
Single-arm Phase II with no randomized comparator; asymptomatic patient selection (Cohort A) represents favorable biology; symptomatic Cohort B had substantially inferior outcomes (iORR 17%, 3-yr OS 27%); corticosteroid dependence likely abrogates immunotherapy efficacy; concurrent BRAF/MEK inhibitors not studied.
Clinical Context
CheckMate 204, together with ABC and NIBIT-M2, established nivo+ipi as the most active systemic regimen for melanoma brain metastases, particularly in asymptomatic patients without corticosteroids. ASCO/ESMO guidelines support systemic dual checkpoint blockade for asymptomatic melanoma brain metastases, with SRS reserved for symptomatic or large lesions. For BRAF-mutant patients, BRAF+MEK inhibition (COMBI-MB) offers faster but less durable intracranial control.