Background
CheckMate 498 was a phase III randomized open-label trial evaluating nivolumab (anti-PD-1) plus radiation therapy (RT) versus temozolomide (TMZ) plus RT in patients with newly diagnosed MGMT-promoter–unmethylated GBM. MGMT-unmethylated GBM represents ~60% of GBM cases and is characterized by relative resistance to alkylating chemotherapy (TMZ), with a particularly poor prognosis (mOS ~13–14 months with STUPP). The rationale for replacing TMZ with nivolumab in this population was that: (1) TMZ provides limited benefit in MGMT-unmethylated tumors, and (2) immunotherapy might be particularly active in the absence of TMZ-induced lymphopenia.
Interventions and follow up
Arm A: Standard TMZ + RT (STUPP regimen): radiation 60 Gy + concurrent TMZ 75 mg/m² → adjuvant TMZ × 6 cycles
Arm B: Nivolumab 240 mg q2w + RT 60 Gy (no temozolomide) → nivolumab maintenance until progression
Primary endpoint: Overall survival (OS); secondary: PFS, ORR, safety, QoL
mFollow up: 42.1 months
Arm B: Nivolumab 240 mg q2w + RT 60 Gy (no temozolomide) → nivolumab maintenance until progression
Primary endpoint: Overall survival (OS); secondary: PFS, ORR, safety, QoL
mFollow up: 42.1 months
Results
mOS: 13.4 months (nivo+RT) vs 14.9 months (TMZ+RT), HR 1.31, 95% CI 1.09–1.58 (nivolumab inferior to TMZ)
mPFS: 6.0 vs 6.3 months, HR 1.12 (NS)
ORR: 7.8% vs 7.5%
PD-L1: did not predict nivolumab benefit
Subgroups: no subgroup showed benefit; worse OS with nivolumab across all prespecified subgroups
mPFS: 6.0 vs 6.3 months, HR 1.12 (NS)
ORR: 7.8% vs 7.5%
PD-L1: did not predict nivolumab benefit
Subgroups: no subgroup showed benefit; worse OS with nivolumab across all prespecified subgroups
Adverse events
Grade ≥3 AEs: 33.7% (nivo) vs 48.5% (TMZ) — fewer Grade ≥3 with nivolumab
Immune-mediated AEs (nivo, all grades): rash 21%, fatigue 23%, endocrinopathies 15%
Grade ≥3 immune-mediated: 11.4%
TMZ arm Grade ≥3 hematologic: neutropenia 8%, thrombocytopenia 10%
Corticosteroid use at baseline: 30% — potentially suppressing nivolumab efficacy
Immune-mediated AEs (nivo, all grades): rash 21%, fatigue 23%, endocrinopathies 15%
Grade ≥3 immune-mediated: 11.4%
TMZ arm Grade ≥3 hematologic: neutropenia 8%, thrombocytopenia 10%
Corticosteroid use at baseline: 30% — potentially suppressing nivolumab efficacy
Conclusions
Nivolumab plus radiotherapy was inferior to temozolomide plus radiotherapy in newly diagnosed MGMT-unmethylated GBM (mOS 13.4 vs 14.9 months, HR 1.31), definitively ruling out PD-1 blockade as a replacement for TMZ-based chemoradiation even in the subgroup where TMZ confers the least biochemical benefit, and reinforcing that GBM represents a uniquely immunotherapy-resistant tumor entity.
Key Limitations
Open-label design; high baseline steroid use (30%) likely impairs nivolumab activity significantly — corticosteroid dependence is a fundamental confound in GBM immunotherapy trials; the immunosuppressive GBM microenvironment (MDSCs, Tregs, TGF-β) may be irreducible by PD-1 alone; MGMT-unmethylated GBM itself may be biologically immunologically cold; TMB in GBM is generally low; no IDH mutation status stratification; nivolumab + RT combinations in other tumor types (head and neck, NSCLC) show benefit — the GBM failure may reflect biology rather than strategy; companion CheckMate 548 (MGMT-methylated, nivo+TMZ) also failed.
Clinical Context
CheckMate 498, together with CheckMate 548 and CheckMate 143, comprehensively established that nivolumab-based immunotherapy fails across all GBM settings (recurrent, newly diagnosed MGMT-unmethylated, newly diagnosed MGMT-methylated). This trilogy of negative results defines the current state of GBM immunotherapy and highlights the unique challenges of the blood-brain barrier, immunosuppressive TME, and steroid dependence in this disease. Standard of care for MGMT-unmethylated GBM remains RT + TMZ (STUPP) ± TTFields (EF-14). The DCVax-L results and early-phase trials of personalized vaccines, oncolytic viruses, and EGFR-targeted CAR-T cells represent the remaining active immunotherapy research frontiers in GBM.