Why This Study Matters
Glioblastoma (GBM) is refractory to standard treatment, with median survival of only 14 months despite maximal safe resection, radiotherapy, and temozolomide chemotherapy. Immune checkpoint inhibitors targeting PD-1 and CTLA-4 have not improved outcomes in GBM, unlike in many other cancers, in part due to a tumor microenvironment marked by profound immune exhaustion and elevated expression of additional checkpoint molecules such as LAG-3. This phase 1 trial tested whether adding an anti-LAG-3 antibody (relatlimab), alone or with the anti-PD-1 antibody nivolumab, could be safely delivered in recurrent GBM.
Study Design
This was an open-label, multicenter, dose-escalation trial (Adult Brain Tumor Consortium study ABTC 1501) with sequential allocation and a window-of-opportunity neoadjuvant arm prior to surgery, alongside a standard adjuvant arm. Patients were assigned to one of two treatment groups: relatlimab monotherapy, dose-escalated from 80 mg to a maximum tolerated dose of 800 mg (23 patients: 17 adjuvant, 6 neoadjuvant); or relatlimab (escalated from 80 mg to 160 mg) plus a fixed 240 mg dose of nivolumab (23 patients: 16 adjuvant, 7 neoadjuvant). The primary objective was safety — determining the maximum tolerated dose of each regimen. Secondary objectives were site-determined 1-year progression-free and overall survival rates, and radiographic response by modified Response Assessment in Neuro-Oncology (mRANO) criteria.
Patient Population
Adults with first-time recurrent glioblastoma after prior standard treatment with radiation therapy and temozolomide were enrolled between September 9, 2016, and April 29, 2020. In total, 46 patients were treated, split evenly between the two cohorts (23 each).
Primary Endpoint Results
The primary safety endpoint was met, establishing a maximum tolerated dose of 800 mg for relatlimab monotherapy and 160 mg relatlimab plus 240 mg nivolumab for combination therapy. Dose-limiting toxicities occurred in 0 of 23 (0%) monotherapy patients versus 6 of 23 (26%) combination-therapy patients. As a phase 1 trial, efficacy endpoints were secondary, descriptive, and not statistically powered: 12-month overall survival was 34.8% with monotherapy versus 52.2% with combination therapy, and no patient had a confirmed partial or complete response by central mRANO review.
The article explicitly cautions that its survival findings "cannot be disentangled from subsequent therapies and selection bias and do not imply clinical benefit," since the study was not designed or powered to formally assess survival. A separate, non-formal historical comparison — a prior ABTC phase 2 trial of nivolumab monotherapy in recurrent GBM, which had a 12-month overall survival rate of 22% — is offered only as context, not as a validated control, given differences in trial design, era, and patient selection.
Subgroup and Exploratory Correlative Analyses
Pseudoprogression — initial radiographic progression followed by stabilization or regression — was common across groups: it occurred in 8 of 17 adjuvant-relatlimab, 9 of 16 adjuvant-combination, 5 of 6 neoadjuvant-relatlimab, and 3 of 7 neoadjuvant-combination patients. Five of the 46 patients (11%) survived beyond 24 months, mostly in the combination arm, where a further breakdown showed 12-month overall survival of 86% for the neoadjuvant combination group versus 38% for the adjuvant combination group. Exploratory, non-formal biomarker work found that tumors with elevated baseline interferon signaling and increased T cell clonality were enriched among patients with durable responses to combination therapy, though no hazard ratio, confidence interval, or p-value was reported for this qualitative finding.
| Analysis | Reported Result |
|---|---|
| IDH-mutant survival range (6 of 46 patients) | 2.7–70.7 months |
| IDH-wildtype survival range (40 of 46 patients) | 2.6–73.6 months |
| Combination arm, neoadjuvant vs adjuvant 12-mo OS | 86% vs 38% |
| Formal statistical subgroup HR / CI values | Not reported — study not powered for subgroup efficacy analysis |
Safety Profile
Grade 3-4 adverse events occurred only in the combination arm (6 of 23 patients; none with monotherapy): cerebral edema (2 patients), grade 3 muscle weakness (1), grade 3 hypertension (1), grade 3 syncope (1), and grade 3 thyroiditis (1). Both cerebral edema cases were managed with a short course of dexamethasone. The most frequent any-grade adverse events with monotherapy were fatigue and headache (5 of 23, 22% each) and lipase elevation (4 of 23, 17%); with combination therapy, fatigue (10 of 23, 43%), lymphopenia (8 of 23, 35%), and nausea (6 of 23, 26%) were most common. No dose-limiting toxicities occurred with monotherapy at any dose level.
| Outcome | Relatlimab Monotherapy (n=23) | Relatlimab + Nivolumab (n=23) |
|---|---|---|
| Dose-limiting toxicities | 0 (0%) | 6 (26%) |
| Grade 3–4 adverse events | 0 (0%) | 6 (26%) |
| Most common any-grade AE | Fatigue / Headache, 5 (22%) each | Fatigue, 10 (43%) |
| Lipase elevation / Lymphopenia | Lipase, 4 (17%) | Lymphopenia, 8 (35%) |
| Cerebral edema (DLT) | 0 | 2 (9%) |
Interpretation and Broader Context
This phase 1 trial demonstrates that relatlimab, alone or combined with nivolumab, has an acceptable safety profile in recurrent glioblastoma at the doses tested, with toxicity manageable through standard supportive care. Because the study was not designed or powered to assess efficacy, the overall survival findings — 34.8% versus 52.2% at 12 months — should be interpreted as hypothesis-generating rather than evidence of a survival benefit; the historical nivolumab-monotherapy comparison (22% 12-month overall survival) is explicitly flagged by the authors as contextual only, not a validated control.
The authors report that a phase 2 trial through the Alliance (A077201/A07221, ClinicalTrials.gov NCT06325683) is planned or under way to formally assess the efficacy of relatlimab plus nivolumab in glioblastoma, informed by the immune correlative findings from this study — elevated baseline interferon signaling and T cell clonality as potential biomarkers of response.
In a phase 1 trial of 46 patients with recurrent glioblastoma, relatlimab (anti-LAG-3) alone or combined with nivolumab (anti-PD-1) met its primary safety endpoint, with dose-limiting toxicities in 0% of monotherapy patients versus 26% of combination-therapy patients and no confirmed radiographic responses under central review. Combination therapy showed a numerically higher 12-month overall survival (52.2% vs 34.8%) alongside exploratory biomarker signals — elevated interferon signaling and T cell clonality — enriched among durable responders, but the authors stress these survival and biomarker findings are descriptive, not proof of clinical benefit. A planned Alliance phase 2 trial (NCT06325683) will formally test whether the combination improves outcomes in glioblastoma.