Non-ICANS neurologic toxicity after BCMA CAR-T therapy: a systematic review and meta-analysis of 4630 patients with...

B-cell maturation antigen (BCMA)-directed chimeric antigen receptor T-cell (CAR-T) therapies have drastically improved outcomes for patients with relapsed or refractory multiple myeloma. Although CAR-T-associated cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome (ICANS) are well characterized, non-ICANS neurologic toxicities (NINTs) remain poorly defined, with limited data regarding incidence and risk factors. We performed a systematic review and meta-analysis of prospective clinical trials and real-world studies reporting neurologic toxicities after BCMA-directed CAR-T therapy, following PRISMA (preferred reporting items for systematic reviews and meta-analyses) guidelines. Random-effects meta-analysis was used to calculate a pooled point estimate of NINTs. Meta-regression was conducted to evaluate treatment-related predictors of risk. Reported NINTs were categorized by clinical phenotype when sufficient detail was available. Fifty-five cohorts comprising 4630 treated patients were included. The pooled point estimate for NINTs was 0.81% (95% confidence interval, 0.37%-1.77%). Incidence differed significantly by product, with a significantly higher frequency of NINTs after ciltacabtagene autoleucel (cilta-cel) than idecabtagene vicleucel (ide-cel; 4.6% vs 0.5%; P = .001) and experimental BCMA-directed constructs (4.6% vs 0.3%; P = .02). Additionally, meta-regression identified cilta-cel as independently associated with increased NINT risk compared with ide-cel. Cranial nerve palsies were the most frequently reported phenotype (32.3%, 43 events), followed by movement and neurocognitive treatment-emergent adverse events (12%, 16 events), then peripheral neuropathies (7.5%, 10 events). This meta-analysis establishes that NINTs are a rare but important toxicity after BCMA-directed CAR-T therapy, particularly cilta-cel. Standardized definitions and improved reporting of NINTs are needed to better characterize risk and inform surveillance strategies.