Real-world assessment of prophylactic anakinra on neurotoxicity after CD19 CAR T-cell therapy in R/R B-cell lymphoma: An inverse probability of treatment weighting analysis

Abstract

Abstract Introduction Immune effector cell-associated neurotoxicity syndrome (ICANS) is a common toxicity of CD19 CAR T-cell therapy (CAR T). Anakinra, an IL-1 receptor antagonist, has been used prophylactically to prevent ICANS, but real-world data remain limited. Our center recently implemented anakinra prophylaxis for high-risk patients over age 65 or receiving CAR products with a CD28 co-stimulatory domain. We conducted a retrospective, inverse probability of treatment weighting (IPTW) cohort study comparing R/R B-cell lymphoma (R/R BCL) patients who received anakinra prophylaxis prior to CAR T at our center to a historical control. Methods Patients with R/R BCL receiving CAR T in the second line or beyond at our center from 2018-2025 were included. Controls were those who would have received anakinra under our center's current policy: those aged ≥65 and/or receiving CD28 CAR T. Anakinra was initiated on day 0 at 100 mg q12h, titrated at the treating physician's discretion, and continued until discharge. The primary endpoint was incidence of grade ≥3 ICANS. Secondary endpoints included CRS/ICANS (any, grade 3+) incidence and duration, infection rates (any, grade 3+), and survival. IPTW was based on a propensity model including age, disease, elevated LDH, prior stem cell transplant, ≥3 prior lines, KPS ≤70, construct (CD28 vs 4-1BB), CNS/extranodal disease, and disease status at infusion. Balance was confirmed via standardized mean differences. Weighted logistic regression estimated adjusted odds ratios (aORs) for toxicity and survival outcomes, controlling for age, elevated LDH, ≥3 prior lines of therapy, KPS ≤70%, CNS or extranodal disease, disease status at infusion, and construct (CD28 vs 4-1BB). Results 171 patients were included (61 anakinra, 110 control). Pre-IPTW, compared to controls, patients receiving anakinra prophylaxis were significantly older (median age 72 vs 67; p = 0.03) and less likely to have ≥3 prior lines (21% vs 38%; p = 0.02). 54% of the anakinra group received 4-1BB CAR T vs 9.1% of controls (p < 0.001); the remainder received CD28 CAR T. When comparing the control and anakinra groups, there was no difference in pre-IPTW rate of disease histology (82% vs 84% DLBCL, 6.4% vs 6.6% FL, 12% vs 9.8% MCL), prior HCT (16% vs 8.2%), KPS ≤ 70 (24% vs 21%), IPI ≥3 (65% vs 61%), CNS involvement (14% vs 15%), extranodal disease (73% vs 62%), or SD/PD status at infusion (70% vs 62%). Post-IPTW, covariates were balanced between groups (all SMDs < 0.1), and patients had similar baseline characteristics. All subsequent reported values are post-IPTW. Rates of grade ≥3 ICANS were similar between anakinra and control groups (13.0% vs. 9.8%; p = 0.57), but anakinra patients had more all grade ICANS (39.6% vs 23.4%, p = 0.046). There was no difference in CRS (all grade/grade 3+), ICANS or CRS duration, time to ICANS or CRS onset, or cumulative steroid dose. Anakinra patients had higher rates of grade 3+ infections by day 30 (22.5% vs 8.0%; p = 0.006) with more grade 3+ bacterial infections than controls (16.5% vs 8.0%), as well as numerically higher rates of all grade infections (32.4% vs 18.1%; p = 0.06). There was no difference in median time to onset of grade 3+ infection for anakinra patients vs controls (5 vs 19 days; p = 0.18). Multivariable IPTW regression confirmed no association between anakinra and grade ≥3 ICANS (p = 0.16), but found increased odds of any grade ICANS (OR 3.05; 95% CI 1.30-7.15), grade 3+ CRS (OR 15.9; 95% CI 1.20-212.15), any grade infection (OR 3.55; 95% CI 1.34-9.40), and grade 3+ infection (HR 10.17; 95% CI 2.53-40.91). Response rates were similar in the control and anakinra groups: ORR 87% vs 93% (p = 0.36), CRR 72% vs 81% (p = 0.27). There was no difference in EFS or OS (p = 0.32 and p = 0.72). Multivariable Cox models showed that no survival impact from anakinra (EFS p = 0.22; OS p = 0.72). Conclusion Prophylactic anakinra did not reduce grade ≥3 ICANS but was associated with increased odds of any-grade ICANS, grade 3+ CRS, and severe infections (HR 10.17). These findings raise important concerns about the routine use of anakinra prophylaxis in CAR T recipients and suggest the need for further prospective evaluation.