Icans mitigation in high-risk elderly patients treated with CD28 co-stimulatory anti-CD19 CAR-T cells using a standardization protocol – a pilot study

Abstract Background CAR-T therapy has transformed the treatment of relapsed or refractory LBCL, with CD28-based products offering rapid responses but carrying a high risk of immune effector cell-associated neurotoxicity syndrome (ICANS). Reported ICANS rates reach 78%, with severe cases in up to 35%. Older adults are particularly vulnerable; yet, standardized mitigation strategies remain lacking, and current treatment is based on consensus guidelines that are not specifically defined and have limited generalizability in older high-risk patients – potentially excluding them from access to CAR-T therapy. We aimed to develop and implement a standardized ICANS mitigation protocol tailored for older adults receiving CD28-based anti-CD19 CAR-T therapy, addressing a critical gap in current clinical practice. Methods We conducted a single-center prospective pilot study at Tel Aviv Sourasky Medical Center between April 2023 and December 2024. Eligible patients were aged ≥75 years or ≥65 years with additional risk factors (neurologic comorbidity, ECOG ≥2, or high disease burden at lymphodepletion). Mitigation Protocol included seizure prophylaxis (levetiracetam and thiamine), early, more intensive corticosteroid treatment adapted by ICANS grade and anakinra for grade ≥3 or refractory ICANS. Refractory ICANS was defined as lack of clinical improvement or grade escalation within 24 hours of treatment initiation. Primary endpoints were ICANS duration and refractoriness. Secondary endpoints included response rates, non-relapse mortality (NRM), and infection rates. Clinical data were analyzed using non-parametric and survival methods. Results Between 04/2023 and 12/2024, 45 patients met eligibility criteria with a median age was 75 years (range 65–86); 78% had ECOG ≥2 and 42% had neurologic comorbidities—reflecting a clinically vulnerable population. ICANS occurred in 67% of patients; 31% developed grade ≥3 events, and 20% met criteria for refractory ICANS. The median onset was 5 days post-infusion (range 0–15), and median duration was 4 days (range 1–15), shorter than historical controls. Anakinra was administered in 24% of patients. Baseline neurologic comorbidity was associated with prolonged ICANS (median 7 vs 3 days; p=0.02), but not with severity. LDH showed a non-significant trend toward longer duration (p=0.07). Other risk factors including CNS involvement, early CRS onset, disease burden at lymphodepletion, and CAR-T expansion – were not associated with ICANS severity or duration. mEASIX showed modest predictive value (AUC=0.68), while ICANS-PSS was not predictive. Higher cumulative steroid dose was significantly associated with infections (p=0.0015) and with non-relapse mortality (p=0.02). Infections occurred in 42%, and 3-month NRM was 26.1%, driven by concomitant ICAHT and infections. CAR-T expansion and efficacy were preserved: 3-month CR rate was 44%, and 6-month PFS and OS were 46% and 59%, respectively. No significant association was observed between cumulative steroid dose and CAR-T expansion (Kendall's τ = 0.086, p = 0.42) Conclusion This prospective pilot study offers preliminary evidence that a structured ICANS mitigation protocol incorporating early, grade-adapted corticosteroids and timely anakinra integration, may reduce neurotoxicity in older high-risk patients undergoing CD28-based CAR-T therapy. The protocol was associated with a notably shorter ICANS duration despite a high incidence of severe events, and known risk factors were not predictive of severity or duration suggesting a potential mitigating effect. Anakinra has previously been used mainly as prophylaxis or salvage for steroid-refractory ICANS. In this study, it was incorporated earlier as part of routine ICANS management, guided by a pragmatic definition of refractoriness – lack of improvement within 24 hours of steroid initiation. This time-sensitive approach may be particularly relevant in older patients with limited physiologic reserve. The protocol was feasible and did not compromise CAR-T expansion or efficacy. However, treatment-related complications – particularly infections and non-relapse mortality were notable. These findings support further evaluation of mitigation-based, steroid-sparing strategies to optimize outcomes for older adults undergoing CAR-T therapy.