The cns-IPI score is predictive of severe immune effector cell-associated neurotoxicity after axicabtagene ciloleucel treatment in patients with large B-cell lymphoma

Abstract Background. Chimeric antigen receptor (CAR) T-cell therapy has significantly improved the prognosis of patients with relapsed/refractory (R/R) large B-cell lymphoma (LBCL). However, it is frequently associated with acute toxicities, such as cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity (ICANS). In this study, we aimed to assess risk factors for ICANS and to evaluate the utility of previously described prognostic scores in predicting its incidence and severity. Methods. Retrospective, single-center study including R/R LBCL patients who received axi-cel from November 2019 until June 2025. Clinical, disease-related and treatment-specific variables were collected, including toxicity and efficacy outcomes following axi-cel treatment. Severe ICANS was defined as grade ³3. We performed a univariate and multivariate logistic regression analysis to identify parameters at time of lymphodepleting chemotherapy (LDC) associated with ICANS development. Results. The study included 130 patients who received axi-cel within the study period. The median age was 58 (range 23-74), most patients (60%) were male and presented an advanced stage of disease (75%). The median number of prior treatment lines was 2 (range 1-5), and 48% were considered primary refractory. Regarding toxicity, CRS and ICANS of any grade occurred in 90% and 56% of patients, respectively; grade ≥3 were reported in 12% (CRS) and 21% (ICANS). Baseline variables associated with an increased rate of ICANS were presence of ≥2 extranodal sites (45% vs. 14%, p<0.001), a history of kidney and/or adrenal involvement (18% vs. 3.5%, p=0.012), a high CAR-HEMATOTOX score (49% vs. 26%, p=0.011) and a high-risk CNS-IPI (≥4) at LDC (29% vs. 3.5%, p<0.001). All patients with a prior history of CNS involvement (n=10) developed ICANS after axi-cel infusion (grade ≥3 in 44%), as did most patients with adrenal and/or kidney involvement at time of LDC (14/15, grade ≥3 in 54%). Among the 27 patients with available total metabolic tumor volume (TMTV) data, we did not identify significant differences according to the occurrence of any-grade ICANS (121.15 vs. 74.10 cm³, p=0.89). Concerning efficacy, overall and complete response rate were 90% and 63%, respectively. With a median follow-up of 19 months (95%CI, 13.6-24.7), 24-month PFS was 60% and 24-months OS was 80%. There were no differences in survival according to any-grade or severe ICANS. In the multivariate analysis, CNS-IPI at time of LDC was identified as an independent risk factor for any-grade ICANS (OR 14.2, 95%CI 3.23-100; p<0.001) and severe ICANS (OR 20, 95%CI 4.17-100; p<0.001). In contrast, elevated ferritin, C-reactive protein, TMTV, lactate dehydrogenase and β2-microglobuline levels were not significantly associated with ICANS. Additionally, established clinical scores such as CAR-HEMATOTOX and EASIX/mEASIX did not predict the occurrence or severity of ICANS in our cohort. Conclusions. A high CNS-IPI at LDC identified patients at increased risk of developing severe ICANS following axi-cel therapy. These findings may allow risk-adapted strategies for the prophylactic and therapeutic management of this adverse event.