Abstract
This real-world prospective observational study across 21 Italian centers (CART-SIE) compares axicabtagene ciloleucel (axi-cel) and tisagenlecleucel (tisa-cel) outcomes in 485 patients with relapsed/refractory large B-cell lymphoma with baseline characteristics matched by stabilized inverse propensity score weighting. Axi-cel versus tisa-cel had higher all-grade cytokine release syndrome (78.6% vs. 89.3%, P = 0.0017) and neurotoxicity (9.9% vs. 32.2%, P < 0.0001) but also superior progression-free survival (PFS) at 1 year (46.5% vs. 34.1%, P = 0.0009). Even among patients who failed bridging therapy, axi-cel PFS was superior to tisa-cel (37.5% vs. 22.7%, P = 0.0059). Differences in overall survival and high-grade immune toxicities were not significant. The CAR-HEMATOTOX score not only predicted hematologic toxicity but also 1-year survival outcomes (51.5% in CAR-HEMATOTOX high vs. 77.2% in CAR-HEMATOTOX low, P < 0.0001). Twenty patients developed second primary malignancies, including two cases of T-cell neoplasms. These findings enable more informed selection of anti-CD19 CAR T-cell therapy, balancing bridging, safety, and efficacy considerations for individual patients. Significance: The findings of this study on 485 patients with relapsed/refractory large B-cell lymphoma treated with commercial axi-cel and tisa-cel indicate axi-cel's superior PFS after propensity score weighting. The predictive utility of CAR-HEMATOTOX in assessing not only toxicity but also outcomes across both CAR T-cell products may guide future risk-stratified management strategies.