Eugenio Galli, Daniele Mannina, Alessandro Corrente, Chiara De Philippis, Ilaria Pansini, Marcello Viscovo, Stefan Hohaus, Patrizia Chiusolo, Federica Sorà, Stefania Bramanti, Simona Sica
InflaMix was externally validated as a predictor of inferior PFS, severe CRS, and ICU admission after CAR-T therapy, and candidates as a practical tool for pre-infusion risk stratification and as a complement to existing scores.
OBJECTIVES: Baseline systemic inflammation has emerged as a determinant of outcome after CD19-directed CAR-T therapy. We evaluated the prognostic performance of the recently developed INFLAmmation MIXture Model (InflaMix) in an independent bicentric cohort of patients with large B-cell lymphoma (LBCL) receiving commercial CAR-T cells, focusing on treatment-related toxicities.
METHODS: We retrospectively analyzed 212 consecutive patients with relapsed/refractory LBCL treated with axi-cel, tisa-cel, or liso-cel between 2020 and 2026. Patients were classified according to the InflaMix inflammatory phenotype. Progression-free survival (PFS), CAR-T-related toxicities, ICU admission, and concordance with CAR-HEMATOTOX and mEASIX were assessed.
RESULTS: An inflammatory InflaMix phenotype was identified in 21% of patients and was associated with inferior PFS (12-month PFS: 30% vs. 56%; p < 0.001). While overall grade 2-4 CRS and ICANS rates were comparable, inflammatory patients showed higher rates of grade 3-4 CRS (11.1% vs. 1.8%; OR 3.18) and ICU admission (24.4% vs. 5.6%; OR 3.00). No association with ICAHT was observed. Concordance was poor with CAR-HEMATOTOX and moderate with mEASIX.
CONCLUSIONS: InflaMix was externally validated as a predictor of inferior PFS, severe CRS, and ICU admission after CAR-T therapy, and candidates as a practical tool for pre-infusion risk stratification and as a complement to existing scores.