P005 | CAR-T CELL DYNAMICS AND MYELOID-CELL ALTERED PROFILE ARE ASSOCIATED WITH CLINICAL OUTCOME IN ANTI-CD19 CAR-T CELL THERAPY FOR RELAPSED/REFRACTORY B-CELL MALIGNANCIES: A REAL-LIFE COHORT STUDY

Anti-CD19 CAR-T cell therapy is effective in recurrent/refractory B-cell malignancies. However, the high rate of treatment related toxicities, including cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), and of disease recurrence, still limit its overall impact. Our aim was to shed light on the biological determinants associated with anti-CD19 CAR-T cell therapy efficacy and toxicity in a real-life setting. The infused product (IP) was analyzed using a 24-color flow cytometry panel. Next, blood samples collected at several timepoints (-7, +3, +7, +10, +14, +21, +30, +45, +60, +90, +180, +360 days) from infusion were analyzed to assess CAR-T pharmacokinetics (PK) and clonality, and to phenotypically characterize PBMC. Among 47 patients enrolled, 20 received axicabtagene ciloleucel (axi-cel), 9 brexucabtagene autoleucel (brexu-cel) and 18 tisagenlecleucel (tisa-cel). The Overall Response Rate was 56% at +90 days. CRS was observed in 44 patients (n=42 grade 1-2, n=2 grade 3), and ICANS in 13 patients (n=9 grade 1-2, n=4 grade 3-4). The infusion of CAR-T with a higher percentage of early-differentiated engineered T cells correlated with a better 1-year clinical response and lower toxicity. Interestingly, high frequencies of infused T cells expressing Inhibitory Receptors correlated with disease relapse. The PK curves showed the association between higher CAR-T counts at +7 days and a better 1-year outcome, threatened by higher toxicities. Integration site analyses performed on PBMCs longitudinally collected from patients who experienced disease relapse at +90 days revealed a significantly lower number of integrations, along with reduced complexity and persistence of circulating CAR-T cells, compared to patients who achieved a complete response (N= 8 patients/group). No differences were observed in the integration datasets from responder and non-responder patients regarding vector integration preferences or enrichment towards specific gene classes. Finally, the unsupervised analysis of PBMC phenotype revealed that a myeloid-cell altered profile, characterized by low expression of HLA-DR and CD86 and high levels of CD163, correlated with severe CRS/ICANS and was already detectable before the infusion. These results indicate the presence of immunological biomarkers within IP. Moreover, the amount and clonality of circulating CAR-T, and myeloid cell phenotype can predict clinical outcomes.