Abstract Introduction: CAR T-cells have significantly improved the prognosis of patients with relapsed or refractory (R/R) large B-cell lymphoma (LBCL). Nevertheless, around 50% of R/R LBCL patients treated with CAR-T cells fail to achieve durable remission, highlighting the need for biomarkers that can predict the efficacy of CAR-T cell therapy, in order to stratify patients and guide timely therapeutic interventions. Recently, several prognostic factors for CAR T-cell therapy have been reported, mostly focused on patient characteristics or disease status (ctDNA, high tumor burden, TP53 genetic alteration). Among potential predictive factors, the composition and phenotype of peripheral blood immune cells at the time of leukapheresis have emerged as promising candidates. These immune profiles may reflect the patient's immune fitness and the quality of the starting material for CAR T-cell production, both of which are likely to influence clinical outcomes. The current study was conducted as an ancillary study of the ALYCANTE trial (NCT04531046), an open-label, phase 2 study designed to assess the efficacy and safety of axicabtagene ciloleucel (axi-cel) (median age 70 years) with R/R LBCL ineligible for autologous stem cell transplantation after one prior line of therapy (Houot R, Nat Med, 2023). Complete metabolic response at 3 months (CMR3) was obtained in 71% of cases. We used high-dimensional mass cytometry (CyTOF) to comprehensively characterize circulating immune cells at the time of leukapheresis. Our objective was to identify early immunological predictors associated with CMR3 and/or PFS12. Methods: Blood samples were prospectively collected within the ALYCANTE trial. Among the 40 patients analyzed, 27 (62.5%) achieved CMR3, and 19 (47.5%) had progressed by 12 months (PFS12 event). No statistical differences were found between groups (including age, sex ratio, ECOG, Ann Arbor, IPI, ctDNA level, or TP53 mutation). Peripheral blood mononuclear cells collected at leukapheresis were cryopreserved and subsequently analyzed using CyTOF. A total of 71 metal-conjugated antibodies were used across two panels, enabling the identification and characterization of 67 immune cell subsets, including T cell, NK cell, and myeloid subpopulations and/or functional states (Roussel M, Cell Reports Med, 2021). In parallel, ctDNA, previously reported as prognostic in this clinical trial (Delfau-Larue MH, EHA2023) has been quantified for the same patients. Results: CMR3 patients had higher counts of circulating myeloid cells (My17, My19), NK cells (NK03), and T cells (T04, T06, T14, and T15) (p 0.05). Consistently, patients without a PFS12 event also showed higher counts of My17 (p 0.05) and T15 (p 0.01). In contrast, the occurrence of a PFS12 event was associated with increased counts of NK06 and T08 (p 0.05). Their phenotype was as follow:populations of monocyte/dendritic cells: My17 (CD68pos cells co-expressing S100A9, CD36, high levels of HLADR, CD36, CCR2dim) and My19 (CD68poscells, CD11b, CD38, CD16, CD45RA, lacking HLADR and CCR2);populations of CD3 T cells including: CD8 T cells: T08 (CD45RApos), T14 (CCR7int), T15 (CD8posCD45ROpos cells expressing CD69 and high levels of CD161) and CD4 T cells (T06, T04, both CD45RAnegCCR7neg);NK cell subsets included NK09 (CD16high, CD57high, CD8, Tbet, Granzyme, CD45RA, TIGIT and TCF1); NK06 (CD16high, CD57high, Ki67high); NK03 (CD7pos, CD57low, Tbetneg, CD117pos). The presence of the cell subsets was independent from ctDNA load and molecular classifier or TP53 mutation. In multivariable analysis, only increased levels of T15 were associated with a lower risk of PFS12 event (OR 0.33). No significant associations were found between immune cell subsets and CMR3. Conclusions: Our study identifies specific immune cell signatures (including myeloid-, T-, and NK- cells) in the blood at leukapheresis that are associated with CMR3 and/or PFS12 following axicabtagene ciloleucel therapy in R/R LBCL. Notably, a subset of T cells characterized by CD8pos CD45ROpos CD69pos, CD161high was independently associated with PFS12, irrespective of baseline clinical characteristics, ctDNA levels, or genetic risk factors. Early immunoprofiling could thus help guide risk-adapted strategies before CAR-T cell infusion.
Gallou et al. (Mon,) studied this question.