Poor neutrophil recovery following CD19-targeted CAR-T cell therapy predicts inferior outcomes in adult patients with B-cell lymphoma

Abstract Background: Although CD19-targeted CAR-T cell therapy has improved outcomes for relapsed/refractory B-cell lymphoma, persistent neutropenia is recognized as one of the serious complications, observed in 24-64% of patients. However, the impact of this complication on long-term treatment outcomes and CAR-T cell efficacy remains unclear. Patients and methods: Of the 88 adult B-cell lymphoma patients who underwent CD19-targeted CAR-T cell therapy between June 2020 and February 2024, 74 patients alive without lymphoma progression until day 28 post-infusion were retrospectively analyzed. Poor neutrophil recovery was defined as a neutrophil count of less than 500/μL at day 28 post-infusion. Severity of CRS and ICANS was evaluated according to ASTCT consensus grading. The CAR-T cell counts in peripheral blood were measured by flow cytometry on days 4, 14, and 28 post-infusion, and the area under the curve (AUC) was calculated. T cells in the apheresis products and CAR-T cells in peripheral blood at day 28 post-infusion were classified into 4 groups, as previously described: naïve (N: CD45RA⁺CD45RO⁻), central memory (CM: CD45RA⁻CD45RO⁺CD62L⁺), effector memory (EM: CD45RA⁻CD45RO⁺CD62L⁻), and terminal effector memory re-expressing CD45RA (TEMRA: CD45RA⁺CD45RO⁻CD62L⁻). Progression-free survival (PFS) was defined as the interval from the date of CAR-T cell infusion to the date of lymphoma progression or death, whichever occurred first. Fisher's exact test and Mann-Whitney U test were used to compare binary variables and continuous variables, respectively. The cumulative incidence (CI) of lymphoma progression was evaluated with Gray's test, considering non-relapse mortality (NRM) as a competing risk. PFS was estimated with the Kaplan-Meier method and compared using the log-rank test. The Cox proportional hazard model and the logistic regression model were used for multivariate analysis of prognostic and risk factors, respectively. Values of p < 0.05 were considered to indicate statistical significance. Results: Of the 74 patients included in this study, 44 were male and 33 were female. The median age at CAR-T cell infusion was 61 years (range, 19–75 years). DLBCL was diagnosed in 56 patients (75%), and 29 (39%) had prior autologous stem cell transplant. Tisa-cel, Liso-cel, and Axi-cel were infused in 47, 25, and 2 patients, respectively. CRS of any grade and ICANS of any grade developed in 55 (74%) and 3 patients (4%), respectively. At day 28 post-infusion, 24 patients (32%) had neutrophil counts below 500/μL. Neutrophil count <1800/μL, Hb <9g/dL, and elevated LDH at baseline were significantly associated with poor neutrophil recovery in univariate analysis, and these factors were identified as independent risk factors in multivariate analysis (neutrophil count: odds ratio [OR] = 5.2; p = 0.02, Hb; OR = 8.5; p < 0.01, and elevated LDH; OR = 4.7; p = 0.03). The PFS rate in patients with poor neutrophil recovery was significantly inferior to those with good neutrophil recovery (49% vs. 78% at 1 year post-infusion; p = 0.02), and poor neutrophil recovery was identified as the sole independent prognostic factor in multivariate analysis (hazard ratio = 2.7; p = 0.02). NRM occurred in only 2 and 1 patient in each group, and although not statistically significant, there was a tendency for the CI of lymphoma progression to be higher in patients with poor neutrophil recovery (46% vs. 22%; p = 0.11). To elucidate potential mechanisms underlying poor outcomes, CAR-T cell expansion, T cell fitness in apheresis products, and CAR-T cell subsets at day 28 post-infusion were analyzed. The median AUC of CAR-T cells until 28 days post-infusion were similar between the 2 groups (547.4 vs. 597.1 cells·day/μL; p = 0.88). No significant difference was observed in the numbers of N + CM (6,176 vs. 3,781/μL; p = 0.07) and EM + TEMRA (7,112 vs. 9,277/μL; p = 0.63) in the apheresis products, and likewise, there were no significant differences in the numbers of any CAR-T cell subsets at day 28 post-infusion, including CD4+, CD8+, double negative, N, CM, EM, and TEMRA.Conclusion: These findings suggest that poor neutrophil recovery is associated with inferior PFS, potentially reflecting impaired CAR-T cell-mediated anti-lymphoma activity. Although no significant differences were observed in CAR-T cell expansion or subset distribution, further studies are warranted to elucidate the biological mechanisms linking hematologic recovery and therapeutic efficacy.