Pre-Treatment Higher Eosinophil Count Is Associated with Improved Survival after Autologous CAR-T Therapy for Hematologic Malignancies

Introduction:

Eosinophils (Eo) play an important role in recruiting T-cells to tumors, and recently were associated with an improved response to immune checkpoint inhibitor therapy. Preclinical studies showed that combining stem cell-derived eosinophil therapy with CAR-T leads to improved treatment efficacy. Early clinical data reported a significant difference in eosinophil counts at two months post-CAR-T therapy between patients who achieved a complete response and those who did not. However, data on this subject thus far have been limited. We therefore conducted a retrospective analysis to examine the impact of elevated peripheral blood eosinophil count before and after autologous CAR-T cell therapy on survival.

Methods:

We included adults (>18 yrs) with leukemia, lymphoma or myeloma treated with commercial CAR-T products targeting CD19 and BCMA between January 2018 and June 2023. We collected patient, disease and treatment characteristics, CAR-T related toxicities, history of autoimmune diseases, and peak absolute and percentage eosinophil counts before CAR-T cell therapy and at ~30, 60, and 100 days after CAR-T cell infusion. Analysis of categorical variables was performed by Chi-square tests and of continuous variables by Wilcoxon rank-sum tests. Recursive partitioning methods were used to find the optimal eosinophil count cutoff based on two-year overall survival (OS) and progression-free survival (PFS).

Results:

We included 106 patients with a median age of 62.3 years (range 19.4-86.2), 61% were male, 66% had DLBCL; 35% received Tisagenlecleucel and 34% Axi-cell. Median peak Eo count pre CAR-T was 4.2% (range 0-26.3%) or 200 cells/μl (range 0-2400). Using recursive partitioning, a peak eosinophil count within 30 days pre CAR-T of 150 determined the high eosinophil group (n=56, 53%) with median eosinophil count of 300 (range 100-2400) and low eosinophil group (n=50, 47%) with median eo count of 100, range 0 - 100 cells/μl). Notably, high eo cohort had significantly higher peak Eo count early post CAR-T (0-30 days) but not at later time points (days 30-100 and beyond) compared to low eo cohort.

Characteristics of the two groups were similar except in the low eo group, fewer patients had DLBCL compare to other histologies (51% vs 78%), more had bone marrow transplant pre-CAR-T (42% vs 18%), more had pre-existing autoimmune condition (34% vs 11%), and high ferritin levels pre-and post-CAR-T therapy. Eo levels (hi vs low) were not associated with CR rate (76% vs 66%) or toxicity. Interestingly, high eo patients experienced improved OS at one year compared to the low eo group (91% [95%CI 77-96%] vs 73% [95%CI 56-85%]; p=0.01). PFS was also higher in the high eo group but did not reach statistical significance (66% [51-77%] vs 51% [36-65%]; p=0.08). There was no difference in ORR or relapse rate at one year post CAR-T therapy. In multivariate analysis adjusted for age, diagnosis and ferritin level, Eo count was associated with improved OS at hazard ratio 0.18 (95% CI 0.05-0.58; p<0.01) but not with PFS (HR 0.59; 95%CI 0.3-1.15; p=0.12). We will discuss the causes of death, UVA and MVA detail.

Conclusion:

Our study suggests that in patients who received commercial autologous CAR-T cell therapy for relapsed/remitted hematologic malignancies, high pre-CAR-T eosinophil count within 4 weeks of CAR-T infusion is associated with overall survival one year after CAR-T treatment. Further studies are necessary to investigate this relationship.

Maakaron:Precision Biosciences: Research Funding; Scripps Research Institute: Research Funding; VOR Bio: Research Funding; Affimed: Research Funding; CRISPR: Research Funding; Atara: Research Funding; Gilead: Research Funding. Bachanova:Miltenyi: Other: DSMB; Allogene: Consultancy; Astra Zeneca: Consultancy; CRISPR: Consultancy; Citius: Research Funding; Incyte: Research Funding; Gamida Cell: Research Funding.