Defining a high-risk population for CAR-T-cell therapy: The impact of clonal hematopoiesis and pre-existing neutropenia on severe outcomes in patients treated for malignant and non-malignant conditions

Abstract

Abstract Importance: CAR-T cells show high efficacy in B-cell neoplasia and non-malignant B-cell driven autoimmune disorders, but therapy-related toxicity can limit treatment success. The identification of biomarkers informing patient selection and pre-emptive measures is important for enhancing treatment safety and improving clinical outcome. Objective: To analyze the influence of pre-treatment cytopenia and clonal hematopoiesis on clinical outcome after B-cell targeting CAR-T-cell therapy in malignant and autoimmune disorders. Design, Participants: This retrospective real-life analysis includes all adult patients with leukemia/lymphoma (L/L, 49), multiple myeloma ((MM, n=38 (n=1 coexistent L)), autoimmune diseases (AID, 11) treated at the University Hospital Tübingen with CD19, CD19/CD22 or BCMA-targeted CAR-T cells between March 2022 and March 2025, with last follow-up in April 2025. We analyzed routine clinical data for blood/CAR-T-cell counts, ICANS, Immune effector Cell-Associated Neurotoxicity Syndrome, CRS, Cytokine Release Syndrome, IEC-HS, Immune Effector Cell–Associated Hemophagocytic Syndrome, overall survival, OS, non-relapse mortality, NRM, remission induction, RI, progression-free survival, PFS and relapse-free survival, RFS, and retrospectively analyzed clonal hematopoiesis (CH) by next generation sequencing (NGS). Main Outcome(s) and Measure(s): Low pre-treatment blood counts most commonly occurring as pre-treatment neutropenia (pNP+) associated with enhanced NRM and reduced OS. Furthermore, pNP+ patients showed higher CAR-T cell expansion but no improvement in disease control. Detection of clonal hematopoiesis (CH+) further impaired outcome in pNP+ patients, but not by itself. AID patients show better CAR-T-cell tolerability than L/L/MM patients, but one AID patient retrospectively identified as pNP+CH+ showed lethal outcome. Results: Twenty-nine of 98 (29.6%) patients showed pNP+ (MM: 47.4%, L/L: 20.4%, AID: 9.1%) and twenty four of 64 (37.5%) patients CH (L/L: 47%, MM: 31%, AID: 45%, with TET2, DNMT3A, PPM1D and TP53 as most mutated genes). pNP+ vs. pNP- (regular pre-treatment neutrophil counts) showed shorter OS (p=0.006) and enhanced non-relapse mortality (NRM, p=0.006), with post-therapy early (p<0.001) and late neutropenia (p=0.001) and infections (p=0.011). Interestingly, similar RI (p=0.174), PFS (p=0.698) and RFS (p=0.605) were observed despite higher CAR-T cell counts (pNP+ vs. pNP-, day +35: 9.5% vs. 0.9%, p=0.0002; day +100: 0.60% vs. 0.17%, p=0.002). CH co-detection impaired OS and NRM in pNP+ patients but not alone or in pNP- (p=0.024, p=0.018) likely due to higher rates of early neutropenia (p=0.002). pNP+CH+ versus pNP+CH- patients furthermore showed enhanced CAR-T cell expansion (day +10, p<0.001, day +35, p=0.042) and B-cell aplasia persistence (day +180, p=0.049). AID patients tolerated CAR-T-cell therapy better than L/L/MM patients showing less posttreatment neutropenia (p=0.029), thrombocytopenia (p=0.009), severe CRS (p=0.035), ICANS (L/L/MM: 14/87; AID: 0/11) and infections (p=0.031). However, one pNP+CH+ AID patient succumbed with cytopenia, severe infections and immune effector cell hemophagocytic syndrome (IEC-HS). Conclusions and Relevance: Combined pre-treatment cytopenia and clonal hematopoiesis should be evaluated as risk factors for CAR-T-cell therapy morbidity and mortality in prospective studies.