Cytopenias and infectious complications following CD19-directed CAR-T cell therapy in patients with B-ALL: A single-center retrospective analysis

Abstract

Abstract Background CD19-directed chimeric antigen receptor (CAR) T-cell therapies have transformed the treatment landscape for relapsed or refractory (r/r) B-cell acute lymphoblastic leukemia (B-ALL), resulting in the regulatory approval of three commercial products. While these therapies offer high rates of remission, immune effector cell–associated hematotoxicity (ICAHT) has emerged as a significant complication with prolonged cytopenias predisposing patients to infectious morbidity and mortality. The recently developed ALL-Hematotox (ALL-HT) score predicts prolonged neutropenia post-CAR T (Nair et al., Blood 2025), but real-world data on the incidence, risk factors, and clinical impact of cytopenias and infections in this setting remain limited. Methods We performed a retrospective analysis of 53 patients with B-ALL who received CD19-directed CAR T-cell therapy (brexu-cel or tisa-cel) between January 2018 and April 2025. Cytopenias (hemoglobin, platelets, ANC) were graded per CTCAE v5.0 at baseline (Day –7 to 0), Day 7, 14, 30, and 90. Infections were classified as early (<30 days) or late (31–100 days), stratified by grade, type (bacterial, viral, fungal), and outcome. Severe infection was defined as one requiring IV antibiotics and/or hospitalization, corresponding to CTCAE grade ≥3. Supportive interventions (IVIG, G-CSF, platelet/pRBC transfusions) were recorded. CRS and ICANS were graded per ASTCT criteria. ALL-HT score was retrospectively calculated; patients with score >4 was considered high risk. Associations between clinical variables and cytopenias or infections were evaluated using univariate logistic regression. Results Among 53 patients (median age 38 years [range 21–79]; 58% male), 46 (87%) reached Day 100 follow-up. Most patients (93%) received brexu-cel and underwent lymphodepletion with fludarabine/cyclophosphamide (Flu/Cy, 53%) or cladribine/cyclophosphamide (Cla/Cy 47%). Marrow blast burden at baseline was ≤5% in 72% of patients, 6–50% in 14%, and >50% in 14%. Prior therapies included blinatumomab (57%), inotuzumab (30%), and alloHSCT (23%). Bridging therapy was administered in 85% of patients. 38 patients were classified as low risk by the ALL-HT score, while 15 were high-risk. CRS occurred in 92% of patients (≥Grade 3 in 11%) and ICANS in 55% (≥Grade 3 in 35%). Tocilizumab, steroids, and anakinra were administered in 64%, 58%, and 12%, respectively. At Day 30, Grade ≥3 anemia, thrombocytopenia, and neutropenia were present in 8.2%, 27%, and 24% of patients, respectively; At Day 90, these declined to 2.6%, 18%, and 21%. IVIG was administered to 45%, G-CSF to 43%, and transfusions occurred in 20% (pRBC) and 9% (platelet) within 7 days, and in 24% and 19%, respectively, between Days 8–100. High-risk ALL-HT scores were significantly associated with Day 30 neutropenia (p=0.011) and thrombocytopenia (p<0.001). Flu/Cy lymphodepletion correlated with both Day 30 thrombocytopenia (p=0.008) and Day 90 neutropenia (p=0.029). Overall, 51% patients developed at least one infection by day 100; 29.8% experienced Grade ≥3 infections. Early infections occurred in 16 patients, 80% of which were bacterial, and 75% were Grade ≥3. Late infections occurred in 17 patients, 72% bacterial, and 41% were Grade ≥3. Infection-related mortality occurred in both early (n=2) and late (n=1) periods. High-risk ALL-HT scores were associated with both early (p=0.004) and late (p=0.029) infections. Additional risk factors for early infections included administration of bridging therapy (p=0.012), day 30 hypogammaglobulinemia (<400 mg/dL; p=0.006), and poor baseline performance status (p=0.033). Severe late infections were associated with day 90 hypogammaglobulinemia (p=0.032), Flu/Cy (p=0.015), and high pre-lymphodepletion blast burden (p=0.020). Conclusion Prolonged cytopenias and infectious complications are frequent and clinically impactful following CD19-directed CAR T-cell therapy for r/r B-ALL, with severe infections occurring in one-third of patients and contributing to early and late mortality. The ALL-HT score effectively stratifies hematologic and infectious risk in this cohort of patients treated with commercial CAR T. Hypogammaglobulinemia was strongly associated with severe infections, highlighting the need for prospective studies evaluating prophylactic IVIG. Compared to Flu/Cy, cladribine-based lymphodepletion may confer a more favorable cytopenia and infection profile and warrants further investigation.