CAR T-cell hematotoxicity: is inflammation the key?

In this issue of Blood, Rejeski et al report a predictive model, CAR-HEMATOTOX, which identifies patients at highest risk of hematologic toxicity following treatment with CD19-directed chimeric antigen receptor (CAR) T-cell therapy for relapsed/refractory large B-cell lymphoma.¹ 

The US Food and Drug Administration has approved 5 CAR T-cell–based products for the treatment of hematologic malignancies. Despite the successes of CAR T-cell therapy, immune-mediated toxicities can lead to morbidity and mortality, limiting the widespread use of this therapy.² Short-term toxicities including cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are well described in the literature. The evolving understanding of the pathophysiology of these toxicities has led to improved toxicity management strategies.^2-4 Long-term toxicities, including B-cell aplasia, immune reconstitution, and prolonged cytopenias, are gaining more recognition. Prolonged cytopenias are common following CAR T-cell therapy, which predispose patients to infectious complications and nonrelapse mortality.^5,6 Several mechanisms contributing to hematopoietic recovery have been proposed, including severe CRS and ICANS, baseline cytopenias, and elevated levels of vascular endothelial growth factor and macrophage-derived chemokines.^5,6 The report by Rejeski et al provides further insight into the etiology of prolonged myelosuppression.

In this multicenter retrospective observational study, the authors analyzed 253 patients with large B-cell lymphoma (LBCL) who were treated with standard-of-care axicabtagene ciloleucel or tisagenlecleucel. They recognized protracted neutropenia in 68% of patients with the majority having an intermittent quick recovery. However, one-quarter of those patient develop prolonged aplasia, increasing hospitalization duration and risk of serious infections. In contrast to Jain et al, the authors did not find a correlation between higher-grade CRS or ICANS with prolonged neutropenia.⁵ Baseline cytopenias and elevated inflammatory markers, characterized by elevated ferritin and C-reactive protein (CRP), were associated with the duration of cytopenias. This observation supports a model proposed by our group in which the severity of CAR T toxicities is influenced by the interaction between a proinflammatory tumor microenvironment and infused CAR T cells.^4,7 The investigators developed the CAR-HEMATOTOX model using baseline features of platelet count, absolute neutrophil count, hemoglobin, CRP, and ferritin in a training cohort. Patients with a score over 2 were stratified as having “high-risk disease.” The authors are to be congratulated for validating the model in 2 independent cohorts, 1 in the United States and 1 in the European Union. Much of CAR T-cell therapy research focuses on the CAR T-cell product. By developing a model using baseline features, the authors support findings that emphasize the critical role of the host in the development of toxicities and treatment resistance.^4,7

The CAR-HEMATOTOX model identifies patients at risk for clinically significant prolonged neutropenia following CAR T-cell infusion. The results of this study are important and easy for clinicians to apply in clinical practice as the model was built using standard-of-care laboratory results obtained prior to CAR T-cell infusion. Due to the retrospective nature of this study, there are several limitations. Few patients had bone marrow samples available for evaluation; such samples are critical for further elucidating mechanisms of delayed hematopoietic recovery such as clonal hematopoiesis in patients who have had multiple lines of cytotoxic chemotherapy. Because this current study only included patients with LBCL, whether this model will apply to patients who primarily have marrow disease such as B-cell acute lymphoblastic leukemia or multiple myeloma is unclear. Furthermore, the financial toxicity of prolonged myelosuppression and its implications, such as infections and longer duration of hospitalization, will need to be further evaluated.

The study by Rejeski et al invites questions regarding the optimal management of prolonged cytopenias beyond supportive care. Standard antimicrobial prophylaxis guidelines including fungal prophylaxis are warranted. Although the use of growth factor and thrombopoietin receptor agonist have been reported, their benefit is unclear and there is no consensus or standardization among centers regarding their use. Although there was no significant prognostic impact of the CAR-HEMATOTOX score, the question of whether patients at high risk of prolonged aplasia benefit from autologous stem cell backup has been raised. The identification of early predictive markers of safety prior to the infusion of CAR T cells, as reported in this study, is vital for designing clinical trials aimed at further reducing the rates of severe toxicities in high-risk patients.^4,7 Moreover, identifying “low-risk” patients who may be safely treated as outpatients is important. Studies examining late toxicities such as prolonged myelosuppression have been largely limited to single-institution experiences. To that end, we support the blueprint set forth by the American Society of Hematology (ASH) task force on immunotherapies, which advocates for real-time data sharing and a centralized biobank.⁸ Multi-institutional collaborations using “real-world” data as reported in this study are critical for advancing the field and increasing the use of CAR T-cell therapy outside of highly specialized academic centers.