Better late than never: risks of late cytopenias after CAR T

In this issue of Blood Advances, Wudhikarn et al¹ evaluated the frequency at which late cytopenias occur ≥day +30 after chimeric antigen receptor T-cell (CAR T) infusion, examined the association between late cytopenias and survival outcomes, and identified risk factors linked to late cytopenias.

The authors observed that grade >3 cytopenias were most common on day +30 (observed in 47% of patients), followed by a progressive decline over time; however, a clinically meaningful proportion of patients experienced cytopenias 1 year after CAR T infusion. Neutropenia occurred in 28%, 25%, 13%, 19%, and 6% of patients on day +30 and at 2, 3, 6, and 12 months after CAR T therapy, respectively.¹ Patients with late cytopenias had inferior 2-year progression-free survival after day +100 (38% vs 67%), 2-year overall survival after day +100 (60% vs 76%), and 1-year nonrelapse mortality (NRM) after day +100 (11% vs 4.4%; see Figure).¹ Infection was the leading cause of death among these patients, and all affected patients were classified as having late cytopenias. Finally, univariable and multivariable analyses demonstrated that a high CAR-HAEMATOTOX score (>2) was associated with prolonged cytopenias at 3 months after CAR T therapy, whereas receipt of bridging therapy, specifically chemotherapy, and the CAR T product (axicabtagene ciloleucel [axi-cel]), were associated with prolonged cytopenias at 6 months after infusion.¹

Late neutropenia after CAR T therapy leads to increased infections and NRM. Professional illustration by Diana Kentell.

The authors of this study provide a valuable assessment of well-recognized CAR T toxicity: late and prolonged cytopenias. Although informative, this analysis also raises several important areas for further consideration.

This work highlights the critical role of prolonged neutropenia as a driver of infections and NRM associated with CAR T therapy. A contemporary meta-analysis of commercial CAR T products previously identified infection as the leading cause of NRM, accounting for 53.4% of deaths, higher than the 27% reported in this study.² Other clinical studies have also demonstrated severe cytopenias leading to frequent infections and increased hospitalizations, although the mechanisms leading to CAR T–associated cytopenias remain poorly understood.³ Although the authors note that reporting of immunoglobulin (IgG) levels was limited, a subset of patients received IV IgG (IVIg), suggesting the presence of hypogammaglobulinemia. These observations raise important clinical questions about whether antimicrobial or IVIg prophylaxis should be applied and, if so, which agents should be used and for how long, after CAR T infusion. Although the American Society for Transplantation and Cellular Therapy has provided infection prophylaxis guidance, perhaps the CAR-HAEMATOTOX score could offer a more personalized approach for high-risk patients.⁴ IVIg replacement is more nuanced, and data from this study could possibly provide a more tailored approach to managing hypogammaglobulinemia after CAR T therapy.⁵

The authors identified axi-cel as being associated with a higher likelihood of 6-month cytopenias compared with lisocabtagene maraleucel (liso-cel) and tisagenlecleucel (tisa-cel). Notably, axi-cel incorporates a CD28 costimulatory domain, in contrast to the 41BB domains, as used in liso-cel and tisa-cel. Although this has not been directly evaluated, axi-cel may be associated with higher NRM secondary to infection than liso-cel and tisa-cel.² These observations raise important clinical considerations regarding product selection when choosing CAR T therapy.

Exposure to bridging chemotherapy was associated with higher rates of late cytopenias compared with other therapies, such as small-molecule inhibitors. This observation raises the question of whether chemotherapy-free bridging regimens should be preferentially chosen when feasible. Studies have explored the use of chemotherapy-free regimens, such as radiation or small-molecule inhibitors, before CAR T infusion to assess efficacy and reduce toxicity, and those studies have shown promising results.^6,7 In situations where a chemotherapy regimen is necessary, consideration of optimized antimicrobial prophylaxis or specific CAR T product selection may be warranted to avoid the sequelae of late cytopenias.

Given that late and prolonged cytopenias correlate with poor overall survival after CAR T infusion, perhaps allogeneic hematopoietic cell transplantation (alloHCT) for high-risk lymphoma should be reconsidered for certain patients. The postcyclophosphamide (PTCy) era has heralded incredible survival outcomes, with minimal NRM.⁸ Although treatment algorithms in lymphoma have increasingly shifted toward earlier use of CAR T therapy, the use of modern reduced-intensity conditioning, PTCy graft-versus-host disease prophylaxis, and maintenance therapy make alloHCT a reasonable alternative for those with high CAR-HAEMATOTOX scores or those anticipating chemotherapy-based bridging therapy. Although the use of alloHCT in lymphoma remains limited, a fresh look at alloHCT in the PTCy era is warranted for patients who are at the greatest risk of late cytopenias and infection-related NRM with CAR T therapy.

Future studies must focus on elucidating the mechanisms of late and prolonged cytopenias after CAR T therapy to inform clinical trial design and prevent infection-related NRM. With next-generation CAR T therapy emerging, improving risk stratification to identify patients who may be the most susceptible to late cytopenias and developing strategies to reduce the associated morbidity and mortality, will be crucial. Novel approaches to minimize exposure to lymphodepleting or bridging chemotherapy, like cytokine-armored CAR T products, may also support efforts to reduce NRM.⁹ This study represents an important step toward that goal.

Conflict-of-interest disclosure: B.C.B. received research support from CTI BioPharma Corp (a Sobi company), VITRAC Therapeutics, and Incyte; holds a patent (WO2015120436A2) related to CD4⁺ T-cell phosphorylated STAT3 as a marker and therapeutic target of acute graft-versus-host disease (GVHD); holds a provisional patent (WO2017058950A1) related to the use of JAK inhibitors for rejection and GVHD prevention; holds a patent (WO2019165156) on CD83 chimeric antigen receptor T cells; and participated in an advisory board event with Bristol Myers Squibb in 2025. A.N. serves on the advisory board of Bristol Myers Squibb and as a consultant for ADC Therapeutics.