A comprehensive study published in Nature Genetics reveals that clonal hematopoiesis (CH) frequently expands following CD19-directed chimeric antigen receptor T cell therapy, a process that contributes to prolonged cytopenias and the development of secondary myeloid cancers. The research, conducted by a team including Mark P. Hamilton and Ash A. Alizadeh, investigated the molecular and clinical mechanisms underlying these toxicities in patients with large B cell lymphoma. By comparing CAR T cell-treated patients with propensity-matched controls who received autologous hematopoietic cell transplantation, the study provides a detailed framework linking immune reconstitution failures, marrow inflammation, and clonal evolution to long-term hematologic complications.
The study evaluated a cohort of patients who received CAR19 therapy, specifically axi-cel, after failing chemoimmunotherapy. The researchers compared these patients to a matched group treated with autologous hematopoietic cell transplantation. This design allowed for a direct assessment of the relative risks associated with cellular therapy versus traditional transplant approaches. The analysis focused on patients treated between January 2018 and July 2022, ensuring a contemporary clinical context. The primary objective was to understand why CAR T cell therapy, despite improving survival rates, is often accompanied by severe and persistent blood count abnormalities.
CAR T cell-treated patients exhibited significantly impaired immune reconstitution compared to the transplant controls. The cumulative incidence of severe infections, defined as grade 3 or higher, was substantially higher in the CAR group. Statistical analysis using Fine-Gray competing risk regression demonstrated a hazard ratio of 2.54 for severe infections in CAR-treated patients relative to controls. Non-relapse mortality was also increased beyond day 30, primarily driven by deaths from infection. These findings highlight the significant clinical burden of immunosuppression that persists after the initial expansion phase of CAR T cell therapy.
Bone marrow biopsies from patients with persistent cytopenias revealed a distinct microenvironment characterized by ongoing B cell production and inflammation. Flow cytometry and single-cell RNA sequencing showed that CD19-positive B cell progenitors remained in the marrow but failed to egress into the peripheral circulation. This suggests that residual CAR T cells continue to target developing B cells, leading to a state of B cell aplasia in the blood despite active production in the marrow. The marrow environment was also marked by interferon-mediated inflammation, with native T cells displaying strong interferon signaling programs.
The study found that clonal hematopoiesis mutations were prevalent in the post-treatment marrow. Among patients with cytopenias, 90.3% had associated mutations detected on clinical sequencing panels. A significant portion of these patients, 41.9%, were ultimately diagnosed with myelodysplastic syndromes or acute myeloid leukemia. The most frequently mutated genes included DNMT3A and TP53. TP53 mutations were particularly significant, as they were strongly associated with the development of therapy-related myeloid neoplasms. In contrast, DNMT3A mutations appeared to be protective against malignant transformation, a finding consistent with their lineage-specific distribution in lymphoid cells.
Longitudinal analysis of cell-free DNA and peripheral blood samples demonstrated that CH clones expand after CAR T cell infusion. The researchers identified clear evidence of gene selection in CH genes such as DNMT3A, TP53, and PPM1D. In many cases, mutations that were present at low variant allele fractions before infusion increased significantly over time. This expansion was most pronounced in the myeloid compartment, suggesting that the inflammatory pressure exerted by the therapy favors the outgrowth of specific clonal populations. The study traced 19 therapy-related myeloid neoplasm-associated mutations to their earliest time points, finding that 15 were detectable in peri-infusion samples.
Single-cell DNA sequencing provided further insight into the clonal architecture of these mutations. The analysis revealed that most CH mutations exist as independent clones rather than co-occurring in the same cell. When multiple mutations were present, they typically followed a hierarchical pattern of an age-associated mutation followed by a DNA damage response mutation. This suggests that chemotherapy-induced DNA damage may confer a fitness advantage to pre-existing age-related clones. The independence of these clones explains the observed clonal competition during the post-infusion period, where different lineages expand and contract in response to the therapeutic environment.
A striking finding was the frequent integration of the CAR vector into pre-existing CH mutations. In peripheral blood samples collected at peak expansion, 8 out of 13 patients with known CH mutations had at least one CAR-positive CH-mutant cell detected. This integration was highly specific to certain genes, with DNMT3A being the most common site of integration. TP53 and PPM1D were largely excluded from CAR vector integration, likely due to their restriction to the myeloid lineage. The presence of CH mutations in expanding CAR T cells suggests that these mutations may drive the persistence and function of the therapeutic cells, although the long-term implications of this interaction remain to be fully understood.
The study also examined the impact of prior chemotherapy on CH development. Comparing pre-chemotherapy and post-chemotherapy samples, the researchers found a significant increase in CH burden after treatment. This effect was most pronounced in DNA damage response genes, indicating that chemotherapy drives the formation of new CH mutations. The presence of pre-infusion CH was associated with reduced recovery of neutrophils and platelets after CAR T cell therapy. This association suggests that CH may serve as a biomarker for underlying DNA damage, which impairs hematopoietic recovery. However, the study notes that it is not yet clear whether CH is a cause or a consequence of impaired hematopoiesis.
Therapy-related myeloid neoplasms following CAR T cell therapy were molecularly and clinically comparable to those arising after autologous transplant. There were no significant differences in age, sex, cytogenetic risk features, or progression-free survival between the two groups. This finding challenges the notion that post-CAR myeloid malignancies represent a unique entity. Instead, it suggests that the risk of secondary malignancy is a shared consequence of the intensive hematologic insults involved in both treatment modalities. The development of these neoplasms was associated with multiple prior cellular therapies, indicating that cumulative damage plays a critical role in malignant transformation.
The researchers propose a model in which post-CAR hematologic toxicity is multifactorial, involving pre-existing DNA damage, persistent B cell aplasia, and sustained marrow inflammation. The inflammatory environment fosters the selective outgrowth of DNA damage response mutant clones, which can lead to marrow failure and, in rare cases, malignancy. While most CH expansions are clinically inert, the study supports the monitoring of CH mutations in patients with persistent cytopenias, particularly those with TP53 mutations. Pre-infusion screening for high-burden CH may be reasonable in some cases, but the presence of CH should not preclude patients from receiving life-saving CAR T cell therapy.
The study has several limitations, including the potential impact of pandemic-era exposures and the relatively small number of patients who developed therapy-related myeloid neoplasms. Larger sample sizes are needed to differentiate the impact of specific CH mutations and to define the frequency of CAR integration into CH clones. Future research will need to determine whether treating patients in earlier lines of therapy, before the accumulation of significant DNA damage, can reduce the risk of CH expansion and secondary malignancies. These findings underscore the need for a deeper understanding of how CAR T cell vectors interact with pre-existing genetic alterations to improve the safety and efficacy of cellular therapies.
Subject of Research: Clonal hematopoiesis and cytopenias following CAR19 therapy
Article Title: Clonal hematopoiesis and cytopenias following CAR19 therapy
Article References: Hamilton, M. P., Phillips, N., Noordenbos, T., Boegeholz, J., Sugio, T., Sworder, B. J., Alig, S. K., Good, Z., Schroers-Martin, J. G., Tamaresis, J., Esfahani, M. S., Lu, Y., Olsen, M., Liu, C. L., Ehlinger, Z., Desai, M., Muffly, L., Negrin, R. S., Arai, S., … Alizadeh, A. A. (2026). Clonal hematopoiesis and cytopenias following CAR19 therapy. Nature Genetics. https://doi.org/10.1038/s41588-026-02795-5
Image Credits: AI Generated
DOI: 10.1038/s41588-026-02795-5
Keywords: Clonal, hematopoiesis, cytopenias, following, CAR19, therapy, scientific research
News Source: Drew Townsend. (November 8, 2026). CAR T Cell Therapy Expands Clonal Hematopoiesis. Scienmag.
