Mechanistic Insights Into Delayed Immune Toxicities Following BCMA CAR T-Cell Therapy
Key Clinical Summary
- Design/Population: An observational translational analysis evaluated 198 patients with multiple myeloma treated with commercial BCMA-directed CAR T-cell therapies to investigate delayed, non-classical toxicities distinct from CRS and ICANS, termed CAR T immune-related adverse events (CIRAEs).
- Key Outcomes: CIRAEs occurred in ~14% of patients, more frequently with cilta-cel than ide-cel and included enterocolitis and neurologic syndromes. Risk factors included cilta-cel exposure, higher CD4:CD8 ratio at apheresis, elevated day-14 ALC, and prior ICANS. Tissue biopsies and correlative studies demonstrated infiltration of BCMA CAR T-cells, predominantly cytotoxic CD4 central memory–like cells, into healthy BCMA-expressing tissues, with transcriptional reprogramming toward cytotoxicity and preferential CD4 trafficking driven by CSF chemokine gradients.
- Clinical Relevance: These findings identify a mechanistic basis for delayed BCMA CAR T-associated immune toxicities, implicating a hyperproliferative cytotoxic CD4 CAR T-cell population, and support risk stratification, monitoring, and mitigation strategies to improve the safety of BCMA CAR T-cell therapy in multiple myeloma.
Julia Han Noll, PhD, University of Pennsylvania, Philadelphia, Pennsylvania, presented mechanistic data identifying a distinct spectrum of delayed toxicities associated with BCMA-directed CAR T-cell therapy at the 2025 ASH Annual Meeting & Exposition.
In an analysis of 198 patients treated with ciltacabtagene autoleucel (cilta-cel) or idecabtagene vicleucel (ide-cel), these toxicities, most commonly enterocolitis and neurologic syndromes, were associated with infiltration of healthy tissues by cytotoxic CD4 CAR T cells, particularly among patients treated with ciltacabtagene autoleucel.
She concluded, “we've shown that these unique toxicities that are associated with BCMA CAR-T cell therapy for multiple myeloma are driven by a cytotoxic CD4 hyperproliferative population.”
Transcript:
My name is Julia Han Noll. I'm a recently graduated PhD student in the lab of Joseph Fraietta at the University of Pennsylvania. I'm at ASH today and I have presented an oral abstract that I submitted on a project that I worked on together with Dr Matthew Ho and Dr Luca Paruzzo, where we investigated BCMA-related CAR T-cell toxicities that are unique compared to classical toxicities.
The mechanism behind those unique delayed toxicities has been so far mechanistically not understood. In our project, we are the first to define mechanistic underpinnings of what goes into the development of these toxicities. We've defined them as CIRAE, so CAR-T immune related adverse events.
We looked at a cohort of 198 patients that were treated with commercial BCMA CAR T-cells, both the cilta-cel or the ide-cel product. About 14% of the patients developed unique toxicities. About a fifth of the cilta-cel cohort developed these toxicities, whereas there was less toxicities in the ide-cel cohort. The toxicities came in 4 different categories. There was rheumatologic, there were lung related toxicities, but the ones that we looked at in more detail in our project mechanistically were enterocolitis and neurological toxicities such as facial palsy or Parkinsonism. That's how they manifest in the clinic.
What we've also seen in the clinic are certain risk factors that are associated with developing CIRAEs, which is just the short form of the CAR-T associated immune related adverse events. The factors that are associated with the development of these toxicities are being treated with cilta-cel, having higher CD4 to CD8 ratio at apheresis, having a high absolute lymphocyte count at day 14, and having developed prior ICANS prior to these toxicities. However, the CRS and prior autoimmunity was not associated with an increased development of these risks.
We found through biopsies of gastrointestinal tissue that the CAR T-cells infiltrated healthy tissue and overlapped with BCMA-expressing healthy tissue, which is what we hypothesized to be directly killing the healthy tissue in the gut and infecting and causing the enterocolitis in these patients. Further immunohistochemistry also confirmed that these cells were primarily CD4s, whereas there were only a few CD8 cells that were present there.
Since we've seen that the CD4 T-cell presence is seen in these patients, we looked at flow cytometry data of immunophenotyping data of the CAR T-cells and found that the patients that developed CRAs had a more CD4 central memory like phenotype, whereas the patients without these toxicities had a CD8 TEMRA phenotype.
Furthermore, we also performed single cell RNA sequencing of CAR T-cells that were in the cerebral spinal fluid of patients with neurological toxicities and felt that not only were the CD4 T-cells more abundant, but also they had a reprogramming at the transcriptional state with enrichment of cytotoxicity pathways and upregulation of genes that are cytotoxics such as GranzymeK or granulysin. We found that the infiltration of the presence of the CAR T-cells industry with spinal fluid was driven by a unique chemokine profile of the CSF through transwell assays, where we saw that when you put in a 1-to-1 ratio of CD4s to CD8 T-cells in the upper chamber of a transwell and then put the CSF serum in the lower chamber, that that causes preferentially CD4 CAR T-cells to migrate through.
In summary, we've shown that these unique toxicities that are associated with BCMA CAR-T cell therapy for multiple myeloma are driven by a cytotoxic CD4 hyperproliferative population.
Source:
Han Noll J, Ho M, Paruzzo L, et al. Delayed neurotoxicity, enterocolitis, and BCMA-CART-associated immune-related adverse events (CirAE) are caused by CD4+ CAR T-cells. Dec 6-9, 2025; Orlando, FL. Abstract: 804
