(PA-042) Peak Absolute Lymphocyte Count as a Biomarker for CAR T-cell Expansion Is Associated with Non-ICANS Neurotoxicities Following Ciltacabtagene Autoleuecel

Ciltacabtagene autoleucel (cilta-cel) is an effective, FDA approved chimeric antigen receptor (CAR) T-cell therapy for relapsed/refractory multiple myeloma (RRMM). Cilta-cel is associated with delayed neurotoxicities which are distinct from immune effector cell-associated neurotoxicity syndrome (ICANS) observed following CAR T-cell infusion. These non-ICANS neurotoxicities (NIN) include but are not limited to cranial nerve palsies (CNP), movement and neurocognitive toxicities (MNT), Guillain-Barré syndrome (GBS), peripheral neuropathy (PN), and altered cognition or personality. Predisposing risk factors for NIN are unknown though peak absolute lymphocyte count (pALC) after infusion may be associated with developing NIN.  

Methods: Retrospective review of patients with RRMM treated with commercial, standard of care (SOC) cilta-cel at a single-center between July 21, 2022 and October 31, 2024. Patient baseline characteristics, outcomes, and NINs were collected. NINs included CNP, MNT, GBS, PN, cognitive changes, or other neurologic symptoms occurring after CAR T-cell infusion, judged distinct from ICANS by the treating clinician and consulting neurologist.  

Results: In total, 109 patients treated with SOC cilta-cel were included. At a median follow-up of 13 months, 12 (11.0%) patients had a NIN, including 5 patients with ≥2 NIN. Overall, 21 NINs were reported, including 13 CNPs (10 CNVII, 2 CNVI, 1 CNIII), 2 MNTs, 2 GBSs, 2 delayed/prolonged ICANS, 1 PN, and 1 patient with major personality change. Median time to NIN was 21 days for patients who had a NIN. A pALC of 3.2×103/µL was determined, using maximally selected Wilcoxon rank statistics, as a significant risk factor for developing NIN. In a univariable analysis, baseline characteristics, including age, sex, race, tumor burden, high-risk cytogenetics, presence of clonal hematopoiesis, steroid exposure, and estimated cumulative fludarabine exposure, were not associated with an increased risk of NIN. In both univariable and multivariable analyses, pALC>3.2×103/µL was associated with a significantly increased risk of NIN (p < 0.0001). Median time to pALC was 12 days after CAR T-cell infusion. Of patients with pALC≤3.2×103/µL, 0/77 (0.0%) developed NIN compared to 12/32 (37.5%) patients with pALC>3.2×103/µL. A pALC>3.2 vs ≤3.2×103/µL was not associated with significant differences in progression-free survival (PFS) or overall survival (OS).   

Conclusions: NINs occurred in 11% of patients treated with cilta-cel with a median symptom onset of 21 days. Peak ALC was associated with NIN and pALC>3.2×103/µL may identify patients at risk of developing NIN. Peak ALC>3.2×103/µL was not associated with longer PFS or OS. Thus, interventions to maintain pALC≤3.2×103/µL may mitigate the risk of NIN without negatively impacting treatment outcomes. T-cell phenotyping analysis via flow cytometry and single-cell transcriptomics from patient peripheral blood samples is ongoing and will be presented at the meeting.