Clonal Hematopoiesis of Indeterminate Potential Is Associated with Decreased Inflammatory Toxicity and Increased Late Cytopenia in Patients with Multiple Myeloma and Non-Hodgkin Lymphoma Treated with Chimeric Antigen Receptor T-Cell Therapy

Background: Chimeric antigen receptor T-cell therapy (CAR-T) is a highly effective therapy in non-Hodgkin lymphoma (NHL) and multiple myeloma (MM). However, the use of CAR-T can be limited by toxicities, including the hyperinflammatory processes of cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), as well as prolonged cytopenia. Clonal hematopoiesis of indeterminate potential (CHIP) is the precursor state in which hematopoietic cells contain leukemogenic driver mutations. CHIP has been shown to impact inflammatory processes and lead to poor outcomes in patients (pts) with NHL and MM undergoing autologous stem cell transplant (ASCT). We aimed to evaluate the prevalence and impact of CHIP on toxicity, response, and outcomes in a cohort of pts receiving commercial CAR-T for MM and NHL.

Methods: All pts who received CAR-T for NHL and MM with blood samples banked prior to CAR-T were included. To identify CHIP, deep next generation sequencing was performed using VariantPlex Core Myeloid panel (Archer) with supplementary modules (for GNAS and PPM1D). Sequencing data were processed using Archer Analysis Software (version 7.3). Alignment to hg19/GCRh37 was performed using a combination of bowtie2 and BWA-MEM; variant analysis was done using LoFreq, Freebayes, and Vision. To limit results to somatic changes, mutations with VAF of 0.45-0.55 and greater than 0.95 were excluded. Only variants previously classified as pathogenic driver mutations were included. Clinical data were extracted, including demographics, response, CRS grade, ICANS grade, cytopenia grade, and any diagnosis of myeloid neoplasm following CAR-T. Toxicity was predefined as CRS ≥2, CRS grade 1 with tocilizumab administration, ICANS ≥2, and/or ICANS grade 1 with steroid administration. The proportion of pts with CHIP was calculated along with the 95% confidence interval using the Exact (Clopper-Pearson) method. Univariable logistic regression was used to study associations between CHIP and the following outcomes: toxicity, prolonged cytopenia (grade ≥1), and response at 30 days post CAR-T infusion.

Results: 138 pts received CAR-T at our center and were evaluated for CHIP from 4/2019 to 5/2024. With median age of 65 (range, 21 to 85), 27% with prior ASCT, and 61.6% male, 28.3% (95% CI: 20.9%-36.5%) harbored CHIP with median VAF 0.062 (range, 0.027-0.80). 52 unique CHIP mutations were found in 12 genes, with DNMT3A (47.2%) the most common, then TET2 (10.9%) and TP53 (9.0%). Of pts with CHIP, all (N=39) had at least 1 CHIP clone, with 25.6% (N=10) having 2 clones and 7.7% (N=3) with 3 clones. Pts with CHIP were older than those without (median age 68 versus 64, P=0.02). Among MM pts (N=82), 32.9% (95% CI: 22.9%-44.2%) had CHIP, and among NHL pts (N=56), 21.4% (95% CI: 11.6%-34.4%) had CHIP. The prevalence of CHIP was similar between the two groups (P=0.1407). Median follow-up was 16.8 months.

Pts receiving CAR-T outside of a clinical trial were considered evaluable for our analysis of CAR-T toxicity and efficacy. Among 107 evaluable pts, 28.0% had CHIP (21.2% in NHL; 34.6% in MM). Of pts with CHIP, 53.3% experienced toxicity, compared to 76.6% of those without CHIP (P=0.02). 54.5% of NHL pts and 52.6% of MM pts with CHIP experienced toxicity. Compared to pts without CHIP, those with CHIP had higher odds of experiencing cytopenia beyond 90 days post CAR-T infusion (OR=3.32; 95% CI 1.10-10.07; P=0.034). No difference in occurrence of cytopenia was observed between pts with CHIP and pts without at 30-90 days post CAR-T. Two pts developed t-MN following commercial CAR-T, one of whom had pre-treatment CHIP. No significant difference was observed in clinical response at 30 days for pts with CHIP versus without.

Conclusions: In this commercial CAR-T recipient cohort, inflammatory toxicities (CRS and ICANS) occurred less frequently in pts with CHIP. However, CHIP was significantly associated with cytopenia >90 days post infusion. With limited follow-up, incidence of t-MN was low. Compared to prior published CHIP in CAR-T experience, a larger number (and proportion) of pts with MM are represented here. Study of a larger cohort of CAR-T pts with longer follow-up is ongoing and necessary to validate our findings, better understand toxicity differences compared to prior published results, and determine whether CHIP is associated with t-MN in the CAR-T population.