Timing antigenic escape in multiple myeloma treated with T-cell redirecting immunotherapies

Abstract

Recent data highlight genomic events driving antigen escape as a recurring cause of chimeric antigen receptor T-cell (CAR-T) and bispecific T-cell engager (TCE) resistance in multiple myeloma (MM). Yet, it remains unclear if these events, leading to clonal dominance at progression, result from acquisition under treatment selection or selection of pre-existing undetectable clones. This differentiation gains importance as these immunotherapies progress to earlier lines of treatment, prompting the need for innovative diagnostic testing to detect these events early on. By reconstructing phylogenetic trees and exploring chemotherapy mutational signatures as temporal barcodes in 11 relapsed refractory MM patients with available whole genome sequencing data before and after CART/TCE treatment, we demonstrated that somatic antigen escape mechanisms for BCMA- and GPRC5D-targeting therapies are acquired post-diagnosis, likely during CART/TCE treatment. Longitudinal tracking of these mutations using digital PCR in 4 patients consistently showed that genomic events promoting antigen escape were not detectable during the initial months of therapy but began to emerge nearly 1 year post therapy initiation. This finding reduces the necessity for a diagnostic panel to identify these events before CART/TCE. Instead, it underscores the importance of surveillance and identifying patients at higher risk of acquiring these events.

Keywords: BCMA; GPRC5D; bispecific T cell engager; chimeric antigen receptor T-cell; genomics; immunotherapy; multiple myeloma.

Figure 1:. Timing antigen escape mechanisms in RRMM treated with CART/TCE.

Selected cases for timing the acquisition of TCE/CART antigen escape mutations. Clusters in the phylogenetic tree were reconstructed using DPClust. Branch length is proportional to the number of single base substitution (SBS) in each cluster. Antigen escape variants are color-coded according to the cluster in which they are found. Mutational signatures analyses were run on clusters with >50 SBS. For example, Cluster #4B in MM-17 had 7 mutations and so mutational signature fitting was not performed. In MM-31 Cluster #0 grouped all the subclonal variants detected at relapse after TCE anti-GPRC5D. Presence of convergent evolution with 4 different and independent subclones was previously demonstrated by phasing reads. This is why the cluster appeared split in the three but is reported as one for the SBS analysis. PFS, Progression-Free Survival.

Figure 2:. Tracking onset of mutations on TNFRSF17 over time in RRMM treated with CART/TCE.

ddPCR reveals the emergence of TNFRSF17 mutants at later time points post TCE therapy initiation. Plots depict the serial measurements of the serum M-protein by serum electrophoresis or free light chain studies (solid black line), the percent of the bone marrow plasma cells (BMPC %, red bars), the donut plots indicate TNFRSF17 WT and mutants VAF and the grey zone reflects the time of biochemical or clinical relapse.