Subclonal genomic heterogeneity shapes immune crosstalk and clinical outcomes in multiple myeloma

Abstract Introduction: Subclonal diversity in multiple myeloma (MM)—defined by genetically distinct populations of malignant plasma cells—plays a pivotal role in disease progression, therapeutic resistance, and relapse. Characterizing subclone-specific transcriptomic profiles, copy number alterations, somatic variants, and their interactions with the immune microenvironment is critical for understanding tumor heterogeneity and identifying novel therapeutic targets. In this study, we analyzed smoldering multiple myeloma (SMM) and newly diagnosed multiple myeloma (NDMM) samples to explore the genomic landscape of subclones and their crosstalk with immune cells. Methods: CD138+ scRNA-seq and scATAC-seq multiomic data from SMM patients (n=10) and NDMM patients (n=21) were obtained from the published study dbGAP phs003220 (PMC11099140). Matched CD138- scRNA-seq data of immune cells were sourced from Sudha et al., IMS 2023. Subclones for each tumor sample were identified using inferCNV (v1.8.1) and a custom-built integration pipeline that utilized both scRNA-seq and scATAC-seq data. Cell-cell communications between subclones and immune cell types were identified using CellChat (v1.6.0) using the ligand-receptor interactions from CellChatDB. Copy Number Abnormality (CNA) profiles of subclones from single cell were compared with whole genome sequencing (WGS) data (phs003220) and similarity scores were computed. Due to differing technologies (RNA-seq vs WGS) and sequencing depth (single cell vs. bulk), this score reflects how well single-cell multiomics captures major chromosomal changes seen in WGS. Subclone-to-immune cell interactions were quantified to identify statistically significant chromosomal abnormalities that change their prevalence. Survival analysis was performed on the most common MM abnormalities to identify if subclonal co-occurrence influences prognosis. For each subclone, BAM files were generated, and somatic variants (SNVs + Indels) were called using Strelka2 with WGS control samples from the same patient. Results: We identified 23 subclones in SMM with a mean of 2.30 (min= 1; max=4); and 79 subclones in NDMM with a mean of 3.59 (min=1; max=8). Subclone and WGS CNAs were comparable to one another with mean Jaccard index (JI) of 0.57 for SMM and 0.69 for NDMM. Samples with low JI tended to have more subclones and more complex CNAs as measured through ploidy (correlation coefficient (r) = -0.069 in SMM and -0.177 in NDMM). NDMM subclones tended to have more CNAs than in SMM. From the IU cohort, Del13 was observed in 60% of the SMM and 45% of the NDMM patients. Gain1q was identified in 50% of the SMM and NDMM patients. Patients with co-occurring Gain1q and Del13q in the same subclone had significantly shorter progression-free survival (PFS) than patients with at least one of these alterations (median PFS 285 days vs 1194 days; P = 0.004). Subclones with Gain1q had a lower average percentage of interactions with CD8T cells (Gain1q=25.435, Norm1q=33.26; P=0.005). In NDMM subclones with Gain1q, there was a lower percentage of MHC-I pathway interactions with CD8T cells (P=0.001) and MHC-II pathway interactions with CD4T cells (P=0.001). Subclones with Del13q showed a higher percentage of MIF pathway (MIF-CD74/CXCR4) interactions in NDMM. SMM subclones with Del13p tended to have a lower percentage of MHC-I interactions in CD8T, CD4T and NK cells. Gain6p subclones had a higher percentage of MK pathway (MDK-NCL/ITGA4+ITGB1) interactions in SMM subclones. Somatic variants were identified for each subclone from the scATAC-seq. When compared with WGS somatic variants in NDMM, a median of 253 variants was found both subclones and WGS while 2752 were unique to subclones. Numerous pathogenic somatic variants were identified in the myeloma driver genes from subclones.Conclusions: The identification of subclone-specific immune interaction networks provides new opportunities for targeted therapeutic intervention. These results establish subclonal CNA profiling as both a prognostic tool and guide for precision immunotherapy in multiple myeloma, particularly for high-risk disease characterized by specific genomic alterations co-occurring in subclones. This study underscores the importance of single-cell analysis to fully capture the clinically relevant genomic heterogeneity that drives disease progression and treatment resistance.