Identification of Molecular Mechanisms Responsible for the Development of Extramedullary Disease in Myeloma and Potential Novel Therapeutic Targets Using Transcriptomic and Exome Profiling

Introduction: Multiple myeloma (MM) is the second most common blood cancer, and despite recent treatment advances, the majority of patients will ultimately die due to progression. Extramedullary disease (EMD) is a less frequent manifestation of MM often occurring during the course of disease where tumor plasma cells become independent of bone marrow microenvironment and infiltrate other tissues and organs. The development of EMD is always considered a high-risk feature with poor prognosis.

Aim: To identify molecular mechanisms responsible for the development of EMD using DNA and RNA sequencing and thus reveal potentially novel druggable targets. Currently, available anti-myeloma agents are not effective in patients with EMD. Thus it is critical to perform this kind of translational research that will ultimately lead to discovering new treatment strategies in this prognostically poor subset of MM patients.

Methods: We collected a unique set of FACS/MACS sorted aberrant plasma cells from 4 freshly obtained EMD samples from relapse and their respective cryopreserved bone marrow samples from diagnosis (available in 3 cases). In the 3 longitudinal ND and EMD samples we analyzed somatic mutations in the whole-exome sequencing data (Illumina, Sure select V6) using a combination of Mutect2 and Strelka2. To analyze differential gene expression, we used Deseq2 R package and Salmon for read mapping and quantification and Cytospace for the pathway enrichment analysis. Besides, to investigate the effect of cryopreservation on the gene expression in longitudinal samples from diagnosis, we compared all 4 EMD samples with 5 fresh unrelated BM samples from ND patients with PET/CT confirmed lack of EMD.

Results: Analysis of somatic mutations revealed an only partial overlap of mutated genes between ND and respective EMD sample (23% on average) and 1.49x increased number of variants in EMD. In our samples (ND and EMD), we identified 11 mutated myeloma drivers (Walker 2018) that mostly affected epigenetic processes and MAPK pathway, with the latter being mutated in all three patients in both stages. Interestingly, we identified two genes, both exclusively mutated only in 2/3 EMD samples and not mutated in any ND sample, DNAJC16 and HERC1. DNAJC16 is a member of heat shock proteins (Hsp40; Kampinga 2009). HERC1 is an E3 ubiquitin ligase that regulates p38 signaling and cell migration (Padrazza 2020), and thus, it represents a potentially interesting target for further analysis of EMD development.

Differential expression analysis of 3 paired ND-EMD samples revealed 131 deregulated transcripts with the TNFSF9 (CD137L) ligand, HECW1 ubiquitin ligase, and UNC13C gene being top upregulated genes. Among the top downregulated were immunoglobulin genes and S1PR4 signaling receptor, MPO myeloperoxidase, and SLAMF1 (CD150) signaling lymphocytic activation molecule. Analysis of a larger cohort of fresh unpaired 5 ND and 4 EMD samples resulted in more distinct clusters of ND/EMD samples and the total number of 673 deregulated genes (Fig. 1). EMD downregulated genes mostly belong to cell adhesion, chemokine signaling pathway, and neutrophil activation process. Upregulated genes belong to proliferation, cell cycle, and DNA damage signaling pathways. Evaluation of currently available druggable targets revealed that only LAG3 (lymphocyte activation 3 protein) is significantly downregulated in EMD. For three upregulated genes (FGFR3, NTRK3, EZH2), we found interaction with potential cancer inhibitors (Erdafitinib, Entrectinib, Tazemetostat). Notably, the latter agent targeting EZH2 methyltransferase gene has been recently approved by the FDA for follicular lymphoma treatment, and clinical trials for MM are ongoing.

Conclusion: Here, we report the preliminary analysis of the genomic and transcriptomic profile of EMD samples with their respective paired samples from the new diagnosis. The genomic alterations proved to be heterogeneous; though, we were able to identify 2 genes exclusively mutated in EMD samples. Moreover, deregulated transcriptome clearly separated ND and EMD stage in unrelated cohorts and pointed to rapid proliferation and increased DNA damage response pathways and pinpointed several potential drug targets. However, our findings warrant a more in-depth analysis on the bigger patient cohort.

Acknowledgement: Work was supported by project ENOCH (No. CZ.02.1.01/0.0/0.0/16_019/0000868)