Landscape of Driver Lesions in Multiple Myeloma and Consequences for Targeted Drug Response

Introduction Multiple myeloma (MM) is a heterogeneous disease that eventually becomes resistant to therapy. Determining the genomic lesions driving each stage of the tumor and identifying actionable items for novel targeted drugs will improve and increase therapeutic options for the malignancy. The aim of the present work is to obtain a comprehensive catalog of driver genomic lesions for both newly diagnosed (NDMM) and refractory/relapsed MM (RRMM) patients by integrating multiple genomic data and linking these to the action of targeted therapeutic approaches. Methods Molecular cytogenetics was assessed by fluorescence in situ hybridization and somatic mutations and copy number changes were identified by performing exome sequencing of DNA from CD138+ cell and skin paired samples collected from 30 MM patients (NDMM n=12; RRMM n=18). In addition, gene expression profiles were obtained by transcriptome sequencing. The proportion of tumor clones bearing a specific mutation was inferred from the variant allele frequency. Genetic alterations involved in the tumorigenesis of each patient (named drivers) were identified by combining an in silico method aimed to score their potential for being malignant with the a priori knowledge retrieved from the identification of complementary signals of positive selection in available tumor cohorts (Tamborero et al. Nat Sci Rep 2013). Selective drug response was assessed by testing the ex vivo sensitivity of patient derived CD138+ cells to 306 oncology drugs and comparing results with responses derived from healthy bone marrow control cells. Results Overall, 0.5 translocations, 3±2.8 mutations and 4.9±2.7 copy number changes per patient were identified as putative drivers. The total number of driver alterations did not differ between NDMM and RRMM samples, and no gene reached statistical significance for being more frequently altered in the latter group. However, the only mutations in RAS genes that appeared at sub-clonal proportions occurred in diagnosed samples, pointing out their positive selection among relapsed patients in which they were present in all clones. Translocations involving IGH@ were observed in 11 (37%) patients, and interestingly 3 other samples exhibited driver alterations in the oncogenes involved in these fusions (i.e. activating mutations in FGFR3 or gene amplification plus peaked overexpression of WHSC1 and CCND1). Recurrent alterations were observed among genes previously associated with MM, including DIS3 (n=15), KRAS (n=11), CYLD (n=8), TRAF3 (n=6) and FAM46C (n=5). Other genes not previously associated with or less-known to be involved in MM pathogenesis were also identified, including the histone methyltransferase MLL, the tumor necrosis factor associated genes FAF1 and TNFRSF13B, the p53-suppressing protein phosphatase PPM1D, and several genes related with blood cell differentiation and B-lymphocyte development (e.g. SOX7, BLK and PRDM1). Overall, the pathways most frequently targeted by driver alterations were MAPK (23 (77%) samples, mostly by mutations), NF-κB (17(57%) samples, mostly by gene copy loss), cell-cycle (18 (60%) samples), and RNA-processing (17 (57%) samples). Comparison of driver lesions to drug response using data derived from ex vivo testing of the same patient samples to different targeted small molecule inhibitors (e.g. PI3K/mTOR and MEK inhibitors) indicated that alterations affecting PI3K and p53 pathways were associated with increased drug sensitivity, while alterations involving activation of FGFR3 and copy loss of TRAF3 were associated with a more resistant phenotype. Conclusions The integration of multiple genomic data by combining different predictive computational tools can comprehensively identify cancer events in individual patients. Applying these tools to genomic data from MM patients identified both known and novel driver lesions, and some of these alterations were associated with the ex vivo response to selective drugs. However, further data is required to confirm biomarkers of response to those novel therapeutics and test potential benefits in MM patients. Disclosures Silvennoinen: Janssen, Sanofi, Celgene: Honoraria; Research Funding of Government Finland, Research Funding from Janssen and Celgene: Research Funding. Porkka:Novartis: Honoraria, Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding. Heckman:Celgene: Research Funding.