ALBERT

Description: Multiple myeloma (MM) is an incurable hematological malignancy characterized by the clonal proliferation of malignant plasma cells in the bone marrow. Like other cancers, MM is a genetically complex and heterogeneous disease. One of its distinctive characteristics is that it is preceded by a pre-malignant condition known as monoclonal gammopathy of undetermined significance (MGUS), which then progresses to asymptomatic (smoldering) multiple myeloma (SMM) and, ultimately, to late-stage MM. Its progression through these stages is determined by a sequence of genomic aberrations, starting with germline events that predispose to the disease, followed by early initiating events and the later acquisition of mutations that contribute to disease progression. Although considerable progress has been made in the past 6 years in cataloguing somatic events underlying MM development and progression, little is known about its genetic predisposition. Therefore, large-scale germline genomic variant studies are urgently needed. Recently, our group has published the largest-scale pan-cancer study of 〉10K adult and 〉1K pediatric cases that revealed new insights on germline predisposition variants across 33 cancer types (853 pathogenic or likely pathogenic variants) (Huang et al., 2018). Here, we aim to apply a similar strategy to MM cases. The CoMMpass study, promoted by MMRF (Multiple Myeloma Research Foundation) is a longitudinal, prospective observational study involving the collection and analysis of sequencing and clinical data from 〉1K MM patients at diagnosis and relapse. We performed germline variant calling on 808 normal samples from this dataset using GenomeVIP (https://github.com/ding-lab/GenomeVIP), which integrates multiple tools: VarScan2 and Genome Analysis ToolKit (GATK) for the identification of single nucleotide variants (SNVs) and indels; and Pindel for indel prediction. Variants were limited to coding regions of full length transcripts obtained from Ensembl release 70 plus the additional two base pairs flanking each exon that cover splice donor/acceptor sites. SNVs were based on the union of raw GATK and VarScan calls. Indels were required to be called by at least two out of the three callers (GATK, Pindel, VarScan). Variant calls from all tools were merged, filtered (allelic depth ≥ 5 for the alternative allele; rare variants with allele frequency ≤ 0.01 in 1000 Genomes and ExAC), and annotated using Variant Effect Predictor (VEP), resulting in an average of 1,653 variants per sample. Further, we applied CharGer (Characterization of Germline Variants, https://github.com/ding-lab/CharGer) to classify the identified germline variants as pathogenic, likely pathogenic, and prioritized variants of unknown significance (VUS). CharGer is an automatic variant classification pipeline developed by our group which adopts ACMG-AMP guidelines specifically for rare variants in cancer. Here, we were able to classify a total of 635 germline variants as pathogenic and 150 as likely pathogenic, affecting 90% of samples. Among pathogenic variants, 28 were found in known cancer predisposition genes including BRCA1 and BRCA2 - which have been previously associated with MM risk - BRIP1, CHEK2, TP53, TERT, and PMS2. Ongoing analyses include: functional characterization of these variants, identifying genes with enriched pathogenic or likely pathogenic variants in our dataset; investigation of LOH and two-hit (biallelic) events; gene and protein expression analyses in carriers of pathogenic germline variants of the respective gene; scanning for rare, germline copy number variations (CNVs); and identification of variants in post-translational modification sites that may affect protein signaling. Additionally, we are currently working on improving our CharGer tool by integrating new tumor associated data, such as DNA-Seq, RNA-Seq, Methyl-Seq and MS proteomics data, to improve variant classification. The preliminary results and analysis strategies described here will allow for efficient and cost-effective discovery of genetic changes relevant to MM etiology. Ultimately, we hope this work will impact our overall understanding of the genetics underlying MM predisposition, allowing for the development of better prevention and early detection strategies. Disclosures Vij: Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Takeda: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Karyopharma: Honoraria, Membership on an entity's Board of Directors or advisory committees; Jazz Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees; Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Jansson: Honoraria, Membership on an entity's Board of Directors or advisory committees.

Description: The Multiple Myeloma Research Foundation (MMRF) CoMMpass trial is the keystone program working towards personalized medicine in multiple myeloma (MM). CoMMpass has characterized over 730 samples from 669 patients, a subset having time-course data, using multiple platforms and has made these data publicly available. Screening for druggable somatic mutations and gene expression outliers revealed great potential for repurposing existing pharmacotherapies. We have developed and curated a Database of Evidence for Precision Oncology (DEPO) that includes 442 mutations and 49 genes with expression changes implicated as drug targets from 34 cancer types. We found 43.6% of CoMMpass samples have one of 26 somatic mutations that could be targeted by 9 different drugs, suggesting many drugs may be repurposed for use in MM. HotSpot3D, a protein-structure-guided analysis tool, showed 3.3% (24/730) of samples have mutations involving BRAF and KRAS, clustering with known drug targets, and another 2.1%(15/730) have mutations in clusters formed from a prior TCGA pan-cancer analysis involving 22 cancer types. This indicates additional potential drug targets and functional mutations in MM. Interestingly, 11 samples (including relapse), have subclonal mutations in both KRAS and BRAF with variable allelic fractions, implicating different treatment requirements for tumor subpopulations and disease stages. Additionally, druggable gene expression outlier analysis of 591 samples reveals an average of nearly 5 outlier genes per sample from among 49 known target genes, with 18.1% (107/591) of samples with MYCN gene outliers as compared to MYC 1.4% (8/591), despite MYC being a known driver of MM. Other top outliers are DLL3 10.2%, EGFR 10.5%, FGFR1 10.2%, and FGFR2 12.5%. Our study highlights candidate drug targets previously omitted in MM. Taken together, it suggests that heterogeneity analysis and pharmaceutical repurposing could lead to substantially improved outcomes. Disclosures No relevant conflicts of interest to declare.