Abstract: Introduction: Multiple myeloma (MM) is a cancer of the plasma cells in the bone marrow, and its clinical course depends on a complex interplay of clinical traits and molecular characteristics of the plasma cells. Since risk-adapted therapy is becoming standard of care, there is an urgent need for a precise risk stratification model to assist in therapeutic decision-making. While progress has been made, there remains a significant opportunity to improve patient stratification to optimize treatment and to develop new therapies for high-risk patients. To accelerate the development and evaluation of such risk models in MM, we formed a DREAM Challenge, a crowd-sourced competition that engages large cross-disciplinary teams of experts to address complex problems in biomedicine. Methods and Data: In collaboration with Multiple Myeloma Genome Project (MGP), clinical variables, patient outcomes, genetic, and gene expression data from thousands of samples were curated and harmonized from multiple public and private studies. In preparation for the challenge, a team of data scientists was assembled to evaluate this data, benchmark public high-risk models, and assess established prediction metrics with regard to progression-free survival (PFS) and overall survival (OS), the clinical endpoints of the challenge. Docker containers will be used to validate submitted models on private data that would otherwise not be available and to facilitate the transition of the best performing predictive signature to a clinical application. The MM DREAM challenge is accessible at: synapse.org. Results: The international staging system (ISS) for myeloma was used as a baseline classifier for high-risk patients (PFS & lt; 18mo). We evaluated published high-risk signatures - UAMS-5, UAMS-17, UAMS-70, and EMC92 - as benchmarks and observed that they consistently outperformed the baseline ISS predictor. High-risk prediction scores from these models were moderately correlated, suggesting published classifiers capture non-overlapping determinants of risk. Development of de novo classifiers by our team integrating clinical and molecular data highlighted opportunities for model refinement and supports rationalization of a crowd-sourced challenge to advance the field. Conclusion: Preliminary analysis of the challenge data suggests there is an opportunity to significantly improve risk stratification models in MM. In addition to the robust benchmarking of existing classifiers, we anticipate new, more accurate models will be proposed through a MM challenge given the scale of the combined data sets. We hope to uncover novel clinical and molecular traits that may yield insight into the pathology of MM and provide direction for follow-up studies. Importantly, this challenge will illustrate the advantages of leveraging public data and crowdsourcing to address therapeutically relevant questions in oncology. In addition, this challenge establishes a community resource for future research and benchmarking of novel classifiers. Citation Format: Michael Mason, Michael Amatangelo, Daniel Auclair, Doug Bassett, Hongyue Dai, Andrew Dervan, Erin Flynt, Hartmut Goldschmidt, Dirk Hose, Konstantinos Mavrommatis, Gareth Morgan, Nikhil Munshi, Alex Ratushny, Dan Rozelle, Mehmet Samur, Frank Schmitz, Ken Shain, Anjan Thakurta, Fadi Towfic, Matthew Trotter, Brian Walker, Brian S. White, Thomas Yu, Justin Guinney. Multiple Myeloma DREAM Challenge: A crowd-sourced challenge to improve identification of high-risk patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4725. doi:10.1158/1538-7445.AM2017-4725

Abstract: While the past decade has seen meaningful improvements in clinical outcomes for multiple myeloma patients, a subset of patients does not benefit from current therapeutics for unclear reasons. Many gene expression-based models of risk have been developed, but each model uses a different combination of genes and often involves assaying many genes making them difficult to implement. We organized the Multiple Myeloma DREAM Challenge, a crowdsourced effort to develop models of rapid progression in newly diagnosed myeloma patients and to benchmark these against previously published models. This effort lead to more robust predictors and found that incorporating specific demographic and clinical features improved gene expression-based models of high risk. Furthermore, post-challenge analysis identified a novel expression-based risk marker, PHF19 , which has recently been found to have an important biological role in multiple myeloma. Lastly, we show that a simple four feature predictor composed of age, ISS, and expression of PHF19 and MMSET performs similarly to more complex models with many more gene expression features included.

Abstract: Multiple myeloma (MM) is a hematological malignancy of terminally differentiated plasma cells residing within the bone marrow with 25,000-30,000 patients diagnosed in the United States each year. The disease's clinical course depends on a complex interplay chromosomal abnormalities and mutations within plasma cells and patient socio-demographic factors. Novel treatments extended the time to disease progression and overall survival for the majority of patients. However, a subset of 15%-20% of MM patients exhibit an aggressive disease course with rapid disease progression and poor overall survival regardless of treatment. Accurately predicting which patients are at high-risk is critical to designing studies with a better understanding of myeloma progression and enabling the discovery of novel therapeutics that extend the progression free period of these patients. To date, most MM risk models use patient demographic data, clinical laboratory results and cytogenetic assays to predict clinical outcome. High-risk associated cytogenetic alterations include deletion of 17p or gain of 1q as well as t(14;16), t(14;20), and most commonly t(4,14), which leads to juxtaposition of MMSET with the immunoglobulin heavy chain locus promoter, resulting in overexpression of the MMSET oncogene. While cytogenetic assays, in particular fluorescence in situ hybridization (FISH), are widely available, their risk prediction is sub-optimal and recently developed gene expression based classifiers predict more accurately rapid progression. To investigate possible improvements to models of myeloma risk, we organized the Multiple Myeloma DREAM Challenge, focusing on predicting high-risk, defined as disease progression or death prior to 18 months from diagnosis. This effort combined 4 discovery datasets providing participants with clinical, cytogenetic, demographic and gene expression data to facilitate model development while retaining 4 additional datasets, whose clinical outcome was not publicly available, in order to benchmark submitted models. This crowd-sourced effort resulted in the unbiased assessment of 171 predictive algorithms on the validation dataset (N = 823 unique patient samples). Analysis of top performing methods identified high expression of PHF19, a histone methyltransferase, as the gene most strongly associated with disease progression, showing greater predictive power than the expression level of the putative high-risk gene MMSET. We show that a simple 4 feature model composed of age, stage and the gene expression of PHF19 and MMSET is as accurate as much larger published models composed of over 50 genes combined with ISS and age. Results from this work suggest that combination of gene expression and clinical data increases accuracy of high risk models which would improve patient selection in the clinic. Disclosures Towfic: Celgene Corporation: Employment, Equity Ownership. Dalton:MILLENNIUM PHARMACEUTICALS, INC.: Honoraria. Goldschmidt:Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; John-Hopkins University: Research Funding; Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Mundipharma: Research Funding; Amgen: Consultancy, Research Funding; Chugai: Honoraria, Research Funding; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Molecular Partners: Research Funding; MSD: Research Funding; Sanofi: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Takeda: Membership on an entity's Board of Directors or advisory committees, Research Funding; Novartis: Membership on an entity's Board of Directors or advisory committees, Research Funding; Adaptive Biotechnology: Membership on an entity's Board of Directors or advisory committees; Janssen: Consultancy, Research Funding; Dietmar-Hopp-Stiftung: Research Funding; John-Hopkins University: Research Funding. Avet-Loiseau:takeda: Consultancy, Other: travel fees, lecture fees, Research Funding; celgene: Consultancy, Other: travel fees, lecture fees, Research Funding. Ortiz:Celgene Corporation: Employment, Equity Ownership. Trotter:Celgene Corporation: Employment, Equity Ownership. Dervan:Celgene: Employment. Flynt:Celgene Corporation: Employment, Equity Ownership. Dai:M2Gen: Employment. Bassett:Celgene: Employment, Equity Ownership. Sonneveld:SkylineDx: Research Funding; Takeda: Honoraria, Research Funding; Karyopharm: Honoraria, Research Funding; Janssen: Honoraria, Research Funding; Celgene: Honoraria, Research Funding; BMS: Honoraria; Amgen: Honoraria, Research Funding. Shain:Amgen: Membership on an entity's Board of Directors or advisory committees; Bristol-Myers Squibb: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees; Janssen: Membership on an entity's Board of Directors or advisory committees; AbbVie: Research Funding; Takeda: Membership on an entity's Board of Directors or advisory committees; Sanofi Genzyme: Membership on an entity's Board of Directors or advisory committees; Adaptive Biotechnologies: Consultancy. Munshi:Abbvie: Consultancy; Takeda: Consultancy; Oncopep: Consultancy; Celgene: Consultancy; Adaptive: Consultancy; Amgen: Consultancy; Janssen: Consultancy. Morgan:Bristol-Myers Squibb, Celgene Corporation, Takeda: Consultancy, Honoraria; Celgene Corporation, Janssen: Research Funding; Amgen, Janssen, Takeda, Celgene Corporation: Other: Travel expenses. Walker:Celgene: Research Funding. Thakurta:Celgene: Employment, Equity Ownership.

Abstract: Introduction: Chromothripsis and chromoplexy are gross structural events that deregulate multiple genes simultaneously and may help explain rapid changes in clinical behavior. Previous screening studies in multiple myeloma (MM) using copy number arrays have identified chromothripsis at a low frequency (1.3%) and suggested it adversely impacts prognosis. Here, using whole genome sequencing (WGS) data we have identified a higher frequency of these events, suggesting they are more common than previously thought. Methods: 10X ChromiumWGS (10XWGS) from 76 newly diagnosed MM (NDMM) patients were analyzed for structural rearrangements using Longranger. Oxford Nanopore long read sequencing was performed on 2 samples. Long insert WGS data from 813 NDMM patient samples from the Myeloma Genome Project (MGP) were analyzed for structural rearrangements using Manta. Whole exome sequencing was available for 712 samples. RNA-seq was available for 643 samples. Chromothripsis was determined by manual curation of breakpoint and copy number data. Chromoplexy was defined as rearrangements within 1 Mb of one another involving 3 or more chromosomes. Results: Chromoplexy was detected in 33/76 (46%) cases using 10XWGS data, and cross validated in the MGP WGS dataset being found in 30% (247/813) of samples and was most frequent on chromosomes 8 (11.7% of samples), 14 (10.6%), 11 (9.6%), 1 (9.5%), 6 (8.0%), 22 (7.6%), 12 (6.7%), and 17 (6.7%). The gene regions most involved in chromoplexy events were MYC (chr8; 7.3%), IGH (chr14, 8.8%), IGL (chr22; 4.6%), CCND1 (chr11; 3.9%), TXNDC5 (chr6; 1.7%), FCHSD2 (chr11; 1.4%), FAM46C (chr1; 1.2%), MMSET (chr4; 1.2%), and MAP3K14 (chr17; 0.7%). Chromoplexy samples involved pairings of super-enhancer donors (IGH, IGL, FAM46C, TXNDC5) and oncogenic receptors (CCND1, MMSET, MAP3K14, MYC) implicating transcriptional deregulation. To confirm, RNASeq showed an elevation of expression over median in the oncogenic receptors when paired with a donor: CCND1 (median expression = 12.0 vs. median expression with donor = 17.9), MAP3K14 (10.8 vs. 14.7), MYC (12.7 vs. 14.1) and MMSET (11.9 vs. 16.7). We also identified elevated expression of PAX5 (8.23 vs. 13.79) and two cases where BCL2 (13.32 vs. 14.68) partnered with MYC, one involved IGH similar to follicular lymphoma. To determine if chromoplexy events were happening on the same allele, we performed long read sequencing using Oxford Nanopore on a sample with a t(2;6;8;11) event. We observed a read mapped to chromosome 2, with secondary alignment to chromosomes 6 and 8. This single 32 kb read was a continuous t(2;6;8) event, proving these events occurred on the same allele. However, despite close proximity, the data did not put the t(8;11) in the same read meaning this event occurred on a different allele or sub-clone, suggesting ongoing genomic instability. Chromothripsis was detected in 16/76 (21%) cases using 10XWGS, and was consistent in MGP data, (170/813; 21%). Chromothripsis occurred on all chromosomes but at different frequencies where chromosome 1 had most events (5.1%), followed by 14 (2.4%), 11 (2.3%), 12 (2.2%), 20 (1.9%), 17 (1.9%), and 8 (1.9%). We hypothesized the presence of both chromoplexy and chromothripsis could be associated with ineffective DNA repair and indeed, using WES data, patients with both events show more mutations in TP53 (19% vs. 5%) and ATM (10% vs. 4%) implicating homologous recombination deficiency as an etiologic mechanism. Gene set enrichment analysis showed significant enrichment and positive normalized enrichment score (NES) for the DNA Repair (P = 0.01; NES = 1.7) and MYC pathways (P = 0.01; NES = 3.2) consistent with previous results. In relation to prognosis, chromoplexy and chromothripsis have a negative impact on progression free survival (28.6 months vs. 42.8 months, P=0.03 and 28.6 months vs. 40.7 months P=0.01, respectively). When patients with both chromoplexy and chromothripsis (9%) were examined there was a pronounced effect on PFS (40.7 months vs. 22.7 months, P 〈 0.001). Conclusion: Complex structural events are seen frequently in MM and could help explain disease progression. Severe cases with both chromoplexy and chromothripsis are associated with acquired genomic instability and an adverse impact on prognosis either directly or due to their association with DNA repair abnormalities. This opens the possibility of specifically therapeutically targeting the underlying DNA abnormalities. Disclosures Flynt: Celgene Corporation: Employment, Equity Ownership. Ortiz:Celgene Research SL (Spain), part of Celgene Corporation: Employment, Equity Ownership. Dervan:Celgene Corporation: Employment, Equity Ownership. Gockley:Celgene Corporation: Employment. Davies:Janssen: Consultancy, Honoraria; TRM Oncology: Honoraria; Abbvie: Consultancy; Celgene: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees; ASH: Honoraria; Takeda: Consultancy, Membership on an entity's Board of Directors or advisory committees; MMRF: Honoraria; Amgen: Consultancy, Membership on an entity's Board of Directors or advisory committees. Thakurta:Celgene Corporation: Employment, Equity Ownership. Morgan:Celgene: Consultancy, Honoraria, Research Funding; Takeda: Consultancy, Honoraria; Bristol-Myers Squibb: Consultancy, Honoraria; Janssen: Research Funding.