Identification of Initiating Trunk Mutations and Distinct Molecular Subtypes: An Interim Analysis of the Mmrf Commpass Study

The Multiple Myeloma Research Foundation (MMRF) CoMMpass trial (NCT145429) is a longitudinal study of 1000 patients with newly-diagnosed multiple myeloma from clinical sites in the United States, Canada, Spain, and Italy. Each patient receives a treatment regimen containing a proteasome inhibitor, immunumodulatory agent, or both.  Clinical parameters are collected at study enrollment and every three months through the five-year observation period. To identify molecular determinants of clinical outcome each baseline and progression tumor specimen is characterized using Whole Genome Sequencing, Exome Sequencing, and RNA sequencing.  This will be the first public presentation of the interim analysis seven cohort with 760 enrolled patients of whom 565 are molecularly characterized.  This cohort of patients includes 14 patients with baseline and secondary samples along with 7 patients with characterized tumor samples from the bone marrow and peripheral blood.

Although the median follow-up time for the cohort is only 260 days the patients on proteasome and IMiD based combinations are currently showing a PFS and OS benefit compared to those receiving combinations with each agent alone. From the raw mutational analysis we identified 24 significant genes that are recurrently mutated and the mutated allele is detectably expressed in all but one, DNAH5.  Suggesting these mutations are likely contributing to myelomagenesis through an unconventional mechanism.  Interestingly, DIS3 mutations are independent of KRAS, NRAS, and BRAF indicating a potential mechanistic link while PRKD2 mutations are associated with t(4;14). To identify events driving the initiation of myeloma we performed a detailed clonality analysis using a bayesian clustering method that corrects for copy number abnormalities and tumor purity to assign mutations into distinct clonal branches versus the initiating trunk mutations. On average 63.8% of mutations are trunk mutations and in 86.7% of patients at least one trunk mutation is associated with somatic hypermutation of an immunoglobulin gene as expected in a late stage B-cell malignancy.  This identified many expressed trunk mutations that did not come out in the classic significance analysis like ATM, EGR1, and CCND1. To identify molecular subtypes we performed unsupervised clustering using a consensus clustering approach on independent discovery and validation cohorts, which identified 12 distinct subtypes, using a combination of silhouette score and cumulative distribution of consensus scores. This analysis identified two distinct groups associated with t(4;14) with mutations in FGFR3 and DIS3 being exclusive to one subgroup.  In addition, this analysis separates patients with cyclin D translocations into three different groups, with one group having the second lowest PFS proportion.  Three patients without CCND1 or CCND3 translocations were found to have IgH translocations targeting CCND2.  The MAF subgroup was associated with the lowest OS and PFS proportion, and the three MAF/MAFB translocation negative patients in the subgroup all had MAFA translocations. The remaining 6 subgroups are associated with hyperdiploid copy number profiles and harbor the majority of the IgH-MYC translocation events. Two of the hyperdiploid groups are associated with a low level of NFKB activation compared to the remaining four, one of these is defined by the highest proliferation index but paradoxically the other has the second worst OS proportion.  Another group is enriched with FAM46C and NRAS mutations. The genomic profiles of the paired tumors isolated from the peripheral blood and bone marrow are highly similar indicating these are not genetically distinct tumor compartments, at least in this subset of seven patients.  Applying our bayesian clustering method to the serial samples resolved additional clonal clusters as mutations with similar cancer cell fractions at diagnosis clearly diverged at later timepoints. These analyses have identified tumor initiating mutations and new subtypes of myeloma, which are associated with distinct molecular events and clinical outcomes.