Oncogene Activation in the Germinal Center is the Primary Determinant of MM Disease Phenotype
Multiple myeloma (MM) is a tumor of mature isotype-switched plasma cells that accumulate in the bone marrow causing anemia, hypercalcemia and bone lesions. MM is a heterogeneous disease. Both a supervised (TC) and an unsupervised analysis of gene expression data have yielded complementary results. These studies definitively show that the primary determinants of this heterogeneity are underlying genetic abnormalities: recurrent immunoglobulin gene translocations and hyperdiploidy. These genetic events identify five homogeneous groups of patients that can be identified today using FISH: MM with t(4;14), FGFR3/MMSET 14%, MM with t(14;16) and variants, c-maf 5%, mafB 2%, mafA 1%, MM with t(11;14) and variants, cyclin D1 16%, cyclin D3 2%, MM with hyperdiploidy, trisomies of chromosomes 3,5,7,9,11,15,19,21 40%, MM not otherwise classified 20%. These are thought to be primary genetic events, with the translocations mediated by germinal center reactions (switch recombination and somatic hypermutation) that each initiate a characteristic cascade of secondary genetic, epigenetic and microenvironmental changes leading to the unique gene expression profile and clinical course. A specific genetic lesion underlying the hyperdiploidy has not been identified, although molecularly it has been associated with a low-level of ectopic, bi-allelic expression of cyclin D1 (TC D1). Interestingly, a number of genetic events that are important in the pathogenesis and prognosis of MM do not have a dominant effect on the global gene expression profile. These include activating mutations of ras, amplifications of 1q, deletions of 13q and deletions of p53. Each of these presumably secondary events is seen (with different frequency) in each of the primary genetic groups above, and represent events that modify the course, but do not define unique diseases. Recently we have identified a promiscuous array of mutations that result in constitutive activation of (primarily) the non-canonical NFKB pathway, present in up to 20% of patients. The most common mutation is inactivation of TRAF3, present in 13% of patients. Preliminary analysis using gene expression as a surrogate for inactivating mutation suggests that this may have important clinical consequences. In the APEX clinical trial, patients with inactivation of TRAF3 appear to have a lower response rate to dexamethasone (10%) and higher response rate to bortezomib (90%), associated with significant prolongation of PFS (83 vs 193 days). In the remaining patients there is no significant difference between dexamethasone and bortezomib in terms of response or PFS. This has important implications for the selection of patients for different treatments, and for the implied mechanism of action of both glucocorticoids and proteasome inhibitors in the treatment of MM. Furthermore, we find that the timing of oncogene activation is the critical determinant of the disease, the nature of the oncogene the critical determinant of the phenotype
Based on a novel mouse model that activates oncogene expression using somatic hypermutation (as occurs in 10% of the translocations in MM) we have developed a faithful model of myeloma. The mouse model uses the c-myc oncogene, which is critically important in human MM, translocated in 15% of newly diagnosed MM, 40% of advanced refractory MM, and over 90% of HMCL. A major clinical implication of these studies is that both the timing of oncogene activation, and the nature of the oncogene, are critical determinants of the resulting disease phenotype.