Sawyer, Jeffrey R.

Abstract Chromosomal translocations involving the immunoglobulin heavy-chain (IgH) locus at 14q32 are genetic hallmarks of human multiple myeloma (MM) and are likely primary oncogenic events in myeloma transformation. Common recurrent translocation partner genes of IgH account for 40–50% of cases and these rearrangements result in the transcriptional dysregulation of various proto-oncogenes such as CCND1, CCND3, FGFR3/MMSET, c-MAF, MAFB, MUM 1(IRF-4), c-MYC and it is becoming clear that these dysregulated genes play important roles in the pathogenesis and the treatment outcome of MM. Although the frequency of other partner genes of IgH translocations may be rare (<1%), it is possible that many genes not currently identified as translocation partners are functionally converged into several signal pathways and play important roles as well. Through global gene expression profiling (GEP) studies of a large population of newly diagnosed MM, we have recently reported that dysregulation of any one of the three D-type cyclins is seen in almost all newly diagnosed MM and most patients with high levels of CCND2 expression was highly correlated with expression of c-MAF, MAFB or FGFR3/MMSET. In order to identify novel translocation partner genes in MM we combined GEP and fluorescence in situ hybridization (FISH) analysis in a large population of newly diagnsoed myeloma. We investigated whether IGH-CCND2 and IGH-MAFA translocations are seen in patients exhibiting high levels of CCND2 expression but not expressing c-MAF, MAFB, or FGFR3/MMSET. We also looked for a breakpoint and targeted gene at a non-random breakpoint at the 11q23 locus in 20 MM patients with 11q23 abnormalities detected by conventional G-banding. Here we report evidence of an IGH-CCND2 fusion in a newly diagnosed case of MM and an IGH-MAFA fusion in a newly diagnosed MM and an MGUS and also mapped the 11q23 breakpoint to within 100 kb in a newly diagnosed MM and a human myeloma cell line, NCU-MM-1 and identified an IGL Pou2AF1 fusion in the NCU-MM-1 line. Importantly these translcoatios do not involve the MLL or ATM genes. These three genes were all highly expressed as detected by GEP or quantitive RT-PCR in the samples in which these genes were translocated to IG loci. Although the frequency of these 3 translocations are very rare, they further the concept of a critical role of CCND2 expression in myelomegenesis and identifed the transcriptional coactivator Pou2AF1, which specifically associates with either OCT1 or OCT2, in the B-cell response to antigens and required for the formation of germinal centers, as a potential new translocation target in myeloma. Finally, our strategy of combining GEP signatures and cytogenetic analyses appears to be a practical means of identifying new IG- mediated translocations in myeloma.

Abstract Background: AT’s have become the standard of care in the management of patients with MM, effecting 10-year survival rates of more than 30%. Long-term sequelae are uncommon, so that even 3rd and 4th AT have been applied. MDS is a well recognized complication of cancer therapy and has been reported in the context of both standard and AT-supported high-dose therapies for lymphoma and MM. It remains unclear whether t-MDS development is caused by prolonged “low” dose chemotherapy or by pre-transplant regimens. The purpose of this study was to survey a cytogenetic (CG) data base (DB) of > 30.000 samples submitted with each bone marrow examination involving 2152 patients receiving at least 1 AT. Patients and Methods: 1886 patients were identified with all relevant information as well as pre- and post-AT CG data available receiving at least 1 AT prior to January 2004. Common criteria for MDS-CA were employed. Routine CG follow-up studies were performed at 3–6 month intervals. MDS-CA development was examined in relationship to age, duration and type of therapy prior to transplant preparative treatment, blood counts, PBSC collection, AT regimen and number, post-transplant platelet and granulocyte recovery, as well as in the context of MM parameters. Interphase FISH analysis for MDS was also performed on PBSC in 42 patients (to detect early MDS cytogenetic changes caused by standard-doses therapies preceding AT) and correlated with MDS-metaphase CA post-AT in 7 patients. Results: Eighty-five patients developed MDS-CA at a median of 33 months after AT (range, 5 to 126 mo); the 10-year incidence estimate was 7%. MDS-metaphase CA development was independently associated with older age (> 55yr), lower CD 34 yield with first collection (< 18 million/kg) and lower platelet recovery 3 mo post-AT. The 10-year probability of MDS-metaphase CA was 1% among 315 patients with 0 risk factor, 6% among 1367 patients with 1 or 2 and 13% among the 199 patients with 3 risk factors (p<.0001). Univariately, >12 mo of pre-AT standard therapy increased the cumulative incidence of MDS-metaphase CA (p=0.04). The median time to MDS-metaphase CA development was shorter (24 mo) in the 55 patients with 2 or 3 risk factors compared to the 28 patients with 0 or 1 risk factors (41 mo; p=0.01). Conclusion: This report is the largest ever on the detection of MDS by metaphase CG, obtained in the context of AT for MM. Results confirm earlier findings that MDS-metaphase CA is infrequent when PBSC are procured early after the start of chemotherapies, in those with robust stem cell collection and prompt platelet recovery post-transplantation (evidence of non-damaged hematopoiesis) and among the younger age group. Thus, t-MDS after AT-supported high-dose therapy can be traced to stem cell damage likely inflicted by and not completely repaired after prolonged pre-AT therapies and accentuated by the post-AT replicative stress, especially in older patients. MDS- FISH data of PBSC will be presented at the meeting.

Two oncogenic pathways have been hypothesized for multiple myeloma (MM) and premalignant monoclonal gammopathy of undetermined significance (MGUS) tumors: a nonhyperdiploid pathway associated with a high prevalence of IgH translocations and a hyperdiploid pathway associated with multiple trisomies of 8 chromosomes. Cyclin D1, D2, or D3 expression appears to be increased and/or dysregulated in virtually all MM tumors despite their low proliferative capacity. Translocations can directly dysregulate CCND1 (11q13) or CCND3 (6p21), or MAF (16q23) or MAFB (20q11) transcription factors that target CCND2. Biallelic dysregulation of CCND1 occurs in nearly 40% of tumors, most of which are hyperdiploid. Other tumors express increased CCND2, either with or without a t(4;14) translocation. Using gene expression profiling to identify 5 recurrent translocations, specific trisomies, and expression of cyclin D genes, MM tumors can be divided into 8 TC (translocation/cyclin D) groups (11q13, 6p21, 4p16, maf, D1, D1+D2, D2, and none) that appear to be defined by early, and perhaps initiating, oncogenic events. However, despite subsequent progression events, these groups have differing gene expression profiles and also significant differences in the prevalence of bone disease, frequency at relapse, and progression to extramedullary tumor.