Our Science – Kuehl Website
W. Michael Kuehl, M.D.
Molecular Pathogenesis of Multiple Myeloma
Multiple myeloma (MM), with a yearly incidence of about 20,000 in the US, accounts for nearly 2% of cancer deaths. MM usually is preceded by a pre-malignant MGUS tumor that can be diagnosed by simple blood tests and is exceptionally common, occurring in about 3% of individuals over the age of 50. The normal counterpart of both tumors is long-lived, post-germinal center bone marrow plasma cells (PC) that have undergone multiple rounds of somatic hypermutation and antigen selection of immunoglobulin (Ig) genes, followed by IgH switch recombination. For both MGUS and MM, the incidence is markedly age dependent, and about two fold higher in American blacks than in Caucasians. Non-IgM MGUS can sporadically progress to MM at an average rate of 1% per year. Currently, there is no way to predict when a particular MGUS tumor will progress to MM. Chromosome content identifies a similar prevalence of hyperdiploid (HRD) MGUS/MM tumors that contain 48-75 chromosomes, typically with over-representation of chromosomes 3, 5, 7, 9, 11, 15, 19, and 21, and non-HRD (NHRD) tumors that contain less than 48 and/or >75 chromosomes.
Our research has accomplished, and in some cases is continuing to study, the following:
1) Characterization of a panel of 53 MM cell lines (MMCL), which are derived mostly from extramedullary MM tumors, and analyses of primary tumor samples, as summarized below (see also diagrams in Gallery link).
2) Evidence showing that most MM tumors are preceded by MGUS tumors.
3) Primary IgH translocations in MGUS and MM tumors. Using a combination of molecular cytogenetics, gene expression profiling, and molecular cloning, we identified seven recurrent IgH translocations (TLC). These mostly simple TLC appear to be mediated by errors in IgH switch recombination or somatic hypermutation as B cells pass through germinal centers, suggesting that they represent very early or primary events in pathogenesis. The seven IgH TLC are present in 40% of MM tumors, including 70% of NHRD tumors but only 12% of NHRD tumors. They comprise three primary TLC groups:
CYCLIN D: 11q13(D1),15%;12p13(D2),<1%;6p21(D3),2%;
MMSET/FGFR3: 4p16(MMSET & FGFR3), 15%.
4) CYCLIN D dysregulation as an early, unifying event in MGUS and MM tumors. Despite a low proliferation index, we showed that virtually all MGUS and MM tumors have dysregulated and/or increased expression of CYCLIN D1, D2, or D3 compared to normal PC. About 25% of tumors have an IgH TLC that directly dysregulates a CYCLIN D gene or a MAF gene that encodes a transcription factor targeting CYCLIN D2. Another 40% of MM tumors are HRD and bi-allelically express CYCLIN D1, unlike normal lymphoid cells or PC that express little or no CYCLIN D1. Most other tumors, including those with a t(4;14) TLC, have increased expression of CYCLIN D2. The HRD tumors that bi-allelically express CYCLIN D1 but not CYCLIN D2 are not represented among our panel of 53 HMCL, suggesting a particularly strong dependence on interactions with bone marrow stromal cells.
5) TC (Translocation-CYCLIN D) classification of MGUS and MM tumors. Based on these early pathogenic events, we proposed eight TC groups: MAF, 8%; 4p (MMSET/FGFR3), 15%; 6p (CYCLIN D3), 2%; 11q (CYCLIN D1), 15%; D1 (bi-allelic CYCLIN D1), 33%; D1+D2, 8%; D2, 17%; and NONE (no increased expression of a CYCLIN D gene compared to normal PC but frequent inactivation of RB1 or significant contamination of tumor sample with normal PC). This can be condensed to six groups by combining 6p/11q and D1/D1+D2. The TC groups have different biological and clinical features, suggesting that they represent different disease entities determined by early pathogenic events. Each of these groups could be divided into sub-groups with high (~15% of newly diagnosed tumors) and low expression proliferation indices, the former being associated with a poorer prognosis.
6) Secondary as progression events. We have identified and characterized secondary TLC/insertions, including all MYC (c->>N->>L-) TLC, most Ig light chain (IgL>>IgK) TLC, most IgH TLC not involving the recurrent partners, and perhaps a small fraction of TLC involving the seven recurrent partners. Secondary TLC often have complex and unbalanced structures, and have a similar prevalence in HRD and NHRD tumors. They can occur at all stages of tumorigenesis, including pre-malignant MGUS, although MYC TLC occur mainly as very late progression events when tumors are becoming more proliferative and less stromal cell dependent. MYC TLC are absent or rare in MGUS, but are present in 15% of newly diagnosed tumors, 40% of advanced MM tumors, and >90% of HMCL. Surprisingly, about 40% of MYC TLC do not involve an Ig locus.
7) The RAS paradox. We have identified K-RAS mutations in 17%, and N-RAS mutations in 14% of 306 MM tumors. Paradoxically, although the prevalence of K-RAS mutations was the same in all TC groups, the prevalence of N-RAS mutations was much higher in tumors that express CYCLIN D1 (22%, but similar for 11q and D1 groups) than in tumors that express CYCLIN D2 (4%). For 39 MGUS tumors, there were no K-RAS mutations but three (6%) N-RAS mutations, so that K-RAS may provide a marker that distinguishes some MM tumors from MGUS tumors.
8)Identifying a critical role for NFKB activation in MM, MGUS, and normal PC. Recently, we found that 40% of HMCL and about 15% of MM tumors have mutations in nine genes that result in constitutive activation of the NFKB pathway, which is associated with a high NFKB index. However, in contrast to HMCL, which usually have a low NFKB index if there are no mutations in this pathway, all normal PC as well as most MGUS and MM tumors have a high NFKB index even without mutations. Others have shown that activation of the NFKB pathway by extrinsic ligands (APRIL, BAFF) is critical for survival of normal mouse bone marrow PC. Therefore, it appears that activation of the NFKB pathway - either by extrinsic signals or intrinsic mutation - is important for the survival and growth of normal PC as well as most MGUS and MM tumors.
9) Increased proliferation associated with additional RB pathway disruption. Dysregulated expression of CYCLIN D is insufficient to cause a high level of proliferation. We found that p18INK4c expression, which is essential for the generation of normal PC, is altered in most proliferative MM tumors. There is homozygous deletion of p18 in 30% of HMCL and 10-20% of proliferative MM tumors. However, most HMCL and proliferative MM tumors have a paradoxical increase in p18 expression, most likely due to increased E2F that causes increased transcription of p18. Therefore, these tumors have become insensitive to increased expression of p18. We cannot explain this paradox except for rare tumors/HMCL that have inactivated RB1.
-10)A current model for the pathogenesis of non-IgM MGUS and MM tumors. There are four partially overlapping events that occur early in pathogenesis: hyperdiploidy, primary IgH TLC, chromosome 13/13q deletion, and dysregulation of a CYCLIN D gene. Progression events include: 1) secondary TLC as well as karyotypic and epigenetic changes can occur at all disease stages; 2) activating RAS or FGFR3 mutations appear to occur close to the time that the tumor progresses from MGUS to MM; 3) the timing of NFKB mutations is not clearly defined but probably occurs beyond the MGUS stage; 4) inactivation or mutation of p53, MYC rearrangements and dysregulation, and additional disruption of the RB pathway (inactivation of p18INK4c or RB1) appear to be relatively late progression events that are associated with increased proliferation and a poor prognosis.
Collaborating with us are P. Leif Bergsagel, Marta Chesi, and Rafael Fonseca, Mayo Clinic Arizona; Bart Barlogie, Jeffrey Sawyer, and John Shaughnessy, University of Arkansas; Brendan Weiss, Walter Reed Army Medical Center; Louis Staudt, Ola Landgren, and Beverley Mock, NCI.
This page was last updated on 6/13/2014.