Abnormalities of 16q in Multiple Myeloma Are Associated with Poor Prognosis: 500K Gene Mapping and Expression Correlations Identify Two Potential Tumor Suppressor Genes, WWOX and CYLD.

Abnormalities of 16q are important recurrent events in multiple myeloma (MM). We performed FISH on CD138 selected plasma cells from 701 newly diagnosed MM patients from the LRF UKMF cytogenetics database. Gene mapping, including paired normal controls, and gene expression analysis was performed on 55 cases using the Affymetrix Human Mapping 500K Array Set and U133 Plus 2.0 Arrays respectively. 16q deletion (del16q) was identified by FISH using probes for cMAF (Abbott Diagnostics) in 131/701 cases (18.7%) and was significantly associated with deletion 17p (16.5% vs. 8.9%, p=0.006), deletion 13 (60.8% vs. 48.5%, p=0.009), deletion of IgH (22.1% vs. 11.1%, p=0.0003) and non-hyperdiploid status (58.3% vs. 42.7%, p=0.006). Del16q showed a trend to poor overall survival, mean survival 43 vs. 61 months (p=0.09), and was associated with significantly worse survival in combination with t(4;14) compared with either t(4;14) or del16q alone, mean survival 15 vs. 26 vs. 45 months respectively (p=0.006). t(14;16) was identified by FISH in 31/701 cases (4.4%) and was associated with poor prognosis, mean survival 29 vs. 54 months (p=0.005). Mapping arrays revealed loss of heterozygosity (LOH) involving all or part of 16q in 20 of 55 cases (36%) in 3 distinct patterns: uniparental disomy (UPD) of chromosome 16 or 16q in 4/55 cases (7%); deletion of chromosome 16 or the whole of 16q in 11/55 cases (20%); and interstitial deletion of small regions of 16q in 5/55 cases (10%), focused on 16q12, the location of CYLD, and 16q23, the location of WWOX. 16q LOH was distributed across translocation groups but was identified in all 4 mapping cases containing 17p deletion, supporting the association identified by FISH. As WWOX is the site of the common fragile site FRA16D and deletions at common fragile sites have been associated with DNA instability in human cancers, we assessed this using gene mapping in these 55 MM cases. Although deletions spanning other common fragile sites were identified, they were not restricted to those with 16q LOH. However, in 2 t(14;16) cases, hemizygous deletions of approximately 100kb could be identified within WWOX at the presumed translocation breakpoint. One of the t(14;16) cases had a similar hemizygous deletion within FHIT, another tumor suppressor gene located within common fragile site FRA3B, consistent with findings in other cancer types. Cases with 16q LOH or t(14;16) all had significantly reduced WWOX expression relative to cases without 16q abnormalities, confirming gene inactivation by either LOH or translocation. Cases with 16q LOH also had significantly reduced expression of two other potential tumor suppressor genes located on 16q, CYLD and RBL2. In summary, our data confirms the adverse prognosis associated with 16q translocation or deletion. Array data reveals 16q LOH occurs due to deletion or UPD with two regions involved, one defined by CYLD and the other by WWOX. WWOX is also inactivated by translocation and is associated with interstitial deletions at this and other common fragile sites. WWOX is a likely candidate gene in MM pathogenesis because of its interaction with TP53 and CYLD via its effects on NF-κB.