The Fusion Gene Landscape in Multiple Myeloma, with Clinical Impact

Background: Gene fusions play an important role in aberrant cellular biology as well as development and progression of cancer. Expression of fusion genes such as PML-RAR drives the transformation in APL and provides important targets for therapy. However, in multiple Myeloma (MM), a heterogeneous disease characterized by genomic instability, frequent gains and losses of DNA, and a diverse mutational landscape, only the well characterized MMSET-IGH fusion product has been reported. Here we investigate the fusion gene landscape in multiple myeloma, and its possible impact on survival.

Method: Deep RNA-Seq was performed on purified MM cells from 430 newly-diagnosed MM patients, 20 normal individuals and 71 cell-lines; data were analyzed for gene expression profiles, long-non coding RNA signatures, and both novel and known fusion genes using two common algorithms: TopHat and MapSplice. MM characteristics, cytogenetic and FISH as well as clinical survival outcomes were also analyzed and correlated with genomic data.

Results: After filtering candidate fusions linking genes belonging to the same family, we identified 416 different candidates in myeloma patients, 40 % of which identified either IGH or Kappa as a partner. IGH fusion partners included the previously described and validated WHSC1 and B2M genes, as well as over 50 new candidates, while more than 70 different partners were found to be fused with Kappa. These genes exhibit functional enrichment of positive regulators of the cytokine-mediated signaling pathway, negative regulation of myeloid cell differentiation, negative regulation of interleukin-6 production, as well as others. 31% of patients presented no fusions, and another 32% presented a single fusion event. The other 37% presented at least 2 fusion candidates, with up to 27 different candidates. Similar patterns were observed in cell-lines, with 196 unique candidates identified, only 16% of which involving IGH or kappa. However, all partners were found in at least one patient as well. Only 12% of cell-lines exhibited no fusion, and another 14% presented only one fusion. On average, 3 fusions were identified per cell-line, with a maximum of 10. Validation of some these fusion genes is required to understand their functional role. Importantly, although having IgH-or kappa-related fusions did not affect patient outcome by themselves, patients with high numbers of fusion candidates had worse event-free survival.

Conclusion: Our data describes a diverse and rich fusion gene landscape in Multiple Myeloma. Similar to mutational profiles, there is no predominant fusion gene driving the disease process. Association of poor prognosis with a higher number of fusions may indicate that genomic instability plays an important role in the biology of Multiple Myeloma.