Evolution and structure of clinically relevant gene fusions in multiple myeloma

Steven M Foltz,Qingsong Gao,Justin King,Li Ding,Christopher J Yoon,Hua Sun,Daniel R Kohnen,Mark A Fiala,Ravi Vij,Reyka G Jayasinghe,Yize Li,Lijun Yao,Song Cao

doi:10.1038/s41467-020-16434-y

Abstract

Multiple myeloma is a plasma cell blood cancer with frequent chromosomal translocations leading to gene fusions. To determine the clinical relevance of fusion events, we detect gene fusions from a cohort of 742 patients from the Multiple Myeloma Research Foundation CoMMpass Study. Patients with multiple clinic visits enable us to track tumor and fusion evolution, and cases with matching peripheral blood and bone marrow samples allow us to evaluate the concordance of fusion calls in patients with high tumor burden. We examine the joint upregulation of WHSC1 and FGFR3 in samples with t(4;14)-related fusions, and we illustrate a method for detecting fusions from single cell RNA-seq. We report fusions at MYC and a neighboring gene, PVT1, which are related to MYC translocations and associated with divergent progression-free survival patterns. Finally, we find that 4% of patients may be eligible for targeted fusion therapies, including three with an NTRK1 fusion.

Highlights

Multiple myeloma is a plasma cell blood cancer with frequent chromosomal translocations leading to gene fusions
We explore the prognostic relevance of fusions by analyzing progression-free survival and find that those with IGH--WHSC1 or PVT1--IGL fusions have significantly worse outcomes. 4% of patients have a fusion annotated as a drug target in a public database
We detected gene fusions from 742 patients from the Myeloma Research Foundation (MMRF) CoMMpass Study, combining RNA and DNA sequencing data with clinical information to form a landscape of fusion events (Fig. 1, Supplementary Fig. 1, Supplementary Data 1–3)

Summary

Results

Total fusion burden was associated with worse prognosis; each additional fusion was associated with a slight decrease in PFS (HR estimate 1.02; 95% CI 1.00–1.04; two-sided z-score p value 0.0178), after controlling for disease stage and patient age (Supplementary Fig. 2d). By Visit 4, mutations in FAM46C, FGFR3, and KRAS were detected at or above 50% VAF, indicating a strong clonal expansion of the new mutations after diagnosis Another patient with an NTRK1 fusion, MMRF 2490, had clonal mutations in well-known myeloma tumor suppressors EGR1 and DIS3, meaning that targeting the NTRK1 fusion alone may not have been sufficient. Those mutations as well as expression levels of the fusion gene indicate tumor stability. Evidence of t(11;14) events is often observed in RNA and scRNA-seq due to upregulation of CCND1, actual IGH--CCND1 fusion transcripts may not be present or reported at the RNA level, and we find a similar low detection rate of chimeric transcripts in scRNA

BM 4 BM

Discussion

Methods

Code availability