Genetics of high risk myeloma

Abstract

Despite significant progress in therapy in the last two decades, the outlook for multiple myeloma (MM) patients is still poor as the disease remains incurable and only 33% of patients survive for up to 10 years after diagnosis. Reasons that underlie poor survival rate include genomic instability, which results in relapse and progression. As a result, there is interest in investigation of genomic aberrations and mutations that drive disease progression. I aimed to determine the genetic events that are associated with high-risk myeloma. I sought to ascertain whether there was an association between poor survival and low incidence genetic markers through performance of meta-analyses of large patient numbers. Additionally, I sought to answer specific questions regarding deletion and mutation of the tumour protein p53 (TP53) gene in terms of clonal heterogeneity at presentation. I investigated whether extreme numbers of copies at the chromosomal arm 1q were associated with survival. Furthermore, I investigated gene expression signatures in myeloma. I addressed these aims by evaluating data from two large randomised controlled trials of newly diagnosed myeloma patients. Through the study of high-risk lesions and clinical annotation, I was able to evaluate these cytogenetic lesions in thousands of uniformly treated patients. The most striking findings were that patients with subclonal TP53 deletion had a similarly poor prognostic association to those with clonal TP53 deletion, and that mono-allelic TP53 alterations had independent prognostic association to those with bi-allelic alterations. We also determined that gain(1q21) was associated with survival rates that were independent of amplification (1q21) and other high-risk lesions. My work provides critical analysis of the methods that are used to determine risk in newly diagnosed myeloma and supports the call for uniform and comprehensive reporting of risk through standardisation of trial report requirements in terms of risk calling.