Deciphering the genomic landscape and evolution in multiple myeloma

Abstract

Multiple myeloma (MM) is the second commonest haematological cancer in Western countries, with most patients dying from progressive disease after relapse. Currently, the molecular mechanisms responsible for the initiation and evolution of MM are poorly understand. The work presented in this thesis aims to characterise novel coding and non-coding drivers, gain insight into the aetiological basis, and understand the genetics of MM evolution and relapse through integrated study of multiple next-generation sequencing datasets. Firstly, using the CoMMpass datatset (>800 patients), multiple regulatory regions were identified as candidate non-coding drivers, including cis-regulatory elements (CREs) of MYC and a PAX5 enhancer. Coding drivers in 40 genes, including 11 novel, were identified. The study revealed that MM oncogenic pathways are targeted somatically through multiple novel mechanisms including coding and non-coding mutations; exemplified by IRF4 and PRDM1, along with BCL6 and PAX5, genes central to plasma cell differentiation. Secondly, coding and non-coding regions were dominated by distinct mutational processes with aging, DNA repair deficiency (DRD), and APOBEC/AID activity characterising MM. Mutational signatures showed subgroup specificity - APOBEC signatures with MAF-translocation t(14;16) and t(14;20) MM; DRD with t(4;14) and t(11;14); and aging with hyperdiploidy. Mutational signatures beyond that associated with APOBEC were independent of established prognostic markers and had relevance to predicting high-risk MM, providing a strong rationale for integration of mutational signatures to tailor therapy. Thirdly, analysis of high-coverage WGS dataset of primary and matched relapsed tumours from Myeloma XI trial validated several recurrently mutated CREs and discovered novel CRE targets (eg. BIRC2 and IGLL5). Relapsed patients were characterised by higher mutational burden, and associated with increased APOBEC/AID activity and DRD. Notably, further acquisition of high-risk large-scaled copy number variants at relapse was also observed, specifically enriched at pre-existing unstable genomic regions. Three major clonal evolutional patterns were identified at relapse: (i) no change in clonal composition (ii) subclonal expansion (iii) emergence of new clones accompanied by decline of primary clones Finally, defective transcription-coupled DNA repairs was observed as predominant mutational process in MM mitochondrial DNA. Relapsed MM was characterised with global positive selection of non-synonymous mutations, most notably in genes encoding the NADH dehydrogenase complex (MT-ND2, MT-ND4 and MT-ND5). Together, these findings provide increased insights into the complex genetic basis underlying MM and its progression to relapse, with potential to support the development of personalised and effective treatment strategies, and predictive biomarkers of therapeutic outcome.