Abstract
Acute myeloid leukaemia (AML) is a highly heterogeneous disease in terms of genetics and at the clonal level. Despite improved understanding of the genetic landscape of AML, pathogenesis and cause of disease relapse remain poorly understood. To address these issues exome sequencing was carried out on 25 diagnostic and 12 relapse samples from 28 cytogenetic standard risk AML patients. More than 500 somatic mutations were identified, including recognised recurrent mutations, and novel variants with driver potential. Copy number variations were also investigated, identifying loss of heterozygosity of chromosome 13 involving the FLT3 locus as a common event acquired at disease relapse. The heterogeneity of disease relapse was also highlighted with the analysis of paired diagnostic and relapse samples.To further develop an understanding of AML clonal architecture and disease evolution, targeted deep sequencing was then carried out using a custom gene panel informed by exome sequencing results. Two cohorts were analysed including 223 diagnostic and 49 relapse samples from NPM1 mutant AML patients, followed by 43 unselected AML patients. In the NPM1 mutant cohort co-operating mutations were identified in 99% of patients, while at least one mutation was detected in 91% of the unselected cohort. Furthermore, quantification of variant allele frequencies allowed detailed analysis of mutational hierarchy and clonal evolution of relapse. The prevalence of mutations in candidate genes was also assessed, notably identifying mutations of MYC as recurring events in NPM1 mutant AML.
Knowledge of AML genetics can also be applied post-treatment to track minimal residual disease (MRD) using molecular markers. Reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) assays were designed and optimised for rare fusion genes and NPM1 mutations. As a model, the NPM1 mutant RT-qPCR assays were used to track MRD in follow up samples from a large cohort of NPM1 mutant AML patients. These data identified a single MRD measurement in the blood following the second course of chemotherapy was the most informative outcome predictor, independent of mutational profile. It was also established that analysis of sequential samples could identify patients destined for relapse providing an opportunity for pre-emptive treatment. In the future a more personalised treatment approach, incorporating both mutational profiling and MRD monitoring, may improve patient outcome.
| Date of Award | 2016 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | David Grimwade (Supervisor) & Ellen Solomon (Supervisor) |