UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterizing gene expression patterns associated with heterogeneity and relapse in pediatric acute myeloid leukemia Wei, Sumin


The effective treatment of patients in certain subgroups of pediatric acute myeloid leukemia (AML) and the management of relapse in this disease still remain unresolved challenges. Our ability to find solutions to these problems has been, in part, confounded by the inter- and intra-tumour heterogeneity observed in this disease, which may also affect how tumours evolve in response to treatment. The limitations in our knowledge of these areas may have been due to the lack of representative patient populations to facilitate the characterization of these different layers of heterogeneity and their relationships to each other. In this thesis, I was able to utilize clinical, genomic and transcriptome data from a diverse and representative cohort of de novo pediatric AML patients enrolled in a large clinical trial, AAML1031, to help better characterize the transcriptional profiles and heterogeneity within established genetic subgroups. Unsupervised clustering on gene expression data of primary samples revealed that genetically-defined subgroups were not completely recapitulated at the transcriptional level. While certain subgroups displayed distinct and relatively uniform transcriptional profiles, others, particularly those marked by fusion oncoproteins involved in epigenetic regulation of gene expression, were associated with activation of diverse transcriptional programs. I also observed that the transcriptional heterogeneity in the KMT2A-rearranged subgroup could be associated with significant differences in clinical outcomes, suggesting that characterization of the transcriptome could yield insights into tumour response to therapy and may also be helpful in patient stratification. I also utilized transcriptome data for primary and matched relapse samples from patients enrolled in AAML1031 to identify possible gene expression patterns associated with relapse. To do this, I compared the transcriptional changes at relapse in patients treated with chemotherapy alone to chemotherapy and bortezomib. Across a number of subgroups, I was able to identify modulation of the immune system to be associated with resistance to bortezomib treatment. This observation indicated that anti-cancer drugs may have off-target effects on non-malignant cells, highlighting the importance of careful evaluation of the impact of treatment on the microenvironment.

Item Media

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International