Introduction: Acute lymphoblastic leukaemia (ALL) is the commonest childhood cancer. In the early trials of ALL treatment, central nervous system (CNS) relapse was a common occurrence - the introduction of CNS-directed therapy in the 1970s was associated with the largest single improvement in outcome for childhood ALL. Today, despite universal intensive CNS-directed therapy - with significant associated toxicity - the CNS is involved in around 50% of ALL relapses, with approximately 50% of these being isolated CNS (iCNS) relapse. Whilst many factors increase risk of CNS relapse, few are specific for CNS relapse. Discovery of specific risk factors for CNS relapse would allow increased therapy for children at high risk, and potentially less CNS-directed therapy for those at low risk. Relatively little is known about the biological differences between systemic and CNS ALL. In the CNS, leukaemic cells form plaques adherent to the leptomeninges, bathed in low-nutrient, low-oxygen cerebrospinal fluid (CSF). It was hypothesised that leukaemic cells adapt metabolically to this nutritionally poor CNS microenvironment, and these metabolic adaptations may be targets for specific therapy and/or specific biomarkers for CNS relapse. Findings: Transcriptional analysis of ALL cell lines from CNS and spleen in a mouse xenograft model, and of ALL cells retrieved from the CSF at CNS relapse of ALL, have shown the upregulation of cholesterol biosynthesis as a key adaptation to the CNS niche. Analysis of transcriptomic data from the bone marrow or peripheral blood from children with ALL at diagnosis have shown the potential for upregulated cholesterol biosynthesis (and, independently, upregulated IL7R) as a significant risk factor for CNS relapse of ALL. To support this finding, metabolomic analysis found evidence of changes in CSF cholesterol in the presence of CNS leukaemia, and of increased mevalonate (a cholesterol precursor) and cholesterol in ALL cells retrieved from the CNS in a xenograft model. Therapeutic targeting of CNS ALL in vivo with statins resulted in a CNS-specific increase ALL disease burden. Untargeted metabolomic analysis of CSF shows differences between children with ALL (either at diagnosis or on maintenance therapy) and non-ALL controls, and between children with ALL at diagnosis and the same children on maintenance therapy. Creatine abundance was significantly different in children with ALL at diagnosis compared with both other groups (1/3 lower at diagnosis than either on maintenance or non-ALL controls). This change in creatine and persisted on analysis of CSF from mice with and without leukaemia. On analysis of CSF from children at CNS relapse with ALL there is evidence of increased reduced creatine at time of CNS relapse in 3 of 4 patients. Conclusions: There is evidence to confirm the hypothesis that ALL cell adapt metabolically to the CNS niche. Cholesterol biosynthesis was identified as a key pathway upregulated in CNS ALL, and upregulated cholesterol biosynthesis in ALL cells at diagnosis was found to be a key risk factor for CNS relapse of ALL. In addition, clear changes in the CSF metabolome related to both ALL and ALL therapy were shown, and a new potential marker for the presence of CNS ALL identified. Prospective analyses in independent cohorts are required to determine the clinical utility of these novel strategies for prediction of CNS relapse risk.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:761967 |
| Date | January 2019 |
| Creators | Cousins, Antony Francis |
| Publisher | University of Glasgow |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://theses.gla.ac.uk/39009/ |
Page generated in 0.2089 seconds