ID #929 Analysis of a Validation Cohort of Cerebrospinal Fluid Samples for the Validation of Biomarkers of Children’s Brain Tumour Relapse

Abstract

Malignant paediatric brain tumours have 30-50% relapse rates. In our first cohort of patients, at day 14 post-surgery, we discovered ten biomarkers of ependymomas and medulloblastomas in cerebrospinal fluid (CSF). To assess variation across the treatment course, we analysed a second set of samples obtained from the UK’s VIVO Biobank (n = 62).

Samples of ependymoma (n = 10), medulloblastoma (n = 49) and ATRT (n = 3) were more diverse than the discovery set. Liquid chromatography-tandem mass spectrometry (LC-MS) was used for targeted metabolomics with high linearity (R2>0.99) and precision (CV ≤ 11%).

Concentrations of metabolites in the validation set were different from those in the discovery set, perhaps due to more varied patient treatments and time points. The concentrations of metabolites in day 0 (the day of surgery) and day 14 medulloblastoma samples were not significantly different. Some medulloblastoma patients had a time course of samples spanning >19 weeks, which showed minimal variation in metabolite levels. The effect of three freeze-thaw cycles on metabolite concentrations was <5%, showing biomarkers stable to some repeat sample handling.

There appear to be baseline levels of some metabolites which do not vary with post-surgery trauma nor treatment time course, suggesting they are robust biomarkers of medulloblastoma presence throughout the treatment journey. We plan to analyse a third set of CSF samples to further validate our first cohort, which are in preparation. We hope this will lead to a liquid biopsy test for brain tumour monitoring.