Abstract
Traditional anti-doping assays target known banned drugs and endogenous molecules. A discovery platform such as metabolomics can be employed to find small molecule biomarkers that are related to the function of drugs of misuse. Biomarkers can be endogenous surrogate metabolites which have the potential to facilitate the discovery of new doping agents with biological activities similar to related named prohibited substances, as well as improving the sensitivity and prolonging the detection time window. Working towards this end, the focus of this thesis was to apply metabolomics to the anti-doping field in order to enhance detection of drug misuse in sport, in particular focusing in methodology that would help discover endogenous biomarkers, namely: sample information, quality, storage, transport, and multivariate/univariate statistics to mine significant biomarkers from metabolic data.The first part of the thesis explored analytical method-development in metabolomics, overall assessing the amount and quality of the metabolic information acquired in small sample volumes using paper spots. Methods for high resolution and high mass accuracy instrument were also studied to aid with the identification of molecules. After this, paper spots were also assessed for storage stability and compared with the conventional sampling method. From these analyses it was concluded that paper spots are a good option when limited sample is available, due to reproducibility, storage and transport advantages, however there is a loss of metabolic information when compared to liquid samples. The high mass accuracy did help with identification although the method was not tested in fragmentation mode.
Then the application of multivariate statistics to mine biomarkers from “doping fingerprints” was studied in detail. Firstly, a small clinical trial with longitudinal samples containing one (endogenous and exogenous) metabolite in urine, GHB, was “fingerprinted”. The aim was to find biomarkers that could be identified for a longer detection window than GHB in urine. A battery of multivariate tools was used to mine the data. Several metabolic features, albeit partly unidentified, showed interesting results, however it was concluded that a new study design with diet control and a placebo group, was needed to validate these features.
Lastly, samples from athletes were analysed and were grouped as salbutamol detectable and non-detectable specimens. Since this study was based upon the shortcomings of the previous GHB study, a strategy to diminish false positive biomarker discovery was included by adding a salbutamol spiked group in the experimental design. Results revealed a potential endogenous biomarker of salbutamol, hypoxanthine, with a ROC of 79 % prediction. These promising results need further validation with a subsequent clinical trial.
Overall these preliminary studies show the potential of metabolic fingerprinting in anti-doping. The findings have shown that metabolomics is a valuable tool in the discovery of surrogate biomarkers in doping.
| Date of Award | 2017 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Cristina Legido Quigley (Supervisor) & David Cowan (Supervisor) |