Biomarker discovery for pre-clinical diagnosis of Tasmanian devil facial tumour disease

Currently, there are no pre-clinical diagnostic tests to test for devil facial tumour disease (DFTD). This disease is a contagious cancer that has spread through the population of Tasmanian devils over 60 % of the island of Tasmania. The disease appears to always prove fatal and has decreased the overall population of Tasmanian devils by 84 %. This research investigated the potential of three mass spectrometry techniques to identify changes in the serum metabolome between DFTD and non-DFTD wild Tasmanian devils. Initially, a pilot study was conducted using a set of Tasmanian devil serum samples obtained from a total of 16 DFTD or non-DFTD devils to investigate the different mass spectrometry techniques. The serum was obtained from both males and females over the course of one year from locations in the Eastern half of Tasmania in known DFTD regions. One of these techniques examined the metabolites directly with mass spectrometry (MS) using electrospray ionisation - quadrupole ‚Äö- time of flight ‚Äö- mass spectrometry (ESI-MS). The other two techniques examined, capillary electrophoresis ‚Äö- mass spectrometry (CEMS) and gas chromatography ‚Äö- mass spectrometry (GC-MS), involved the separation of metabolites prior to MS detection. The CE-MS method was unable to detect as many compounds as the other two methods so only the ESI-MS and GC-MS techniques were investigated further. The ESI-MS and GC-MS techniques were used to examine the metabolites in three separate sets of samples obtained from wild Tasmanian devils that contained increasing amounts of sample variability using principal component analysis (PCA). Using the set with the least sample variability, both of the methods were successful at showing variability between all of the non-DFTD and all of the DFTD samples. Each of these two techniques was also able to classify devils with DFTD up to six months prior to visible tumours and this ability to identify DFTD devils pre-clinically continued with the other sample sets. The ESI-MS method was used to analyse a set of ~100 samples obtained from male devils in the Forrester Peninsula. This method provided a sensitivity of 66 % and a specificity of 80 %. The lower sensitivity was caused by uncertain classification of 6 of the DFTD samples and one DFTD sample incorrectly classified as non-DFTD. Unfortunately, because of the contagiousness of DFTD, all of the blinded samples in this sample set either had other diseases or developed DFTD within 12 months so the suitability of diagnosing healthy devils could not be determined. The final sample set examined included samples from both male and female devils obtained from locations throughout Tasmania. The results were improved by separating the samples according to gender. When the female samples were analysed, a sensitivity of 95 % was obtained using both methods, with specificities of 80 and 72 % for the GC-MS and ESI-MS method respectively. The PCA results for the male samples were not as successful at showing variability between the DFTD and non-DFTD devils which may have been caused by the bias in the sampling locations of the male DFTD devils. The results showed the presence of changes in the serum metabolome of DFTD Tasmanian devils that appear prior to clinical signs of the disease. With the correct validation, these methods could be used to diagnosis DFTD up to at least 6 months prior to clinical signs.