P08.06.B TARGETED PLASMA METABOLOMICS DETECTS A METABOLIC SIGNATURE POTENTIALLY RELEVANT FOR DIAGNOSIS AND PROGNOSIS IN PATIENTS WITH GLIOBLASTOMA
M Pantalone, C Söderberg Naucler, G StragliottoAbstract
BACKGROUND
Glioblastoma diagnosis is made initially with radiological examination and pathological examination of the resected tumor. Debut symptoms may vary and be mild in the beginning therefore diagnosis can be delayed. Additional un-invasive diagnostic strategies can be useful for diagnosis and clinical follow up. Moreover, despite glioblastoma being almost invariably fatal, there are no clinically established plasma biomarkers to define prognosis in patients with glioblastoma.
MATERIAL AND METHODS
In this study we employed flow injection analysis tandem mass spectrometry to compare the plasma metabolic profile of patients with GBM (n=51) and healthy controls (n=24).
RESULTS
Principal Component Analysis (PCA) and Partial Least Discriminant Analysis (PLS-DA) highlighted a distinctive glioblastoma signature. The best confusion matrix from the ROC curve built on two metabolites, GLU and GLN-LYS, provided specificity of 87% and sensitivity of 92%, with positive and negative predictive values of 95% and 88%, respectively. Multivariate survival analyses revealed that higher levels of ALA, MET, ORN, PHE, TYR were to be associated with a shorter overall survival.
CONCLUSION
These results are currently being evaluated prospectively and at different stages of disease in a larger cohort of patients with glioblastoma to identify plasma metabolic markers relevant for recurrence development.