P59DIAGNOSTIC APPLICATION OF VIBRATIONAL SPECTROSCOPY FOR THE SEGREGATION OF BRAIN TUMOUR CELL LINES COUPLED WITH MULTIVARIATE STATISTICS

摘要

INTRODUCTION: Brain tumours are caused by the proliferation and multiplication of single cells which go through molecular mutations. These neoplasms are identified histologically through an increased cellular grouping and disorder in tissue giving it a unique morphological appearance. Histopathology has been shown to be highly subjective and pathological discrepancies have been shown to occur in 1-43% of specimen. Raman and Infrared spectroscopy are types of vibrational spectroscopy which have a potential to be used as diagnostic tools in neuropathology. They provide an intrinsic molecular fingerprint of the sample based on the interaction of light and the sample. Our study investigates the potential application of vibrational spectroscopy in the segregation of different types of brain tumours using two tumour cell lines and a control. METHOD: U87MG (WHO grade IV), 1321N1 (WHO grade III) and SVGP12 (Control) cells were cultured based on the manufacturers protocols and fixed on CaF2 slides using 10% formalin and left to dry at room temperature. Spectra were collected using ATR-FTIR and Raman spectroscopy from three biological replicates of each cell line, n = 30. RESULTS: PCA, DFA and SVM models were able to separate the three cell lines based on their chemical information with very high sensitivities, 1321N1 = 100%, SVGP12 = 100%, U87MG =90% and specificities 1321N1 = 100, SVGP12 = 100% and U87MG =94.7%. Separation based on the loading was mostly due to C-C/C-O phospholipids, CH3 CH2 wagging mode of collagen, nucleic acids, proteins and lipids. CONCLUSION: Both Infrared and Raman spectroscopy can be used as diagnostic tools to purge chemical information on biological samples with good reproducibility.