OP19SERUM SPECTROSCOPY OF BRIAN TUMOURS: A RAPID AND ACCURATE DIAGNOSTIC TOOL

INTRODUCTION: The ability to diagnose brain tumours rapidly from serum would allow for rapid testing and decreased time to results providing a responsive diagnostic environment. ATR-FTIR (Attenuated Total Reflection - Fourier Transform Infrared Spectroscopy) generates a serum sample - fingerprint - since biomolecules exhibit different responses to different wavelengths of light. We have previously used ATR-FTIR to discriminate high grade brain cancer, low-grade brain cancer and normal serum to sensitivities and specificities on average of 93.75 and 96.53 % respectively. This paper reports the analysis of 429 patients. METHOD: 1 µl of serum was analysed in triplicate from 429 patients via ATR-FTIR. This dataset was analysed using pattern recognition algorithms for the following models 1) Cancer (n = 314) vs Non-cancer (n = 122), 2) Metastatic Brain Cancer (n = 192) vs Brain Cancer (n = 124) vs Non-Cancer (n = 122), 3) Organ of tumour origin, 4)High grade glioma (n = 52) vs low grade glioma (n = 24) vs meningioma (n = 47) 5) Subtype of Brain Cancer. RESULTS: ATR-FTIR spectroscopy was able to discriminate different groupings with high accuracy. The predicted range for each diagnostic layer was between 78.26 - 100.00 % sensitivity and specificity respectively. Interrogation of the loadings plots showed differences in Amide and lipid content of the serum. CONCLUSION: A robust reproducible method has been developed for the diagnosis of small volumes of serum. In addition multiple layers of diagnostic information can be provided from the same spectral dataset. ATR-FTIR has shown to be an excellent clinical tool for rapid diagnosis enabling a dynamic and responsive clinical environment. Future work will investigate spectroscopic changes during treatment.