Screening of biochemical/histological changes associated to C6 glioma tumor development by FTIR/PCA imaging

Abstract

Principal components analysis (PCA) combined to Fourier transform infrared (FTIR) microspectroscopic imaging was used to screen biochemical changes associated to C6 glioma tumor from 7 to 19 days growth. Normal brain and tumor obtained at 7, 12, 19 days after C6 cell injection were used to develop a diagnostic model of brain glioma based on PCA analysis. This classification model was validated using extra-measurements on normal and tumor at 9 and 15 days post-implantation. The spatial and biochemical information obtained from FTIR/PCA maps can be used to improve the discrimination between normal and grading human glioma. The first 4 PCs which account for more than 93.6% of total spectral variance were used to construct pseudo-color scores maps and compared each map to the corresponding hematoxylin and eosin (H&E) staining. Our results reported that by correlating pseudocolor map scores with H&E staining it was possible to screen histological changes associated with tissue transformation. In fact, PC1 and PC4 were associated to the tumor, surrounding tumor and necrosis. Indeed, at day 7 after tumor implantation, FTIR investigations displayed a very small abnormal zone associated with the proliferation of C6 cells in the caudate putamen (CP). PC2 and PC3 described normal brain structures such as white matter (corpus callosum (CC) and commissura anterior (CA)) and some cortex layers (grey matter). After delipidation of the tissues, we were still able to differentiate between different tissue features based on nucleic acid and protein content. By comparing the patterns of the PC loads with the spectra of lipids extracted from white and gray matters, and DNA, we have identified some biochemical changes associated with tissue transformation. This work demonstrated that our classification model provides a successful histological classification of different brain structures.