Optical detection and grading of lung neoplasia by Raman microspectroscopy

The aim of this study was to investigate whether Raman spectroscopy could be used to identify and potentially grade lung neoplasia in cell samples. Normal human bronchial epithelial cells (HBEpCs) were analyzed by Raman spectroscopy and compared with (i) HBEpCs expressing human papillomavirus (HPV) type 16 E7 or CDK4; (ii) the immortalized bronchial epithelial cell line BEP2D and (iii) its asbestos-transformed derivative AsbTB2A. Overall, Raman spectroscopy, in combination with a linear discriminant analysis algorithm, was able to identify abnormal cells with a sensitivity of 91% and a specificity of 75%. Subdivision of the cell types into 3 groups, representing normal cells (HBEpCs), cells with extended lifespan (HBEpCs expressing HPV 16 E7 or CDK4) and immortalized/transformed cells (BEP2D and AsbTB2A) showed that Raman spectroscopy identifies cells in these categories correctly with sensitivities of 75, 79 and 87%, and specificities of 91, 85 and 96%, respectively. In conclusion, Raman spectroscopy can, with high sensitivity, detect the presence of neoplastic development in lung cells and identify the stage of this development accurately, suggesting that this minimally invasive optical technology has potential for lung cancer diagnosis. (C) 2008 Wiley-Liss. Inc.