Early detection of cervical neoplasia by Raman spectroscopy

Early detection of malignant tumours, or their precursor lesions, improves patient outcome. High risk human papillomavirus (HPV), particularly HPV16, infection can lead to the development of uterine cervical neoplasia, and therefore, the identification in clinical samples of the effects of HPV infection may have clinical value. In this report, we apply Raman microspectroscopy to live and fixed cultured cells to discriminate between defined cell types. Raman spectra were acquired from primary human keratinocytes (PHK), PHK expressing the E7 gene of HPV 16 (PHK E7) and CaSki cells, an HPV16-containing cervical carcinoma-derived cell line. Averaged Raman spectra showed variations, mostly in peaks originating from DNA and proteins, consistent with HPV gene expression and cellular changes associated with neoplasia, in both live and fixed cells. Principal component analysis produced good discrimination between the cell types, with sensitivities of up to 100% for the comparison of fixed PHK and CaSki. These results demonstrate the ability of Raman spectroscopy to discriminate between cell types representing different stages of cervical neoplasia. More specifically, this technique was able to identify cells expressing the HPV 16 E7 gene accurately and objectively, suggesting that this approach may be of value in diagnosis. Moreover, the ability to detect the effects of the virus in fixed samples also demonstrates the compatibility of Raman spectroscopy with current cervical screening methods. (c) 2007 Wiley-Liss, Inc.