Discrimination of exfoliative cytology samples employing ATR-FTIR microspectroscopy.

The aetiology of cervical cancer is strongly associated with viral infection (i.e. human papillomavirus (HPV)); other risk factors include low socio-economic class, multiple sexual partners, smoking and poor diet. Cervical cancer screening employing smears of exfoliative cytology for the visual identification of atypical cells has successfully led to a fall in mortality from this disease. However, false negatives occur when a smear is determined normal although atypical cells (be they pre-malignant or malignant) are present. The use of attenuated total reflection (ATR)-FTIR microspectroscopy to interrogate samples of exfoliative cytology is a novel approach to screening that may allow the rapid profiling of a pathological status. Changes in derived vibrational spectra may be associated with a biochemical biomarker of dyskaryosis. To investigate this possibility, we interrogated eight cervical samples of exfoliative cytology using ATR-FTIR microspectroscopy. Two were normal, two were low-grade cervical intraepithelial neoplasia (CIN)1, two were high-grade (CIN2/3) and two were invasive carcinoma samples. Cellular material was applied to an ATR crystal and vibrational spectra were obtained using a Bruker Vector 22 FTIR spectrometer with Helios ATR attachment. At least 10 spectra per sample of exfoliative cytology were acquired in the ATR mode using Bruker OPUS software; these spectra were then individually baseline corrected. These derived spectra were subsequently analysed using principal component analysis (PCA). Cluster plots allowed the identification of any spatial separation (pointing to increasing dissimilarity) as a function of spectral differences between different samples in a 3-D ??hyperspace??. Apparently independent of oncogenic HPV genotype, successful discrimination between normal, CIN1, CIN2/3 and invasive carcinoma samples was obtained. Normal and invasive carcinoma samples exhibited 100% segregation in ??hyperspace?? and the samples classified as low- or high-grade clustered variously between, pointing towards a model of progression. Retrospective histological analysis of smears confirmed the diagnostic potential of this approach. ATR-FTIR microspectroscopy potentially could be used in routine cervical cancer screening in an objective approach less prone to the current limitation of false-negative occurrence. The combined application of ATR-FTIR microspectroscopy and PCA may also allow for the tracking of biochemical markers of progression from normal to invasive carcinoma. We are currently interrogating larger cohorts to determine the applicability of IR microspectroscopy for routine screening