AJISTRACTInfrared spectra were obtained from exfoliated cervical cells from 156 females, of whom 136 were normal, 12-had cervical cancer, and 8 had dysplasia.

Infrared spectra of the normal connective, the normal epithelial, and the malignant epithelial tissues of cervix from seven patients have been measured as a function of pressure. Extremely high quality spectra of these tissue samples have been obtained. Consequently, structural differences at the molecular level among these three types of cervical tissues have been extracted from their pressure-tuning infrared spectra in the regions of the symmetric and antisymmetric stretching modes of phosphodiester groups, the C-O stretching mode, the CH2 bending mode, and the amide I mode. Significant differences in many features between the infrared spectra of the normal and the malignant cervical tissues and cells suggest that the infrared spectra of exfoliated cells and the biopsy of cervical tissues may be used in rapid evaluation of cervical cancer or in screening of large-volume normal cervical specimens. The infrared spectrum of the normal connective tissue of cervix in the frequency region 950 to 1100 cm−1 is similar to that of the malignant cervical tissue and cells. Therefore, if only this region of the spectrum is examined, the normal connective tissue will be misinterpreted as malignant tissue. However, the normal connective tissue can be differentiated unambiguously from the malignant tissue or the normal epithelial tissue by the infrared spectra in the frequency region 1200 to 1500 cm−1, where several well-defined sharp bands are unique for the normal connective tissue.

The confounding variables that can potentially lead to a misinterpretation of FTIR spectroscopy of exfoliated cervical cells is described. A detailed account of the spectral effects of the following variables in FTIR spectroscopic screening of exfoliated cervical cells is presented: polymorphs; Cell degradation; and impurities such as endocervical columnar cells, metaplastic cells, cervical mucus, red cells, and debris. The interpretation of the spectra of exfoliated cervical cells must be done with subtraction analysis, which includes these factors. This is essential to prevent unacceptable false-positive rates. The above techniques are subsequently applied to two clinic populations: a dysplasia clinic in follow-up patients with negative cytology and two general gynecology clinics with patients with negative cytology. In the dysplasia clinic group 250 sequential patients with negative smears were tested. Thirty had false-positive smears as defined by the IR spectroscopy using the above methodology. Twenty of those patients subsequently had one follow-up and six had a positive abnormal smear. In the community clinic group 656 sequential patients were examined who had negative smears, of which 27 had false-positive FTIR spectra.