Study of Correlation of Cervical Epithelial Thickness With the Grade of Colposcopic Abnormality.

Abstract

Low epithelial thickness has been identified as the cause for nonvisualization of high-grade cervical intraepithelial neoplasia (CIN) on colposcopy in an earlier study. Multiple random biopsies are recommended by some authors to detect these "thin" CIN lesions in absence of colposcopic abnormalities. The present study was conducted to evaluate the correlation between the severity of colposcopic impression and the thickness of the epithelium so that the results of previous study could be validated. The cross-sectional study examined 209 histopathology slides with normal, human papillomavirus, or CIN diagnosis from a population-based study. Average epithelial thickness was measured by obtaining mean of the thicknesses at thinnest and thickest areas. Average thickness of dysplastic layer was also measured. These values were correlated with age, human papillomavirus status, colposcopic appearance and histopathology. Mean epithelial thicknesses were 212.8 μm for normal (N=28), 297.3 μm for human papillomavirus changes (N=48), 245.3 μm for CIN1 (N=46), 191.4 μm for CIN2 (N=50), and 218.5 μm for CIN3 (N=37). Within each histologic category, no correlation was observed between epithelial thickness and severity of colposcopic appearance. Mean epithelial thickness of CIN1/CIN2 lesions with normal colposcopy was more than that of CIN1/CIN2 lesions with high-grade appearance on colposcopy. Thickness of CIN3 lesions with high-grade abnormalities was higher than those without visible colposcopic abnormality but the difference was not statistically significant. Thickness of dysplasia increased with higher grades of CIN but did not have any relation to colposcopic appearance. Colposcopic appearance does not depend on the thickness of the epithelium affected by CIN. False-negative colposcopy in presence of high-grade CIN is likely due to failure of detecting small or predominantly endocervical lesions rather than "thin" CIN.