Agreement between methods for diagnosing HPV-induced anal lesions in women with cervical neoplasia.

Objective: To evaluate agreement between 3 methods for screening anal intraepithelial lesions: anal cytology, anoscopy and human papillomavirus (HPV) detection by PCR.

Study Design: This prospective, cross-sectional study screened 324 women with cervical neoplasia for anal neoplasia. Agreement between methods was calculated using the κ coefficient.

Results: Of 324 anal cytologies performed, 31.5% (n = 102) were found to be abnormal: low-grade anal lesions were detected in 19.1% (n = 62) of cases, high-grade lesions in 3.1% (n = 10) and atypical squamous cells of undetermined significance in 9.3% (n = 30). With respect to the biopsies, 25.7% (n = 20) were positive, consisting of 7 cases of HPV infection, 5 anal intraepithelial neoplasia (AIN) grade 1, 6 AIN grade 2, and 2 AIN grade 3. Twenty-one samples (6.5%) were inadequate for HPV analysis. Of the 303 adequate samples, 84.2% (n = 255) tested positive for HPV. Agreement between cytology and anoscopy was fair (κ = 0.31). Agreement between PCR for HPV and cytology was slight (κ = 0.08) and no agreement was found between PCR for HPV and anoscopy (κ = 0.00).

Conclusion: Agreement between the different methods of diagnosing HPV-induced anal lesions is slight to fair; however, anal cytology permits identification of cases in which lesions are present, allowing them to be referred for anoscopy and biopsy.