Purpose: to compare histopathology and polymerase chain reaction (PCR) for the diagnosis of human papillomavirus (HPV) in cervical lesions of human immunodeficiency virus (HIV)-seropositive women. Methods: fifty-two HIV-seropositive women with suspected HPV cervical lesions were studied. Cervical scrapes were collected for PCR and colposcopy-guided biopsy was made for the histopathologic study. Three samples were disqualified for PCR, leaving a study population of 49 women. Results: the prevalence of HPV was 53% by histopathology and 85.7% by PCR. Among the 42 patients in whom HPV was detected by PCR, 26 were confirmed by histopathology (sensitivity = 61.9%). This method gave no false-positives (specificity = 100%), with 100% of positive prediction. Compared to PCR, the histopathology had: positive predcitive value = 100% and negative predcitive value = 30.4%. Among the 26 patients with HPV-positive biopsy, 15 (57.7%) had cervical intraepithelial neoplasia (CIN); relative risk = 13.3. Conclusion: histopathology was 100% correct for HPV-infection diagnosis. It means that when the biopsy is positive, HPV will be present, confirming the clinical suspicion. However, the low sensitivity excludes histopathology as a screening examination in this group of women.
Histology; Polymerase chain reaction; Human; HIV infection; Uterine cervix
