Cervical specimens collected in liquid buffer are suitable for both cytologic screening and ancillary human papillomavirus testing

BACKGROUND. Several new techniques have been developed to improve the sensitivity of cervical carcinoma screening and reduce equivocal cytologic diagnoses referred to as atypical squamous cells of undetermined significance (ASCUS). This study evaluates the effectiveness of combining two newly introduced diagnostic techniques: preparation of thin-layer cytologic slides from ThinPrep liquid buffer and selected Hybrid Capture testing for human papillomavirus (HPV) DNA. Because HPV DNA detection has been strongly associated with the presence of a cervical carcinoma precursor ('squamous intraepithelial lesion,' or SIL), HPV testing might be useful for identifying women with ASCUS who have an underlying SIL. METHODS. Two hundred specimens demonstrating diverse cervical abnormalities were selected from a prospective population-based study of 9174 women conducted in Costa Rica. The entire cohort had been screened with conventional cervical smears; ThinPrep slides made from liquid buffer; PAPNET, a computerized slide reading system; and Cervicography. Patients with any abnormal screening test were referred for colposcopy, punch biopsy, and loop excision of cases with high grade cytologic abnormalities not explained by punch biopsy. For this investigation, the results of ThinPrep cytology and HPV testing alone and in combination were compared with the final diagnoses, with an emphasis on the detection of carcinoma and high grade SIL. RESULTS. The 200 subjects studied included 7 women with a final diagnosis of carcinoma, 44 with high grade SIL, 34 with low grade SIL, 51 with a variety of equivocal diagnoses, and 64 with normal diagnoses. A ThinPrep cytologic diagnosis of SIL or carcinoma was made in 39 (76%) of the 51 women with final diagnoses of high grade SIL or carcinoma. Hybrid Capture testing detected carcinoma-associated types of HPV DNA in 100% of women with carcinoma, 75% with high grade SIL, 62% with low grade SIL, 20% with equivocal final diagnoses, and 12% of normal women. If colposcopy referral had been limited to women with a ThinPrep diagnosis of SIL or a diagnosis of ASCUS associated with the detection of carcinoma- associated HPV DNA from the same vial, 100% of women with carcinoma and 80% with high grade SIL would have been examined. To achieve this high sensitivity in the entire population of 9174 women would have required the referral of about 7% of the population. The combined screening strategy would have performed marginally better than optimized conventional screening with referral of any abnormal cytology (ASCUS and above). CONCLUSIONS. A cervical carcinoma screening technique which uses a single sample for cytopathology and HPV testing to triage equivocal diagnoses may be promising if it proves to be cost-effective.