We compared clinical performance of p16/Ki-67 dual-stained cytology and HPV genotyping, via different algorithms-alone, or in combination with cytology-to identify cervical intraepithelial neoplasia grade 2 or worse (CIN2+) and grade 3 or worse (CIN3+) in women referred to colposcopy. We included 492 cervical specimens (134 normal, 130 CIN1, 99 CIN2, 121 CIN3, 8 cancers) randomly selected from 1158 specimens with valid conventional cytology, HPV (cobas 4800 HPV Test) and biopsy results. Dual-stained cytology was retrospectively performed (CINtec PLUS assay) on PreservCyt material; slides were read by a cytologist and confirmed by two pathologists, blinded to cytology, biopsy, and genotyping results. Sensitivity and specificity (95% confidence intervals in parentheses) of dual-stained cytology to detect CIN2+ and CIN3+ were compared with other screening tests available for the same women. Positivity rate for dual-stained cytology increased with histological severity; 30.6% in normal, 41.5% in CIN1, 72.7% in CIN2, 86.8% in CIN3, and 87.5% in cancer. Dual-stained cytology alone had lower sensitivity than HPV testing for CIN2+ [80.7% (75.0-85.6) vs 89.9% (85.3-93.5)] and CIN3+ [86.8% (79.7-92.1) vs 92.3% (86.2-96.2)]. However, corresponding specificity values were higher [64.0% (57.9-69.8) vs 56.1% (49.8-62.1) for CIN2+; 54.0% (48.7-59.2) vs 44.4% (39.2-49.6) for CIN3+]. Combining dual-stained cytology with an ASC-US abnormality threshold decreased specificity to 31.4% (25.9-37.4) for CIN2+ and 24.2% (19.9-29.0) for CIN3+. Corresponding values considering an LSIL threshold were 42.8% (36.8-49.0) and 35.0% (30.1-40.1). Dual-stained cytology and HPV testing had similar performance, although the former improved the specificity by 7.9% and 9.6% for CIN2+ and CIN3+, respectively. This article is protected by copyright. All rights reserved.
This article is protected by copyright. All rights reserved.