The p16/Ki-67 dual-staining is increasingly applied to increase diagnostic accuracy in detecting high-grade cervical lesions, including cervical intraepithelial neoplasia Grade 2 (CIN2+) and CIN3+. To compare the diagnostic performance of p16/Ki-67 dual-staining with the human papillomavirus (HPV) tests in the triage of women with atypical squamous cells of undetermined significance (ASC-US) and low-grade squamous intraepithelial lesions (LSIL) cytology results. Publications before April 27, 2024, were identified through PubMed, Embase, Web of Science, and Cochrane Library. The studies have head-to-head comparison of p16/Ki-67 dual-staining and HPV testing in detecting high-grade cervical lesions. Two researchers independently screened articles by title, abstract, and full text. The GRADE and QUADAS-2 tool was used for quality evaluation. The pooled sensitivity and specificity for CIN2+ and CIN3+ were estimated using random effects models, with results and heterogeneity assessments presented in forest plots. Pretest-posttest probability (PPP) plots were constructed to evaluate the detection rates of CIN3+ in ASC-US and LSIL patients. Twenty-one studies with 6394 participants were included. The pooled specificity of p16/Ki-67 dual-staining was higher than that of HPV tests (CIN2+: 0.73 95% confidence interval [CI 0.65-0.80] versus 0.41 95% CI 0.33-0.50; CIN3+: 0.61 95% CI 0.53-0.69 versus 0.33 95% CI 0.23-0.45). Summary receiver operating characteristic curve analysis demonstrated p16/Ki-67 dual-staining had better diagnostic accuracy for CIN2+ than HPV tests (area under the curve: 0.88 95% CI 0.85-0.90 versus 0.79 95% CI 0.75-0.82). Pretest-posttest probability (PPP) plots highlighted the superior performance of p16/Ki-67 dual-staining for colposcopy referrals in LSIL patients. P16/Ki-67 dual-staining offers greater specificity for detecting CIN2+/CIN3+ in ASC-US and LSIL triage, particularly for LSIL. It holds potential as an alternative to HPV tests in resource-limited settings or LSIL cases, warranting further research to refine its application in diverse populations.
Zhang et al. (Mon,) studied this question.