Relationship between p16/ki67 immunoscores and PAX1/ZNF582 methylation status in precancerous and cancerous cervical lesions in high-risk HPV-positive women

Abstract

Background The risk of cervical cancer progression in high-risk human papillomavirus (HR-HPV)-positive women is associated with cervical lesion severity and molecular heterogeneity. Classification systems based on p16 and Ki67 expression cumulative scores (0–3 each)—p16/Ki67 collectively known as an immunoscore [IS]—are an accurate and reproducible method for grading cervical intraepithelial neoplasia (CIN) lesions. Meanwhile, DNA methylation is an early event in the development of cervical cancer. Hence, this study evaluated the relationship among CIN, p16/Ki-67 IS, and PAX1/ZNF582 methylation. Methods In this study, 414 HPV-positive paraffin-embedded specimens were collected, and PAX1/ZNF582 methylation and the p16/ki67 IS were determined. A total of 43 invalid samples were excluded and 371 were included in the statistical analyses. There were 103 cervicitis, 95 CIN1, 71 CIN2, 89 CIN3, and 13 squamous cell carcinoma (SCC) cases. The association between PAX1/ZNF582 methylation and p16/Ki6 immunohistochemical staining scores was analyzed. Results The ΔCp of PAX1m (PAX1 methylation) and ZNF582m (ZNF582 methylation) decreased with cervical lesion severity (Cuzick trend test, all P < 0.001). The severity of the cervical lesions and p16, Ki67, and p16/Ki67 IS showed an increasing trend (Multinomial Cochran-Armitage trend test, all P < 0.001). The prevalence of PAX1m/ZNF582m increased with an increase in the IS of p16, Ki67, and p16/Ki67 (Cochran-Armitage trend test, all P < 0.001). In cervical SCC, the IS was 5–6, and the PAX1m/ZNF582m was positive. Meanwhile, heterogeneity was observed in CIN lesions: 10 cases had an IS of 3–4 and were PAX1m/ZNF582m-positive in ≤ CIN1; 1 case had an IS of 0–2 and was PAX1m/ZNF582m-positive in CIN2/3. Conclusions Significant heterogeneity was observed in CIN lesions for p16 and Ki67 immunohistochemical staining scores and PAX1/ZNF582 methylation. This may help clinicians personalize the management of CIN based on the predicted short-term risk of cancer progression, minimizing the rate of missed CIN1 diagnoses and incorrect treatment of CIN2/3.