Cervical cancer (CC) is the only cancer that the World Health Organization (WHO) has proposed to eliminate, intending to reduce the incidence of CC to below 4/100,000 by the year 2030 [1]. Squamous cell carcinoma (SCC) of the cervix is the most common type of cervical cancer, accounting for 75 % of cases [2]. The development of SCC is a long process that includes human papillomavirus (HPV) infection [3], [4], cervical intraepithelial neoplasia (CIN) grade 1 (CIN1) lesion, CIN grade 2 (CIN2) lesion, CIN grade 3 (CIN3) lesion and eventually cancer [5]. WHO-recommended interventions include HPV vaccination, CC screening and treatment of cervical lesions. Screening-positive patients underwent colposcopy. Given the high immediate risk of CIN3 lesions, with more than 30 % of CIN3 lesions [6] progressing to CC within 30 years, guidelines were published to treat patients diagnosed with CIN3 after CC screening [7]. Screening and treatment of patients with cervical lesions [8], [9] have greatly reduced the incidence of CC. However, a subset of CIN3 lesions were downgraded in postoperative pathological results, suggesting that CIN3 would regress.

A proportion (14.1 %-19.4 %) of patients diagnosed with CIN3 on colposcopy or histology had their postoperative pathology downgraded to inflammation or CIN1 (CIN1-) [10], [11], [12]. Notwithstanding the presence of CIN3 lesions evident on colposcopy or identified on biopsy histology, residual lesions were found to be CIN1-, as indicated by postoperative pathology results [13]. There were possible explanations for the observed downgrades. Firstly, the lesions were localized and the most severe lesions had already been removed at the time of the biopsy. Secondly, a considerable interval occurred between the colposcopy and the surgery, which could have initiated an immune response after the biopsy, resulting in the regression of the lesions to CIN1-[14]. Alternatively, the CIN3 lesions localized in the deep cervical canal were not removed, which is highly unlikely [15]. It was possible that CIN3 lesions could be cured by immune responses or regress spontaneously [16], but colposcopy or histopathology could not ascertain this. To mitigate the risk of CC, the most prudent approach was to intervene in all CIN3 lesions, mainly by cold knife conization (CKC) and loop electrosurgical excision procedure (LEEP). Despite advances in surgical techniques, cervical surgery remained associated with surgical complications and adverse obstetric outcomes [17], [18]. Therefore, biomarkers were needed to identify CIN3 lesions that would be regressed to CIN1- on the postoperative pathology to reduce overtreatment.

Unfortunately, there were currently no reliable biomarkers that could accurately differentiate regressing CIN3 lesions from persistent or progressing CIN3 lesions. Virological HPV DNA testing and/or morphological cytology, as traditional screening methods, had limited accuracy in predicting the progression of CIN3 lesions. HPV16 infection was the strongest predictor of CC in women. However, CIN lesions might still regress after infection with HPV16, and therefore, HPV16 could not determine the developmental trend of CIN lesions [19]. Similar to virological HPV DNA testing, cytology could not differentiate the outcome of CIN3 lesions based on morphology. Methylation was an epigenetic modification that played an important role in tumorigenesis and progression. Methylation was a promising biomarker that was not limited by subjective factors inherent in cytology, revealing lesion severity at the molecular level. It could be used not only to screen for CC but also to guide individualized management to avoid overtreatment of CIN2/CIN3 lesions.

Paired box gene 1 methylation (PAX1m) had been reported to triage in women with CIN3 [20] before the CKC procedure. In our previous study [21], zinc finger protein 671 methylation (ZNF671m) had superior efficiency in molecular-based screening methods, including HPV16/18 and PAX1m, and the specificity of the ZNF671m test was significantly higher than that of the Thinprep cytologic test (TCT). We observed that the levels of ZNF671m were more concentrated in the CC group, whereas they were more widely distributed in the CIN3 group. We speculated that the level of ZNF671m might be able to distinguish regressing CIN3 lesions from persistent ones at the screening stage. For ethical reasons, all CIN3 patients still needed surgery. Therefore, this study evaluated the correlation between ZNF671m and CIN3 regression by examining the postoperative pathological downgrading of CIN3 patients. This information would help patients and physicians make more personalized treatment decisions.