Globally, cancers of the female reproductive organs (vulva, vagina, cervix, uterus, ovaries) represent approximately 15% of all female cancer cases and deaths (1). In February 2024, the International Agency for Research on Cancer (IARC) reported that 2022 witnessed roughly 1,473,427 new cases and 680,372 deaths in female reproductive organs cancers worldwide (2). Data from China’s National Central Cancer Registry and the World Cancer Research Fund International (WCRF International) for 2022 revealed a persistent increase in new cervical cancer cases in China, with an estimated 150,700 new cases, positioning China as the country with the highest incidence of new cervical cancer cases globally. During the same period, there were approximately 55,700 cervical cancer-related deaths in China (3, 4).

Over 90% of cervical cancers are caused by human papillomavirus (HPV) infection (5). Currently, high-risk HPV (hrHPV) DNA testing is recommended as the primary screening method for cervical cancer in China, with liquid-based cytology testing (LBC) used in resource-limited settings (6). Following infection with hrHPV, patients undergo epigenetic changes that can lead to the development of cervical intraepithelial neoplasia (CIN) 1-3 and, ultimately, cervical cancer (7). HrHPV infection induces mild to moderate cellular abnormalities, which are histologically classified as CIN 1 or CIN 2, whereas most HPV-positive cases are transient and resolve spontaneously, with only a minority progressing to CIN2 or worse (CIN2+). High-grade lesions CIN 2 and CIN 3, both of which are classified by the American Society for Colposcopy and Cervical Pathology (ASCCP) as primary screening targets in precancers, may take decades to progress to cervical cancer (7–9). Thus, HPV positivity for some individuals, especially for women in their 20s, may only signify an infection stage rather than a precancerous state, while misinterpretation of HPV positivity potentially causes undue anxiety, unnecessary treatment, and even increased risk of obstetric complications among screened women (6, 10). Despite high specificity and immediate risk assessment capabilities, cytology’s lower sensitivity and negative predictive value (NPV) limit its effectiveness in long-term risk prediction compared to hrHPV DNA testing (11). While introduction of HPV vaccination has notably reduced cervical cancer caused by vaccine-covered hrHPV genotypes (12–14), the rising incidence of non-vaccine hrHPV genotypes and HPV-unrelated adenocarcinomas is still noteworthy (15, 16). Therefore, within the framework of current guidelines recommending screening methods and comprehensively promoting HPV vaccination, there remains a continuing need to identify appropriate biomarkers as auxiliary screening indicators.

Epigenetic modification mainly involves alterations in gene expression levels driven by non-sequence-based changes. Dysregulated DNA methylation, one of the primary epigenetic mechanisms, contributes considerably to cancer development, invasion, and metastasis (8, 17). Accumulation of DNA methylation in specific genes, acting as an early indicator of malignant transformation, offers new avenues for early prevention, recurrence monitoring, and prognosis assessment (18, 19). Paired box gene 1 (PAX1) methylation emerged as a triage tool for detecting CIN3+ among hrHPV-positive women, demonstrating comparable clinical performance with LBC and superior testing efficacy compared to HPV16/18 (20, 21). Additionally, junctional adhesion molecule 3 (JAM3) methylation showed good diagnostic accuracy for CIN2+, serving as an effective triage marker for hrHPV-positive patients and stratification method for patients with atypical squamous cells of undetermined significance (ASC-US) (22–24). The combination of these dual genes has also been validated as a tool for cervical cancer screening and stratification. In a multicenter prospective study, PAX1/JAM3 methylation demonstrated superior clinical efficacy, especially the area under the curve (AUC), for testing CIN3+ compared to cytology, leading to a reduction in unnecessary colposcopy referrals (25). Furthermore, elevated PAX1/JAM3 methylation levels correlated significantly with persistent HPV infections lasting over three years, suggesting their potential utility in identifying continuous HPV infection (26). In this report, we conducted a real-world study to evaluate the triage efficacy of PAX1/JAM3 methylation (CISCER) test among hrHPV-positive patients.

This study evaluated the clinical efficacy of PAX1/JAM3 methylation detection in triaging HPV16/18-positive and non-16/18 hrHPV-positive patients within a cohort of 436 enrolled women, including 396 who were hrHPV-positive. Our findings demonstrated robust performance of CISCER test with sensitivity and specificity exceeding 92%. In addition, PAX1 and JAM3 methylation detection, whether alone or in combination, achieved a more balanced sensitivity and specificity profile and higher AUC values compared to these of cytology. Importantly, our study highlighted that the CISCER tests successfully identified all cervical cancer cases among the enrolled patients, including two adenocarcinoma cases that were hrHPV- negative.

Cervical cancer stands as the most predominant malignancy affecting the female reproductive system and a leading cause of female cancer-related deaths worldwide, accounting for approximately 6.9% of all cancer cases and 8.1% of fatalities as of February 2024 (2). Recent data from the National Cancer Center of China reveal a rising trend in cervical cancer incidence and mortality rate among Chinese women from 2010 to 2018 (4). Adding to the problem, cervical cancer screening coverage remains inadequate, with only 36.8% of women aged 35-64 undergoing screening between 2018 and 2019 in China, particularly in rural areas, as well as in the central and western regions (31). While cytology offers high specificity, its limited sensitivity (50-70%) poses challenges in detecting high-grade lesions and contributes to potential underdiagnosis issues (32, 33). Since 97% of precancerous lesions are HPV-positive, there is limited additional information that can be provided by performing both cytology and HPV DNA testing compared to HPV DNA testing alone (34, 35).

DNA methylation has garnered international recognition as a pivotal tool in cervical cancer screening, triage, and management, due to its reliable performance in efficiently detecting precancerous lesions (7, 36, 37). The latest expert consensus released by the Tumor marker committee of Chinese anti-cancer association in 2024 proposes that DNA methylation testing could play a role in cervical cancer screening management and post-treatment monitoring (38). Previous research, as evidenced by study including conditional inference tree model, had shown that methylation assay was more pertinent to the pathological diagnosis than cytology in diagnosing both high- and low-grade cervical lesions, and the assay can effectively differentiate high-grade cervical lesions regardless of cytology test (39). DNA methylation testing could mitigate overtreatment in patients diagnosed as CIN2+, where a positive result for specific methylation markers indicates a heightened risk of short-term progression to cervical cancer (7). In this study, we observed that over 35% of CIN2+ patients had negative cytology or low-grade lesions (ASC-US and LSIL) but positive CISCER results, including 3 cancer cases. Additionally, more than 50% of CIN2- patients was abnormal cytology (ASC-US+) but negative for CISCER tests. Moreover, CISCER’s immediate CIN3+ risk for negative results in hrHPV-positive women was much lower than LBC. These finding indicates that CISCER tests held a lower missed diagnosis rate in high-grade lesions and also resulted in fewer referrals for colposcopy in women with low-grade lesions. This highlights the potential of PAX1/JAM3 methylation to complement hrHPV tests in triaging HPV-positive women rather than relying on cytology detection.

Cervical cancer is frequently diagnosed in populations lacking adequate screening (40). Despite hrHPV DNA test being recommended as the primary screening tool for cervical cancer (6, 41), not all cases are associated with HPV infection, such as gastric-type endocervical adenocarcinoma (G-EAC). In our study cohort, two adenocarcinoma patients tested negative for hrHPV DNA test, but were successfully identified by CISCER tests. Specifically, these two cases exhibited notably high levels of PAX1 and JAM3 methylation (Patient 1: ΔCtPAX1 = 2.52, ΔCtJAM3 = 5.28, LBC: NILM; Patient 2: ΔCtPAX1 = 1.82, ΔCtJAM3 = 3.72, LBC: adenocarcinoma). Thus, DNA methylation tests in our study provided more comprehensive identification of cervical cancer patients compared to hrHPV DNA tests, consistent with previous findings (42). Therefore, PAX1/JAM3 methylation detection could not only reduce the number of patients who were missed or misdiagnosed by cytology, but also offers a highly sensitive marker panel for cancer, especially those unrelated to HPV infection.

We observed that CISCER had higher sensitivity in women aged≥50 and higher specificity in women aged<30. An analysis of cervical cancer trends among Chinese women from 1990 to 2019 reveals that the highest incidence and mortality rates were observed in the 50-54 age group, followed by women aged 55 and older (43). This pattern may be influenced by hormonal changes in postmenopausal women, which lead to genital tract atrophy and adhesions and then complicate the collection of adequate samples during gynecological examinations (44–46), which may cause anxiety in healthy women. Even worse, the squamous-columnar junction of the cervix shifts upward for elderly women, causing common sites of cervical cancer and precancerous lesions to retract into the cervical canal. This anatomical change often results in missed or inaccurate diagnoses during cervical cytology, colposcopy biopsy, or endocervical curettage (ECC), leading to delays in the initiation of appropriate treatment (47). These changes have introduced significant challenges for cervical cancer screening in postmenopausal women. The high sensitivity of CISCER tests in our study suggests its effectiveness in accurately identifying high-risk patients, which is essential for optimizing cervical cancer screening strategies for postmenopausal women.

Efforts to eliminate cervical cancer in China include extensive HPV vaccination campaigns nationwide, aiming at increasing vaccination rates among women (48, 49). A national survey identified HPV 52, 58, 53, 16, and 51 as the predominant genotypes, varying significantly by age groups and geographic locations in China (50). HPV vaccination efforts expand, necessitating effective cervical cancer screening methods for vaccinated and non-vaccinated individuals and also those infected with non-vaccinated hrHPV genotypes. A long-term follow-up of Dutch women vaccinated with the bivalent vaccine in adolescence revealed high-grade cervical lesions were still detectable in women who were vaccinated early, and the main types of HPV infection were HPV52, 59, 51, 58, and 33, which were non-vaccinated HPV genotypes (12). We observed consistent performance of PAX1 and JAM3 methylation in screening hrHPV-positive patients with CIN2+ regardless of HPV16/18 or non-16/18 hrHPV infections, and PAX1/JAM3 methylation levels were comparable between high-grade cervical lesions caused by non-16/18 hrHPV and HPV16/18. Given the direct correlation where higher levels of methylation correspond to more advanced cervical disease (51), it suggested that the severity of cervical disease did not vary based on hrHPV genotypes once high grade lesions had emerged. Hence, cervical diseases caused by non-16/18 hrHPV genotypes are also noteworthy, as women may not experience reduced severity of cervical lesions due to infection with these genotypes. Methylation tests hold the ability to detect high-grade cervical lesions caused by non-vaccinated hrHPV genotypes, which underscores the potential for them to perform well in screening programs as more vaccinated individuals enter the screening cohort in the future.

Limitations of our study included its simple size, which might limit the generalizability of the findings due to the lack of extensive research data supporting the results. In addition, the histopathological results relied on biopsies from patients referred to colposcopy outpatients rather than surgical pathological results, potentially impacting the accuracy of the histopathological assessments. Future multicenter prospective studies with larger sample sizes and rigorous clinical evaluations are needed to further validate the clinical utility of DNA methylation testing in cervical cancer screening.

Our findings indicate that PAX1/JAM3 methylation test significantly outperforms cytology in triaging hrHPV-positive patients. This could help mitigate the issue of missed diagnoses in postmenopausal women and address fertility preservation concerns in younger women. Furthermore, PAX1/JAM3 methylation test could be valuable for detecting cervical lesions associated with non-vaccine HPV types.