Cytological DNA methylation for cervical cancer screening: a validation set

Linghua Kong, Linhai Wang, Ziyun Wang, Xiaoping Xiao, Yan You, Huanwen Wu, Ming Wu, Pei Liu, Lei LiFrontiers in Oncology 2023 · DOI

Abstract

The host EPB41L3 and JAM3 gene methylation assay in cervical cytology had favorable diagnostic accuracy for CIN2+ and was highly specific for residual CIN2+ lesions. The methylation assay is a promising triage tool in hrHPV+ women, or even an independent tool for cervical cancer screening. The methylation status in cervical cytology could also serve as a prognostic biomarker. Its role in detecting endometrial carcinomas is worthy of further exploration.

Introduction

Cervical cancer screening currently relies on two primary methods: cervical cytology (Pap smears) and high-risk human papillomavirus (hrHPV) testing. However, both have significant drawbacks. Cytology is subjective and prone to human error, while hrHPV testing, though highly sensitive, lacks specificity—many women test positive for HPV but will never develop significant lesions, leading to unnecessary anxiety and invasive follow-up procedures.

There is an urgent need for objective, molecular biomarkers that can accurately distinguish between a transient HPV infection and a true precancerous lesion (CIN2+). DNA methylation—an epigenetic change where methyl groups are added to DNA, often silencing tumor-suppressor genes—offers a promising, stable, and objective alternative.

Methods and Materials

This study was a prospective validation trial involving 307 participants at a gynecologic oncology unit in China. The research team focused on the methylation status of two host genes: EPB41L3 (Erythrocyte Membrane Protein Band 4.1 Like 3) and JAM3 (Junctional Adhesion Molecule 3).

Liquid-based cervical swabs were collected from patients scheduled for surgery (conization or hysterectomy). These samples underwent both the DNA methylation assay (using TaqMan-based real-time PCR) and standard hrHPV genotyping. Results were compared against confirmed histological findings from the surgical specimens, providing a definitive "ground truth" for diagnostic accuracy.

Flow diagram of the study cohort selection and final pathology.

The study utilized a newly collected validation dataset of 307 eligible patients recruited between June and September 2019. This cohort included patients with benign conditions (n=41), CIN2/3 (n=101), cervical cancer (n=66), and other tumors, such as ovarian or endometrial malignancies (n=99).

Key Results

The most striking takeaway is the assay's specificity (96.7%). In a screening context, high specificity reduces "false alarms," ensuring that only women at genuine risk undergo invasive biopsies or surgery. While methylation sensitivity (68.7%) was lower than hrHPV testing (86.1%), the Positive Predictive Value (PPV) was significantly higher (0.963 vs. 0.780).

Comparison of primary diagnosis, surgical procedures, and actual cervical pathology.

The assay also performed well for cervical adenocarcinoma (ADC), a harder-to-detect cancer subtype, with a sensitivity of 73.7% and specificity of 92.7%. Furthermore, positive cervical DNA methylation was strongly associated with endometrial carcinoma (OR 15.5), suggesting that methylated DNA from the upper reproductive tract can "shed" into the cervix, potentially opening a door for a non-invasive endometrial cancer screen.

Discussion and Insights

The assay was exceptionally good at identifying residual disease. In patients who had already undergone a biopsy that removed the lesion, the methylation signal disappeared (100% specificity for identifying those without residual CIN2+). This suggests it could be used post-procedure to confirm a "clean" result. Additionally, the assay maintained favorable sensitivity (56.3%) and high specificity (96.9%) even in patients who tested negative for hrHPV, suggesting it could catch rare HPV-negative cancers that standard screening might miss.

Limitations and Future Work

The authors acknowledge several limitations, including cohort bias due to the hospital setting and small sample sizes for specific subgroups like ovarian cancer. Future work will involve testing the effectiveness of the assay in broader, community-based populations to see if the high specificity holds up in a low-prevalence environment. The researchers also suggest that the role of this assay in detecting endometrial carcinomas deserves a dedicated, larger-scale exploration.

Conclusion

Researchers have validated a DNA methylation assay targeting the EPB41L3 and JAM3 genes as a highly specific tool for detecting high-grade cervical lesions (CIN2+). While less sensitive than standard hrHPV testing, the assay’s superior specificity makes it a powerful candidate for triaging HPV-positive women and potentially screening for endometrial carcinomas.

Summary updated

Abstract

The host <i>EPB41L3</i> and <i>JAM3</i> gene methylation assay in cervical cytology had favorable diagnostic accuracy for CIN2+ and was highly specific for residual CIN2+ lesions The methylation assay is a promising triage tool in hrHPV+ women, or even an independent tool for cervical cancer screening. The methylation status in cervical cytology could also serve as a prognostic biomarker. Its role in detecting endometrial carcinomas is worthy of further exploration.
BibTeX
@inproceedings{kong2023cytological,
  author = {Linghua Kong and Linhai Wang and Ziyun Wang and Xiaoping Xiao and Yan You and Huanwen Wu and Ming Wu and Pei Liu and Lei Li},
  title = {{Cytological DNA methylation for cervical cancer screening: a validation set}},
  booktitle = {Frontiers in Oncology},
  year = {2023},
  doi = {10.3389/fonc.2023.1181982},
}
Figure