1Background & Objective
- Problem: post-conization residual 5–15% & recurrence 3–10% — TCT/HPV follow-up is limited.
- Objective: evaluate PAX1/JAM3 methylation to predict post-op residual & recurrence.
2Study Design & Cohort
120
HSIL patients
12 mo
follow-up
41.9y
mean age
- Setting: Guilin MCH Hospital, Jan 2023 – Dec 2024; conization (CKC 45 / LEEP 75).
- Assay: MSP methylation of PAX1 / JAM3 from pre-op cervical cells.
- Outcome: residual or recurrence within 12 months (22/120).
- HPV: hrHPV+ 87.5% (HPV16 53 · HPV18 26 · other 26).
87.5%
hrHPV+
53
HPV16
26
HPV18
26
other hrHPV
45
CKC conization
75
LEEP conization
37
canal involvement
- Inclusion: first-diagnosed HSIL, conization, pre-op methylation, complete records.
- Exclusion: other cancer, surgical contraindication, pregnancy, immunosuppression, invasive cancer.
- Follow-up: 12 months — TCT/HPV + colposcopy as needed.
- Statistics: χ² / multivariate logistic / ROC.
13
residual
9
recurred
22
adverse outcomes
- Adverse link: methylation(+) significantly associated with post-op residual/recurrence.
- Best predictor: dual-gene methylation — highest OR & prediction accuracy.
- Mechanism: methylation persists in residual disease — a molecular trace.
Univariate → multivariate logistic; OR & ROC evaluation.
3Methylation Positivity
42
single-gene positive
23
dual-gene positive
4Independent Risk Factors
Multivariate: PAX1m OR 5.892 (1.985–17.596, P=0.001) · dual OR 8.654 (2.783–26.927, P<0.001).
5Dual-Gene Prediction
91.3%
Sensitivity
45.4%
Specificity
30.2%
PPV
94.3%
NPV
- High NPV: methylation(−) → low risk of residual/recurrence — safe follow-up.
- AUC 0.765: higher than single-gene tests — practical post-op monitoring.
- Actionable: dual-gene(+) → intensified surveillance & earlier intervention.
Methylation complements TCT/HPV in post-conization follow-up.