Clinical Epigenetics · 2024;16:108 · Multicenter Prospective Study · China

PAX1m/JAM3m Triage in Non-16/18 hrHPV-Positive Women: A Multicenter Prospective Study in China

PAX1m/JAM3m 双基因甲基化在非 16/18 型高危 HPV 阳性女性中的分诊效能——中国多中心前瞻性研究
1,851 women with histology · sensitivity comparable to cytology (84.8% vs 90.1%), specificity far higher (88.5% vs 26.7%), colposcopy referrals cut 57.2%
Chen X, Jin X, Kong L, Liou Y, Liu P, Dong Z, Zhou S, Qi B, Fei J, Chen X, Xiong G, Hu Y, Liu S, Zhou J, Shou H, Li L (correspondence)
Multicenter · 6 hospitals · China · CISCER® DNA Methylation Kit (NMPA-approved) · doi: 10.1186/s13148-024-01731-w
84.8%
Sensitivity · PAX1m/JAM3m
88.5%
Specificity · PAX1m/JAM3m
0.866
AUC · highest of all triage
98.5%
NPV · negative safety
−57.2%
Colposcopy referrals

1Background & Objective

  • Burden: non-16/18 hrHPV types drive much HSIL+/cancer in Asia; triage still relies on cytology.
  • Cytology limits: subjective, lab-dependent, low specificity (26.7%) → colposcopy overreferral.
  • Objective: validate CISCER® PAX1m/JAM3m for CIN3+ triage in non-16/18 hrHPV+ women.

2Study Design & Cohort

30,084
screened · 6 hospitals
4,735
non-16/18 hrHPV+
1,851
with histology
2
prospective cohorts
6
hospitals
42y
median age
HPV genotyping+ LBC cytology+ PAX1m/JAM3m
METHY3 · 2020.11–2021.12
14 mo
METHY4 · 2022.05–2022.10
6 mo
  • Cohorts: METHY3 + METHY4, opportunistic screening; lab blinded.
  • Endpoint: CIN3+ detection; comparator LBC ≥ASC-US; colposcopy/surgery histology.
  • Statistics: ROC/AUC, McNemar; SPSS 26 / R 4.1.2.
  • Inclusion: ≥18 y, intact cervix, no severe immunodeficiency.
  • Assay: bisulfite conversion + qPCR (GAPDH control, ABI 7500).
  • Outcome: 151 CIN3+ (131 CIN3 + 20 cancers) among 1,851.
82.3%
Normal/CIN1
9.6%
CIN2
7.1%
CIN3
1.1%
Cancer (n=20)

3Key Results — Signal & Performance

PAX1/JAM3 ΔCt distributions by pathology and cytology
Fig. 2 PAX1/JAM3 ΔCt by histology & LBC grade — rises with severity (P<0.05).
0%25%50%75%100% Sensitivity Specificity P/J(+) LBC≥ASC-US P/J(+) LBC≥ASC-US 84.8 90.1 88.5 26.7
Fig. 3 Se vs Sp by triage (data: Table 2) — methylation: high both; cytology: high Se, very low Sp.
Odds ratio for CIN3+ PAX1m/JAM3m JAM3m PAX1m LBC ≥ASC-US 42.6 38.4 29.1 3.3 OR (95% CI) vs marker-negative · P < 0.001 · data: Table 3
Fig. 4 OR for CIN3+ (Table 3) — P/J(+) 42.6 vs LBC(+) 3.3.
Dose–response: ΔCt falls as lesions progress — all 20 cancers were PAX1m/JAM3m(+); positive = ΔCt PAX1 ≤ 6.6 or ΔCt JAM3 ≤ 10.0.

4CIN3+ Detection & Referral Burden

MethodSens. %Spec. %AUCNPV %
PAX1m/JAM3m (+)84.888.50.86698.5
LBC ≥ ASC-US90.126.70.58496.8
LBC or P/J (+)98.024.80.61499.3
OR for CIN3+: P/J(+) 42.6 (27.1–69.6) vs LBC(+) 3.3 (2.0–5.9), P<0.001.
TriageReferred% of 1,851
PAX1m/JAM3m (+)32317.45%
LBC ≥ ASC-US1,38274.66%
LBC or P/J (+)1,42777.09%
Methylation alone keeps the lowest referral burden.
LBC ≥ ASC-US
74.7%
PAX1m/JAM3m
17.5%
Referral rate — P/J(+) triage cuts referrals by 57.2%.

5Cancer Safety & Risk-Based Management

20 / 20
cancers caught by P/J(+)
100%
16 / 20
caught by LBC ≥ASC-US
80% — 4 missed
42.6
OR for CIN3+ if P/J(+)
98.5%
NPV if P/J(−)
  • Missed by cytology: 3 SCC + 1 ADC — all four were P/J(+).
  • P/J(+) → direct colposcopy (high CIN3+ odds); P/J(−) → safe follow-up (NPV 98.5%).
  • Standalone: adding cytology gives no triage advantage — methylation alone is the better strategy.
Clinical Significance