HPV Genotype Distribution, Vaccine Protection Gaps, and Co-Infection Network Centrality Among 8,515 Clinical Women in Wuxi, China: A Retrospective Cross-Sectional Study

1. Introduction

Cervical cancer kills approximately 350,000 women annually; GLOBOCAN 2022 recorded 660,000 new cases globally [1], with China contributing an estimated 150,000 new diagnoses and 55,000 deaths in the same year [2]. All invasive cervical cancers are causally attributable to persistent infection with oncogenic HPV types [3,18]; IARC classifies 13 types as definite (Group 1) or probable (Group 2A) human carcinogens [4]. The WHO Global Strategy to Accelerate the Elimination of Cervical Cancer (2020) sets a 90% vaccination coverage target for adolescent girls by 2030 [5], making accurate local genotype data a prerequisite for rational policy design.

Three prophylactic vaccines are licensed in China: bivalent (HPV16/18, 2016), quadrivalent (HPV6/11/16/18, 2017), and nonavalent (HPV6/11/16/18/31/33/45/52/58, 2018) [7,33]. The nonavalent formulation is estimated to prevent approximately 90% of HPV-attributable cervical cancers globally [8,16], but its real-world protection in East Asian populations—where HPV52 and HPV58 predominate rather than HPV16 and HPV18 [6,29]—remains incompletely characterised at the city level.

Regional surveillance studies have been published for Qingdao [9], Weifang [10], Guangzhou [11], Suzhou [12], Changsha [13], Shanghai Jinshan [14], Shanghai Xuhui [15], Chengdu [31,32], and 37 cities nationwide [28], but no large-scale, multi-type HPV epidemiological study exists for Wuxi or the broader Yangtze River Delta outside Shanghai. Wuxi Xishan District had 892,400 registered residents at end of 2024; Xishan People’s Hospital of Wuxi City is its sole tertiary-level general hospital (Grade III Class B), with an annual outpatient volume exceeding 1.14 million visits and centralised cervical HPV testing covering all inpatient and outpatient departments.

A second evidence gap concerns age distribution. Most published series focus on women aged 25–60, leaving adolescents (≤20 years) and older women (≥61 years) poorly characterised [14,15]. Adolescents represent the primary target for pre-exposure vaccination; older women constitute a dual blind spot—ineligible for vaccination (upper limit 45 years in China) and progressively disengaging from screening programmes.

A third gap concerns co-infection network topology. HPV multiple-type infection is associated with substantially elevated risk of high-grade cervical intraepithelial neoplasia [22,35], and the structure of genotype co-occurrence networks may reveal biologically important interactions. Yet formal network centrality analysis is absent from most Chinese HPV surveillance studies.

This study addressed four questions using comprehensive 23-type data from 8,515 consecutive clinical women: (1) the Wuxi-specific HPV genotype prevalence profile, with type-resolved 95% confidence intervals; (2) quantification of vaccine protection gaps across all three licensed formulations; (3) infection characteristics at both age extremes (≤20 and ≥61 years); and (4) co-infection network centrality of HPV52 as a candidate hub genotype, using degree centrality metrics and observed-versus-expected co-infection ratios. This study was conducted and reported in accordance with the STROBE guidelines for cross-sectional studies.

2. Materials and Methods

2.1. Study Design and Setting

This retrospective cross-sectional study enrolled all women who underwent cervical HPV genotyping at Xishan People’s Hospital of Wuxi City between 1 January 2024 and 30 June 2025. The Clinical Laboratory Department centralises all cervical HPV nucleic acid testing hospital-wide; results feed directly into the Laboratory Information System (LIS). Of 8,516 LIS records exported, the final entry was a system-generated null row (all genotype fields negative; age and department blank) and was excluded, leaving 8,515 analysable records.

Data cleaning resolved two coding irregularities: 99 records carrying department code "1163" were verified via the Hospital Information System (HIS) as gynaecology outpatient visits and merged accordingly; 9 records coded as "internal medicine health examination" were merged with the health examination centre group (final n = 1,732). As the exported dataset contained no unique patient identifiers or test dates, the analysis unit was the individual testing record (see Limitations).

Inclusion criteria: female sex, age ≥ 16 years, cervical exfoliated cell sample submitted for HPV genotyping as the primary clinical indication. Exclusion criteria: documented pregnancy, confirmed history of cervical surgery in HIS records, or missing or implausible age data. Eligibility was verified by HIS–LIS record linkage. This study was approved by the Institutional Review Board of Xishan People’s Hospital of Wuxi City (No. 2026-K089-01); individual informed consent was waived under Article 39 of China's Measures for Ethical Review of Biomedical Research Involving Human Subjects given the retrospective design and use of de-identified data.

2.2. Specimen Collection and HPV Genotyping

Qualified gynaecologists or nurses collected cervical exfoliated cells by rotating a cervical brush five times clockwise at the external os. Specimens were immediately transferred to proprietary preservation solution (Shengxiang Biotechnology, Changsha, China), stored at 4 °C, and processed within 24 h of collection.

Genotyping used the Shengxiang 23-type HPV nucleic acid detection kit (fluorescent multiplexed PCR) on QuantStudio™ 5 (Thermo Fisher Scientific, Waltham, MA, USA) and SLAN-96S (Shanghai Hongshi Medical Technology, Shanghai, China) real-time PCR systems. The panel covered 13 high-risk types (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68), 5 intermediate-risk types (HPV26, 53, 66, 73, 82), and 5 low-risk types (HPV6, 11, 42, 43, 81), classified per current IARC criteria [4]—HPV51 as Group 1 (definite carcinogen) and HPV68 as Group 2A (probable carcinogen). Positive and negative controls were included in every run.

2.3. Variable Definitions

Age groups: ≤20, 21–30, 31–40, 41–50, 51–60, and ≥61 years. Infection multiplicity: single-type (one type detected), dual-type (two types), and multiple-type infection (≥2 types).

Three vaccine coverage indices were calculated per formulation: (i) at least partial coverage (APC)—≥1 positive type falls within vaccine targets; (ii) mixed gap—≥1 positive type falls outside vaccine targets; (iii) complete unprotected (CU)—all positive types fall outside vaccine targets. APC and CU are mutually complementary (APC + CU = total positives). For bivalent and quadrivalent vaccines, CU equals mixed gap since these vaccines cover only 2 or 4 types. The nonavalent vaccine targets HPV6, 11, 16, 18, 31, 33, 45, 52, and 58.

Co-infection network metrics: degree of a genotype was defined as the total number of co-infection events in which it appeared (i.e., sum of co-detected partner types across all multiple-infected patients). Normalised degree = degree / [(N − 1) × M], where N is the number of genotypes in the panel (23) and M is the number of multiple-infected patients (585). Expected co-infection frequency for a genotype pair (A, B) was calculated under the independence assumption as: E(A,B) = P(A) × P(B) × N_positive, where P(X) is the marginal prevalence of type X in the full cohort (n = 8,515). The observed/expected (O/E) ratio quantifies enrichment above independence-expected co-occurrence.

2.4. Statistical Analysis

Analyses used SPSS 26.0 (IBM Corp., USA) and Python 3.11 (network metrics). HPV prevalence and 95% confidence intervals (CIs) were estimated by the normal approximation. Between-group prevalence comparisons used Pearson's χ² test, or Fisher's exact test when any expected cell count was < 5. Co-infection pairs were enumerated by pairwise combination of all detected types per patient. All tests were two-tailed; significance threshold α = 0.05.

3. Results

3.1. Overall Prevalence and Genotype Distribution

Among 8,515 women (age 16–87 years; mean ± SD 43.4 ± 11.9 years), 1,944 (22.83%; 95% CI 21.9–23.8%) tested positive for ≥1 HPV type. High-risk, intermediate-risk, and low-risk HPV prevalence were 17.63%, 3.86%, and 5.55%, respectively. All 23 genotypes were detected. HPV52 ranked first (4.82%; 95% CI 4.36–5.27%), followed by HPV58 (3.14%), HPV16 (2.96%), HPV53 (2.50%), and HPV42 (2.08%). Crucially, HPV18 (0.74%; 95% CI 0.56–0.92%)—the second antigen target in all licensed vaccines—was less prevalent than the unvaccinated IARC Group 1 carcinogen HPV51 (1.76%) and the Group 2A carcinogen HPV68 (1.68%). Complete type-specific data with 95% CIs are presented in Table 1.

3.2. Vaccine Protection Gaps

Of 1,944 HPV-positive women, 57.0% (n = 1,109) had at least partial nonavalent vaccine coverage (APC), while 43.0% (n = 835) were completely unprotected (CU)—all detected types absent from 9vHPV—and 60.4% (n = 1,174) harboured ≥1 non-9vHPV type (mixed gap). For 2vHPV and 4vHPV, the CU rate equalled the mixed-gap rate (84.0% and 79.4%, respectively), as these formulations cover only 2 or 4 types (Table 2).

Among the 14 genotypes excluded from 9vHPV, the five highest-burden in positive women were HPV53 (10.96%), HPV42 (9.10%), HPV81 (8.54%), HPV51 (7.72%), and HPV68 (7.36%). HPV51 and HPV68 together accounted for 15.08% of all positive cases—exceeding the combined prevalence-weighted contribution of the 9vHPV-covered types HPV31 (3.66%) and HPV45 (0.98%), and underscoring a carcinogen-coverage asymmetry in current vaccine formulations.

3.3. Age-Stratified Prevalence: U-Shaped Distribution

Prevalence differed significantly across the six age groups (χ² = 123.27, df = 5, P < 0.001; Figure 1), tracing a U-shaped curve. It peaked in the ≤20-year group (52.94%; 95% CI 39.2–66.6%; n = 51; exploratory finding—see Table 3 footnote), fell to a nadir in the 31–40-year group (19.32%), then rose progressively, reaching 34.66% (95% CI 31.3–38.0%) in women aged ≥61 years. High-risk HPV prevalence in the ≥61-year group was 29.90%—the highest among the non-exploratory strata. Triple-or-higher type infection was most frequent at both extremes: 13.73% (7/51) in ≤20 years and 5.28% (41/776) in ≥61 years. The nonavalent CU rate was lowest in the ≥61-year group (31.6%) versus 43.3–47.9% in the 21–60-year groups, consistent with higher background prevalence of 9vHPV-targeted types in older women (Table 3 and Table 4).

3.4. Co-Infection Network Centrality of HPV52

Of 1,944 positive women, 1,359 (15.96%) had single-type and 585 (6.87%) had multiple-type infection: dual 400 (4.70%), triple 134 (1.57%), quadruple 30 (0.35%), quintuple 12 (0.14%), sextuple 6 (0.07%), septuple 1 (0.01%), and octuple 2 (0.02%).

Network centrality analysis across 585 multiply infected women revealed HPV52 as the highest-degree node (degree = 281 co-infection events; normalised degree 0.0218), substantially exceeding HPV58 (degree 212), HPV53 (207), HPV81 (181), and HPV16 (174) (Table 5; Figure 2). HPV52 co-detected with all other 22 panel genotypes, yielding 21 distinct co-infection partner types. Among the five most frequent HPV52-anchored pairs, observed co-infection frequencies exceeded independence-expected values by 6.5–14.8-fold (Table 6), indicating non-random co-occurrence enrichment. HPV52 appeared in 8 of the 15 most frequent dual-type pairs. Notably, the pairs HPV53 + HPV68 (n = 21) and HPV58 + HPV68 (n = 16) featured HPV68, an unvaccinated Group 2A carcinogen.

3.5. Department-Stratified Prevalence

Prevalence differed significantly across the six departments (χ² = 183.0, df = 5, P < 0.001; Table 7). The cervical specialty clinic recorded the highest rate (33.09%; 95% CI 29.2–37.0%), 2.4-fold the health examination centre reference (13.63%; 95% CI 12.0–15.2%). A monotonic gradient was observed tracking clinical referral intensity: cervical specialty clinic (33.09%) > gynaecology OPD (26.37%) > gynaecology inpatient (16.42%) > health examination centre (13.63%). The gynaecology inpatient–health-exam-centre difference did not reach statistical significance (P = 0.056), limiting causal interpretation. An additional 85 cases from low-volume non-gynaecological departments were retained in the overall prevalence numerator but excluded from between-department comparisons; the six analysed departments totalled 8,430 women.

3.6. Risk Stratification

Of 8,515 women, 6,571 (77.17%) were HPV-negative. Among 1,944 positive women: high-risk only 1,201 (14.10%); intermediate-risk only 159 (1.87%); low-risk only 260 (3.05%); high-risk + intermediate-risk 111 (1.30%); high-risk + low-risk 154 (1.81%); intermediate-risk + low-risk 24 (0.28%); all three risk tiers 35 (0.41%).

4. Discussion

Four findings define the clinical and public-health significance of these data. First, HPV52 is the dominant high-risk type across all age strata and displays the highest co-infection network degree centrality, with observed co-infection frequencies 6.5–14.8-fold above independence-expected values—consistent with a hub-genotype architecture. Second, 43.0% of HPV-positive women in this clinical population carry types entirely absent from the nonavalent vaccine, with the IARC-classified carcinogens HPV51 and HPV68 collectively accounting for 15.08% of positive cases. Third, HPV prevalence at both age extremes (≤20 and ≥61 years) substantially exceeds the 31–40-year nadir, identifying two currently underserved prevention populations. Fourth, a 2.4-fold inter-departmental gradient demonstrates that the overall 22.83% prevalence reflects case-mix enrichment; the health examination centre rate (13.63%) provides a closer community-level approximation.

HPV52 dominance and hub-genotype topology. HPV52 displacement of HPV16/18 from the top prevalence position is consistent across East Asian epidemiology [6,9,10,11,12,13,20,27,29,32] and has been attributed to genomic and immunological characteristics that facilitate immune persistence in East Asian host populations [36]. In our network analysis, HPV52 co-detected with all 22 other panel types and achieved a normalised degree of 0.0218—29% above HPV58 (0.0165). Observed co-infection rates for HPV52 with its four most frequent partners were 6.5–14.8× the independence expectation, suggesting co-occurrence enrichment that may reflect either shared exposure windows or impaired type-specific immune clearance. This contrasts with hub-genotype analyses in other populations that implicate HPV16 as the dominant co-infection node [37]. The continued dominance of HPV52 post-9vHPV introduction warrants prospective type-replacement surveillance [10].

Vaccine protection gaps and antigen prioritisation. The 43.0% complete-unprotected rate exceeds figures reported from Suzhou (21.7%) [12], Beijing [38], and several other Chinese cities [17,25,26], reflecting both our broader 23-type panel and the referral-enriched sample. HPV51 (1.76%) and HPV68 (1.68%) together constituted 15.08% of positive cases—individually exceeding the cohort-wide prevalence of 9vHPV-covered HPV31 (0.83%) and HPV45 (0.22%) and collectively surpassing their sum. Both meet IARC carcinogenicity criteria [4,30], and epidemiological studies have attributed 3–5% of cervical cancers globally to HPV51 and HPV68 [39]. Broader-spectrum vaccines incorporating these types are in development [19]; our local frequency and O/E co-occurrence data provide a quantitative rationale for prioritising HPV51 and HPV68 in next-generation antigen selection for the East Asian market.

U-shaped age distribution: implications for prevention policy. The trough at 31–40 years and re-elevation at ≥61 years (HR-HPV 29.90%) replicate a pattern documented in Shanghai [14,15], Weifang [10], and Chengdu [31,32], and mirrors the age-specific multiple-infection pattern reported nationally [21]. Immunosenescence-driven viral reactivation and impaired T-cell-mediated clearance are the most widely cited mechanisms [24,40]; the lower 9vHPV CU rate in women ≥61 (31.6% versus 43.3–47.9% in younger groups) is consistent with higher background carriage of vaccine-targeted types accumulated over longer exposure histories. Given that Chinese recommendations cap HPV vaccination at age 45 and that elderly women progressively disengage from screening, routine high-risk HPV screening continuation beyond age 65 and colposcopy referral for HPV-positive women aged ≥61 are warranted, consistent with current Chinese cervical cancer screening guidelines [23]. The ≤20-year finding (52.94%; n = 51) is exploratory; all cases were clinical attendees with inherent selection bias, and the wide 95% CI (39.2–66.6%) reflects the small sample. Nevertheless, a similarly elevated rate in this stratum has been reported in Shanghai [14,15] and Chengdu [31], and the HPV16 prevalence of 15.69%—the highest of all age strata—is biologically plausible given rapid high-risk exposure shortly after sexual debut [41]. These data support clinician-initiated HPV risk assessment for sexually active patients below age 25 in clinical settings, without implying a need to revise national screening age thresholds.

Department gradient and external validity. The 2.4-fold prevalence gradient directly tracking clinical referral intensity confirms that overall hospital-level HPV rates are strongly confounded by case-mix. The health examination centre prevalence (13.63%) closely approximates community-level estimates from Suzhou (10.2%) [12] and Chengdu health-check cohorts (~14%) [32], suggesting that the underlying community HPV burden in Wuxi is not markedly elevated relative to adjacent cities, and that the 22.83% overall figure reflects referral enrichment rather than an intrinsically higher population prevalence. This distinction is crucial for policy translation: local cervical cancer prevention strategies should be calibrated to the community baseline (approximately 13–14%), not the referral-enriched clinical rate.

Limitations. This single-centre retrospective cross-sectional study has six limitations. First, generalisation to the broader Wuxi population requires caution. Second, retrospective data preclude adjustment for key confounders: vaccination history, number of sexual partners, contraceptive use, and smoking status. The absence of vaccination status data is a particular limitation given that this study period (2024–2025) coincides with expanding 9vHPV uptake in China; we cannot determine whether the 43.0% CU rate would differ between vaccinated and unvaccinated subgroups. Third, the ≤20-year group (n = 51, 0.60% of cohort) consists entirely of clinical attendees; prevalence is exploratory. Fourth, the cross-sectional design precludes inference on infection persistence, temporal sequence, or causal direction in co-infection pairs. Fifth, the exported dataset lacked unique patient identifiers and test dates, so the analysis unit was the individual testing record rather than the individual patient; a small number of women who submitted repeat samples during the study period were therefore counted more than once, which may have marginally influenced prevalence estimates. Sixth, LIS diagnostic field completion was 42.3%, preventing systematic HPV–cytology linkage; integration of TCT and histopathology results, alongside HPV genotyping, would substantially strengthen the clinical interpretation of these epidemiological findings [34].

5. Conclusions

This 23-type retrospective cross-sectional study of 8,515 consecutive clinical women in Wuxi establishes four actionable conclusions. HPV52 is the dominant high-risk genotype across all age strata, achieves the highest co-infection network degree centrality (degree 281; normalised degree 0.0218), and displays 6.5–14.8-fold enriched co-occurrence with its top partners—consistent with hub-genotype behaviour. The nonavalent vaccine leaves 43.0–60.4% of positive women incompletely protected; the IARC carcinogens HPV51 (Group 1) and HPV68 (Group 2A) constitute the principal unmet need, each exceeding the prevalence of 9vHPV-covered HPV31 and HPV45. A U-shaped age-prevalence distribution, with high-risk HPV rates of 29.90% in women ≥61 years and exploratory rates of 45.10% in women ≤20 years, reveals two currently underserved populations at the extremes of the age spectrum. Department-level stratification shows that the health examination centre baseline (13.63%) approximates community-level prevalence, while the 22.83% overall rate reflects case-mix enrichment from referral patients. These data support HPV51 and HPV68 as priority antigens for next-generation vaccine design and call for sustained cervical cancer screening coverage across the full clinically active age spectrum.