2.1 Patient Samples and Definitions

To ensure the inclusion of a population-based series of non-vaccinated women, we selected women aged 30–63 years who participated in the national PBS program in the North of the Netherlands between July 2020 and March 2022, as the first group of vaccinated women would not be eligible to participate until 2023. Their samples were obtained using a self-sampling device at home (Evalyn Brush, Rovers Medical Devices B.V., Oss, the Netherlands), and after transporting to the central laboratory at UMCG resuspended and preserved in 20 mL of ThinPrep medium (Hologic Inc., Marlborough, MA, USA). All samples in this study tested positive for hrHPV using the Cobas 4800 clinical hrHPV test (Roche Diagnostics, Alameda CA, USA) within the regular PBS. Self-collected samples were stored at room temperature at the UMCG and were processed for DNA extraction and genotyping between August 2023 and February 2024. Histopathological classification of the tissue/biopsies taken by the gynaecologist during colposcopy was used as the gold standard for defining disease stage. Histology and cytology results were retrieved from the the Dutch Nationwide Pathology Databank (PALGA). Histology was categorized as CIN0, CIN1, CIN2, CIN3 and cancer. Women with two consecutive normal cytology results (at primary screening [t = 0 months] and ~6 months of follow-up screening [control smear], and therefore not referred to the gynaecologist for colposcopy) were considered as hrHPV-positive women without disease (i.e., NILM).

2.2 DNA Extraction

DNA extraction was performed on the same samples used for routine hrHPV testing in the PBS program. The Promega Maxwell RSC Viral Total Nucleic Acid Kit (Promega Corporation, Madison, WI, USA) was used according to the manufacturer protocol. The ThinPrep vial was vortexed briefly to resuspend cells, following which 300 µL of the sample was transferred to a 1.5 mL microtube. Subsequently, 300 µL of the lysis buffer and 30 µL of proteinase K were added to the sample and vortexed for 10 s. The sample was then incubated at 56°C for 10 min. Following the sample incubation, the entire volume of lysate was transferred to the Maxwell RSC cartridge and processed in the Maxwell RSC Instrument using the Viral Total Nucleic Acid method. The isolated DNA was then eluted in 50 µL Nuclease-Free Water [14]. For quality control of the extracted DNA, DNA from the first 100 samples were also amplified using the BIOMED-2 PCR protocol [15]. All 100 samples contained sufficient amplified DNA and revealed PCR-products of at least 400 bp indicating high quality of DNA. Because all the included samples also showed sufficient DNA in the Cobas 4800 clinical hrHPV test, we assumed that the rest of the samples would also have sufficient good-quality DNA. The INNO-LiPA HPV Genotyping Extra II assay (see next paragraph) also includes an internal control for sufficient DNA input and all samples were valid.

2.3 HPV Genotyping

The INNO-LiPA HPV Genotyping Extra II assay (Innogenetics N.V., Ghent, Belgium) was utilized to identify 32 different HPV genotypes individually (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 26, 53, 66, 70, 73, 82, 6, 11, 40, 42, 43, 44, 54, 61, 81, 62, 67, 83, and 89), providing individual results for each genotype, and ensuring high analytical sensitivity. These analyses were conducted following the instructions of the manufacturer. This assay involves an initial PCR amplification step, followed by a line probe assay based on reverse hybridization [16]. In short, 10 µL of the isolated DNA was subjected to a short PCR fragment assay using biotinylated consensus primers, amplifying a 65 bp fragment in the conserved L1 region of multiple alpha HPV types [16]. Additionally, a primer set targeting the human HLA-DPB1 gene was included to assess the quality of extracted DNA. To identify the HPV type-specific sequences, the resulting biotinylated amplicons were denatured and hybridized to filter-strips each containing various HPV-specific oligonucleotide probes. The INNO-LiPA assay, including sample incubation, stringent wash, and color development, was performed using the TENDIGO system according to manufacturer's protocol (INNO-LiPA; Fujirebio.com). The results of each sample were independently scored manually by two researchers using the INNO-LiPA HPV Genotyping Extra II Reading Card. Intensely colored line probes were scored as 1, while those faintly colored but still visible to the naked eye and after capturing a picture were scored as 2. A score of 0 was assigned when there was no evidence of color after capturing the pictures. For this study, scores of 1 and 2 were considered positive for the specific genotype line probe, while score 0 was considered negative. If multiple infections were present within the same sample, we assessed each genotype individually. Samples that did not show positive results for any of the 32 genotype line probes were classified as HPV-negative according to the INNO-LiPA test. Contamination of amplification products was prevented by all standard precautions, using separate laboratories for pre- and post-PCR handling in a NEN-ISO15189 accredited laboratory.

2.4 Statistical Analysis

All statistical analyses were conducted using R version 4.3.1, with the following packages: gtsummary, ggplot2, ComplexUpset, binom, and nnet. Firstly, the prevalence of all 32 HPV genotypes was computed for each histologic diagnosis, cytology results, and age group. The prevalence of each genotype was calculated both when it appeared alone (single infection) and when it appeared with other types (multiple infections) in the total population. This was done overall, as well as for CIN2+ (CIN2, CIN3, and cancer) and CIN3+ (CIN3 and cancer). The results were plotted in a stacked bar plot. This plot shows the segments representing single and multiple infections stacked on top of each other. The total length of the bar represents the total prevalence. To visualize the most common patterns (either single or multiple) of HPV infection in our population, we applied upset plots (a plot representing a unique pattern or combination of categories, in this case HPVs, and the length of the bar indicating the prevalence of the specific combination). Secondly, we explored the prevalence of HPV using three different classifications: (1) HPV vaccine types (HPV16/18 [2vHPV], HPV6/11/16/18 [4vHPV], and HPV6/11/16/18/31/33/45/52/58 [9vHPV]). (2) Based on the cancer risk classification established by the International Agency for Research on Cancer (IARC) [11], we employed a hierarchical system where groups were mutually exclusive, with each participant assigned to only one group. HrHPV genotypes were categorized into five risk-based groups: (a) HPV16, (b) HPV18/45, (c) HPV33/31/52/58/35, (d) HPV39/51/59/56/68, and (e) negative for hrHPV. (3) We used the number of hrHPV types within each sample, categorizing participants as positive for one, two or three and/or more hrHPV genotypes. The prevalence of each classification per histological diagnosis and age group was estimated with 95% confidence intervals (95% CI), using the normal approximation of the binomial distribution. Finally, to evaluate the relationship between high-risk genotypes included in the 9vHPV vaccine, age, and histologic diagnosis, and to determine if age distribution influences the association between hrHPV of the 9vHPV and histologic diagnosis, we performed univariable and multivariable logistic regression analyses. Age and histologic diagnosis were considered independent variables, while the presence of high-risk genotypes included in the 9vHPV (16/18/31/33/45/52/58) was the dependent variable.

3.1 Study Population and Overall HPV Genotype Prevalence

In total, 1200 hrHPV-positive women from the Dutch cervical cancer screening program were included in our population-based series for HPV genotyping. Among them, 400 cases were classified as NILM, 315 as CIN0/1, 164 as CIN2, 308 as CIN3, and 13 as cervical cancer. 56% of the total population were between 30 and 39 years old, and most CIN3+ cases (51%) occurred within this age group (Supporting Information S1: Table 1).

In the total cohort, at least one of any 32 HPV genotypes was identified in 1185 out of the 1200 samples (98.8%). The top three most prevalent HPV genotypes were HPV16 (33%), HPV31 (18%), and HPV52 (17%) (Figure 1). 52% (624/1200) of the samples had multiple infections (considering all 32 genotypes evaluated). The most prevalent pattern of HPV infection was a single infection with HPV16 (n = 180), followed by a single infection with HPV52 (n = 74), and a single infection with HPV51 (n = 36) (Figure 2). Although half of the samples had co-infections, the frequency of a specific combination of HPVs was relatively low. The combination of HPV31 and HPV52 was the most frequent co-infection, observed in 17 participants, whereas only 5 participants had co-infections of HPV16 and HPV31, and only 4 participants had co-infections of HPV52 and HPV66 (Figure 2).

3.2 Prevalence of HPV Genotypes by Histology Diagnosis

The prevalence of HPV16 increased with the severity of the underlying lesion. Although HPV31 was the second most common type, its prevalence was relatively high in high-grade lesions but low in cancer cases (only one case (out of 13) was positive for HPV31, and it was in a co-infection with HPV18) (Supporting Information S1: Table 2). The prevalence of HPV16 was 57% in CIN2+ cases and 33% in < CIN2 cases. Similarly, HPV16 prevalence was 67% in CIN3+ cases and 20% in < CIN3 cases. Approximately 60% of HPV16 infections in both CIN2+ and CIN3+ cases were single infections. Most of the low-risk genotypes were presented as multiple infections only (Figure 1).

3.3 Distribution of hrHPV by Different Classifications

When using the IARC classification, we observed that 1116 (93%) of the total samples tested positive for hrHPV. Of the 84 (7%) hrHPV-negative samples, four of them were CIN3 (1.3%, 95% CI: 0.42–3.5). HPV16 was present in 66% (95% CI: 61–71) of CIN3 cases and 85% (95% CI: 54–97) of cancer cases. Additionally, we observed an increasing prevalence of HPV16 as the severity of the lesions increased. Conversely, there was a decreasing tendency in the prevalence of HPV59/39/68/51/56 with the increasing severity of the lesions (p < 0.001) (Table 1).

With the number of hrHPV genotypes, we observed that overall, 63% (95% CI: 60–66) of the samples had only one hrHPV genotype, and around 30% of the samples had two or more hrHPV types. The distribution of the number of hrHPV genotypes per histologic diagnosis seems quite consistent (Table 1).

Using the hrHPV vaccine types classification, 40% (479/1200) of the samples were positive for HPV types present in the 2vHPV, with 64% (305/479) of these cases classified as CIN2+ and 48% (231/479) as CIN3+. 42% (507/1200) of the samples were positive for HPV types present in the 4vHPV, showing similar distribution of cases. Additionally, 75% (905/1200) of the samples had HPV types present in the 9vHPV, with 49% (447/905) classified as CIN2+ and 33% (302/905) as CIN3+. In other words, HPV16/18 (2vHPV) is present in 23% of NILM, 27% of CIN0/CIN1, 45% of CIN2, 71% of CIN3, and 92% of cervical cancer cases. When including all HPV types covered by the 4vHPV (HPV6/11/16/18), these percentages remain the same as for the 2vHPV (25% for NILM, 31% for CIN0/CIN1, 46% for CIN2, 72% for CIN3, and 92% for cervical cancer). When considering all types included in the 9vHPV (HPV6/11/16/18/31/33/45/52/58), these percentages increase to 60% for NILM, 69% for CIN0/1, 88% for CIN2, 94% for CIN3, and 100% for cancer cases (Table 1).

3.4 Prevalence of HPV Genotypes by Age

The prevalence of HPV16 and HPV31 decreased with age, while other genotypes, such as HPV18, decreased from ages 30 to 40, then increased, and decreased again after age 50 (Supporting Information S1: Table 3). The percentage of women aged 30-34 was higher among those with an HPV16 infection, regardless of whether the infection was single or a coinfection (Supporting Information S1: Figure 1).

Analyzing hrHPV prevalence using the IARC classification, we observed that the prevalence of other hrHPV genotypes different than HPV16 either increased or remained stable with increasing age (Table 2). Additionally, the proportion of individuals negative for hrHPV genotypes (0 hrHPV) increased with age, while the presence of one or two hrHPV genotypes remained relatively stable throughout life. In contrast, the occurrence of three or more high-risk genotypes decreased from ages 30 to 49 but increased again starting at age 50 (Table 2). Using the classification of all vaccine types (2vHPV, 4vHPV, and 9vHPV), the percentage of these types tended to decrease with age (Table 2).

3.5 Association Between With hrHPV Vaccine Types, Age, and Histologic Diagnosis

When we performed logistic regression to evaluate the association between 9vHPV types, age and histologic diagnosis (Table 3), we found that in a model with only age, for each additional year of age, the risk of having 9vHPV types decreased (OR: 0.97; 95% CI: −0.95,0.98). In a model with histological diagnosis (< CIN3 vs. CIN3 + ) women with CIN3+ were more likely to have the 9vHPV types compared to those < CIN3 (OR: 7.37; 95%CI: 4.71,12.19). In a model combing age and histological diagnosis, the association between CIN3+ and 9vHPV types remained (OR: 6.85; 95% CI: 4.36, 11.37).