Population-based prevalence, type- and age-specific distribution of HPV in women before introduction of an HPV-vaccination program in Denmark


  • Conflict of Interest: SKK is a member of the Scientific Steering Committee for Merck. She has received travel and research grants from Merck and Sanofi Pasteur MSD. CM has received support for travel and conference participation from Merck. GB and MW formerly worked for SP MSD that commercializes Gardasil in Europe.


Knowledge about the prevalence of human papillomavirus (HPV) on a population level is important. We conducted a large population-based study in Denmark to determine the overall and age-specific HPV prevalence, and HPV type distribution in women. Liquid-based cytology samples (SurePath®) were collected consecutively. HPV testing was performed with Hybrid Capture 2 (HC2; Digene) (high-risk and low-risk probes), and LiPA (Innogenetics) was used for genotyping. We analyzed samples from 11,617 women; 94.0% had normal cytology, 4.3% atypical squamous cells of undetermined significance or low-grade squamous intraepithelial lesion and 1.6% had high-grade squamous intraepithelial lesion (HSIL). The HPV prevalence was 26.4% with a peak in women 20–24 years (50.2%) and then decreased without a second peak in older women. Among the youngest women (15–19 years), 14% had HPV 16/18 and 16% had HPV 6/11. Prevalence of high-risk HPV types increased from 19.2% in women with normal cytology to 100% in women with cervical intraepithelial neoplasia grade 3 (CIN3)/cervical cancer. HPV 16 was the most prevalent type (6.0% of all women), and was also the most prevalent in women with HSIL (35.1%) and CIN3 (53.2%). Other common HPV types in women with CIN3 included HPV 52, 51, 31, 33 and 18. HPV 16/18 alone was present in 23% of CIN3 lesions and 67% of cervical cancers, and HPV 16/18 together with other high-risk HPV types was present in 41% of CIN3 lesions. This suggests that an efficacious HPV 16/18 vaccine will have a substantial preventive potential in the general female population. © 2008 Wiley-Liss, Inc.

Cervical cancer is the second most common cancer in women worldwide; more than 490,000 women are diagnosed with invasive cervical cancer each year.1 There is a strong causal relationship between human papillomavirus (HPV) infection, which is sexually transmitted,2 and the development of cervical cancer, with 99.7% of cervical cancers containing HPV DNA.3 The lifetime risk of acquiring HPV infection is high as the majority of sexually active individuals will be infected at some time in their lives.4 HPV infections are usually transient, but HPV infection can persist and can progress to high-grade cervical intraepithelial neoplasia, and, in the worst situation, to cervical cancer.5

Approximately 40 HPV types are known to infect the anogenital tract, of which more than 15 are thought to be associated with the development of cervical cancer (high-risk (HR) HPV types).6 It has been estimated that 2 of these types, HPV 16 and 18, are associated with 70–75% of invasive squamous cell cervical cancers and 80% of adenocarcinomas of the cervix.7, 8 HPV 16 and 18 are also associated with the development of other neoplasias of the genital tract such as vulvar and vaginal cancers and the preceding precancerous lesions.9 The low-risk (LR) HPV types 6 and 11 are mainly associated with low-grade cervical lesions and the majority of genital warts (90%).10

In the early 1990s work started on the development of prophylactic vaccines against specific HPV types. Data on the effectiveness of these vaccines has been exceptionally encouraging. A quadrivalent HPV vaccine containing types 6, 11, 16 and 18 has been licensed in the US, Europe, Australia and Latin America. Recent results from large-scale clinical trials have shown that the vaccine is well tolerated and highly efficacious against cervical, vaginal and vulvar precancerous lesions and genital warts.11, 12 A bivalent HPV 16 and 18 vaccine, recently licensed in Australia and Europe, has also produced encouraging clinical efficacy results.13

To assess the potential impact of the papillomavirus vaccines, it is necessary to determine the overall burden of HPV infection before large-scale vaccination programs are implemented. We report the results of a population-based study conducted in Denmark to determine the prevalence of HPV infection for different grades of cervical neoplasia as determined by cytology and histology, as well as the overall age-specific prevalence of HPV infection in the general population of women undergoing cervical cancer screening. In addition, we assessed the type-specific HPV prevalence to estimate the preventive potential of an HPV 16/18 VLP vaccine in preventing CIN2 or worse.

Material and methods

Study population

A population-based study of prevalent HPV infection was conducted in Copenhagen, Denmark. In Denmark, cervical cancer screening is recommended every 3 years starting from age 23 years, but younger women are often screened opportunistically.14 From October 2004 to June 2005 consecutive liquid-based cytology samples were collected from the Department of Pathology at the Copenhagen University Hospital in Hvidovre, Denmark. This department receives and analyses all cervical cytology and histology samples obtained from both routine and opportunistic screening in the Greater Copenhagen area. In this area, all cervical samples are routinely collected by general practitioners and gynecologists using the SurePath® liquid-based cytology system (TriPath Imaging, Burlington, NC) and sent to the Department of Pathology at Hvidovre Hospital within 2 days of collection. Upon arrival, all samples are assigned a microscopy number and then processed for cytological examination. Certified cytotechnicians and cytopathologists make the diagnosis. In the event of subsequent biopsy or conization, the tissue is also sent to the Pathology Department at Hvidovre Hospital.

For the present study, the sample remaining after the cytological examination was collected by the Department of Virus, Hormones and Cancer at the Danish Cancer Society on a weekly basis, marked with a study number (the microscopy number) and sent to the Medical Virology Department, University Hospital of Tübingen, Germany for HPV testing, within 2–3 weeks after the smear was taken. A total of 12,829 samples were collected by the Danish Cancer Society during the study period. Of those, 802 samples were excluded as they were follow-up smears after an abnormal cervical cytology/histology diagnosis within 12 months of the index sample. The cytology results of the samples that were excluded were as follows: 0.3% inadequate, 65.1% normal, 24.9% atypical squamous cells of undetermined significance (ASCUS)/low-grade squamous intraepithelial lesion (LSIL) and 9.7% high-grade squamous intraepithelial lesion (HSIL).

Finally, 9 samples were excluded because the sample had been taken more than 7 weeks before the HPV test. Thus the Medical Virology Department at the University Hospital of Tübingen received 12,018 samples for HPV analysis. One hundred sixty-seven samples were excluded due to technical errors or inadequacy of the sample. Furthermore, 234 samples were excluded from the analyses as they were duplicate samples from the same women (in this case the first sample was used). Thus, the final study population included 11,617 women (with 1 sample each) with a mean age of 36.4 years (range: 15–93 years).

Cytological and histological diagnoses

The Central Population Register

In Denmark, every citizen has a unique 10-digit personal identification number which is registered in the computerized Central Population Register. This register contains information on date of birth and gender. The identification numbers, which are universally used in Danish society, including by the public administration, enable reliable linking between different registers.

The Pathology Data Bank

This is a nationwide computerized pathology register containing information on all cervical cytology (organized and opportunistic, normal and abnormal) and all cervical histology (biopsies, conizations and hysterectomies) performed in Denmark. An on-line, real-time system enables communication between the various pathology departments in Denmark and the Pathology Data Bank.

The cohort in the present study was linked to the Danish Pathology Data Bank using the unique personal identification number and the microscopy number as key identifiers. Through this procedure, we retrieved for each study participant information on all cervical cytology and histology in the year prior to the date of obtaining the index sample, the cytology result of the index sample as well as information on subsequent cervical histology related to an abnormal index sample.

HPV testing

Hybrid Capture 2 Testing

All cervical samples received at the laboratory in Germany were marked only with the study number, and thus HPV testing was performed without knowledge of the clinical or demographic data related to the sample. Briefly, the cells were denatured and a sample was analyzed by Hybrid Capture 2 (HC2; Digene Corporation, Gaithersburg, MD), according to the manufacturer's instructions. HC2, which is based on an immunochemiluminescent reaction, uses 2 probe mixtures to detect HPV: a HR probe that detects 13 individual HR HPV types (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68) and a LR probe that detects 5 individual LR HPV types (6, 11, 42, 43 and 44).15, 16 Samples were considered positive if they attained or exceeded the FDA-approved threshold of 1.0 pg/ml of HPV DNA, which corresponds to 1.0 RLU.

HPV genotyping

HPV genotype testing was performed using a line probe assay (LiPA; Innogenetics, Gent, Belgium). Total DNA was isolated from the remaining denatured cervical samples that had been found to be positive by HC2 using the MagnaPure device (Roche, Indianapolis, IN) and analyzed using the INNO-LiPA v2HPV prototype assay according to the manufacturer's instructions. All PCR manipulations were performed in a separate room according to ISO 15189 diagnostic guidelines. An aliquot of the amplified PCR product was hybridized to the LiPA hybridization strips using an Auto LiPA device and the resulting strip was analyzed using a flatbed scanner and the LiRAS prototype software, which reports the patterns and intensity of positive bands as gray tone values between 0.1 and 1.0. The LiPAv2 test can identify 24 HPV genotypes (6, 11, 16, 18, 31, 33, 35, 39, 40, 42, 43, 44, 45, 51, 52, 53, 54, 56, 58, 59, 66, 68, 70 and 74). Samples that were found to be positive in the LiPA assay, but for which a specific HPV type could not be identified where included in the analyses as HC2 positive and LiPA positive samples of undetermined HPV type.

Statistical analysis

The overall prevalence and type-specific HPV prevalence and their 95% confidence intervals (CI) were estimated for all women, and were also estimated by age group (15–19, 20–24, 25–29, 30–34, 35–39, 40–44. 45–49, 50–54, 55–59, 60–64 and ≥ 65 years), cytology result (normal, ASCUS/LSIL and HSIL) and histology result (cervical intraepithelial neoplasia (CIN1, CIN2, CIN3) and cervical cancer). Logistic regression, adjusted for age, was used to evaluate the role of infection with HPV 16/18 alone (i.e., without the presence of other HR HPV types), infection with other HR HPV types alone (i.e., without the presence of HPV 16/18) and infection with both HPV 16/18 and other HR HPV types (with HPV 16/18 negative, other HR HPV negative as reference group) for the development of different grades of cytological and histological abnormalities. The association was measured as an odds ratio (OR) with 95% CI. HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66 and 68 were categorized as high risk types, whereas HPV types 6, 11, 40, 42, 43, 44, 54, 70 and 74 were categorized as low risk.6


Among the 11,617 women in the study, a total of 10,918 (94.0%) had normal cytology, 494 (4.3%) had a cytological diagnosis of ASCUS or LSIL and 188 (1.6%) were diagnosed with HSIL. The remaining 17 women had samples that were not adequate for analysis. These 17 samples as well as 1 other sample where the HC2 low risk result was missing were excluded from the relevant analyses. Overall the results from the HC2 assay showed that 3,069 women (26.4%; 95% CI: 25.6–27.2) were HPV positive (HR and/or LR HPV) with 8.2% (95% CI: 7.7–8.7) harboring LR HPV types and 22.8% (95% CI: 22.0–23.5) HR HPV types. HR HPV types only were detected in 2,113 women (18.2; 95% CI: 17.5–18.5), LR HPV types only in 425 (3.7; 95% CI: 3.3–4.0) and both LR and HR HPV types were detected in 531 (4.6; 95% CI: 4.2–5.0).

HPV prevalence according to age

The overall prevalence of any type of HPV peaked in women aged 20–24 years (50.2; 95% CI: 48.2–52.2) and then decreased rapidly until about 40 years old. After this the prevalence did not change significantly with increasing age (Fig. 1a). Stratification of the HC2 results by type (HR and LR) and age showed that about 40% (39.5, 95% CI: 28.9–50.2) of women in the age category 15–19 years were infected with a HR HPV type (Fig. 1a). The prevalence of HR HPV infections was highest in women 20–24 years of age (44.7; 95% CI: 42.7–46.7) and then decreased markedly with increasing age. A similar age-prevalence relationship was observed for LR HPV types with a peak prevalence of 27.2% (95% CI: 17.5–36.9) in women 15–19 years of age. Infections with HR HPV types only were the most prevalent infections in all age groups (data not shown). No significant second prevalence peak was observed in older women for either HR HPV or LR HPV types.

Figure 1.

(a) Prevalence (%) of high-risk and low-risk HPV types according to age among women 15–93 years of age. (b) Prevalence (%) of HPV types 16/18 and HPV 6/11 according to age among women 15–93 years of age.

The prevalence of HPV 16/18 infection in women aged 15–19 years old was 13.6% (95% CI: 6.1–21.0) and this increased to 18.5% (95% CI: 16.9–20.0) in women 20–24 years old, and then decreased markedly with age (Fig. 1b). Although the prevalence was higher in the oldest age group (≥ 65 years) (3.0; 95% CI: 0.8–5.8), compared to that in the 45–64 years old group (1.0; 95% CI: 0.4–2.8) it was not statistically significantly higher. The HPV 6/11 prevalence peaked in the youngest age group (15–19 years), where 16.1% (95% CI: 8.1–24.0) were positive, and then decreased strongly with increasing age.

HPV prevalence in relation to diagnosis

Overall HPV prevalence

HC2 testing (Table I) revealed that 22.9% of women with normal cervical cytology were HPV positive with 19.2% having HR HPV types and 7.4% having LR types (3.8% had LR types only, 15.5% had HR types only and 3.7% had both HR and LR types). As the severity of the cervical lesion increased, the prevalence of HPV infection also increased, reaching 86.7% in women with a HSIL diagnosis. This increase was mainly due to an increase in infections with HR HPV types alone; 15.5, 56.9 and 76.1% in samples from women with normal, ASCUS/LSIL and HSIL cytology diagnoses, respectively.

Table I. Prevalence (%) of HPV Infection (HC2) and Prevalence of Single and Multiple HPV Types (LIPA) by Cytological and Histological Diagnosis
 Cytological diagnosisHistological diagnosis
Normal (n = 10,918)ASCUS/LSIL (n = 494)HSIL (n = 188)CIN1 (n = 86)CIN2 (n = 36)CIN3 (n = 141)Cervical cancer (n = 6)
  • SD, standard deviation.

  • 1

    Figures do not add up to total HPV pos. as some samples could not be typed, and thus it was not possible to define whether the sample contained a single or multiple HPV types.

Mean age (yrs) (±SD)36.3 (±12.4)30.0 (±10.1)33.2 (±11.3)27.6 (±6.6)28.0 (±8.4)29.4 (±7.0)34.8 (±8.1)
HPV pos. (any type)2,501 (22.9)405 (82.0)163 (86.7)80 (93.0)35 (97.2)141 (100.0)6 (100.0)
HR HPV2,091 (19.2)390 (78.9)163 (86.7)77 (89.5)35 (97.2)141 (100.0)6 (100.0)
LR HPV812 (7.4)124 (25.1)20 (10.6)25 (29.1)5 (13.9)23 (16.3)1 (16.7)
 LR HPV only410 (3.8)15 (3.0)03 (3.5)000
 HR HPV only1,689 (15.5)281 (56.9)143 (76.1)55 (64.0)30 (83.3)118 (83.7)5 (83.3)
 LR + HR HPV402 (3.7)109 (22.1)20 (10.6)22 (25.6)5 (13.9)23 (16.3)1 (16.7)
Single HPV type1 (% among infected)1,024 (9.3)144 (29.1)67 (35.6)24 (27.9)15 (41.7)54 (38.3)4 (66.7)
Multiple HPV types1 (% among infected)1,324 (12.1)248 (50.2)93 (49.5)54 (62.8)19 (52.8)85 (60.3)2 (33.3)
Multiple LR HPV types1 (% among infected)137 (1.3)18 (3.6)4 (2.1)3 (3.5)1 (2.8)2 (1.4)1 (16.7)
Multiple HR HPV types1 (% among infected)1,035 (9.5)209 (42.3)86 (45.7)43 (50.0)18 (50.0)80 (56.7)1 (16.7)

Histological examination was performed in 378 women (Table I). As for the cytology, the overall prevalence of HPV infection and the percentage of infections with HR HPV types alone increased with increasing severity of the histological diagnosis. Among the 86 women diagnosed with CIN 1 lesions, 93.0% were positive for HPV (3.5% LR types only, 64.0% HR types only and 25.6% HR and LR types). Among the 36 women who were diagnosed with CIN 2 lesions, 97.2% were HPV positive (83.3% had HR types only and 13.9% had both HR and LR types). All 141 women with CIN 3 lesions and the 6 women with cervical cancer were HPV positive (83.7 and 83.3% had HR types only and 16.3 and 16.7% had both HR and LR types for CIN 3 and cervical cancer lesions, respectively). None of the CIN 2 or CIN 3 lesions or cervical cancers had only LR HPV types (Table I).

Prevalence of single and multiple HPV infections

Multiple HPV infections were detected in 14.5% (95% CI: 13.7–15.0) of all women and in 54.3% (52.5–56.0) of the women with HPV positive samples (data not shown). Multiple HPV infections were more prevalent than single HPV infections for all cytological and histological grades, except for cervical cancer, but there were only 6 samples (Table I). Multiple HPV infections were more frequently observed in women with abnormal cytology (ASCUS/LSIL (50.2%) and HSIL (49.5%)) samples than in women with normal cytology (12.1%). However, multiple HPV infections (including multiple HR HPV infections and multiple LR HPV infections) types did not seem to be further related to the cytological or histological grade. Multiple HR HPV infections were more common in samples from younger women, both when considering all women and when considering HPV positive women only, e.g., the prevalence of multiple HR HPV types in HPV positive samples from 20 to 24 years old women was 60.9% (95% CI: 58.0–63.9) whereas it was only 23.5% in HPV positive samples from women 65 years or older (95% CI: 3.4%–43.7%) (data not shown).

Type-specific HPV prevalence

The most common HPV type detected in all women was HPV 16 (6.0; 95% CI: 5.5–6.4). Although significantly less frequent than HPV 16, HPV 31, 52 and 51 were also relatively common (prevalence range: 4.4–4.5%). These were followed by HPV types 68, 18, 39, 53, 66, 33, 45, 56 and 6, all with a prevalence ranging between 1.9 and 2.8%, whereas the prevalence of the other HPV types was 1.5% or lower.

HPV 16 was the most prevalent type independent of cytology diagnosis (4.8% in normal, 20.6% of ASCUS/LSIL and 35.1% of HSIL samples) (Table II). As the severity of the neoplasia increased, the percentage of samples containing HPV 18 also tended to increase. HPV 31, 51 and 52 were common in both LSIL lesions (14.2, 18.0 and 14.8%, respectively) and in HSIL (17.6, 13.3 and 22.3%), whereas HPV 45 was less common (7.4%) The most common LR type associated with all grades of cytology was HPV 6, being present in 1.6% of normal samples, 8.3% of ASCUS/LSIL samples and 4.3% of HSIL samples. Finally, HPV 11 was present in 0.5% of the women with normal cytology, 1.4% of the women with ASCUS/LSIL lesions and in none of the women with HSIL lesions.

Table II. Prevalence of Specific HPV Types1 and Groups of Types According to Cytology and Histology Diagnosis
 Prevalence (%) (95% CI)
Cytology diagnosisHistology diagnosis
Normal (N = 10,918)LSIL (N = 494)HSIL (N = 188)CIN 1 (N = 86)CIN 2 (N = 36)CIN 3 (N = 141)Cancer (N = 6)
  • 1

    Samples initially tested with HC2 (high-risk and low-risk probes). Samples positive to one or both probes were genotyped using Lipa PCR.

HR types
164.8 (4.4–5.2)20.6 (17.1–24.2)35.1 (28.3–41.9)12.8 (5.7–19.9)27.8 (13.2–42.4)53.2 (45.0–61.4)33.3 (0–71.1)
182.2 (1.9–2.5)9.3 (6.8–11.9)14.9 (9.8–20.0)9.3 (3.2–15.0)8.3 (0–17.4)13.5 (7.8–19.2)33.3 (0–71.1)
313.8 (3.5–4.2)14.2 (11.1–17.3)17.6 (12.1–23.0)15.1 (7.6–22.7)33.3 (17.9–48.7)19.9 (13.3–26.4)16.7 (0–49.5)
331.6 (1.4–1.9)10.1 (7.5–12.8)11.7 (7.1–16.3)7.0 (1.6–12.4)25.0 (10.9–39.2)16.3 (10.2–22.4)16.7 (0–49.5)
350.8 (0.7–1.0)3.0 (1.5–4.6)5.3 (2.1–8.5)2.3 (0–5.5)2.8 (0–8.2)5.7 (1.0–9.5)0
392.2 (1.9–2.5)9.3 (6.8–11.9)8.0 (4.1–11.9)12.8 (5.7–19.9)07.1 (2.9–11.3)0
451.8 (1.5–2.0)5.9 (3.8–7.9)7.4 (3.7–11.2)8.1 (2.4–13.9)2.8 (0–8.2)9.9 (5.0–14.9)0
513.6 (3.2–3.9)18.0 (14.6–21.4)13.3 (8.4–18.2)19.8 (11.4–28.2)11.1 (0.8–21.4)18.4 (12.0–24.8)16.7 (0–49.5)
523.7 (3.3–4.0)14.8 (11.7–17.9)22.3 (16.4–28.3)18.6 (10.4–16.8)27.8 (13.2–42.4)22.0 (15.2–28.8)0
531.9 (1.6–2.2)11.1 (8.4–13.9)5.9 (2.5–9.2)10.5 (4.0–16.9)03.5 (0.5–6.6)0
561.7 (1.5–2.0)8.3 (5.9–10.7)3.2 (0.7–5.7)12.8 (5.7–19.9)13.9 (2.6–25.2)5.7 (1.0–9.5)0
581.1 (0.9–1.3)4.7 (2.8–6.1)3.7 (1.0–6.4)4.7 (0.2–9.1)16.7 (4.5–28.8)5.0 (1.4–8.6)0
591.1 (0.9–1.3)3.4 (1.8–5.1)1.1 (0–2.5)2.3 (0–5.5)02.1 (0–4.5)0
661.9 (1.6–2.1)11.7 (8.9–14.6)7.4 (3.7–11.2)17.4 (9.4–25.5)5.6 (0–13.4)7.1 (2.9–11.3)0
682.4 (2.2–2.7)7.9 (5.5–10.3)7.4 (3.7–11.2)7.0 (1.6–12.4)2.8 (0–8.2)9.9 (5.0–14.9)0
LR types
61.6 (1.4–1.9)8.3 (5.9–10.794.3 (4.1–11.9)11.6 (4.9–18.4)8.3 (0–17.4)6.4 (2.4–10.4)16.7 (0–49.5)
110.5 (0.3–0.6)1.4 (0.4–2.5)01.2 (0–3.4)000
400.4 (0.3–0.6)2.0 (0.8–3.3)02.3 (0–5.5)000
420.8 (0.6–0.9)1.2 (0.3–2.2)02.3 (0–5.5)000
430.5 (0.4–0.6)1.0 (0.1–1.9)02.3 (0–5.5)000
441.5 (1.3–1.8)2.0 (0.8–3.3)1.1 (0–2.5)1.2 (0–3.4)02.8 (0.1–5.6)0
540.9 (0.8–1.1)2.8 (1.4–4.3)2.7 (0.4–5.0)2.3 (0–5.5)2.8 (0–8.2)0.7 (0–2.1)16.7 (0–49.5)
701.3 (1.1–1.5)6.5 (4.3–9.3)2.7 (0.4–5.0)10.5 (4.0–16.9)8.3 (0–17.4)2.8 (0.1–5.6)0
741.4 (1.1–1.6)3.4 (1.8–5.1)3.2 (0.7–5.7)3.5 (0–7.4)02.1 (0–4.5)0
Groups of types
6/111.9 (1.7–2.2)9.3 (6.8–11.9)4.3 (4.1–11.9)11.6 (4.9–18.4)8.3 (0–17.4)6.4 (2.4–10.4)16.7 (0–49.5)
16/186.6 (6.1–7–0)26.5 (22.6–30.4)46.8 (39.7–53.9)19.8 (11.4–28.2)36.1 (20.4–51.8)63.1 (55.2–71.1)66.7 (29.0–100.0)
6/11/16/187.8 (7.3–8.4)33.6 (29.4–37.8)48.4 (41.3–55.6)29.1 (19.5–38.7)38.9 (23.0–54.8)66.0 (58.1–73.8)66.7 (29.0–100.0)
6/11/16/18/3111.3 (10.7–11.9)43.7 (39.4–48.1)58.5 (55.5–65.6)37.2 (27.0–47.4)66.7 (51.3–82.1)75.9 (68.8–83.0)83.3 (53.8–100.0)
6/11/16/18/31/4512.8 (12.1–13.4)47.2 (42.8–51.6)61.2 (54.2–68.1)45.3 (34.8–55.9)66.7 (51.3–82.1)80.1 (73.6–86.7)83.3 (53.8–100.0)

In women with CIN 1 or CIN 2, HPV 16, 31, 33, 39, 51, 52 and 56 were the predominant types (Table II). However, in samples from women with CIN 3 HPV 16 was by far the most prevalent HPV type (53.2%) compared to the other common HR HPV types, i.e., HPV 31 (19.9%), HPV 33 (16.3%), HPV 51 (18.4%) and HPV 52 (22.0%). The percentage of women with HPV 18 present increased, although not significantly, from 9.3% (3.2–15.0) in CIN1 to 33.3% (0–71.1) in the 6 samples from women with cancer. Among the LR HPV types, HPV 6 and HPV 70 were the most frequent types, and were mainly present in CIN1 (11.6 and 10.5%, respectively) and CIN2 (8.3% for both HPV types) lesions.

The occurrence of respectively HPV 16 and/or HPV 18 alone, HPV 16 and/or HPV 18 together with other HR HPV types and other HR HPV types alone (i.e., without HPV 16/18) in CIN2+ lesions is presented in Figure 2. The prevalence of HPV 16/18 alone increased from 12% in CIN2 to 67% in the cervical cancer cases. The simultaneous presence of HPV 16/18 and other HR types only occurred in CIN2 (26%) and in CIN3 (41%), but not in the 6 cases of cervical cancer.

Figure 2.

Prevalence (%) of HPV 16/18 alone, HPV 16/18 with other high-risk HPV types, and other high-risk HPV types (i.e., without HPV 16/18) in CIN2+ lesions. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

The age-adjusted associations between different grades of cytological/histological abnormalities and different combinations of HPV 16/18 and other HR HPV types are shown in Table III. Compared to women who were negative for HPV 16/18, women infected with HR HPV types other than HPV 16/18 (HPV16/18 negative/other HR HPV positive) had an increased risk of ASCUS/LSIL (OR = 12.0; 95% CI: 9.4–15.2). The risks for ASCUS/LSIL associated with HPV 16/18 alone (HPV 16/18 positive/other HR HPV negative) or together with other HR HPV types (HPV 16/18 positive/other HR HPV positive) were similar. The risk of CIN 1 was nonsignificantly higher in women who were HPV16/18 negative/other HR HPV positive samples (OR = 19.9; 95% CI: 10.6–37.4) than in women who were infected with only HPV 16/18 or with both HPV 16/18 and other HR HPV types (OR = 10.1 (95% CI: 3.8–27.1) and 10.5 (95% CI: 4.6–24.4), respectively). In contrast, the risk of CIN 3 or worse was lowest in women who were HPV16/18 negative/other HR HPV positive (OR = 55.0; 95%: 23.1–130.9) compared to women who were HPV 16/18 positive (OR = 212.7; 95% CI: 87.5–514.1) or women with both HPV 16/18 and other HR HPV types (OR = 193.3; 95% CI: 80.0–467.0). We also performed the same analysis in women who were negative for LR HPV types, and the results showed similar trends (data not shown).

Table III. Age-Adjusted1 Odds Ratios (OR) with 95% Confidence Intervals (CI) for Associations Between Different Categories of High-Risk (HR) HPV Types (HPV16/18 and Other HR HPV Types)2 for Different Grades of Cytological/Histological Abnormalities
  ASCUS/LSIL vs. normalHSIL vs. normalCIN1 vs. normalCIN2 vs. normal≥CIN3 vs. normal
  • 1

    Adjusted for age (as a continuous variable).

  • 2

    Samples initially tested with HC2 (high-risk and low-risk probes). Samples positive to one or both probes were genotyped using Lipa PCR.

HPV 16/18Other HR types     
NegativePositive12.0 (9.4–15.2)19.0 (12.1–29.8)19.9 (10.6–37.4)56.8 (12.9–250.3)55.0 (23.1–130.9)
PositiveNegative11.4 (8.0–16.4)46.3 (27.6–77.8)10.1 (3.8–27.1)51.1 (9.1–288.0)212.7 (87.5–514.1)
PositivePositive11.9 (8.7–16.3)50.7 (30.9–83.3)10.5 (4.6–24.4)69.5 (14.2–341.1)193.3 (80.0–467.0)


Our results show that HPV infection is very common among Danish women. The overall prevalence of HPV, which was 22.9% in women with normal cytology, may have been underestimated as we excluded samples from women with an abnormal cytology/histology within the 12 months preceding inclusion in the study. The impact of this underestimation may have been greater with respect to the low-grade cervical lesions, and thus potentially for HPV 6 and 11. However, our results are similar to those in other recent large-scale studies, for example, studies carried out in the UK, the US and Costa Rica.17–19 The HPV prevalence among women with normal cytology in a previous study we conducted in Denmark20 was 15.9% (95% CI: 15.2–16.6). Our study tested cervical swab samples, using only a HR probe HC2 test. One reason why the prevalence is higher in our study could be that the previous study was performed more than 10 years ago (1991–1993) and HPV prevalence might have been lower then. In addition, it has been suggested that HPV may be a little higher when tested in LBC samples than in cervical swabs.

The prevalence of HPV infection increased with the severity of the cytological grade (reaching 87% in HSIL samples) and also with the severity of the histological diagnosis (reaching 100% in CIN 3 and cervical cancer samples). This is consistent with the large body of evidence that infection with HPV is a necessary cause of invasive cervical cancer worldwide.3, 9 Women diagnosed with ASCUS/LSIL or CIN 1 had evidence of infection with HR as well as LR HPV types, which is consistent with the fact that low-grade neoplasia can be caused by LR types, such as HPV 6 and 11, as well as HR types.21 As expected, none of the women with HSIL cytology, CIN 2/3 or cervical cancer histology diagnoses had evidence of infection with LR HPV types alone.

The most common HPV type in all cytology samples and high-grade histology samples was HPV 16, which is in agreement with many other studies.19, 21–23 Although significantly less frequent than HPV 16, both HPV 31, 52 and 51 were also common in the present study. HPV 31 has previously been found to be widespread in Europe.21, 23 Our finding that HPV 51 is frequently found is in line with several studies19, 21 but is in contrast to a recent pooled analysis based on 294 HPV infections among 5,141 women with normal cytology from 3 European countries (Italy, The Netherlands, Spain).23 HPV type 52 has been found to be associated with the development of cervical cancer predominantly in Asia.24 The differences in the prevalence of individual HPV types observed in our study compared to other studies may be because the LiPA test is more sensitive for some HPV types than other HPV detection methods.25–27

In Europe, the 2 HPV types most frequently associated with cervical cancer are HPV 16 and 18 with HPV 16 being the predominant type in squamous cell carcinomas and HPV 18 the predominant type in adenocarcinomas and adenosquamous carcinomas of the cervix.7, 9 Our study revealed that HPV 16 and/or 18 were present in 6.6% of samples from women with normal cytology, 26.5% of ASCUS/LSIL samples, 19.8% of samples from women with CIN 1 lesions and 63.1% of women with CIN 3 lesions, which is similar to previously published results.7, 19, 21, 22 If we assume a 100% efficacy and coverage of an HPV 16/18 vaccine, our results suggest that the vaccine could prevent at least 23% of all CIN 3 lesions. If it is also anticipated that HPV 16/18 is the causal agent even in the presence of other HR HPV types, the preventive potential could be >60% of CIN 3 cases. Our results corroborate those in a recent study from France,28 and are of importance to help us have a more clear understanding of how much disease can potentially be avoided by vaccinating women against HPV.

We found a clear trend for infections with HPV 16/18 alone being associated with a higher age-adjusted risk of ≥ CIN3 lesions compared to infections with other HR HPV types alone. In addition, we found that the risk of ≥ CIN3 was not higher when the infection was with HPV 16/18 and other HR HPV types together compared to infection with HPV 16/18 alone. The presence of LR HPV types did not change this picture. These results are consistent with recent results from Costa Rica and although the CI were overlapping, the results may provide additional support for the carcinogenic potential of HPV 16 and 18.19

In agreement with most published data, we found that the prevalence of both HR HPV and LR HPV is age-dependent with HPV prevalence decreasing significantly with age.17–19, 29 We did not observe a second peak in the HPV prevalence in postmenopausal women. This is in agreement with the results from other European studies,17, 29 but in contrast to those from studies in Latin American countries like Costa Rica, Mexico, Chile and Colombia.19, 29 The results from our study show that ∼40 and 44%, respectively, of sexually active women 15–19 and 20–24 years of age have evidence of infection with HR HPV types, and 25 and 16%, respectively, have evidence of infection with LR HPV types. We found that 14% of women aged 15–19 years were infected with HPV 16/18 and 16% were infected with HPV 6/11. The corresponding figures for women aged 20–24 years were 19% (HPV 16/18) and 6% (HPV 6/11). Although this gives an estimate of the percentage of women exposed, it is based on a single time point and does not provide any indication of previous exposure, and therefore the percentage of already exposed women is likely to be higher. It is important to determine the age at which individuals first encounter HPV infection in determining at what age vaccination would be appropriate. As the vaccines are prophylactic, vaccination should ideally occur before exposure to HPV (i.e., before sexual debut) to ensure effective prophylaxis. This means that the vaccine would typically be provided through children/adolescent vaccination programs. However, to obtain an effect of the HPV vaccine as soon as possible, it has also been suggested that catch-up vaccination of subjects older than 13–14 years should take place.

The results of our study also show that infection with multiple HPV types is frequent. In agreement with previously published results, we observed that these multiple HPV infections are more frequent in women with cervical neoplasia (∼50%) than in those with normal cervical samples (12.1%), but that the occurrence of multiple HPV infection (and HR HPV infections) is similar for the different grades of cervical neoplasia.19, 30 Our results showed that infection with both HR and LR HPV types is most frequent in women with low-grade lesions (22.1% in ASCUS/LSIL and 25.6% in CIN1). The consequences of multiple HPV infection are not clear, but the results from one study reported that multiple infections (except for those infections including HPV 16) were associated with an increased risk of high-grade cervical lesions and cervical cancer.19 However, another study reported that this risk was comparable for women with single and multiple HPV infection.31 Further studies on the potential role of multiple HPV infections in carcinogenesis are, therefore, warranted.

Denmark was among the first countries in the world to introduce organized screening programs against cervical cancer, but the incidence and mortality rates for cervical cancer in Denmark are still among the highest in Western Europe. The results of the present study indicate a relatively high prevalence of HPV infection compared to some other European countries such as The Netherlands, Spain and Italy.29

The present population-based study included women from the general population in Denmark undergoing organized as well as opportunistic cervical screening. Women in Denmark between 23 and 59 years of age automatically receive an invitation every 3 years to participate in the organized screening program.32 Through this approach, ∼75% of eligible women are routinely screened.32 Opportunistic screening of women younger than 23 years of age is also common. However, because women younger than 23 years may be more likely to have a reason to request or undergo the screening test (as they are not yet included in the organized program), our estimation for the 15–22 year age range may not be an accurate reflection of the true HPV prevalence in this age group. Depending on the nature of the individual's behavior (e.g., greater interest in health or engagement in high risk behavior), the HPV prevalence could be either an over- or an underestimation of the true HPV prevalence.

In conclusion, it is well recognized that HPV infection is the necessary cause of cervical cancer. Until recently, there were no primary prevention strategies available that targeted the direct cause of the disease, but in 2006, the first HPV vaccine was licensed. To understand the overall benefits provided by such a vaccine it is essential to understand the burden of disease caused by HPV. Our study provides valuable baseline knowledge about the prevalence and type distribution of HPV prior to the introduction of HPV vaccination. In addition, our data clearly indicate that HPV 16/18 VLP vaccination is expected to substantially reduce the burden of CIN2 or worse.


We would like to thank Ms. Barbara Holz, Ms. Betti Schopp, Ms. Nazife Kilic, Ms. Nurgül Düzenli, Ms. Annette Rothe at the University Hospital of Tübingen, the study staff at Hvidore Hospital and the Danish Cancer Society for their technical support that made this study possible. We also thank Ms. Margaret Haugh for assistance with preparation of the manuscript, and Ms. France Goullet and Mr. Bennett Lee (SP MSD) for their comments.