Human papillomavirus infection: an Anonymous Prevalence Study in South Wales, UK

Abstract The objective of this study was to describe human papillomavirus (HPV) prevalence in South Wales in relation to age, cytology and social deprivation. This was an unlinked, prospective, anonymous, population-based study. DNA was purified from 1911 liquid-based cytology samples (mean age 37.7 years, cytology 93.2% negative, social deprivation average score 17.9) using quality assured techniques and the presence of virus determined by PCR-Enzyme Immuno Assay (PCR-EIA). 209 (10.9%) samples contained high-risk (HR) HPV infection of which 36.4% had multiple HR-HPV types. The most frequent HR types were HPV 16 (19.6%), HPV 35 (9.5%), HPV 66 (9.2%), HPV 59 (8.5%) and HPV 56 (7.6%). There was a strong association between HPV infection and cytological abnormality. Significantly more HR-HPV infections were detected in women under the age of 30 years (68.9% of all HR-HPV infections Fisher's exact test P=0.0001) compared to 30 years and above. There was no difference in HPV prevalence between different socioeconomic groups. The data presented suggest a different HPV type distribution in South Wales in comparison to that reported for other populations. Main Cervical cancer is the second most prevalent female cancer worldwide (Parkin et al, 2005). HR-human papillomavirus (HPV) infection plays a central role in cervical carcinogenesis, with HPV DNA identified in 99.7% of invasive cervical carcinomas (Walboomers et al, 1999). However, HPV infection is common and in the majority of cases self-limiting. An estimated 70% of infections are transient and cleared within 18 months, with less than 1–2% of high-risk infections resulting in cervical cancer. There is a large discordance between the number of women infected (up to 80% cumulative lifetime risk), and those who develop cervical cancer. Further research is required to define prognostic biomarkers that will identify women in whom HPV infection is likely to progress to cervical neoplasia. Factors that may influence the outcome of infection include HPV type, viral load and integration status. The prevalence of HPV- and HPV-type-specific distribution varies between different geographical regions worldwide (Clifford et al, 2005a). HPV 16 and 18 are associated with an increased risk of progression to cervical neoplasia (Clifford et al, 2003, 2005b). However, the incidence of different HPV types in cervical cancers has been found to vary with different geographical populations (Bosch et al, 1995). Only a limited number of studies have investigated HPV prevalence within the UK (Cuschieri et al, 2004; Peto et al, 2004) with the results of the largest study to date expected in November 2008 (Smyth et al, 2004). HPV prevalence is age dependent with a peak in women below the age of 25 (Schiffman et al, 1993; Petignat et al, 2005). In some populations, however, a second peak in women over 55 has been observed (Herrero et al, 2000; Molano et al, 2002). Social deprivation has also been associated with cervical cancer incidence, but the link is likely to be complex, as other factors such as smoking, education and reduced participation in screening must be considered (Sweetnam et al, 1981; Murphy et al, 1990; de Sanjose et al, 1997; Krieger et al, 1999; Thomas et al, 2001; Parikh et al, 2003; McFadden et al, 2004). However, not all studies have been able to confirm this correlation (Parazzini et al, 1998). Prophylactic HPV vaccines are currently in phase III trials and should be available within 5 years. Prevalence studies are required to identify baseline data against which efficacy of vaccination can be compared, elucidate if type-replacement occurs following vaccination and determine what fraction of disease is prevented. Current vaccines only protect against HPV16 and 18 (Harper et al, 2004; Villa et al, 2005). Our laboratory has demonstrated that other HPV types are linked with anogenital neoplasia in South Wales (KW Hart, personal communication), underlining the need to identify type-specific prevalence in this region. Determination of HPV prevalence requires a nonselected population, such as prospective sampling of screening groups, although this will not capture those who fail to attend for screening. When coverage by the screening programme is high, sampling is likely to reflect the true prevalence of HPV infection within a population. Studies of HPV infection with informed consent may result in bias, as women may refuse testing or be excluded for logistical reasons. For these reasons prospective anonymous sample collection was chosen to investigate HPV prevalence within South Wales. Here, we describe the first HPV prevalence study in Wales and assess if there is an association between social deprivation, age or cytology and HPV type. Materials and methods Clinical cohort In total, 2023 consecutive screening samples were collected, with the assistance of Cervical Screening Wales (CSW), over a 5-month period in 2004. Cervical Screening Wales manages an organised call–recall system inviting women aged 20–65 years for 3 yearly cervical screening. The residual material of liquid-based cytology (LBC) samples from women attending for routine cervical screening were used for this study. Inadequate cytology samples or those from colposcopy clinics were excluded. This study was approved by South East Wales Local Research Ethics Committee. Sample processing The LBC samples (Thinprep, Cytyc Corp, Boxborough, MA, USA) were processed and analysed by the Cytology Laboratory at Llandough Hospital, Cardiff, Wales. The residual specimen was anonymised and transported to the HPV Laboratory, Wales College of Medicine, Cardiff University. Samples were centrifuged, washed with 10 mM Tris-HCL (pH 7.4) and resuspended in 1 ml 10 mM Tris-HCL (pH 7.4) and stored at −80°C until required for further analysis. DNA purification and PCR-EIA Of 10 mg ml−1 proteinase K (Boehringer Mannheim) 10 μl was added to a 100 μl aliquot from each sample and incubated at 56°C for 2 h, followed by 100°C for 10 min. Samples were allowed to cool, then centrifuged at 13 000 r.p.m. for 10 min, and the supernatant transfered to an appropriately labelled tube. Standard HPV typing of LBC samples was performed using the PCR-Enzyme Immuno Assay (PCR-EIA) method of Walboomers et al (Jacobs et al, 1997) with minor modifications. PCR reactions were performed in a final volume of 25 μl and PCR cycling conditions were 94°C – 4 min, then 40 cycles of 94°C – 30 s, 40°C – 90 s, 72°C – 60 s followed by 72°C – 4 min. Positive (CaSki) and negative (water) PCR/ELISA controls were included in every experiment, as were positive and negative DNA extraction controls. A standard PCR for the house-keeping gene β-globin was also performed on each sample to ensure PCR viability. 10% of samples analysed were repeated to determine reproducibility. Selection criteria for samples included in analysis Detailed analysis was only performed on samples that conformed to the following criteria: β-G (...truncated)