• ervical cancer

Cervical cancer screening by Papanicolaou (Pap) smear has shown its use in reducing both incidence and mortality. Nowadays, cervical tumours are mostly diagnosed in women who were not, or not properly, screened. The invasive sampling method of screening is one of the reasons why women do not participate. The efficiency of cervical cancer screening could be increased if a less invasive test were available. Today, there is extensive scientific evidence that infection with high risk human papillomavirus (HPV) is associated with the development of cervical cancer. An international survey of more than 1000 cervical cancers showed that HPV DNA was present in 93% of all tumours.¹ Further investigation of the HPV negative carcinomas showed that, with improved methodology, 99.7% of all cervical tumours contained HPV DNA.² Recently, it has been suggested that self sampled vaginal material can be used for HPV detection. Several investigations—on a limited number of women—have shown a good correlation between self sampled vaginal material and a cervical sample taken by a professional.^3,4

This study aimed to investigate the HPV prevalence in cervix and vagina on samples taken by a professional. Between October 2001 and March 2003, 159 women were enrolled in this study. Of these women, 96 visited their GP for a routine Pap smear, whereas 63 women, working as prostitutes, visited an STI clinic. The study protocol was approved by the medical ethics board of Antwerp University. The GP or STI doctor first took a vaginal sample using a polyurethane tipped swab (Culturette EZ, Becton Dickinson) and then, after inserting a speculum, a cervical sample using a Cervex-Brush (Rovers, Oss, Netherlands). Samples were treated as described previously.⁵ HPV DNA amplification was performed using the GP5+/6+ HPV polymerase chain reaction (PCR).⁶ Detection of PCR products was performed in an enzyme immunoassay format.⁷ After detection of HPV with a HPV probe cocktail, typing analysis was performed for HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68. The correlation of HPV in the vagina and the cervix was determined with an unweighted kappa statistic to determine the percentage of correlation beyond that expected by chance.

In 145 of the 159 women HPV detection was complete. The overall HPV prevalence in this study was 22.8%; 14.1% in the general population, and 35.0% in sex workers. There was excellent agreement between HPV prevalence in vaginal samples and in cervical samples (table 1). The overall agreement was 94.5% (kappa 0.83, 95% CI 0.77 to 0.89). The HPV prevalence was slightly higher in the cervix than in the vagina (21.4% versus 18.6%, respectively). In all but one positive case at least one HPV type was present in both sites.

In conclusion, we have shown that there is a very high concordance between vaginal and cervical HPV prevalence when a polyurethane tipped swab is used by a professional. We are currently performing a field study to investigate the impact of self sampling by the women as well as that of transport and storage.