At present, screening for cervical cancer
is done mainly by reading Pap smears
for abnormal cells. Although this subjective method has led to a decrease in incidence and mortality attributable to cervical cancer in developed countries, false-positive and false-negative screening results happen regularly. With the insight that infection with high-risk human papillomavirus (HPV)
is obligatory for the development of cervical cancer, and the introduction of sensitive HPV testing methods, it was foreseen in 1992 that testing for HPV had implications as a primary technique for cervical screening.1
Despite the overwhelming evidence that testing for HPV results in earlier detection of cervical intraepithelial neoplasia (CIN
) grade 3 and cervical cancer, it has not yet been implemented in regular screening programmes. Resistance from cytopathologists against a molecular test and the fear of higher colposcopy referral rates due to the 4—6% lower specificity of the HPV test are among the main reasons.
In The Lancet Oncology
, the study by Hormuzd Katki and colleagues2
provides extensive data about the cumulative incidence of CIN2 or worse and CIN3 or worse after HPV and cytology
co-testing of more than 300 000 women aged 30 years or older participating in the Kaiser Permanente
cervical screening programme. Although similar conclusions from randomised trials have been published earlier, the strength of the report lies in the reliable risk estimates for CIN2 or worse and CIN3 or worse for rare combinations of HPV and cytology results and the routine setting wherein the results have been obtained.
An important conclusion is that the 1-year and 4-year cumulative incidences of CIN3 or worse of sole HPV testing and HPV and cytology co-testing are similar, which questions the value of co-testing, still being recommended in the USA
for women aged 30 years or older. Moreover, the CIN3 or worse risk after a negative HPV test is much lower than after a negative cytology test, providing justification for an extension of the screening interval. Another important conclusion is that women positive for HPV but negative by cytology have a substantial risk of CIN3 or worse. This finding, usually underestimated by clinicians, necessitates closer surveillance. A third important conclusion is that testing for HPV detects more adenocarcinomas and precursors than cytology, providing hope that the incidence of adenocarcinoma will decrease in the future.
From these and other data, it becomes clear that guidelines for cervical screening have to be changed in the near future. The expectation is that the primary screening method will be a clinically validated HPV test and the role of cytology will be to triage women positive for HPV to restrict the number of colposcopy referrals. Cytologically negative women can be followed up by HPV testing after 3 years3
or by cytology after 6—12 months.4
HPV16/18 genotyping has also been suggested as a triage method but will lead to many more colposcopy referrals.4
For triage of women positive for HPV, cytology will face growing competition from other methods that detect viral (eg, E6/E7 region mRNA
) or non-viral markers. The latter includes an immunocytochemical assay, which relies on dual-staining of cervical cells for p16INK4a and Ki67 antigens and has revealed promising data.5
Another promising candidate triage tool involves measurement of promoter methylation of CADM1
that has the advantage that it is a molecular technique and hence not troubled by subjective interpretation.
Finally, it must be remarked that screening refusal is an important problem in several countries, and efforts should be made to maximise the uptake of screening. For women that refuse to respond to a screening invitation, devices for self-collection of cervicovaginal material for HPV testing have been developed and data are accumulating suggesting that the sensitivity for high-grade CIN can be similar to that of testing for HPV on a physician-collected cervical smear. However, unlike the HPV test, cytology on self-sampled vaginal material is not reliably applicable, and triage of women positive for HPV by molecular markers seems indicated. We foresee that in the future, cervical screening will be done by the testing of HPV on either a smear taken by a physician or a self-sampled specimen, dependent on the preference of the woman, followed by triage by a molecular marker, making cervical screening objective and reproducible.
CJLMM has served on an advisory board for Qiagen and has received speakers’ fees from Qiagen and Roche. JB has received an unrestricted research grant of GlaxoSmithKline. PJFS has served on advisory boards for Gen-Probe, Roche, and GlaxoSmithKline.