Henry Kitchener: Primary cervical screening with human papillomavirus testing

According to our latest research paper, screening for high risk human papillomavirus (HR-HPV) infection is more sensitive than cytology testing. Together with prophylactic HPV vaccination this provides a compelling rationale to replace cytology with HPV testing in the NHS Cervical Screening Programme. Our paper suggests that not only would less cervical intraepithelial neoplasia (CIN) remain undetected, but screening intervals could also be extended, providing a “win win” of greater effectiveness and reduced costs. The major difference to the screening programme would be the need to manage a new class of screen positive women; those with negative cytology and who test positive HR-HPV infection. To refer all of these women on to specialist services would overwhelm the colposcopy service.

Cervical screening in England is provided to a national specification to which all providers must adhere. Previous HPV based changes to the NHS Cervical Screening Programme in England have been preceded by large pilot studies. This approach ensures that the results of clinical trials can be replicated on a population base in the “real world.” allowing the development of guidance that covers laboratory staffing, IT requirements, training, and technology approval. This also enables a smooth national roll-out.

Clinical trials show that HR-HPV screening leads to earlier detection of cervical lesions than liquid-based cytology, and therefore NHS England and Public Health England are working towards a national roll-out of HPV screening by the end of 2019.

To ensure that these trial results would work in the real world a decision was made in 2012 to initiate a pilot of routine HPV and LBC testing ahead of full implementation. It was carried out in six NHS laboratories across England—the results of this pilot are presented in our research paper.

There were two key issues that we had to resolve; the early recall protocol, and whether to screen with HR-HPV from age 25. In order to minimise unnecessary colposcopy, it was decided to recall at 12 months, and if HR-HPV persisted, to recall again at 24 months. With regard to the age threshold, some authorities in the field advocate restricting HR-HPV primary screening to age 30 and above, because of the very high prevalence below age 30. The steering group took the view that women aged 25-29 should not be denied a more sensitive test, and the same protocol across the age range would maintain simplicity.

In order to have a valid comparison with cytology, and to avoid prematurely deskilling the participating cytology laboratories, the conversion to primary HPV testing was partial. Six laboratories were chosen to participate in the pilot. Three in the south of England and three in the north of England. They had previously demonstrated a willingness and ability to accommodate a pilot study. Great credit is due to them for absorbing the additional work required to implement the new protocols alongside existing ones.

The pilot began in mid-2013.  We delayed the initial report until now, to allow us to report not only the entire prevalence round with early recall, but also early data from the next screening round. The key objectives were threefold: 1) to demonstrate practicability in overall terms of acceptability including the acceptability to women of early recall, and colposcopy capacity; 2) to confirm the expected increase in detection from HR-HPV testing; and 3) to confirm the expected very low incidence in the next screening round, which would support an extended screening interval.

While early recall of cytology negative/HR-HPV positive women did achieve an uplift of around one third relative to baseline detection, there was a 70% increase in colposcopy referral. This proved manageable due to partial conversion, the resilience of the service, and because it was spread over two years. This increase will be a one-off, as the prevalence of HR-HPV can be expected to drop considerably in subsequent rounds due to the sensitive detection and treatment of underlying CIN lesions, which account for a high proportion of HR-HPV infections. There is also the prospect of a reduction in HR-HPV prevalence and consequent CIN, when the HPV vaccinated cohort enter the English screening programme next year, aged 25. Such an effect has been reported in Scotland, where until recently, screening began aged 20. [1] The pilot study has otherwise gone remarkably smoothly with no pushback from either women undergoing screening or primary care staff who are responsible for screening women. All of the anticipated outcomes have been confirmed, which give us confidence that a national switch to primary HR-HPV testing, with the screening interval extended to five years, can be rolled out.

There are a number of areas that have not yet been reported. A psychosocial study will report soon on the impact of the HPV testing protocol on screened women. A study of the value of HPV genotyping is due to report. It will be important to monitor interval cancers, which can be expected to become much rarer than they currently are. There is also a need for a new IT system to support the new programme and to integrate cervical screening and HPV vaccination status. Finally, the vaccinated cohort of women who will have entered the English screening programme by next year, will be at lower risk of cervical cancer and consideration of further extending screening intervals for these women will be warranted. 

Henry C Kitchener, professor emeritus, Division of Cancer Sciences, University of Manchester and Manchester NIHR BRC.

Competing interests: See research paper.


1] Cameron RL, Kavanagh K, Cameron Watt D, Robertson C, Cushieri K, Ahmed S, Pollock KG. (2017) J Epidemiol Community Health, 71; 954-960