Follow-up after treatment for cervical intraepithelial neoplasia

BMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e7186 (Published 01 November 2012) Cite this as: BMJ 2012;345:e7186
  1. Maaike C G Bleeker, pathologist 1,
  2. Chris J L M Meijer, pathologist1,
  3. Johannes Berkhof, statistician2
  1. 1Department of Pathology, VU Medical Centre, PO Box 7075, Amsterdam, Netherlands
  2. 2Department of Epidemiology and Biostatistics, VU Medical Centre
  1. mcg.bleeker{at}vumc.nl

Combined cytology and HPV testing at six and 24 months should be sufficient

To prevent cervical cancer, many women are treated for cervical intraepithelial neoplasia (CIN) with excision of the lesion. Although this treatment is considered to be highly effective, 10-15% of treated women will have residual or recurrent disease.1 2 3 Accurate assessments of the risk of cervical cancer after such treatment can guide the development of the most effective post-treatment surveillance strategy. In a linked paper (doi:10.1136/bmj.e6855), Rebolj and colleagues found a fourfold greater risk of cervical cancer in women who completed their post-treatment follow-up compared with those with a normal primary smear.4

To judge the importance of these findings in relation to post-treatment surveillance strategies, it is important to understand what is being measured. Firstly, a substantial proportion of post-treatment cancers probably result from so called residual disease. In such cases the lesion was not completely excised and human papillomavirus (HPV) infection persists. If the residual lesion was small, it might take some time for it to regrow to a detectable size. Secondly, women treated for CIN whose lesion is completely excised may have an increased risk of developing recurrent cervical disease by contracting a new HPV infection. Women who have developed CIN before may be more susceptible than those who have never had CIN to developing an incident lesion, particularly after contracting a new HPV infection. Although such biological scenarios probably contribute to the increased risk of cervical cancer in post-treatment women it is difficult to distinguish between them. This makes it difficult to work out the “optimal” post-treatment surveillance strategy.

Post-treatment surveillance should target residual disease from unsuccessful treatment and also identify women with persistent HPV infection, who have an increased short term risk of incident CIN. Given these aims, it is reasonable to question whether short term surveillance with three negative cytological tests at six, 12, and 24 months after treatment for CIN constitutes sufficient follow-up before patients return to the usual screening programme. Cytology has limited sensitivity for detecting CIN. Consequently, long term follow-up with frequent post-treatment cytological examinations is currently advocated in some countries.3 5 6 7 Such an approach has its drawbacks: for many women long term follow-up is unnecessary, and the longer the follow-up the higher the chance that incident CIN will be detected that arises from a new HPV infection. Moreover, screening related harm should be considered. Many women do not complete follow-up; only 45-60% of treated women complete the entire follow-up scheme in the first two years after treatment,8 9 and in settings with longer follow-up the attendance rate at later invitations is probably no better.

A better and more efficient short term surveillance scheme would be expected if tests for HPV were included.10 Kocken and colleagues showed that post-treatment HPV testing had a much higher sensitivity than cytology to detect high grade CIN (0.92 v 0.79), with comparable specificity.11 This result supports a strategy of using HPV testing in post-treatment follow-up. In a second linked paper (doi:10.1136/bmj.e7086), Legood and colleagues found that inclusion of an HPV test at six months might be more effective and less costly than long term cytological follow-up of women treated for CIN.12

Post-treatment surveillance has not been developed specifically to target CIN arising from newly acquired infections, yet detection of new CIN will be a consequence of long term surveillance. However, long term surveillance might be justified if women successfully treated for CIN have an excessively high absolute risk of developing cancer. Rebolj and colleagues did not report on absolute risk of cancer, but it can be calculated from the data in table 2.4 The absolute risk of cancer after a third consecutive cytologically negative smear is about 0.3% after 10 years. Although this risk is low, it may be even lower when using HPV testing. Looking at absolute numbers, a total of 20 cervical cancers was seen after three cytologically negative smears.4 Over 10 years, the total number of cervical cancer cases in the Netherlands after a normal smear was 1613; this implies that the proportion of cervical cancers in post-treatment women with completed negative follow-up accounts for only 1% of the cervical cancer cases preceded by normal cytology. In line with this observation, Kocken and colleagues found that women with negative results for both cytology and HPV at six months and 24 months after treatment had a similar five year risk of high grade CIN (1%) to women with normal cytology in population based screening.2 Thus, there is no evidence to support long term intensive monitoring of women successfully treated for CIN over and above national screening programmes.

Sufficient evidence exists for current guidelines on surveillance of women treated for CIN to be modified. Short term monitoring post-treatment of women with combined HPV testing and cytology at six months and 24 months seems to be sufficient to detect post-treatment cervical disease. After this time, adequate follow-up would be provided by a return to the national screening programme.


Cite this as: BMJ 2012;345:e7186


  • Research, doi:10.1136/bmj.e6855
  • Research, doi:10.1136/bmj.e7086
  • Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; CJLMM is member of the scientific advisory board of Qiagen, has received speakers’ fees from GlaxoSmithKline, and his institution had received board membership and consultancy fees from Qiagen; JB has received speakers’ fees from Qiagen and consultancy fees from GSK and Sanofi Pasteur; CJLMM is a shareholder of Self-screen; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.