Intended for healthcare professionals

Editorials

At what age should cervical screening stop?

BMJ 2009; 338 doi: https://doi.org/10.1136/bmj.b809 (Published 24 April 2009) Cite this as: BMJ 2009;338:b809
  1. Björn Strander, director
  1. 1Cervical Screening Oncology Centre, Sahlgren’s University Hospital, SE-413 45 Göteborg, Sweden
  1. bjorn.strander{at}oc.gu.se

    Negative tests are no reason to stop screening earlier

    Ever since the first organised cervical screening programmes started in Europe more than 40 years ago discussion about the upper age limit for effective screening has been ongoing. The debate is still relevant because mass vaccination of pre-adolescent girls against two or more types of human papillomavirus (HPV) will not affect the incidence of cancer in girls born around the turn of the millennium until 2050-60. In the linked study (doi:10.1136/bmj.b1354), Rebolj and colleagues report that the incidence of cervical cancer is similar in 218 847 women aged 45-54 years and 445 382 women aged 30-44 after their third negative smear.1

    Evidence suggests that repeating smear tests in women aged 60-65 whose previous tests have been normal has little, if any, benefit,2 and some researchers have proposed that the age limit should be lowered to 50.3 4 In all European programmes, cervical cancer screening stops at a lower age than breast cancer screening, and in some programmes screening intervals between age 50 and 60 are prolonged.5

    With reservations for biological variations, this strategy is based on the following reasoning. Women acquire oncogenic HPV infections in their 20s. Some women whose infections are not cleared develop high grade cervical lesions in their 30s, which can progress to cancer in their late 40s or 50s. But women who have had several normal smears at that age have good protection for the rest of their life. This biological model is largely consistent with epidemiological data.

    The paper by Rebolj and colleagues does not refute the old concepts completely, but it does pose some questions.1 The researchers confirm that women with three normal smears have a low absolute risk of developing cancer; however, they also show that this is independent of whether the woman is above or below 45. Moreover, they find an equally accelerated increase in incidence more than 10 years after the last normal smear—the protective effect of a history of repeated negative smears has a “best before date,” even at 50 years of age.

    Curves that show age specific incidences for cervical cancer in areas with established cervical screening programmes typically have two peaks. One is at around 40 years of age, and the other at around 75.6 In Sweden, this second peak has decreased only modestly with time, although today all older women have received screening invitations during their earlier life. The incidence of cervical cancer in Sweden has fallen by 65-70% in the 45-60 age groups, when the period 1970-5 is compared with the most recent period (2003-7), but the decrease has been only 22-30% in women over 75.7 Furthermore, most advanced cancers occur in women over 70, whereas invasive cancer in younger women is more often subclinical.8

    Should screening programmes be extended beyond 60-65 years of age? As yet, we have insufficient evidence to support this idea. We still do not know what protection is offered to older women by screening with cytology. The finding by Rebolj and colleagues of fewer precursor lesions in the older age group, although screening intensity and cancer rates were similar, agrees with other studies. However, some studies show a good protective effect from normal smears in older age groups.8

    It has been suggested that as women leave the screening programme they should be tested for HPV (“exit test”),9 and that surveillance should be continued in women who are HPV positive only, but evidence to support such a strategy is scarce. We know little, not only about the efficacy of testing, but also about basic HPV biology in this age group. We must also remember that studies on screening effects in different age groups are retrospective and reflect sexual behaviour and HPV transmission many years ago. In 1995 the median accumulated number of sexual partners for a woman was three times as high as in 1967.10 We have to pay close attention to developments in invasive cancer in age groups above the cut-off point for screening and be prepared to adjust the screening ages as we learn more.

    With modern computer technology we could tailor screening invitations to the individual. Rebolj and colleagues and others have found that women with several normal smears are protected against cervical cancer. This finding, and the knowledge that women with previous precursor lesions are at high risk up to old age,11 12 could be used to individualise call-recall systems by extending or shortening intervals and setting the upper age limit. Such information could be incorporated into future algorithms that include results on HPV tests and documentation of HPV vaccination. Resources could then be allocated away from women who would not benefit from additional smears within a certain number of years to those who would and the question of whether to screen above the age of 60 could be answered—yes, for those who benefit the most from it.

    Notes

    Cite this as: BMJ 2009;338:b809

    Footnotes

    References

    Log in

    Log in through your institution

    Subscribe

    * For online subscription