Cervical intraepithelial neoplasia and higher long term risk of cancer

Abstract

Women treated for CIN3 should have long term regular screening, even if they are beyond the normal age limit

In some countries, such as the United Kingdom, women have long term intensive surveillance after conservative treatment for high grade cervical intraepithelial neoplasia, whereas in others, such as the Netherlands and Finland, they return to regular screening after a few years. The second choice is based on the consistent observation that most recurrences of cervical intraepithelial neoplasia occur in the first two to three years after treatment.¹

In this week's BMJ, Strander and colleagues provide strong evidence that women treated for cervical intraepithelial neoplasia grade 3 have a long lasting excess risk of invasive cervical cancer.² Among more than 130 000 women with this condition, the age adjusted incidence was more than double that of the general population up to 20-25 years after diagnosis. The excess incidence was greater in women treated at older age and in recent years.

How do these findings translate into clinical recommendations? The important question is how different follow-up schedules compare in terms of effectiveness and cost in preventing the excess risk of cancer. Unfortunately, we have no direct evidence to answer this question, as no large study of the risk of cancer has also reported women's individual follow-up history.^{2 3 4}

Long term surveillance with intensive cytology might be an option. However, the relative protection afforded by different frequencies of testing in the general population⁵ cannot be assumed to be the same for women treated for high grade cervical lesions. Indeed, without data on individual follow-up the long term excess risk of cancer may be explained by cytology becoming less accurate in the long term in treated women or by women not attending long term intensive follow-up or even regular screening.¹ It would take a long time for prospective studies of women with different follow-up regimens to identify the most effective regimen. Case-control studies comparing the follow-up of women who developed cancer with those who did not might provide crucial information in the short term. The interval since the last normal cytology result is the key variable, and studying the risk of cancer according to this interval and time since treatment might identify the best regimen.

Testing for human papillomavirus DNA has consistently been shown to be more sensitive than cytology for detecting short term occurrence of cervical intraepithelial neoplasia in treated women.⁶ Long term schedules need to be defined. Repeated testing for human papillomavirus may be needed to control the long term risk of cancer, because residual infection might be difficult to detect and new infection needs to be identified. Increased risk of cancer could be the result of a new or persistent infection rather than treatment failure. Women who were previously infected could still have the same risk factors such as behaviour or susceptibility to infection. The high negative predictive value of human papillomavirus means that women could be tested less often compared with cytology. The scarce data available on follow-up by colposcopy do not show a substantial advantage over cytology.⁷

One clear indication is that women treated for cervical intraepithelial neoplasia stage 3 should continue surveillance beyond the age limit of regular screening. Such age limits have been adopted for the general population as cancer risk drops to negligible values in previously regularly screened women, but Strander's study² shows that this is not the case in women who had cervical intraepithelial neoplasia stage 3.

According to Strander's data,² any follow-up policy that can confer the same cancer risk as that seen in the general population would avoid 21.5 cancers per 100 000 person-years (including an unknown but possibly large proportion of microinvasive carcinomas) in these high risk women. Related disadvantages are not only financial but include stress related to long term intensive surveillance and the risk of having to be retreated unnecessarily as a result of false positive histology. Increased frequency of testing will lead to an increased number of biopsies and, in the general population, about 15% of diagnoses of cervical intraepithelial neoplasia grades 2-3 are not confirmed at review.⁸

The higher incidence of recurrence of cervical neoplasia in women treated in recent years correlates with the use of more conservative treatments. As for all ecological correlations, alternative explanations are difficult to exclude. For example, the higher prevalence of human papillomavirus in recent years might have made persistent hazardous sexual behaviour increasingly risky. Again, comparing by calendar period the type and characteristics of treatment (for example, if excision margins were free of dysplasia) in women who did and did not develop cancer would provide precious information.

In the meantime, returning to the widespread use of hysterectomy for cervical intraepithelial neoplasia is clearly unacceptable, especially as only some high grade cervical lesions progress to cancer⁹ and the incidence of false positive histological diagnosis is relatively high.⁸ Regarding excision, no significant difference in obstetric outcomes has been shown between cold knife conisation and other excisional techniques,¹⁰ although short term complications were more frequent with cold knife conisation.¹¹

Current evidence calls for high quality conservative treatment—this might be more achievable in centres with a high workload volume. In addition, more attention should be paid to the completeness of excision, especially in older women who have a higher risk of cancer.