We wish to thank Drs Bamford and Hall for some interesting and very relevant comments.1 We agree with their statement that a small and shallow LLETZ (large loop excision of the transformation zone) cone biopsy may do little or no harm in relation to preterm delivery. The size of a cone can be difficult to measure, which may account for why different studies show different results. However, our study2 agrees with many others that there is a positive correlation between the depth of the cone and the risk of subsequent preterm delivery, and we think this agreement is unlikely to have occurred by chance. It therefore seems logical that the clinician should try to avoid a deep cone when this can be done without putting women at risk for cervical cancer. Bamford and Hall also speculate that the depth of the cone biopsies in our study was greater than is usual in UK practice. The decision to make a deep or shallow biopsy is related to the colposcopic findings—particularly the location of the transformation zone, the severity of the dysplasia and the histology—as for example with adenocarcinoma in situ. In our study the indications for cone biopsy were cervical intraepithelial neoplasia, stage 3 (CIN3) for 89% of cases and, CIN1 or 2 in only 11% of the women. Only lesions located in a visible transformation zone were treated conservatively. The severity of the lesions seen in our practice may have caused us to make wider and deeper cones, and thereby created a higher risk of preterm delivery. However, this risk is likely to represent a continuum rather than be ‘all or nothing’. The fact that we recorded a preterm delivery in every third woman who had two conisations emphasises the importance of keeping the size of any cone to a minimum, consistent with removing the lesion. Because of our findings, our current practice is to make as shallow a cone biopsy as possibly, although we do not hesitate to make the cone as wide as is needed to include the entire transformation zone.
The preparation of the cone biopsy for histological examination may lead to changes in its length and width. In our study, all cone biopsies were treated in the same way so that even though the biopsy may have been stretched or shrunk, this cannot explain our findings that a deeper cone increases the risk of preterm delivery.
The purpose of screening and treatment of CIN is to prevent cervical cancer. As CIN2 lesions may regress in approximately half of our patients, we suggest that CIN2 in young women should be observed every 4 months, with colposcopy and cervical smear, until the lesion either disappears or progresses to CIN3, at which point we then suggest a cone biopsy. We also consider that women should be informed of the increased risk of preterm delivery before a conisation is performed.
We hope that our results will encourage the introduction of, and participation in, nationwide human papillomavirus (HPV) vaccination programmes. The option of HPV vaccination in young women having their first cone biopsy should be considered, in order to minimise the need for a second, with its consequent substantial risk of preterm labour, although we recognise that the benefit of vaccination for this group still needs to be shown in a randomised study.