Editor’s Choice

Preterm birth and previous conisation of the cervix

In January 2007 we published a paper by Bruinsma et al. (BJOG 2007;114:70–80) that highlighted the increased risk of subsequent preterm birth associated with excisional (rather than ablative) techniques for treating carcinoma in situ of the cervix. In the unit where I trained in the 1970s, the introduction of colposcopy was pioneered by a pathologist, Malcolm Anderson. He emphasised the importance of accurate diagnosis before treatment. When I carried out a knife conisation (following the instructions for lateral extent and depth that he had carefully documented in the notes) it seemed likely to me that this would have a significant effect on the subsequent function of the cervix. Technically, knife conisation is challenging, and it was easy not to take enough tissue, or to take too much. This led to the development of large-loop excision of the transformation zone (LLETZ). In my Editor’s choice of January 2007 I recounted my growing concern over the last 20 years that the cheapness and ease of use of LLETZ would lead to overintervention. Although many women were told that the procedure had few long-term effects, the study by Bruinsma reported that cone biopsy, LLETZ and diathermy were all associated with subsequent preterm birth; only ablative treatments did not carry this risk. In the same issue, Paraskevaidis et al. (BJOG 2007;114:3–4) commented that ‘reversing the trend of using LLETZ to that of low morbidity procedures, such as laser ablation, should reduce women’s anxieties by allowing us to provide effective treatment, while at the same time, reassuring them that there is a minimal, if any, subsequent risk of future adverse pregnancy effects’. Their commentary was backed up by a very influential BJOG podcast (still online for those who have not heard it). In this month’s issue of BJOG, we carry three further articles developing this theme. First, on page 243, Siobhan Quenby writes that ‘There is increasing evidence that excisional procedures for treatment of cervical intraepithelial neoplasia (CIN) are associated with preterm delivery and subsequent perinatal mortality’, and says that the evidence now supports ‘the concept that the more cervix that is removed the worse the risk of preterm labour’. Moreover, she points out that at present we do not have an effective strategy to combat the effects of excising cervical tissue on subsequent pregnancy, and concludes that although “policies of ‘see and treat’ may have been attractive in organisational terms.… more conservative approaches (listed in detail in her article) may now be more appropriate”. Providing more evidence for this point of view are the papers by Ørtoft et al. on page 258 and by Poppe et al. on page 268. Ørtoft et al. report that in their study of 721 births after one conisation and 37 after two conisations, the relative risks of preterm delivery were 2.8 (95% confidence interval 2.3–3.5) and 9.9 (95% confidence interval 6–17). Translated to absolute terms, this means that approximately one in nine women will experience preterm birth following a single conisation, rising to one in three following two conisations. Moreover, not only was the risk of preterm birth increased, they were able to show in the single conisation group that it was associated with a statistically significant 2.8-fold increased risk of perinatal death. Poppe et al. studied 72 pregnancies in 47 women following conisation, and found that one in four pregnancies ended in preterm birth, with almost half of these before 34 weeks of gestation. They concluded that ‘because of adverse pregnancy outcomes after conisation, expectant management of all lesions in young women, and of low-grade lesions in older women is recommended’, and that ‘women with a definite desire to have children, should be informed about the possible risks associated with excisional treatment of CIN’, before the procedure is carried out. They should also be treated as high risk in subsequent pregnancies. Given the evidence in recent meta-analyses (Kirgiou et al. Lancet 2006;367:489–498 and Arbyn et al. BMJ 2008;337:a1284), these conclusions are now surely incontrovertible.

Preterm birth and previous threatened first-trimester miscarriage

The incidence of first-trimester miscarriage is about 30% if one includes all chemically diagnosed pregnancies and about 15% if one includes only pregnancies that are clinically apparent. The majority of these (perhaps as many as 80%) are due to chromosomal abnormalities, and I have commonly reassured women that this is ‘nature’s way of dealing with pregnancies that were wrong from the beginning’. But what of women who have first-trimester bleeding per vaginam but then go on to have a continuing pregnancy? In the past I have reassured women that, provided the bleeding settles down, the outcome is likely to be good. But how good? This question is addressed by Saraswat and colleagues in their systematic review on page 245. It is perhaps not unexpected that there is an increased incidence of antepartum haemorrhage due to placenta praevia (odds ratio 1.62, 95% confidence interval 1.19–2.2) or of unknown origin (odds ratio 2.47, 95% confidence interval 1.52–4.02), and that the risk of preterm prelabour rupture of membranes is also greater (odds ratio 1.78, 95% confidence interval 1.28–2.48). However, the findings of an increased risk of preterm delivery (odds ratio 2.05, 95% confidence interval 1.76–2.4) and intrauterine growth restriction (odds ratio 1.54, 95% confidence interval 1.18–2.0), and a raised perinatal mortality (odds ratio 2.15, 95% confidence interval 1.41–3.27) are perhaps not so widely known. Clearly women with first-trimester miscarriage bleeding per vaginam need monitoring carefully during pregnancy, ideally with serial ultrasound scans.

Preterm birth and maternal hypertension and diabetes

Both damage to the cervix and abnormal placentation are likely to lead to an increased risk of preterm birth. But can there be something inherent in the mother that predisposes to preterm birth? While established hypertension and diabetes obviously carry the risk that iatrogenic preterm delivery will be needed, the study by Lykke et al. on page 274 of 782 287 Danish women with a first singleton delivery and 536 419 women with a first and second singleton delivery has shown that preterm birth is a marker of an increased incidence of subsequent type 2 diabetes (hazard ratios 1.58–2.30 depending on the number of preterm births) and cardiovascular morbidity in the mother (particularly thromboembolism, hazard ratios 1.18–1.8). Preterm birth is known to be associated with elevated prepregnancy and pregnancy lipid concentrations and inflammatory markers, as well as later dyslipidaemia and cardiovascular morbidity. It seems likely that these have some causal role in the aetiology of both preterm birth and subsequent ill health, but mechanisms remain to be elucidated. Further observations on the role of circulating angiogenic factors and adipocytokines are detailed on page 314 by Masuyama and colleagues and the role of the anti-angiogenic factor endoglin is described on page 321 by Buhimschi and colleagues.

Radical trachelectomy for carcinoma of the cervix

In BJOG we have published several papers from the group at St Bartholomew’s in London, documenting their experience in removing carcinoma of the cervix surgically in such a way as to allow subsequent pregnancy (Shepherd et al. BJOG 2001;108:882–885; BJOG 2006;113:719–724) and from the Budapest group (Ungar et al. BJOG 2005;112:366–369). On page 340, we publish an account by Kim and colleagues from Seoul in Korea of 32 such procedures planned using a laparoscopic approach. The procedure was completed laparoscopically in 27. There has subsequently been one death from recurrence, and three successful conceptions, with one live birth. If you want to know more technical details of how this operation was carried out, log on to the paper on Wiley Interscience at www3.interscience.wiley.com/journal/118523178/home and view the video that accompanies the paper (once you have found the article, click on ‘Supporting Information’).

Environmental pollution and endometrial cancer

As humankind has spread across the globe, pollution of the environment has been an ever growing problem. Initially, largely because of contamination with human biological waste (one thinks particularly of contamination of the water supply with faecal organisms), chemical pollution grew rapidly during the Industrial Revolution. An infamous example from 50 years ago is the pollution of the Shiranui Sea off the shores of Japan, where the release of mercury from a chemical factory killed almost 1800 people, and over 10 000 people received financial compensation for neurological damage. In 1984, the release of methyl isocyanate from a Union Carbide pesticide plant in Bhopal India killed an estimated 25 000 people. The apparent 50% reduction in average sperm counts over the last century has also been attributed to environmental pollutants, particularly those with estrogenic activity. However, despite widespread concern about pollution and rising rates of cancer, it is generally estimated by organisations such as the World Health Organization, and the International Agency for Research into Cancer, that environmental pollutants are only responsible for about 3% of cancers. But given the high and rising level of pollutants with estrogenic activity, could the uterus be a particularly vulnerable target? On page 348, Dey and colleagues hypothesised that urban pollution with estrogens in a developing country would be of particular concern, and so they carried out a study using a population cancer registry based in Tanta, the capital of Gharbiah province in Egypt. The incidence rate ratio for uterine cancer in urban areas compared with rural areas was 6.07 (95% confidence interval 4.17–8.85), twice the differential seen for cancer of the cervix or ovary. Other studies from this group have also shown a higher urban rate of breast cancer (Breast Cancer Res Treat, 2009 June 23 epub DOI 10.1007/s10549-009-0427-9). Whereas the reduction of male fertility secondary to Xenestrogens (industrially made compounds that have an estrogenic activity) might be seen as useful in helping to moderate population growth, if they cause cancer, methods for reducing their production or release into the environment need to be developed.

Reaching the millennium development goal for the reduction of maternal mortality

For two decades or more, the high rates of maternal death in the developing world have been hitting the headlines. However, recent studies have made it increasingly clear that it is not that we do not know how this terrible toll might be reduced, or even that improvement would be impossibly expensive. Rather, it is the ‘know-do gap’ that is responsible for progress being slow. The ‘know-do gap’ is defined on page 304 by Hussein and her colleagues as ‘the disparity between what is known, and the application of that knowledge in policy and practice’. They studied maternal mortality reduction programmes in Afghanistan, Bangladesh, India, Nepal and Pakistan, to identify initiatives that worked in practice and not just in theory. They found a number of key themes; ‘task shifting’ (training support workers to carry out common procedures such as administration of oxytocics, manual removal of the placenta and assisted vaginal delivery, previously carried out by doctors); implementation of communication helplines and emergency transport; financial incentives to promote desired behaviour in health workers; providing families with money to pay for key services; supporting and encouraging health workers including hospital cleaners and ambulance drivers; and providing health workers with input into policy development at a national level. It is sometimes quipped that development aid is money sent from poor people in rich countries to rich people in poor countries. It is clear from Hussein and colleagues’ study that finance is just one part of the solution. Trustworthy leadership and improved organisation, together with the encouragement and support of local workers, are just as necessary.