Ambulatory blood pressure monitoring in pregnancy

Preeclampsia remains a major cause of maternal death and perinatal mortality [1]. Globally it is estimated that 40 000 women die each year from this condition, mostly from uncontrollable seizures, pulmonary oedema or intracerebral haemorrhage. That many of these deaths are eminently preventable, is evidenced by a recent Confidential Enquiry into Maternal Deaths in the UK [2], which reported that only two women (less than one in every million deliveries) died of this condition in the UK in 2012–2014. Factors contributing to this dramatic improvement include prophylactic use of low-dose aspirin from early gestation in high-risk individuals, anticonvulsant therapy including magnesium sulfate, appropriate use of antihypertensive therapy and protocol-driven antenatal and general obstetric care, including planned delivery. Preeclampsia is also a major contributor to perinatal morbidity and mortality, including an estimated annual global incidence of 2.6 million stillbirths [3]. Much of the burden of morbidity and mortality from preeclampsia falls upon resource-poor populations, where access to obstetric care is very limited. Admittedly, interventional treatments for various manifestations of preeclampsia may have relatively greater impact on maternal than foetal outcomes. Nevertheless, efforts to identify, by relatively inexpensive means, those patients at greatest risk and meriting greatest antenatal attention, deserve our interest. Preeclampsia is a multisystem disease characterized by derangements in cardiovascular, renal, hepatic, cerebral, placental and haematological systems. Endothelial dysfunction may be a common denominator. Hypertension is a key element, not only as a manifestation of the underlying vascular abnormality but also as a driver of maternal and placental complications. Although the cause(s) are not fully understood, it would appear that the process has its origins very early in pregnancy, whereby a faulty placentation initiates or perpetuates a ‘slow burning’ endothelial dysfunction that eventually manifests itself as the clinical syndrome characterized by rising blood pressure, proteinuria, intrauterine growth restriction and the other features that can cause so many problems in later pregnancy. Not surprisingly, great efforts have been made to develop means of predicting which pregnant women may be at greatest risk of developing this condition and in women who do, which of those are most likely to progress to the various clinical patterns of the illness. One of the difficulties, of course, is that the many potential ‘end points’ make interpretation of results rather more difficult than for a more homogenous condition. Over the years, a wide range of methods has been proposed to do this, either singly or as combined risk scores. These include data on demographics, haemodynamics, biochemistry, endocrinology and more recently a plethora of biomarkers. It is outwith the remit of this commentary to consider these in detail, but one thing is clear – no single marker has proved sufficiently accurate to gain acceptance as a gold standard. The successful use of low-dose aspirin from 12 weeks gestation in high-risk individuals [4] demonstrates several important points. The coagulation system is important in the pathogenesis of preeclampsia. Early intervention can make a difference to outcomes, although the numbers needed to treat remain high. It is possible to conduct large trials in obstetrics! Even so, and despite a number of metanalyses and Guidelines, uncertainties remain about the optimal dose and starting date. As yet, the same confidence cannot be expressed about antihypertensive treatment. Yes, treatment of moderate-to-severe hypertension reduces the risk of progression to more severe hypertension and probably reduces the risk of maternal stroke and pulmonary oedema, but there is as yet no convincing evidence that it otherwise alters the course of preeclampsia or that it benefits foetal outcomes. Nevertheless, knowledge of the blood pressure is an important component of antenatal care, and if it can be used to help identify women who might benefit from more intensive monitoring, planning of delivery, and antepartum use of corticosteroids to aid foetal lung maturation, that may be a worthwhile aim in itself. This really has three components – first identifying a predictor that is accurate and reliable, second, ensuring that such a predictor is clinically pragmatic and accessible to vulnerable populations, and third, demonstrating that the predictor has clinical utility when tested prospectively. Such an apparently basic and ubiquitous observation as blood pressure measurement remains fraught with technical problems. The difficulties with sphygmomanometry have been well documented [5,6]. In order to circumvent some of these, various methods of automated or semiautomated measurement have been developed, including ambulatory blood pressure measurement (ABPM). In obstetric practice, there has been variable uptake of these new techniques [7] for blood pressure measurement. Innumerable reports have described the use of blood pressure measurements, by various methods and at various times during gestation, as a predictor of preeclampsia and its various manifestations and complications. In summary, there is reasonable evidence that tracking of blood pressure occurs even from first trimester [8], that blood pressure measurements alone are of limited value in predicting adverse outcomes in an individual patient but that adding more blood pressure measurements from about 28 weeks may add to the predictive value [9]. Two recent contributions from Salazar et al. have added usefully to our knowledge by using ABPM in patients considered to be at high risk of developing preeclampsia. In 2016, they [10] reported that ABPM at 28 weeks was able to identify those women who subsequently developed preeclampsia with much greater accuracy than clinic blood pressure. ‘Masked’ hypertension (normal clinic pressure, mean 24 h pressure ≥130/80 mmHg) and ‘nocturnal’ hypertension (normal clinic and mean daytime ambulatory pressure, mean night-time BP ≥120/70 mmHg) both identified those at risk, though nocturnal hypertension was the better predictor. In this issue, Salazar et al.[11] provide further evidence that ABPM may provide a more accurate means of predicting the subsequent development of preeclampsia than clinic measurements. Ninety-two women considered to be at high risk of the condition underwent ABPM using a Spacelabs 90207 monitor (Spacelabs, Redmond, Washington, USA) at an average gestion of 23 weeks. The overall incidence of preeclampsia was 15%. When the ABPM was normal, the risk of preeclampsia was 7.7%, increasing to 25.7% when the ABPM showed hypertension at any time and to 29% when the ABPM showed nocturnal hypertension. Nocturnal hypertension at the end of the second trimester thus markedly increased the relative risk of preeclampsia [odds ratio (OR) 5.32, 95% confidence interval (CI) 1.48–19.10]. One interesting post hoc observation was that the risk of preeclampsia for those with nocturnal hypertension was twice as high in women not taking aspirin at the 20th week of gestation. The attractions of this study include that it gives further support to the superiority of ABPM over clinic blood pressure (BP) as a means of predicting adverse outcomes in pregnancy. It is perhaps not surprising that a more accurate methodology will prove a better predictor than ‘office’ measurements, with all their limitations and inaccuracies. This study also gives some hope that even earlier prediction of risk of preeclampsia may be possible. The technique is reliable, noninvasive and widely available, though sadly not yet in many of the higher risk populations. Outwith pregnancy, it is many years now since it was reported [12] that ABPM more accurately predicted major cardiovascular outcomes than conventional clinic measurements and indeed that mean nocturnal SBP seemed to be the best predictor. Perhaps there is also a wider message in as far as choosing a method to reduce observer influence and ‘white-coat’ effect, and the period during the 24 h with lowest external influences, may lead to superior predictive power, not only in terms of cardiovascular disease but also within the completely different timescale of preeclampsia. On the downside, there are several important limitations. The study was undertaken on a small number of selected patients in a single centre. The investigators cannot be criticized for this – these were the patients available for the study and they had to start somewhere. However, it does raise real concerns about the generalizability of results and in particular whether ABPM would retain its predictive advantage when applied to patients not already known to be at increased risk. Whilst not exactly ‘cutting edge’ technology, and though the instrument used for ABPM has been widely deployed in many countries, the technique employed for ABPM is still not available in many less privileged communities. The particular issue of small sample size has bedevilled research in pregnancy for far too long – witness the real paucity of reliable data on the use of antihypertensive drugs in pregnancy – a sad indictment of the medical profession, given the huge scale of such trials in nonpregnant patients! In respect of the management of preeclampsia, aspirin [4] and magnesium sulfate [13] are two honourable exceptions, confirming – as if confirmation was ever needed – the value of trials of adequate power. As a further illustration, consider a recent meta-analysis [14] of the effects of blood pressure lowering in hypertension (nonpregnancy) published in this Journal in 2014. That meta-analysis included 68 trials, 245 885 participants and 1 058 117 patient-years of treatment. 82% of the studies included were rated ‘high quality.’ By stark contrast, in their Guidelines on Hypertension in Pregnancy 2010, NICE [15] was able to reference for preeclampsia with severe hypertension a total of 24 randomised controlled trials, including 2949 patients. Only three of these trials included more than 100 patients and in only five was there adequate blinding. In moderate–severe hypertension, the total number of trials was 21 with 1085 participants. Four of these trials had adequate blinding and 11 had adequate randomization. It is not surprising that in a more recent Cochrane overview [16] of antihypertensive drug therapy for mild-to-moderate hypertension in pregnancy, it was concluded that ‘It remains unclear whether antihypertensive drug therapy …… is worthwhile.’ There thus remain huge uncertainties about the most appropriate thresholds for treatment, the most appropriate target values for BP control, and the most appropriate individual drugs. Should we not expect that blood pressure management in pregnancy be guided by more robust evidence? I fear an analogous situation may arise in respect of ABPM monitoring in pregnancy. It is over 25 years since ABPM was proposed as a screening test for preeclampsia [17]. Whereas ABPM is now established as a necessary component in the diagnosis of hypertension in routine practice [18,19,20] it has not yet been accepted as a necessary contributor to obstetric practice. What is needed, in this and other aspects of hypertension management in pregnancy, are large-scale prospective multicentre trials to confirm the tantalizing data from these useful but limited smaller studies. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.