How to manage blood pressure in acute stroke

The study of how best to manage blood pressure in acute stroke has been long and fraught. Although it is appreciated that high blood pressure is common (approximately 75–80%) in patients with acute stroke, whether of ischaemic or haemorrhagic type, the management of this physiological disequilibrium has been debated for many years [1,2]. High blood pressure has been found to be associated with a poor outcome in some, but not all, studies. Most of these studies have been small, and are based on a few tens to hundreds of inpatients admitted to a single centre. Accordingly, they have limited statistical power to assess the relationship between blood pressure and outcome, and cannot adjust for important covariates. A meta-analysis of these observational studies supported the observation that high blood pressure is associated with a poor outcome [3]. It took a large observational study nested within a mega-trial to confirm this finding, which showed that a low or low–normal blood pressure (systolic blood pressure < 120 mmHg) is also associated with a poor outcome in patients with ischaemic stroke, and that these observations are independent of prognostic factors such as age and stroke severity [4]. In this issue of the journal, Okumura et al. [5] extend these findings and add important new information. The authors performed a relatively large hospital-based study, thereby differing from earlier smaller studies, and confirmed the ‘U-shaped’ curve relating early blood pressure and outcome (death in this study) in 1004 patients with ischaemic stroke, as found previously [4,6,7]. They also found the same relationship in the 1097 patients with haemorrhagic stroke, as reported previously [8]. Most importantly, Okumura et al. [5] found that the ‘U-shaped’ relationship between blood pressure and death was shifted to the right in patients with previous high blood pressure compared to those without such a history. This finding, which mirrors the right-shift seen in cerebral autoregulation in hypertensive individuals, reminds us that patients should be treated as individuals when considering blood pressure management plans. The weaknesses in the study of Okumura et al. [5], such as the potential for bias due to the inclusion of only 44% of the identified subjects, do not appear to detract from the main findings. Furthermore, death alone as an outcome was studied, whereas the combination of death or dependency (or disability) is most relevant to patients with acute stroke, and their carers and healthcare professionals. Nevertheless, the findings for functional outcome are likely to mirror death, at least on the basis of previous work [4]. If a high blood pressure is both common and independently associated with an increase in death, and combined death or dependency, then acute lowering of blood pressure might improve outcome after stroke. However, concerns about the effects of altering blood pressure on cerebral blood flow (CBF) have limited research activity. Cerebral autoregulation normally maintains cerebral perfusion in the face of changing blood pressure but becomes dysfunctional in acute stroke, such that CBF could decrease as blood pressure is reduced, as observed with nicardipine, a calcium channel blocker [9]. Recent pilot studies have suggested that CBF is not adversely affected with mild to moderate reductions in blood pressure with oral angiotensin-converting enzyme inhibitors (oral perindopril) [10], angiotensin receptor antagonists (oral losartan) [11] and nitric oxide donors (intravenous sodium nitroprusside [12]; transdermal glyceryl trinitrate, Willmot and Bath, unpublished data). The potentially favourable findings seen in the ACCESS phase II trial of candesartan [13] have also boosted the confidence of clinical researchers. As a result, two phase III trials assessing the management of blood pressure in acute stroke are in progress. The academic ENOS trial (www.enos.ac.uk/) aims to assess the safety and efficacy of transdermal glyceryl trinitrate versus a control in patients with ischaemic or haemorrhagic stroke [14]. Using a partial factorial design, ENOS also aims to compare the effect of continuing or temporarily stopping existing antihypertensive therapy. The latter question is the primary focus of another non-commercial trial, COSSACS (www.le.ac.uk/cv/research/COSSACS/COSSACShome.html). A third phase III trial, SCAST, will test candesartan, as used in ACCESS. Thus, the data provided by Okumura et al. [5] may signal the end-of-the beginning of this saga, whereas the ongoing trials will offer the opportunity to answer this key stroke question in the moderately near future.