Shared mechanisms of ischaemic and haemorrhagic stroke: still a lot to learn.

T he risk of intracerebral haemorrhage (ICH) is increased in patients with ischaemic cerebrovascular disease and vice versa. Risk factors shared by both stroke subtypes, such as hypertension, age, and smoking, partly explain this observation. While some researchers stress that most ischaemic stroke subtypes have similar risk factor profiles, one can go further and claim that most strokes (ischaemic and haemorrhagic) are the result of a common final pathway: arterial wall damage with accumulation of abnormal cells and proteins, inflammation, molecular changes, and eventual breakdown of the intima and media. Hypertensive cerebral arteriolopathy and cerebral amyloid angiopathy are examples where ischaemic and haemorrhagic brain damage frequently coexist. Leukoaraiosis and microbleeds may both be markers of a single disease process affecting small cerebral arteries, putting patients at risk of both ischaemia and haemorrhages. This double risk may also explain why aggressive antithrombotic treatments fail in ischaemic stroke patients: the haemorrhage rate matches or exceeds the reduction in ischaemic events. Does the study by Ariesen et al in this issue of the journal (see pages 92–4) support the hypothesis that ischaemic and haemorrhagic strokes have a common pathogenesis? The authors found that in patients with previous ischaemic strokes, older age, hypertension, and current antihypertensive treatment are associated with a subsequently increased ICH risk. Most of their patients were on antithrombotic treatment, although the exact numbers are not stated. These results strongly support the presence of shared mechanisms of ischaemic and haemorrhagic events. The finding that higher blood glucose levels were associated with decreased ICH rates is unexpected. It is at odds with a previous meta-analysis by the same authors, but not with our recent observations. Why should diabetics who have had an ischaemic stroke be more resistant to ICH? They have more white matter ischaemia, brain atrophy, and cognitive decline than non-diabetics. 5 Some of this is likely due to pathology in small arteries, but it remains unknown whether the arteriolopathy of diabetics without hypertension is different from the arteriolopathy in hypertensive patients. Even if the protective effects of hyperglycaemia are confirmed by other studies, this by no means allows physicians to liberally prescribe antithrombotics: in a randomised controlled trial where 68% of patients were diabetic, combination antiplatelet treatment increased ICH significantly. Which patients are at very high risk of haemorrhage after ischaemic stroke and may therefore be poor candidates for antithrombotic treatment? Based on the findings by Ariesen et al, one might conclude that elderly hypertensive patients should not receive antithrombotic treatment after an ischaemic stroke. Unfortunately, hypertension and age are not only major risk factors for ICH but also for recurrence of ischaemic stroke. As recurrences after ischaemic stroke are mostly ischaemic, withholding antithrombotics in such patients will likely allow more ischaemic events to occur than it will avoid haemorrhages. Adding other markers that were not studied by Ariesen et al, such as alcohol intake, cognitive decline, leukoaraiosis, or microbleeds, may allow more precise definition of a very high risk group in the future. But again, only randomised controlled trials on the risk of antithrombotics in such patients will be able to answer these questions. Fortunately, several interventions have been shown to simultaneously reduce ischaemic and haemorrhagic strokes, probably by reducing damage to arterial walls: blood pressure lowering in hypertensive, normotensive, and elderly patients, 7 cessation of smoking, and regular physical activity. Until more is known, physicians will have to strictly apply this knowledge to prevent stroke recurrences. For patients suffering from both ischaemic and haemorrhagic cardiocerebrovascular events, use of antithrombotics will still have to be decided upon on an individual basis, given the lack of evidence that may last for many more years to come.