Secondary prevention for stroke and transient ischaemic attacksBMJ 2004; 328 doi: https://doi.org/10.1136/bmj.328.7435.297 (Published 05 February 2004) Cite this as: BMJ 2004;328:297
- Keith W Muir, senior lecturer in neurology ()
- Division of Clinical Neurosciences, University of Glasgow, Institute of Neurological Sciences, Southern General Hospital, Glasgow G51 4TF
Even patients with normal blood pressure and cholesterol levels may benefit
Two key trials—the perindopril protection against recurrent stroke study (PROGRESS)1 and the heart protection study (HPS)2—have expanded options for secondary prevention after stroke or transient ischaemic attack and have also mandated a fundamental change in thinking about risk. The concepts of hypertension and hypercholesterolaemia may now be irrelevant or even harmful in this population.w1
In observational epidemiology studies no demonstrable floor exists for the relation between blood pressure and risk of stroke, with risk continuing to halve for every 10 mm Hg fall in diastolic pressure even at conventionally normotensive values.3 Meta-analyses of small trials of antihypertensive treatment in patients with a stroke showed a 28% reduction in relative risk for stroke regardless of baseline blood pressure.4 w2 This was supported by the post-stroke antihypertensive study (PATS),5 in which indapamide gave an absolute reduction in stroke risk by 2.9% compared with placebo over three years (number needed to treat (NNT) = 34) in 5665 patients with prior stroke or transient ischaemic attack and mean baseline blood pressure of 154/93 mm Hg.
In PROGRESS, 6105 patients who had had a stroke or transient ischaemic attack an average of six months previously were randomised to get 4 mg of perindopril, a long acting angiotensin converting enzyme (ACE) inhibitor prodrug, in combination with indapamide 2 mg or 2.5 mg (at the discretion of the treating clinician—not randomised), or to matching placebo(s), in addition to other antihypertensive drugs. Over four years the relative risk of stroke was reduced as predicted by 28%, with a similar reduction in all cause cardiovascular morbidity. In PROGRESS, patients with intracerebral haemorrhage benefited perhaps even more than those with ischaemic events. Overall the absolute benefit was 3.7% over four years (NNT = 27).
In the lowest tertile of blood pressure at entry to PROGRESS mean blood pressure was 128/77 mm Hg. These patients had an identical reduction in relative risk compared with higher tertiles and an absolute reduction for stroke of 4.3% (NNT = 23) while taking combination treatment.
The heart protection study postulated on the basis of comparable observations of the continuous relation of cholesterol concentration and risk of coronary artery disease that prior medical history would define individual risk more than any artificial threshold and similarly found equal benefit of simvastatin 40 mg across all tertiles of baseline cholesterol, down to a threshold of only 3.5 mmol/l total cholesterol. In the heart protection study 3280 patients (15.7% of the study population) had prior stroke or transient ischaemic attack with or without coexisting coronary disease. Major vascular events were reduced by 5.1% over five years (NNT = 20), and nearly identical benefit was confirmed in the subgroup of 1820 patients with only cerebrovascular disease.
In the face of these results definitions of hypertension and hypercholesterolaemia in any patient with stroke or transient ischaemic attack seem artificial. Irrespective of starting levels of blood pressure almost all patients may benefit from treatment to reduce blood pressure and cholesterol.
Several questions remain unanswered. It is unclear whether the specific combination of perindopril and indapamide is superior to other agents. Since the reduction in the risk of stroke is consistent with predictions from meta-regression analysis of antihypertensive trials,6 any equally effective regimen (resulting in a mean blood pressure reduction of 9/4 mm Hg) may yield similar benefit. However, differences among classes of antihypertensive drugs that are not necessarily explained by blood pressure lowering effectiveness are beginning to emerge.7 w3 Several factors favour the PROGRESS regimen. Firstly, the combination was well tolerated; 90% of patients continued to take treatment for four years or more (although around 1000 patients intolerant of perindopril were excluded in the run-in phase). Secondly, dose titration of perindopril is rapid and simpler than other ACE inhibitors. Thirdly, perindopril may reduce blood pressure without impairing global cerebral blood flow, even in patients with moderate to severe carotid stenosis.8 Fourthly, indapamide differs pharmacologically from thiazides, with less propensity for adverse metabolic effects and some vasodilating actions.
One concern arises from a planned subgroup analysis in PROGRESS that shows no benefit from perindopril monotherapy compared with placebo, despite modest antihypertensive action (mean blood pressure reduction of 5/3 mm Hg). Although the comparison was non-randomised no differences were found between groups to explain away the finding statistically. Some reassurance comes from stroke prevention with ramipril with a nearly identical reduction (4/3 mm Hg) in blood pressure in the heart outcomes prevention evaluation (HOPE) study (similarly irrespective of starting blood pressure).9 HOPE, however, included only 1013 patients with a prior cerebrovascular event—too few for confident conclusions regarding the efficacy of monotherapy with ACE inhibitors in this population. Indapamide, on the other hand, has evidence of benefit as monotherapy from PATS. Overall combined ACE inhibitor and diuretic treatment has the best supporting evidence.
These developments expand secondary preventive options greatly: as many as 84% of patients with transient ischaemic attack and stroke are now prescribed antihypertensive drugs, statins, or non-aspirin antiplatelet agents in addition to any pre-existing medication.w4 However, the probability that lifelong multidrug treatment is recommended places added responsibility on clinicians to improve diagnostic accuracy since only about half of those referred as transient ischaemic attack have actually had one.w5 Since typically 30% of patients referred to special clinics for transient ischaemic attacks have other neurological diagnoses, training in the United Kingdom—where few neurologists participate in stroke services—must take this into account. While the prospect of routine polypharmacy is daunting, both for younger patients unaccustomed to illness and for elderly patients with multiple comorbidities who may be vulnerable to drug interactions and symptomatic hypotension, the consequence of failing to translate the results of this trial into practice will be avoidable disabling or fatal strokes.
That transient ischaemic attack and minor stroke are medical emergencies is reinforced by new community based data that confirm a far higher risk of subsequent stroke than has conventionally been appreciated. The seven day risk of stroke is between 8-12%: other studies have found risk to be as high as 20% in some patient groups.10 11 Whether the benefits of secondary preventative pharmacotherapy extend to this very early period is unknown, but will be tested in ongoing trials of combination antiplatelet therapy (fast assessment of stroke and transient ischaemic attack to prevent early recurrence—FASTER and prevention regimen for effectively avoiding second strokes—PRoFESS), statins (FASTER), and blood pressure lowering (PRoFESS). If these trials find benefits similar to those from PROGRESS and HPS, then the time will have arrived to abandon the still prevalent tendency to dismiss events as “just a TIA” and to advance into an era of acute cerebrovascular syndromes meriting treatment as aggressive as cardiologists currently employ for the coronary equivalent.
Primary care p 326
References w1-w5 are on bmj.com
Competing interests KM has received honorariums for speaking at educational meetings sponsored by Servier and has received a grant from Servier (value £7000) towards a community study of the prevalence of stroke.
Log in using your username and password
Log in through your institution
Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
Sign up for a free trial