Matters arising 255 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.2.255 on 1 February 1994. Downloaded from stroke, Orgaran is superior to low-dose aspirin should be used as well as streptoki- ignored. Until a treatment is found that heparin. nase, or avoided. These presuppositions are works, acute ischaemic stroke treatment Our main conclusion, that the available foolish, especially as we already have the trials should proceed in the most practical randomised trials included in our overview example of the large myocardial infarction and sensible manner possible by adopting were wholly inadequate to determine trials in which, by testing thrombolysis and wide entry criteria and avoiding presupposi- whether or not antithrombotic therapy with aspirin beyond six hours from symptom tions about the effects of treatment. or are onset, and in a very heterogeneous group of J M WARDLAW aspirin, heparin, other agents safe Department ofClinical Neuroradiology, and effective when used in patients with patients with acute myocardial infarction, it Southern General Hospital, acute stroke, is unaltered and will remain so was possible to establish the true time 1345 Govan Road, Glasgow until the large trials in progress (IST, window to treatment (12 hours), the effect G51 4TF, UK TOAST, National Study of Stroke in of age (benefit at all ages) and that throm- China, MAST-I) are completed. bolysis and aspiin together work better 1 Morris AD, Grosset DG, Squire IB, Lees KR, Bone I, Reid JL. The experiences of an PAG SANDERCOCK than either individually.23 acute stroke unit-implications for multi- RI LINDLEY No wonder 95 5% of the patients in the J SLATTERY centre acute stroke trials. J Neurol Neurosurg Neurosciences Trials Unit, study by Morris et al were excluded from Psychiatry 1993;56:352-5. their streptokinase trial: 50% presented 2 ISIS-2 Collaborative Group: Randomised trial Department ofClinical Neurosciences, of intravenous streptokinase, oral aspirin, University ofEdinburgh, after six hours; 23% had a previous cere- both or neither among 17187 cases of acute Bramwell Don Building, brovascular accident with residual clinical Western General Hospital, Crewe Road, myocardial infarction. Lancet 1988;2: deficit; 15-5% had a lacunar infarct; 5-5% 349-60. Edinburgh EH4 2XU, UK 3 Gruppo Italiano per lo Studio della had posterior circulation ischaemia; 22-5% Streptochinasi nell'Infarto Miocardico 1 Sandercock PAG, van den Belt AGM, Lindley had other serious systemic illness (nature in (GISSI): Effectiveness of intravenous RI, Slattery J. Antithrombotic therapy not thrombolytic treatment in acute myocardial acute ischaemic stroke: an overview of the specified). In fact only 32 (haemorrhage on CT), two (tumour on CT), one infarction. Lancet 1986;i:397-402. completed randomised trials. J Neurol 4 Wardlaw JM, Warlow CP. Thrombolysis in Neurosurg Psychiatry 1993;56:17-25. (streptokinase in the past year), seven acute ischaemic stroke-does it work? 2 Gent M, Sackett DL. The qualification and (warfarin treatment), two (?pregnant), 13 disqualification of patients and events in Stroke 1992;23: 1826-39. (bleeding tendency or DU), nine (transient 5 Sandercock PAG, van den Belt AGM, Lindley long-term cardiovascular clinical trials. RI, Slattery J. Antithrombotic therapy in Thromb Haemost 1979;41:123-34. ischaemic attack), 12 (not clinical stroke) acute ischaemic stroke: an overview of the 3 Turpie AG, Gent M, Cote R, et al. A low- had true contraindications to streptokinase molecular-weight heparinoid compared with completed randomised trials. J7 Neurol were 1993;56:17-25. unfractionated heparin in the prevention of and most of these potentially-ineligible Neurosurg Psychiatry deep vein thrombosis in patients with acute for the IST for the same reasons. ischemic stroke. A randomized, double- Who are we clinicians to decide, on the blind study. Ann Intern Med 1992;117: no Dr et 353-7. basis of evidence whatsoever,4 that Lees al reply: patients with a previous cerebrovascular We thank Dr Wardlaw, who is the United accident, or who reach hospital after six Kingdom representative of MAST-I, for her hours (most patients with stroke in the letter. Despite her criticism of the cautious Mast trials United Kingdom) or who have a lacunar or entry criteria for the intemational version mild cortical infarct, etc, are unlikely to of MAST, more patients have been ran- The study by Morris et al on recruitment benefit from a particular acute stroke treat- domised to MAST than to MAST-I in the for acute stroke treatment trials of patients ment, never mind thrombolysis? If treat- United Kingdom. with stroke admitted to hospital illustrates, ments are not tested in a practical manner We agree that stroke trials should adopt not treatment eligibility, should these treat- in a representative group of patients, then wide entry criteria without prejudging the ments be proven to work, but simply how the trial result will never be applicable to results. We also believe, however, in an artificially restrictive selection process the generality of patients who suffer an restricting exposure to potentially danger- can hinder trial recruitment. acute ischaemic stroke, and important ous treatments to patients in whom the In their study, the two trials compared benefits may be missed. risk/benefit ratio justifies intervention. We have very different exclusion criteria, not It is important to understand that the are not prepared to disregard evidence dictated simply by the standard contra- MAST trial described by Morris et al is not regarding treatment from experimental indications to the treatments being tested, the same as the Multicentre Acute Stroke studies, large clinical studies of thromboly- and are addressing very different questions. Trial-Italy (MAST-I). MAST-I is the sis after acute myocardial infarction and

In the International Stroke Trial (IST) all largest randomised controlled trial of pilot studies after stroke. The selection of a http://jnnp.bmj.com/ types of acute ischaemic stroke are eligible thrombolysis in acute ischaemic stroke so homogeneous group of patients without up to 48 hours after onset unless severely far, with more than 440 patients prior disability is aimed at maximising the disabled or there is a clear contraindication randomised (most in Italy but some in the chance of a statistically meaningful result. to aspirin or heparin, such as active duo- United Kingdom) and strong encourage- Experimental evidence suggests that the denal ulcer. The aim of the IST is to answer ment from its Data Monitoring Committee therapeutic window for successful neuro- the simple question: "Do aspirin, or not only to continue, but also to expand to protection through reperfusion is under six heparin, or both improve outcome after enhance recruitment. MAST-I is testing hours.' Although the ISIS-3 study reported acute ischaemic stroke"? The trial was streptokinase, aspirin, both or neither (like intracerebral haemorrhage in under 1% of designed to include as heterogeneous a the Italian Group Studying Streptokinase in patients treated with thrombolysis after on September 27, 2021 by guest. Protected copyright. group of patients with acute stroke as myocardial infarction (GISSI) and ISIS-2)2 3 myocardial infarction, the incidence of fatal possible, so that in future, physicians would in all types and severities of acute ischaemic intracranial haematoma in pilot studies of know accurately the risks and benefits of stroke. It has a six-hour time window to thrombolysis after stroke has been up to aspirin and heparin treatment when treating treatment which is likely to be extended in 10%.23 Haemorrhage was less common in almost any such patient. the near future. At the end of MAST-I, a patients treated within 90 minutes of stroke In contrast, in the Multicentre Acute physician faced with a patient with stroke onset.4 Outcome after stroke is variable, but Stroke Trial (MAST), to which Morris et al will have useful information on the risks generally much better after lacunar or small referred, a very restricted question is being and benefits of streptokinase and aspirin, cortical infarcts than after large MCA asked: "Does streptokinase improve out- together and separately, applicable to that infarction. Inclusion of patients with come after major middle cerebral artery individual patient. inevitably bad outcome due to preexisting (MCA) territory ischaemic stroke if started Clinical trials should be designed to disability, or a high probability of good out- within six hours?" Consequently the result answer practical questions on the risks and come due to minor stroke, would confound of this trial will only apply to a very benefits of treatmnent for as many patients as assessment of outcome. restricted group of patients with acute possible, especially for conditions as We consider that it is responsible to await stroke-those with major MCA occlusions common as acute ischaemic stroke. Let us evidence that thrombolysis is of benefit reaching hospital in time to be examined, not make the mistake of equating trial under optimal conditions before progressing investigated and treatment started within six eligibility with treatment eligibility, nor to milder forms of stroke, treated late. This hours. In other words, the trial design pre- make assumptions about when promising, is not prejudice; it is caution. Other treat- supposes that streptokinase will not work but largely untested, treatments4 5 are likely ments that have a better safety profile after six hours, or in small cortical, or to work. The lessons from the acute may be tested in wider groups of patients; lacunar, or posterior circulation strokes. It myocardial infarction trials of thrombolytic in our Acute Stroke Unit at the Westem will not yield any information on whether and antithrombotic drugs should not be Infirmary we give 10 patients at random