Interpretation of IST and CAST stroke trials

Abstract

IST Collaborative Group's reply SIR—We are delighted that, in general, your correspondents agree that the successful completion of the IST and CAST “megatrials” is a step forward in the treatment of acute stroke. Moreover, these two trials demonstrate that the use of aspirin in early ischaemic stroke produces a small but real (2p=0·001) reduction of nine deaths or recurrent strokes per 1000 patients treated for the first few weeks. By contrast, in IST, subcutaneous heparin (both doses combined) was not associated with a reduction in early death or recurrent stroke, with a small decrease in recurrent ischaemic strokes offset by a similar-sized increase in haemorrhagic strokes. In her commentary on IST, Marie-Germaine Bousser noted that the low-dose heparin regimen was associated with a reduction in early death or recurrent stroke of 12 per 1000. However, unlike the clear evidence of benefit with aspirin, this was of only marginal statistical significance (2p=0·03). Furthermore, she was not correct when she stated that “low-dose heparin had a better benefit-risk ratio than aspirin”. In IST low-dose heparin was associated with significant increases in both major extracranial bleeds (0·6% vs 0·4%, 2p=0·05) and haemorrhagic strokes (0·7% vs 0·4%, 2p=0·02) compared with no heparin. The excess of extracranial bleeds with aspirin in the absence of heparin was only two per 1000 patients treated and was not significant (2p=0·1). And, in contrast with the trend to improved outcome at 6 months with aspirin, patients allocated low-dose heparin had a somewhat worse long-term outcome (five extra deaths [NS] and two extra dead or dependent [NS] for every 1000 patients treated with low-dose heparin). It also seems unwise to recommend that full-dose anticoagulation should be started early in acute stroke in patients with atrial fibrillation (AF), given the definite risk of major haemorrhage and the lack of short or long-term benefit seen in the medium-dose heparin group in IST (in patients with and without AF). In the European Atrial Fibrillation Study,1EAFT (European Atrial Fibrillation Trial) Study GroupSecondary prevention in non-rheumatic atrial fibrillation after transient ischaemic attack or minor stroke.Lancet. 1993; 342: 1255-1262Summary PubMed Scopus (1809) Google Scholar which showed that anticoagulation was more effective than aspirin in long-term secondary prevention in patients with AF, most patients were not anticoagulated within 14 days of their stroke or transient ischaemic attack (when the risk of intracerebral haemorrhage is highest). Taking IST and CAST together, the beneficial effect of aspirin on early death or recurrent stroke among patients with AF is not significantly different from the effect in patients without AF. We would therefore suggest that patients with acute ischaemic stroke and AF be initially started on aspirin and that, in those suitable, long-term anticoagulation is started after 10–14 days. Anna Czlonkowska and Grazyna Maria Szpak and Marie-Germaine Bousser felt the IST results were biased against heparin, chiefly because there was no blinding of heparin with a placebo. It seems unlikely, however, that this lack of blinding materially altered the main findings. The short-term effects of heparin in IST, including the effect on symptomatic intracranial haemorrhage, were similar to those in previous placebo-controlled trials and were also similar to the preliminary results of the TOAST study (a placebo-controlled trial of a heparinoid in acute ischaemic stroke) presented at the European Stroke Conference.2James A Stroke treatment trials yield disappointing results.Lancet. 1997; 349: 1673Summary Full Text Full Text PDF Google Scholar Bousser also criticised the lack of anticoagulant monitoring in IST, but in his letter Adam Cohen recognises the practical difficulties of doing so. We, and no doubt Cohen too, were disappointed that TOAST showed that even carefully dose-adjusted intravenous anticoagulation with a more refined agent did not produce net benefit in acute ischaemic stroke. There is no good evidence, from IST or other trials, that any anticoagulant regimen produces any net long-term clinical benefit when used early in acute stroke, either overall or in patients with AF (or indeed in any other identifiable category of patient). Writing before publication of CAST Bousser said “whether the results of that [CAST] trial, and a meta-analysis that includes the IST data, show that aspirin is effective remains to be seen”. Now that the results of both trials are available it is clear, as Stefano Ricci and colleagues point out, that aspirin is the preferable antithrombotic regimen in acute ischaemic stroke.