1.      Intravenous rt-PA within 3 hours after the onset of symptoms in patients with acute ischaemic stroke is a highly effective treatment in selected patients (grade A evidence). The use of rt-PA is supported by results from randomised controlled trials, pooled analysis of six randomised rt-PA trials, meta-analyses and post-approval data.  According to meta-analyses, for patients given rt-PA within 3 hours of ischaemic stroke approximately 1 out of 10 more will be independent, 1 of 22 will suffer symptomatic haemorrhage and 1 in 100 fewer may die as a result of the treatment (grade A evidence).

 2.      The pooled analysis of the 2776 patients of the ATLANTIS, ECASS and NINDS rt-PA trials suggest benefit up to 4 ½ hours after the onset of an ischaemic stroke but the treatment benefit decreases with the time. Odds ratios for favourable outcomes with rt-PA in the pooled analyses were 2.8 (95%CI 1.8-4.5) in the 0-90, 1.5 (95%CI 1.1-2.1) in the 91-180, 1.4 (95%CI 1.1-1.9) in the 181-270 and 1.2 (95%CI 0.9-1.5) in the 271-360 minute time windows. The use of rt-PA up to 6 hours is supported by the results of a meta-analysis of tabular data (grade A evidence) but further randomised trials will be required to determine the latest time window (grade A evidence).

 3.      Outside randomised controlled trials, it is recommended that the therapeutic use of intravenous rt-PA should follow published guidelines (with local modifications as appropriate), and the labelling approved by Health Authorities and should be subject to continuous quality control.

 4.      In most open studies the safety and efficacy of intravenous rt-PA in routine clinical practice is comparable to randomised studies. A meta-analysis of 12 post-approval open series with 2529 patients treated within 3 hours revealed comparable rates of symptomatic haemorrhages (5.1 vs 6.4 % in the NINDS trial), of favourable outcome as defined by mRS 0-1 pts (37,1 vs 39% in the NINDS trial) and of initial stroke severity as defined by the baseline NIHSS (14 vs 14 in the NINDS trial).

 5.      Thrombolysis within 3 hours has been approved in the United States, Canada, Brazil, Germany and recently also in the European Union. The evidence strongly supports that rt-PA is made available for routine clinical use to treat stroke patients in adequately resourced and experienced centres. However, there is an obvious need for continuous education and for trained stroke specialists to guarantee safe and efficient use of thrombolysis. The development of hospital services designed to deliver early thrombolysis 24 hours a day for acute ischaemic stroke is encouraged.

 6.      The approval of rt-PA in ischaemic stroke by the European Union Health Authorities was conditional. The first condition was that all treated patients should be included in the Safety Implementation of Thrombolysis in Stroke - The Monitoring Study register (SITS-MOST). The second condition was a double-blind, randomised, placebo-controlled trial of IV rt-PA in ischaemic stroke where thrombolysis is initiated between 3 and 4 hours after stroke onset (ECASS III).

 7.      The heterogeneity for good outcome in meta-analyses implies that more data are needed on how to identify the patients most likely to benefit and least likely to be harmed by thrombolysis. The role of patient characteristics, stroke severity, stroke subtype, vascular lesions, stroke pathophysiology, concomitant disease, and drug treatments should be further evaluated in future trials, meta-analyses, phase IV studies and the SITS-MOST register.

 8.      It is recommended that some future trials of safety and efficacy of thrombolysis should assess modern imaging techniques as a part of the protocol to help in patient selection and in monitoring of the effects of therapy. Extended ischaemic oedema as detected by CT should weigh against the use of thrombolysis (grade B evidence). MR diffusion and perfusion weighted imaging may reveal brain tissue at risk that can be salvaged with thrombolysis in individual patients within a 3-hour time window and possibly even longer (grade C evidence). Other imaging modalities including perfusion CT, MR angiography, SPECT and TCD may be useful in selecting patients, in assessing arterial pathology and in monitoring the effects of therapy (grade C evidence). However, further data from randomised trials incorporating these new imaging modalities are required, as - although each may provide extra prognostic information - they may also increase time to treatment and hence reduce the benefit of rt-PA.

 9.      The limitation of the 3-hour time window is the major factor, which prevents wider application of thrombolysis in ischaemic stroke. ECASS III, IST-3 and EPITHET aim at extending the time window. This may have an important public health impact in Europe. It is recommended that new thrombolytic agents, wider time windows and thrombolysis together with neuroprotective agents, should be evaluated in future randomised trials to try to increase the effectiveness and to decrease the risks of thrombolysis.

 10. Another factor that prevents the wider application of thrombolysis in ischaemic stroke is the concern that brain haemorrhage may harm the patient. In the pooled analysis (ATLANTIS, ECASS and NINDS), the rate of large space occupying haematoma (PH) was 1.1% in the placebo group and 5.9% in the rt-PA group. Despite this 5.4-fold risk, rt-PA significantly increased the overall chance of not being disabled or dead after stroke. This observation suggests, that not all of these haemorrhages were symptomatic. Accordingly, the definition of symptomatic brain haemorrhage should fulfil the following criteria: “Clinical deterioration by > 4 p NIHSS (or equivalent) combined with brain haemorrhage without other pathology, or clinical deterioration by > 4 p NIHSS (or equivalent) combined with cerebral haemorrhage within an ischaemic lesion where any mass effect is largely attributed the haemorrhage. Any haemorrhage in large space-occupying infarction should not be considered as “symptomatic haemorrhage” but rather “ischaemic oedema”, and data should be collected on ischaemic oedema in future trials.

 11. The impact of antiplatelet treatment prior to the stroke on the risk of symptomatic haemorrhage following stroke needs to be studied further although existing data do not indicate an increased risk of symptomatic haemorrhage (grade B evidence). Furthermore, the safety and efficacy of postthrombolytic antithrombotic treatment and anticoagulation to prevent re-occlusion needs to be analyzed in randomised trials.

  12. Future studies should include collection of data to allow the assessment of the impact on health economics and on quality of life of rt-PA in acute ischaemic stroke. The public should be educated of the value of early expert assessment and treatment.