Source: Emberson J, Lees KR, Lyden P, et al. Effect of treatment delay, age, and stroke severity on the effects of intravenous thrombolysis with alteplase for acute ischaemic stroke: a meta-analysis of individual patient data from randomised trials. Lancet 2014; 384(9958): 1929-35.
Study Population: 6756 patients in nine randomized trials comparing alteplase with placebo or open control for the treatment acute ischemic stroke. Trials of other thrombolytic agents (e.g. streptokinase, urokinase) were excluded.
Efficacy Endpoints: Excellent functional outcome defined as a modified Rankin Scale (mRS) score of 0-1 assessed at 90-180 days after stroke. The mRS is an inclusive efficacy and harm outcome because it includes harm related to poor outcomes.
Harm Endpoints: Fatal intracerebral hemorrhage, all-cause mortality. (Note that harm endpoints are included in the efficacy endpoint because the mRS includes all-cause mortality as a poor outcome)
Narrative: Stroke is a devastating condition that leaves most survivors with permanent neurological disability. Worldwide, stroke is the second leading cause of death and has a declining age-adjusted incidence in high-income countries but a rapidly rising incidence and impact in lower- and middle-income countries. Ischemic stroke is the most common stroke type, accounting for roughly 85% of all strokes. Ischemic stroke is caused by an acute occlusion of an intracranial artery or one of four extracranial cervical arteries leading to the brain. Treatment with thrombolytic drugs targets the occlusion and is designed to restore blood flow to the brain. Thrombolytic drugs are effective in recanalizing occluded brain arteries only some of the time. Large proximal artery occlusions (e.g. carotid artery or middle cerebral artery stem) are opened 10-35% of the time, and smaller more distal arteries, up to 60-80% of the time. Large proximal artery occlusions are best treated with a combination of thrombolysis and endovascular thrombectomy.1, 2
Initial thrombolytic trials, which assessed streptokinase, were uniformly negative; streptokinase was shown to be harmful, resulting in a higher rate of fatal intracranial hemorrhage and overall greater mortality compared to placebo. As a consequence, streptokinase has been abandoned and should no longer be considered for stroke. Individual trials of alteplase have been both positive and neutral. Treatment efficacy is highly time-dependent, while treatment harm is independent of time, within a 6-hour treatment window. The therapeutic index (i.e, benefit vs harm) is therefore much larger when treatment is administered quickly after stroke onset, and much smaller when given at later times. More recent trials have examined comparisons of alteplase dose (0.6mg/kg vs. 0.9mg/kg) and the use of advanced brain imaging to select patients for treatment who fall outside the usual time-based rules.3, 4, 5
Stroke outcomes are assessed using a global outcome which includes both efficacy (good neurological functional outcome) and harm (death, fatal ICH and ICH leading to poor functional neurological status) outcomes in a single composite measure. The modified Rankin Scale assesses neurologic functional outcome on a 7-point scale from 0 (no symptoms at all) to 5 (bedbound) and 6 (death). Outcome is typically assessed at 90-180 days after stroke onset. The outcome is validated and reproducible. The score is commonly characterized and evaluated statistically as follows: mRS 0-1 = excellent functional outcome; mRS 0-2 = independent functional outcome; mRS 3-6 = dependency or death.
This review summarizes the pooled individual-patient meta-analysis of randomized phase-3 trials comparing alteplase to placebo, comprising 6756 patients.6 Assessments of streptokinase or other thrombolytic agents are not included because they are known to be harmful. The treatment benefit is most pronounced when treatment is delivered rapidly after stroke onset (NNT = 10 for treatment within 3 hours of stroke onset). Beyond 4.5 hours from stroke onset, there is no average benefit. With increasing awareness of the importance of rapid evaluation and treatment and the establishment of acute stroke teams that include emergency physicians and specialist stroke physicians, door-to-treatment times of under 30 minutes can be routinely achieved. Treatment of ischemic stroke due to large vessel occlusion within 30 minutes of hospital arrival, as compared to >30 minutes after hospital arrival, is associated with a NNT of 5; for every 5 patients with large vessel occlusion treated with intravenous alteplase within 30 minutes of hospital arrival, one additional patients will have an independent functional outcome (mRS 0-2).
However, only one third of stroke patients achieve an excellent neurologic outcome with thrombolysis; thus, a majority still fare poorly with only alteplase treatment. [Note that the rising proportion of patients with a good outcome in the control group in later time windows reflects that fact that stroke severity is lower at later time points; more severe stroke presents early.] The combination of alteplase plus endovascular thrombectomy for severe strokes due to proximal large vessel occlusion improves the proportion of patients with independent functional outcome to 50%.1
Stroke treatment with alteplase is associated with an increased risk of fatal intracranial hemorrhage (2.7% vs. 0.4%) in the first 7 days (SITS-MOST criteria type 2 parenchymal hemorrhage - which defines clinical symptomatic and large radiological ICH). The absolute risk increase is 2.0-2.5% (NNH = 40-50).6 Additional non-fatal symptomatic hemorrhage in the first 7 days (SITS-MOST criteria type 2 parenchymal hemorrhage) is increased (1.0% vs 0.2%). The absolute increase is 0.8% (NNH = 125). When hemorrhage is defined radiologically only as PH-2 type hemorrhage (including all fatal, symptomatic and asymptomatic large parenchymal hematoma), there is an absolute increase of PH-2 type hemorrhage in the first 7 days of 5.5% (6.8% vs 1.3%). In addition, the natural history of ischemic stroke is to evolve some degree of asymptomatic (usually petechial) hemorrhage. Asymptomatic hemorrhage can be detected by magnetic resonance imaging in nearly all ischemic strokes, with and without thrombolysis treatment, and is not associated with poorer outcome. Poor outcomes (death or disability) due to hemorrhage are included in the global outcome assessment using the mRS. It is important not to conflate acute hemorrhage with poor outcome at 90 days. While acute hemorrhage is often serious and even fatal (~50% of symptomatic ICH is fatal), it is not necessarily the dominant cause of poorer long-term outcomes; overall, the major prevalent cause of poor outcomes is large ischemic stroke, which are reduced with thrombolysis. The NNH for all-cause mortality is 71.6
There is a tradeoff involved in stroke thrombolysis. Treatment, particularly fast treatment, results in a greater number of patients with independent neurologic functional outcome, but with a small risk of early death due to fatal intracranial hemorrhage.
Caveats: Over the two decades since the first ischemic stroke thrombolytic trials were reported, it has become clear that acute ischemic stroke treatment is best delivered by a team with expertise in emergency critical care, rapid neurological assessment, interpretation of brain imaging, and access to stroke unit care. It is within this kind of structure that rapid, expert treatment can be delivered, and in multiple hospitals around the world this approach has resulted in improved outcomes for stroke patients. Skepticism about thrombolytic treatment has arisen mainly in the emergency medicine literature. However, discussion of the strengths or limitations of the original trials is beyond the scope of this evidence-based summary. When treatment has been attempted in the absence of a collaborative team structure, trial results have not been duplicated. Without the necessary structure, real-world treatment has been either neutral or frankly harmful. To apply this therapy well and duplicate the results of published trials requires the support of a dedicated stroke team. We encourage dialogue and teamwork between stroke neurologists and emergency physicians.
A limitation of meta-analyses is the selection criteria for the included trials. An advantage of pooling individual patient data is that a time-effect can be estimated. Characteristics of the individual trials should be noted. The majority of data on treatment within the early time window (0-3h) originates from two parallel trials conducted in the United States only. Most data from the later-time window (3-6h) come from European trials. Few data are available on patients from Asia. All but one trial used a dose of 0.9 mg/kg alteplase. Four of the trials were exclusively funded by industry; five were funded by public granting organizations. These results are corroborated by the Cochrane meta-analysis that examined trial-level outcomes.7
Treatment of ischemic stroke with intravenous alteplase is strongly recommended (Level 1 or Grade A) by all major stroke guidelines. However, it is an intervention that is best provided by a well-organized and expert stroke team comprising emergency physicians and specialist stroke physicians in a setting that can provide excellent subsequent care in a dedicated stroke unit.
Author: Michael D. Hill, MD; Noreen Kamal, PhD; Eddy Lang, MD
See theNNT.com's previous reviews of this topic: Thrombolytics for Acute Ischemic Stroke, March 25, 2013
Published/Updated: January 11, 2019
The title bar is color-coded with our overall recommendation.