The New England Journal of Medicine December 14, 1995

Tissue Plasminogen Activator for Acute Ischemic Stroke

The National Institute of Neurological Disorders and Stroke (NINDS) rt-PA Stroke Study Group

Bottom Line

A landmark, randomized trial demonstrating that intravenous tissue plasminogen activator (t-PA) administered within 3 hours of acute ischemic stroke onset significantly improves functional outcomes at 3 months, despite an increased risk of early symptomatic intracerebral hemorrhage.

Key Findings

1. In Part 1, there was no significant difference between the t-PA and placebo groups in the percentage of patients with early neurologic improvement (≥4-point NIHSS improvement or resolution of deficit) at 24 hours.

2. In Part 2, patients treated with t-PA had a significantly higher likelihood of a favorable clinical outcome across all four assessment scales at 3 months, with a global odds ratio for favorable outcome of 1.7 (95% CI, 1.2 to 2.6) [1.1].

3. Compared to placebo, the t-PA group demonstrated a 12% absolute increase in the number of patients achieving minimal or no disability on the Barthel index (score of 95-100) and an 11% absolute increase for those achieving an NIHSS score of 0 or 1 at 3 months.

4. Overall, patients treated with t-PA were at least 30% more likely to have minimal or no disability at 3 months compared to those receiving placebo.

5. Symptomatic intracerebral hemorrhage within 36 hours occurred in 6.4% of patients given t-PA compared to only 0.6% of patients given placebo (P < 0.001).

6. Despite the increased bleeding risk, mortality at 3 months was similar between the groups (17% in the t-PA group vs. 21% in the placebo group, P = 0.30).

Study Design

Design

RCT

Double-Blind

Sample

624

Patients

Duration

3 mo

Median

Setting

Multicenter, US

Population Adults presenting with an acute ischemic stroke with a measurable neurologic deficit on the NIHSS, capable of receiving treatment within 3 hours of symptom onset, and with a baseline non-contrast head CT showing no evidence of intracranial hemorrhage.

Intervention Intravenous recombinant tissue plasminogen activator (t-PA, alteplase) at a dose of 0.9 mg/kg (maximum 90 mg), with 10% administered as an initial bolus and the remainder infused over 60 minutes.

Comparator Matching intravenous placebo administered in an identical fashion.

Outcome Part 1 (n=291): Neurologic improvement of ≥4 points over baseline NIHSS score or complete resolution of neurologic deficit at 24 hours. Part 2 (n=333): Global test statistic assessing favorable clinical outcome at 3 months based on four scales: the Barthel index, modified Rankin scale, Glasgow outcome scale, and NIHSS.

Study Limitations

• The intervention carries a significant 6.4% risk of symptomatic intracerebral hemorrhage, emphasizing the need for strict adherence to inclusion and exclusion criteria.

• The strict 3-hour therapeutic window requires robust, highly coordinated prehospital and emergency department stroke systems, limiting the proportion of stroke patients eligible for therapy.

• Patients with severe hypertension (>185/110 mm Hg), recent bleeding, or rapidly improving symptoms were excluded, limiting the generalizability of the findings to broader patient populations.

• The two-part study design, in which Part 1 failed to meet its primary endpoint of early neurologic improvement at 24 hours, fueled years of debate within the emergency medicine community regarding the drug's efficacy and safety profile.

Clinical Significance

This landmark study fundamentally revolutionized stroke neurology. By proving that acute ischemic stroke is a treatable emergency if intervened upon quickly, it led to the FDA approval of intravenous alteplase. This catalyzed the creation of comprehensive stroke centers, emergency 'stroke alert' protocols, and the modern 'time is brain' paradigm in clinical practice.

Historical Context

Prior to 1995, ischemic stroke was viewed largely as an untreatable condition managed with supportive care and rehabilitation. Early thrombolytic trials in stroke using streptokinase were halted prematurely due to unacceptably high rates of fatal intracerebral hemorrhage. The NINDS trial was the first to successfully demonstrate that by using a highly specific recombinant agent (rt-PA) and restricting the treatment window to under 3 hours, the long-term functional benefits outweighed the acute hemorrhagic risks.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

What is the mechanism of action of tissue plasminogen activator (t-PA), and why did the NINDS trial establish a strict 3-hour time window for its administration?

Key Response

t-PA is a thrombolytic agent that converts plasminogen to plasmin, an enzyme that degrades fibrin clots to restore blood flow. The 3-hour window is critical because of the ischemic penumbra; beyond this timeframe, the brain tissue undergoes irreversible infarction, and the breakdown of the blood-brain barrier significantly increases the risk of hemorrhagic transformation, causing the risks of t-PA to outweigh its benefits.

Resident

Based on the NINDS trial criteria, what are the absolute contraindications to administering IV t-PA, and what initial steps must be taken in the emergency department to rule them out?

Key Response

Residents must rapidly assess for bleeding risks. Key steps include obtaining a non-contrast head CT to rule out intracranial hemorrhage, checking a point-of-care blood glucose to rule out hypoglycemia (a common stroke mimic), and ensuring blood pressure is strictly below 185/110 mm Hg. History must be quickly screened for recent trauma, surgery, bleeding diathesis, or active anticoagulant use.

Fellow

The NINDS trial demonstrated improved functional outcomes but no overall mortality benefit, alongside a 6.4% rate of symptomatic intracerebral hemorrhage (sICH). How do you counsel a patient's family using the Number Needed to Treat (NNT) versus Number Needed to Harm (NNH) under hyperacute time pressure?

Key Response

Fellows must master high-stakes shared decision-making. The NNT for a near-complete recovery (mRS 0-1) is approximately 8, while the NNH for a devastating sICH is roughly 17. Effectively communicating this nuanced risk/benefit ratio—highlighting the chance of functional independence against the risk of bleeding without a change in overall mortality—is a crucial neurovascular skill.

Attending

In the pursuit of increasingly short 'door-to-needle' times driven by the NINDS trial's findings, patients with 'stroke mimics' inevitably receive t-PA. How does the pathophysiology of stroke mimics influence the risk profile of t-PA in these cases, and how does this shape your clinical threshold to treat?

Key Response

This addresses a major real-world attending dilemma. The teaching point is that the risk of hemorrhagic transformation in stroke mimics (e.g., complex migraines, seizures, conversion disorder) is remarkably low because there is no actual ischemic brain tissue undergoing blood-brain barrier breakdown. This physiological fact justifies the aggressive 'treat first' paradigm when clinical suspicion of stroke is high, as delaying for certainty costs irreplaceable neurons.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The NINDS trial uniquely utilized a 'global statistic' combining four different outcome scales (Barthel Index, modified Rankin Scale, Glasgow Outcome Scale, and NIHSS) to assess efficacy at 3 months. What are the statistical advantages and methodological assumptions of using a global odds ratio compared to a single dichotomized primary endpoint?

Key Response

A global statistic increases statistical power and captures a broader, more robust spectrum of neurological recovery across multiple validated domains. However, it relies on the proportional odds assumption (that the treatment effect is consistent across all cut-points of the scales) and assumes the scales move in a correlated direction, which can complicate simple clinical interpretability compared to a traditional single dichotomous endpoint.

Journal Editor

Critics of the NINDS trial historically flagged a baseline imbalance in stroke severity (NIHSS scores) between the t-PA and placebo groups in Part 2 of the study. As a peer reviewer, how would you evaluate the threat this confounding variable poses to the trial's internal validity, and what specific analyses would you demand prior to publication?

Key Response

An editor would immediately flag baseline imbalances in severity as a major threat to internal validity, as milder strokes naturally have better outcomes. The necessary editorial demand would be a multivariable logistic regression adjusting for baseline NIHSS scores to ensure the treatment effect of t-PA remains statistically significant and robust independent of the initial severity imbalance (which subsequent re-analyses of NINDS successfully confirmed).

Guideline Committee

How did the results of the NINDS trial inform the initial AHA/ASA Class 1 guidelines for acute ischemic stroke, and how must current guidelines synthesize the NINDS 3-hour window with subsequent ECASS III data and modern endovascular thrombectomy (EVT) trials?

Key Response

NINDS established the foundational Class I, Level A recommendation for IV alteplase within 3 hours. Guidelines later integrated ECASS III to extend the window to 4.5 hours for eligible patients (with stricter exclusion criteria). Current guidelines mandate that IV t-PA should still be administered without delay even if the patient is eligible for mechanical thrombectomy (bridging therapy), though committees are actively reviewing recent direct-to-thrombectomy non-inferiority trials that challenge this sequential paradigm.

Clinical Landscape

Noteworthy Related Trials

2008

ECASS III Trial

n = 821 · NEJM

Tested

Intravenous Alteplase (0.9 mg/kg)

Population

Acute ischemic stroke patients 3 to 4.5 hours after symptom onset

Comparator

Placebo

Endpoint

Disability at 90 days (modified Rankin scale score 0 or 1)

Key result: Intravenous alteplase administered between 3 and 4.5 hours after the onset of symptoms significantly improved clinical outcomes compared to placebo.

2012

IST-3 Trial

n = 3035 · Lancet

Tested

Intravenous Alteplase (0.9 mg/kg)

Population

Acute ischemic stroke patients within 6 hours of symptom onset

Comparator

Open-label control (standard care)

Endpoint

Proportion of patients alive and independent (OHS score 0-2) at 6 months

Key result: While the primary endpoint was not significantly different, ordinal analysis showed improved functional outcomes, particularly benefiting patients over 80 years old.

2018

WAKE-UP Trial

n = 503 · NEJM

Tested

Intravenous Alteplase (0.9 mg/kg)

Population

Acute ischemic stroke patients with unknown time of onset and MRI mismatch

Comparator

Placebo

Endpoint

Favorable outcome at 90 days (modified Rankin scale score 0 or 1)

Key result: Intravenous alteplase guided by a mismatch between clinical deficit and MRI findings resulted in significantly better functional outcomes at 90 days compared to placebo.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis