PDF: ENCHANTED (low dose tPA)
(2016)Objective
To determine whether low-dose intravenous alteplase (0.6 mg/kg) would be noninferior to standard-dose (0.9 mg/kg) with respect to death or disability at 90 days, and whether it would be superior with respect to symptomatic intracerebral hemorrhage.
Study Summary
• The trial did not show noninferiority of low-dose alteplase to the standard dose for the primary outcome of death or disability (OR, 1.09; 95% CI, 0.95 to 1.25).
• Low-dose alteplase was associated with significantly fewer major symptomatic intracerebral hemorrhages (1.0% vs. 2.1%; P=0.01).
• Mortality at 90 days was not significantly different between the two groups (8.5% vs. 10.3%; P=0.07).
• Low-dose alteplase was associated with significantly fewer major symptomatic intracerebral hemorrhages (1.0% vs. 2.1%; P=0.01).
• Mortality at 90 days was not significantly different between the two groups (8.5% vs. 10.3%; P=0.07).
Intervention
Low-dose intravenous alteplase (0.6 mg/kg) vs. standard-dose intravenous alteplase (0.9 mg/kg).
Inclusion Criteria
Patients 18 years of age or older with acute ischemic stroke who met guideline-recommended criteria for treatment with intravenous alteplase.
Study Design
Arms: Low-Dose Alteplase, Standard-Dose Alteplase
Patients per Arm: 1654 in the low-dose group and 1643 in the standard-dose group.
Outcome
• Primary outcome: death or disability (mRS scores 2-6) at 90 days.
• Secondary outcome: symptomatic intracerebral hemorrhage.
• Other outcomes included mortality and functional recovery on the mRS.
• Secondary outcome: symptomatic intracerebral hemorrhage.
• Other outcomes included mortality and functional recovery on the mRS.
Bottom Line
The trial did not establish the noninferiority of low-dose alteplase for the primary outcome of death or disability, but it did show a significant reduction in symptomatic intracerebral hemorrhage compared to the standard dose.
Major Points
- ENCHANTED addressed a critical clinical question, particularly relevant in Asia: can a lower dose of tPA (0.6 mg/kg) achieve similar outcomes to the standard Western dose (0.9 mg/kg) while reducing the feared complication of intracerebral hemorrhage?
- 2-by-2 quasi-factorial design: randomized both tPA dose (0.6 vs 0.9 mg/kg) AND blood pressure management (intensive vs guideline) — the BP arm was published separately. 3,310 patients randomized to the dose comparison.
- Primary endpoint (death or disability mRS 2–6 at 90 days): FAILED to meet noninferiority — 53.2% low-dose vs 51.1% standard-dose (OR 1.09, 95% CI 0.95–1.25, p=0.51 for noninferiority at 1.14 margin). The upper CI crossed the noninferiority margin.
- Dramatically reduced symptomatic ICH: low-dose had half the rate — 1.0% vs 2.1% by SITS-MOST criteria (OR 0.48, p=0.01), and 5.7% vs 8.0% by NINDS criteria (p=0.02). Fatal ICH also halved: 1.3% vs 2.5% (p=0.02).
- Mortality trended lower with low-dose tPA: 7-day death 3.6% vs 5.3% (p=0.02); 90-day death 8.5% vs 10.3% (p=0.07) — the mortality signal was consistent but the trial wasn't powered for this secondary endpoint.
- Ordinal shift analysis (secondary, more sensitive endpoint) showed no significant difference in functional outcome distribution across the entire mRS spectrum — suggesting the practical clinical difference between doses was minimal.
- Predominantly Asian population (63%) — Japan had long used 0.6 mg/kg based on the J-MARS registry, and this trial was designed to validate or refute that practice. The Asian predominance limits direct generalizability to non-Asian populations.
- Low-dose regimen: 0.6 mg/kg, 15% as bolus, 85% as 60-min infusion, max 60 mg. Standard-dose: 0.9 mg/kg, 10% as bolus, 90% as 60-min infusion, max 90 mg. The low-dose has a higher bolus fraction (15% vs 10%) to achieve faster initial recanalization despite lower total dose.
- Median NIHSS was 8 (moderate severity) — relatively mild strokes. The trial included patients with NIHSS as low as 1, where the benefit-risk ratio of any tPA dose is debatable. Post-hoc analyses suggested more severe strokes may have had a wider efficacy gap between doses.
- Influenced Japanese and some Asian-Pacific guidelines that already used 0.6 mg/kg — but Western guidelines (AHA/ASA) maintained 0.9 mg/kg as standard, citing the failed noninferiority result. The 'right' dose remains debated in clinical practice.
Study Design
- Study Type
- International, multicenter, prospective, randomized, open-label trial with blinded outcome evaluation.
- Randomization
- Yes
- Blinding
- Outcome assessors and central hemorrhage adjudicators were blinded to treatment assignments.
- Sample Size
- 3310
- Follow-up
- 90 days.
- Centers
- 111
- Countries
- Australia, UK, Brazil, South Korea, China, Japan, Taiwan, Singapore, Vietnam, Canada, Colombia, Italy, India
Primary Outcome
Definition: Combined end point of death or disability at 90 days (mRS scores of 2 to 6).
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 51.1% (817/1599) | 53.2% (855/1607) | - (0.95 to 1.25) | 0.51 for noninferiority |
Limitations & Criticisms
- Failed to meet noninferiority for the primary endpoint — the upper CI (1.25) crossed the pre-specified noninferiority margin of 1.14, meaning a clinically meaningful efficacy loss with low-dose tPA cannot be excluded.
- Predominantly Asian population (63%) limits generalizability — Asian patients have different stroke subtypes (more small vessel, more ICH risk), lower body weight, and potentially different fibrinolytic response. Western neurologists cannot directly apply these results.
- Median NIHSS of 8 (moderate, not severe) — the most clinically relevant population for this dose question (severe strokes with large vessel occlusion) was underrepresented. Post-hoc analyses suggest the efficacy gap widens with stroke severity.
- Open-label design (PROBE) — treating physicians knew the dose, which could influence post-tPA management decisions (BP targets, antiplatelet timing, monitoring intensity) and introduce performance bias.
- The noninferiority margin of 1.14 (OR) was generous — accepting up to 14% relative increase in death/disability. A stricter margin (e.g., 1.10) would have made noninferiority even harder to achieve.
- 2-by-2 factorial design with BP co-randomization complicated interpretation — potential interaction between tPA dose and BP management was difficult to fully disentangle.
- mRS primarily assessed by telephone at 90 days — less reliable than in-person assessment, though assessors were blinded to treatment assignment.
- Enrolled during 2012–2015, mostly before the thrombectomy era — post-2015, the treatment paradigm shifted dramatically for large vessel occlusions, making the low-dose question less relevant for the most severe strokes (which now go to thrombectomy).
- Half the sICH rate with low-dose is clinically important but was not the primary endpoint — the trial's formal conclusion was 'failed noninferiority,' though many clinicians view the mortality trend and halved ICH as meaningful safety advantages.
Citation
N Engl J Med 2016;374:2313-23. DOI: 10.1056/NEJMoa1515510