Neurology Clinical Trial Database
ATACH-2
Antihypertensive Treatment of Acute Cerebral Hemorrhage II
Year of Publication: 2016
Authors: Qureshi AI, Palesch YY, Barsan WG, ..., Hanley DF
Journal: The New England Journal of Medicine
Citation: Qureshi AI, et al. N Engl J Med. 2016;375(11):1033–1043.
PDF: Clinical Question
Does intensive systolic blood pressure lowering (110–139 mm Hg) improve outcomes versus standard treatment (140–179 mm Hg) in patients with acute intracerebral hemorrhage?
Bottom Line
Intensive blood pressure lowering did not reduce death or disability compared to standard treatment, and was associated with more renal adverse events.
Major Points
- ATACH-2 is the definitive ICH blood pressure trial establishing that INTENSIVE BP lowering (target SBP 110–139 mmHg) does NOT improve outcomes vs STANDARD (target 140–179 mmHg) and causes MORE renal injury — setting the lower safety boundary for ICH BP management.
- 1,000 patients with spontaneous supratentorial ICH <60 cm³, GCS ≥5, and SBP ≥180 mmHg, randomized within 4.5 hours of onset across 110 centers in 6 countries (USA, Japan, China, Taiwan, South Korea, Germany).
- Primary outcome (death or disability mRS 4–6 at 3 months): 38.7% intensive vs 37.7% standard (RR 1.04, 95% CI 0.85–1.27, p=0.72) — completely null result. Trial stopped early for FUTILITY after planned interim analysis.
- CRITICAL COMPARISON with INTERACT2: Both tested ICH BP targets, but ATACH-2 mandated IV nicardipine protocol with STRICTER targets (110–139 vs INTERACT2's <140) and FASTER achievement (within 2h). INTERACT2 was borderline positive (ordinal p=0.04); ATACH-2 was null — together they suggest ≥140 mmHg is the appropriate target, not lower.
- Renal adverse events significantly higher with intensive treatment: 9.0% vs 4.0% (p=0.002). This renal harm signal was the most important safety finding — rapid aggressive BP lowering in ICH may compromise renal perfusion.
- Hematoma expansion (≥33% volume increase) trended lower with intensive treatment: 18.9% vs 24.4% (p=0.08) — suggesting BP lowering DOES reduce hematoma growth but this biological effect does NOT translate to clinical benefit.
- Achieved SBP was 128.9 mmHg in intensive vs 141.1 mmHg in standard — the 12 mmHg difference between groups was relatively small, raising questions about whether the trial tested a meaningful therapeutic window.
- High proportion had relatively mild ICH: GCS 15 in ~60% of both groups, median hematoma volume ~10–11 mL — enriching for patients with good prognosis regardless of treatment, potentially limiting ability to detect benefit.
- Mandated IV nicardipine protocol — unlike INTERACT2 which allowed flexible agent choice. This rigid protocol may not reflect real-world ICH BP management and limits generalizability to settings without nicardipine.
- Together with INTERACT2, ATACH-2 shaped 2022 AHA/ASA ICH guidelines recommending target SBP 130–150 mmHg for ICH presenting with SBP 150–220 mmHg — avoiding both extremes of too aggressive (<120) and too lenient (>180).
Design
Study Type: Multicenter, randomized, open-label, blinded-endpoint
Randomization: 1
Blinding: Blinded outcome assessment
Enrollment Period: May 2011 – September 2015
Follow-up Duration: 3 months
Centers: 110
Countries: USA, Japan, China, Taiwan, South Korea, Germany
Sample Size: 1000
Analysis: Intention-to-treat
Inclusion Criteria
- Age ≥18 years
- Spontaneous supratentorial ICH <60 cm³
- GCS score ≥5
- Systolic BP ≥180 mm Hg on presentation
- Able to randomize within 4.5 hours of symptom onset
Exclusion Criteria
- Systolic BP <140 mm Hg before randomization
- ICH requiring surgical evacuation
- Secondary causes of hemorrhage
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| Mean age (yr) | 62 | 62 |
| Female sex (%) | 38 | 38 |
| Mean SBP at presentation (mmHg) | 200 | 201 |
| GCS 15 (%) | ~60 | ~60 |
| Median hematoma volume (mL) | 10.3 | 10.2 |
| Intraventricular hemorrhage (%) | 25 | 26 |
| History of hypertension (%) | ~80 | ~80 |
| Prior antihypertensive use (%) | ~60 | ~60 |
| Diabetes (%) | ~15 | ~15 |
| Asian race (%) | ~55 | ~55 |
Arms
| Field | Intensive BP Lowering | Control |
|---|---|---|
| Intervention | Target SBP 110–139 mmHg using IV nicardipine as first-line (dihydropyridine calcium channel blocker, continuous infusion 5–15 mg/hr with rapid titration). Labetalol allowed as second-line. Target achieved within 2 hours and maintained for 24 hours, then transitioned to oral antihypertensives. | Target SBP 140–179 mmHg using IV nicardipine. Treatment initiated only if SBP exceeded 179 mmHg. Same agents as intensive group but with higher trigger threshold and less aggressive titration. |
| Duration | Intensive IV management for 24 hours, then oral antihypertensives | IV management as needed for 24 hours, then oral antihypertensives |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Death or disability (mRS 4–6) at 3 months | Primary | 37.7% | 38.7% | 1.04 | 0.72 |
| Hematoma expansion ≥33% | Secondary | 24.4% | 18.9% | 0.78 | 0.08 |
| Any serious adverse event (3 months) | Secondary | 20.0% | 25.6% | 1.3 | 0.05 |
| Renal adverse events (7 days) | Adverse | 4.0% | 9.0% | 0.002 | |
| Treatment-related SAE (72 hours) | Adverse | 1.2% | 1.6% |
Criticisms
- Stopped early for FUTILITY — only 1,000 of planned 1,280 patients enrolled. While futility stopping was prespecified, it raises the possibility of missing a small but real benefit detectable only with the full sample.
- Relatively SMALL hematomas enriched the population: median volume ~10 mL, GCS 15 in ~60% — these patients have relatively good prognosis regardless of treatment, creating a ceiling effect that limits the ability to detect improvement.
- Narrow BP separation between groups: achieved SBP 128.9 vs 141.1 mmHg (only ~12 mmHg difference) — the therapeutic contrast may have been insufficient to produce a clinical difference. Both groups received active BP treatment.
- Open-label treatment with blinded outcome assessment (PROBE design) — treating physicians knew allocation, potentially affecting post-ICH care decisions, rehabilitation intensity, and withdrawal-of-care discussions.
- Mandated IV nicardipine — not universally available globally and limits generalizability to settings using labetalol, clevidipine, or other agents. INTERACT2's flexible agent approach may be more clinically relevant.
- Renal harm signal (9.0% vs 4.0%, p=0.002) was the most concerning safety finding — rapid aggressive BP lowering may compromise organ perfusion, particularly in elderly patients with pre-existing renovascular disease.
- Excluded large hematomas (>60 mL) and comatose patients (GCS <5) — precisely the population that might benefit most from aggressive hematoma expansion prevention, but was deemed too critically ill for the trial.
- Asian-predominant enrollment (~55%) — while ICH is more common in Asian populations, the drug metabolism, baseline BP patterns, and cerebral autoregulation may differ from other populations.
- Hematoma expansion trended lower with intensive treatment (18.9% vs 24.4%, p=0.08) — showing the BIOLOGICAL effect works but doesn't translate to clinical benefit, suggesting other factors (renal injury, ischemic events) counterbalance the reduced hematoma growth.
Funding
National Institute of Neurological Disorders and Stroke; National Cerebral and Cardiovascular Center
Based on: ATACH-2 (The New England Journal of Medicine, 2016)
Authors: Qureshi AI, Palesch YY, Barsan WG, ..., Hanley DF
Citation: Qureshi AI, et al. N Engl J Med. 2016;375(11):1033–1043.
Content summarized and formatted by NeuroTrials.ai.