New England Journal of Medicine JUNE 20, 2013

Rapid blood-pressure lowering in patients with acute intracerebral hemorrhage

Craig S. Anderson, Else Charlotte Sandset, et al. (INTERACT2 Investigators)

Bottom Line

The INTERACT2 trial found that early intensive blood pressure lowering (systolic target <140 mmHg) did not significantly reduce the primary composite outcome of death or major disability at 90 days compared to guideline-based management (systolic target <180 mmHg), although a secondary ordinal analysis suggested a potential benefit in functional recovery.

Key Findings

1. The primary endpoint of death or major disability (modified Rankin Scale scores 3-6) occurred in 52.0% of the intensive treatment group compared to 55.6% in the guideline-recommended group (OR 0.87; 95% CI 0.75-1.01; P=0.06).
2. A prespecified ordinal analysis of the modified Rankin Scale scores across the entire range (0-6) demonstrated a significant favorable shift in functional outcome favoring the intensive treatment strategy (OR 0.87; 95% CI 0.77-1.00; P=0.04).
3. No significant differences were observed between groups in rates of death or serious adverse events, suggesting that intensive blood pressure lowering is safe in the acute setting.
4. Patients in the intensive treatment group reported better health-related quality of life (EQ-5D utility scores) at 90 days compared to those in the standard care group (0.60 vs. 0.55; P=0.002).

Study Design

Design
RCT
Open-Label, Assessor-Blinded
Sample
2,839
Patients
Duration
90 days
Median
Setting
Multicenter, Global
Population Adults with spontaneous intracerebral hemorrhage within 6 hours of onset and elevated systolic blood pressure (150–220 mmHg).
Intervention Intensive blood pressure lowering with a target systolic blood pressure of <140 mmHg within 1 hour, maintained for 7 days.
Comparator Guideline-recommended blood pressure management with a target systolic blood pressure of <180 mmHg.
Outcome Death or major disability at 90 days, defined as a score of 3-6 on the modified Rankin Scale.

Study Limitations

The trial failed to reach statistical significance on its primary dichotomous outcome, likely due to a lower-than-anticipated treatment effect or cross-over of treatment protocols.
The intensive target was not always achieved within the planned timeframe, potentially diluting the observed clinical benefit.
The study utilized an open-label design, which may introduce observer bias, particularly in the assessment of disability using the modified Rankin Scale.
There was a higher-than-expected rate of protocol deviations where some patients in the standard care group received early blood pressure lowering.

Clinical Significance

While the primary result did not achieve formal statistical significance, the favorable ordinal shift in disability and the demonstrated safety of the protocol suggest that early intensive blood pressure management (target <140 mmHg) is a reasonable and safe strategy for patients presenting with acute spontaneous intracerebral hemorrhage, influencing subsequent clinical practice and guidelines.

Historical Context

Following the pilot INTERACT1 trial, which showed the feasibility of rapid blood pressure reduction and its association with reduced hematoma expansion, INTERACT2 was designed as a definitive Phase 3 trial to assess whether this approach improved long-term clinical outcomes in patients with spontaneous intracerebral hemorrhage.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

What is the physiological rationale for lowering blood pressure acutely in spontaneous intracerebral hemorrhage (ICH), and how does this differ from the management of acute ischemic stroke?

Key Response

The primary goal in ICH is to reduce the hydrostatic pressure exerted on the ruptured vessel to prevent 'hematoma expansion,' which occurs in up to 30% of patients and is a major predictor of poor outcome. In contrast, in acute ischemic stroke (unless thrombolysis is planned), the blood pressure is often allowed to remain higher (permissive hypertension) to maintain collateral flow to the ischemic penumbra.

Resident
Resident

Based on the INTERACT2 results, if you achieve a systolic blood pressure (SBP) of 135 mmHg in a patient within 6 hours of ICH onset, what are the expected clinical benefits and the potential safety concerns compared to a standard target of <180 mmHg?

Key Response

INTERACT2 suggests that intensive lowering (<140 mmHg) is safe and may improve functional recovery (as shown in the shift analysis of the modified Rankin Scale). Unlike the later ATACH-II trial, INTERACT2 did not show a significant increase in renal adverse events, though it also failed to show a significant reduction in the primary outcome of death or major disability (mRS 3-6).

Fellow
Fellow

INTERACT2 reported a p-value of 0.06 for its primary dichotomous outcome (death or major disability) but a p-value of 0.04 for the secondary ordinal analysis (mRS shift). How should a subspecialist interpret these conflicting signals regarding the efficacy of intensive BP lowering?

Key Response

The discrepancy highlights the loss of statistical power when dichotomizing continuous functional scales. The shift analysis suggests a consistent 'slide' toward better outcomes across all levels of disability, which may be more clinically representative than an arbitrary cutoff at mRS 3. However, since the primary endpoint was not met, the results must be viewed as hypothesis-generating or supportive rather than definitive evidence of superiority.

Attending
Attending

When reconciling the INTERACT2 findings with the ATACH-II trial results, how do you determine the 'ideal' blood pressure target for a patient with ICH in the first 24 hours of admission?

Key Response

The 'sweet spot' appears to be achieving a target of 140 mmHg without aggressive overshooting. ATACH-II used more intensive intravenous nicardipine to reach targets faster and showed higher rates of renal injury without benefit. INTERACT2 achieved targets more gradually. Practically, this suggests aiming for 140 mmHg is beneficial, but the speed and intensity of pharmacotherapy should be balanced to avoid acute kidney injury and extreme hypotension (<120 mmHg).

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The INTERACT2 trial utilized an ordinal analysis (proportional-odds model) for the modified Rankin Scale as a secondary endpoint. What are the statistical assumptions required for this model to be valid, and why might this approach be more sensitive than a binary logistic regression in stroke trials?

Key Response

The proportional-odds assumption requires that the treatment effect (odds ratio) is the same across all possible cut-points of the mRS. Ordinal analysis is more sensitive because it utilizes the full distribution of the data, capturing improvements such as a move from mRS 5 to 4, which a binary '0-2 vs 3-6' analysis would count as a failure, thereby increasing the effective power of the study.

Journal Editor
Journal Editor

INTERACT2 was an open-label trial where the primary outcome (mRS at 90 days) was assessed by observers who were aware of the treatment allocation in some cases. As a reviewer, how would you weigh the threat of 'expectation bias' against the 'near-significant' primary p-value of 0.06?

Key Response

In an open-label design, even with blinded adjudicators, the lack of a placebo control for BP management can lead to subtle differences in nursing care or rehabilitation efforts. A p=0.06 is particularly vulnerable to this bias, as even a small number of misclassified mRS scores due to observer enthusiasm for the intensive arm could shift the result from non-significant to significant, demanding a conservative interpretation.

Guideline Committee
Guideline Committee

Current AHA/ASA guidelines (2022) provide a Class 2a recommendation for acute BP lowering to 140 mmHg in patients presenting with SBP between 150-220 mmHg. Why does INTERACT2's lack of success on its primary endpoint prevent this from being a Class 1 recommendation?

Key Response

A Class 1 (Level A) recommendation typically requires multiple large RCTs with significant results on the primary endpoint. Because INTERACT2's primary endpoint was p=0.06 and the subsequent ATACH-II trial was stopped for futility (showing no benefit and potential renal harm), the evidence remains 'moderate' (Class 2a), indicating that while the intervention is reasonable and likely beneficial, the evidence base lacks the definitive strength for a mandatory standard of care.

Clinical Landscape

Noteworthy Related Trials

2010

ENIGMA Trial

n = 114 · Stroke

Tested

Early intensive blood pressure lowering (target SBP <140 mmHg)

Population

Patients with acute intracerebral hemorrhage

Comparator

Standard care (target SBP <180 mmHg)

Endpoint

Hematoma growth at 24 hours

Key result: Early intensive blood pressure management was associated with reduced hematoma expansion compared to standard care.
2013

INTERACT2 Trial

n = 2839 · NEJM

Tested

Intensive blood pressure lowering (target SBP <140 mmHg within 1 hour)

Population

Patients with acute spontaneous intracerebral hemorrhage within 6 hours of onset

Comparator

Guideline-based blood pressure lowering (target SBP <180 mmHg)

Endpoint

Death or major disability at 90 days (modified Rankin scale scores 3–6)

Key result: There was no significant difference in the primary outcome between the intensive and guideline-based groups, though functional outcomes were better in the intensive group on ordinal analysis.
2016

ATACH-2 Trial

n = 1000 · NEJM

Tested

Intensive blood pressure lowering (target SBP 110-139 mmHg)

Population

Patients with acute intracerebral hemorrhage within 4.5 hours of onset

Comparator

Standard blood pressure lowering (target SBP 140-179 mmHg)

Endpoint

Death or disability at 3 months (modified Rankin scale scores 4–6)

Key result: Intensive blood pressure lowering did not result in a lower rate of death or disability compared to standard care and was associated with a higher incidence of adverse events.

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