INTERACT2

Clinical Context

The onset of intracerebral hemorrhage (ICH) is frequently accompanied by an acute hypertensive response. This surge in blood pressure can reach very high levels and is associated with the detrimental expansion of the hematoma and consequently, poorer patient outcomes. This observed relationship fostered the hypothesis that rapidly lowering blood pressure might mitigate the risk of hematoma expansion, thereby reducing the likelihood of a poor clinical outcome.

However, the potential benefits of aggressive blood pressure reduction must be carefully weighed against the perceived risk of compromising cerebral perfusion, particularly in patients with a history of chronic hypertension.

Prior to INTERACT2, existing AHA/ASA Guidelines (2010) recommended maintaining a systolic blood pressure (SBP) below 180 mmHg and/or a mean arterial pressure (MAP) below 110 mmHg. These recommendations aimed to balance the theoretical reduction in hematoma expansion risk with the need to maintain adequate cerebral perfusion pressure. Notably, this guidance was not founded on robust evidence from high-quality randomized trials.

Published in 2013, INTERACT2 was the first large-scale randomized trial designed to rigorously evaluate the safety and efficacy of early, intensive blood pressure lowering on clinical outcomes in patients with acute ICH.

Patient Population

The trial enrolled patients who presented within six hours of ICH onset and had an initial SBP between 150 and 220 mmHg. Key exclusion criteria included an admission Glasgow Coma Scale (GCS) score of 3-5, a perception by clinicians of an inevitably poor prognosis, or if early surgical evacuation of the hematoma was planned.

The study population had a mean age of approximately 63 years, with two-thirds being male. A significant portion (about two-thirds) was recruited from China, with other participants enrolled from centers in Europe, the United States, India, and South America.

Most patients presented with mild to moderate ICH, as indicated by a median GCS of 14 (IQR 12–15), a median NIH Stroke Scale (NIHSS) score of 10 (IQR 6–15), and a median hematoma volume of 11 mL (IQR 6–19). Approximately one-third of bleeds exhibited intraventricular extension, and the vast majority (83%) were in a deep hematoma location. Relatively few patients required advanced interventions at baseline: only about 7% were intubated, and 3% had an external ventricular drain (EVD).

The average SBP on presentation was 179 mmHg. This baseline SBP is notably lower than that observed in the subsequent ATACH-II trial, where patients presented with an average SBP of 200 mmHg, highlighting potential differences in the populations or baseline management approaches.

Study Design

INTERACT2 was a multicenter, international trial employing a Prospective, Randomized, Open-label, Blinded Endpoint (PROBE) design, involving over 100 sites.

Protocol Details

Patients were randomized into one of two treatment arms:

• Intervention Group: Assigned to intensive blood pressure control with a target SBP of <140 mmHg. This target was informed by previous observational studies suggesting that an SBP <140-150 mmHg was associated with less frequent hematoma expansion compared to higher targets.
• Control Group: Assigned to guideline-recommended blood pressure control with a target SBP of <180 mmHg, based on contemporary AHA guidelines. No specific lower limit was mandated for this group.

For both groups, the protocol recommended achieving the target SBP within one hour of randomization and maintaining this goal for the subsequent seven days. The choice of specific antihypertensive medications was left to pre-specified local protocols, though it was suggested that the initial agent for the intensive group be an intravenous medication, while the guideline group could start with oral agents. All patients received care in an acute stroke unit.

A planned sub-study aimed to include 600 patients (300 Asian and 300 non-Asian) who would undergo repeat brain imaging at 24 hours to assess for differences in hematoma expansion between the groups.

Power Analysis

The trial was powered (at least 90%) to detect a 14% relative risk reduction (RRR), corresponding to a 7% absolute risk reduction (ARR), in the primary outcome of death or major disability (mRS 3-6) at 90 days. This was based on an assumed event rate of 50% in the control group and 43% in the intervention group. Aiming for a 7% ARR was considered a clinically relevant and reasonable expectation for an intervention of this nature.

Risk of Bias Analysis

(Content for this section needs to be added if you plan to include it.)

Results

Achieved Blood Pressure Control

A critical aspect of trials manipulating physiological parameters is the actual separation achieved between groups. In INTERACT2, there was a clear and rapid divergence in SBP levels. The intensive group achieved a mean SBP of 150 mmHg within one hour, which further decreased towards the <140 mmHg target. The guideline treatment group also experienced a reduction in SBP, with a mean SBP around 150 mmHg by 24 hours, substantially lower than their <180 mmHg target. This observation suggests that clinicians, even in the control arm, may have been targeting SBP levels lower than prevailing guideline recommendations, possibly influenced by emerging data or the pilot INTERACT trial. This smaller-than-anticipated difference in achieved SBP between groups is an important consideration when interpreting the trial’s outcomes.

Primary Outcome: Death or Major Disability

The primary endpoint was the proportion of patients who died or had major disability (mRS score 3–6) at 90 days post-symptom onset. In the intensive treatment group, 52.0% of patients experienced this outcome, compared to 55.6% in the guideline treatment group. This resulted in an odds ratio (OR) of 0.87 (95% Confidence Interval [CI] 0.75–1.01; p=0.06), indicating a trend towards benefit with intensive lowering that did not reach conventional statistical significance for this dichotomized outcome.

Secondary Outcome: Ordinal Analysis of mRS

As a pre-specified secondary outcome, an ordinal analysis of the full range of mRS scores at 90 days was performed. This analysis suggested a statistically significant shift towards better functional outcomes with intensive treatment (common OR 0.87, 95% CI 0.77–1.00; p=0.04).

The consistency of the odds ratios between the dichotomized primary outcome and the ordinal analysis of mRS scores (both OR 0.87) suggests a small but potentially favorable effect of intensive blood pressure lowering across the spectrum of disability.

Hematoma Expansion Sub-study

The imaging sub-study, which included 473 patients in the guideline group and 491 in the intensive group with repeat imaging, did not find a statistically significant difference in absolute or proportional hematoma growth at 24 hours between the two treatment arms.

Safety Outcomes

There was no evidence of a difference in the rates of neurologic deterioration within the first 24 hours or in the overall incidence of other serious adverse events between the intensive and standard treatment groups. This suggests that rapid intensive blood pressure lowering to a target of <140 mmHg was generally safe in this population.

Final Thoughts & Critical Appraisal

INTERACT2 was a landmark achievement in ICH research, being the first large RCT to directly address the optimal acute blood pressure target. Its robust design and execution have significantly influenced clinical practice and subsequent guideline recommendations.

Several points warrant further discussion:

• Interpreting the “Non-Significant” Primary Outcome: It’s a common misinterpretation to conclude “no difference” or “no effect” solely based on a p-value exceeding 0.05 for the primary dichotomized outcome. The confidence interval (0.75–1.01) for the primary outcome is largely on the side of benefit for intensive treatment and includes clinically important reductions in death or major disability. This highlights the “absence of evidence is not evidence of absence” principle.
• Consistency of Ordinal vs. Dichotomized mRS: The ordinal analysis of mRS, which was pre-specified, yielded a p-value of 0.04, suggesting a favorable shift in outcomes. The similar odds ratios between the two analyses (0.87) reinforce the likelihood of a small, positive treatment effect. The slight difference in p-values across the 0.05 threshold should not be over-interpreted as a contradictory finding.
• Magnitude of Effect: While there appears to be a probable benefit to intensive blood pressure control, the effect size is likely modest. However, even a small absolute risk reduction can be meaningful at a population level for a common and devastating condition like ICH. Given that intensive BP management is feasible in most acute care settings, often without substantial marginal cost, this modest benefit may still be worthwhile.
• Influence of Achieved BP Differences: The relatively small difference in achieved SBP between the groups (particularly after the first few hours, with the control group also achieving SBPs around 150 mmHg) might have attenuated the observed treatment effect. A larger separation in BP might have yielded a more pronounced difference in outcomes.
• Hematoma Expansion Findings: The lack of a clear signal for reduced hematoma expansion in the imaging sub-study was somewhat unexpected, given the presumed mechanism of benefit. This could be due to the sub-study being underpowered for this specific endpoint, the timing of imaging, or other complex factors influencing hematoma growth.

How do we reconcile the results with other studies?

(Content for this section needs to be added if you plan to include it.)

Found an error or want to suggest an improvement? Edit this page on GitHub.