By Eric Walter, MD, MSc

Pulmonary and Critical Care Medicine, Northwest Permanente and Kaiser Sunnyside Medical Center, Portland, OR

Dr. Walter reports no financial relationships relevant to this field of study.

Spontaneous intracerebral hemorrhage (ICH) is an unfortunately common and devastating event. In 2010, an estimated 5.3 million patients worldwide experienced ICH.1 Mortality at one month is approximately 40%, and survivors often exhibit significant long-term deficits, with independency rates of only 12-39%.2 Between 50-70% of patients presenting with ICH have a history of arterial hypertension, making this the most common risk factor for ICH.3 Furthermore, blood pressure often increases acutely in the setting of ICH.4 Arguments have been made for and against the treatment of hypertension following ICH. Patients exhibiting chronic hypertension may demonstrate less ability to autoregulate cerebral blood flow, necessitating higher systemic blood pressure to maintain cerebral perfusion. There has been concern that aggressive blood pressure control following ICH may decrease blood flow to the ischemic penumbra.3 On the other hand, aggressive blood pressure control may decrease hematoma expansion and cerebral edema.

Recently, two large randomized, controlled trials compared intensive blood pressure control with permissive hypertension in the setting of acute ICH.5,6 The authors of the two trials reached seemingly differing conclusions, leading to confusion on how to best manage patients. Intensive Blood Pressure Reduction in Acute Cerebral Hemorrhage 2 (INTERACT2) was published in 2013 and Antihypertensive Treatment of Acute Cerebral Hemorrhage 2 (ATACH2) in 2016. Both studies were large, international, multicenter trials. The creators of both trials randomized patients with acute spontaneous ICH to a goal systolic blood pressure (SBP) of < 140 mmHg (intensive treatment) or < 180 mmHg (guideline or standard treatment). Results from INTERACT2 suggested that intensive blood pressure control improved functional outcomes at three months.5 However, ATACH2 did not show any benefit, and a post-hoc analysis suggested intensive control increased adverse events.6


Patients were eligible for inclusion in INTERACT2 (n = 2,839) if they presented with spontaneous ICH and a SBP 150-220 mmHg. In ATACH2, patients (n = 1,000) were enrolled if they exhibited at least one SBP reading 180 mmHg. Exclusion criteria in both studies were large ICH, deep coma (Glasgow Coma Scale < 6 for INTERACT2, < 5 for ATACH2), or a structural cerebral cause for the ICH (including trauma). Planned early surgery was an exclusion criterion for INTERACT2. The demographic and clinical characteristics of both populations were well matched across study groups and strikingly similar between studies. The mean age was just over 60 years, and about 60% of patients were male. Both studies recruited from more than 100 sites worldwide, with the majority of sites in Asian countries (~67% of patients in INTERACT2 were from China, and ~50% of patients in ATACH2 were from Asia). Central and East Asia demonstrate the highest incidence of ICH in the world.1 The median National Institutes of Health Stroke Scale score (11) and hematoma volume (11 cm3) were similar in both studies. Twenty-eight percent of patients in INTERACT2 experienced intraventricular extension of hemorrhage. In ATACH2, 25% of patients in the intensive treatment group and 29% of patients in the standard treatment group experienced intraventricular extension.


Researchers differed in their approaches in how to treat early and aggressive blood pressure. In INTERACT2, the goal was to reduce SBP to < 140 mmHg within one hour after randomization and to maintain this level for seven days for patients in the intensive treatment group. Patients in the guideline treatment group received treatment only if SBP was > 180 mmHg. Treatment could be initiated within six hours of symptom onset. In practice, the median time to first treatment was 4-4.5 hours, with nearly half of patients in the intensive treatment group not receiving any antihypertensives for more than four hours after symptom onset. Even with treatment, the goal SBP often was not achieved. Sixty-seven percent of patients in the intensive treatment group did not meet the SBP goal of < 140 mmHg within one hour. Despite these failures, at one hour there was a separation in mean SBP between treatment groups (150 mmHg in the intensive treatment group vs. 164 mmHg in the guideline treatment group).

ATACH2 was designed to offer even earlier and more aggressive blood pressure control. The treatment goal was to reduce SBP to 110-139 mmHg in the intensive treatment group and 140-179 in the standard treatment group. SBP goals were to be maintained for the first 24 hours after randomization, as opposed to INTERACT2, during which BP goals were maintained for seven days. The initial study design randomized patients who presented within three hours of symptom onset. This was extended to 4.5 hours after additional data became available that suggested a treatment effect of blood pressure control out to 4.5 hours and no difference in hematoma expansion between zero to three hours and three to 4.5 hours.7 In practice, the median time to first treatment was three hours in both groups. Thus, in comparison to INTERACT2, in which nearly half the patients underwent randomization after four hours, all patients started treatment within 4.5 hours in ATACH2. The mean minimum SBP during the first two hours after randomization was 128.9 mmHg in the intensive treatment group and 141.1 in the standard treatment group. The treatment failure rate was much lower, with only 12.2% of patients in the intensive treatment group failing to achieve the SBP goal.


The primary outcome in both studies was death or severe disability. Severe disability was defined as a modified Rankin scale score of 3-6 in INTERACT2 and 4-6 in ATACH2. The modified Rankin scale score assesses disability and dependence in daily activities and ranges from 0-6 (0 = no symptoms; 1 = symptoms, but no significant disabilities; 2 = slight disability, able to look after one’s own affairs without assistance; 3 = moderate disability, able to walk without assistance; 4 = moderate to severe disability, unable to walk and attend to own bodily needs without assistance; 5 = severe disability, bedridden and requiring constant nursing care; and 6 = death).

In INTERACT2, 52% of subjects in the intensive treatment group and 55.6% in the guideline treatment group died or were severely disabled at three months. This difference did not quite meet statistical significance (odds ratio [OR], 0.87; 95% confidence interval [CI], 0.75-1.01; P = 0.06). However, in a secondary outcome measure, Rankin scores were ranked rather than dichotomized. Ranking provides increased power to detect a true difference between study arms. Using ranked scores, intensive treatment was associated with a significantly favorable shift in the distribution of scores (pooled OR, 0.87; 95% CI, 0.77-1.00; P = 0.04). This essentially means that more patients who received intensive treatment received favorable modified Rankin scores. In other secondary analyses, patients in the intensive treatment group also reported fewer problems and better quality of life. There were no significant differences in adverse effects between the two groups. The authors had hypothesized that aggressive blood pressure control would lead to improved outcomes by decreasing hematoma expansion. However, the authors did not observe this. In a subgroup of patients in which the hematoma change was measured, there was no significant difference in hematoma expansion (absolute change, 1.4 mL; 95%CI, -0.6 to 3.4; P = 0.18). Overall, INTERACT2 generally was viewed as a positive study, given the near significant difference in primary outcome, the favorable phase shift in ranked modified Rankin scores, and the lack of adverse events associated with intensive blood pressure control in ICH. Since half of patients were not randomized until after four hours and administered the modest decrease in blood pressure achieved, many believed that earlier and more aggressive blood pressure control would produce an even more impressive improvement in outcomes. However, despite achieving these goals in ATACH2, intensive treatment of blood pressure was not associated with better outcomes.

The creators of ATACH2 intended to enroll 1,280 participants but ended the trial early after 1,000 patients were enrolled due to futility. Death or disability was observed in 38.7% of patients in the intensive treatment group and 37.7% of patients in the standard treatment group (relative risk [RR], 1.04; 95% CI, 0.85-1.27; P = 0.72). When modified Rankin scores were ranked, there also was no significant difference between the two study groups. Intensive blood pressure treatment was associated with more serious adverse events within three months (RR, 1.30; 95% CI, 1.00-1.69; P = 0.05). There were no differences in quality of life. Similar to INTERACT2, there was no significant difference in hematoma expansion.


INTERACT2 and ATACH2 were the types of trials ICU clinicians seek: large, well-designed, randomized, controlled trials addressing a relevant clinical question. However, despite these publications, clinicians remain left to weigh the results of a seemingly positive study with a negative study that ended early due to futility. Although not definite, results seem to suggest that moderate blood pressure control following spontaneous ICH, targeting a SBP of 140-150 mmHg, may provide benefit. INTERACT2 suggested that a blood pressure of 150 mmHg (intensive treatment) was better than 164 mmHg (guideline treatment). In ATACH2, the blood pressure in the standard treatment group (mean SBP 141.1 mmHg) was similar to the intensive treatment group of INTERACT2. Pushing blood pressure lower, as in the intensive treatment group in ATACH2 (mean SBP 128.9 mmHg), does not seem to provide added benefit, possibly harm. This idea of moderate blood pressure control following ICH, with a goal SBP of around 140 mmHg8 or 140-180 mmHg,9 has been recommended by others following these publications as well. A moderate rather than an intensive blood pressure goal may have implications beyond just the individual patient. Targeting a more moderate reduction in blood pressure may avoid the need for IV drips, thus allowing for an earlier transfer out of the ICU or avoidance of some ICU admissions altogether. Furthermore, moderate blood pressure control was not associated with an increase in adverse events in INTERACT2.

If blood pressure control leads to better outcomes, the mechanism through which this occurs remains unanswered. It does not seem to be through preventing hematoma expansion. In both studies, hematoma expansion was not affected by blood pressure control. Other factors, such as anticoagulation or location of the hemorrhage, may be more important determinants of hematoma expansion. Finally, it is important to remember that patients who presented in a deep coma, with large hematomas, with intracranial pressure elevation, or with hemorrhage due to trauma, were excluded from both trials. Very few patients were on anticoagulation, and virtually all bleeds were supratentorial. Results from INTERACT2 and ATACH2 may not apply to these patients, and the optimal blood pressure in these situations remains undetermined.


  1. Krishnamurthi RV, Feigin VL, Forouzanfar MH, et al. Global and regional burden of first-ever ischaemic and haemorrhagic stroke during 1990-2010: Fndings from the Global Burden of Disease Study 2010. Lancet Glob Health 2013;1:e259-e281.
  2. van Asch CJ, Luitse MJ, Rinkel GJ, et al. Incidence, case fatality, and functional outcome of intracerebral haemorrhage over time, according to age, sex, and ethnic origin: A systematic review and meta-analysis. Lancet Neurol 2010;9:167-176.
  3. de Oliveira Manoel AL, Goffi A, Zampieri FG, et al. The critical care management of spontaneous intracranial hemorrhage: A contemporary review. Crit Care 2016;20:272.
  4. Qureshi AI, Ezzeddine MA, Nasar A, et al. Prevalence of elevated blood pressure in 563,704 adult patients with stroke presenting to the ED in the United States. Am J Emerg Med 2007;25:32-38.
  5. Anderson CS, Heeley E, Huang Y, et al. Rapid blood-pressure lowering in patients with acute intracerebral hemorrhage. N Engl J Med 2013;368:2355-2365.
  6. Qureshi AI, Palesch YY, Barsan WG, et al. Intensive blood-pressure lowering in patients with acute cerebral hemorrhage. N Engl J Med 2016;375:1033-1043.
  7. Qureshi A, Palesch Y, Investigators AI. Expansion of recruitment time window in Antihypertensive Treatment of Acute Cerebral Hemorrhage (ATACH) II trial. J Vasc Interv Neurol 2012;5:6-9.
  8. Rothwell PM. Stroke research in 2016: When more medicine is better, and when it isn’t. Lancet Neurol 2017;16:2-3.
  9. Starke RM, Peterson EC, Komotar RJ, Connolly ES. A randomized clinical trial of aggressive blood pressure control in patients with acute cerebral hemorrhage. Neurosurgery 2016;79:N17-N8.

Table 1: Comparison of Key Differences and Outcomes Between INTERACT2 and ATACH2


Inclusion Criteria (Systolic Blood Pressure in mmHg)

Treatment Goals (Systolic Blood Pressure in mmHg)

Time to

Mean Systolic Blood Pressure (mmHg) Achieved

Treatment Failure Rate in Intensive Treatment Group

Death or Severe Disability
at 3 Months



Intensive: < 140
Guideline: < 180

Within 6 hours from symptom onset

Intensive: 150
Guideline: 164


Intensive: 52.9%
Guideline: 55.6%
P = 0.06


≥ 180

Intensive: 110-139
Standard: 140-179

Within 4.5 hours of symptom onset

Intensive: 141.1
Standard: 128.9


Intensive: 38.7%
Guideline: 37.7%
P = 0.72