Premorbid hypertension modifies predicted optimal blood pressure targets for intracerebral hemorrhage

Background Prior trials found an inconsistent benefit of uniform, intensive blood pressure reduction after intracerebral hemorrhage (ICH). Untreated hypertension impairs cerebral autoregulation, increasing vulnerability to cerebral hypoperfusion. We sought to determine whether premorbid hypertension treatment modifies systolic blood pressure (SBP) targets associated with optimal neurologic outcomes. Methods Using the Antihypertensive Treatment of Acute Cerebral Hemorrhage-2 trial, patients with premorbid hypertension were categorized as treated or untreated based on antihypertensive use within 30 days prior to ICH. The primary outcome was ninety-day poor functional outcome defined as modified Rankin Scale > 2. Multivariable mixed-effects regression using restricted cubic splines modeled the joint interaction between SBP achieved at 24 h and hypertension treatment. A secondary analysis assessed the interaction between hypertension treatment and SBP reduction intensity on early neurologic decline using Glasgow Coma Scale (GCS) measured hourly during the initial 24 h. Results Among 732 participants, 299 had treated hypertension and 433 had untreated hypertension. Poor ninety-day outcome occurred in a total of 416 (57%) patients. The primary analysis demonstrated the association between achieved SBP and ninety-day outcome was modified by hypertension treatment (joint Wald interaction test χ²=20.45, p < 0.001). Spline models observed a J-shaped relationship among patients with untreated hypertension and a predicted optimal target of 130–140 mmHg. Patients with treated hypertension demonstrated continued reduction in the probability of poor outcome with lower SBP and a predicted optimal target of ≤ 120 mmHg. In the secondary analysis, hypertension treatment modified the relationship between SBP reduction and change in GCS, such that extreme SBP reduction was associated with GCS decline in those with untreated hypertension (joint Wald interaction test χ²=9.73, p = 0.008). Conclusion Premorbid hypertension treatment modifies the predicted optimal SBP targets after ICH. Personalized antihypertensive strategies that account for chronic hypertension warrant further study. Trial registration www.clinicaltrials.gov (Unique Identifier: NCT01176565). First submitted August 2010.