Predictors of cerebral infarction in patients with aneurysmal subarachnoid hemorrhage

Abstract

Objective: Cerebral infarction would be expected to be associated with poor outcome after aneurysmal subarachnoid hemorrhage (SAH), although there are few data on which to base this assumption. The goals of this study were to determine the impact of cerebral infarction on outcome and to examine predictors of infarction in these patients.

Methods: Univariate and multivariable statistical methods were used to examine the impact of cerebral infarction on the Glasgow Outcome Scale score 3 months after SAH among 3567 patients entered into four prospective, randomized, double-blind, placebo-controlled trials of tirilazad conducted in neurosurgical centers around the world between 1991 and 1997. Patient demographics, clinical variables, radiographic characteristics, and treatment variables associated with cerebral infarction were also determined by the same methods.

Results: Seven hundred and seven (26%) out of 2741 patients with complete data had cerebral infarction on computed tomographic scans 6 weeks after SAH. Multivariable logistic regression showed that cerebral infarction increased the odds of unfavorable outcome by a factor of 5.4 (adjusted odds ratio, 5.4; 95% confidence interval, 4.2-6.8; P < 0.0001), which was a higher odds ratio than all other factors associated with outcome. The proportion of explained variance in outcome was also highest for cerebral infarction and accounted for 39% of the explained variance. Multivariable analysis found that cerebral infarction was significantly associated with increasing patient age, worse neurological grade on admission, history of hypertension or diabetes mellitus, larger aneurysm, use of prophylactically or therapeutically induced hypertension, temperature more than 38 degrees C 8 days after SAH, and symptomatic vasospasm.

Conclusion: Cerebral infarction was strongly associated with poor outcome after aneurysmal SAH. The most important potentially treatable factor associated with infarction was symptomatic vasospasm.