Multivariate analysis of predictors of cerebral vasospasm occurrence after aneurysmal subarachnoid hemorrhage

Abstract

Background and purpose: The role of type of treatment on cerebral vasospasm occurrence after aneurysmal subarachnoid hemorrhage (SAH) has not been studied. Through multivariate analysis we determined the independent prognostic factors of the occurrence of symptomatic vasospasm following aneurysmal SAH in a study cohort of 244 patients undergoing either surgical or endovascular treatment. The prognostic factors of sequelae after aneurysmal SAH were studied as well.

Methods: Symptomatic vasospasm was defined as the association of deterioration in a patient's neurological condition between 3 and 14 days after SAH with no other explanation and an increase in mean transcranial Doppler velocities of >120 cm/s. The prognostic factors were registered on admission and during the intensive care stay.

Results: Symptomatic vasospasm occurred in 22.2% surgical patients compared with 17.2% endovascular treatment patients (P=0.37). Multivariate analysis revealed that the probability of occurrence of symptomatic vasospasm decreased with age >50 years (relative risk [RR], 0.47 [0.25 to 0.88]) and severe World Federation of Neurological Surgeons (WFNS) grade measured on admission (RR, 0.43 [0.20 to 0.90]) and increased with hyperglycemia occurring during the intensive care stay (RR, 1.94 [1.04 to 3.63]). No difference in risk of symptomatic vasospasm could be identified between surgical and endovascular treatment. Symptomatic vasospasm (OR, 4.73 [CI, 1. 77 to 12.6]) as well as WFNS grade of >2 (OR, 8.95 [3.46 to 23.2]), treatment complications (OR, 8.39 [3.16 to 22.3]), and secondary brain insults were associated with an increased risk of 6-month sequelae.

Conclusions: Age <50 years, good neurological grade, and hyperglycemia were all associated with an increased risk of cerebral vasospasm whereas treatment was not. This provides a basis for future clinical prospective randomized trials comparing both treatments.