Endovascular treatment or neurosurgical clipping of ruptured intracranial aneurysms: effect on angiographic vasospasm, delayed ischemic neurological deficit, cerebral infarction, and clinical outcome.

BACKGROUND AND PURPOSE The effects of aneurysm treatment modality (clipping or coiling) on the incidence of cerebral vasospasm and infarction after subarachnoid hemorrhage have not been clearly defined. We hypothesized that there may be a difference in angiographic and clinical vasospasm, cerebral infarction, and clinical outcome between patients undergoing clipping compared to coiling. METHODS A retrospective, exploratory analysis of 413 patients randomized into the CONSCIOUS-1 trial was conducted. Patients underwent baseline and follow-up catheter angiography and computed tomography, as well as clinical assessments. Radiology end points were adjudicated by central blinded review, and angiographic vasospasm was quantified by measurements of arterial diameters on catheter angiography. The effect of method of aneurysm treatment (clipping [n=199] or coiling [n=214]) on angiographic vasospasm, delayed ischemic neurological deficit, cerebral infarction, and clinical outcome was analyzed using univariate and multivariate logistic regression. Propensity matching was used to adjust for differences in baseline risk factors between clipped and coiled patients. RESULTS In all patients and the propensity-matched subset, aneurysm coiling was associated with a significantly reduced risk of angiographic vasospasm and delayed ischemic neurological deficit compared to clipping. Cerebral infarction and clinical outcome were not associated with clipping or coiling. CONCLUSIONS In this exploratory analysis, aneurysm coiling was associated with less angiographic vasospasm and delayed ischemic neurological deficit than surgical clipping, whereas no effect on cerebral infarction or clinical outcome was observed. Whether this is attributable to differences in baseline risk factors between clipped and coiled patients or a true difference cannot be proven here.