Effect of clipping, craniotomy, or intravascular coiling on cerebral vasospasm and patient outcome after aneurysmal subarachnoid hemorrhage

Abstract

Objective: Although several recent studies have suggested that the incidence of vasospasm after aneurysmal subarachnoid hemorrhage is lower in patients undergoing aneurysmal coiling as compared with clipping, other studies have had conflicting results. We reviewed our experience over 8 years and assessed whether clipping, craniotomy, or coiling affects patient outcomes or the risk for vasospasm.

Methods: We included 515 patients with aneurysmal subarachnoid hemorrhage, identified prospectively from November 2000 to February 2003 (243 patients) and retrospectively from November 1995 to October 2000 (272 patients), by using International Classification of Diseases, 9th Revision, codes for subarachnoid hemorrhage. We classified patients as follows: clipping (413 patients), coiling (79 patients), and craniotomy (436 patients, including all 413 patients who underwent clipping plus 23 who underwent coiling as well as craniotomy for various reasons). We studied four outcome measures: total vasospasm, symptomatic vasospasm, poor outcome (modified Rankin score 3-6), and in-hospital mortality. To assess the risk of total vasospasm and symptomatic vasospasm, we performed multivariate regression analyses adjusting for age, Fisher grade, Hunt and Hess grade, aneurysm location (anterior versus posterior circulation), and aneurysm treatment modality. To assess the risk for poor outcome and in-hospital mortality, we adjusted for all the above variables as well as for total and symptomatic vasospasm.

Results: In the clipping group there was 63% total vasospasm and 28% symptomatic vasospasm; in the coiling group there was 54% total vasospasm and 33% symptomatic vasospasm; and in the craniotomy group there was 64% total vasospasm and 28% symptomatic vasospasm. In the multivariate analysis, age <50 years (P = 0.0099) and Fisher Grade 3 (P < 0.00001) predicted total vasospasm, and Fisher Grade 3 (P < 0.000001) and Hunt and Hess Grade IV or V (P = 0.018) predicted symptomatic vasospasm. Predictors of poor outcome were age >or=50 years (P < 0.0001), Fisher Grade 3 (P = 0.0072), Hunt and Hess Grade IV or V (P < 0.00001), symptomatic vasospasm (P < 0.0001), and coiling (P = 0.0314 versus clipping and P = 0.045 versus craniotomy). Predictors of in-hospital mortality were age >or= 50 years (P = 0.0030), Hunt and Hess Grade IV or V (P = 0.0001), symptomatic vasospasm (P < 0.00001), and coiling (P = 0.008 versus clipping and P = 0.0013 versus craniotomy). There was no significant difference in total vasospasm or symptomatic vasospasm when patients who underwent clipping or craniotomy were compared with patients who underwent coiling. In patients with Hunt and Hess Grade I to III ("good grade"), clipping and craniotomy were associated with better outcome and less in-hospital mortality, but there was no difference in total vasospasm or symptomatic vasospasm versus coiling. In patients with Hunt and Hess Grade IV or V ("poor grade"), there was no difference in any outcome measure among the treatment groups.

Conclusion: In a single-center, retrospective, nonrandomized study, performance of clipping and/or craniotomy had significantly better outcome and lower mortality at discharge than coiling in good-grade patients but had no effect on total vasospasm or symptomatic vasospasm in good- or poor-grade patients.