Clinical course and treatment of vertebrobasilar dolichoectasia: a systematic review of the literature

Objectives: Vertebrobasilar dolichoectasia is a rare condition with unsettled clinical course and treatment. We performed a systematic review of the literature on clinical course and treatment of vertebrobasilar dolichoectasia.Methods: We systematically searched PUBMED, EMBASE, and the Cochrane Library and hand searched citations for studies on vertebrobasilar dolichoectasia with an average follow-up duration of at least 1 year and fulfilling predefined inclusion criteria. We pooled data to calculate proportions with corresponding 95% confidence intervals (CIs) for 5 year incidences of: (1) transient ischemic attack; (2) ischemic stroke; (3) subarachnoid hemorrhage; (4) other intracranial hemorrhage; (5) (progressive) brainstem compression; (6) clinical deterioration from (progressive) hydrocephalus; and (7) case fatality. We used the Cochrane collaboration five-item tool to assess risk of bias in treatment studies.Results: From 958 citations identified, we included 12 manuscripts describing nine cohorts with 375 patients. Follow-up ranged 1·5–11·7 years. Estimated 5 year complication risks were for ischemic stroke 17·6% (95%CI: 12·4–22·8), brainstem compression 10·3% (95%CI: 6·4–14·1), transient ischemic attack 10·1% (95%CI: 6·3–14·0), hemorrhagic stroke 4·7% (95%CI: 2·4–8·3), hydrocephalus 3·3% (95%CI: 1·4–6·4), and subarachnoid hemorrhage 2·6% (95%CI: 1·1–5·1). The estimated five year case fatality was 36·2% (95%CI: 30·6–41·8). The 5 year prognosis is more favorable in patients who are asymptomatic at time of diagnosis. No randomized intervention trials were identified. In the included studies, treatment specifications were often not available for individual patients. None of the studies fulfilled any of the criteria to reduce risk of bias.Conclusion: Patients with vertebrobasilar dolichoectasia are at high risk of ischemic stroke, brainstem compression, and death. Hemorrhagic complications are less common. Risk of bias in the included studies was too high to make any recommendation regarding treatment.