Effect of choroidal collateral vessels on de novo hemorrhage in moyamoya disease: analysis of nonhemorrhagic hemispheres in the Japan Adult Moyamoya Trial

Objective: Following hemorrhagic stroke in moyamoya disease, de novo intracranial hemorrhage can occur in the previously unaffected nonhemorrhagic hemisphere. In the present analysis the authors intended to determine whether the presence in the nonhemorrhagic hemisphere of choroidal collateral vessels, which have been the focus of attention as a source of bleeding, affects the risk of de novo hemorrhage.

Methods: The subject of focus of the present cohort study was the nonhemorrhagic hemispheres of adult patients with hemorrhagic moyamoya disease enrolled in the Japan Adult Moyamoya Trial and allocated to the nonsurgical arm. The variable of interest was the presence of choroidal collaterals (also termed choroidal anastomoses), identified with baseline angiography and represented by a connection (anastomosis) between the anterior or posterior choroidal arteries and the medullary arteries. The outcome measure was de novo hemorrhage during the 5-year follow-up period, assessed in all nonhemorrhagic hemispheres. The incidence of de novo hemorrhage in the collateral-positive and -negative groups was compared.

Results: Choroidal collaterals were present in 15 of 36 (41.7%) nonhemorrhagic hemispheres analyzed. The overall annual risk of de novo hemorrhage was 2.0%. Three de novo hemorrhages occurred in the collateral-positive group, whereas no hemorrhage occurred in the collateral-negative group. The annual risk of de novo hemorrhage was significantly higher in the collateral-positive group than in the collateral-negative group (5.8% per year vs 0% per year; p = 0.017). All hemorrhage sites corresponded to the distribution of choroidal collaterals.

Conclusions: The present preliminary results suggest that the presence of choroidal collaterals affects the risk of de novo hemorrhage in the nonhemorrhagic hemisphere, subject to verification in larger studies. Further studies are needed to determine the optimal treatment strategy for nonhemorrhagic hemispheres and asymptomatic patients.