Introduction
It is well known that dissection in the vertebrobasilar (VB) systems shows various clinical features such as aneurysmal dilatation and consequential hemorrhagic stroke, or occlusive change and consequential ischemic stroke 1 . These features have close relationships with the clinical outcome. The purpose of the present study was to clarify which factors had an influence on aneurysm formation in patients with dissection in the VB system.
Methods
This study included 20 consecutive patients with a dissection in the VB system. All patients underwent cerebral digital subtraction angiography (DSA), MRI including diffusion weighted image (DWI), MR angiography (MRA), and duplex ultrasonography. We adopted the diagnostic criteria for VB dissection as follows: 1) intimal flap, double lumen, pearl and string sign, or string sign was observed in DSA, MRA, or duplex ultrasonography, or 2) pearl sign or tapered occlusion was observed in DSA and changes in these findings were documented in the serial studies. In DWI, sites and number of infarct were evaluated. In duplex ultrasonography, diameter and flow velocity of the cervical portion of the vertebral artery (VA) were measured. Dominant VA was defined when diameter ratio (ipsilateral/contralateral) was more than 1.4 2 or when contralateral VA had occluded. According to the criteria, dissecting VA was classified into 3 groups: dominant VA, non-differentiated VA, and hypoplastic VA. We investigated the relationships between the clinical and radiological factors and dissecting aneurysm formation.
Results
On DWI, brain infarction was observed in 14 patients; 12 in the posterior inferior cerebellar artery (PICA) territory (cerebellum in 10 and the lateral medulla in 6), 2 in the medial medulla, 1 in the pons, 1 in the cerebellum of the superior cerebellar artery territory, and 1 in the posterior cerebral artery territory. One patient had dissection in the bilateral VAs. The remaining 19 patients had the unilateral dissection. Sites of dissection were VA before branching PICA in 16, VA after branching PICA in 3, and PICA in 2. Dominant VA was dissected in 5 patients, non-differentiated VA in 11, and hypoplastic VA in 4. Dissecting aneurysm was observed in 5 patients (1 in dominant VA and 4 in non-differenciated VA). Aneurysmal change was absent in the patients with dissection in hypoplastic VA. The remaining 15 patients had VA or PICA occlusive lesions due to the dissection. The diameter ratio of the VA tended to be higher in patients with dissecting aneurysm than in patients with occlusive lesions (1.25±0.43 vs 0.88±0.32, p<0.1). Patients with VA dissection in the thicker side tended to more frequently have aneurysm formation than patients with VA dissection in the narrower side (50% vs 20%, p<0.1).
Conclusions
VA diameter or dominancy evaluated using duplex ultrasonography may be an important factor for aneurysm formation in patients with VA dissection.