Magnetic resonance angiography and magnetic resonance tomography in dissection of the vertebral artery

Vertebral artery dissection (VAD) is an important cause of posterior circulation stroke in young adults. Initial symptoms are often non-specific and diagnostic arteriography is not performed until neurological deficits are obvious. Since magnetic resonance tomography (MRT) is superior in the diagnosis of vertebrobasilar ischemia, we retrospectively analyzed the role of MRT and MR angiography (MRA) in the detection of dissections of the vertebral artery. Between 1989 and 1995 we identified 24 patients with a vertebral artery dissection and 1 patient with a basilar artery dissection (8 females and 17 males, 23-60 years of age, mean 41.2 years). The diagnosis of VAD (14 left VAD, 9 right VAD, 1 bilateral VAD, 1 basilar artery dissection) was established by specific arteriographical findings (DSA) or clinical and neuroradiological course. All patients underwent a combined MRT/MRA examination protocol at 1.5T that consisted of spin-echo imaging and time of flight MRA of the intra- and extracranial arteries using 2D Flash and 3D Fisp sequences. The MRT/MRA findings were correlated to DSA and ultrasound results. During the acute and subacute stage, MRT/MRA revealed abnormal findings in 21 of 22 dissected vessels (95.5%). There was one false-negative MRT/MRA in a patient with a V1 dissection (intimal flap without peripheral flow disturbances). In 7/22 VAD the MRT/MRA findings were rated specific (double lumen n = 1, mural hematoma n = 4, pseudoaneurysm n = 2). DAS was sensitive in 100% and ultrasound in 77.3%. Specific results were obtained by DSA in 8/ 22 VAD (36.4%) and in 7/22 VAD (30.4%) by MRT/MRA. When MRT/MRA and DSA results were combined, the specific findings increased to 43.5%. Follow-up examinations revealed recanalization in 52% of initially stenosed or occluded vertebral arteries; four patients developed a pseudoaneurysm, and two of them underwent ligation of the VAD. With this retrospective approach, we were able to show a high sensitivity of MRT/ MRA for the presence of disturbed flow in the dissected vertebral artery. The MRA projections tended to overestimate stenosis and were inferior to DSA in the appreciation of irregularities of the vessel wall. Identification of high-grade stenosis, especially in the presence of distal occlusion, was improved on the MRA source images. During the acute and subacute stage, the diagnosis of luminal thrombus can be difficult, because signal ambiguities exist between hemoglobin breakdown products and flow effects and adjacent fat tissues. The differentiation between luminal thrombus and mural hematoma requires interpretation of MRA source images, together with flow compensated spin-echo images. Additional fat suppressed images and flow presaturation may be required at the appropriate levels. The identification of mural hematoma is important, because this finding is considered specific and cannot be obtained with DSA. There is a complementary role of MRT/MRA and DSA for an improved overall specificity for vertebral artery dissection. A negative MRT/MRA result in a patient with appropriate symptoms, however, cannot exclude a dissection and should prompt DSA. On the other hand, a suggestive MRT/MRA result in the appropriate clinical context can replace DSA. The advantage of MRT/MRA is that the method offers a simultaneous diagnosis of posterior fossa ischemia and vertebral artery abnormalities. Therefore, MRT/MRA should be recommended in patients with suspected VAD and especially in those who have no definite neurological deficit. These patients will benefit greatly from early diagnosis and therapy. The fact that all our patients were diagnosed after neurological symptoms and that 64% of them have residual deficits gives an ethical and economical rationale for advocating early MRT/MRA in these patients.