Comparison of 3D TOF-MRA and 3D CE-MRA at 3T for imaging of intracranial aneurysms.

Background: Risks and complications associated with the treatment of intracranial aneurysms have decreased with the growing use of less in- vasive endovascular techniques. Nevertheless, some risk remains and must be balanced with the risk of aneurysm rupture. Both the risk of rupture and selection of treatment are largely based on factors related to specific characteristics of the particular aneurysm obtained from imaging examinations. Because magnetic resonance angiography (MRA) is noninvasive, lacks use of ionizing radiation, and is less costly, it is largely replacing conventional angiography for the diagnosis and follow-up of intracranial aneurysms. Studies have shown that nonenhanced 3D time-of-flight- (TOF-) MRA is satisfactory for follow-up of aneurysms, and that TOF-MRA performed at 3T provides improved depiction of aneu- rysms compared to 1.5T. Whether the use of MRA with gadolinium contrast provides any added benefit is less clear, with some studies showing that 3D contrast-enhanced MRA (CE-MRA) is better for visualizing aneurysmal morphology, while other studies have shown no such benefit, particularly at 3T. Here we compare 3D TOF-MRA and 3D CE-MRA, both at 3T, for their ability to provide detailed characterization of intracranial aneurysms. Methods: Twenty-nine patients (12 male; 17 female) with known or suspected intracranial aneurysms underwent both unenhanced and contrast-enhanced MRA using an In- tera 3 Tesla magnet (Philips Medical Systems, The Netherlands) with an 8-channel SENSE head coil in a single session. The unenhanced acquisition was carried out with a 3D TOF-MRA sequence (3D FFE, TR 2.3, TE 3.5, FOV 250, matrix 1024x1024, SENSE factor 2.5, 180 slice, 4 chunk, voxel size 0.5x0.5x1 mm), focused on the Circle of Willis and carried out in the axial plane, with fat suppression. Con- trast-enhanced images were acquired with a 3D ultrafast FE sequence (TR 5.9, TE 1.8, FOV 220, matrix 304, re- construction 512, SENSE factor 3, 80 slice, voxel size 0.72x0.72x0.80 mm) in the axial plane, focused on the an- eurysm, using CENTRA for k-space elliptical mapping. A total dose of 0.1 mmol/kg gadobenate dimeglumine (MultiHance; Bracco SpA, Italy) was injected at 2 mL/sec, followed by a 25- mL saline flush injected at the same rate. Both maximum intensity projec- tion (MIP) and volume rendering (VR) techniques were used for image analy- sis. The following features were com- pared for the 2 imaging techniques: the location and length of the aneu- rysm, the sac shape, the presence and measurement of the aneurysm neck, the detection of arterial branches orig- inating from the sac or the neck of the aneurysm, and any other associated circulatory abnormalities. Results: A total of 41 aneurysms were identified with both TOF-MRA and CE-MRA techniques: 20 in the in- ternal carotid artery, 10 vertebrobasi- lar, 7 in the anterior communicating artery, and 4 in the middle cerebral artery. There were no differences be- tween the techniques in terms of detec- tionoftheaneurysms,location,oridentification of the parent artery. There were no differences between TOF-MRA and CE-MRA in terms of assessment of the aneurysm dimensions: 23 were <7mm, 11 were 7-12 mm, 3 were 13-24mm, and 4 were >24 mm. TOF-MRA data detected 12 aneurysms with an irregular sac, whereas CE-MRA acqui- sitions detected 19 irregularly-shaped aneurysms. Both imaging techniques identified 5 fusiform aneurysms with no neck. Of the remaining 36 aneu- rysms, TOF-MRA reconstructions did not enable detection of the neck and therefore, did not permit measurement of the sac/neck ratio, in 10 aneurysms. With CE-MRA, the neck was detected and measured in all 36 nonfusiform aneurysms, permitting calculation of the sac/neck ratio. CE-MRA detected 15 aneurysms with branches origi- nating from the sac and/or the neck, whereas the TOF-MRA sequence ena- bled recognition of branches in only 12 of the 15 aneurysms. Finally, of the 41 aneurysms investigated, abnormali- ties of the Circle of Willis were found in 4 aneurysms with both imaging techniques. In 3 cases, A1 aplasia was found, and in the fourth case, TOF- MRA and CE-MRA both documented aplasia of the left posterior communi- cating artery. The diameter of all 4 of these aneurysms was <13 mm: in 2 of them it was less than 7 mm and in the other 2 it was between 7 and 13 mm. Conclusion: 3D CE-MRA and 3D TOF-MRA at 3T are both excellent imaging techniques for determining the presence, location, and length of intracranial aneurysms. However, CE-MRA is superior to TOF-MRA for detailed visualization of certain aneu- rysmal features that impact treatment selection, including the sac shape, neck measurements, and the presence of arterial branching.