MR angiography after coiling of intracranial aneurysms

Introduction Endovascular occlusion with detachable coils has become an alternative treatment to neurosurgical clipping of intracranial aneurysms over the last two decades. Its minimal invasiveness is the most important advantage of this treatment compared to clipping. The disadvantage of occlusion with coils is an approximately 20% risk of reopening of the aneurysm as a result of coil impaction, dissolution of thrombus, or growth of the aneurysm and 10% of coiled patients need additional treatment. As a consequence, patients need to be followed up after coiling to detect and treat reopening and thereby prevent rupture of a reopened aneurysm. The standard technique to follow up coiled patients is intra-arterial digital subtraction angiography (IA-DSA) but this is an invasive and irradiating procedure for which a short hospitalization is required. Magnetic resonance angiography (MRA) could be a good alternative because this imaging technique is non-invasive, non-irradiating, cheaper than IA-DSA, and does not require hospitalization. The aim of this thesis is to evaluate whether the diagnostic performance of MRA in coiled patients is sufficient to replace IA-DSA as the first-choice follow-up modality for these patients. Methods We first calculated test characteristics of MRA with IA-DSA as the reference test, then we evaluated the added value of contrast-enhanced MRA to unenhanced MRA, and finally we assessed the cost-effectiveness of follow-up with MRA compared to follow-up with IA-DSA. Besides, we studied coil-artifact production on MRA for which we varied field strength, scan parameters, and coil materials. We also evaluated whether blood-vessel anatomy influences the risk of aneurysm reopening. Finally, the long-term bleeding risk from adequately coiled aneurysms at 6-months follow-up has been assessed. Results We found good test characteristics of MRA with a high negative predictive value and similar treatment decisions on MRA and IA-DSA. Test characteristics were similar for MRA at 1.5 Tesla and 3.0 Tesla and contrast-enhanced MRA did not have additional value to unenhanced MRA. MRA appeared to be cost-effective compared to IA-DSA. We furthermore found that intra-voxel dephasing, which is influenced by the echo time, is the dominant mechanism in coil-artifact production on unenhanced MRA. At 3.0 Tesla, artifact reduction through the possible echo-time shortening even compensated for artifact enlargement through increased field disturbances at higher field strength. Then, aneurysms at sharper bifurcation angles, thus more flow deviation at the aneurysm base, seem to carry a higher risk for reopening but this finding needs to be confirmed in a larger cohort. And finally, the long-term bleeding risk for patients with an adequately coiled aneurysm at 6-months follow-up is very small. Conclusions Patients with coiled intracranial aneurysms should be followed up by MRA instead of IA-DSA. For MRA, generally no contrast enhancement is required and it can be performed at either 1.5 Tesla or 3.0 Tesla as long as the echo time is as short as possible while keeping acceptable SNR. Thus far, we cannot select patients who are at risk for reopening after coiling so we need to follow up all coiled patients. Future studies could focus on risk profiles of patients for tailored follow-up schedules