Abstract 164: Comparison Between Transcranial Doppler and Functional MRI With Breath Holding Test for Cerebral Vasoreactivity Assessment in Patients with Carotid Stenosis

Abstract

Background: Impaired cerebral vasoreactivity (CVR) detected by Transcranial Doppler (TCD) is predictive of stroke in patients with carotid stenosis. Functional MRI (fMRI) can evaluate CVR either. Few studies have compared both methods for CVR assessment already. However, none of them used breath holding test (BHT) as the vasodilatory stimulus. Objectives: To assess CVR through BHT using TCD and fMRI in patients with carotid stenosis and controls in order to verify whether fMRI can detect any difference on BOLD-signal intensities between these groups. Additionally, we aimed to verify if fMRI can discriminate correctly any subject with impaired CVR as compared to TCD. Methods: Fifteen patients with carotid stenosis defined as stenosis ≥50% by angiography or ≥70% by ultrasound or ≥70% by CTA or MRA (11men;52-90yo;mean:67.7yo) and 7 matched-controls (5men:51-85yo;mean:64.6yo) performed BHT during TCD. Impaired CVR was diagnosed if mean percentage increase of MCA velocities ≤31% on 3-consecutive intervals of 30s of apnea intercalated by 4-minute-interval of normal breathing. During fMRI, BHT was performed using a traffic lights-based code for instruction on when it was time to stop breathing and was composed by 3-consecutive intervals of 24s of apnea intercalated by 4-minute-intervals of normal breathing. BOLD-signal intensity was measured in the lentiform nucleus ipsilateral to the carotid stenosis. In controls, the mean signal intensity of both lentiform nuclei was considered. Results: Patients and controls showed significantly different BOLD-signal intensities (p:0,0053). Based on CVR defined by TCD, there was an agreement between TCD and fMRI of 6/7 (85,7%) in controls, 11/15 (73.3%) in patients, and 17/22 (77,2%) overall. Conclusion: fMRI showed different BOLD signal intensities in patients and controls using BHT as the vasodilatory stimulus for CVR assessment. Additionally, there was a good agreement between fMRI and TCD on detecting patients with impaired CVR. Further studies are needed to evaluate whether impaired CVR on fMRI with BHT is predictive of stroke in patients with carotid stenosis, but these results suggest that fMRI may be a potentially useful tool for stroke risk assessment in clinical practice.