Efficacy of patient selection strategies for carotid endarterectomy by contrast-enhanced MRA on a 1 T machine and duplex ultrasound in a regional hospital

Aim

To investigate whether contrast-enhanced magnetic resonance angiography (CE-MRA) and duplex ultrasound (DUS) could replace digital subtraction angiography (DSA) for diagnosing internal carotid artery (ICA) stenosis in regional centres with less specialized technicians and equipment, such as a 1 Tesla MRI machine.

Materials and methods

Sixty-six consecutive, symptomatic patients with ICA stenosis, as evidenced using DSA, were included. In the first 34 patients DUS was validated and cut-off criteria were established. Data were analysed by receiver operating characteristic curve and logistic regression. Two observers analysed the DUS and CE-MRA results of 32 patients. Stenoses were categorized in accordance with North American Symptomatic Endarterectomy Trial (NASCET) measurement criteria.

Results

Peak systolic velocity (PSV) in the ICA was defined as a better parameter for defining stenosis than end diastolic velocity and the PSV ICA:common carotid artery ratio. The optimal PSV threshold was 230cm/s. Four ICAs were not interpretable on DUS, and one on CE-MRA. Two patients did not undergo CE-MRA. The sensitivities and specificities were calculated: for DUS these were 100% and 68% respectively; for observer 1 on CE-MRA these were 93% and 89%, respectively; for observer 2 these were 92% and 87%, respectively. The sensitivity and specificity for combined DUS/CE-MRA were 100% and 85%, respectively. Seventy-eight percent of CE-MRA and DUS correlated. The weighted Kappa for CE-MRA and DSA were 0.8 and 0.9, respectively.

Conclusion

DUS and CE-MRA are effective non-invasive methods for selecting patients with ICA stenosis for carotid endarterectomy in non-specialized centres using a 1T machine. The present results suggest that no referrals to more specialized centres for non-invasive diagnostic work-up for carotid artery stenoses will be necessary.