Carotid artery duplex velocity criteria might be equivocal after left ventricular assist device implantation

Background: Although conventional angiography remains the reference standard for the grading of carotid stenosis, carotid duplex ultrasound (CDUS) is the most commonly used modality for determining the degree of carotid stenosis. The validity of CDUS findings for patients after left ventricular assist device (LVAD) implantation is questionable, because the velocities are often altered secondary to the continuous flow nature of the devices. Methods: A retrospective review was performed of all patients who had undergone LVAD implantation from January 2007 to December 2019. All patients who had undergone CDUS before and after LVAD implantation were included. Patients receiving extracorporeal membrane oxygenation, those with unusable carotid imaging studies, and those with internal carotid artery (ICA) occlusion were excluded. The peak systolic velocity (PSV) and end-diastolic velocity (EDV) in the ICA and common carotid artery (CCA) and the ICA/CCA ratios were compared before and after LVAD implantation. Results: A total of 36 patients (mean age 59 years; 30 men; 6 women) had undergone CDUS both before and after LVAD implantation (mean, 647 days between imaging studies). A total of 61 ICAs had met the criteria for inclusion. Before LVAD, 7 carotid arteries (13%) had had >50% carotid stenosis and 53 (87%) had had 0% to 50% stenosis. The mean changes in the velocities after LVAD were as follows. The ICA PSV had decreased by 6.12 ± 4.34 cm/s, and the ICA EDV had increased by 13.44 ± 4.23 cm/s. The CCA PSV had decreased by 17.22 ± 4.95 cm/s, and the CCA EDV had increased by 10.83 ± 2.59 cm/s. The mean ICA/CCA ratio had increased by 0.18 ± 0.05. All the mean changes in velocity were significant (P <.01), except for the ICA PSV (P =.167). Among four patients with known stenosis of 60% to 69%, the degree of increase in the ICA and CCA EDVs (75.8 and 13.3 cm/s, respectively) was significantly greater than that for patients with <50% or no stenosis. Carotid artery laterality did not significantly affect the differences in mean velocity. Centrifugal LVADs resulted in a significantly larger increase in the ICA EDV compared with axial LVADs (26.0 vs 6.3 cm/s; P <.01). Conclusions: LVADs were associated with significant changes in CCA PSV, ICA and CCA EDV, and ICA/CCA ratios. However, the magnitude of these changes in patients with <50% stenosis was minimal and might not be clinically significant. The LVAD type might only have an effect on EDV measurements in the CCA, and the left and right carotid arteries did not appear to have different degrees of change in velocity. The currently used criteria for determining carotid stenosis might result in an under- or overestimation of carotid stenosis in patients with an LVAD.