Defining Duplex Ultrasound Criteria for In-Stent Restenosis of the Carotid Artery Using Computed Tomographic Angiography

Abstract

Introduction Duplex ultrasound (DUS) velocity measurement is the preferred method for evaluating carotid artery stenosis. However, velocity criteria based upon native carotid arteries may not apply to internal carotid artery stents. Previously, catheter-based angiography was used to determine DUS velocity criteria for in-stent restenosis (ISR), but conventional angiography is invasive and can be limited. This study sought to define duplex ultrasound velocity criteria for predicting internal carotid artery in-stent restenosis by correlating in-stent velocities with computed tomographic angiography (CTA) measurements of percent stenosis. Methods A retrospective chart review was conducted on all patients who underwent internal carotid artery (ICA) stenting within our health system between January 2013 and February 2020. Thirty-eight surveillance DUS studies from 32 patients were found to have CTA performed within 30 days. Centerline reconstructions of internal carotid artery stents were created using Aquarius iNtuition software (TeraRecon, Durham, NC, USA). Two independent observers measured percent stenosis by three built-in methods. Stenotic areas were matched to DUS-measured peak systolic velocities (PSV) and end-diastolic velocities (EDV). Internal carotid artery PSV (stent) to common carotid artery (CCA) PSV ratios (ICA/CCA) were calculated, and receiver operating characteristic (ROC) curves were generated. The optimal DUS velocity criteria in the stented ICA were determined by maximizing Youden's index. Results Mean vessel diameter measurement of percent stenosis resulted in the most accurate model for all DUS velocity parameters (PSV, EDV, and ICA/CCA ratio) and was used for threshold determinations (area under the receiver operating characteristics (AUROC): 0.99, 0.96, and 0.96, respectively). A PSV cutoff of 240 cm/s for ≥60% ISR resulted in the highest Youden's index (97%) with 100% sensitivity and 97% specificity. Secondary DUS parameters included an EDV ≥50 cm/s (Youden's index 84%) and an ICA/CCA ratio ≥ 2.2 (Youden's index 91%). Conclusions Velocity criteria to predict internal carotid artery ISR is needed to inform decisions for possible reintervention. Using CTA, we found that a PSV ≥240 cm/s on carotid DUS can predict ≥60% ISR with high sensitivity and specificity. This value can be used as an alternative to current velocity criteria based on native carotid arteries. However, the optimal thresholds for EDV and ICA/CCA ratio were similar to native carotid arteries.

Keywords: carotid artery stenting; carotid in-stent restenosis; ct angiogram; duplex ultrasound; restenosis.

Figure 1. Centerline reconstruction of stented internal carotid artery

Centerline measurements of the ICA from CTA in (a) sagittal, and (b) coronal projection with the most stenotic portion noted at the red hash compared with normal artery diameter noted by the blue hash. Axial illustration of the stent with flow channels selected at the (c) normal distal portion of the ICA, and (d) the most stenotic in-stent portion.

Figure 2. Receiver operating characteristic curves for three measures of percent stenosis on centerline CTA

Receiver operating characteristic curves showing peak systolic velocity (PSV, solid line), internal carotid artery and common carotid artery (ICA/CCA) ratio (dashed line), and end-diastolic velocity (EDV, dotted line) area under the curve (AUC) for three different techniques to measure in-stent restenosis. (a) Mean vessel diameter: PSV AUC, 0.99; ICA/CCA ratio AUC, 0.96; EDV AUC, 0.96 (b) Minimum vs maximum diameter: PSV AUC, 0.98; ICA/CCA ratio AUC, 0.90; EDV AUC, 0.92 (c) Percent area reduction: PSV AUC, 0.72; ICA/CCA ratio AUC, 0.77; EDV AUC, 0.65