Discussion

Dr Ali F. AbuRahma (Charleston, WV). Carotid artery stenting has become an accepted treatment modality for carotid stenosis, particularly in high-risk patients. However, there is an ongoing debate regarding which duplex ultrasound criteria to use to determine the rate of in-stent restenosis. This study is by Dr Armstrong and his group, and they are well respected and nationally known for their advocacy for duplex scan surveillance after vascular interventions. This study reports on duplex scan surveillance after carotid artery stenting and the rationale for the definition of in-stent restenosis.

In this study, Dr Armstrong and his group analyzed their experience using a previously validated surveillance protocol after carotid endarterectomy and applied these criteria to carotid artery stenting surveillance. In a similar study that we presented at the Eastern Vascular Society meeting in Washington, DC in September 2006, which is presently in press, when we applied the old duplex ultrasound velocity criteria for nonstented carotid arteries, around 50% of our patients were interpreted to have ≥30% restenosis, as defined by a peak systolic velocity of >120 cm/s, however, when we applied new duplex ultrasound criteria for stented arteries, a peak systolic velocity of >155 cm/s was consistent with ≥30% restenosis in only 33% of patients at a mean follow-up of 2 years.

With this in mind, I have the following questions and/or comments for Dr. Armstrong.

First, it is noted in your study that you used criteria for carotid artery stent surveillance that was somewhat similar with some modification to the criteria your group published in the Journal of Vascular Surgery in 1999 for carotid endarterectomy surveillance. Specifically, to define a <50% stenosis category, the peak systolic velocity was increased from 125 cm/s to <150 cm/s, with an ICA/CCA ratio of <2; and for 50% to <75% stenosis, the peak systolic velocity was raised from 125 cm/s to >150 cm/sec, with a ratio of 2. For stenosis >75%, the criteria were left the same (eg, a peak systolic velocity of >300 cm/s with a ratio of >4 and an end diastolic velocity >125 cm/s. Did you validate these criteria on patients with carotid artery stenting? In other words, did you obtain any other modality, specifically, carotid angiography or CTA, to verify the degree of in-stent restenosis? If so, did you conduct any ROC curves to detect the sensitivity, specificity, positive and negative predictive values for specific velocities that were consistent with the various classifications you propose in your study?

Second, did you obtain immediate duplex ultrasounds after completion of the carotid stenting to compare normal angiography after carotid stenting to the peak systolic velocities and/or the end diastolic velocities?

Third, in your presentation, only six patients were found to have ≥75% in-stent restenosis that required intervention. In view of the earlier discussion, is it possible that there are other patients who have similar stenoses that were missed because of your present criteria? It is my understanding that these are the only six patients who had their stenoses confirmed by arteriography.

I enjoyed your presentation, and I am looking forward to seeing additional work on this very important clinical subject from your well-respected institution.

Dr Paul A. Armstrong: Certainly one criticism of this review may be the lack of a structured validation method (ie, angiography), but the validation for carotid duplex criteria does have an established track record, including recent work done by Dr Hobson and others. We believe these data to be valid based on the recent information provided by the core duplex reading center from the CREST Trial, which documented strongly favorable positive predictive values for defining moderate and high-grade lesions using similar duplex criteria.

Our current surveillance protocol includes immediate duplex scanning the day of the procedure. In addition we are now using IVUS as part of stent implantation to provide in formation on vessel diameters, plaque morphology, and stent apposition.

In answer to whether or not we are missing high-grade lesions, we would say no. As you are aware, some investigators are supportive of accepting higher stent velocities based on the stent characteristics, and in general, we know that there is a tendency of duplex to overestimate stenosis. Therefore the chance of missing high-grade asymptomatic lesions when applying similar criteria in an accredited vascular laboratory is not likely.