Ultrasound surveillance following carotid endarterectomy “costly and unlikely to improve outcomes”

Shira Strauss

According to a recent study, ultrasound surveillance following carotid endarterectomy is “costly and unlikely to improve outcomes or affect management,” even if over 80% ipsilateral restenosis is detected. The investigators suggest that “further research should determine the utility of routine contralateral carotid duplex ultrasound surveillance in this population, who may be at higher risk for progression to severe contralateral carotid disease”.

Shira Strauss and colleagues at The Ottawa Hospital (Ottawa, Canada) base this conclusion on the finding that severe carotid restenosis following carotid endarterectomy is “rare and not associated with worse outcomes,” and note that, despite historical recommendation, The Ottawa Hospital’s management of severe restenosis has been conservative based on its “benign natural history.”

Strauss recently presented these results at the Society for Vascular Surgery (SVS) 2019 Vascular Annual Meeting in National Harbor, Maryland (VAM; 12–15 June).

The investigators collected data from the period 2003–2018, during which 1,190 carotid endarterectomies were performed in 1,129 patients at The Ottawa Hospital. The majority of these cases (78%) were for symptomatic carotids ranging from 70–99% stenosis. A total of 769 surgeries in 729 patients had duplex ultrasound surveillance data available for a mean of 3.52 years, which detected over 80% ipsilateral restenosis in 31 patients (4.03%).

There were 32 postoperative strokes, report Strauss and colleagues, of which 25 (73.5%) occurred within 30 days of surgery—prior to the initial surveillance duplex ultrasound—and 16 occurred within the first 24 hours. They note that no stroke beyond 30 days was attributed to ipsilateral carotid restenosis.

As per The Ottawa Hospital post-carotid endarterectomy protocol, the estimated cost of duplex ultrasound surveillance in this cohort was US$359,925.20 in total, or US$11,610.49 per restenosis. Duplex ultrasound surveillance did not lead to a single ipsilateral redo surgery during this period.

Duplex ultrasound was completed at The Ottawa Hospital’s accredited Vascular Diagnostic Center as per protocol, which currently involves performing surveillance at three and 12 months postoperatively and yearly thereafter.

Strauss elaborated on follow-up protocol at The Ottawa Hospital: “In the earlier years of the study period the protocol was a lot more extensive—at three months, six months, nine months, 12 months and then yearly thereafter, which eventually decreased to ultrasound at three months and yearly thereafter post-CEA. We are now in a transition period where some of the staff are starting to slow that down even further.”

Follow-up duplex ultrasound and clinical data were obtained from Vascubase and electronic medical records (vOacis). Primary outcomes included time to over 80% restenosis, post-carotid endarterectomy stroke rate, and number of surgical reinterventions, and statistical analyses involved survival analysis, χ2 and Fisher’s exact tests, and cost effectiveness analysis.

Moderator Brajesh Lal (University of Maryland School of Medicine, Baltimore, USA) applauded the amount of “due diligence” carried out in collecting follow-up data in this “very well done study”, however also noted that about a third of the study population were lost, asking why these people did not return. In response, Strauss remarked: “It is a flaw within any retrospective study that we cannot account for those who are lost to follow-up. One of the drawbacks of the study is we would not know if a patient presented to one of the peripheral hospitals for a stroke or a TIA, so certainly those would be lost to follow-up”.

In addition, Lal mentioned a paper which reports results contrary to the conclusions of the present study. “In the meta-analysis of the ICSS data and the CREST data with four years of follow-up there is actually a strong relationship between restenosis after carotid endarterectomies but no relationship to stroke after post-stenting restenosis,” he said, going on to ask how Strauss would respond to this data.

“From what I found in my literature review, even when most patients do develop greater than 80% restenosis, most of them do not actually develop ischaemic events and therefore there is not really much reason to intervene,” replied Strauss, adding that “once again, with a retrospective study it is hard to assess whether or not most patients go on to develop ischaemic attacks.”


Please enter your comment!
Please enter your name here