How Duplex Ultrasound Screening Can Lead to Overuse of Carotid Interventions Gregory L. Moneta, M.D. Chief, Division of Vascular Surgery Department of Surgery Knight Cardiovascular Institute Oregon Health & Science University Portland, Oregon No Disclosures Prevalence >70% Asymptomatic ICA Stenosis* Low Prevalence of Disease Low Virulence of Disease/Low Therapeutic Index of Intervention Poor Specificity Screening Test 8 P r 7 e 6 v 5 a l 4 e 3 n c 2 e 1 % 0 0.1 0.0 0.2 0.1 0.8 1.0 0.9 0.2 <50 50-59 60-69 70-79 >80 Age (years) 2.1 Men Women 3.1 *Stroke 2010;41:1294-1297
Prophylactic Carotid Intervention Statins and Ischemic Stroke - Background of better medical management of atherosclerosis. - Increased use of Plavix, statin medications, beta blockers and ace inhibitors. - Differences between surgical and medical management of ICA high-grade stenosis may therefore be even less pronounced than those suggested by the random trials. Meta analysis of 90,056 participants in 14 randomized trials of statin therapy. 22% reduction in first ischemic stroke per mmol/l reduction in LDL cholesterol. Independent of previous MI or CAD. Lancet 2005; 366:1267-1278 Prophylactic Carotid Intervention - The estimated current risk of stroke in patients with asymptomatic carotid stenosis is 0.5% per year. - Lower than the risk of CEA in CREST or stenting in CREST and the recently published ACT 1 trial. - Unlike symptomatic disease, the severity of stenosis in patients with asymptomatic stenosis does not reliably predict stroke risk. Prophylactic Carotid Intervention - Following results of the randomized trials of carotid surgery use carotid endarterectomy increased. - Carotid stenting initially also took off driven primarily by industry. - Up to 92% of carotid interventions in the United States are now performed for asymptomatic disease.
Prophylactic Carotid Intervention - There is a at best only a small benefit of prophylactic carotid intervention in asymptomatic patients. - Carotid intervention is often based only on duplex scanning without other preprocedure diagnostic modalities. Threshold Level vs Categories of Stenosis The use of relatively broad categories of stenosis reflects the fact duplex cannot predict stenosis in fine increments. One can therefore develop categories of stenosis with acceptable sensitivities and specificities where changes between categories are potentially clinically meaningful and reproducible. Natural history of the disease. Threshold Levels vs Categories of Stenosis Categories of Stenosis: The degree of stenosis is identified as being between high and low values. Threshold Levels of Stenosis: The degree of stenosis is identified as being above or below a threshold value. Threshold Level vs Categories of Stenosis Threshold levels of stenosis can be used to determine whether a patient is an anatomic candidate for an intervention using duplex scanning as the sole imaging modality prior to the intervention. -Above the line: yes! -Below the line: no! Predictive values and the therapeutic index of the procedure are more important when using threshold levels.
SRU Carotid Duplex Consensus Conference (Determination of Percent Stenosis) NASCET Based Criteria for ICA Stenosis Carotid Duplex Consensus Conference Society of Radiologists in Ultrasound 16 panelists: Radiology, Vascular surgery, Cardiology and Neurology Goals of conference: Develop recommendations for performance of exam Standardization of technical factors Recommend defined diagnostic criteria for ICA stenosis Carotid Duplex Consensus Conference Carotid Duplex Consensus Conference Andrei Alexandrov J. Dennis Baker Carol Benson * Ed Bluth Barbara Carroll Michael Eliasziw John Gocke Barbara Hertzberg Participants Moderator: Ed Grant * Sandra Katanick Gregory Moneta * Larry Needleman John Pellerito Joe Polak Ken Rholl Doug Wooster Gene Zierler * = Writing Committee
Carotid Duplex Consensus Conference (SRU Suggested Criteria for Duplex Diagnosis of ICA Stenosis) Society Radiologists in Ultrasound (SRU) Consensus Conference Radiology 2003; 229: 340-346 Sens Spec PPV NPV Accuracy* ICA PSV > 140 98.3 61.9 54.3 98.7 73.4 190 96.6 72.2 61.5 97.8 79.9 210 94.8 78.6 67.1 97.1 83.7 230 93.1 81.0 69.2 96.2 84.8 250 91.4 82.5 70.7 95.4 85.3 265 89.7 84.9 73.2 94.7 86.4 280 84.5 86.5 74.2 92.4 85.9 295 84.5 88.1 76.6 92.5 87.0 325 82.8 90.5 80.0 91.9 88.0 345 70.7 92.9 82.0 87.3 85.9 375 65.5 96.0 88.4 85.8 86.4 400 51.7 96.8 88.2 81.3 82.6 440 41.4 98.4 92.3 78.5 80.4 * >70% ICA Stenosis OHSU Criteria for >60 99% ICA Stenosis (n = 352 angio/ duplex comparisons) (ICA PSV 290cm/s and EDV 80cm/s) Sensitivity 78% Specificity 96% PPV 95% NPV 92% Accuracy 88%
Carotid Duplex Consensus Conference A PSV of 230 cm/sec maximizes sensitivity for detection of >70% ICA stenosis likely at the expense of higher numbers of false positives. For symptomatic patients one can argue this is irrelevant as NASCET also indentified >50% to 69% symptomatic ICA as benefiting from CEA. For asymptomatic patients this my be unacceptable as potentially many patients with less than 70% stenosis could be subject to a procedure of now marginal or even uncertain or no benefit! Follow-up Asymptomatic Carotid Stenosis (OHSU Study) NIH sponsored trial of the effects of folate supplementation on PAD 562 subjects were entered and randomized 87 subjects met inclusion criteria with an ICA PSV between >230 and 300 cm/s Mean follow-up 27.4 months; range 4-102 months Follow-up Asymptomatic Carotid Stenosis (OHSU Study) - Subjects were followed every 6 months with clinical history and physical exam screening for new neurologic symptoms - Carotid artery duplex scanning every 6 months - Subjects assessed for progression of ICA PSV to >300 cm/s Follow-up Asymptomatic Carotid Stenosis (OHSU Study) - 44/87 (50.6%) subjects remained <290 cm/s during follow-up (Non-progression group) (Mean F/U = 34.8m) - 43/87 (49.4%) subjects progressed beyond 290 cm/s during follow-up (Progression group) (Mean F/U = 32m) - Mean time to progression: 19.9 months (range 6 60 months)
Vascular Laboratory Endpoint and Operative Intervention Clinical Endpoint Progression Percent % 100 80 60 40 20 0 48.8 51.1 Progression 13.6 86.4 OHSU Operative Nonprogression Nonoperative Percent % 100 80 60 40 20 0 3/87 (3.4%) Ipsilateral CVA, TIA, AF 84/87 (96.6%) Asymptomatic Ipsilateral Stroke - 3/87 subjects had ipsilateral stroke - 2 strokes occurred after progression of ICA PSV to >300 cm/s - One patient refused operative intervention - One patient was unfit for intervention Age Gender Multivariate Logistic Regression Analysis for Progression to >300 cm/s Diabetes (OR=.25) Tobacco Hypertension Statin Ipsilateral PSV Contralateral PSV (OR=.99) p=.72 p=.29 p=.03 p=.70 p=.53 p=.39 p=.21 p=.04 0.0 0.5 1.0 1.5 2.0 2.5
Conclusions Diabetes and an elevated contralateral PSV are independent risk factors associated with vascular laboratory determined progression of ICA stenosis. In subjects with ICA PSV 230 to 300 cm/s the 2-year risk of ipsilateral stroke or TIA is <2%. Conclusions Categories of stenosis as carotid criteria have proven useful in natural history studies of carotid atherosclerosis and in clinical monitoring of individual patient s level of ICA stenosis Use of threshold levels of stenosis as carotid stenosis criteria may facilitate clinical decision making for potential intervention. Conclusions Questions? The therapeutic index for intervention is likely too low to justify PCI for an ICA PSV >230 cm/s as the minimum cut point. Use of the SRU Consensus Conference recommendation of ICA PSV >230 cm/s as diagnostic of ICA stenosis >70% in asymptomatic patients will result in overuse of prophylactic carotid intervention. Columbia River, Portland Oregon