An Unjustified Return to the Past
To the Editor:
We read with interest the article by Qureshi et al1 regarding the role of conventional arteriography (CA) in evaluating patients with internal carotid artery (ICA) stenosis assessed by Doppler ultrasound (DUS) in general practice. Because we recently looked at this issue,2 and because we performed all carotid endarterectomies (CEAs) on the basis of DUS findings alone, we would like to comment on certain concerns raised by this article.
To determine whether it is appropriate to select patients on the basis of DUS studies performed at 20 different neurovascular laboratories, Qureshi et al1 used CA to evaluate 130 patients referred to their endovascular service with symptomatic (≥50%) or asymptomatic (≥60%) ICA stenoses. The positive predictive value of DUS in identifying symptomatic and asymptomatic candidates for ICA intervention was 80% and 59%, respectively, with a false-positive value of 20% and 41%, respectively. In addition, the authors1 reported that an analysis on a subset of 41 patients who underwent DUS at their laboratory, accredited by the Intersocietal Commission for the Accreditation of Vascular Laboratories, revealed a false-positive value of 20% (8/41). In the light of these findings, Qureshi et al suggested that a large proportion of the patients referred to their service on the basis of DUS in general practice did not in fact have significant stenosis likely to benefit from ICA intervention, and that inappropriate ICA intervention could result in 1 in every 5 patients, even when studies were performed at the best ultrasound laboratories. While we might agree with their first point, we have some doubts as to the reliability of the DUS performed at most of the laboratories considered in this study.
At many institutions of many countries, DUS currently provides adequate physiological and imaging information for clinical decision-making and operative treatment in most patients with ICA stenosis, without the use of CA. The accuracy of DUS in classifying the severity of disease is higher than 90% in accredited laboratories, with experienced technicians and a physician’s careful review. After a technically adequate DUS (the likelihood of error has been even further reduced by the introduction of echo-contrast agents that have significantly helped in distinguishing between an occlusion and a pseudo-occlusion3), preoperative CA has rarely induced any changes to patient management, and in as few as 1 to 2% of CEA cases in several recent series.4–6⇓⇓ The DUS criteria adopted vary according the angiographic standard used7–11⇓⇓⇓⇓ and should be tailored to specific institutional and individual situations. If DUS is used as a screening tool, a high sensitivity is required to avoid missing patients who might benefit from CEA. On the other hand, if DUS is used as the only preoperative test for ICA disease, a high specificity is also required in order to avoid performing any unnecessary CEA. However, DUS relies on the technical skills and experience of the operator.8,12⇓ Each laboratory should validate its own criteria for grading ICA stenosis against CA before DUS can be considered a reliable diagnostic tool for patient selection. Criteria for grading ICA stenosis must be both machine- and laboratory-specific, bearing in mind that both velocity criteria and frequency shift criteria differ considerably in different duplex machines. Moreover, applying the same diagnostic criteria to different equipment is a potential source of inaccuracy, and using more than one make of scanner at the same laboratory involves applying different duplex criteria to each machine. In an era of continuous renewal of technologies and turnover of human resources, DUS machines and technologists also change, possibly affecting the outcome of DUS. Regular internal validation is consequently mandatory within each laboratory to achieve and maintain the highest level of DUS accuracy. The “alarming” findings reported by Qureshi et al on the inaccuracy of DUS in general practice should not cast a shadow on the effectiveness of the DUS method as a stand-alone screening and diagnostic tool; they are merely the expression of the laboratories’ failure to implement regular internal validation, even if they are accredited.
Moreover, Qureshi et al stated that it is their policy to perform CA for all patients with suspected ICA lesion, and not just for patients whose DUS is considered inadequate. Since the authors observed no complications in 94 patients who underwent CA alone, they concluded that “given the present low procedure-related morbidity and high degree of accuracy in centers with documented low morbidity, CA may be considered in each patient before a decision is made regarding ICA intervention”.1 Although this represents the preferred procedural tendency at many institutions where ICA stenting is routinely performed, this suggestion constitutes a return to the past and is frankly misleading. In fact, if we consider all 130 patients who underwent CA, including those treated by means of ICA stenting in the same session, the overall complication rate was 8.4% (11/130), with a neurological complication rate of 3.8% (5/130). This confirms the fact that CA carries a definite neurovascular risk in the evaluation of patients before CEA, even when it is performed by experienced hands.13 In the Asymptomatic Carotid Atherosclerosis Study14 in particular, the CA-related stroke rate was 1.2%, but this apparently low incidence is clinically relevant considering that all patients were asymptomatic and that CA alone accounted for nearly one half of all perioperative strokes. Qureshi et al1 stated that “all complications were directly related to use of larger catheters and delivery devices used for stenting,” but this is just a hypothesis. Many symptomatic and asymptomatic patients (on the basis of DUS findings in general practice) come to our university vascular service from various parts of the country. We find it more ethical, faster, safer, more cost-effective, and less invasive to repeat DUS at our laboratory rather than exposing patients (especially if they are asymptomatic) to an unnecessary additional CA-related risk, reserving the CA procedure for a few selected circumstances. We believe that, for ICA surgery to be justified, the risks have to be minimized as much as possible, so the current diagnostic trend must surely be noninvasive, as the Harvard group also recently pointed out.15
- ↵Qureshi AI, Suri FK, Ali Z, Kim SH, Fessler RD, Ringer AJ, Guterman LR, Budny JL, Hopkins LN. Role of conventional angiography in evaluation of patients with carotid artery stenosis demonstrated by Doppler ultrasound in general practice. Stroke. 2001; 32: 2287–2291.
- ↵Ballotta E, Da Giau G, Abbruzzese E, Saladini M, Renon L, Scannapieco G, Meneghetti G. Carotid endarterectomy without angiography: can clinical evaluation and duplex ultrasonographic scanning alone replace traditional arteriography for carotid surgery work-up? A prospective study. Surgery. 1999; 126: 20–27.
- ↵Meents H, Burkard A. Contrast-enhanced colour duplex sonography of carotid arteries. Eur Radiol. 1994; 4: 533–537.
- ↵Alexandrov AV, Bladin CF, Maggisano R, Norris JW. Measuring carotid stenosis: time for a reappraisal. Stroke. 1993; 24: 1292–1296.
- ↵Thiele BL, Jones AM, Hobson RW, Bandyk DF, Baker WH, Sumner DS, Rutherford RB. Standards in noninvasive cerebrovascular testing: report from the Committee on Standards for Noninvasive Vascular Testing of the Joint Council of the Society for Vascular Surgery and the North American chapter of the International Society for Cardiovascular Surgery. J Vasc Surg. 1992; 15: 495–503.
- ↵Neale ML, Chambers JL, Kelly AT, Connard S, Lawton MA, Roche J, Appleberg M. Reappraisal of duplex criteria to assess significant stenosis with special reference to reports from the North American Symptomatic Carotid Endarterectomy Trial and the European Carotid Surgery Trial. J Vasc Surg. 1994; 20: 642–649.
- ↵Hankey GJ, Warlow CP, Sellar RJ. Cerebral angiographic risk in mild cerebrovascular disease. Stroke. 1990; 21: 209–222.
We appreciate the insightful comments by Ballotta and colleagues. As they correctly pointed out, variations in the quality of results among Doppler ultrasonographic laboratories preclude effective generalization of any results derived from a single site. The relative inaccuracy of carotid Doppler ultrasound in detecting carotid stenosis observed in our study may not be seen at other centers, as mentioned by Ballotta and colleagues.1,2⇓ However, the converse may be true as well in that the high degree of accuracy reported by some centers may not be reproducible in general practice surroundings, where referrals are made on the basis of studies performed at many laboratories. Johnston and Goldstein3 compared the results of carotid Doppler ultrasound and contrast angiography in 569 consecutive patients evaluated for carotid endarterectomy. The rate of misclassification of carotid stenosis among patients undergoing carotid endarterectomies that were performed on the basis of Doppler ultrasound results alone was found to be 28%. Similar misclassification rates were found for patients evaluated at academic medical centers and community hospitals in the study performed by Johnston and Goldstein. The investigators recommended that surgical decisions be made with caution if they are based on the results of Doppler ultrasound alone. As Ballotta and colleagues correctly pointed out, each setting must have rigorous internal validations performed continuously to ensure the accuracy of Doppler ultrasound assessment of carotid stenosis. We agree that cerebral angiography does carry an inherent risk of complications that ranges from 0.5% to 4%.4–6⇓⇓ These include ischemic stroke, allergic reaction to contrast material, and local vascular complications related to femoral artery catheterization. Major complications are reported in 0.1% to 0.5% of the cases.4–6⇓⇓ However, carotid endarterectomy carries a higher risk of major complications (5% to 7% in large cross-sectional surveys), which varies according to patient population, surgeons, and centers.7,8⇓ Therefore, our recommendation is that each center rigorously evaluate the accuracy of its noninvasive diagnostic practices to ensure that only patients who can potentially benefit from carotid endarterectomy are treated. This avoids exposure of patients to potential treatment-related complications in the absence of any anticipated benefit.
- ↵Erdoes LS, Marek JM, Mills JL, Berman SS, Whitehill T, Hunter GC, Feinberg W, Krupski W. The relative contributions of carotid duplex scanning, magnetic resonance angiography, and cerebral arteriography to clinical decision making: a prospective study in patients with carotid occlusive disease. J Vasc Surg. 1996; 23: 950–956.
- ↵Ballotta E, Da Giau G, Abbruzzese E, Saladini M, Renon L, Scannapieco G, Meneghetti G. Carotid endarterectomy without angiography: can clinical evaluation and duplex ultrasonographic scanning alone replace traditional arteriography for carotid surgery workup? A prospective study. Surgery. 1999; 126: 20–27.
- ↵Johnston DC, Goldstein LB. Clinical carotid endarterectomy decision making: noninvasive vascular imaging versus angiography. Neurology. 2001; 56: 1009–1015.
- ↵Qureshi AI, Suri MF, Ali Z, Kim SH, Fessler RD, Ringer AJ, Guterman LR, Budny JL, Hopkins LN. Role of conventional angiography in evaluation of patients with carotid artery stenosis demonstrated by Doppler ultrasound in general practice. Stroke. 2001; 32: 2287–2291.
- ↵Dawson DL, Zierler RE, Strandess DE Jr, Clowes AW, Kohler TR. The role of duplex scanning and arteriography before endarterectomy: a prospective study. J Vasc Surg. 1993; 18: 673–683.
- ↵Dion JE, Gates PC, Fox AJ, Barnett HJ, Blom RJ. Clinical events following neuroangiography: a prospective study. Stroke. 1987; 18: 997–1004.
- ↵McCrory DC, Goldstein LB, Samsa GP, Oddone EZ, Landsman PB, Moore WS, Matchar DB. Predicting complications of carotid endarterectomy. Stroke. 1993; 24: 1285–1291.