Carotid Endarterectomy for Transient Monocular Visual Loss and Other Ocular Ischemic Conditions

Evidence from randomized clinical trials (RCTs) is a rare and prized commodity in medicine. When it does come along, as in the case of carotid endarterectomy (CE) for stroke prevention, it evokes an amazing variety of responses (1-5). The evidence may be rejected because it does not conform to preconceptions. It may be dismissed as out of date, especially when techniques have improved. The fine print or follow-up reports may be ignored and the evidence misapplied to subgroups that differ substantially from the general cohort. It may be applied to types of patients who were not even in the original cohort, and, most often, “statistically significant” benefits may be surmised as “clinically meaningful” without any consideration of cost. In this issue of the Journal of Neuro-Ophthalmology, three articles (6-8) deal with the use of CE for transient monocular visual loss (TMVL) and other ophthalmic indications. It is astounding how differently the same evidence is interpreted and applied. In a Viewpoint article, Wolintz (6) finds little robust evidence to support a meaningful benefit of CE in TMVL. The basis of her argument is an after-the-fact subset North American Symptomatic Carotid Endarterectomy Trial (NASCET) analysis by Benavente et al (9) who found that TMVL patients required at least three of the following features to benefit from CE: 1) male sex, 2) age of 75 years or older, 3) history of hemispheric transient ischemic attack (TIA) or stroke, 4) history of intermittent claudication, 5) ipsilateral internal carotid artery stenosis of 80%-94%, and 6) absence of collateral vessels on angiography. Only 20% of the NASCET cohort met those criteria. Yet, in another Viewpoint article, Nicolle and Hachinski (7) report that at the University of Western Ontario, the home of the authoritative NASCET study, CE is offered to patients of either sex and of any age if the ipsilateral cervical carotid stenosis is high grade (>70%) without consideration of how many of the above risk factors are present in the patients. In a Point Counter Point debate on CE for TMVL, Caplan and Hertzer (8) both support the use of CE for TMVL, dismissing the findings of Benavente et al (9) for different reasons. Caplan argues that, as a large study, the NASCET could not consider individual patient characteristics such as the echographic features of the plaque, now believed to refine stroke risk. Hertzer rejects the subgroup analysis of TMVL in the NASCET report as being statistically flawed. There is no doubt that RCT data support CE for stroke prevention in patients with transient hemispheric TIA and stroke (2,3). But do patients with clinical manifestations of ocular ischemia belong in that group? In its abstract summarizing a review of articles on CE and stroke prevention published between 1990 and 2002, the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology (10) affirms that “evidence supports carotid endarterectomy for severe (70 to 99%) symptomatic stenosis” but makes no exception for patients with TMVL or retinal stroke. Yet, deep within the text of the article, the authors acknowledge that “patients with hemispheric TIA/stroke had greater benefit from CE than patients with retinal ischemic events” (10). This difference in benefit is based on the fact that medically managed patients with retinal ischemic events clearly have a much lower risk of major hemispheric stroke than do patients with hemispheric TIA or minor stroke (11). What about the value of CE for other ophthalmic indications? Nicolle and Hachinski (7), Caplan (8), and Hertzer (8) consider central retinal artery occlusion (CRAO) as much of an indication for CE as TMVL. Yet CRAO need not be the result of an embolic event; it often appears pathologically as thrombosis within a retrolaminar atherosclerotic plaque (12). In one study of 34 patients with CRAO, only 12% had ipsilateral carotid stenosis of 80% or more (13); in another study of 77 patients, only 20% had stenosis of 50% or more (14). Moreover, although the risk of hemispheric stroke in patients with CRAO has not been rigorously documented, it seems to be lower than in patients with recent hemispheric TIA or stroke (15). Finally, what about CE for patients with asymptomatic Hollenhorst plaque? Nicolle and Hachinski (7) do not offer CE to such patients, but Hertzer (8) does, and Caplan (8) might. Is there a compelling reason to believe that the asymptomatic patient with high-grade carotid stenosis and a Hollenhorst plaque should be managed differently from the asymptomatic patient with high-grade carotid stenosis but no such plaque? An older study (16) pointed out that early death from myocardial infarction is the major risk in patients with retinal plaque. No study has directly addressed the value of CE in such patients. But the American Academy of Neurology report of 2005 (10) finds that for “asymptomatic patients with 60 to 99% stenosis, the benefit/risk ratio is smaller compared to symptomatic patients,” suggesting that in this group, “individual decisions must be made.” The authors cite the following sobering facts in relation to CE in asymptomatic high-grade carotid stenosis: 1) the Mayo study (17) was terminated early because of a high rate of myocardial infarction (22%) in the surgical group; 2) the Veterans Administration study (18) found no significant benefit of CE; 3) the Asymptomatic Carotid Artery Study (19) found that CE reduced the relative risk of ipsilateral hemispheric stroke by 53%, but the five-year risk was only 11% in the medically treated group. The small benefit would have disappeared if the peri-operative stroke rate had been more than 2.3%, a rate much lower than has been found in other studies (20); and 4) the Asymptomatic Carotid Surgery Trial (21) found that patients undergoing early CE had a five-year stroke risk of 6.4%, whereas patients undergoing delayed CE had a five-year stroke risk of 11.8%. But these data totaled all strokes, including those in the contralateral carotid and vertebrobasilar distributions. This information leads to some inescapable conclusions: 1) TMVL should not be lumped with hemispheric TIA or stroke; the stroke risk without CE is relatively small, and the benefit from CE is not established; 2) CRAO should not be lumped with a hemispheric TIA or stroke; carotid stenosis is often mild, suggesting a non-embolic mechanism; and 3) asymptomatic patients with Hollenhorst plaques and high-grade carotid stenosis may have a stroke risk equal to that of patients without such plaques; if so, they should not undergo CE because the reduction in future stroke is too low, even if the peri-operative complications are minimal. Two other facts ought to caution the use of CE in ocular ischemia. First, patients probably are at just as high a risk of myocardial infarction and cardiac death as are those with hemispheric TIA and stroke (14-16), which is a risk that cannot possibly be made better by CE. Second, the benefit of CE, even in symptomatic patients, depends on a peri-operative stroke and death rate less than 5% (2). In most hospitals, this rate is not tallied or made known to non-surgeons (22); when it is tallied, it is often more than an acceptable level (23). For all these reasons, it is time to remove ophthalmic ischemic conditions from the indications for CE.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.