Portal de Periódicos da CAPES

The Management of Transient Monocular Visual Loss

2005; Lippincott Williams & Wilkins; Volume: 25; Issue: 4 Linguagem: Inglês

10.1097/01.wno.0000190062.33744.68

1536-5166

Louis R. Caplan, Norman Hertzer,

Cerebral Venous Sinus Thrombosis

A 60-year-old man develops sudden complete transient loss of vision OD three days ago lasting 30 seconds. It occurred while he was reading. He has a history of medication-controlled hypertension and has smoked 1/2 pack of cigarettes daily for many decades. His medical history is otherwise unremarkable. General physical, ophthalmological, and neurological examinations are normal.Figure: Louis R. Caplan, MD Professor of Neurology Harvard Medical School Chief of Section in Cerebrovascular Disease Beth Israel Deaconess Medical Center Boston, MassachusettsFigure: Norman R. Hertzer, MD Emeritus Chairman Department of Vascular Surgery The Cleveland Clinic Foundation Cleveland, OhioWhat is the probable cause of the deficit, and what is the next step in evaluation? Louis R. Caplan, MD: The most likely cause of his transient monocular visual loss (TMVL) is retinal ischemia (1-3). Occasionally, transient optic nerve ischemia can present similar symptoms (3). Most ocular ischemia is caused by embolism to the ophthalmic artery, and its branches that feed the retina and optic nerve. Sources of embolism to the eye include the heart, aorta, ipsilateral extracranial and siphon portions of the internal carotid artery, and the ophthalmic artery. In this patient, a normal ophthalmologic examination excludes impending central retinal vein occlusion (4), optic disc drusen (5), and pre-retinal loops (6), which are occasional causes of ocular ischemia. There were no Hollenhorst plaques, white platelet-fibrin thrombi, or calcific emboli within the retinal arteries-clues to the origin of emboli. The risk factors in this patient make internal carotid artery (ICA) disease the most likely source of TMVL. Atheromatous disease of the aorta or heart is another important possibility. Patients with polycythemia and thrombocytosis can occasionally develop TMVL. Temporal arteritis is an infrequent cause. I would begin by ordering a complete blood count and sedimentation rate, an electrocardiogram (EKG), a duplex ultrasound study of the neck arteries, and a transcranial Doppler (TCD) exploration of the cranial arteries. The duplex examination should detect significant ICA disease at the common carotid artery bifurcation. TCD can help detect disease of the carotid siphon proximal to the ophthalmic artery, disease of the ophthalmic artery itself, and help to quantify any reduction in blood flow related to ICA disease in the neck (7). If the duplex findings in the neck are equivocal, 20-30 minutes of TCD monitoring of the intracranial middle cerebral arteries (MCAs) might help identify the presence and source of emboli (8,9). If the source is the ipsilateral ICA, embolic signals would be detected only in the ipsilateral MCA; if the source is cardiac or aortic, emboli should be detected bilaterally. If a four-probe embolic detection system is available, the ICAs can also be probe sites. Emboli from the heart or aorta would first be detected in the ICA and shortly thereafter in the ipsilateral MCA. If the emboli arise from the ICA, no embolic signals would be heard in the neck. I usually order a brain and head-and-neck vascular imaging study (computed tomography [CT]/CT angiography [CTA] or MRI/MR angiography [MRA]) along with the ultrasound. Some patients have had unexpected brain infarcts; their distribution can give a clue as to their cause. Brain infarction ipsilateral to ICA disease also tells of the biological activity of the lesion despite lack of symptoms (10,11). CTA and MRA examinations of the neck and head can show the presence, extent, and morphology of an ICA lesion in the neck and provide images of occlusive disease that might involve the pharyngeal portion of the ICA (such as fibromuscular dysplasia or dissection), the ICA siphon, and its ophthalmic artery branch. An EKG should have detected atrial fibrillation and recent or past myocardial infarction. A transesophageal echocardiogram (TEE), which searches for embolic sources in the cardiac ventricles, atria, interatrial septum, cardiac valves, and aorta, might be indicated depending on the results of the blood tests, EKG, neck ultrasound, TCD, and brain and vascular imaging (12). Norman R. Hertzer, MD: Despite its unusually brief duration, this episode of TMVL almost certainly represents retinal microembolization, which, in this case, probably was related to platelet thrombi given the transitory nature of the event and the absence of any atheromatous debris observed in the retina. Until proven otherwise, atherosclerotic stenosis or ulceration in the ipsilateral carotid bifurcation represents the most likely embolic source in this patient, especially considering his long history of tobacco use. If cervical carotid disease can be excluded by appropriate studies, however, then cardio-embolic and other less common etiologies would have to be investigated (13). In an effort to prevent any further micro-embolic events while the diagnosis is being established, the patient should immediately be instructed to begin taking aspirin at least 81 mg/day (14). The next step in the evaluation would be a carotid duplex scan, preferably at a facility that has been certified for accuracy by the Intersocietal Commission for the Accreditation of Vascular Laboratories (www.iacvl.org). Provided high-grade (70%-99% or 80%-99%) extracranial carotid stenosis is identified on the basis of reliable duplex scanning, carotid endarterectomy (CE) could legitimately be recommended without the small additional risk of a confirmatory arteriogram (15). But if the duplex scan reveals only mild to moderate cervical carotid disease, MRA or even catheter arteriography may be required to exclude non-stenotic plaque ulceration, aortic arch lesions, or intracranial carotid siphon disease proximal to the origin of the ophthalmic artery. An echocardiogram might be indicated if the underlying etiology still remains in doubt. The patient undergoes a duplex cervical carotid ultrasound that reveals 80% cross sectional stenosis on the right and 50% on the left. A brain MRI is negative. What do you advise as the next step? Louis R. Caplan, MD: I would first review the results with the patient, indicating that he has significant carotid artery disease and is at risk for stroke. I would explain that the alternatives are maximal medical therapy, CE, or carotid angioplasty/stenting. Maximal medical therapy, which consists of an antiplatelet agent, a statin in the appropriate dose, and an ACE inhibitor or ACE receptor blocker, would be prescribed. Although the risk of stroke after a retinal transient ischemic attack (TIA) is about one third the risk of stroke after a hemispheric TIA (16), it is likely to be more than 10% during the ensuing year or two (16). If he opted for medical treatment and was to have a hemispheric TIA or minor stroke, then the risk of major stroke would increase considerably, and aggressive therapy would be mandated. If medical treatment has not been optimal, consideration should be given to adding a statin or increasing the dose of a statin if one is already being used. Studies have shown that large doses of statins can reduce plaques or at least stabilize them and reduce the frequency of stroke (17-20). ACE inhibitors and ACE receptor blockers have also been shown to have salutary effects on the endothelium and could be prescribed (21,22). Some antiplatelet agents, especially an aspirin-dipyridamole combination or cilostazol, have effects on platelets and vascular endothelium and may be more protective than aspirin alone or clopidogrel. These medical prescriptions can be given while further investigations are performed and while the patient is deciding on a preferred strategy of treatment. The nature of the carotid plaque might influence prognosis and treatment. The main features are plaque echogenicity, regularity, homogeneity, location, and the thinness of the fibrous cap overlying the plaque (23-25). Hypoechoic plaques contain more macrophages and lipid materials and are more likely to become symptomatic (and to be reducible by statins). Hyperechoic material indicates calcium and stability. Heterogeneity and irregularity of the surface, especially with ulceration, carry a worse prognosis for symptom development. These features can now be detected by high-quality B-mode ultrasound and, in some centers, by modern CT and MRI imaging of plaques using cross-sectional views of the artery. Brain imaging is also useful. The presence of brain infarcts on CT or MRI attributable to the stenotic artery gives evidence of biological activity of the carotid disease despite the absence of hemispheric symptoms and increases the risk of stroke (10). Embolic monitoring also provides prognostic data. The number and frequency of microemboli (high-intensity transient signals on TCD) and their response to medical treatment is predictive of symptom development (26-28). If the patient makes up his mind to undergo CE or stenting, I would refer him to the appropriate specialist. If he is opposed to a procedure, I would maximize medical treatment and instruct him about hemispheric attacks. If he is uncertain and believes that more data would help him decide, I would schedule further investigations such as cranial MRI, high-quality duplex ultrasound, and TCD of the intracranial arteries, along with embolic monitoring. Norman R. Hertzer, MD: For over a decade, the importance of identifying symptomatic carotid disease primarily has been predicated on the conclusions of the North American Symptomatic Carotid Endarterectomy Trial (NASCET). In this prospective, multicenter clinical trial, patients having a recent history of TMVL, transient hemispheric ischemia, or non-disabling stroke were randomized to receive CE plus “optimal” medical management (predominantly anti-platelet therapy) or medical management alone for either 50%-69% or 70%-99% carotid stenosis documented on the appropriate side of the neck by arteriography. The initial NASCET results were reported in 1991 for 70%-99% stenosis (29) and demonstrated that CE significantly reduced the 2-year ipsilateral stroke rate as compared with medical management alone (9% versus 26%; P < 0.001). Interestingly, subset analysis later revealed that patients in the medical cohort who originally presented with transient hemispheric TIA or stroke had a higher 2-year stroke rate than did patients who presented with TMVL (43.5% ± 6.7% versus 16.6% ± 5.6%; P = 0.002), representing a relative risk ratio of 3.23 (95% confidence interval, 1.47-7.12) (16). It took substantially longer for the NASCET trialists to reach conclusions regarding the efficacy of CE in symptomatic patients who had only 50%-69% carotid stenosis. Their 1998 report (30) showed that, whereas still statistically significant, the benefit of surgical treatment was much less convincing. The 5-year ipsilateral stroke rate for CE in these patients was 15.7% versus 22.2% in medically treated patients (P = 0.045). CE provided no advantage in comparison to medical management for an exceedingly small group of only 36 patients who were randomized because of TMVL or retinal stroke. Provided the duplex scan was performed in a reliable noninvasive vascular laboratory, I would recommend right CE by a qualified surgeon or participation in one of the current randomized trials comparing CE to percutaneous transluminal angioplasty (PTA) and stenting with the use of a cerebral protection device. The choice between these two options would depend in part on their availability at the center where he receives his care. I should also add that obtaining an MRI brain scan seems to have been an unnecessary extravagance in our patient, especially after only a single 30-second episode of TMVL. The chance that such a study would have revealed any useful information seems remote. In 2001, Benavente et al (31) finally reported the outcomes of the subgroup of TMVL patients from the NASCET trial who were randomized to CE versus best medical treatment. As you recall, they found that a meaningful benefit of CE was restricted to those who had at least three of the following characteristics: age >75 years, male, history of hemispheric TIA or stroke, history of intermittent calf claudication, ipsilateral internal carotid stenosis >80%, or no collaterals on angiography. Our patient meets only two of these criteria. Is CE indicated? Louis R. Caplan, MD: The cited NASCET data (31) must be put in the perspective of the study. The analysis was retrospective. It considered effectiveness as reducing strokes (not just those related to the ipsilateral carotid artery). This means that factors related to strokes and vascular disease in general, such as hypertension, diabetes, hyperlipidemia, and smoking, were heavily weighted. The data analyses used indirect risk factors (age, sex, and claudication) as surrogates for estimating the seriousness of vascular disease. The percentage reduction of luminal size on catheter cerebral angiograms and the presence of collateral circulation were the main direct measures of the nature and severity of the carotid artery disease. The details of the composition and morphology of the stenosing plaques, as shown by high-quality ultrasound, were not included. Several factors were not included or sufficiently emphasized, as is true in many large trials that must lump disparate patients. For example, the presence of brain infarcts in the territory of the stenosed artery and other medical therapies. Moreover, since the NASCET study was planned, advances in technology and treatment have occurred. There are new drugs and information about effective dosages of statins, ACE inhibitors, ACE receptor blockers, and non-aspirin platelet inhibitors, plaque morphology, as studied by modern ultrasound, the importance of brain infarcts seen on MRI, the ability to detect emboli by TCD, and the availability of carotid angioplasty/stenting. Norman R. Hertzer, MD: So many subset analyses from the NASCET have been published since the initial disclosures from this trial in 1991 that it has become increasingly difficult to keep track of all of them. Now that I have read the article by Benavente et al (31), I find that its data regarding the treatment of TMVL are based on 107 patients having none of the five stated risk factors, 150 patients having two of the risk factors, and 103 patients having three or more of the risk factors. These sample sizes apparently do not detract from the statistical validity of the conclusions, but they do parse the total NASCET population (n = 2,885) into some pretty small pieces. I personally would be very uncomfortable using this information to withhold CE from a symptomatic patient who has 80%-99% carotid stenosis merely because he is only 60 years old and does not claudicate. Furthermore, it seems hard to know how the NASCET proportional hazards model applies to our patient because he has not yet had an arteriogram to demonstrate the presence or absence of collateral circulation distal to his high-grade cervical carotid lesion. An arteriogram presumably would have to be done in order to determine where he fits into the NASCET analysis, but findings from the Asymptomatic Carotid Atherosclerosis Study (ACAS) indicate that an arteriogram also would add a small but measurable risk to his management (32). Five (26%) of the 19 strokes or deaths contributing to the combined stroke and mortality rate of 2.3% in the surgical cohort of the ACAS were directly related to the preoperative arteriogram, an observation that undoubtedly has influenced a growing reliance on noninvasive testing alone in the selection of patients for CE. At the Massachusetts General Hospital, for example, only 10% of patients underwent preoperative arteriography before CE in 1998-1999 compared with 87% in 1989-1990 (33). As a practical matter, two other large randomized trials-the ACAS and the more recent Asymptomatic Carotid Surgery Trial-have shown that our patient already has carotid stenosis of sufficient severity to justify prophylactic CE even if he had no symptoms whatsoever (32,34). For all of these reasons, I believe that he is an appropriate candidate for CE despite the fact that he has only two of the NASCET risk factors described by Benavente et al (31). Of the two options, CE and angioplasty/stenting, which would you choose for this patient, or for others? Louis R. Caplan, MD: I would place some weight on the findings from high quality ultrasound and vascular imaging. Since the NASCET study, there have been a number of studies of the risks and benefits of angioplasty/stenting with and without protective devices (35-37). In a high carotid bifurcation or very long smooth plaque, stenting is likely better. In a very focal ulcerated lesion, surgery is likely better. If there is a tandem lesion distal to the bifurcation, then angioplasty/stenting might treat both lesions. The availability, experience, and complication rates of the surgeons and interventionalists are also important in deciding which treatment is preferable. I would discuss the options with the patient, attempting to share the risks and benefits of each approach in his individual case and in relation to the individuals available to provide the treatments. All things being equal, I usually favor stenting or entering the patient into a randomized trial that compares stenting with surgery (38). Norman R. Hertzer, MD: First let me give a brief overview of catheter-based intervention for carotid bifurcation disease. Isolated case reports and experience with small series of patients suggested nearly two decades ago that angioplasty of the carotid artery might prove to be an appropriate alternative to traditional surgical treatment for high internal carotid lesions near the skull base, recurrent stenosis after previous CE, or a history of cervical irradiation (39). Intraluminal stenting was introduced in the early 1990s to reduce the incidence of peri-procedural embolic events and recurrent stenosis that were associated with angioplasty alone, and over-the-wire cerebral protection devices were later developed in a further attempt to prevent strokes caused by atheromatous emboli dislodged during stent deployment. These refinements have widened the indications for catheter-based intervention, but as so often occurs with rapidly advancing technology, they also have turned outcome assessment into a moving target. The elusive nature of this target can be illustrated by two relatively recent examples involving symptomatic patients. In 2000, Golledge et al (40) conducted a meta-analysis of 33 single-center studies that had been reported from 1990 through 1999 and calculated that the risks for any related stroke (7.1% versus 3.3%; P < 0.001) or stroke or death (7.8% versus 4.0%; P < 0.001) were significantly higher for angioplasty than for CE. In 2001, the results of a multicenter, prospective randomized trial (CAVATAS) conducted in Europe from 1992 to 1997 indicated that, whereas the 30-day combined stroke and mortality rates (CSM) of angioplasty and CE were uniformly poor (10% versus 9.9%), the 1-year incidence of recurrent 70% stenosis was significantly higher (14% versus 4.0%, P < 0.001) for angioplasty (41). However, the conclusions of both of these investigations became obsolete nearly as soon as they were published for two reasons: 1) stenting had been done in only 44% of the angioplasty cases reviewed in the meta-analysis and in just 26% of the angioplasty cohort of the CAVATAS; and 2) cerebral protection devices were not yet available during either of the two study periods. Therefore, the evidence base for contemporary angioplasty/stenting with intra-procedural cerebral protection is limited to the past five years and still is evolving. Two industry-funded trials now have shown outcomes for angioplasty/stenting with cerebral protection that are equivalent to CE in a mix of symptomatic and asymptomatic patients drawn from the United States. The CARESS study (Boston Scientific, Natick, MA; Medtronic AVE, Santa Rosa CA) is a Phase I clinical trial in which 397 typical patients were treated at 14 sites by angioplasty/stenting (n = 143) or CE (n = 254) (42). In that study, there were no significant differences in the 30-day CSM (2.1% versus 2.4%) or the CSM-plus-myocardial infarction (MI) rates (2.1% versus 3.1%). The other study is the Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial (Cordis Corporation, Warren, NJ), in which 334 patients with medical comorbidities or anatomic features considered unfavorable for CE were randomized to angioplasty/stenting (n = 167) or CE (n = 167) at 29 participating centers (35). Using conventional end points consisting of the 30-day CSM plus ipsilateral stroke or death within 1 year, the SAPPHIRE trial had equivalent outcomes for angioplasty/stenting and CE (5.5% versus 8.4%; P = 0.36). When increased peri-procedural cardiac enzyme levels also were considered, the CSM/MI plus 1-year event rates were lower for angioplasty/stenting than for CE (12% versus 20%; P = 0.05). The ongoing Carotid Revascularization Endarterectomy versus Stent Trial (CREST) is a multicenter, prospective, randomized investigation that is supported by the National Institutes of Health with plans to recruit 1,200 to 1,600 symptomatic patients (43). No outcome data have been disclosed for approximately 650 randomizations, but information has been reported for 749 non-randomized patients who were treated by angioplasty/stenting with adjunctive cerebral protection during a lead-in phase used for credentialing catheter-based interventionalists at 51 sites (44). These results substantiate that the CSM for angioplasty/stenting is correlated with advancing age, ranging from 1.4% in 349 patients aged less than 70 years to 7.0% in 400 patients aged 70 years or older (P = 0.0006). Others (45) have made similar observations regarding age and the early risk of angioplasty/stenting, despite the use of anti-embolic devices. Now, which option would I choose for this patient? He does not represent a predictable early risk for either approach at only 60 years of age, and therefore, he should be given the opportunity to enroll in the CREST if a trial center is nearby. If not, his anticipated life expectancy becomes a consideration with respect to the risk for recurrent stenosis, which may be greater for angioplasty/stenting although this has not been rigorously verified for contemporary angioplasty/stenting techniques. In the absence of further data, I favor CE with carotid patching on the basis of my own experience. But the key point is: what are the comparative results of CE and angioplasty/stenting at the hospital where the patient will be treated? A decision concerning management options cannot be made objectively without this information. Would your management differ if, instead of transient visual loss, this patient had presented with acute persistent unilateral visual loss owing to a central retinal artery occlusion (CRAO)? Would management differ if the patient had presented without visual symptoms but had been found to have a Hollenhorst plaque in one eye? Louis R. Caplan, MD: My suggestions for management would not be different had the patient presented with a CRAO. If he was asymptomatic and had a Hollenhorst plaque discovered incidentally, I would be heavily guided by the nature and extent of the carotid artery lesion. I would also lean more toward surgery if brain imaging disclosed a brain infarct in the carotid artery territory. Norman R. Hertzer, MD: None of the prospectively randomized trials of CE or angioplasty/stenting has generated data specifically concerning either CRAO or incidental Hollenhorst plaques, but independent case series suggest that these findings are much less likely than TMVL to be associated with serious underlying carotid stenosis. For example, three studies (46-48) have collectively reported at least 60% ipsilateral carotid stenosis by duplex scanning in 31 (37%) of 84 patients with TMVL, compared to just seven (11%) of 63 patients with Hollenhorst plaques and two (7.4%) of 27 patients with CRAO. The observation that CRAO is a poor predictor of carotid artery disease is not surprising because it often is merely an ophthalmic manifestation of systemic illness, such as uncontrolled diabetes, arteritis, or a variety of coagulation disorders (49). For this reason, medical consultation probably should be obtained as one of the first steps in the evaluation of CRAO. If another plausible etiology is discovered, investigation of the carotid arteries may not be required provided, of course, that the onset of sudden monocular blindness had not been preceded by earlier episodes of classic TMVL in the same eye. With a history of previous TMVL, one might be more likely to suspect ipsilateral carotid emboli and proceed with a duplex scan. Asymptomatic Hollenhorst plaques can originate from the heart or the aortic arch as well as from the carotid bifurcation, in which case additional retinal emboli and/or neurologic events may still occur even if CE already has been performed for what was thought to be the responsible lesion (50). This emphasizes the importance of a thorough ophthalmologic evaluation, because the presence of bilateral retinal atheroemboli would clearly imply that they share a common proximal source in the chest. Another dilemma in the management of Hollenhorst plaques is the fact that they may persist for as long as three years (51). Therefore, unless serial ophthalmologic examinations have documented that a Hollenhorst plaque is recent, it could have originated in an atheromatous ulcer in the ipsilateral carotid artery or elsewhere that has healed and is no longer a threat for further embolization. Despite all of these uncertainties, a unilateral Hollenhorst plaque remains an appropriate indication for carotid duplex scanning if only because one does not want to overlook an easily accessible lesion for which anti-platelet therapy alone would offer no guarantee against future athero-embolic complications. It seems prudent, however, to limit CE or angioplasty/stenting strictly to patients who, like our own, have such severe carotid stenosis that they would be candidates for carotid intervention anyway. If you were this patient, with the same scenario, what would you wish to have done? Louis R. Caplan, MD: Stenting with a competent experienced interventionalist using a protective device. Norman R. Hertzer, MD: I would first seek out a qualified surgeon, preferably one who holds the Certificate of Special or Added Qualifications in Vascular Surgery issued by the American Board of Surgery and who practices in a hospital at which the outcome of CE is continuously audited for quality assessment. I would then opt for CE with a patch because so many randomized and non-randomized studies now have shown that patching reduces the risks for postoperative death, postoperative stroke, late stroke, recurrent stenosis, and re-operations as compared with primary arteriotomy closure. In general, CE/patching is associated with a CSM less than 2%, a 5% incidence of recurrent stenosis, and a re-intervention rate of 3% or less (33, 52-56). It may be that angioplasty/stenting will someday match these results and, at certain centers of excellence, perhaps it already does. At the present time, however, the technology of catheter-based carotid intervention and the learning curve of its practitioners are in their maturation phases. For this reason, I believe that CE with patching currently represents the best available approach throughout the countless communities, large and small, where most patients receive their care. Rebuttals Louis R. Caplan, MD: It comes as no surprise that Dr. Hertzer, a renowned vascular surgeon, would choose vascular surgery over angioplasty/stenting. We tend to choose the more familiar over the less familiar. No doubt Dr. Hertzer would have access to a very capable, experienced vascular surgeon with a superb track record for CE, a luxury our patient may not have. Dr. Hertzer's discussion and some of the queries emphasize trial data. Caring for individual patients is different from trials. The CREST and other trials pit CE against angioplasty/stenting with the naïve idea that one approach is always superior to the other. Actually, one approach may be better for some lesions and some patients with coexisting conditions and risks. The location, content, extent, and morphology of the occlusive plaque are important determinants for the choice of treatments. The carotid artery investigations reported on this patient do not sufficiently detail this information. Also, trials assume equipoise between the surgeon and the interventionalist. As Dr. Hertzer notes, this is not always the case at the institution where the patient will be treated and heavily influences selection of strategy. Most important, and again not included in trials, is the preference of the patient and the referring physician. General anesthesia poses a very definite risk and many doctors and patients would avoid it. CE is customarily performed under general anesthesia; angioplasty/stenting is not. Patients live in different socioeconomic-psychological milieus. Given the same information, they make different decisions. Some find the knowledge of a severe carotid lesion very difficult to live with. Told that they are at imminent risk of having a disabling stroke, they feel that a sword (the stenotic artery) is hovering over them. Some will gamble on removal of the sword. Others fear the knife and opt for no incision and remaining awake. Still others have heard of risk to their health from these interventions and will not let knowledge of their carotid artery disease interfere with their lives; they are content to pursue medical therapy and clinical monitoring. They may have decided that a stroke during surgery or stenting would affect them now, whereas the risk of stroke without aggressive intervention is spaced over the years ahead. Patients should be provided with information and the pros and cons of each alternative. Doctors should convey their own advice and opinion, but ultimately, the patient has the right to choose. Trial data that average out results may not heavily influence some individuals. Norman R. Hertzer, MD: Aside from our personal preferences for CE or percutaneous angioplasty/stenting, the main areas in which I disagree with Dr. Caplan regarding the management of this patient involve the extent of his initial work-up and his willingness to continue medical management even after severe carotid stenosis has been documented by noninvasive imaging. Dr. Caplan apparently would obtain a lot more tests than I feel are necessary for the evaluation of an otherwise healthy 60-year-old smoker who has had a single brief episode of TMVL. It would not cross my mind to order a sedimentation rate, a TCD examination, a head-and-neck CT/CTA, or an MRI/MRA for this patient before the duplex scan had even been performed. Once a reliable duplex scan has shown 80%-99% stenosis in the ipsilateral carotid bifurcation, the cost effectiveness of the other studies becomes very questionable. I believe they should be reserved for patients who have less than severe carotid stenosis by duplex scanning or some other specific reason for additional testing. Dr. Caplan considers “maximal medical therapy” to be a reasonable alternative to CE or angioplasty/stenting in this patient. He then states that “if medical treatment has not been optimal” (that more symptoms have occurred), “consideration should be given to adding a statin” or some other refinements in medical management “while further investigations are performed and while the patient is deciding on a preferred strategy of treatment.” Assuming for a moment that this patient, like most others, is content to follow whatever evidence-based advice is given to him, I think this is all a waste of time and merely places the patient at risk to have the stroke for which his TMVL was a warning event. His original symptom and the severity of his carotid stenosis already satisfy the criteria for intervention that have been established by several major randomized trials (NASCET (29), ACAS (32), and ACST (34)) comparing CE to best medical management. In my opinion, the patient should get on with it. Editor's Summary These two eminent physicians-a neurologist specializing in stroke and an experienced vascular surgeon-appear to agree on one fundamental point: that under the right circumstances, CE or angioplasty/stenting is indicated in this typical case of TMVL. They differ in other critical aspects. Both experts accept the idea that surgical opening of the ipsilateral cervical carotid artery will improve the patient's welfare. Dr. Caplan, the neurologist, cites the 1995 report (16) in which follow-up data from the NASCET study showed that patients with TMVL had a 16.6% 2-year risk of ipsilateral hemispheric stroke if they were managed medically. But he does not mention that not one of the TMVL patients in this study had a major stroke in the follow-up period! When confronted with the evidence from the 2001 NASCET report by Benavente et al (31) suggesting that this patient would probably not meet criteria for definite benefit of CE, Dr. Caplan points out that the study does not incorporate the new and improved methods of predicting stroke risk from carotid artery disease. Dr. Hertzer, the surgeon, acknowledges the better outcome of medically treated TMVL patients with high-grade carotid stenosis. But to perform this analysis, he believes that the NASCET cohort was excessively “parsed.” As a result, he “would be very uncomfortable using this information to withhold endarterectomy.” After agreeing on that fundamental issue, the experts diverge. Given the finding of hemodynamically significant ipsilateral carotid stenosis, Dr. Caplan does not want to make a recommendation until he knows more about the patient's medical risk factors and how they are being managed, the nature of the arterial lesion, heart, intracranial circulation, and brain parenchyma. In asking for this information, he seems to be saying that: 1) he is not utterly convinced that the impressive carotid stenosis is the cause of the TMVL; 2) he wants to fractionate the risk of stroke better; and 3) he needs to know if there is room to improve medical management. Dr. Hertzer, on the other hand, says that if you find convincing carotid stenosis with ultrasound, stop fussing and get on with the procedure. Only if carotid echography is equivocal would he extend the work-up. In considering other ophthalmic indications for carotid intervention, the debaters also seem to differ. Dr. Caplan approaches a patient with new CRAO exactly as he does a patient with new TMVL. Dr. Hertzer points out that data show that CRAO patients are much less likely to have high-grade ipsilateral carotid stenosis than are TMVL patients, but he would still recommend a carotid procedure if “no other plausible cause” were found, particularly if the patient had had a recent flurry of TMVL episodes. For asymptomatic Hollenhorst plaque, Dr. Caplan would be cautious, being “heavily guided by the nature and extent of the carotid artery lesion” and more inclined to favor surgery if “brain imaging disclosed a brain infarct in the carotid artery territory.” Dr. Hertzer apparently considers asymptomatic Hollenhorst plaque a perfectly good indication for a carotid procedure provided the plaque is new. The problem: there is usually no way to know if the plaque is new. Finally, in choosing between CE and angioplasty/stenting, the experts acknowledge that there is not yet enough information to make a meaningful decision. Both doctors suggest enrolling the patient in a randomized trial comparing the two approaches but acknowledge that this choice is often not convenient. Without that option, Dr. Caplan favors angioplasty/stenting for distal, long smooth, or tandem plaques, and CE for very focal ulcerated lesions, but he emphasizes the importance of “the availability, experience, and complication rates of the surgeons and interventionalists.” In providing a comprehensive review of the trials of CE and angioplasty, Dr. Hertzer points out that the risk of recurrent stenosis, originally a concern in angioplasty, may not apply for angioplasty plus stenting. Also, the older data on the higher peri-operative rate of stroke and other mortality after angioplasty may be obsolete if brain protective devices are used in the procedure. For this patient, he favors CE. For other opinions on this topic in this issue of the journal, see the editorial on page 259 and the viewpoints starting on page 295.