User login
POINT: Medical treatment ends need for CEA or CAS.
By Anne L. Abbott, M.D.
The medical profession and the wider community must be congratulated upon their sustained efforts over recent decades that have seen an 80% fall in the average annual risk of stroke associated with moderate and severe (50%-99%) asymptomatic carotid stenosis (ACS).1-6 This has been achieved by better medical treatment which consists of encouraging healthy lifestyle habits and appropriate use of medication. This major impact results from the combined effect of addressing all vascular risk factors in individual patients and efforts to use the best medical treatment available at the time. Rates are now so low (around 0.5% per year for ipsilateral stroke) that procedures, such as carotid endarterectomy (CEA), are now more likely to harm than help patients.7
Even if procedures were always completely risk-free, improved medical treatment may mean we have now reached the point where carotid procedures for ACS are safe but essentially ineffective for reducing stroke risk. The latest measurements of stroke risk using medical treatment alone indicate that only about 2.5% of patients with 50%-99% ACS will have a an ipsilateral stroke due to the carotid lesion during their remaining lifetime if they are receiving pretty good quality, current medical treatment alone. This is because the average age of identifying patients with ACS in past studies was about 70 years and the average survival following diagnosis was 10 years.1 Further, only about half the strokes occurring in the distribution of an internal carotid artery with >60% proximal stenosis are due to the carotid lesion.8
Guidelines recommendations for CEA for 50%-99% or 60%-99% ACS rely on marginal, 20- to 30-year-old differences in stroke rates between patients given medical treatment alone versus those given additional CEA in best practice settings.9-11 Such recommendations are not relevant to current clinical practice largely because the medical treatment used in these studies is obsolete. Multiple independent observations regarding the improved stroke prevention efficacy of medical treatment,1-6,12 and the additional observations below, provide ample evidence that current medical treatment alone is the only routine-practice (nontrial) approach we should use for patients with 50%-99% ACS and any future role of carotid procedures in these patients could only apply to very small minority subgroups:
i. Current optimal medical treatment for patients with ACS has not been defined nor its impact measured. This means that it is likely we can lower the risk of stroke and other vascular complications in patients with ACS stenosis even further than has been achieved in the most recent studies. The definition of current optimal medical treatment will vary from patient to patient depending on which vascular risk factors they have and what has been shown effective in modifying these to reduce the risk of any complications of vascular disease.
ii. The 30-day peri-operative risk of stroke or death (and other significant complications) remains above 0% in the most recent results of trials and registries13-15 and is usually not measured in routine practice. Latest measurements of average annual ipsilateral stroke risk with medical treatment alone are about 2-3 times lower than for patients who had CEA or CAS in the Asymp tomatic Carotid Atherosclerosis Study (ACAS)10 or the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST).7, 13
iii. Patients with 50%-99% ACS receiving current optimal medical treatment and with a sufficiently high average annual risk of ipsilateral stroke, indicating they may benefit from CEA, have not been identified. This rate would need to be in excess of at least 2.5%-3.0%, using results from ACAS,10 to expect any surgical benefit in routine practice. Studies of baseline degree of ACS within the 50%-99% range,10,11,16 plaque echolucency17 and most studies of detecting asymptomatic stenosis progression18-21 show that these parameters (used separately) confer a relative risk of stroke of only about 2.0-2.5. Therefore, a higher degree of baseline stenosis within the 50%-99% range, the detection of predominantly echolucent carotid plaques or asymptomatic progression are too weak on their own to identity patients likely to benefit from an additional carotid procedure. Combinations of risk markers are required for sufficient risk discrimination. For instance, results from the Asymptomatic Carotid Stenosis and Risk of Stroke Study (largest study so far of medically managed patients with moderate or severe ACS) showed that a combination of clinical features, baseline degree of stenosis and standardized ultrasonic plaque characteristics can achieve average annual ipsilateral stroke risk stratification ranging from <1.0% to 10%.16 However, like all stroke risk stratification studies performed so far, this study was performed before the era of current medical treatment and the results have not been independently tested.
iv. Even if patients with sufficiently higher than average annual risk of ipsilateral stroke are one day reliably identified, randomized trials of an additional carotid procedure will be required to determine if, and to what extent, that procedure is likely to reduce ipsilateral stroke risk in routine practice.
v. The available evidence from randomized trials and registries indicates that CAS causes about twice as many strokes or deaths as surgery (just like it does for symptomatic carotid stenosis). Therefore, currently CAS cannot be recommended.7 In conclusion, the available evidence clearly indicates that current medical treatment alone now offers the best chance of reducing the risk of ipsilateral stroke in patients with 50%-99% ACS. There is no current evidence of benefit from CEA or CAS in these patients overall, or in any particular subgroups. However, there is much evidence regarding procedural risk and unaffordable cost. Risk of ipsilateral stroke is now so low without carotid procedures it is time to shift from the historic approach of identifying ACS primarily to administer CEA. Rather, it is time to properly recognize that carotid stenosis is a risk factor of all complications of vascular disease, more than it is for ipsilateral stroke.22
The priority is to define current optimal medical treatment as best we can, recognising that patients with ACS are a risk-heterogenous population. Then we need quality independent measurements of its impact on risk of all vascular disease complications using quality prospective cohort studies. Risk stratification models should be used to identify those who may benefit from trials of more intensive medical treatment, motivational strategies, plus/minus the safest carotid procedures. If identifying patients with ACS for CEA in routine practice is to be feasible, this needs to be done within well organized environments that support patients with a wide range of stenosis severity with the primary aim of implementing current optimal medical treatment to prevent all vascular complications. Finally, it must be accepted that as medical treatment and its implementation continue to improve, the added value of carotid procedures, including for symptomatic carotid stenosis, will continue to recede until we can say, ‘good job - it is finally fixed and it is time to move on to other major health issues.’
Dr. Abbott is a neurologist and an associate professor at Monash University, Melbourne, Australia.
References
1. Stroke. 2009;40:e573-583
2. Eur J Vasc Endovasc Surg. 2009;37:625-632
3. Nat Rev Cardiol. 2011;9:116-124
4. Stroke. 2013;44
5. Management of asymptomatic carotid stenosis: Technology assessment report. 2012:83
6. Annals of Internal Medicine. 2013;158:676-685
7. Stroke. 2013;44:1186-1190
8. N Engl J Med. 2000;342:1693-1700
9. N Engl J Med. 1993;328:221-227.
10.JAMA. 1995;273:1421-1428
11. Lancet. 2004;363:1491-1502
12. Arch Neurol. 2010;67:180-186
13. N Engl J Med. 2010;363:11-23
14. J Vasc Surg. 2009;49:71-79
15. J Vasc Surg. 2011;53:307-315
16. J Vasc Surg. 2010;52:1486-1496 e1481-1485
17. Gupta A, Kesavabhotla K, Barbadaran H, Kamel H, Panda A, Giambrone A, et al. Plaque echolucency and stroke risk in asymptoamtioc carotid stenosis: A systematic review and meta-analysis. Stroke. 2014 in press.
18. J Vasc Surg. 1999;29:208-214; discussion 214-206
19. J Vasc Surg. 2013;58:128-135 e121
20 Stroke. 2013;44:792-794
21. J Vasc Surg. 2014;59:956-967
22. N Engl J Med. 1986;315:860-865.
COUNTERPOINT: Medical therapy alone is not always enough.
By Mark F. Conrad, M.D., M.MSc., and Richard P. Cambria M.D.
Stroke remains the 3rd leading cause of death in the United States.1 It is estimated that 10% to 20% of ischemic strokes can be attributed to an ipsilateral, typically high-grade carotid stenosis, thus, asymptomatic carotid bifurcation stenosis remains a potentially significant public health problem.2 Epidemiologic studies indicate that 5% to 6% of the population >65 years of age will harbor an asymptomatic and potentially surgically significant carotid stenosis.1 The modern literature linking the degree of stenosis of the internal carotid artery and risk of ipsilateral stroke dates to the natural history studies of Chambers and Norris published in the New England Journal of Medicine in 1986.3 These investigators demonstrated a significant correlation of stroke risk with a >70% ipsilateral carotid artery stenosis and progression under observation.
Carotid endarterectomy (CEA) has been the standard of care for the prevention of stroke in patients with severe (>70%) asymptomatic carotid artery stenosis (ACS) for five decades. This is supported by level 1 evidence from multiple randomized trials4, 5 and consensus guideline recommendations.2, 6 Yet the use of CEA in patients with ACS has been recently challenged by a widely publicized review article7 whose author has embarked upon an anti-CEA crusade with the zealous fervor of one who worships at the altar of the statin. In this review, 11 prospective studies of medical therapy of patients with ACS (defined by the author as >50% stenosis) were stratified by date of publication such that the four series published from 2000-2007 had a lower raw data stroke rate with medical management (.6- 1.3%) than the 1.5% stroke rate reported after CEA in the ACAS study.4,7 The authors concluded that medical management alone is “at least 3-8” times more cost effective than CEA despite a complete lack of cost data in the cited articles. In an effort to debunk this revisionist history, we will begin by addressing the four aforementioned trials and finish with a discussion of the recent literature including an observational study from our own institution.
The first study of the four followed the asymptomatic contralateral carotid artery of all patients enrolled in the North American Symptomatic Carotid Endarterectomy Trial (NASCET) for 5 years to determine the risk of stroke in medically managed patients.8 There were 2,377 patients of whom 216 (9%) had a stenosis >60% but only 113 (4.7%) had a stenosis of 75-99%. The 5-year risk of ipsilateral stroke in the 75-99% asymptomatic stenosis cohort was 18.5% or 3.7% per year (this did not include other neurologic events such as TIA or amaurosis fugax).8 However, in the meta-analysis, no stroke rate was included in the raw data column (which was the basis of the final calculated overall stroke risk) and the final estimate quoted used the 60%-99% cohort such that the annual stroke rate was reported as 3.2%.7
The Asymptomatic Carotid Stenosis and Risk of Ipsilateral Hemispheric Ischemic Events Study (ACSRS) was a multicenter study that followed 1,115 patients with ACS >50% by duplex scanning for 6-84 months (mean 37.1) with the goal of stratifying patients into cohorts of high and low risk for future neurologic events.9 They concluded that the annual stroke rate in high-risk patients was 4.3% versus 0.7% in low-risk patients. There were 453 patients with 70%-99% stenosis by NASCET criteria with a raw stroke rate of 5.7% (1.9%/ year over an average 3-year follow-up) and a 5-year ipsilateral event rate of about 18%.9 However, when the patients with 50-69% were added, the raw stroke rate decreased to 1.3%.7
The Asymptomatic Stenosis Embolus Detection (ASED) study was a prospective trial that tested the theory that transcranial Doppler embolic signal detection would identify increased risk of ipsilateral neurologic events in patients with >60% ACS.10 Of the 240 arteries studied, 115 (48%) had a stenosis of 70%-99% but 10 of these patients were censored because they underwent CEA during the follow-up period. Their outcomes were not further stratified by degree of stenosis but the average ipsilateral carotid event rate was 3.1% per year with a 1% stroke risk per year.7, 10
The final, and most damning study in the meta-analysis, included a cohort of patients from the Second Manifestations of Arterial disease (SMART) study which is a registry of patients from the Netherlands with risk factors for, or symptoms of, arterial disease.
This study attempted to determine the risk of new vascular event in patients with ACS but 996 (27% of the registry) patients with a history of cerebrovascular disease (undefined) were excluded from analysis.11 They identified 221 patients (8% of 2684 eligible patients) with ACS >50% and reported an ipsilateral stroke event rate of .6%/year.7
However, in the 147 patients with a 70-99% stenosis, the hazard ratio for ischemic stroke was 1.7 but it was not possible to separate the patients with a moderate (50-69%) stenosis from the outcomes of those with more severe (>70%) disease.11
The major flaw with this study is that the authors excluded the 996 patients with a history of cerebrovascular disease who were at highest risk of having an ACS and subsequently suffering a stroke.
The flaw with combining the results of the above studies to conclude that medical therapy is the best way to prevent stroke in patients with severe ACS is that only half (828/1,754, 47%) of the patients included had a severe (70-99%) stenosis that would warrant CEA in the United States.6, 7
The majority of the patients studied would have been treated with best medical therapy and serial Duplex scanning. Indeed, when the moderate patients were excluded, the yearly stroke rates for patients with actual severe stenosis ranged from 2.0%-3.7%; substantially higher than the 1.5% stroke rate associated with CEA.8, 9
Randomized prospective trials such as the asymptomatic carotid artery stenosis (ACAS) study published in 1995 and the more recently updated asymptomatic carotid surgery trial (ACST) indicated a quite similar annual stroke risk in the 2% range for patients treated with medical therapy as opposed to those randomized to carotid endarterectomy (this study was excluded from the meta-analysis because patients with a remote (>6 months before entry) history of neurologic events were included).4,12 In addition, the 10-year follow-up data in the ACST trial demonstrated a sustained benefit for endarterectomy over optimal medical therapy.5 It is important to emphasize that in this trial some 80% of patients were on optimal medical therapy (aspirin plus statin agents) in the later years of the trial.
These long-term data indicate that while the protective effect of endarterectomy was more pronounced in those not on appropriate lipid-lowering therapy (the 10-year advantage of endarterectomy over medical therapy in the prevention of stroke was 5.8% in patients taking a statin and 6.2% in those who were not) the protective effect of endarterectomy over optimal medical therapy was statistically significant (P=0.002).5
This is of course relevant because clinicians in the modern era have typically used data from ACST to counsel patients about the stroke risk of asymptomatic high-grade carotid stenosis.
Finally, we followed at a cohort of 115 patients with severe (>70%) ACS who did not undergo CEA for a variety of reasons.
The average follow-up was 27 months and 86% of patients were on statin therapy.
The 5-year ipsilateral ischemic event rate was 30%, and 48% of these were strokes. When stratified by degree of stenosis, the patients with a 90%-99% stenosis had an ipsilateral neurologic event rate of 55%. This is a single center experience but the stroke rate of 3% per year is consistent with contemporary series and reiterates the reality of this risk for medically treated patients.
The question remains. When the efficacy of a CEA for the prevention of stroke in patients with ACS is supported by level 1 evidence from multiple prospective, randomized trials, why would anyone who treats patients with carotid disease allow a meta-analysis of natural history studies (that redefines a severe stenosis as >50%) convince them otherwise?
We believe that the answer lies in the unsubstantiated conclusion that “current medical intervention was estimated at least 3 to 8 times more cost effective” than CEA.7
As health care in the United States continues to evolve, we are faced with the issue of how best to spend limited resource dollars.
Is it practical or cost efficient to perform preventative surgery?
In England, an annual stroke rate of 2%-3% in a small subset of the population may be acceptable but we are not ready to concede that in the United States.
For now, we will follow the SVS practice guidelines for the management of ACS as they are evidenced based and definitive.
CEA in conjunction with medical therapy remains the best way to prevent stroke in patients with severe (>70%) ACS.
Dr. Conrad is an assistant professor of surgery, Harvard Medical School, Boston. Dr. Cambria is The Robert R. Linton MD Professor of Vascular and Endovascular Surgery, Harvard Medical School, Boston.
References
1.Neurology. 2006;67:1390-1395
2.Circulation. 2012;125:188-197
3.N Engl J of Med. 1986;315:860-865
4.JAMA. 1995;273:1421-1428
5.Lancet. 2010;376:1074-1084
6.Journal of Vascular Surgery. 2011;54:e1-31
7.Stroke. 2009;40:e573-583
8.N Engl J Med. 2000;342:1693-1700
9. Eur J Vasc Endovasc Surg. 2005;30:275-284
10.Stroke. 2005;36:1128-1133
11. Stroke. 2007;38:1470-1475
12. Lancet. 2004;363:1491-1502
POINT: Medical treatment ends need for CEA or CAS.
By Anne L. Abbott, M.D.
The medical profession and the wider community must be congratulated upon their sustained efforts over recent decades that have seen an 80% fall in the average annual risk of stroke associated with moderate and severe (50%-99%) asymptomatic carotid stenosis (ACS).1-6 This has been achieved by better medical treatment which consists of encouraging healthy lifestyle habits and appropriate use of medication. This major impact results from the combined effect of addressing all vascular risk factors in individual patients and efforts to use the best medical treatment available at the time. Rates are now so low (around 0.5% per year for ipsilateral stroke) that procedures, such as carotid endarterectomy (CEA), are now more likely to harm than help patients.7
Even if procedures were always completely risk-free, improved medical treatment may mean we have now reached the point where carotid procedures for ACS are safe but essentially ineffective for reducing stroke risk. The latest measurements of stroke risk using medical treatment alone indicate that only about 2.5% of patients with 50%-99% ACS will have a an ipsilateral stroke due to the carotid lesion during their remaining lifetime if they are receiving pretty good quality, current medical treatment alone. This is because the average age of identifying patients with ACS in past studies was about 70 years and the average survival following diagnosis was 10 years.1 Further, only about half the strokes occurring in the distribution of an internal carotid artery with >60% proximal stenosis are due to the carotid lesion.8
Guidelines recommendations for CEA for 50%-99% or 60%-99% ACS rely on marginal, 20- to 30-year-old differences in stroke rates between patients given medical treatment alone versus those given additional CEA in best practice settings.9-11 Such recommendations are not relevant to current clinical practice largely because the medical treatment used in these studies is obsolete. Multiple independent observations regarding the improved stroke prevention efficacy of medical treatment,1-6,12 and the additional observations below, provide ample evidence that current medical treatment alone is the only routine-practice (nontrial) approach we should use for patients with 50%-99% ACS and any future role of carotid procedures in these patients could only apply to very small minority subgroups:
i. Current optimal medical treatment for patients with ACS has not been defined nor its impact measured. This means that it is likely we can lower the risk of stroke and other vascular complications in patients with ACS stenosis even further than has been achieved in the most recent studies. The definition of current optimal medical treatment will vary from patient to patient depending on which vascular risk factors they have and what has been shown effective in modifying these to reduce the risk of any complications of vascular disease.
ii. The 30-day peri-operative risk of stroke or death (and other significant complications) remains above 0% in the most recent results of trials and registries13-15 and is usually not measured in routine practice. Latest measurements of average annual ipsilateral stroke risk with medical treatment alone are about 2-3 times lower than for patients who had CEA or CAS in the Asymp tomatic Carotid Atherosclerosis Study (ACAS)10 or the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST).7, 13
iii. Patients with 50%-99% ACS receiving current optimal medical treatment and with a sufficiently high average annual risk of ipsilateral stroke, indicating they may benefit from CEA, have not been identified. This rate would need to be in excess of at least 2.5%-3.0%, using results from ACAS,10 to expect any surgical benefit in routine practice. Studies of baseline degree of ACS within the 50%-99% range,10,11,16 plaque echolucency17 and most studies of detecting asymptomatic stenosis progression18-21 show that these parameters (used separately) confer a relative risk of stroke of only about 2.0-2.5. Therefore, a higher degree of baseline stenosis within the 50%-99% range, the detection of predominantly echolucent carotid plaques or asymptomatic progression are too weak on their own to identity patients likely to benefit from an additional carotid procedure. Combinations of risk markers are required for sufficient risk discrimination. For instance, results from the Asymptomatic Carotid Stenosis and Risk of Stroke Study (largest study so far of medically managed patients with moderate or severe ACS) showed that a combination of clinical features, baseline degree of stenosis and standardized ultrasonic plaque characteristics can achieve average annual ipsilateral stroke risk stratification ranging from <1.0% to 10%.16 However, like all stroke risk stratification studies performed so far, this study was performed before the era of current medical treatment and the results have not been independently tested.
iv. Even if patients with sufficiently higher than average annual risk of ipsilateral stroke are one day reliably identified, randomized trials of an additional carotid procedure will be required to determine if, and to what extent, that procedure is likely to reduce ipsilateral stroke risk in routine practice.
v. The available evidence from randomized trials and registries indicates that CAS causes about twice as many strokes or deaths as surgery (just like it does for symptomatic carotid stenosis). Therefore, currently CAS cannot be recommended.7 In conclusion, the available evidence clearly indicates that current medical treatment alone now offers the best chance of reducing the risk of ipsilateral stroke in patients with 50%-99% ACS. There is no current evidence of benefit from CEA or CAS in these patients overall, or in any particular subgroups. However, there is much evidence regarding procedural risk and unaffordable cost. Risk of ipsilateral stroke is now so low without carotid procedures it is time to shift from the historic approach of identifying ACS primarily to administer CEA. Rather, it is time to properly recognize that carotid stenosis is a risk factor of all complications of vascular disease, more than it is for ipsilateral stroke.22
The priority is to define current optimal medical treatment as best we can, recognising that patients with ACS are a risk-heterogenous population. Then we need quality independent measurements of its impact on risk of all vascular disease complications using quality prospective cohort studies. Risk stratification models should be used to identify those who may benefit from trials of more intensive medical treatment, motivational strategies, plus/minus the safest carotid procedures. If identifying patients with ACS for CEA in routine practice is to be feasible, this needs to be done within well organized environments that support patients with a wide range of stenosis severity with the primary aim of implementing current optimal medical treatment to prevent all vascular complications. Finally, it must be accepted that as medical treatment and its implementation continue to improve, the added value of carotid procedures, including for symptomatic carotid stenosis, will continue to recede until we can say, ‘good job - it is finally fixed and it is time to move on to other major health issues.’
Dr. Abbott is a neurologist and an associate professor at Monash University, Melbourne, Australia.
References
1. Stroke. 2009;40:e573-583
2. Eur J Vasc Endovasc Surg. 2009;37:625-632
3. Nat Rev Cardiol. 2011;9:116-124
4. Stroke. 2013;44
5. Management of asymptomatic carotid stenosis: Technology assessment report. 2012:83
6. Annals of Internal Medicine. 2013;158:676-685
7. Stroke. 2013;44:1186-1190
8. N Engl J Med. 2000;342:1693-1700
9. N Engl J Med. 1993;328:221-227.
10.JAMA. 1995;273:1421-1428
11. Lancet. 2004;363:1491-1502
12. Arch Neurol. 2010;67:180-186
13. N Engl J Med. 2010;363:11-23
14. J Vasc Surg. 2009;49:71-79
15. J Vasc Surg. 2011;53:307-315
16. J Vasc Surg. 2010;52:1486-1496 e1481-1485
17. Gupta A, Kesavabhotla K, Barbadaran H, Kamel H, Panda A, Giambrone A, et al. Plaque echolucency and stroke risk in asymptoamtioc carotid stenosis: A systematic review and meta-analysis. Stroke. 2014 in press.
18. J Vasc Surg. 1999;29:208-214; discussion 214-206
19. J Vasc Surg. 2013;58:128-135 e121
20 Stroke. 2013;44:792-794
21. J Vasc Surg. 2014;59:956-967
22. N Engl J Med. 1986;315:860-865.
COUNTERPOINT: Medical therapy alone is not always enough.
By Mark F. Conrad, M.D., M.MSc., and Richard P. Cambria M.D.
Stroke remains the 3rd leading cause of death in the United States.1 It is estimated that 10% to 20% of ischemic strokes can be attributed to an ipsilateral, typically high-grade carotid stenosis, thus, asymptomatic carotid bifurcation stenosis remains a potentially significant public health problem.2 Epidemiologic studies indicate that 5% to 6% of the population >65 years of age will harbor an asymptomatic and potentially surgically significant carotid stenosis.1 The modern literature linking the degree of stenosis of the internal carotid artery and risk of ipsilateral stroke dates to the natural history studies of Chambers and Norris published in the New England Journal of Medicine in 1986.3 These investigators demonstrated a significant correlation of stroke risk with a >70% ipsilateral carotid artery stenosis and progression under observation.
Carotid endarterectomy (CEA) has been the standard of care for the prevention of stroke in patients with severe (>70%) asymptomatic carotid artery stenosis (ACS) for five decades. This is supported by level 1 evidence from multiple randomized trials4, 5 and consensus guideline recommendations.2, 6 Yet the use of CEA in patients with ACS has been recently challenged by a widely publicized review article7 whose author has embarked upon an anti-CEA crusade with the zealous fervor of one who worships at the altar of the statin. In this review, 11 prospective studies of medical therapy of patients with ACS (defined by the author as >50% stenosis) were stratified by date of publication such that the four series published from 2000-2007 had a lower raw data stroke rate with medical management (.6- 1.3%) than the 1.5% stroke rate reported after CEA in the ACAS study.4,7 The authors concluded that medical management alone is “at least 3-8” times more cost effective than CEA despite a complete lack of cost data in the cited articles. In an effort to debunk this revisionist history, we will begin by addressing the four aforementioned trials and finish with a discussion of the recent literature including an observational study from our own institution.
The first study of the four followed the asymptomatic contralateral carotid artery of all patients enrolled in the North American Symptomatic Carotid Endarterectomy Trial (NASCET) for 5 years to determine the risk of stroke in medically managed patients.8 There were 2,377 patients of whom 216 (9%) had a stenosis >60% but only 113 (4.7%) had a stenosis of 75-99%. The 5-year risk of ipsilateral stroke in the 75-99% asymptomatic stenosis cohort was 18.5% or 3.7% per year (this did not include other neurologic events such as TIA or amaurosis fugax).8 However, in the meta-analysis, no stroke rate was included in the raw data column (which was the basis of the final calculated overall stroke risk) and the final estimate quoted used the 60%-99% cohort such that the annual stroke rate was reported as 3.2%.7
The Asymptomatic Carotid Stenosis and Risk of Ipsilateral Hemispheric Ischemic Events Study (ACSRS) was a multicenter study that followed 1,115 patients with ACS >50% by duplex scanning for 6-84 months (mean 37.1) with the goal of stratifying patients into cohorts of high and low risk for future neurologic events.9 They concluded that the annual stroke rate in high-risk patients was 4.3% versus 0.7% in low-risk patients. There were 453 patients with 70%-99% stenosis by NASCET criteria with a raw stroke rate of 5.7% (1.9%/ year over an average 3-year follow-up) and a 5-year ipsilateral event rate of about 18%.9 However, when the patients with 50-69% were added, the raw stroke rate decreased to 1.3%.7
The Asymptomatic Stenosis Embolus Detection (ASED) study was a prospective trial that tested the theory that transcranial Doppler embolic signal detection would identify increased risk of ipsilateral neurologic events in patients with >60% ACS.10 Of the 240 arteries studied, 115 (48%) had a stenosis of 70%-99% but 10 of these patients were censored because they underwent CEA during the follow-up period. Their outcomes were not further stratified by degree of stenosis but the average ipsilateral carotid event rate was 3.1% per year with a 1% stroke risk per year.7, 10
The final, and most damning study in the meta-analysis, included a cohort of patients from the Second Manifestations of Arterial disease (SMART) study which is a registry of patients from the Netherlands with risk factors for, or symptoms of, arterial disease.
This study attempted to determine the risk of new vascular event in patients with ACS but 996 (27% of the registry) patients with a history of cerebrovascular disease (undefined) were excluded from analysis.11 They identified 221 patients (8% of 2684 eligible patients) with ACS >50% and reported an ipsilateral stroke event rate of .6%/year.7
However, in the 147 patients with a 70-99% stenosis, the hazard ratio for ischemic stroke was 1.7 but it was not possible to separate the patients with a moderate (50-69%) stenosis from the outcomes of those with more severe (>70%) disease.11
The major flaw with this study is that the authors excluded the 996 patients with a history of cerebrovascular disease who were at highest risk of having an ACS and subsequently suffering a stroke.
The flaw with combining the results of the above studies to conclude that medical therapy is the best way to prevent stroke in patients with severe ACS is that only half (828/1,754, 47%) of the patients included had a severe (70-99%) stenosis that would warrant CEA in the United States.6, 7
The majority of the patients studied would have been treated with best medical therapy and serial Duplex scanning. Indeed, when the moderate patients were excluded, the yearly stroke rates for patients with actual severe stenosis ranged from 2.0%-3.7%; substantially higher than the 1.5% stroke rate associated with CEA.8, 9
Randomized prospective trials such as the asymptomatic carotid artery stenosis (ACAS) study published in 1995 and the more recently updated asymptomatic carotid surgery trial (ACST) indicated a quite similar annual stroke risk in the 2% range for patients treated with medical therapy as opposed to those randomized to carotid endarterectomy (this study was excluded from the meta-analysis because patients with a remote (>6 months before entry) history of neurologic events were included).4,12 In addition, the 10-year follow-up data in the ACST trial demonstrated a sustained benefit for endarterectomy over optimal medical therapy.5 It is important to emphasize that in this trial some 80% of patients were on optimal medical therapy (aspirin plus statin agents) in the later years of the trial.
These long-term data indicate that while the protective effect of endarterectomy was more pronounced in those not on appropriate lipid-lowering therapy (the 10-year advantage of endarterectomy over medical therapy in the prevention of stroke was 5.8% in patients taking a statin and 6.2% in those who were not) the protective effect of endarterectomy over optimal medical therapy was statistically significant (P=0.002).5
This is of course relevant because clinicians in the modern era have typically used data from ACST to counsel patients about the stroke risk of asymptomatic high-grade carotid stenosis.
Finally, we followed at a cohort of 115 patients with severe (>70%) ACS who did not undergo CEA for a variety of reasons.
The average follow-up was 27 months and 86% of patients were on statin therapy.
The 5-year ipsilateral ischemic event rate was 30%, and 48% of these were strokes. When stratified by degree of stenosis, the patients with a 90%-99% stenosis had an ipsilateral neurologic event rate of 55%. This is a single center experience but the stroke rate of 3% per year is consistent with contemporary series and reiterates the reality of this risk for medically treated patients.
The question remains. When the efficacy of a CEA for the prevention of stroke in patients with ACS is supported by level 1 evidence from multiple prospective, randomized trials, why would anyone who treats patients with carotid disease allow a meta-analysis of natural history studies (that redefines a severe stenosis as >50%) convince them otherwise?
We believe that the answer lies in the unsubstantiated conclusion that “current medical intervention was estimated at least 3 to 8 times more cost effective” than CEA.7
As health care in the United States continues to evolve, we are faced with the issue of how best to spend limited resource dollars.
Is it practical or cost efficient to perform preventative surgery?
In England, an annual stroke rate of 2%-3% in a small subset of the population may be acceptable but we are not ready to concede that in the United States.
For now, we will follow the SVS practice guidelines for the management of ACS as they are evidenced based and definitive.
CEA in conjunction with medical therapy remains the best way to prevent stroke in patients with severe (>70%) ACS.
Dr. Conrad is an assistant professor of surgery, Harvard Medical School, Boston. Dr. Cambria is The Robert R. Linton MD Professor of Vascular and Endovascular Surgery, Harvard Medical School, Boston.
References
1.Neurology. 2006;67:1390-1395
2.Circulation. 2012;125:188-197
3.N Engl J of Med. 1986;315:860-865
4.JAMA. 1995;273:1421-1428
5.Lancet. 2010;376:1074-1084
6.Journal of Vascular Surgery. 2011;54:e1-31
7.Stroke. 2009;40:e573-583
8.N Engl J Med. 2000;342:1693-1700
9. Eur J Vasc Endovasc Surg. 2005;30:275-284
10.Stroke. 2005;36:1128-1133
11. Stroke. 2007;38:1470-1475
12. Lancet. 2004;363:1491-1502
POINT: Medical treatment ends need for CEA or CAS.
By Anne L. Abbott, M.D.
The medical profession and the wider community must be congratulated upon their sustained efforts over recent decades that have seen an 80% fall in the average annual risk of stroke associated with moderate and severe (50%-99%) asymptomatic carotid stenosis (ACS).1-6 This has been achieved by better medical treatment which consists of encouraging healthy lifestyle habits and appropriate use of medication. This major impact results from the combined effect of addressing all vascular risk factors in individual patients and efforts to use the best medical treatment available at the time. Rates are now so low (around 0.5% per year for ipsilateral stroke) that procedures, such as carotid endarterectomy (CEA), are now more likely to harm than help patients.7
Even if procedures were always completely risk-free, improved medical treatment may mean we have now reached the point where carotid procedures for ACS are safe but essentially ineffective for reducing stroke risk. The latest measurements of stroke risk using medical treatment alone indicate that only about 2.5% of patients with 50%-99% ACS will have a an ipsilateral stroke due to the carotid lesion during their remaining lifetime if they are receiving pretty good quality, current medical treatment alone. This is because the average age of identifying patients with ACS in past studies was about 70 years and the average survival following diagnosis was 10 years.1 Further, only about half the strokes occurring in the distribution of an internal carotid artery with >60% proximal stenosis are due to the carotid lesion.8
Guidelines recommendations for CEA for 50%-99% or 60%-99% ACS rely on marginal, 20- to 30-year-old differences in stroke rates between patients given medical treatment alone versus those given additional CEA in best practice settings.9-11 Such recommendations are not relevant to current clinical practice largely because the medical treatment used in these studies is obsolete. Multiple independent observations regarding the improved stroke prevention efficacy of medical treatment,1-6,12 and the additional observations below, provide ample evidence that current medical treatment alone is the only routine-practice (nontrial) approach we should use for patients with 50%-99% ACS and any future role of carotid procedures in these patients could only apply to very small minority subgroups:
i. Current optimal medical treatment for patients with ACS has not been defined nor its impact measured. This means that it is likely we can lower the risk of stroke and other vascular complications in patients with ACS stenosis even further than has been achieved in the most recent studies. The definition of current optimal medical treatment will vary from patient to patient depending on which vascular risk factors they have and what has been shown effective in modifying these to reduce the risk of any complications of vascular disease.
ii. The 30-day peri-operative risk of stroke or death (and other significant complications) remains above 0% in the most recent results of trials and registries13-15 and is usually not measured in routine practice. Latest measurements of average annual ipsilateral stroke risk with medical treatment alone are about 2-3 times lower than for patients who had CEA or CAS in the Asymp tomatic Carotid Atherosclerosis Study (ACAS)10 or the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST).7, 13
iii. Patients with 50%-99% ACS receiving current optimal medical treatment and with a sufficiently high average annual risk of ipsilateral stroke, indicating they may benefit from CEA, have not been identified. This rate would need to be in excess of at least 2.5%-3.0%, using results from ACAS,10 to expect any surgical benefit in routine practice. Studies of baseline degree of ACS within the 50%-99% range,10,11,16 plaque echolucency17 and most studies of detecting asymptomatic stenosis progression18-21 show that these parameters (used separately) confer a relative risk of stroke of only about 2.0-2.5. Therefore, a higher degree of baseline stenosis within the 50%-99% range, the detection of predominantly echolucent carotid plaques or asymptomatic progression are too weak on their own to identity patients likely to benefit from an additional carotid procedure. Combinations of risk markers are required for sufficient risk discrimination. For instance, results from the Asymptomatic Carotid Stenosis and Risk of Stroke Study (largest study so far of medically managed patients with moderate or severe ACS) showed that a combination of clinical features, baseline degree of stenosis and standardized ultrasonic plaque characteristics can achieve average annual ipsilateral stroke risk stratification ranging from <1.0% to 10%.16 However, like all stroke risk stratification studies performed so far, this study was performed before the era of current medical treatment and the results have not been independently tested.
iv. Even if patients with sufficiently higher than average annual risk of ipsilateral stroke are one day reliably identified, randomized trials of an additional carotid procedure will be required to determine if, and to what extent, that procedure is likely to reduce ipsilateral stroke risk in routine practice.
v. The available evidence from randomized trials and registries indicates that CAS causes about twice as many strokes or deaths as surgery (just like it does for symptomatic carotid stenosis). Therefore, currently CAS cannot be recommended.7 In conclusion, the available evidence clearly indicates that current medical treatment alone now offers the best chance of reducing the risk of ipsilateral stroke in patients with 50%-99% ACS. There is no current evidence of benefit from CEA or CAS in these patients overall, or in any particular subgroups. However, there is much evidence regarding procedural risk and unaffordable cost. Risk of ipsilateral stroke is now so low without carotid procedures it is time to shift from the historic approach of identifying ACS primarily to administer CEA. Rather, it is time to properly recognize that carotid stenosis is a risk factor of all complications of vascular disease, more than it is for ipsilateral stroke.22
The priority is to define current optimal medical treatment as best we can, recognising that patients with ACS are a risk-heterogenous population. Then we need quality independent measurements of its impact on risk of all vascular disease complications using quality prospective cohort studies. Risk stratification models should be used to identify those who may benefit from trials of more intensive medical treatment, motivational strategies, plus/minus the safest carotid procedures. If identifying patients with ACS for CEA in routine practice is to be feasible, this needs to be done within well organized environments that support patients with a wide range of stenosis severity with the primary aim of implementing current optimal medical treatment to prevent all vascular complications. Finally, it must be accepted that as medical treatment and its implementation continue to improve, the added value of carotid procedures, including for symptomatic carotid stenosis, will continue to recede until we can say, ‘good job - it is finally fixed and it is time to move on to other major health issues.’
Dr. Abbott is a neurologist and an associate professor at Monash University, Melbourne, Australia.
References
1. Stroke. 2009;40:e573-583
2. Eur J Vasc Endovasc Surg. 2009;37:625-632
3. Nat Rev Cardiol. 2011;9:116-124
4. Stroke. 2013;44
5. Management of asymptomatic carotid stenosis: Technology assessment report. 2012:83
6. Annals of Internal Medicine. 2013;158:676-685
7. Stroke. 2013;44:1186-1190
8. N Engl J Med. 2000;342:1693-1700
9. N Engl J Med. 1993;328:221-227.
10.JAMA. 1995;273:1421-1428
11. Lancet. 2004;363:1491-1502
12. Arch Neurol. 2010;67:180-186
13. N Engl J Med. 2010;363:11-23
14. J Vasc Surg. 2009;49:71-79
15. J Vasc Surg. 2011;53:307-315
16. J Vasc Surg. 2010;52:1486-1496 e1481-1485
17. Gupta A, Kesavabhotla K, Barbadaran H, Kamel H, Panda A, Giambrone A, et al. Plaque echolucency and stroke risk in asymptoamtioc carotid stenosis: A systematic review and meta-analysis. Stroke. 2014 in press.
18. J Vasc Surg. 1999;29:208-214; discussion 214-206
19. J Vasc Surg. 2013;58:128-135 e121
20 Stroke. 2013;44:792-794
21. J Vasc Surg. 2014;59:956-967
22. N Engl J Med. 1986;315:860-865.
COUNTERPOINT: Medical therapy alone is not always enough.
By Mark F. Conrad, M.D., M.MSc., and Richard P. Cambria M.D.
Stroke remains the 3rd leading cause of death in the United States.1 It is estimated that 10% to 20% of ischemic strokes can be attributed to an ipsilateral, typically high-grade carotid stenosis, thus, asymptomatic carotid bifurcation stenosis remains a potentially significant public health problem.2 Epidemiologic studies indicate that 5% to 6% of the population >65 years of age will harbor an asymptomatic and potentially surgically significant carotid stenosis.1 The modern literature linking the degree of stenosis of the internal carotid artery and risk of ipsilateral stroke dates to the natural history studies of Chambers and Norris published in the New England Journal of Medicine in 1986.3 These investigators demonstrated a significant correlation of stroke risk with a >70% ipsilateral carotid artery stenosis and progression under observation.
Carotid endarterectomy (CEA) has been the standard of care for the prevention of stroke in patients with severe (>70%) asymptomatic carotid artery stenosis (ACS) for five decades. This is supported by level 1 evidence from multiple randomized trials4, 5 and consensus guideline recommendations.2, 6 Yet the use of CEA in patients with ACS has been recently challenged by a widely publicized review article7 whose author has embarked upon an anti-CEA crusade with the zealous fervor of one who worships at the altar of the statin. In this review, 11 prospective studies of medical therapy of patients with ACS (defined by the author as >50% stenosis) were stratified by date of publication such that the four series published from 2000-2007 had a lower raw data stroke rate with medical management (.6- 1.3%) than the 1.5% stroke rate reported after CEA in the ACAS study.4,7 The authors concluded that medical management alone is “at least 3-8” times more cost effective than CEA despite a complete lack of cost data in the cited articles. In an effort to debunk this revisionist history, we will begin by addressing the four aforementioned trials and finish with a discussion of the recent literature including an observational study from our own institution.
The first study of the four followed the asymptomatic contralateral carotid artery of all patients enrolled in the North American Symptomatic Carotid Endarterectomy Trial (NASCET) for 5 years to determine the risk of stroke in medically managed patients.8 There were 2,377 patients of whom 216 (9%) had a stenosis >60% but only 113 (4.7%) had a stenosis of 75-99%. The 5-year risk of ipsilateral stroke in the 75-99% asymptomatic stenosis cohort was 18.5% or 3.7% per year (this did not include other neurologic events such as TIA or amaurosis fugax).8 However, in the meta-analysis, no stroke rate was included in the raw data column (which was the basis of the final calculated overall stroke risk) and the final estimate quoted used the 60%-99% cohort such that the annual stroke rate was reported as 3.2%.7
The Asymptomatic Carotid Stenosis and Risk of Ipsilateral Hemispheric Ischemic Events Study (ACSRS) was a multicenter study that followed 1,115 patients with ACS >50% by duplex scanning for 6-84 months (mean 37.1) with the goal of stratifying patients into cohorts of high and low risk for future neurologic events.9 They concluded that the annual stroke rate in high-risk patients was 4.3% versus 0.7% in low-risk patients. There were 453 patients with 70%-99% stenosis by NASCET criteria with a raw stroke rate of 5.7% (1.9%/ year over an average 3-year follow-up) and a 5-year ipsilateral event rate of about 18%.9 However, when the patients with 50-69% were added, the raw stroke rate decreased to 1.3%.7
The Asymptomatic Stenosis Embolus Detection (ASED) study was a prospective trial that tested the theory that transcranial Doppler embolic signal detection would identify increased risk of ipsilateral neurologic events in patients with >60% ACS.10 Of the 240 arteries studied, 115 (48%) had a stenosis of 70%-99% but 10 of these patients were censored because they underwent CEA during the follow-up period. Their outcomes were not further stratified by degree of stenosis but the average ipsilateral carotid event rate was 3.1% per year with a 1% stroke risk per year.7, 10
The final, and most damning study in the meta-analysis, included a cohort of patients from the Second Manifestations of Arterial disease (SMART) study which is a registry of patients from the Netherlands with risk factors for, or symptoms of, arterial disease.
This study attempted to determine the risk of new vascular event in patients with ACS but 996 (27% of the registry) patients with a history of cerebrovascular disease (undefined) were excluded from analysis.11 They identified 221 patients (8% of 2684 eligible patients) with ACS >50% and reported an ipsilateral stroke event rate of .6%/year.7
However, in the 147 patients with a 70-99% stenosis, the hazard ratio for ischemic stroke was 1.7 but it was not possible to separate the patients with a moderate (50-69%) stenosis from the outcomes of those with more severe (>70%) disease.11
The major flaw with this study is that the authors excluded the 996 patients with a history of cerebrovascular disease who were at highest risk of having an ACS and subsequently suffering a stroke.
The flaw with combining the results of the above studies to conclude that medical therapy is the best way to prevent stroke in patients with severe ACS is that only half (828/1,754, 47%) of the patients included had a severe (70-99%) stenosis that would warrant CEA in the United States.6, 7
The majority of the patients studied would have been treated with best medical therapy and serial Duplex scanning. Indeed, when the moderate patients were excluded, the yearly stroke rates for patients with actual severe stenosis ranged from 2.0%-3.7%; substantially higher than the 1.5% stroke rate associated with CEA.8, 9
Randomized prospective trials such as the asymptomatic carotid artery stenosis (ACAS) study published in 1995 and the more recently updated asymptomatic carotid surgery trial (ACST) indicated a quite similar annual stroke risk in the 2% range for patients treated with medical therapy as opposed to those randomized to carotid endarterectomy (this study was excluded from the meta-analysis because patients with a remote (>6 months before entry) history of neurologic events were included).4,12 In addition, the 10-year follow-up data in the ACST trial demonstrated a sustained benefit for endarterectomy over optimal medical therapy.5 It is important to emphasize that in this trial some 80% of patients were on optimal medical therapy (aspirin plus statin agents) in the later years of the trial.
These long-term data indicate that while the protective effect of endarterectomy was more pronounced in those not on appropriate lipid-lowering therapy (the 10-year advantage of endarterectomy over medical therapy in the prevention of stroke was 5.8% in patients taking a statin and 6.2% in those who were not) the protective effect of endarterectomy over optimal medical therapy was statistically significant (P=0.002).5
This is of course relevant because clinicians in the modern era have typically used data from ACST to counsel patients about the stroke risk of asymptomatic high-grade carotid stenosis.
Finally, we followed at a cohort of 115 patients with severe (>70%) ACS who did not undergo CEA for a variety of reasons.
The average follow-up was 27 months and 86% of patients were on statin therapy.
The 5-year ipsilateral ischemic event rate was 30%, and 48% of these were strokes. When stratified by degree of stenosis, the patients with a 90%-99% stenosis had an ipsilateral neurologic event rate of 55%. This is a single center experience but the stroke rate of 3% per year is consistent with contemporary series and reiterates the reality of this risk for medically treated patients.
The question remains. When the efficacy of a CEA for the prevention of stroke in patients with ACS is supported by level 1 evidence from multiple prospective, randomized trials, why would anyone who treats patients with carotid disease allow a meta-analysis of natural history studies (that redefines a severe stenosis as >50%) convince them otherwise?
We believe that the answer lies in the unsubstantiated conclusion that “current medical intervention was estimated at least 3 to 8 times more cost effective” than CEA.7
As health care in the United States continues to evolve, we are faced with the issue of how best to spend limited resource dollars.
Is it practical or cost efficient to perform preventative surgery?
In England, an annual stroke rate of 2%-3% in a small subset of the population may be acceptable but we are not ready to concede that in the United States.
For now, we will follow the SVS practice guidelines for the management of ACS as they are evidenced based and definitive.
CEA in conjunction with medical therapy remains the best way to prevent stroke in patients with severe (>70%) ACS.
Dr. Conrad is an assistant professor of surgery, Harvard Medical School, Boston. Dr. Cambria is The Robert R. Linton MD Professor of Vascular and Endovascular Surgery, Harvard Medical School, Boston.
References
1.Neurology. 2006;67:1390-1395
2.Circulation. 2012;125:188-197
3.N Engl J of Med. 1986;315:860-865
4.JAMA. 1995;273:1421-1428
5.Lancet. 2010;376:1074-1084
6.Journal of Vascular Surgery. 2011;54:e1-31
7.Stroke. 2009;40:e573-583
8.N Engl J Med. 2000;342:1693-1700
9. Eur J Vasc Endovasc Surg. 2005;30:275-284
10.Stroke. 2005;36:1128-1133
11. Stroke. 2007;38:1470-1475
12. Lancet. 2004;363:1491-1502
