Meta-Analysis
APR 14SillesenEuropean Journal of Preventive Cardiology
This editorial refers to ‘High-risk features, covert brain infarcts and cognitive performance: a cross-sectional analysis of the Carotid Artery Multi-modality imaging Prognostic study’, by G. Masini et al. https://doi.org/10.1093/eurjpc/zwaf778 Until recently, management of asymptomatic carotid stenosis was ‘evidence based’ from two historic trials, ACAS enrolment in the United States and published in 1995 and the European ACST1 published in 2004.1,2 Both studies included neurologically asymptomatic patients with 60% stenosis or greater, and both found that carotid endarterectomy (CEA) reduced the incidence of new, ipsilateral stroke by approximately 50%. Guidelines in Europe (ESVS) recommends CEA with a IIa (level B evidence) only in case of one or more imaging or clinical parameters associated with increased risk.3 Carotid stenting (CAS) was recommended with IIb (level B evidence). Both CEA and CAS should or may be considered, provided 30-day stroke/death rates are 3% and patient life expectancy exceeds 5 years. The European Society for Cardiology (ESC) has a 2b recommendation for invasive treatment of carotid stenosis of 60% or more, equal for CEA and CAS.4 This discrepancy is explained by the facts that the latter authors consider medical therapy stronger and more efficient to prevent events than in late 90s and the progress in CAS technology and experience leading to case-by-case decision if revascularization is considered. In Europe, the majority of cases of carotid stenosis treated are symptomatic, although there is some heterogeneity, with Northern European countries like Norway, Sweden, Denmark, and the UK, where only 0–20% of cases are asymptomatic, whereas in more Southern parts of Europe, the proportion exceeds 50%, depending also on the health insurance coverage. Very recently, the CREST-2 trial reported that intervention with CAS or CEA plus best medical treatment (BMT) reduced the risk of any stroke or death or any stroke ipsilateral in the perioperative period (surgery until 44 days post-op) or ipsilateral stroke in the remaining 4-year period, as compared to BMT alone.5 CREST-2 was a dual-arm trial testing on one side if medical treatment + CAS reduced the risk of ipsilateral stroke and compared to BMT alone, and similarly if BMT + CEA was superior to BMT alone for the same endpoints. Most notably, the 4-year ipsilateral stroke rate in both the medical arms was very low: 1.7% for medical treatment vs. the CAS arm and 1.3% vs. the CEA arm. Numbers needed to treat (NNT) with CAS to prevent one event in 4 years was 31 patients. Interestingly, to put the low stroke rate into perspective, the overall death rate was approximately the double of the ipsilateral stroke rate. Compared to the ACST1 trial,2 which also randomized asymptomatic patients with carotid stenosis to BMT + CEA or BMT only, but almost two decades earlier, the event rate in the medical arm in CREST-2 was lower (although direct comparisons are difficult due to differences in inclusions criteria and endpoints) most certainly reflecting better medical treatment nowadays. In the original ACST trial, the NNT was 20 CEAs to prevent one stroke in 5 years, and in many countries, this was found cost inefficient and therefore invasive treatment for asymptomatic carotid stenosis was in general not encouraged.2 Surely, the higher NNT in CREST-2 will not change that perception, viewing asymptomatic carotid disease primarily as a condition to be treated medically. However, considering the accumulating evidence that clearly indicates risk stratification may identify people with much higher risk than average, selective revascularization may be justified. The accompanying editorial by Brown and Bonati similarly concluded that the results of CREST-2 should not change practice with respect to primarily offer medical treatment for asymptomatic carotid stenosis, but that trials evaluating if selection based on imaging criteria (primarily IPH) would identify the few that would benefit from CAS and CEA.6 In the present issue, Masini et al. report the results of the CAMP study, a cross-sectional imaging centred study in patients referred to a tertiary vascular clinic, of which 30% had diabetes, 29% chronic coronary syndrome, and 12% lower extremity arterial disease.7 All patients had moderate carotid disease (40–60% stenosis) without neurological symptoms, and all underwent duplex ultrasound scanning including both B-mode imaging and Doppler ultrasound evaluation in addition to CT angiography for assessment of ulceration and MRA for intraplaque haemorrhage (IPH) and lipid rich necrotic core (LRNC). MRI also included identification of signs of embolism (ipsilateral non-lacunar infarcts). Among the 138 participants, 51% revealed one or more ‘high risk’ features on one of the imaging methods, most prevalent was hypoechoic plaque in 39% followed by ulceration (37%), LRNC (24%), and IPH (23%). These findings corroborate the prevalences found in the largest meta-analysis published so far, including more than 20 000 people with asymptomatic carotid stenosis, although this meta-analysis, for good reasons, had heterogeneity with respect to some parameters: inclusion criteria, test and reading methods, and study methodology.8 However, the overall important findings were that high-risk imaging features are relatively frequent and they increase the risk of an ipsilateral neurological event (both TIAs and strokes were included) by a factor 2–3 or more depending on various details. In the present, cross-sectional study, only 11% showed signs of previous relevant cerebral infarction (non-lacunar infarct), underlining the fact that not nearly all, maybe only a small proportion of people with asymptomatic carotid disease, will ever suffer an ipsilateral stroke, obviously depending on risk factor exposure and modification in addition to pharmaceutical treatment level prescribed and how it is adhered too. However, the high-risk features were associated with the non-lacunar infarcts with considerable odds ratios (with wide confidence intervals due to low numbers), also in line with meta-analyses referred to above.7,8 The CREST2 findings combined with the accumulating data showing that the degree of stenosis only informs a little about the risk of the individual, there is an urgent need for undertaking new trials randomizing people with carotid stenosis according to high-risk features such as those already summarized in current guidelines (ESVS, ESC) and presented in the study of Masini et al. Only by such studies it will be possible to identify those who truly will benefit from invasive treatment of asymptomatic carotid disease. Henrik H. Sillesen (Conceptualization, Writing—original draft [equal]) and Victor Aboyans (Conceptualization, Writing—original draft [equal]).
Editorial
SillesenEuropean Journal of Preventive Cardiology
This editorial refers to ‘High-risk features, covert brain infarcts and cognitive performance: a cross-sectional analysis of the Carotid Artery Multi-modality imaging Prognostic study’, by G. Masini et al. https://doi.org/10.1093/eurjpc/zwaf778 Until recently, management of asymptomatic carotid stenosis was ‘evidence based’ from two historic trials, ACAS enrolment in the United States and published in 1995 and the European ACST1 published in 2004.1,2 Both studies included neurologically asymptomatic patients with 60% stenosis or greater, and both found that carotid endarterectomy (CEA) reduced the incidence of new, ipsilateral stroke by approximately 50%. Guidelines in Europe (ESVS) recommends CEA with a IIa (level B evidence) only in case of one or more imaging or clinical parameters associated with increased risk.3 Carotid stenting (CAS) was recommended with IIb (level B evidence). Both CEA and CAS should or may be considered, provided 30-day stroke/death rates are 3% and patient life expectancy exceeds 5 years. The European Society for Cardiology (ESC) has a 2b recommendation for invasive treatment of carotid stenosis of 60% or more, equal for CEA and CAS.4 This discrepancy is explained by the facts that the latter authors consider medical therapy stronger and more efficient to prevent events than in late 90s and the progress in CAS technology and experience leading to case-by-case decision if revascularization is considered. In Europe, the majority of cases of carotid stenosis treated are symptomatic, although there is some heterogeneity, with Northern European countries like Norway, Sweden, Denmark, and the UK, where only 0–20% of cases are asymptomatic, whereas in more Southern parts of Europe, the proportion exceeds 50%, depending also on the health insurance coverage. Very recently, the CREST-2 trial reported that intervention with CAS or CEA plus best medical treatment (BMT) reduced the risk of any stroke or death or any stroke ipsilateral in the perioperative period (surgery until 44 days post-op) or ipsilateral stroke in the remaining 4-year period, as compared to BMT alone.5 CREST-2 was a dual-arm trial testing on one side if medical treatment + CAS reduced the risk of ipsilateral stroke and compared to BMT alone, and similarly if BMT + CEA was superior to BMT alone for the same endpoints. Most notably, the 4-year ipsilateral stroke rate in both the medical arms was very low: 1.7% for medical treatment vs. the CAS arm and 1.3% vs. the CEA arm. Numbers needed to treat (NNT) with CAS to prevent one event in 4 years was 31 patients. Interestingly, to put the low stroke rate into perspective, the overall death rate was approximately the double of the ipsilateral stroke rate. Compared to the ACST1 trial,2 which also randomized asymptomatic patients with carotid stenosis to BMT + CEA or BMT only, but almost two decades earlier, the event rate in the medical arm in CREST-2 was lower (although direct comparisons are difficult due to differences in inclusions criteria and endpoints) most certainly reflecting better medical treatment nowadays. In the original ACST trial, the NNT was 20 CEAs to prevent one stroke in 5 years, and in many countries, this was found cost inefficient and therefore invasive treatment for asymptomatic carotid stenosis was in general not encouraged.2 Surely, the higher NNT in CREST-2 will not change that perception, viewing asymptomatic carotid disease primarily as a condition to be treated medically. However, considering the accumulating evidence that clearly indicates risk stratification may identify people with much higher risk than average, selective revascularization may be justified. The accompanying editorial by Brown and Bonati similarly concluded that the results of CREST-2 should not change practice with respect to primarily offer medical treatment for asymptomatic carotid stenosis, but that trials evaluating if selection based on imaging criteria (primarily IPH) would identify the few that would benefit from CAS and CEA.6 In the present issue, Masini et al. report the results of the CAMP study, a cross-sectional imaging centred study in patients referred to a tertiary vascular clinic, of which 30% had diabetes, 29% chronic coronary syndrome, and 12% lower extremity arterial disease.7 All patients had moderate carotid disease (40–60% stenosis) without neurological symptoms, and all underwent duplex ultrasound scanning including both B-mode imaging and Doppler ultrasound evaluation in addition to CT angiography for assessment of ulceration and MRA for intraplaque haemorrhage (IPH) and lipid rich necrotic core (LRNC). MRI also included identification of signs of embolism (ipsilateral non-lacunar infarcts). Among the 138 participants, 51% revealed one or more ‘high risk’ features on one of the imaging methods, most prevalent was hypoechoic plaque in 39% followed by ulceration (37%), LRNC (24%), and IPH (23%). These findings corroborate the prevalences found in the largest meta-analysis published so far, including more than 20 000 people with asymptomatic carotid stenosis, although this meta-analysis, for good reasons, had heterogeneity with respect to some parameters: inclusion criteria, test and reading methods, and study methodology.8 However, the overall important findings were that high-risk imaging features are relatively frequent and they increase the risk of an ipsilateral neurological event (both TIAs and strokes were included) by a factor 2–3 or more depending on various details. In the present, cross-sectional study, only 11% showed signs of previous relevant cerebral infarction (non-lacunar infarct), underlining the fact that not nearly all, maybe only a small proportion of people with asymptomatic carotid disease, will ever suffer an ipsilateral stroke, obviously depending on risk factor exposure and modification in addition to pharmaceutical treatment level prescribed and how it is adhered too. However, the high-risk features were associated with the non-lacunar infarcts with considerable odds ratios (with wide confidence intervals due to low numbers), also in line with meta-analyses referred to above.7,8 The CREST2 findings combined with the accumulating data showing that the degree of stenosis only informs a little about the risk of the individual, there is an urgent need for undertaking new trials randomizing people with carotid stenosis according to high-risk features such as those already summarized in current guidelines (ESVS, ESC) and presented in the study of Masini et al. Only by such studies it will be possible to identify those who truly will benefit from invasive treatment of asymptomatic carotid disease. Henrik H. Sillesen (Conceptualization, Writing—original draft [equal]) and Victor Aboyans (Conceptualization, Writing—original draft [equal]).
