Assessment of Subclinical Atherosclerosis in Asymptomatic People In Vivo: Measurements Suitable for Biomarker and Mendelian Randomization Studies

Abstract

Background: One strategy to reduce the burden of cardiovascular disease is the early detection and treatment of atherosclerosis. This has led to significant interest in studies of subclinical atherosclerosis, using different phenotypes, not all of which are accurate reflections of the presence of asymptomatic atherosclerotic plaques. The aim of part 2 of this series is to provide a review of the existing literature on purported measures of subclinical disease and recommendations concerning which tests may be appropriate in the prevention of incident cardiovascular disease.

Methods: We conducted a critical review of measurements used to infer the presence of subclinical atherosclerosis in the major conduit arteries and focused on the predictive value of these tests for future cardiovascular events, independent of conventional cardiovascular risk factors, in asymptomatic people. The emphasis was on studies with >10 000 person-years of follow-up, with meta-analysis of results reporting adjusted hazard ratios (HRs) with 95% CIs. The arterial territories were limited to carotid, coronary, aorta, and lower limb arteries.

Results: In the carotid arteries, the presence of plaque (8 studies) was independently associated with future stroke (pooled HR, 1.89 [1.04-3.44]) and cardiac events (7 studies), with a pooled HR, 1.77 (1.19-2.62). Increased coronary artery calcium (5 studies) was associated with the risk of coronary heart disease events, pooled HR, 1.54 (1.07-2.07) and increasing severity of calcification (by Agaston score) was associated with escalation of risk (13 studies). An ankle/brachial index (ABI) of <0.9, the pooled HR for cardiovascular death from 7 studies was 2.01 (1.43-2.81). There were insufficient studies of either, thoracic or aortic calcium, aortic diameter, or femoral plaque to synthesize the data based on consistent reporting of these measures.

Conclusions: The presence of carotid plaque, coronary artery calcium, or abnormal ankle pressures seems to be a valid indicator of the presence of subclinical atherosclerosis and may be considered for use in biomarker, Mendelian randomization and similar studies.

Keywords: aorta; atherosclerosis; carotid arteries; myocardial infarction; phenotype.

Figure 1

Forest plots illustrating the findings of a meta-analysis of included studies. a) Association of carotid plaque presence on ultrasound with risk of stroke in 8 studies, n=23,792; including 12,635 women and 11,157 men; b) Association of carotid plaque presence on ultrasound with risk of coronary heart related events in 7 studies, n=24,115; including 12,392 women and 11,723 men.

Figure 2

Forest plots illustrating the findings for coronary artery calcification (a) meta-analysis of 4 studies for the association of CAC severity with risk of future Coronary Heart Disease Events (n=21717). Studies included in this analysis: Erbel, LaMonte, Bhatia, Hoffman. For the Heinz Nixdorf study relative risk (RR) was used. For the Heinz Nixdorf and Cooper Clinic studies, results are presented per unit of log-transformed CAC. For MESA studies, results are presented per 1-standard deviation increase in log-transformed CAC. (b) a meta-analysis of the association of category of CAC (low moderate or severe) with risk of future events Coronary Heart Disease Events. The number of studies in each category ranged from 2 (n=10190) to 4 (n=16144). For the Heinz Nixdorf study relative risk (RR) was used.

Figure 3

Forest plots illustrating the findings of a meta-analysis of included studies. a) Association of Thoracic Aortic Calcification by Agaston score with risk of coronary heart disease events in 3 studies, n=9204; including 12,635 women and 11,157 men; b) with cardiovascular events

Figure 4

Forest plots illustrating the findings of a meta-analysis of included studies with risk of future events for people with abnormally low ABI, (a) shows studies using the outcome of cardiovascular death with 7 studies n=23890 and (b) for the outcome of cardiovascular events only with 4 studies n=14395. CVD cardiovascular death, CVE cardiovascular events. The Cardiovascular Health, Edinburgh Artery, Limburg and Strong Heart studies were include in the earlier ABI collaboration.