Coronary artery calcium score (CAC)
Coronary arterial calcification is part of the development of atherosclerosis, occurring almost exclusively in atherosclerotic arteries being absent in normal vessel walls [
67]. Atherosclerotic plaque proceeds through progressive stages where instability and rupture can be followed by calcification, providing stability to an unstable lesion [
68]. Coronary artery calcium score (CAC) is determined by electron-beam (EBCT) and multi-detector (MDCT) computed tomography [
69]. It has a strong correlation with the total coronary atherosclerotic burden and is able to define CHD risk, being an independent predictor of cardiovascular disease [
70,
71].
The MESA Study [
23] compared the improvement in prediction of incident CHD and CVD between six risk markers in 6814 patients, in whom 1330 were at intermediate risk, determined by the Framingham risk score (FRS) at baseline. After a 7.6-year median follow up, there were 94 CHD and 123 CVD events. Baseline accuracy ROC curve for CAC afforded the highest increment of sensitivity and specificity to the FRS compared to all other markers. CAC score has an incremental relationship with higher event rates when compared to a CAC zero score. A metanalysis [
70] evaluating the prognostic value of CAC from 4 studies in asymptomatic subjects, indicated a linear relationship between CAC and CHD events. The summary adjusted relative risk ratios for scores ranging from CAC 1–100 compared to CAC zero was 2.1 (95% CI 1.6–2.9). The RR for CAC 101–400 and CAC >400 compared to CAC zero ranged from 3.0 to 17.0, but varied significantly among studies.
In patients with diabetes, cross-sectional studies have shown higher prevalences and extent of coronary calcium compared with non-diabetic patients, with a great heterogeneity [
71‐
73]. In one series, of 155 asymptomatic individuals with diabetes, 72% had positive CAC scores and 48% had a CAC score >102
72.
Coronary arterial calcification is predictive for cardiovascular end-points in asymptomatic patient with T2DM. The PREDICT study [
74], aimed to evaluate the CAC score as a predictor of cardiovascular events in type 2 diabetes. They included 589 T2DM patients with no history of cardiovascular disease, mean age 63 years, predominantly overweight male, who had CAC score measured at baseline. Patients were followed for a median of 4 years for cardiovascular endpoints. CAC was a highly significant independent predictor of events (p < 0.001). A doubling in CAC was associated with a 32% increase in risk of events. There was a progressive increase in hazard ratio according with the CAC score level, comparing to CAC <10. Moreover, the area under the ROC curve of Framingham risk score and UKPDS-Risk Engine increased significantly with the addition of CAC score.
Coronary arterial calcification is also predictive for mortality in asymptomatic T2DM. In a large cohort study with a mean follow-up of 5 years [
13], including 10,377 asymptomatic individuals, with 903 T2DM, the mean CAC score for individuals with and without diabetes were respectively 281 ± 567 and 119 ± 341 Agatston units (p < 0.0001). The death rate was 3.5% in T2DM and 2% in non-DM respectively (p < 0.0001). The increase in mortality was proportional to increases in CAC. Interestingly, the absence of coronary calcium (CAC = 0) conferred a similar survival rate for both groups with or without diabetes [
72].
These associations were confirmed by a systematic review and metanalysis of 8 cohort studies [
12] that investigated the association of CAC with all cause mortality and cardiovascular events in T2DM. The study included 6.521 T2DM patients with a mean follow up of 5.18 years. They compared the number of events in patients with CAC above and below 10. There were a total of 802 cardiovascular events. CAC below 10 was present in 28.5% of patients. The relative risk for all-cause mortality or CV events was 5.47 (95% CI 2.59–11.53 p < 0.001). In relation to the main outcome, a CAC score >10 presented respectively a sensitivity of 94% (95% CI 89–96) and a specificity of 34% (24–44%). Because people with a CAC < 10 were 6.8 times less likely to have a cardiovascular event, the authors suggest that the negative predictive value of CAC <10 may be useful to discriminate T2DM to a lower risk category.
The long-term predictive value of CAC score for all cause mortality in asymptomatic patients with diabetes was recently addressed in a 15-year cohort study [
75]. Baseline CAC was determined in 9715 non-diabetic individuals and in 810 T2DM patients, predominantly male, with mean age of 53 years. In 34% of T2DM, baseline CAC score was zero (CAC = 0). The cumulative mortality rate over 15 years according to baseline CAC score was greater in T2DM than in non-diabetic individuals. The adjusted HR (95% CI) for mortality at 15 years was respectively: for CAC [0]: 2.53 (1.74–3.69); CAC [1–399]: 2.07 (1.64–2.62); CAC [>400]: 1.88 (1.41–2.51). Interestingly, a CAC zero conferred a similar mortality rate between T2DM and non-DM patients for the first 5 years. After 5 years, however, the risk of mortality increased significantly for diabetic patients even in the presence of a baseline CAC = 0 [
75].
Although the importance of CAC in risk stratification has increased significantly in the last years, ADA 2016 [
15] still does not recommend CAC score for routine use in risk stratification of patients with diabetes, due to still open questions in cost-effectiveness. Further studies should address this point specially taking into account the risks of excessive exposition to radiation and costs.
Carotid-wall intima-media thickness (CIMT) is the distance from the lumen-intima interface to the media-adventitia interface of the artery wall, determined by a carotid artery ultrasound [
76]. CMIT is a surrogate marker for new acute myocardial infarction and stroke in individuals above 65 years old [
77], when maximal IMT is above 1.11 mm, both in common and in internal carotid arteries. Increased CIMT above 1 mm is also predictive for CHD in younger individuals without previous cardiovascular events [
78]. The addition of maximal CMIT measurement of the internal carotid to the Framingham Risk Score (FRS) only modestly (7.6% p < 0.001) improves its accuracy for predicting cardiovascular events [
79]. In patients with T2DM, CMIT above 1.9 mm is predictive of coronary artery stenosis, improving the FRS and UKPDS risk engine scores accuracy in Japanese population [
80]. Interestingly, CMIT seems to perform better in obese than in lean T2DM patients [
81].
A carotid plaque is defined as the thickness of the intima above 1.0 mm [
82], 1.1 mm [
79] or even 1.5 mm [
83]. The total plaque area determination is a simple and highly reproducible method to quantify atherosclerosis. It improves significantly the sensitivity of FRS as a screening tool to reclassify intermediate and high risk in non-diabetic patients [
82]. In asymptomatic patients with T2DM, the sum of the maximum plaque thickness above 1.1 mm from both sides of the carotid wall carotid plaque, increases the predictive value for detecting coronary stenosis greater than 50% (obstructive CAD). This seems to be independent of age, hypertension, hyperlipidemia and HbA1c [
83].
Although promising, CIMT and carotid plaque detection are currently not recommended for routine use for risk assessment by AHA/ACC 2013 guidelines [
14,
18]. The panel considers that additional research is needed to quantify the cost effectiveness and the impact of imaging for subclinical atherosclerosis on cardiovascular risk factor management and patient outcomes.
Ankle-braquial index (ABI)
Ankle-braquial index is obtained by measuring systolic blood pressure in the supine position, in bilateral brachial arteries, dorsalis pedis arteries and in posterior tibial arteries using Doppler with a 5-mHz probe. The highest value of each blood pressure measurement is used. Most of studies have used a cut-off point of <0.90 [
84‐
87]. A low ABI score is associated with elevated cardiovascular risk. In the MESA study [
23], ABI was compared to other cardiovascular risk markers in relation to Framingham risk score. ABI was superior to Framingham Score alone and was considered an independent risk predictor of incident CHD/CVD beyond traditional risk factors. In a systematic review including 9 studies, the sensitivity and specificity of a low ABI as a predictor of future CVD events were respectively 16.5% and 92.7% for coronary heart diseases, 16.0% and 92.2%, for incident stroke and 41.0% and 87.9% for cardiovascular mortality. Thus ABI has a high specificity but a very low sensitivity, limiting its utility as a screening test for CAD [
86]. The ACC/AHA Expert Opinion guidelines [
14,
18] currently recommend ABI threshold of <0.9 for considering a patient at high risk, in the cardiovascular risk assessment of asymptomatic adults at intermediate risk.
In patients with diabetes data is less available. In a Chinese study in T2DM patients, ABI <0.9 was independently associated with high risk for all-cause mortality and CVD mortality [
88]. In that study, decreasing ABI scores below 0.9 presented a progressive association with mortality.