In general, good clinical practice considers screening for any disease appropriate only when an effective treatment is available. Hence the underlying assumption of a study wishing to evaluate if screening for CVD is worthwhile in terms of survival, is that an effective treatment is available and the study design should reflect this by testing screening and treatment together. This is not the case looking at the literature of screening for coronary artery disease (CAD) in people with diabetes (8,9). On the other hand, The Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial (10) and the subsequent Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial (11) represent the landmark trials in the field of stable CAD treatment. Both studies found no benefits in terms of survival of revascularization (surgical or percutaneous) over medical therapy in stable people with documented coronary artery stenosis. Of note, participants with markedly positive stress test were excluded in COURAGE. The Does coronary Atherosclerosis Deserve to be Diagnosed earlY in Diabetic patients? (DADDY) study main finding reported that screening for CAD and revascularization did not affect the occurrence of a first cardiac event in people with diabetes (12). These results are in line with the Detection of Ischemia in Asymptomatic Diabetes (DIAD), COURAGE and BARI 2D studies and confirm that to date there is no proven indication, in daily practice, to search for ischemia in people with diabetes without symptoms. However, when one is clinically suspicious of the presence of CVD, different modalities can be used to assess the presence of CAD in people with diabetes.
Role of stress testing
Exercise stress testing is useful in people with diabetes at high risk of CAD for the assessment of prognosis and the identification of individuals who may benefit from coronary artery revascularization to improve long-term survival. The most predictive clinical observation for CAD in the person with or without diabetes is a history of chest pain or discomfort, but these features will be absent in a significant proportion (20% to 50%) of individuals with diabetes (13–19). Clinical findings, such as dyspnea on exertion, resting electrocardiogram (ECG) abnormalities or multiple CVD risk factors for atherosclerosis, may also indicate the presence of CAD. Recognition of such features is of clinical importance, as the outcome of CAD events is worse in people with diabetes when shortness of breath is the primary symptom (13).
The presence of CAD risk factors and resting ECG abnormalities identify people with diabetes at increased risk of important CAD burden and abnormal stress ECG or perfusion imaging results (20). A resting ECG at the time of diagnosis of diabetes also provides a baseline to which future ECGs can be compared. In people with diabetes considered to be at high risk for CAD, a repeat resting ECG may detect changes that result from silent MI and lead to earlier detection of critical CAD. There is evidence that early screening and intervention in people with diabetes and silent ischemia is beneficial and may improve long-term survival (16,21). Screening with exercise ECG stress testing will find 3-vessel CAD in 13% to 15% of those with abnormal stress test findings (22) and lead to angiography with revascularization in 1% to 3% of asymptomatic individuals (22–24). Similar findings were reported recently in The For Asymptomatic Obstructive Coronary Artery Disease Among High-Risk Diabetic Patients Using CT Angiography, Following Core 64: A Randomized Control Study (FACTOR-64) trial. It randomized 900 participants to coronary computerized tomography angiography (CTA) screening (n=452) or standard care (n=448). CTA showed no CAD in 31%, mild stenosis in 46%, moderate in 12% and severe stenosis in 11% of the participants. Although there was no significant reduction in CHD events in this 900-person study, the authors concluded that coronary CTA screening led to more aggressive risk factor modification in 70% of participants, including improvements in statin use and more serum lipids and systemic BP (25).
The Definition of Ischemia in Asymptomatic Diabetes (DIAD) study prospectively investigated the value of routine adenosine stress myocardial perfusion scanning in asymptomatic people with type 2 diabetes ≥55 years for the prevention of coronary events (19). The baseline study showed either perfusion defects or stress-induced ECG abnormalities in 22% of participants and large defects in 6%. In this study, multiple risk factors for CAD did not help identify people with positive functional tests for CAD. Of note, baseline ECG was normal in all participants. A substantial portion of the DIAD population was defined as having intermediate/high baseline CV risk. Nevertheless, their annual CV event rate was low and not altered by routine screening for inducible ischemia. Yet, a randomized pilot study on the impact of stress testing to screen for CAD in asymptomatic subjects with diabetes suggested a significant reduction in cardiac death and MI (26). Larger and adequately powered studies are necessary to support this provocative observation before clinical practice is changed. In the Basel Asymptomatic high-Risk Diabetics' Outcome Trial, almost one-quarter of the 400 asymptomatic participants with type 2 diabetes had silent myocardial ischemia, which was associated with a worse outcome (27). The yield of myocardial perfusion imaging can be improved by selecting a higher-risk group of people with diabetes with symptoms, peripheral vascular disease (PVD), chronic kidney disease (CKD), an abnormal ECG or a high coronary artery calcium (CAC) score (e.g. >400 Agatston score) (28). The choice of initial stress test should be based on evaluation of the resting ECG, the individual's ability to exercise, and local expertise and technology. Thus, the yield of stress testing in asymptomatic people with diabetes can be improved by selecting people based on the pre-test probability of CAD. The retrospective studies that showed a high prevalence of stress test abnormalities included people with abnormal ECGs (43% with Q waves) and vascular disease (28%) (28).
Data using diverse imaging technology have been reviewed and reported recently (28), but the additional benefit of imaging on prognosis and quality of life is not clear. Studies using coronary CTA in asymptomatic people with diabetes mostly concluded that these people have a high prevalence of coronary atherosclerosis and obstructive CAD, as well as a higher prevalence of plaques with features of instability compared with subjects without diabetes. Furthermore, it is important to emphasize that a normal ECG does not offer a long-term warranty from CVD events in people with type 2 diabetes. It is the same with stress echocardiography and myocardial perfusion imaging where no events were recorded in the first 2 years of follow up among people with a normal stress echocardiography or normal nuclear scan but significantly increased thereafter (6,28–30).
People with diabetes without evidence of CAD seen on computed tomography coronary angiography have an excellent prognosis, with no cardiac events at 62-month follow up. Thus, this imaging modality can be a useful tool to reassure people with diabetes with suspected CAD regarding their outcome, with a warranty period of at least 5 years in the presence of a normal result (5,31). Of note, coronary CTA is often performed in addition to a standard diagnostic work-up. This approach may be particularly useful in specific subsets of people with diabetes with unknown CAD and equivocal or uninterpretable stress tests or in case of a discrepancy between clinical presentation and stress test results. Owing to the high prevalence of CAD, the role of coronary imaging in people with diabetes may be not to document the presence of coronary atherosclerosis but rather to identify those people with more extensive disease vs. those without any atherosclerosis. Although CT coronary angiography is able to predict the prognosis of people with diabetes on the basis of the presence/extent of CAD and plaque type, coronary imaging by computed tomography coronary angiography is not, as in case of invasive angiography, able to predict which plaque may progress to destabilization and rupture, potentially causing a clinical event.
ECG abnormalities that limit the diagnostic accuracy of a stress ECG include resting ST depression (1 mm), left bundle branch block or right bundle branch block, an intraventricular conduction defect with QRS duration >120 ms, ventricular paced rhythm or pre-excitation. Individuals with these resting ECG findings should have a stress test with an imaging modality, such as scintigraphic myocardial perfusion imaging or echocardiography. The role of other imaging modalities (anatomical imaging), such as coronary CT, calcium score, etc., in comparison to functional imaging, needs to be determined in individuals with diabetes.
Exercise stress testing can identify people with diabetes with silent ischemia; however, whether at large exercise testing results in improved outcomes in people with diabetes has not been demonstrated. The strongest and most consistent prognostic marker identified during exercise ECG stress testing is the person's maximum exercise capacity (13). Although exercise capacity is decreased in individuals with diabetes (32–34), it is still of prognostic importance (13). Silent ischemia is most likely to occur in individuals with diabetes who are older (mean age 65 years) and have elevated total cholesterol and proteinuria (23).
An ECG with ST-T abnormalities at rest has been shown to be most predictive for silent ischemia (Odds Ratio 9.27, 95% CI 4.44–19.38) and was the only significant predictor of silent ischemia in women (23). The relevance of ST-T abnormalities as a predictive factor for silent ischemia emphasizes the importance of recording a resting ECG in most individuals with type 2 diabetes. An abnormal ECG may indicate the need for further investigations and result in the earlier detection and more aggressive management of CAD (23). An abnormal exercise ECG is associated with an annual CAD event rate of 2.1%, compared with 0.97% in subjects with normal exercise ECG (26).
Myocardial ischemia (whether silent or symptomatic) detected during exercise stress testing in individuals with diabetes is associated with poorer long-term survival compared to individuals without diabetes (16). Silent MI is common (40%) in older asymptomatic individuals with type 2 diabetes, but is more frequent (65%) in those with diabetes who also have microalbuminuria (35). People with diabetes and silent ischemia have an annual event rate for CAD of 6.2% (50% of events were new-onset angina and 50% were cardiac death or MIs) (36). Thus, silent MI is a prelude not only to symptomatic ischemia, but also to potentially fatal events. Also, it has been shown in a randomized trial in people with silent ischemia (the vast majority of whom did not have diabetes) that long-term anti-ischemic drug therapy (11 years follow up) reduces cardiac events (cardiac death, nonfatal MI, acute coronary syndrome or revascularization) with preservation of ejection fraction (37). In a retrospective study analyzing 14,849 consecutive people (3,654 with diabetes and 11,195 without diabetes) undergoing a combination of exercise stress and pharmacologic stress testing (combined protocol received intravenous dipyridamole [0.56 mg/kg] infusion over 4 minutes followed shortly by symptom-limited treadmill exercise), it was observed that, despite significant perfusion defects, people with diabetes who achieve ≥5 metabolic equivalents (METs) during exercise stress single-photon emission-computed tomography (SPECT) myocardial perfusion imaging (MPI) have significantly reduced risk for future cardiac events. People with diabetes who achieved a high workload (≥10 METS) had a low annualized event rate of 0.9% (38). The importance of low exercise capacity associated to worse CVD outcomes has been also observed in a smaller study (39).
Exercise capacity is frequently impaired in people with diabetes due to the high prevalence of obesity, sedentary lifestyle, peripheral neuropathy (both sensory and motor) and unknown vascular disease in this population. Individuals who cannot adequately exercise on a stress test have a poorer prognosis than those who can, regardless of the reason for this incapacity. Perfusion imaging also provides important prognostic information. Myocardial perfusion imaging has similar predictive value for cardiac death and nonfatal MI in individuals with diabetes as in those without diabetes (40). For those unable to perform an exercise ECG stress test, pharmacologic stress imaging, using dipyridamole, adenosine or dobutamine testing, is required. Stress echocardiography and stress nuclear imaging have similar values for cardiac events in the general population (41), but no comparative data are available for the person with diabetes. In a meta-analysis of perfusion imaging, an abnormal scan was predictive of future CAD events in subjects with and without diabetes. However, the cardiac event rate in individuals with diabetes was significantly greater than in those without diabetes (41). The choice of the optimal imaging modality to detect stress-induced MI is best determined by local availability and expertise.
The utility of newer CAD diagnostic modalities, such as coronary CTA, CAC scoring and cardiac magnetic resonance imaging, is currently unknown in terms of guiding management decisions in patients with type 2 diabetes (42). Coronary CTA has emerged as a noninvasive tool for the diagnosis of CAD that enables assessment of the vascular lumen together with the arterial wall. Multidetector coronary CTA allows assessment of coronary atherosclerosis at an earlier stage compared with imaging techniques that help evaluate myocardial perfusion. The long-term prognostic value of coronary CTA in a large population of people with diabetes without chest pain syndrome was investigated (43). Coronary CTA demonstrated a high prevalence of CAD (85%), mostly non-obstructive CAD (51%). People with events were more often classified in a higher CAC-risk category but coronary CTA performed better than the CAC-score regarding the events prediction (43).
Studies have demonstrated that increased CAC in persons with diabetes is associated with increased prevalence of ischemic events and mortality and is a better predictor than the Framingham risk score (28). Also, it was reported in 392 people with type 2 diabetes that the best predictors of progression were baseline CAC score, statin use and A1C >7.0% during follow up (44). Of importance, people with diabetes but with no CAC demonstrated a survival rate similar to that of people without diabetes and no detectable calcium (5,7).