In patients with diabetes, the utility of diagnostic screening cardiac tests in subjects without clinical coronary artery disease remains controversial. The aim of this study was to assess the prognostic meaning of dual-imaging stress echocardiography (conventional wall motion analysis and Doppler-derived coronary flow velocity reserve [CFVR] of the left anterior descending coronary artery) in high-risk asymptomatic individuals with diabetes.
This was a prospective analysis of 230 asymptomatic patients with diabetes (128 men; mean age, 66 ± 9 years) with no clinical evidence of coronary artery disease, no Q waves or deep negative waves on the electrocardiogram, and no wall motion abnormalities on resting echocardiography. Of these subjects, 147 (64%) had target organ damage and 83 (36%) had two or more associated cardiovascular risk factors. All patients underwent dipyridamole stress echocardiography with CFVR assessment of the left anterior descending coronary artery by transthoracic Doppler, and test results were entered into a database at the time of testing for a clinical and outcome follow-up (mean, 4.6 ± 2.7 years).
Inducible ischemia and reduced CFVR (≤2) were detected in six and 52 patients, respectively. A total of 54 subjects (23%) had abnormal test results (ischemia or reduced CFVR). During follow-up, 39 major adverse cardiac events (MACEs) occurred: 22 hard events (18 deaths and four nonfatal myocardial infarctions) and 17 coronary revascularizations. The yearly incidence rates of hard events and MACEs in the entire study population were 2.1% and 3.6%, respectively. Abnormal test results were the only multivariate indicator of both hard events (hazard ratio, 3.69; 95% CI, 1.54–8.80) and MACEs (hazard ratio, 6.12; 95% CI, 3.22–11.62).
Abnormal test results were obtained in one of four cases and were a strong and independent predictor of future hard events and MACEs.
Diabetes mellitus is responsible for 21% of all deaths in the United States, and coronary artery disease (CAD) is the leading cause of morbidity and mortality in patients with diabetes. Because the cardiac event rate in patients with diabetes with no histories of CAD is similar to that in patients without diabetes with CAD, individuals with diabetes should be managed as risk-equivalent CAD patients. Of note, asymptomatic individuals with diabetes without clinical CAD have an increased prevalence of coronary atherosclerosis compared with patients without diabetes, as well as a high incidence of death, myocardial infarction, and need for coronary revascularization over follow-up periods ranging from 3 to 10 years. In the past, screening with specialized testing was proposed in subjects with diabetes with two or more associated cardiovascular risk factors, aiming at a change in clinical management on the basis of test results that could improve outcomes. However, prospective studies using myocardial perfusion imaging provided conflicting results in high-risk asymptomatic patients with diabetes. In fact, the Detection of Ischemia in Asymptomatic Diabetics (DIAD) trial reported a very low cardiac event rate that was not reduced by screening for myocardial ischemia. Conversely, the Basel Asymptomatic High-Risk Diabetics’ Outcome Trial (BARDOT) documented effective prognostication by test results with, respectively, 3 and 7 times greater frequency of cardiac events and progression of disease at 2-year follow-up in patients with inducible myocardial ischemia. The uncertainty regarding the clinical utility of test screening can be summarized by the inconsistent recommendations proposed by scientific societies: whereas the American Diabetes Association currently does not recommend routine screening, the European Society of Cardiology’s guidelines on diabetes attribute a class IIb recommendation to the screening of patients with diabetes at particularly high risk, such as those with peripheral artery disease, high coronary artery calcium scores, or proteinuria, emphasizing, however, the need to better define the characteristics of subjects to be screened.
Dual-imaging stress echocardiography, combining conventional wall motion analysis with two-dimensional echocardiography and coronary flow velocity reserve (CFVR) with pulsed Doppler flow measurement of the mid-distal left anterior descending coronary artery (LAD), is the recommended technique during vasodilator stress echocardiography, having demonstrated feasibility >90% and the ability to improve the diagnostic and prognostic value of standard stress echocardiography. Of interest, in patients with diabetes with no stress-induced ischemia, reduced CFVR of the LAD was found to be a strong predictor of death or myocardial infarction in both an unselected sample and patients with chest pain and angiographically normal or near normal coronary arteries.
The aim of this prospective, observational study was to assess the prognostic implications of dual-imaging stress echocardiography in high-risk asymptomatic subjects with diabetes with no clinical evidence of CAD.
This was a prospective analysis of 230 asymptomatic patients (128 men; mean age, 66 ± 9 years) with diabetes mellitus lasting 13 ± 10 years who underwent dipyridamole stress echocardiography with CFVR assessment of the LAD by transthoracic Doppler from January 2007 to December 2012 at two Italian cardiology institutions (in Lucca and Cesena). Inclusion criteria were (1) no history of chest pain, dyspnea, or syncope; (2) no history of CAD (i.e., acute coronary syndrome, coronary revascularization, and/or angiographic evidence of ≥50% diameter coronary stenosis); (3) target organ damage, as defined by macrovascular disease (peripheral or carotid artery disease) or microvascular disease (albuminuria, retinopathy, or peripheral neuropathy), or the presence of two or more associated cardiovascular risk factors; (4) no pathologic Q waves or deep negative waves on the electrocardiogram; (5) no wall motion abnormalities on resting echocardiography; (6) no significant valvular or congenital heart disease; (7) no prognostically relevant noncardiac diseases (cancer, end-stage renal or liver disease, or severe obstructive pulmonary disease); and (8) adequate acoustic window for imaging of the left ventricle (for two-dimensional echocardiography) and LAD flow Doppler (for CFVR assessment). Follow-up information was available for all patients. Therapy was discontinued 72 hours before the test in 40 of 76 patients who were taking β-blockers and 24 hours before the test in 37 of 53 patients who were taking calcium antagonists. Accordingly, 47 individuals (20%) were evaluated under β-blockers ( n = 31), calcium antagonists ( n = 11), or both medications ( n = 5) ( Table 1 ). Phylline-containing drugs and beverages were discontinued ≥24 hours before testing. All patients gave written informed consent when they underwent stress echocardiography. When patients signed the informed consent document, they also authorized physicians to use their clinical data. Stress echocardiographic data were collected and analyzed by stress echocardiographers not involved in patient care. Test results were available to referring physicians, and indication for coronary revascularization was determined on the basis of several factors, of which test result was only one. Echocardiographers performing stress imaging were not involved in management and decision making for these patients.
|Normal test result ( n = 176)||Abnormal test result ( n = 54)||P|
|Age (y)||65 ± 9||67 ± 8||.19|
|Men||94 (53%)||34 (63%)||.22|
|Body mass index (kg/m 2 )||28 ± 3||28 ± 4||.30|
|Duration (y)||12 ± 9||17 ± 14||<.0001|
|1||5 (3%)||4 (7%)||.13|
|2||171 (97%)||50 (93%)||.13|
|Glycated hemoglobin (mg/dL)|
|Mean||7.5 ± 1.1||7.4 ± 1.3||.74|
|≤6.5||49 (28%)||14 (26%)||.78|
|6.6–8.0||85 (48%)||28 (52%)||.65|
|>8.0||42 (24%)||12 (22%)||.80|
|Oral glucose-lowering agent only||126 (72%)||29 (54%)||.01|
|Insulin only||22 (13%)||16 (30%)||.003|
|Both oral glucose-lowering agent and insulin||28 (16%)||9 (16%)||.89|
|Statin||116 (66%)||39 (72%)||.39|
|ACE inhibitor/ARB||108 (61%)||35 (65%)||.65|
|β-blocking agent||56 (32%)||20 (37%)||.48|
|Calcium antagonist||38 (22%)||15 (28%)||.34|
|Aspirin||104 (59%)||32 (59%)||.98|
|Creatinine (mg/dL)||0.95 ± 0.40||1.15 ± 0.53||.006|
|Blood pressure at the time of test (mm Hg)|
|Systolic||136 ± 16||141 ± 19||.03|
|Diastolic||78 ± 10||79 ± 10||.84|
|Cardiovascular risk factors|
|Family history of premature CAD||56 (32%)||7 (13%)||.007|
|Arterial hypertension||147 (84%)||46 (85%)||.77|
|Hyperlipidemia||109 (62%)||33 (61%)||.91|
|Smoking habit||54 (31%)||18 (33%)||.71|
|Overweight or obesity||151 (86%)||48 (89%)||.56|
|Number of associated risk factors||3.0 ± 0.9||2.8 ± 0.9||.32|
|At least three risk factors||134 (76%)||34 (63%)||.06|
|Target organ damage|
|Macrovascular disease||81 (46%)||32 (59%)||.09|
|Peripheral artery disease||52 (30%)||24 (44%)||.04|
|Carotid artery disease||72 (41%)||26 (48%)||.35|
|Microvascular disease||66 (38%)||33 (61%)||.002|
|Albuminuria||35 (20%)||19 (35%)||.02|
|Macroalbuminuria||7 (4%)||3 (6%)||.62|
|Microalbuminuria||28 (16%)||16 (30%)||.02|
|Retinopathy||35 (20%)||19 (35%)||.02|
|Peripheral neuropathy||35 (20%)||14 (26%)||.34|
|At least one target organ damage||105 (60%)||42 (78%)||<.0001|
|Left ventricular ejection fraction (%)||60 ± 4||60 ± 4||.32|
|Test performed on antianginal therapy|
|β-blocking agent||21 (12%)||15 (28%)||.005|
|Calcium antagonist||11 (6%)||5 (9%)||.45|
|At least one antianginal medication||30 (17%)||17 (32%)||.02|
|Ischemic test result||0||6 (11%)||<.0001|
|CFVR of LAD||2.69 ± 0.64||1.70 ± 0.26||<.0001|
|CFVR of LAD ≤ 2||0||52 (96%)||<.0001|
|Resting velocity of LAD (cm/sec)||32 ± 10||44 ± 26||<.0001|
|Peak velocity of LAD (cm/sec)||85 ± 31||74 ± 43||.17|
|Both ischemic test result and CFVR of LAD ≤ 2||0||4 (7%)||.0003|
Arterial hypertension, hyperlipidemia, overweight or obesity, family history of premature CAD, and smoking habit were taken as associated cardiovascular risk factors and defined according to standard definitions. Peripheral artery disease was assessed using Doppler ultrasound and/or arteriography, and carotid artery disease was defined using Doppler techniques. Microalbuminuria was a level of albumin ranging from 30 to 299 mg in a 24-hour urine collection, while macroalbuminuria was a urinary albumin excretion of ≥300 mg/24 h. Diabetic retinopathy was defined according to the criteria of the Early Treatment of Diabetic Retinopathy Study. Nerve conduction test and electromyography were adopted for refining the diagnosis of peripheral neuropathy.
Transthoracic stress echocardiographic studies were performed using a commercially available ultrasound machine (iE33 [Philips Medical Systems, Andover, MA] and Vivid 7 [GE Healthcare, Little Chalfont, United Kingdom]) equipped with multifrequency phased-array sector scan probes (S3–S8) and with second-harmonic technology. Two-dimensional echocardiography and 12-lead electrocardiographic monitoring were performed in combination with high-dose dipyridamole (up to 0.84 mg over 6 min). Echocardiographic images were semiquantitatively assessed using a 17-segment, four-point scale model of the left ventricle. A wall motion score index was derived by dividing the sum of individual segment scores by the number of interpretable segments. Ischemia was defined as stress-induced wall motion abnormality. CFVR was assessed during the standard stress echocardiographic examination by an intermittent imaging of both wall motion and LAD flow. Coronary flow in the mid-distal portion of the LAD was searched in the low parasternal long-axis section under the guidance of color Doppler flow mapping. All studies were digitally stored to simplify offline review and measurement. Coronary flow parameters were analyzed offline using the built-in calculation package of the ultrasound unit. Flow velocities were measured at least twice for each study: at baseline and at peak stress (before aminophylline injection). At each time point, three optimal profiles of peak diastolic Doppler flow velocities were measured, and the results were averaged. CFVR was defined as the ratio between hyperemic peak and basal peak diastolic coronary flow velocities. A value of CFVR ≤2 was considered reduced. The test result was considered normal in the case of no ischemia and CFVR >2 and abnormal in the case of ischemia and/or CFVR ≤2. All investigators from contributing centers passed quality control criteria for regional wall motion and Doppler interpretation before entering the study, as previously described. The previously assessed intra- and interobserver variability for measurements of Doppler recordings and regional wall motion analysis assessment were <10%.
Outcomes were determined from patient interviews at the outpatient clinic, hospital chart reviews, and telephone interviews with patients, their close relatives, or referring physician. Death, nonfatal myocardial infarction, and coronary revascularization (surgery or percutaneous intervention) were registered as clinical events. To avoid misclassification of the cause of death, overall mortality was considered. Myocardial infarction was defined by typical symptoms, electrocardiographic, and cardiac enzyme changes. Follow-up data were analyzed for the prediction of hard events (death or nonfatal myocardial infarction) and major adverse cardiac events (MACE) (death, nonfatal myocardial infarction, or coronary revascularization).
Continuous variables are expressed as mean ± SD. Differences between groups were compared using Student’s t and χ 2 tests, as appropriate. Event rates were estimated using Kaplan-Meier curves and compared using the log-rank test. Patients undergoing coronary revascularization were censored at the time of the procedure. Only the first event was taken into account. Annual event rates were obtained from Kaplan-Meier estimates to take censoring of the data into account. The associations of selected variables with outcomes were assessed using a Cox proportional hazard model using univariate and stepwise multivariate procedures. Significance of .05 was required for a variable to be included in the multivariate model, while .10 was the cutoff value for exclusion. According to a stepwise selection process, variables were entered into, or removed from, the regression equation on the basis of a computed significance probability value (maximized partial-likelihood ratio). Hazard ratios with corresponding 95% CIs were estimated. The parameters included in the Cox model were established a priori to assess how conventional risk factors, diabetes-related complications, and stress echocardiographic results were able to identify patients at risk for events. Moreover, clinical findings, resting wall motion score index, and ischemia on stress echocardiography were sequentially included in the model. The global χ 2 value of the model was calculated from the log likelihood ratio; a significant increase after the addition of further variables indicated incremental prognostic value. Statistical significance was set at P < .05. SPSS version 16.0 (SPSS, Chicago, IL) was used for analysis.
The main clinical and echocardiographic findings of the study population are reported in Table 1 . One hundred forty-seven patients (64%) had target organ damage, and 83 (36%) had two or more associated risk factors. Macrovascular and microvascular disease were documented in 113 (49%) and 99 (43%) patients, respectively ( Table 1 ). Sixty-five individuals (28%) had both macrovascular and microvascular disease. The mean number of associated risk factors in the entire study group was 3.0 ± 0.9, similar in subjects with and without organ damage ( P = .96).
Of 230 patients, 75 (35%) were taking insulin alone or in combination with oral glucose-lowering agents, 155 (67%) were taking statins, 143 (62%) were taking angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers, 136 (59%) were taking aspirin, and 76 (33%) were taking β-blockers ( Table 1 ).
The mean resting echocardiographic ejection fraction was 60 ± 4%. No complication or limiting side effects occurred during the tests. Mean coronary flow velocities were 34.8 ± 16.4 cm/sec at rest and 82.3 ± 34.7 cm/sec at peak stress. Mean CFVR in the entire population was 2.46 ± 0.71. The feasibility of CFVR assessment was 93%, consistent with our previous results. Inducible ischemia and CFVR ≤2 were detected in six and 52 patients, respectively. Four patients showed both ischemia and reduced CFVR. Accordingly, test results were abnormal in 54 subjects (23%) and normal in the remaining 176. Patients with abnormal test results had longer duration of diabetes, had a higher mean value of creatinine, were more often taking insulin, and had greater frequencies of peripheral artery disease, microalbuminuria, retinopathy, and at least one target organ damage ( Table 1 ). In addition, they were more frequently tested while on β-blocker therapy ( Table 1 ).
Coronary Angiographic Results
Fifty-two patients underwent coronary angiography. Fifteen had significant CAD (stenosis > 70%), seven had intermediate stenosis (50%–70%), and 30 had normal or near normal coronary arteries (stenosis < 50%).
During a mean follow-up period of 4.6 ± 2.7 years, there were 22 hard events (18 deaths and four nonfatal myocardial infarctions) and 17 coronary revascularizations (five surgical and 12 percutaneous interventions). Hard events and MACEs occurred in 10 (18.5%) and 23 (42.6%) patients with abnormal test results, respectively, and in 12 (6.7%) and 16 (9.0%) patients with normal test results. Of 17 revascularizations, nine were performed in the first 3 months after testing (15 ± 14 days) and eight after 3 months (1,065 ± 972 days). The subset of patients who underwent early revascularization had CFVR significantly lower than those with late revascularization (1.54 ± 0.24 vs 2.05 ± 0.35, P = .003).
Annual hard event rate and MACE rate in the study population were 2.1% and 3.6%, respectively. The yearly hard event and MACE incidence rates by main clinical and stress echocardiographic features are shown in Figures 1 and 2 . Among clinical variables, age ≥ 65 years, insulin therapy, macrovascular disease, microvascular disease, and end-organ damage were associated with ≥2 times higher hard event rate and MACE rate per year ( Figure 1 and 2 ). An abnormal stress echocardiographic result predicted a 4.1 times higher yearly hard event rate ( Figure 1 ) and a 6.8 times higher yearly MACE rate ( Figure 2 ).