Background
Type 2 diabetes mellitus (T2DM) is a chronic metabolic disease that represents a major public health concern. T2DM affects more than 382 million people world-wide and the prevalence is expected to increase substantially [
1]. Cardiovascular (CV) disease is the most common cause of death in patients with T2DM [
2,
3]. The prevalence of undiagnosed coronary artery disease (CAD) among asymptomatic patients with T2DM is high and independent assessments using different diagnostic techniques, i.e., invasive coronary angiography (ICA) [
4] or maximum stress-test [
5] have found that more than 1 in 5 adults with T2DM have significant CAD. The diagnosis of CAD may be missed or delayed in these patients since the typical symptoms of CAD are often absent in patients with long-standing T2DM, which in turn further increases their risk for CV events. To potentially prevent CV events it may therefore be important to detect subclinical CAD in T2DM to enable appropriate intervention to reduce the risk of fatal and non-fatal cardiac events.
Findings to date to not support widespread screening for CAD in patients with T2DM, but this may partly be related to lack of a reliable and cost-effective screening tool [
6].
ICA has been the accepted gold standard method for assessing the presence, localization, and severity of CAD, but has a substantial procedural cost and is an invasive method associated with a risk for complications. Coronary computed tomography angiography (CCTA) has over recent years emerged as an alternative to ICA for CAD assessment and the 64-slice generation scanners multidetector computer tomography (MDCT) is now considered to have high diagnostic performance for detection of significant coronary stenosis [
7] in different populations. So far only few studies have assessed the role of MDCT in patients with T2DM [
8,
9], and its full potential is not fully understood, in particular in light of studies being suggestive of a reduced diagnostic performance of MDCT compared to non-T2DM populations [
10]. Thus, the role of CCTA is considered uncertain in asymptomatic high-risk patients like patients with T2DM [
11].
In this study we aimed to assess the role of CCTA in detecting and characterizing CAD in patients with T2DM without cardiac symptoms. To provide robustness of using CCTA, all patients underwent both CCTA and gold standard ICA, and to our knowledge this is the first study to employ both strategies in asymptomatic patients with T2DM.
Results
Demographics
Of 93 patients consenting to participate in ABCD-2, 56 patients had no contraindications to the imaging procedures and consented to perform both CCTA and ICA. In eight patients CT angiography was not performed, because of calcium score >1000, rendering 48 patients eligible for analysis. The patient population studied was obese (mean BMI 29.6 kg/m2) and middle-aged (mean age 64 years) with a long history of T2DM (mean duration 15.6 years), not at optimal glycemic control (mean HbA1c 7.4 %). The majority were males (75 %). Further characteristics are given in Table
1.
Table 1
Baseline characteristics of the study participants
Number | 48 |
Background information |
Gender (male, n (%) | 36 (75 %) |
Age (years), mean ± SD (range) | 64.0 ± 7.3 (41.4–76,7) |
Body Mass Index (kg/m2), mean ± SD (range) | 29.6 ± 4.3 (21.7–39.7) |
T2DM duration (years), mean ± SD (range) | 14.6 ± 6.4 (6.8–37.7) |
Medical history and medications |
History of smoking, n (%) | 19 (39.6 %) |
Family history of premature CAD (1st degree relative), n (%) | 24 (50 %) |
Any medication for hypertension, n (%) | 36 (75 %) |
Any lipidlowering medication, n (%) | 38 (79 %) |
Any blood glucose lowering medication, n (%) | 46 (96 %) |
-Any use of insulin, n (%) | 9 (18.7 %) |
Laboratory and clinically assessment |
HbA1c (%), mean ± SD | 7,4 ± 1.1 |
Serum creatinine (μmol/L), mean ± SD (range) | 71.4 ± 16.7 (42–117) |
Total cholesterol (mmol/L), mean ± SD | 4.3 ± 1.0 |
LDL cholesterol (mmol/L), mean ± SD | 2.4 ± 0.8 |
HDL cholesterol (mmol/L), mean ± SD | 1.4 ± 0.4 |
Triglycerides (mmol/L), mean ± SD | 1.3 ± 0.5 |
Systolic blood pressure (mmHg), mean ± SD | 138 ± 17 |
Diastolic blood pressure (mmHg), mean ± SD | 79 ± 8 |
Heart rate (bpm), mean ± SD | 73 ± 13 |
Exercise ECG | |
- Positive, n (%) - Inconclusive, n (%) - Negative, n (%) | 4 (8.3 %) 11 (22.9 %) 33 (68.8 %) |
Imaging assessments
In total 99 % (588 out of 594) of coronary segments with a luminal diameter ≥1 mm from the full patient population were eligible for analysis with CCTA. Only 6 segments were non-interpretable, mainly because of severe artifacts from pacemaker electrodes. Mean Agatston score was 269, but ranged from 0 to 976 and this did not correlate with neither diabetes duration (r = 0.05, p = 0.75) or HbA1c (r = -0.14, p = 0.35). Image quality and degree of calcifications are detailed in Table
2. In analyzing the 588 segments, CCTA had a sensitivity of 90 % (9 of 10) and a specificity of 96 % (557 of 578) for detection of coronary stenosis ≥50 % using ICA as gold standard (Table
3). The positive predictive value was 30 % (9 of 30) and the negative predictive value 99 % (556 of 557). Degree of stenosis per segment analysis and maximum degree of any stenosis per patient analysis are given in Table
3. CCTA wrongly classified 21 segments (3.8 %) as having significant stenosis. Of these 21 segments ICA classified 10 segments with 25-49 % stenosis and the remaining 11 segments without any stenosis (≤24 % lumen reduction). All these 21 segments were found to have either severe calcifications or a combination of calcifications and minor artifacts on CCTA.
Table 2
Coronary CTA, scan characteristics and results
Intravenous β-blocker given, n (%) | 18 (37.5 %) |
Heart rate during scan, mean ± SD (range) | 63.7 ± 7.6 (49–88) |
Radiation dose (mSv), mean ± SD (range) | 3.8 ± 0.7 (2.5–5.0) |
Agatston score, mean ± SD (range) | 269.0 ± 292.8 (0–976) |
Image quality per segment, n (%) | |
1. Good 2. Adequate 3. Decreased | 420 (71 %) 132 (23 %) 36 (6 %) |
Degree of calcifications per segment, n (%) | |
1. None 2. Minor 3. Severe | 374 (64 %) 171 (29 %) 43 (7 %) |
Table 3
Results of CCTA and ICA
Diagnostic accuracy of CCTA using ICA as gold standard | Per segment | Per patient |
- Sensitivity, % (CI) - Spesificity, % (CI) - Positive predictice value, % (CI) - Negative predictive value, % (CI) | 90 % (54–99) 96 % (94–98) 30 % (15–50) 99 % (99–100) | 100 % (60–100) 78 % (61–89) 47 % (24–72) 100 % (86–100) |
Degree of stenosis n (%), per segment analysis | CCTA | ICA |
0. No stenosis or ≤24 % lumen reduction 1. 25–49 % stenosis 2. 50–74 % stenosis 3. ≥75 % stenosis | 447 (76.0 %) 110 (18.7 %) 29 (4.9 %) 2 (0.4 %) | 553 (94.0 %) 25 (4.3 %) 9 (1.5 %) 1 (0.2 %) |
Maximum degree of any stenosis, n (%), per patient analysis | CCTA | ICA |
0. No stenosis or ≤24 % lumen reduction 1. 25–49 % stenosis 2. 50–74 % stenosis 3. ≥75 % stenosis | 10 (20.8 %) 21 (43.8 %) 15 (31.3 %) 2 (4.2 %) | 28 (58.3 %) 12 (25.0 %) 7 (14.6 %) 1 (2.1 %) |
In a per patient analysis, 17 % (8 of 48) had at least one stenosis ≥50 % in any segment at ICA. CCTA correctly identified all these patients, corresponding to a sensitivity of 100 %. Specificity was 78 %, positive predictive value 47 %, and negative predictive value 100 %. Estimated Kappa was 0,53, giving moderate agreement between the methods.
At the maximum exercise test, the mean achieved work capacity was 143 ± 44 Watts. 15 patients (31 %) had signs of silent ischemia with positive or inconclusive test. When comparing silent ischemia as detected on the maximum exercise test versus CCTA and ICA estimated Kappa were 0,16 and 0,17 respectively, yielding poor agreement.
Discussion
This study assessed the applicability of using a modern non-invasive imaging modality to detect significant CAD in an intermediate CV risk cohort of patients with T2DM and compared its results with the gold standard for assessing this, namely ICA. This study, to the best of our knowledge, is the first that have screened asymptomatic patients with both CCTA and ICA. The main finding of our study was that 256-slice CCTA with step-and-shoot technique provides excellent sensitivity and yields a high negative predictive value for excluding significant CAD in patients with T2DM, whereas the specificity and positive predictive value were lower (78 % and 47 %, respectively). These data are in agreement with results of other assessments on sensitivities in patients with low degree of vasculopathy, i.e. low calcium scores [
19]. Our study also shows that the technique provides good feasibility with nearly all coronary segments eligible for analysis. Radiation dose is lower than with earlier generation CT scanners [
20] and may be further reduced with newer techniques like iterative reconstruction [
21,
22]. Our data suggested that despite a mean T2D duration of 15 years, mean Agatston score was 269, indicating that the population still was at intermediate CV risk, and not too advanced, as one could have expected based on the lengthy diabetes duration [
23].
CCTA is well known to yield many false positives, but the low positive predictive values in our study seems inferior to most other reports although the results are varying [
24]. There may be several contributions to this. Positive and negative predictive values are generally influenced by prevalence of disease, and the low prevalence of significant CAD in this study would contribute to lower positive predictive value than in studies with higher prevalence. The coronary plaque burden and coronary calcium score are generally higher in diabetic than in non-diabetic patients [
19,
25] and since calcified plaques can lead to overestimation of lesion severity [
26] this may contribute to lower positive predictive value for CCTA in a diabetic population than in a non-diabetic population. Some studies have shown smaller coronary vessel calibre in patients with T2DM as compared to those without [
27,
28], and this may affect the diagnostic accuracy of CCTA. Reduced diagnostic performance of CCTA in patients with T2DM is reported in the only known study comparing with patients without T2DM [
10], but this study included mainly symptomatic patients and the prevalence of obstructive CAD was high. On the other side, a mean CAC value of 269 and 18 % patients with zero CAC score would also render CCTA a suboptimal test to rule out obstructive CAD among symptomatic patients [
19].
The major established indications for use of CCTA are within groups of symptomatic patients with low or intermediate pretest probability of obstructive CAD, and in some pre- and postoperative settings [
11]. A recent meta-analysis supports the use of CCTA as a first imaging test for low- and intermediate-risk patients presenting to the emergency department with chest pain [
29]. The role of CCTA remains uncertain in asymptomatic high-risk patients like patients with T2DM, and even if there are reports finding a prognostic value of CCTA [
8,
9] there is another recent study reporting that use of CCTA to screen asymptomatic patients with T2DM do not improve clinical outcome [
30]. One study in patients with T2DM and mild anginal complaints demonstrated a crucial impact of ischemia on cardiac event rate and showed a prognostic value of myocardial perfusion scintigraphy (MPS [
31], and a possible strategy may be to perform supplementary MPS in asymptomatic patients with positive CCTA. Another possible future improvement of screening CCTA may be a combination with computed tomography myocardial perfusion imaging, as there are recent promising reports about this method [
32,
33], or to combine results of CCTA with prognostic biomarkers [
34].
The relatively small number of included patients is a limitation of the generalizability of the study results, including the relatively low proportion of female patients.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Competing interests
APO and OEJ are employees of Boehringer Ingelheim. The authors declare that they have no other competing interests.
Authors’ contributions
GRU, APO, LG, OEJ and AB contributed in the conception and design of the study and in drafting of the manuscript. GRU and AB performed and interpreted all CCTA examinations. KE interpreted all ICA examinations. GRU, APO, OEJ and AB contributed in analysis of study data. GRU and OEJ drafted the manuscript. All authors revised and approved the final manuscript.