Clinical Predictors of Coronary Artery Plaque Progression by Quantitative Serial Assessment Using 320-Row Computed Tomography Coronary Angiography in Asymptomatic Patients with Type 2 Diabetes Mellitus

Open ArchivePublished:June 06, 2020DOI:https://doi.org/10.1016/j.jjcc.2020.05.004

      Highlights

      • We prospectively analyzed coronary plaque in diabetic patients by serial coronary computed tomography.
      • Increase in hemoglobin A1c (HbA1c) was an independent predictor for coronary plaque progression.
      • Homeostasis model assessment of insulin resistance was also associated with coronary plaque progression.
      • Increase in low-density lipoprotein-cholesterol, but not HbA1c, correlated to increase in necrotic core.

      Abstract

      Background

      Natural history of coronary plaque progression (PP) in patients with diabetes mellitus (DM) remains unclear. This study aimed to investigate the clinical predictors of coronary PP in patients with DM.

      Methods

      In this prospective observational study, we analyzed 70 asymptomatic patients (age, 64.4 years; male, 67%) with type 2 DM without prior history of coronary artery disease who underwent serial 320-row computed tomography coronary angiography with an interscan interval of more than 24 months (median 37.7 months). Study endpoint was PP, which was defined if coronary plaque volumes (PVs) at follow-up minus PVs at baseline was >0. We evaluated plaque composition using the Hounsfield Unit thresholds and insulin resistance estimated by the homeostasis model assessment of insulin resistance (HOMA-IR).

      Results

      Thirty-nine patients who showed PP had a higher increase in hemoglobin A1c (⊿HbA1c) from baseline to follow-up than those without PP (0.3% ± 0.8% vs −0.4% ± 1.1%; p = 0.01), although there was no statistical difference in HbA1c at baseline (7.1 ± 0.5% vs. 7.3 ± 1.4%; p = 0.24). In multivariable analysis, ⊿HbA1c [odds ratio (OR): 3.05; 95% confidence interval (CI): 1.39–6.67; p = 0.001] was an independent predictor for PP.
      Increase in low-density lipoprotein cholesterol (⊿LDL-C), not ⊿HbA1c, was significantly correlated to percent change in necrotic core (NC) volume (β-coefficients: 0.04; 95% CI: 0.004 - 0.08; p = 0.03). Among 48 patients without insulin therapy, patients with PP (n = 28) had a higher increase in HOMA-IR than those without PP (n = 20) (0.95 ± 2.00 vs. −0.63 ± 1.31; p = 0.003).

      Conclusions

      Increase in HbA1c and HOMA-IR was associated with PP in asymptomatic patients with type 2 DM, whereas increase in LDL-C was correlated to increase in NC.

      Keywords

      Introduction

      Diabetes mellitus (DM) remains a significant independent cardiovascular risk factor [
      • Emerging Risk Factors Collaboration
      • Sarwar N.
      • Gao P.
      • Seshasai S.R.
      • Gobin R.
      • Kaptoge S.
      • et al.
      Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies.
      ]. Additionally, despite the progress in cardiovascular disease treatment and diagnosis, coronary artery disease (CAD) is still the leading cause of death in DM patients [
      • Fox C.S.
      • Coady S.
      • Sorlie P.D.
      • D’Agostino Sr, R.B.
      • Pencina M.J.
      • Vasan R.S.
      • et al.
      Increasing cardiovascular disease burden due to diabetes mellitus: the Framingham Heart Study.
      ]. Recent studies have shown that silent myocardial ischemia affects 20%–35% of asymptomatic patients with DM and that approximately 25% of myocardial infarctions in diabetic patients are asymptomatic [
      • Wackers F.J.
      • Young L.H.
      • Inzucchi S.E.
      • Chyun D.A.
      • Davey J.A.
      • Barrett E.J.
      • et al.
      Detection of silent myocardial ischemia in asymptomatic diabetic subjects: the DIAD study.
      ,
      • Soejima H.
      • Ogawa H.
      • Morimoto T.
      • Okada S.
      • Sakuma M.
      • Nakayama M.
      • et al.
      One quarter of total myocardial infarctions are silent manifestation in patients with type 2 diabetes mellitus.
      ]. Once CAD is symptomatic, morbidity and mortality are high and significantly worse in DM patients than in non-DM patients [
      Influence of diabetes on 5-year mortality and morbidity in a randomized trial comparing CABG and PTCA in patients with multivessel disease: the Bypass Angioplasty Revascularization Investigation (BARI).
      ,
      • Poutanen O.
      • Mattila A.
      • Seppala N.H.
      • Groth L.
      • Koivisto A.M.
      • Salokangas R.K.
      Seven-year outcome of depression in primary and psychiatric outpatient care: results of the TADEP (Tampere Depression) II Study.
      ]. Hence, preventing coronary atherosclerosis in asymptomatic patients with DM is important.
      Coronary computed tomography angiography (CCTA) is a well-established non-invasive method for the assessment of coronary atherosclerosis. CCTA allows the quantification of coronary plaque volume (PV) and evaluation of coronary plaque characteristics, such as necrotic core (NC), similar to intravascular ultrasound (IVUS) [
      • Motoyama S.
      • Ito H.
      • Sarai M.
      • Kondo T.
      • Kawai H.
      • Nagahara Y.
      • et al.
      Plaque characterization by coronary computed tomography angiography and the likelihood of acute coronary events in mid-term follow-up.
      ,
      • de Graaf M.A.
      • Broersen A.
      • Kitslaar P.H.
      • Roos C.J.
      • Dijkstra J.
      • Lelieveldt B.P.
      • et al.
      Automatic quantification and characterization of coronary atherosclerosis with computed tomography coronary angiography: cross-correlation with intravascular ultrasound virtual histology.
      ,
      • Boogers M.J.
      • Broersen A.
      • van Velzen J.E.
      • de Graaf F.R.
      • El-Naggar H.M.
      • Kitslaar P.H.
      • et al.
      Automated quantification of coronary plaque with computed tomography: comparison with intravascular ultrasound using a dedicated registration algorithm for fusion-based quantification.
      ]. Recently, the PARADIGM (Progression of Atherosclerotic Plaque Determined by Computed Tomographic Angiography Imaging) investigators [
      • Kim U.
      • Leipsic J.A.
      • Sellers S.L.
      • Shao M.
      • Blanke P.
      • Hadamitzky M.
      • et al.
      Natural history of diabetic coronary atherosclerosis by quantitative measurement of serial coronary computed tomographic angiography: results of the PARADIGM study.
      ] retrospectively demonstrated the benefit of using CCTA to evaluate coronary plaque progression (PP) and changes in plaque characteristics in DM patients and revealed predictors of coronary PP. The identification of predictors of PP using CCTA images and the therapeutic strategy for preventing PP can lead to clinical cardiac event reduction in asymptomatic patients with DM. However, the target of glycemic control for preventing PP is still unclear.
      We prospectively investigated the association between glycemic control and coronary PP or changes in coronary plaque constitution in asymptomatic patients with type 2 DM by serial assessment using 320-row CCTA.

      Methods

       Study design

      A single-center prospective study recruited 120 asymptomatic patients aged >45 years with type 2 DM from October 2014 to July 2015. The included patients had a previous diagnosis of type 2 DM or received treatment with oral hypoglycemic agents or insulin. Exclusion criteria were history of percutaneous coronary intervention or coronary artery bypass graft, severe heart failure (New York Heart Association classification ≥3), bronchial asthma, allergy to iodinated contrast agent, and known severe renal failure (estimated glomerular filtration rate <30 mL/min/1.73 m2).
      The patients underwent 320-row CCTA at enrolment, and repeat CCTA was performed at follow-up with an interscan interval of >24 months. We excluded those patients from follow-up CCTA if they presented with allergy to iodinated contrast agents and severe renal failure after CCTA at baseline.
      The present study was conducted in accordance with the Declaration of Helsinki. All patients provided informed consent prior to study enrolment. The study was approved by the Mitsui Memorial Hospital ethics committee and registered as UMIN000014765 in the UMIN (University Hospital Medical Information Network) Clinical Trials Registry.

      Measurement of clinical and confounding variables

      We systemically collected information on the presence of cardiac risk factors for each individual at baseline CCTA scan. Height and body weight were measured at baseline and follow-up CCTA scans. Laboratory data including hemoglobin A1c (HbA1c), fasting plasma glucose and insulin, total cholesterol, triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) levels were measured within 1 month of the baseline and follow-up CT scans. During the follow-up periods, the physicians examined these laboratory data at least four times every year. The patients received the national guideline-based standard diabetes care from the individual clinicians. Among the patients without insulin therapy, insulin resistance (IR) was estimated by the homeostasis model assessment of IR (HOMA-IR) values, calculated as [fasting insulin (mU/mL) × fasting plasma glucose (mg/dl)/405] [
      • Matthews D.R.
      • Hosker J.P.
      • Rudenski A.S.
      • Naylor B.A.
      • Treacher D.F.
      • Turner R.C.
      Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man.
      ].

      Scan protocol and image reconstruction

      CCTA scans were obtained using a 320-slice multidetector CT scanner (Aqualion One, Canon Medical Systems Ltd., Tochigi, Japan) according to a previously described protocol [
      • Kishi S.
      • Giannopoulos A.A.
      • Tang A.
      • Kato N.
      • Chatzizisis Y.S.
      • Dennie C.
      • et al.
      Fractional flow reserve estimated at coronary CT angiography in intermediate lesions: comparison of diagnostic accuracy of different methods to determine coronary flow distribution.
      ]. Scanning parameters were as follows: collimation of 320 rows × 0.5 mm, rotation time of 350–400 ms, and tube voltage of 120 kV. The tube current (270–400 mA) was selected according to the standard deviation of the noise level measured on the CT projection radiograph. Electrocardiogram-gated non-enhanced scans were obtained, followed by contrast-enhanced scans using Iopamiron 370 mg/mL (Bayer Schering Pharma, Osaka, Japan) injected at a speed of 4.0–5.0 mL/s over 10 s. The scanner’s “arrhythmia rejection” software automatically removed irregular beatings using the multi-segment acquisition technique. We semi-automatically determined the appropriate cardiac phase with minimum cardiac motion for the CT axial image reconstruction using the Phase-NAVI scanner software (Canon Medical Systems Ltd.). All data acquired with 0.25 mm-thick slices were reconstructed using a standard kernel of FC04.

      CCTA analysis

      Quantitative plaque analysis was performed using a QAngio CT work station (Research Edition, version 2.0.5; Medis Medical Imaging Systems, Leiden, the Netherlands) by two cardiologists as previously described [
      • Kato N.
      • Kishi S.
      • Arbab-Zadeh A.
      • Rybicki F.J.
      • Tanimoto S.
      • Aoki J.
      • et al.
      Relative atherosclerotic plaque volume by CT coronary angiography trumps conventional stenosis assessment for identifying flow-limiting lesions.
      ]. The coronary tree was automatically extracted, and each major vessel (left anterior descending artery, left circumflex artery, right coronary artery) was individually analyzed from the ostium to the point at which the internal vessel caliber decreased to <2.0 mm, exclusive of focal stenosis. Segmentation was performed using the American Heart Association 16-segment classification [
      • Austen W.G.
      • Edwards J.E.
      • Frye R.L.
      • Gensini G.G.
      • Gott V.L.
      • Griffith L.S.
      • et al.
      A reporting system on patients evaluated for coronary artery disease. Report of the ad hoc committee for grading of coronary artery disease, council on cardiovascular surgery, American Heart Association.
      ].
      Automated longitudinal contouring of the inner lumen and outer wall was performed; results were manually adjusted when clear deviations were noted (Online Fig. 1) and reviewed on transverse reconstructed cross-sections of the artery on a section-by-section basis at 0.5-mm increments. PV was calculated by subtracting the lumen volume from the outer wall volume. Total PV was the sum of each PV in all three major coronary vessels. Furthermore, PP was defined as the PV difference of >0, which was calculated by subtracting the total PV at baseline CCTA from the total PV at follow-up CCTA.
      In the plaque constitution analysis, we used an adaptive threshold previously reported [
      • de Graaf M.A.
      • Broersen A.
      • Kitslaar P.H.
      • Roos C.J.
      • Dijkstra J.
      • Lelieveldt B.P.
      • et al.
      Automatic quantification and characterization of coronary atherosclerosis with computed tomography coronary angiography: cross-correlation with intravascular ultrasound virtual histology.
      ] and divided PV into the following four classifications: fibrous, fibro-fatty, NC, and dense calcium volumes. In this approach, the Hounsfield Unit (HU) thresholds were adapted according to the lumen attenuation value. These dynamic thresholds were automatically derived and user independent. Percent fibrous plaque (%FP), fibro-fatty plaque (%FFP), NC (%NC), and dense calcium (%DC) were defined as the volume of each parameter divided by total PV. Longitudinal changes in coronary plaque constitution, such as changes in %FP (⊿%FP), %FFP (⊿%FFP), %NC (⊿%NC), and %DC (⊿%DC), were defined as each component at follow-up minus that at baseline. Previous studies demonstrated excellent intra- and inter-observer, and interscan reproducibility in plaque analysis of CCTA using semiautomated software [
      • de Graaf M.A.
      • Broersen A.
      • Kitslaar P.H.
      • Roos C.J.
      • Dijkstra J.
      • Lelieveldt B.P.
      • et al.
      Automatic quantification and characterization of coronary atherosclerosis with computed tomography coronary angiography: cross-correlation with intravascular ultrasound virtual histology.
      ,
      • Boogers M.J.
      • Broersen A.
      • van Velzen J.E.
      • de Graaf F.R.
      • El-Naggar H.M.
      • Kitslaar P.H.
      • et al.
      Automated quantification of coronary plaque with computed tomography: comparison with intravascular ultrasound using a dedicated registration algorithm for fusion-based quantification.
      ,
      • Kato N.
      • Kishi S.
      • Arbab-Zadeh A.
      • Rybicki F.J.
      • Tanimoto S.
      • Aoki J.
      • et al.
      Relative atherosclerotic plaque volume by CT coronary angiography trumps conventional stenosis assessment for identifying flow-limiting lesions.
      ]. The coronary artery calcium (CAC) score was calculated using a workstation (Ziostation 2, Ziosoft, Tokyo, Japan) and expressed as an Agatston score as previously reported [
      • Agatston A.S.
      • Janowitz W.R.
      • Hildner F.J.
      • Zusmer N.R.
      • Viamonte Jr, M.
      • Detrano R.
      Quantification of coronary artery calcium using ultrafast computed tomography.
      ].

      Sample size calculation

      At the time of study design and commencement, no studies had assessed PP in DM patients using CT. Therefore, we assumed that total plaque volume increased 8% per year according to a previous study evaluating PP in DM patients using IVUS [
      • Inaba S.
      • Okayama H.
      • Funada J.
      • Higashi H.
      • Saito M.
      • Yoshii T.
      • et al.
      Impact of type 2 diabetes on serial changes in tissue characteristics of coronary plaques: an integrated backscatter intravascular ultrasound analysis.
      ], and that enrollment of 108 patients would achieve 80% power at an α level of 0.05. An overall sample size of approximately 120 subjects was deemed necessary for study enrollment when permitting a 20% drop-out rate.

      Statistical analysis

      Participants’ baseline data were presented as mean ± standard deviation or as median [first quartile, third quartile] for continuous variables and as frequencies or proportions for categorical variables. The Spearman rank correlation test or Wilcoxon test for non-parametric parameters were used to assess the correlation between two variables. Interobserver and intraobserver reproducibility of PV measurement was tested by Pearson’s linear regression analysis and Bland–Altman plotting (mean of the 2 assessments on the x-axis and ratio of the 2 assessments on the y-axis) for 20 randomly selected cases. Multivariable logistic models or multivariable linear regression models were used to assess the predictors for PP. Factors with a p-value < 0.10 by univariate analysis were included in the multivariate analysis. The strength of the association is presented as odds ratios [ORs, 95% confidence interval (CI)]. The association between changes in plaque constitution and each variable was reported using β coefficients. Statistical significance was set at p < 0.05. All analyses were conducted using JMP (version 13, for Windows, SAS Institute, Inc., Cary, NC, USA) and R software (version 3.2.2; R Foundation for Statistical Computing, Vienna, Austria).

      Results

       Patients’ baseline characteristics

      We enrolled 120 patients at baseline. Fifty patients were excluded because 2 patients died, 15 had cardiovascular events (1 patient suffered from non-fatal myocardial infarction, 2 had cerebral infarction, 12 had coronary revascularization), 3 patients had allergy to iodinated contrast agent after baseline CCTA, 25 patients withdrew consent for follow-up CCTA, and 5 were lost to follow-up. Seventy patients were analyzed for PP (Online Fig. 2). At the follow-up CCTA, 2 of 70 patients had atypical chest symptoms that were not associated with CAD, since the follow-up CCTA did not reveal any significant coronary stenosis.
      Thirty-nine patients showed PP at follow-up CCTA, whereas 31 patients had PV reduction at follow-up CCTA. Patients’ baseline characteristics are summarized in Table 1. The mean patient age was 64 years, and 70% of the patients were male. Median follow-up period was 37.7 [36.3, 41.3] months. Regarding laboratory data at enrolment, the mean HbA1c and LDL-C levels were 7.2% and 97 mg/dl, respectively. Twenty patients received insulin therapy and 42 patients had statins at enrolment.
      Table 1Baseline patients characteristics.
      AllPlaque progressionNon-Plaque progression
      Parametersn = 70n = 39n = 31p-Value
      Age (years)64.4 ± 8.265.0 ± 8.963.6 ± 7.40.51
      Male, n (%)48 (69)26 (67)22 (71)0.55
      BMI at baseline (kg/m2)25.1 ± 4.226.1 ± 4.323.7 ± 3.70.01
      Hypertension, n (%)37 (53)22 (56)15 (48)0.42
      Dyslipidemia, n (%)50 (71)27 (69)23 (74)0.49
      Current smoker, n (%)15 (21)9 (23)6 (19)0.79
      Family history, n (%)22 (31)12 (31)10 (32)0.95
      Follow-up period (months)37.7 [36.3, 41.3]37.2 [36.2, 40.9]39.1 [36.5, 42.3]0.28
      Laboratory data at baseline
       HbA1c (%)7.2 ± 1.07.1 ± 0.57.3 ± 1.40.24
       LDL-C (mg/dl)97 ± 2791 ± 24103 ± 280.04
       HDL-C (mg/dl)55 ± 1254 ± 1157 ± 120.32
       TG (mg/dl)129 ± 83136 ± 92120 ± 690.43
       eGFR (mL/min/1.73 m2)75.5 ± 13.275.1 ± 12.875.9 ± 13.90.81
       HS-CRP (mg/dl)0.06 [0.02, 0.13]0.05 [0.02, 0.13]0.06 [0.02, 0.15]0.80
      Laboratory data at follow-up
       HbA1c (%)7.2 ± 0.87.3 ± 0.87.0 ± 0.60.04
       LDL-C (mg/dl)92 ± 2689 ± 2495 ± 290.20
       HDL-C (mg/dl)60 ± 1558 ± 1463 ± 150.21
       TG (mg/dl)134 ± 101141 ± 115126 ± 800.55
       eGFR (mL/min/1.73 m2)74.9 ± 15.273.4 ± 15.076.8 ± 15.60.37
       HS-CRP (mg/dl)0.07 [0.04, 0.14]0.08 [0.02, 0.11]0.06 [0.03, 0.08]0.15
      Medication at baseline
       Insulin, n (%)20 (29)9 (23)11 (37)0.22
       Biguanides, n (%)43 (61)24 (62)19 (61)0.90
       α-Glucosidase inhibitor, n (%)5 (7)2 (5)3 (10)0.44
       Sulfonylurea, n (%)31 (44)17 (44)14 (45)0.98
       Thiazolidine, n (%)10 (14)6 (15)4 (13)0.81
       DPP-4 inhibitor, n (%)46 (66)26 (67)20 (65)0.77
       SGLT-2 inhibitor, n (%)3 (4)2 (5)1 (3)0.72
       Statin, n (%)42 (60)23 (59)19 (61)0.93
      Medication at follow-up
       Insulin, n (%)21 (30)11 (28)10 (33)0.65
       Biguanides, n (%)47 (67)27 (69)20 (65)0.61
       α-Glucosidase inhibitor, n (%)5 (7)2 (5)3 (10)0.44
       Sulfonylurea, n (%)29 (41)14 (36)15 (48)0.37
       Thiazolidine, n (%)9 (13)5 (13)4 (13)0.95
       DPP-4 inhibitor, n (%)50 (71)27 (69)23 (74)0.71
       SGLT-2 inhibitor, n (%)14 (20)7 (18)7 (23)0.83
       Statin, n (%)52 (74)28 (72)24 (77)0.65
      The data represent the mean ± SD, the medians [interquartile ranges], or n (%).
      BMI, body mass index; HbA1c, hemoglobin A1c; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; eGFR, estimated glomerular filtration rate; HS-CRP, high-sensitivity C-reactive protein; DPP-4, dipeptidyl peptidase-4; SGLT-2, sodium-glucose cotransporter-2.
      At follow-up CCTA, the mean HbA1c and LDL-C levels were 7.2% and 92 mg/dl, respectively, which did not significantly change from those at enrolment. Although the number of patients with insulin therapy did not significantly increase (20 patients vs. 21 patients), those of patients taking statins significantly increased at follow-up CCTA (42 patients vs. 52 patients).
      Patients with PP had significantly higher body mass index (BMI) and lower LDL-C level (26.1 ± 4.3 kg/m2 vs. 23.7 ± 3.7 kg/m2, p = 0.01: 91 ± 24 mg/dl vs. 103 ± 28 mg/dl, p = 0.03, respectively). There were no significant differences in the follow-up period and medication use at both enrolment and follow-up between groups.

       Analysis of glycemic control and insulin resistance

      The glycemic control among the total study population was stable, and HbA1c level did not significantly change at baseline and follow-up, including the average HbA1c level at each year (Table 2A). Although there was no significant difference in HbA1c level at baseline and average HbA1c level at each year between patients with and without PP, patients with PP had higher HbA1c level (7.3% ± 0.8% vs. 7.0% ± 0.6%, p = 0.04) and higher increase in HbA1c level (⊿HbA1c) (0.3% ± 0.8% vs. −0.4% ± 1.1%, p = 0.01) at follow-up.
      Table 2Analysis of glycemic control and insulin resistance.
      A. Analysis of HbA1c level
      AllPlaque progressionNon-Plaque progression
      Parametersn = 70n = 39n = 31p-Value
      HbA1c level
       At baseline (%)7.2 ± 1.07.1 ± 0.57.3 ± 1.40.24
       Average in the 1 st year (%)7.1 ± 0.67.1 ± 0.47.1 ± 0.70.91
       Average in the 2nd year (%)7.1 ± 0.67.2 ± 0.57.1 ± 0.60.39
       Average in the 3rd year (%)7.2 ± 0.77.2 ± 0.87.0 ± 0.60.24
       At follow up (%)7.2 ± 0.87.3 ± 0.87.0 ± 0.60.04
      Average HbA1c during the follow-up (%)7.1 ± 0.57.2 ± 0.57.0 ± 0.60.29
      ⊿HbA1c (%)0.01 ± 1.00.3 ± 0.8−0.4 ± 1.10.01
      B. Analysis of insulin resistance
      All patients without insulin therapyPlaque progressionNon-Plaque progression
      n = 48n = 28n = 20p-Value
      At baseline
       Glucose (mg/dl)140 ± 29138 ± 23142 ± 370.63
       IRI (μU/mL)6.4 ± 3.96.8 ± 4.15.9 ± 3.70.45
       HOMA-IR2.2 ± 1.42.3 ± 1.32.1 ± 1.40.60
      At follow-up
       Glucose (mg/dl)147 ± 28156 ± 28133 ± 210.004
       IRI (μU/mL)6.6 ± 3.78.3 ± 3.94.4 ± 1.50.001
       HOMA-IR2.5 ± 1.73.2 ± 1.81.4 ± 0.50.001
      ⊿ HOMA-IR0.29 ± 1.900.95 ± 2.00−0.63 ± 1.310.003
      The data represent the mean ± SD or n (%).
      HbA1c, hemoglobin A1c; ⊿HbA1c, increase in hemoglobin A1c; IRI, immunoreactive insulin; HOMA-IR, homeostasis model assessment insulin resistance; ⊿HOMA-IR, increase in homeostasis model assessment insulin resistance.
      Table 2B shows IR analysis results using HOMA-IR among 48 patients without insulin therapy. In the whole cohort, HOMA-IR increased during the follow-up period. Patients with PP had higher fasting insulin levels (8.3 ± 3.9 mU/mL vs. 4.4 ± 1.5 mU/mL, p = 0.001) and HOMA-IR (3.2 ± 1.8 vs. 1.4 ± 0.5, p = 0.001) at follow-up than those without PP. Increase in HOMA-IR (⊿HOMA-IR) was also higher in patients with PP (0.95 ± 2.00 vs. −0.63 ± 1.31, p = 0.008).

       Analysis of coronary plaque volume and characteristics

      Linear regression analysis of 20 cases revealed excellent interobserver and intraobsever correlations for PV (r = 0.94; 95% CI: 0.87–1.01, p < 0.001; r = 0.92; 95% CI: 0.87−0.97, p < 0.001, respectively) (Online Fig. 3). Bland–Altman analysis implied excellent interobserver and intraobserver concordance. The mean values of interobserver and intraobserver differences were 75.5 ± 133.1 mm2 with 95% limits of agreement ranging from −185.4 mm2 to 336.5 mm2 and −10.2 ± 117.5 mm2 with 95% limits of agreement ranging from −240.4 mm2 to 220.1 mm2.
      Table 3 shows changes in coronary plaque volume and characteristics. The PV in the study population increased at follow-up CCTA (⊿PV = 150 ± 460 mm3, p = 0.02). In accordance with PV, CAC score significantly increased at baseline CCTA (from 96 [0, 443] to 127 [13, 559]). CAC score was correlated to total PV at baseline and follow-up (r = 0.45, 95% CI: 0.23−0.62, p = 0.001; r = 0.53, 95% CI: 0.33−0.68, p < 0.001, respectively). However, the differences between groups were not significant at both baseline and follow-up.
      Table 3Analysis of coronary computed tomography angiography. CAC, coronary artery calcium.
      AllPlaque progressionNon-Plaque progression
      Parametersn = 70n = 39n = 31p-Value
      Baseline vessel-analysis
       Total vessel volume (mm3)5189 ± 13255178 ± 14405200 ± 11870.91
       Total lumen volume (mm3)2296 ± 7242376 ± 7962199 ± 6260.37
       Total plaque volume (mm3)2892 ± 6742803 ± 7103000 ± 6200.24
       CAC score96 [0, 443]122 [20, 522]60 [0, 303]0.19
      Follow-up vessel-analysis
       Total vessel volume (mm3)5203 ± 14735501 ± 16084829 ± 12080.06
       Total lumen volume (mm3)2185 ± 7492291 ± 8272044 ± 6240.18
       Total plaque volume (mm3)3020 ± 8323215 ± 9042774 ± 6580.03
       CAC score127 [13, 559]176 [39, 933]78 [8, 445]0.23
      ⊿ Total plaque volume (mm3)150 ± 460438 ± 370−212 ± 245<0.001
      Baseline plaque characteristics
       Fibrous plaque (%)49.9 [47.3, 56.4]52.5 [47.7, 61.4]48.7 [46.7, 52.7]0.04
       Fibro-fatty plaque (%)31.8 [30.0, 34.3]30.3 [28.3, 33.5]32.5 [30.6, 34.9]0.06
       Necrotic core (%)13.8 [10.5, 17.6]12.5 [8.1, 15.0]16.3 [13.2, 18.2]0.003
       Dense calcium (%)1.2 [0.2, 5.7]1.4 [0.3, 6.9]0.8 [0.1, 4.9]0.27
      Follow-up plaque characteristics
       Fibrous plaque (%)48.7 [45.0, 54.1]49.1 [45.0, 53.8]48.3 [44.1, 55.8]0.81
       Fibro-fatty plaque (%)32.0 [26.5, 35.1]30.7 [26.3, 35.0]32.9 [29.3, 36.2]0.12
       Necrotic core (%)14.8 [18.1, 11.1]14.9 [10.9, 18.0]14.7 [11.2, 19.2]0.66
       Dense calcium (%)2.1 [0.4, 8.2]2.5 [0.6, 9.8]1.3 [0.4, 5.1]0.27
      Plaque characteristics change
       ⊿Fibrous plaque (%)−3.0 ± 9.1−5.1 ± 9.0−0.4 ± 8.50.04
       ⊿Fibro-fatty plaque (%)0.2 ± 5.20.2 ± 5.50.2 ± 4.80.98
       ⊿Necrotic core (%)1.4 ± 5.63.3 ± 5.4−1.0 ± 5.00.001
       ⊿Dense calcium (%)1.4 ± 1.71.5 ± 1.71.1 ± 2.20.39
      The data represent the mean ± SD, the medians [interquartile ranges], or n (%) CAC, coronary artery calcium.
      NC was greater in patients without PP than in those with PP (16.3% [13.2%, 18.2%] vs. 12.5% [8.1%, 15.0%], p = 0.003) at baseline CCTA, but the difference was not significant at follow-up CCTA (14.7% [11.2%, 19.2%] vs. 14.9% [10.9%, 18.0%], p = 0.66). Contrarily, FP was greater in patients with PP (52.5% [47.7%, 61.4%] vs. 48.7% [46.7%, 52.7%], p = 0.04). At follow-up CCTA, both patients with and without PP had a similar burden of FP (49.1% [45.0%, 53.8%] vs. 48.3% [44.1%, 55.8%], p = 0.81). The distribution of FFP and DC was similar between groups.
      Patients with PP had a higher increase in NC (⊿NC; 3.3% ± 5.4% vs.−1.0% ± 5.0%, p = 0.001), whereas FP was significantly decreased in patients with PP (−5.1% ± 9.0% vs. −0.4% ± 8.5%, p = 0.04). In the linear regression analysis, ⊿LDL-C was significantly correlated to ⊿NC% (β-coefficient: 0.04, 95% CI: 0.004−0.08, p = 0.03) and inversely correlated to ⊿FP% (β-coefficient: −0.01, 95% CI: −0.1 to −0.01, p = 0.03) (Table 4). In contrast, ⊿HbA1c and ⊿HOMA-IR were not significantly associated with changes in coronary plaque characteristics.
      Table 4Analysis of changes in plaque composition.
      ⊿ Fibrous plaque (%)⊿ Fibro-fatty plaque (%)⊿ Necrotic core (%)⊿ Dense calcium (%)
      β95% CIp-Valueβ95% CIp-Valueβ95% CIp-Valueβ95% CIp-Value
      ⊿HbA1c (%)−0.5−2.6-1.70.680.03−2.3-2.40.980.4−1.0-1.70.58−0.2−0.7-0.20.25
      ⊿LDL-C (mg/dl)−0.1−0.1-−0.010.030.04−0.03-0.10.250.040.004-0.080.03−0.01−0.02-0.0010.07
      ⊿ HOMA-IR0.1−1.6-1.80.910.5−1.3-2.20.610.2−0.8-1.20.68−0.2−0.5-0.10.18
      Statin treatment

      at follow up
      0.8−1.7-3.30.50−2.8−5.5-−0.20.04−0.4−1.9-1.20.610.3−0.2-0.70.28
      CT interval (months)−0.4−0.9-0.10.130.4−0.1-0.90.150.2−0.1-0.50.21−0.01−0.1-0.10.76
      HbA1c, hemoglobin A1c; LDL-C, low-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment insulin resistance; CT, computed tomography.

       Predictors for coronary plaque progression

      The univariate analysis demonstrated that BMI at baseline (kg/m2) (OR: 1.17, 95%CI: 1.03–1.34, p = 0.01), ⊿HbA1c (%) (OR: 2.18, 95%CI: 1.14–4.19, p = 0.02), and LCL-C at baseline (mg/dl) (OR: 0.98, 95%CI: 0.96−0.99, p = 0.03) were associated with PP (Table 5). In the multivariate analysis, independent predictors for PP were BMI at baseline (kg/m2) (OR: 1.23, 95% CI: 1.05–1.45, p = 0.005) and ⊿HbA1c (%) (OR: 3.05, 95%CI: 1.39–6.67, p < 0.001). Other factors including HbA1c level at follow-up and CT scan interval were not independent predictors.
      Table 5Uni- and multivariate analysis of predictors for plaque progression.
      Univariate analysisMultivariate analysis
      ParametersOR95% CIp-ValueOR95% CIp-Value
      Age (years)1.020.96 – 1.080.50
      Male0.730.26 – 2.070.55
      BMI at baseline (kg/m2)1.171.03 – 1.340.011.231.05 – 1.450.005
      Hypertension1.480.57 – 3.850.42
      Dyslipidemia0.680.23 – 2.030.49
      Current smoker1.200.37 – 3.840.76
      Family history1.040.37 – 2.920.95
      ⊿ HbA1c (%)2.181.14 – 4.190.023.051.39 – 6.67<0.001
      ⊿ LDL-C (mg/dl)1.010.99 – 1.020.36
      LDL-C at baseline (mg/dl)0.980.96 – 0.990.030.980.96 – 1.010.15
      Statin treatment at follow up0.770.26 – 2.320.65
      CT scan interval (months)0.970.87 – 1.080.59
      BMI, body mass index; HbA1c, hemoglobin A1c; LDL-C, low-density lipoprotein cholesterol; CT, computed tomography.

      Discussion

      The main findings are as follows: (i) increase in HbA1c level, but not HbA1c level at baseline or average HbA1c level, was an independent predictor of PP in asymptomatic patients with type 2 DM; (ii) although increase in HbA1c level was not correlated to changes in coronary plaque constitution, increase in LDL-C level was associated with increase in NC; and (iii) in patients without insulin therapy, worsening of IR, such as elevation of HOMA-IR, was shown in those with PP.
      Most of the patients fully achieved the goal of primary prevention of CAD; however, approximately 60% of the patients had PP despite the primary prevention. The increase in HbA1c level, not HbA1c level at baseline nor average HbA1c level during the follow-up, was significantly associated with PP. These findings suggest that increase in HbA1c level can be more important in coronary atherosclerosis in DM patients than the absolute HbA1c level, which is in line with the results of prior studies [
      • Kim U.
      • Leipsic J.A.
      • Sellers S.L.
      • Shao M.
      • Blanke P.
      • Hadamitzky M.
      • et al.
      Natural history of diabetic coronary atherosclerosis by quantitative measurement of serial coronary computed tomographic angiography: results of the PARADIGM study.
      ,
      • Shin S.
      • Park H.B.
      • Chang H.J.
      • Arsanjani R.
      • Min J.K.
      • Kim Y.J.
      • et al.
      Impact of intensive LDL cholesterol lowering on coronary artery atherosclerosis progression: a serial CT angiography study.
      ]. Using glycemic-lowering and antihyperlipidemic drugs was not a significant predictor in the present study.
      Several pathogeneses of the impact of DM on coronary artery atherosclerosis have been reported. One important mechanism is hyperglycemic damage, mainly driven by free-radical accumulation, which activates vascular inflammation and endothelial dysfunction [
      • Nishikawa T.
      • Edelstein D.
      • Du X.L.
      • Yamagishi S.
      • Matsumura T.
      • Kaneda Y.
      • et al.
      Normalizing mitochondrial superoxide production blocks three pathways of hyperglycaemic damage.
      ]. Furthermore, IR can be also an important factor in coronary atherosclerosis. IR decreases the effects produced by the normal activity of insulin and causes hyperinsulinemia, resulting in an increased risk of coronary heart disease [
      • Despres J.P.
      • Lamarche B.
      • Mauriege P.
      • Cantin B.
      • Dagenais G.R.
      • Moorjani S.
      • et al.
      Hyperinsulinemia as an independent risk factor for ischemic heart disease.
      ]. A previous study demonstrated the association between a decrease in IR and suppression of coronary atherosclerosis progression [
      • Nissen S.E.
      • Nicholls S.J.
      • Wolski K.
      • Nesto R.
      • Kupfer S.
      • Perez A.
      • et al.
      Comparison of pioglitazone vs glimepiride on progression of coronary atherosclerosis in patients with type 2 diabetes: the PERISCOPE randomized controlled trial.
      ]. This may support our finding that IR elevation, rather than the index status of IR, plays a more important role in PP in asymptomatic patients with DM. With the aim of primary prevention, careful monitoring and management of HbA1c and HOMA-IR may be necessary even in patients with relatively stable glycemic control. Considering the risk of hypoglycemia, the therapy to prevent an increase in HbA1c level may be more effective than that to reduce aggressively the HbA1c level to a certain target. Additionally, the finding of our study may lead to the early detection of CAD in clinical settings. If patients have a gradual increase in HbA1c level, it may be necessary for the clinicians to evaluate silent myocardial ischemia using CCTA or other imaging modalities before increasing CAD severity.
      The large NC size is among the characteristics of vulnerable plaques, which could cause plaque rupture and acute myocardial infarction [
      • Finn A.V.
      • Nakano M.
      • Narula J.
      • Kolodgie F.D.
      • Virmani R.
      Concept of vulnerable/unstable plaque.
      ]. To reduce NC size and stabilize coronary plaque, intensive statin and lipid-lowering therapies are effective in patients with established acute coronary syndrome and stable angina pectoris. However, its effectiveness for asymptomatic patients with non-established CAD is still unclear. In the present study, although neither HbA1c level nor IR had a significant correlation to plaque composition, the decrease in LDL-C level was significantly correlated to NC reduction. Contrarily, LDL-C and statin therapy were not statistically correlated to PP. Furthermore, we could not demonstrate the effectiveness of statin therapy in plaque composition, which may be due to the small sample size, higher statin use, and relatively good LDL-C control. Furthermore, initiation of statin might have negative effects on coronary plaques through changes in glucose homeostasis [
      • Barkas F.
      • Elisaf M.
      • Liberopoulos E.
      • Liamis G.
      • Ntzani E.E.
      • Rizos E.C.
      Atherogenic dyslipidemia increases the risk of incident diabetes in statin-treated patients with impaired fasting glucose or obesity.
      ].
      The PARADIGM investigators evaluated the predictors of PP and the natural course of coronary atherosclerosis using CCTA images of DM patients [
      • Kim U.
      • Leipsic J.A.
      • Sellers S.L.
      • Shao M.
      • Blanke P.
      • Hadamitzky M.
      • et al.
      Natural history of diabetic coronary atherosclerosis by quantitative measurement of serial coronary computed tomographic angiography: results of the PARADIGM study.
      ]. However, the study was retrospective and did not fully assess the serial glycemic control. Thus, to the best of our knowledge, the present work was the first prospective study that evaluated PP using serial CCTA images and frequent monitoring of glycemic control. In addition to the PARADIGM study, Stone et al. evaluated the association between coronary plaque and major adverse cardiovascular events (MACE) using IVUS [
      • Stone G.W.
      • Maehara A.
      • Lansky A.J.
      • de Bruyne B.
      • Cristea E.
      • Mintz G.S.
      • et al.
      A prospective natural-history study of coronary atherosclerosis.
      ]. These previous studies demonstrated that the plaque burden at baseline was an independent predictor for PP or MACE. However, the total PV at baseline was not a predictor for PP in our study. This might be because of the different analysis method. The two previous studies analyzed the lesions identified at baseline assessment using CT or IVUS, whereas we assessed the total PVs of three coronary vessels. Patients with DM have various coronary plaques at multiple stages of atherosclerosis. Then, considering not only the development of identified plaques at baseline but also the new plaque formation and diffuse wall thickening without stenosis, we measured the total PV.
      Compared with another previous study that evaluated the total PVs of the whole three coronary arteries [
      • Auscher S.
      • Heinsen L.
      • Nieman K.
      • Vinther K.H.
      • Logstrup B.
      • Moller J.E.
      • et al.
      Effects of intensive lipid-lowering therapy on coronary plaques composition in patients with acute myocardial infarction: assessment with serial coronary CT angiography.
      ], PV at baseline in the present study was somewhat higher than that of the previous study. There may be two reasons. First, the previous study excluded the PVs at segments with stent or severe calcification, which might have more PVs compared to other segments. Second, patients with DM were approximately 10% of the population in the previous study, which was much less compared to population size in our study.
      CCTA is a well-established non-invasive method for coronary atherosclerosis assessment. SCOT-HEART investigators demonstrated the efficacy of CCTA-guided management in patients with suspected stable CAD [
      • SCOT-HEART Investigators
      CT coronary angiography in patients with suspected angina due to coronary heart disease (SCOT-HEART): an open-label, parallel-group, multicentre trial.
      ]; however, the CCTA-guided management in asymptomatic patients is controversial. In the FACTOR-64 trial [
      • Muhlestein J.B.
      • Lappe D.L.
      • Lima J.A.
      • Rosen B.D.
      • May H.T.
      • Knight S.
      • et al.
      Effect of screening for coronary artery disease using CT angiography on mortality and cardiac events in high-risk patients with diabetes: the FACTOR-64 randomized clinical trial.
      ], using routine CCTA screening and CCTA-guided aggressive therapy in asymptomatic patients with DM could not reduce MACE during approximately the 4-year follow-up. Thus, the routine use of CCTA in asymptomatic patients with DM is not supported. There are several explanations for this negative result. One of the explanations is the relatively short follow-up period and low event-rate to evaluate the effects on clinical events. Thus, we evaluated the serial change in coronary PV using CCTA images during a similar follow-up period to the FACTOR-64 trial. Additionally, we assessed the changes in the coronary plaque composition that were not evaluated in the previous study. Our results demonstrated the natural course under sufficient primary prevention and the characteristics of high-risk patients for coronary atherosclerosis progression. Then, it may be necessary not to apply CCTA-guided therapy to the whole asymptomatic patients, but to focus on those high-risk patients.
      We assessed coronary PVs and their characteristics by using CCTA instead of using invasive coronary imaging modalities. In addition to its advantage of less invasiveness, compared to invasive imaging modalities, CCTA enables a simultaneous assessment of all coronary plaques in the entire coronary tree, unlike invasive coronary imaging [
      • Lee S.E.
      • Villines T.C.
      • Chang H.J.
      Should CT replace IVUS for evaluation of CAD in large-scale clinical trials: effects of medical therapy on atherosclerotic plaque.
      ]. Moreover, the technical aspects of CCTA are highly standardized with clearly established acquisition protocols, and fully or semi-automated software for evaluation of coronary plaques is available without highly trained operators or observers. Contrarily, its disadvantage is its lower spatial resolution compared to those of invasive imaging modalities.
      This study has several limitations. First, this study involving a small sample size was conducted at a single center. Thus, we were unable to identify statistically significant predictors for PP, including LDL-C level and medical therapy such as statin use. Second, the proportion of patients who underwent follow-up CCTA was relatively low. We could not perform the follow-up CCTA in approximately 30% of the patients, including those with cardiovascular events. Although 27 patients refused to undergo follow-up CCTA, we could still follow them up to determine whether they experienced cardiovascular events. Finally, the initiation or precision of glucose-lowering drugs depended on the physicians’ decision. Although the patients were treated according to the current guideline, there might be several biases in the management of these patients.

      Conclusion

      In asymptomatic DM patients, an increase in HbA1c level is correlated to coronary PP, and an increase in LDL-C level is correlated to an increase in NC. In DM patients without insulin therapy, IR worsening is also associated with PP. These findings may suggest that asymptomatic patients with type 2 DM with increasing HbA1c level or worsening IR have a higher risk of coronary atherosclerosis progression. Hence, these patients might need intensive medical treatments for glycemic and lipid control and routine imaging assessment of the coronary artery.

      Conflicts of interest

      The authors report no conflicts of interest.

      Appendix A. Supplementary data

      The following are Supplementary data to this article:

      References

        • Emerging Risk Factors Collaboration
        • Sarwar N.
        • Gao P.
        • Seshasai S.R.
        • Gobin R.
        • Kaptoge S.
        • et al.
        Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies.
        Lancet. 2010; 375: 2215-2222
        • Fox C.S.
        • Coady S.
        • Sorlie P.D.
        • D’Agostino Sr, R.B.
        • Pencina M.J.
        • Vasan R.S.
        • et al.
        Increasing cardiovascular disease burden due to diabetes mellitus: the Framingham Heart Study.
        Circulation. 2007; 115: 1544-1550
        • Wackers F.J.
        • Young L.H.
        • Inzucchi S.E.
        • Chyun D.A.
        • Davey J.A.
        • Barrett E.J.
        • et al.
        Detection of silent myocardial ischemia in asymptomatic diabetic subjects: the DIAD study.
        Diabetes Care. 2004; 27: 1954-1961
        • Soejima H.
        • Ogawa H.
        • Morimoto T.
        • Okada S.
        • Sakuma M.
        • Nakayama M.
        • et al.
        One quarter of total myocardial infarctions are silent manifestation in patients with type 2 diabetes mellitus.
        J Cardiol. 2019; 73: 33-37
      1. Influence of diabetes on 5-year mortality and morbidity in a randomized trial comparing CABG and PTCA in patients with multivessel disease: the Bypass Angioplasty Revascularization Investigation (BARI).
        Circulation. 1997; 96: 1761-1769
        • Poutanen O.
        • Mattila A.
        • Seppala N.H.
        • Groth L.
        • Koivisto A.M.
        • Salokangas R.K.
        Seven-year outcome of depression in primary and psychiatric outpatient care: results of the TADEP (Tampere Depression) II Study.
        Nord J Psychiatry. 2007; 61: 62-70
        • Motoyama S.
        • Ito H.
        • Sarai M.
        • Kondo T.
        • Kawai H.
        • Nagahara Y.
        • et al.
        Plaque characterization by coronary computed tomography angiography and the likelihood of acute coronary events in mid-term follow-up.
        J Am Coll Cardiol. 2015; 66: 337-346
        • de Graaf M.A.
        • Broersen A.
        • Kitslaar P.H.
        • Roos C.J.
        • Dijkstra J.
        • Lelieveldt B.P.
        • et al.
        Automatic quantification and characterization of coronary atherosclerosis with computed tomography coronary angiography: cross-correlation with intravascular ultrasound virtual histology.
        Int J Cardiovasc Imaging. 2013; 29: 1177-1190
        • Boogers M.J.
        • Broersen A.
        • van Velzen J.E.
        • de Graaf F.R.
        • El-Naggar H.M.
        • Kitslaar P.H.
        • et al.
        Automated quantification of coronary plaque with computed tomography: comparison with intravascular ultrasound using a dedicated registration algorithm for fusion-based quantification.
        Eur Heart J. 2012; 33: 1007-1016
        • Kim U.
        • Leipsic J.A.
        • Sellers S.L.
        • Shao M.
        • Blanke P.
        • Hadamitzky M.
        • et al.
        Natural history of diabetic coronary atherosclerosis by quantitative measurement of serial coronary computed tomographic angiography: results of the PARADIGM study.
        JACC Cardiovasc Imaging. 2018; 11: 1461-1471
        • Matthews D.R.
        • Hosker J.P.
        • Rudenski A.S.
        • Naylor B.A.
        • Treacher D.F.
        • Turner R.C.
        Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man.
        Diabetologia. 1985; 28: 412-419
        • Kishi S.
        • Giannopoulos A.A.
        • Tang A.
        • Kato N.
        • Chatzizisis Y.S.
        • Dennie C.
        • et al.
        Fractional flow reserve estimated at coronary CT angiography in intermediate lesions: comparison of diagnostic accuracy of different methods to determine coronary flow distribution.
        Radiology. 2018; 287: 76-84
        • Kato N.
        • Kishi S.
        • Arbab-Zadeh A.
        • Rybicki F.J.
        • Tanimoto S.
        • Aoki J.
        • et al.
        Relative atherosclerotic plaque volume by CT coronary angiography trumps conventional stenosis assessment for identifying flow-limiting lesions.
        Int J Cardiovasc Imaging. 2017; 33: 1847-1855
        • Austen W.G.
        • Edwards J.E.
        • Frye R.L.
        • Gensini G.G.
        • Gott V.L.
        • Griffith L.S.
        • et al.
        A reporting system on patients evaluated for coronary artery disease. Report of the ad hoc committee for grading of coronary artery disease, council on cardiovascular surgery, American Heart Association.
        Circulation. 1975; 51: 5-40
        • Agatston A.S.
        • Janowitz W.R.
        • Hildner F.J.
        • Zusmer N.R.
        • Viamonte Jr, M.
        • Detrano R.
        Quantification of coronary artery calcium using ultrafast computed tomography.
        J Am Coll Cardiol. 1990; 15: 827-832
        • Inaba S.
        • Okayama H.
        • Funada J.
        • Higashi H.
        • Saito M.
        • Yoshii T.
        • et al.
        Impact of type 2 diabetes on serial changes in tissue characteristics of coronary plaques: an integrated backscatter intravascular ultrasound analysis.
        Eur Heart J Cardiovasc Imaging. 2012; 13: 717-723
        • Shin S.
        • Park H.B.
        • Chang H.J.
        • Arsanjani R.
        • Min J.K.
        • Kim Y.J.
        • et al.
        Impact of intensive LDL cholesterol lowering on coronary artery atherosclerosis progression: a serial CT angiography study.
        JACC Cardiovasc Imaging. 2017; 10: 437-446
        • Nishikawa T.
        • Edelstein D.
        • Du X.L.
        • Yamagishi S.
        • Matsumura T.
        • Kaneda Y.
        • et al.
        Normalizing mitochondrial superoxide production blocks three pathways of hyperglycaemic damage.
        Nature. 2000; 404: 787-790
        • Despres J.P.
        • Lamarche B.
        • Mauriege P.
        • Cantin B.
        • Dagenais G.R.
        • Moorjani S.
        • et al.
        Hyperinsulinemia as an independent risk factor for ischemic heart disease.
        N Engl J Med. 1996; 334: 952-957
        • Nissen S.E.
        • Nicholls S.J.
        • Wolski K.
        • Nesto R.
        • Kupfer S.
        • Perez A.
        • et al.
        Comparison of pioglitazone vs glimepiride on progression of coronary atherosclerosis in patients with type 2 diabetes: the PERISCOPE randomized controlled trial.
        JAMA. 2008; 299: 1561-1573
        • Finn A.V.
        • Nakano M.
        • Narula J.
        • Kolodgie F.D.
        • Virmani R.
        Concept of vulnerable/unstable plaque.
        Arterioscler Thromb Vasc Biol. 2010; 30: 1282-1292
        • Barkas F.
        • Elisaf M.
        • Liberopoulos E.
        • Liamis G.
        • Ntzani E.E.
        • Rizos E.C.
        Atherogenic dyslipidemia increases the risk of incident diabetes in statin-treated patients with impaired fasting glucose or obesity.
        J Cardiol. 2019; 74: 290-295
        • Stone G.W.
        • Maehara A.
        • Lansky A.J.
        • de Bruyne B.
        • Cristea E.
        • Mintz G.S.
        • et al.
        A prospective natural-history study of coronary atherosclerosis.
        N Engl J Med. 2011; 364: 226-235
        • Auscher S.
        • Heinsen L.
        • Nieman K.
        • Vinther K.H.
        • Logstrup B.
        • Moller J.E.
        • et al.
        Effects of intensive lipid-lowering therapy on coronary plaques composition in patients with acute myocardial infarction: assessment with serial coronary CT angiography.
        Atherosclerosis. 2015; 241: 579-587
        • SCOT-HEART Investigators
        CT coronary angiography in patients with suspected angina due to coronary heart disease (SCOT-HEART): an open-label, parallel-group, multicentre trial.
        Lancet. 2015; 385: 2383-2391
        • Muhlestein J.B.
        • Lappe D.L.
        • Lima J.A.
        • Rosen B.D.
        • May H.T.
        • Knight S.
        • et al.
        Effect of screening for coronary artery disease using CT angiography on mortality and cardiac events in high-risk patients with diabetes: the FACTOR-64 randomized clinical trial.
        JAMA. 2014; 312: 2234-2243
        • Lee S.E.
        • Villines T.C.
        • Chang H.J.
        Should CT replace IVUS for evaluation of CAD in large-scale clinical trials: effects of medical therapy on atherosclerotic plaque.
        J Cardiovasc Comput Tomogr. 2019; 13: 248-253