The role of insulin resistance and metabolic

hoa,

rea

Article history:Received 2 February 2015Received in revised form 15 April 2015Accepted 18 April 2015Available online 22 April 2015

Keywords:Insulin resistanceNecrotic coreVirtual histology-intravascular ultrasound

ns could be a potentialstratication [3].ysiological role in ath-to adverse cardiovas-

International Journal of Cardiology 190 (2015) 5662

Contents lists available at ScienceDirect

International Journ

j ourna l homepage: www.e lcular outcome [410]. It has been reported that metabolic syndromewas associated with the lipid-rich plaque in non-culprit coronary le-sions and lesions in pre-intervention on three coronary vessels [11,12]

area;DC,densecalcium;EEM,externalelasticmembrane;FBS, fastingbloodsugar;FF,bro-fatty;HbA1c,glycatedhemoglobin;hsCRP,highsensitivityC-reactiveprotein; IFG, impairedfastingglucose; IR, insulinresistance; IVUS, intravascularultrasound;MI,myocardial infarc-severity and spasm lead to acute coronary syndrome [1]. Recently, spec-tral analysis of virtual histology intravascular ultrasound (VH-IVUS)

events and to quantify the amount of NC in lesiomeasure for plaque vulnerability and further risk

Insulin resistance (IR) plays a major pathopherosclerotic cardiovascular diseases and is related

Abbreviations:BMI,bodymass index;CAD, coronaryarterydisease;CSA,crosssectional1. Introduction

Vulnerable plaques are high-risk atherosclerotic lesions and compli-cations of these plaques such as plaque rupture, luminal and muralthrombosis, intraplaque hemorrhage, rapid progression in stenosis

radiofrequency data has demonstrated potential to provide detailedquantitative information on plaque morphology and component:brous, bro-fatty, dense calcium and lipid-rich necrotic core (NC) andhas been validated in studies of explanted human coronary segments[2]. Unstable, lipid-rich plaques are believed to play a key role in thesetion; NC, necrotic core; QUICKI, using quantitative insulingion of interest; TCFA, thin cap broatheroma; VH-IVUS,ultrasound. Corresponding author at: Gachon University Gil Med

Namdong-gu, Incheon 405-760, South Korea.E-mail address: shhan@gilhospital.com (S.H. Han).

http://dx.doi.org/10.1016/j.ijcard.2015.04.1630167-5273/ 2015 Elsevier Ireland Ltd. All rights reservedable plaque pathogenesis. 2015 Elsevier Ireland Ltd. All rights reserved.was an independent factor for the NC at the ROI (beta coefcient = 0.44, p = 0.003), but QUICKI was not (betacoefcient =0.01, p = 0.94).Conclusions: Instead of a singlemeasurement of IR index or eachmetabolic risk factor, clustering of risk factors forIR plays an important role on plaque vulnerability.Condensed abstract:We investigated the role of insulin resistance (IR) on culprit coronary plaque. Patientswith IRhad a greater amount of necrotic core in culprit coronary lesions than in patients without IR. Rather than a singlemeasurement of IR index or each metabolic risk factor, clustering of metabolic risk factors for IR plays an impor-tant role in plaque vulnerability in patientswith coronary artery disease. Our study demonstrates the role of IR onculprit coronary plaque and highlights the importance of the clustering of metabolic risk factors for IR in vulner-Background:Detailed relationships between insulin resistance (IR) and vulnerable plaque are not clear, therefore,we sought the role of IR and metabolic risk factors on culprit coronary plaque.Methods: Plaque components at a region of interest (ROI, 10mm)were analyzed by virtual histology intravascu-lar ultrasound. IR was dened as quantitative insulin sensitivity check index (QUICKI) 0.33. Seven metabolicrisk factors (5 risk factors for metabolic syndrome dened by ATP III, history of smoking, and hsCRP) for IRwere determined.Results: The data for 150 (males 104) patients were analyzed. Patients with IR (n=69) had greater necrotic core(NC) at the ROI (21.2 15.8mm3 vs 15.7 11.9mm3, p= 0.02) than in patients without IR (n= 81). The NC atthe ROI was correlated with QUICKI (r =0.16, p = 0.05), HbA1c (r = 0.24, p b 0.01), body mass index (r =0.17, p= 0.04), presence of diabetes mellitus (r= 0.29, p b 0.001), hsCRP (r= 0.17, p= 0.04) and the numbersof risk factors for IR (r=0.41, p b 0.001). Themultivariate analysis revealed that the numbers of risk factors for IRa b s t r a c ta r t i c l e i n f ocoronary plaque

Yae Min Park a, Seung Hwan Han a,, Jong Goo Seo a, SiKyounghoon Lee a, Soon Yong Suh a, Woong Chol Kanga Division of Cardiovascular Disease, Gachon University Gil Hospital, Incheon, South Koreab Division of Endocrinology and Metabolism, Gachon University Gil Hospital, Incheon, South Kosensitivity check index; ROI, re-virtual histology intravascular

ical Center, 1198 Kuwol-dong,

.risk factors on culprit

on Lee b, Pyung Chun Oh a, Kwang Kon Koh a,Taehoon Ahn a, In Suck Choi a, Eak Kyun Shin a

al of Cardiology

sev ie r .com/ locate / i j ca rdand hyperinsulinemia and abnormal glucose regulationwere associatedwith lipid rich coronary plaque by intracoronary imaging methods [10,13]. However, there is no data for the independent role of IR indexand detailed relationships between IR including metabolic risk factorsand plaque vulnerability in culprit coronary lesions. In the present

57Y.M. Park et al. / International Journal of Cardiology 190 (2015) 5662study, we compared the plaque characteristics between patientswith IRandwithout IR and sought the role of IR includingmetabolic risk factorson culprit coronary plaque.

2. Patients and methods

2.1. Patients and study design

This study was a cross sectional study which used the data fromthe registry in patients with coronary artery disease (diameterstenosis N 50%) who underwent VH-IVUS before percutaneous coro-nary intervention by operators' discretion. Between August 2008and September 2011, 258 consecutive patients were enrolled. Weexcluded patients with previous coronary stent insertion, manualpullback of IVUS catheter, coronary artery bypass graft failure, andpatients with inadequate IVUS images. Also patients with a left ven-tricular ejection fraction of less than 35% and severe hepatic andrenal disease were excluded. To avoid the effects of insulin on insu-lin sensitivity index, diabetic patients who required insulin treat-ment were also excluded. In patients who underwent multi-vesselVH-IVUS, the lesion with the worst diameter stenosis and morecomplex morphology was selected for VH-IVUS analysis. Finally,150 patients with analyzable IVUS images of native coronary ves-sels, pullback length greater than 10 mm, and with complete clinicaland laboratory value, were enrolled to this analysis. Hospital recordsof patients were reviewed to obtain information on clinical demo-graphics. The local Institutional Review Board approved this study,and written informed consents were obtained from all patients.

2.2. IVUS procedure and analysis

Before the performance of gray-scale and VH-IVUS examination, pa-tients were administered an intracoronary 0.2 mg of nitroglycerin toprevent coronary spasm. A 20-MHz, 2.9 F IVUS imaging catheter(Eagle Eye, Volcano Corp., Rancho Cordova, CA, USA) was advancedmore than 10mmbeyond the lesion; and automated pull-backwas per-formed to a point more than 10mmproximal to the lesion at a speed of0.5 mm/s.

Quantitative volumetric gray-scale and VH-IVUS analyses were per-formed across the entire lesion segment, and cross-sectional analysiswas performed at the region of interest (ROI) and at the minimallumen site. The IVUS region of interest (ROI) was the most diseased10 mm segment, identied by summarizing plaque volume in contigu-ous cross sections over an axial distance of 10 mm. Therefore, the seg-ment with the greatest plaque volume constituted the most diseased10 mm.

Conventional quantitative volumetric gray-scale IVUS analysis wasperformed according to the American College of Cardiology Clinical Ex-pert Consensus Document on Standards for Acquisition, Measurementand Reporting of Intravascular Ultrasound Studies [14]. External elasticmembrane (EEM) and lumen cross-sectional areas (CSAs) were identi-ed using automatic edge detection and manually corrected when nec-essary. Plaque plusmedia CSAwas calculated as EEMminus lumen CSA;and plaque burdenwas calculated as plaque plusmedia divided by EEMCSA.

The IVUS-VH data were stored on a CD-ROM for ofine analysis.Subsequently, VH-IVUS analysis classied the color-coded tissue intofour major components: (brous [F, labeled green], bro-fatty [FF, la-beled greenish-yellow], dense calcium [DC, labeled white] and necroticcore [NC, labeled red]) [3]. VH-IVUS analysis was reported as absoluteplaque amounts and as percentages (relative amounts). Thin-capbroatheroma (TCFA) was dened as focal, NC-rich (10% of the CSA)plaques being in contact with the lumen in a plaque burden 40%over three consecutive frames. Analyses were conducted by 2 indepen-

dent investigators unaware of the clinical data and the mean value wascalculated. Inter-observer correlations was excellent, with correlationcoefcients (r) being 0.90.

2.3. Coronary risk factors and lipids, metabolic parameters

Diabetes mellitus was dened as fasting glucose 126 mg/dL or 2 hpostprandial glucose 200 mg/dl or glycated hemoglobin (HbA1c) 6.5%, or if they were already being treated for this condition. Hyperten-sion was dened as systolic blood pressure 140 mm Hg or diastolicblood pressure 90 mm Hg, or if they were already being treated forthis condition. Body mass index (BMI) was calculated as body weightin kilograms divided by the square of height in meters (kg/m2).

Blood samples for laboratory assays were obtained at the timeof coronary angiography following overnight fasting for at least 8 h.Total cholesterol and triglycerides were analyzed with enzymaticmethods (Shinyang Chemical, Seoul, Korea), and high density lipopro-tein (HDL) cholesterol by a direct immunoinhibition method (WakoPure Chemical, Osaka, Japan). LDL cholesterol was calculated using theFriedewald equation [15]. Fasting blood sugar (FBS) was determinedby the hexokinase method (Shinyang Chemical, Seoul, Korea) using aHitachi 7600-110. Assays for plasma insulin levels were performed byimmunoradiometric assay (INSULIN-RIABEAD II, TFB, Inc., Tokyo,Japan). Assays for glycated hemoglobin (HgA1c) were measured byhigh performance liquid chromatography assay (VARIANT II TUR BO,BIORAD, Inc., Hercules, California). High sensitivity C-reactive protein(hsCRP) levels were determined with a turbidimetric assay (DenkaSeiken, Tokyo, Japan) using the Hitachi 7600-110. History of smokingwas obtained from all patients.

2.4. Insulin resistance (IR) index, numbers of risk factors for IR

IR index was determined from plasma glucose and insulin con-centrations, using quantitative insulin sensitivity check index (QUICKI)and calculated by using the formula; 1 / (log insulin (U/ml) + log glu-cose (mg/dL)) [16]. Patientswith IRwas dened asQUICKI 0.33 by theprevious studies [17,18].

Numbers of risk factors for IR (07) were derived from the sum ofrisk factors which were related with IR from the previous reports[1924]. These include 1) high BMI N 25 (kg/m2), 2) impaired fastingglucose (IFG, FBS 110 mg/dL) or diabetes mellitus, 3) hypertension,4) hypertriglyceridemia (triglyceride 150mg/dL), 5) lowHDL choles-terol (male b 40 mg/dL, female b 50 mg/dL), 6) history of smoking, and7) high hsCRP N 1.0 mg/L. The diagnosis for metabolic syndrome wasmade by patientswho hadmore than 3 risk factors among 15) risk fac-tors for IR. Each criterion for metabolic syndrome was slightly modiedfor this study.

2.5. Cardiovascular diagnosis

Acute coronary syndromes included unstable angina, non-ST ele-vation myocardial infarction (MI), or ST elevation MI according toAmerican College of Cardiology/American Heart Association guide-lines [25]. The diagnosis of acute MI was based on elevation of atleast 1 positive biomarker (creatine kinase, creatine kinase-MB, ortroponin T), characteristic electrocardiogram changes, and a historyof prolonged acute chest pain. Unstable angina pectoris was denedas either angina with a progressive crescendo pattern or angina thatoccurred at rest. Stable angina pectoris was dened as no change inthe frequency, duration, or intensity of symptoms within 4 weeksbefore the intervention.

2.6. Data analysis

Patients were divided into two groups according to the presenceof IR. Variables were analyzed to compare the characteristics of pa-

tients with or without IR. Continuous variables were expressed as

3.4. Independent predictors for the amount of NC at the ROI

The multivariate linear regression analysis (including all possibleparameters and the number of risk factors for IR) showed that numberof risk factors for IR was an independent factor for the amount of NC atthe ROI (beta coefcient= 0.44, p= 0.003). In othermodels (includingall possible parameters and the presence of metabolic syndrome), the

Table 1Baseline clinical characteristics.

IR pValue

Yes (n = 69) No (n = 81)

QUICKI 0.298 0.023 0.362 0.028 0.00

Age (years) 58.7 12.2 61.1 10.8 0.19Male, n (%) 48 (69.6%) 56 (69.1%) 0.96Clinical history, n (%)Diabetes mellitus 25 (36.2) 17 (21.0) 0.04Hypertension 43 (62.3) 41 (50.6) 0.15Current or ex-smoker 30 (43.4) 27 (33.3) 0.20

Height (cm) 163.4 9.3 163.8 8.3 0.80Weight (kg) 69.2 10.8 64.3 9.6 0.004

2

Insulin (mU/L) 22.7 17.4 6.7 2.5 0.00

Lipid levels (mg/dL)Total cholesterol 173.0 33.6 172.4 38.6 0.92Triglyceride 162.0

(91.8227.0)130.0(96.0205.0)

0.21

HDL cholesterol 43.0 10.7 44.0 11.1 0.55Non-HDL cholesterol 129.8 34.0 128.4 37.1 0.81LDL cholesterol 94.3 31.9 96.8 38.8 0.67

Hs CRP (mg/L) 2.0 (0.710.7) 1.6 (0.67.4) 0.39Hemoglobin (g/dL) 13.7 1.9 13.6 1.8 0.67White blood cells (103/mm3) 8.04 2.65 7.33 2.56 0.10Platelet count (103/mm3) 248.0 59.8 248.4 65.8 0.97Creatinine (mg/dL) 1.31 2.10 0.94 0.45 0.16Metabolic syndrome (%) 63.8 34.6 0.00

Numbers of risk factors for IR(number)

3.6 1.3 2.6 1.3 0.00

Culprit vessel 0.63Left main 1 (1.5) 2 (2.5)Left anterior descending 35 (50.7) 46 (56.8)Left circumex 11 (15.9) 16 (19.8)Obtuse marginal 2 (2.9) 1 (1.2)Right 20 (29.0) 16 (19.8)

Data are means SD or number (%) or median (25 percentile75 percentile).IR = insulin resistance; QUICKI = quantitative insulin sensitivity check index; BMI =body mass index; MI = myocardial infarction; BNP = brain natriuretic peptide; HDL =

high- -density lipoprotein; LDL = low- -density lipoprotein; HsCRP = high sensitivity Creactive protein. p b 0.00.

Table 2Data for conventional intravascular ultrasound.

IR p Value

Yes (n = 69) No (n = 81)

ROIEEM volume (mm3) 167.8 60.8 153.1 45.2 0.13Lumen volume (mm3) 54.4 17.8 57.1 22.4 0.46P & M volume (mm3) 113.4 51.4 96.0 35.3 0.03Plaque volume (%) 65.6 9.7 62.3 10.7 0.08

Data are means SD.IR= insulin resistance; ROI= region of interest; EEM= external elastic membrane; CSA=

58 Y.M. Park et al. / International Journal of Cardiology 190 (2015) 5662mean SD or median (25 percentile75 percentile) and comparedby Student t test or MannWhitney U test to evaluate differencesbetween mean values. Categorical variables were expressed as per-centages and frequencies and compared by chi-square test or Fisherexact test as appropriate. Correlations between the levels of risk fac-tors and the VH-IVUS derived amount of NC were tested usingPearson's coefcient of correlation. To determine the independentparameters for the amount of NC, multivariate linear regressionanalysis was performed. Values of p b 0.05 were considered signi-cant. All tests were 2-sided.

3. Results

3.1. Study population and baseline characteristics (Table 1)

Data for a total 150 consecutive patients (104 males with a medianage of 61 11.5, range 30 to 85) were analyzed. The prevalence of IRwas 69 patients (46.0%) among the patients. The baseline clinical char-acteristics are summarized in Table 1. Forty two patients (28%) had pre-vious history of diabetes mellitus and the proportion was signicantlyhigher in patients from the group with IR than without IR (36.2% vs.21.0%, p = 0.04). BMI was signicantly higher in patients with IR(25.9 3.3 kg/m2 vs 23.9 2.5 kg/m2, p b 0.001). The proportion ofthe patients with acute coronary syndrome was similar between twogroups. Patients with IR showed higher triglyceride and hsCRP withoutstatistical signicance. Otherwise, there were no signicant clinical andlaboratory differences between two groups except for glucose, insulinlevel and IR index. The proportion of metabolic syndrome and the num-bers of risk factors for IR were signicantly greater in patients with IRthan in patients without IR (63.8% vs 34.6%, p b 0.001, 3.6 1.3 vs2.6 1.3, p b 0.001, respectively).

3.2. Quantitative parameters of gray-scale and VH-IVUS

Gray-scale IVUS ndings are summarized in Table 2. The representa-tive cases of VH-IVUS ndings on culprit coronary lesion in patientswith IR and without IR are illustrated in Fig. 1A and B. The volume oftotal plaque plus media at the ROI was signicantly greater in patientswith IR than in patients without IR (113.4 51.4 mm3 vs 96.0 35.3 mm3, p = 0.03) although plaque burden (%) did not differ signi-cantly between two groups (Table 2).

At the ROI, the volume of brous tissue (48.2 24.8mm3 vs. 38.7 19.7 mm3, p = 0.01) and NC (21.2 15.8 mm3 vs. 15.7 11.9 mm3,p = 0.02) was signicantly greater in patients with IR than in patientswithout IR. But the volume of FF and DC were comparable betweenthe two groups (Fig. 2).

3.3. Correlation betweenmetabolic, lipid parameters and the amount of ne-crotic core

The amount of NC at the ROI had borderline correlation with thelevel of QUICKI (r =0.16, p = 0.05, Fig. 3A), and signicant correla-tionwithHbA1c (r= 0.24, p b 0.01), BMI (r= 0.17, p=0.04), presenceof diabetes mellitus (r = 0.29, p b 0.001), hsCRP (r = 0.17, p = 0.04)and the numbers of risk factors for IR (r = 0.41, p b 0.001, Fig. 3B). Alllipid proles were not correlated with the volume of NC at the ROI(0.06 r 0.05, 0.15 p 0.84).

In patients with IR, the amount of NC at the ROI was signicantlycorrelatedwith the level of HbA1c (r=0.27, p=0.03), but not correlat-ed with other metabolic and lipid parameters (0.05 r 0.13,0.30 p 0.95).

In patients without IR, the amount of NC at the ROI was signicantlycorrelated with the levels of BMI (r= 0.22, p b 0.05), but not correlatedwith other metabolic and lipid parameters (0.12 r 0.34,

0.07 p 0.86).BMI (kg/m ) 25.9 3.3 23.9 2.5 0.00Cardiovascular diagnosis, n (%) 0.35Acute coronary syndrome 38 (55.1) 35 (43.2)ST segment elevation MI 9 (13.0) 7 (8.6)Non-ST segment elevation MI 14 (20.3) 11 (13.6)Unstable angina pectoris 15 (21.7) 17 (21.0)

Stable angina pectoris 31 (44.9) 46 (56.8)Ejection fraction (%) 60.0 12.6 61.1 11.3 0.56Pro BNP (pg/mL) 122 (47318) 124 (49300) 0.66Fasting blood sugar (mg/dL) 137.7 60.0 100.3 24.2 0.00cross sectional area; P & M= plaque plus media.

59Y.M. Park et al. / International Journal of Cardiology 190 (2015) 5662Apresence of metabolic syndrome and history of smokingwere indepen-dent predictors for the amount of NC at the ROI (beta coefcient= 0.35,0.33, p = 0.02, 0.02, respectively). However, QUICKI was not an inde-pendent predictor for the amount of NC at the ROI in two models(beta coefcient =0.03,0.21, p = 0.83, 0.16, respectively).

3.5. Correlation between the presence of metabolic risk factors and plaque(Table 3)

Table 3 revealed the correlations between the presence of metabolicrisk factors and VH-IVUS derived plaque. Plaque plus media volume atthe ROI signicantly correlated with the presence of IR (r = 0.21,p b 0.01), IFG or diabetes mellitus (r = 0.17, p b 0.01), metabolic syn-drome (r = 0.22, p b 0.01) and the numbers of risk factors for IR (r =0.22, p b 0.01). Plaque burden % signicantly correlatedwith the presenceof IFG or diabetes mellitus (r = 0.20, p b 0.05) and metabolic syndrome

B

Fig. 1.The representative cases of VH IVUSndings on culprit coronary lesion in patientswith IRat themid LAD artery. VH-IVUS showed that the amount of NC at the ROIwas 27.3mm3 and NCstenosis at themid LAD. VH-IVUS revealed that the amount of NC at the ROIwas 5.2mm3 and Nresistance; LAD = left anterior descending; NC = necrotic core; ROI = region of interest; MLD(r = 0.22, p b 0.01) and the numbers of risk factors for IR (r = 0.22,p b 0.01).

The amount of NC at the ROI was signicantly correlated with thepresence of IR (r = 0.20, p b 0.05), BMI (r = 0.18, p b 0.05), IFG or dia-betes mellitus (r = 0.18, p b 0.05), hypertension (r = 0.22, p b 0.01),history of smoking (r = 0.30, p b 0.001), and metabolic syndrome(r = 0.32, p b 0.001) and the number of risk factors for IR (r = 0.41,p b 0.001).

The correlation between the presence of metabolic syndrome andthe amount ofNC at the ROIwas also signicant in subgroup analysis ac-cording to the presence of IR (r= 0.30, p= 0.01 in patients with IR, r=0.27, p=0.02 in patientswithout IR). Similarly, the correlation betweenthe number of risk factors and the amount of NC at the ROIwas also sig-nicant in subgroup analysis whether the patients had IR (r=0.31, p=0.01) or not (r = 0.46, p b 0.001).

In addition, the amount of DC at the ROI was signicantly correlatedwith the presence of IFG or diabetes mellitus (r = 0.20, p b 0.05),

andwithout IR. (A) In a patientwith IR, coronary angiography revealed signicant stenosisatMLDwas 4.5mm2. (B) In a patientwithout IR, coronary angiography showed signicantC atMLDwas 0.7mm2. VH-IVUS= virtual histology intravascular ultrasound; IR= insulin= minimal luminal diameter.

plaques and adverse cardiovascular outcome are not surprising because

Fig. 2. Comparisons of VH-IVUS parameters between patients with IR and without IR. The

Table 3Correlations between insulin resistance index, metabolic risk factors and plaque.

P plus Mvolume

Plaqueburden(%)

NCvolume

DCvolume

FFvolume

Fvolume

Insulin resistance 0.21 0.16 0.20 0.11 0.08 0.21

High BMI 0.09 0.07 0.18 0.13 0.09 0.11IFG or diabetes mellitus 0.17 0.20 0.18 0.20 0.03 0.15Hypertension 0.10 0.12 0.22 0.25 0.11 0.06Hypertriglyceridemia 0.04 0.02 0.12 0.05 0.08 0.04Low HDL cholesterol 0.03 0.04 0.11 0.002 0.08 0.05History of smoking 0.16 0.02 0.30 0.17 0.10 0.11hsCRP N 1.0 mg/dL 0.06 0.08 0.10 0.08 0.05 0.08Metabolic syndrome 0.22 0.22 0.32 0.25 0.04 0.18Numbers of risk factorsfor IR

0.22 0.22 0.41 0.31 0.17 0.21

Data are Pearson's correlation coefcients.P plusM= plaque plusmedia; NC= necrotic core; DC= dense calcium; FF= brofatty;F = fatty; BMI = body mass index; IFG = impaired fasting glucose; HDL = high density

60 Y.M. Park et al. / International Journal of Cardiology 190 (2015) 5662hypertension (r = 0.25, p b 0.01), history of smoking (r = 0.17,p b 0.05), metabolic syndrome (r = 0.25, p b 0.01) and the numbersof risk factors for IR (r = 0.31, p b 0.001).

3.6. Thin cap broatheroma (TCFA)

The presence of TCFA was not signicantly different whether thepatients had IR or not, however patients with metabolic syndromeshowed a signicantly higher incidence of TCFA than patients withoutmetabolic syndrome [(40/72) 55.6% vs. (28/78) 35.9%, p = 0.016]. Bythe numbers of risk factors for IR, the presence of TCFAwas signicantlydifferent (number 0=0%, number 1=10%, number 2=46.7%, number3 = 51.4%, number 4 = 50%, number 5 = 68.4%, number 6 = 50%,p b 0.001).

4. Discussion

The current study investigated the role of IR and metabolic riskfactors on coronary plaque vulnerability which was assessed by theamount of NC in culprit coronary lesions using VH-IVUS. Patients withIR showed a greater amount of NC at the ROI compared with patients

patients with IR had a greater amount (volume) of brous tissue, and NC at the ROI thanin patients without IR. VH-IVUS = virtual histology intravascular ultrasound; IR = insulinresistance; ROI = region of interest; F = brous; FF = bro-fatty; NC = necrotic core;DC = dense calcium. p b 0.05.without IR. The amount ofNC in culprit coronary plaquewas signicant-ly correlatedwith the levels of HbA1c, C-reactive protein, BMI, the pres-ence of metabolic syndrome and the numbers of risk factors for IR andcorrelated with IR index with borderline signicance. Of interest, thepresence of metabolic syndrome and the numbers of risk factors for IR

0.20 0.25 0.30 0.35 0.40 0.45 0.50

0

20

40

60

80

100

QUICKI

Volu

me

of N

C at

RO

I (mm3

) r= -0.160 p= 0.051

Volu

me

of N

C at

RO

I (mm3

)

A B

Fig. 3.Correlations of IR index (A) and the numbers of risk factors for IR (B)with the amount of Nof interest; QUICKI = quantitative insulin sensitivity check index.were independent predictors for the amount of NC in culprit coronaryplaque. Although previous studies have shown the relationships be-tween IR and plaque vulnerability, our study is the rst to report thatclustering of risk factors for IR has amore important role for plaque vul-nerability rather than each risk factor for IR including IR index.

Previous pathologic studies suggested that the decisive factor deter-mining plaque vulnerability was plaque composition, rather than thedegree of luminal narrowing [26]. The culprit lesion in patients withacute coronary syndrome was shown to be a relatively minor stenosis(b50% of the percentage diameter stenosis) [27]. Furthermore, the NCof atherosclerotic plaques was meaningful for ow restoration andST-segment elevation resolution after ST-segment elevationmyocardialinfarction [28,29]. Early detection of vulnerable plaque before rupture isan important clinical goal for the prevention of catastrophic events suchas acute coronary syndrome or sudden death and can be a guide for anadjunctive pharmacological or device-based treatment plan [30]. In ourcurrent study, we assessed plaque vulnerability by the amount of NC inthe culprit coronary lesion using VH-IVUS.

Regarding the associations of IR with increasing atherosclerotic

lipoprotein cholesterol; hsCRP = high sensitivity C reactive protein, IR = insulinresistance. p b 0.05. p b 0.01. p b 0.001.IR has been well established as risk factors for cardiovascular disease.However, few data are available for assessing the associationwith plaquecomponents and IR. Hyperinsulinemia and abnormal glucose regulationwas associated with lipid rich coronary plaque by intracoronary imagingmethods [10,13]. Similarly, patients with abnormal glucose regulation,

0 1 2 3 4 5 6 7

0

20

40

60

80

100

The numbers of risk factors for IR

r= 0.412P

61Y.M. Park et al. / International Journal of Cardiology 190 (2015) 5662including impaired glucose regulation and diabetes mellitus, presentedmore lipid rich plaque which may be related to the increased IR [10].Taken together, there are still limited data for the detailed relationshipsbetween metabolic risk factors and plaque components in patients withcoronary artery disease (CAD) and there are no data for the independentrole of IR index on culprit coronary artery plaque. Therefore,we comparedthe plaque characteristics between patients with IR and without IR andassessed the detailed relationships between metabolic risk factors andthe amount of VH-IVUS derived NC in culprit coronary artery plaque inpatients with CAD.

As we expected, patients with IR had a greater amount of NC than inpatients without IR. In addition, each risk factor for IR such as the levelsof HbA1c, hsCRP, and BMI was signicantly associated with the amountof NC in culprit coronary plaque. However, the correlation between IRindex and the amount of NC showed borderline signicance. Interest-ingly, our study demonstrated that the clustering of risk factors for IRin terms of the presence of metabolic syndrome and greater numbersof risk factors for IR were independent risk factors for the amount ofNC in culprit coronary plaque. These results correspond with a recentstudy using Integrated Backscatter-IVUS or VH-IVUS, on the impact ofmetabolic syndrome identied by National Cholesterol Education Pro-gram in Adult Treatment Panel III criteria on tissue characteristics ofthe coronary plaques which showed that patients with metabolic syn-drome had a signicantly higher prevalence of lipid-rich plaque [11,12]. In our current study, we also found that the correlation co-efciency of the numbers of risk factors for IR with the amount of NCwas greater than in the presence of metabolic syndrome. Our resultssuggest that the increased clustering of metabolic risk factors representa greater possibility of plaque vulnerability. These ndings are also sup-ported by patients with metabolic syndrome showing a greater inci-dence of TCFA than in patients without metabolic syndrome in ourcurrent study.

Taken together, our results demonstrated that the clustering of riskfactors for IR plays an important role on culprit coronary plaque vulner-ability instead of a single measurement of IR index or each metabolicparameter.

In addition, the amount of DC at the ROI signicantly correlatedwiththe presence of IFG or diabetes mellitus, hypertension, history ofsmoking, metabolic syndrome and the numbers of risk factors for IR.In these correlations, correlation co-efciency with DC at the ROI andthe numbers of risk factors for IR was much greater. These results areconsistent with previous ndings which showed the presence of highcalcication in patients with diabetes and metabolic syndrome [31,32]. This result also suggests that patients with abnormal metabolicrisk factors may have a higher incidence of cardiovascular event ratewhich was demonstrated by greater cardiovascular event rates in pa-tients with a high calcication score by CT study [33].

Our study has some limitations. First, this was a retrospective analy-sis. Our analysis consisted of one vessel per patient, and a full segment ofone coronary tree was not evaluated. We dened ROI as the most dis-eased 10 mm segment which comprises only a small part of the entirecoronary arteries and thus selection of the ROI might include somebias. Our results of plaque components were not conrmed by histolo-gy. Second, this single center study had a relatively small number of pa-tients, thus possibly posing a risk of patient selection bias. A similarrelative amount of each plaque component at the ROI in patients withIR andwithout IR can also be caused by a relatively small number of pa-tients. Third, a single measurement of IR index may not represent thelong term status of IR. Lastly, there are some differences in patient char-acteristics between two groups according to the presence of IR. Howev-er, the intimate relationship is well known between insulin resistanceand metabolic syndrome including its risk factors. A large number ofprospective studies are warranted to conrm these data in the future.

In conclusion, patients with IR had a greater amount of NC in culpritcoronary lesions than in patients without IR. Rather than a single mea-

surement of IR index or metabolic risk factors, clustering of risk factorsfor IR plays an important role in plaque vulnerability in patients withcoronary artery disease.

Conicts of interest

The authors declared no conict of interest.

Acknowledgments

This was partly supported by the unrestricted grant from Sano-Aventis Korea (20115113). The authors of this manuscript havecertied that they comply with the Principles of Ethical Publishing inthe International Journal of Cardiology.

References

[1] R. Virmani, A.P. Burke, A. Farb, F.D. Kolodgie, Pathology of the vulnerable plaque, J.Am. Coll. Cardiol. 47 (2006) C13C18.

[2] A. Nair, B.D. Kuban, E.M. Tuzcu, P. Schoenhagen, S.E. Nissen, D.G. Vince, Coronaryplaque classication with intravascular ultrasound radiofrequency data analysis,Circulation 106 (2002) 22002206.

[3] G.W. Stone, A. Maehar, A.J. Lansky, et al., PROSPECT investigators, N. Engl. J. Med.364 (2011) 226235.

[4] R.A. DeFronzo, E. Ferrannini, Insulin resistance. A multifaceted syndrome responsi-ble for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovas-cular disease, Diabetes Care 14 (1991) 173194.

[5] S.H. Han, I. Sakuma, E.K. Shin, K.K. Koh, Antiatherosclerotic and anti-insulin resis-tance effects of adiponectin; basic and clinical studies, Prog. Cardiovasc. Dis. 52(2009) 126140.

[6] S.H. Han, M.J. Quon, J.A. Kim, K.K. Koh, Adiponectin and cardiovascular disease; re-sponse to therapeutic intervention, J. Am. Coll. Cardiol. 49 (2007) 531538.

[7] S.H. Han, M.J. Quon, K.K. Koh, Reciprocal relationships between abnormal metabolicparameters and endothelial dysfunction, Curr. Opin. Lipidol. 18 (2007) 5865.

[8] K.K. Koh, S.H. Han, M.J. Quon, Inammatory markers and the metabolic syndrome:insights from therapeutic interventions, J. Am. Coll. Cardiol. 46 (2005) 19781985.

[9] M. Suzuki, K. Shinozaki, A. Kanazawa, et al., Insulin resistance as an independent riskfactor for carotid wall thickening, Hypertension 28 (1996) 593598.

[10] T. Amano, T. Matsubara, T. Uetani, et al., Abnormal glucose regulation is associatedwith lipid-rich coronary plaque: relationship to insulin resistance, JACC Cardiovasc.Imaging 1 (2008) 3945.

[11] T. Amano, T. Matsubara, T. Uetani, et al., Impact of metabolic syndrome on tissuecharacteristics of angiographically mild to moderate coronary lesions integratedbackscatter intravascular ultrasound study, J. Am. Coll. Cardiol. 49 (2007)11491156.

[12] M. Zheng, S.Y. Choi, S.J. Tahk, et al., The relationship between volumetric plaquecomponents and classical cardiovascular risk factors and the metabolic syndromea 3-vessel coronary artery virtual histology-intravascular ultrasound analysis, JACCCardiovasc. Interv. 4 (2011) 503510.

[13] T. Mitsuhashi, K. Hibi, M. Kosuge, et al., Relation between hyperinsulinemia andnonculprit plaque characteristics in nondiabetic patients with acute coronary syn-dromes, JACC Cardiovasc. Imaging 4 (2011) 392401.

[14] G.S. Mintz, S.E. Nissen, W.D. Anderson, et al., American College of Cardiology clinicalexpert consensus document on standards for acquisition, measurement andreporting of intravascular ultrasound studies (IVUS). A report of the AmericanCollege of Cardiology Task Force on Clinical Expert Consensus Documents, J. Am.Coll. Cardiol. 37 (2001) 14781492.

[15] W.T. Friedewald, R.I. Levy, D.S. Fredrickson, Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentri-fuge, Clin. Chem. 18 (1972) 499502.

[16] A. Katz, S.S. Nambi, K. Mather, et al., Quantitative insulin sensitivity check index:a simple, accurate method for assessing insulin sensitivity in humans, J. Clin.Endocrinol. Metab. 85 (2000) 24022410.

[17] S. Lee, S. Choi, H.J. Kim, et al., Cutoff values of surrogate measures of insulin resis-tance for metabolic syndrome in Korean non-diabetic adults, J. Korean Med. Sci.21 (2006) 695700.

[18] L.M. Hettihawa, S. Palangasinghe, S.S. Jayasinghe, S.W. Gunasekara, T.P.Weerarathna, Comparison of insulin resistance by indirect methods HOMA,QUICKI and McAuley with fasting insulin in patients with type 2 diabetes inGalle, Sri Lanka: a pilot study, Online J. Health Allied Sci. 1 (2006) 2 (Available athttp://www.ojhas.org/issue17/2006-1-2.htm).

[19] R.H. Eckel, S.M. Grundy, P.Z. Zimmet, The metabolic syndrome, Lancet 365 (2005)14151428.

[20] S. Attvall, J. Fowelin, I. Lager, H. Von Schenck, U. Smith, Smoking induces insulin re-sistance a potential link with the insulin resistance syndrome, J. Intern. Med. 233(1993) 327332.

[21] F.S. Facchini, C.B. Hollenbeck, J. Jeppesen, Y.D. Chen, G.M. Reaven, Insulin resistanceand cigarette smoking, Lancet 339 (1992) 11281130.

[22] J.S. Yudkin, C.D.A. Stehouwer, J.J. Emeis, S.W. Coppack, C-reactive protein in healthysubjects: associations with obesity, insulin resistance, and endothelial dysfunction,

Arterioscler. Thromb. Vasc. Biol. 19 (1999) 972978.

[23] A. Festa, R. D'Agostino Jr., G. Howard, L. Mykkanen, R.P. Tracy, S.M. Haffner, Chronicsubclinical inammation as part of the insulin resistance syndrome: the insulinresistance atherosclerosis study (IRAS), Circulation 102 (2000) 4247.

[24] T. McLaughlin, F. Abbasi, C. Lamendola, L. Liang, G. Reaven, P. Schaaf, P. Reaven,Differentiation between obesity and insulin resistance in the association withC-reactive protein, Circulation 106 (2002) 29082912.

[25] J.L. Anderson, C.D. Adams, E.M. Antman, et al., ACC/AHA 2007 guidelines for themanagement of patientswith unstable angina/non-ST-elevationmyocardial infarction:a report of the American College of Cardiology/American Heart Association Task Forceon Practice Guidelines (Writing Committee to Revise the 2002 Guidelines for theMan-agement of Patients With Unstable Angina/Non-ST-Elevation Myocardial Infarction)developed in collaborationwith theAmerican College of Emergency Physicians, the So-ciety for Cardiovascular Angiography and Interventions, and the Society of ThoracicSurgeons endorsed by the American Association of Cardiovascular and Pulmonary Re-habilitation and the Society for Academic Emergency Medicine, J. Am. Coll. Cardiol. 50(2007) e1e157.

[26] R. Virmani, A.P. Burke, F.D. Kolodgie, A. Farb, Pathology of the thin-capbroatheroma: atype of vulnerable plaque, J. Interv. Cardiol. 16 (2003) 267272.

[27] J.A. Ambrose, M.A. Tannenbaum, D. Alexopoulos, et al., Angiographic progression ofcoronary artery disease and the development of myocardial infarction, J. Am. Coll.Cardiol. 12 (1988) 5662.

[28] G. Giannopoulos, L. Pappas, A. Synetos, et al., Association of virtual histology charac-teristics of the culprit plaque with post-brinolysis ow restoration in ST-elevationmyocardial infarction, Int. J. Cardiol. 174 (2014) 678682.

[29] K. Ohshima, S. Ikeda, H. Kadota, et al., Impact of culprit plaque volume and compo-sition on myocardial microcirculation following primary angioplasty in patientswith ST-segment elevation myocardial infarction: virtual histology intravascularultrasound analysis, Int. J. Cardiol. 167 (2013) 10001005.

[30] Y.K. Cho, S.H. Hur, Practical application of coronary imaging devices in cardiovascu-lar intervention, Korean Circ. J. 45 (2015) 8795.

[31] Y.H. Choi, Y.J. Hong, I.K. Park, et al., Relationship between coronary artery calciumscore by multidetector computed tomography and plaque components by virtualhistology intravascular ultrasound, J. Korean Med. Sci. 26 (2011) 10521060.

[32] K. Nasu, E. Tsuchikane, O. Katoh, et al., Plaque characterisation by virtual histologyintravascular ultrasound analysis in patients with type 2 diabetes, Heart 94 (2008)429433.

[33] M.O. Versteylen, I.A. Joosen, M.H. Winkens, et al., Combined use of exercise electro-cardiography, coronary calcium score and cardiac CT angiography for the predictionof major cardiovascular events in patients presenting with stable chest pain, Int. J.Cardiol. 167 (2013) 121125.

62 Y.M. Park et al. / International Journal of Cardiology 190 (2015) 5662

The role of insulin resistance and metabolic risk factors on culprit coronary plaque1. Introduction2. Patients and methods2.1. Patients and study design2.2. IVUS procedure and analysis2.3. Coronary risk factors and lipids, metabolic parameters2.4. Insulin resistance (IR) index, numbers of risk factors for IR2.5. Cardiovascular diagnosis2.6. Data analysis

3. Results3.1. Study population and baseline characteristics (Table1)3.2. Quantitative parameters of gray-scale and VH-IVUS3.3. Correlation between metabolic, lipid parameters and the amount of necrotic core3.4. Independent predictors for the amount of NC at the ROI3.5. Correlation between the presence of metabolic risk factors and plaque (Table3)3.6. Thin cap fibroatheroma (TCFA)

4. DiscussionConflicts of interestAcknowledgmentsReferences