Association of high-risk coronary atherosclerosis at CCTA with clinical and circulating biomarkers: insight from CAPIRE study
Edoardo Conte, MD, Daniele Andreini, MD, PhD, Marco Magnoni, MD, Serge Masson, PhD, Saima Mushtaq, MD, Sergio Berti, MD, Mauro Canestrari, MD, Giancarlo Casolo, MD, Domenico Gabrielli, MD, Roberto Latini, MD, Paolo
ABSTRACT
Background
High-risk coronary atherosclerosis features evaluated coronary CT angiography (CCTA) were suggested to have a prognostic role. The present study aimed to evaluate the association of circulating biomarkers with high-risk plaque features assessed by CCTA.
Methods
A consecutive cohort of subjects who underwent CCTA because of suspected CAD was screened for inclusion in the CAPIRE study. Based on risk factors (RF) burden patients were defined as having a low clinical risk (0-1 RF with the exclusion of patients with diabetes mellitus as single RF) or an high clinical risk (≥3 RFs). In all patients, measurement of inflammatory biomarkers and CCTA analysis focused on high-risk plaque features were performed. Univariate and multivariate logistic regression analysis were used to evaluate the relationship between clinical and biological variables with CCTA advanced plaque features.
Results
528 patients were enrolled in CAPIRE study. Older age and male sex appeared to be predictors of qualitative high-risk plaque features and associated with the presence of elevated total, noncalcified and low-attenuation plaque volume. Among circulating biomarkers only hs-CRP was found to be associated with qualitative high-risk plaque features (OR 2.02, p=0.004 and 2.02, p=0.012 for LAP and RI>1.1, respectively) with borderline association with LAP-Vol (OR 1.52, p=0.076); HbA1c and PTX-3 resulted to be significantly associated with quantitative high-risk plaque features (OR 1.71, p=0.003 and 1.04, p=0.002 for LAP-Vol, respectively).
Conclusions
Our results support the association between inflammatory biomarkers (hs-CRP, PTX-3), HbA1c and high-risk atherosclerotic features detected by CCTA. Male sex and older age are significant predictors of high-risk atherosclerosis.
Keywords
Coronary atherosclerosis; Coronary plaque; Cardiac Computed Tomography; Inflammation; Cardiovascular Prevention; High-risk plaque features
INTRODUCTION
The pivotal role of atherosclerotic plaque features in the pathophysiology of acute coronary syndrome (ACS) has been previously well established(1-3). However, the detection of subclinical coronary atherosclerosis has still a limited role in the primary prevention setting(4). Recent data from PROSPECT trial suggested coronary plaque burden and thin-cap fibroatheroma evaluated by intravascular ultrasound (IVUS) were associated with future ACS, beyond lumen stenosis severity(5).
Cardiovascular computed tomography angiography (CCTA) has recently emerged as a promising tool for non-invasive assessment of coronary atherosclerotic disease, over lumen stenosis quantification(6). Recent findings from the ICONIC trial confirmed that non-stenotic plaques at CCTA may represent precursors of ACS culprit lesions which are often characterized by high-risk plaque features such as positive remodeling, low-attenuation, napkin ring sing and elevated fibrofatty and non-calcified plaque volume(7). Data from GISSI-CAPIRE study confirmed the prognostic value of atherosclerosis assessment by coronary CTA, demonstrating high noncalcified plaque volume as the most ACS-predictive parameter in patients with extensive CAD (8). Limited data are still available regarding the association between clinical or circulating biological variables and high-risk atherosclerosis detected by CCTA(9-11). Thus, the present study aimed to investigate the association of traditional risk factors and circulating inflammatory biomarkers with coronary atherosclerosis and high-risk plaque features evaluated by CCTA.
METHODS
CAPIRE study
The CAPIRE study (ClinicalTrials.gov Identifier:NCT02157662) is part of the GISSI Outlier Project jointly promoted by the Heart Care Foundation Onlus, Italian Association of Hospital Cardiologists (ANMCO) and Mario Negri Institute of Pharmacological Research, Milan. The CAPIRE study was designed as a prospective, observational, international multicenter study involving a cross-sectional comparison of clinical, imaging, and biomolecular variables of extreme selected populations(11).
Study population
From January 2011 and June 2013 a consecutive cohort of subjects, both male and female, aged 45 to 75 years, who underwent 64-slice (or superior) CCTA because of suspected stable CAD in the outpatient clinics of the 11 centers involved in the study, were screened. Exclusion criteria of the CAPIRE study were previous cardiovascular events (acute myocardial infarction, unstable angina, chronic stable angina, previous percutaneous or surgical coronary revascularization, heart failure), previous heart disorders documented or identified at CCTA, previous documented acute or chronic peripheral vascular disease and active inflammatory or neoplastic disease as previously described (11). It should be underlined that, in accordance with the original study design, patients with CAD but having a segment involvement score (SIS)<5 at CCTA and patients with diabetes as single risk factor were not enrolled (11).
Cardiovascular risk factors (RF) presence was prospectively recorded as previously described(11). Based on RF burden patients were defined as having a low clinical risk (0-1 RF with the exclusion of patients with type 1 or type 2 diabetes mellitus as single RF) or high clinical risk (3 or more RFs.)(11).
Laboratory analysis
A peripheral venous blood sample was collected from each patient. The samples were immediately processed to obtain separate aliquots of whole blood, plasma, and serum and stored at -70°C in a dedicated biological bank (SATURNE-1, Mario Negri Institute of Pharmacological Research, Milan).
Circulating biomarkers were measured in a central laboratory, in a single batch, by personnel unaware of patients' characteristics, previously reported(11-12). More specially glycated hemoglobin (HbA1c) and lipids were measured with standard, automated laboratory methods. Plasma levels of pentraxin-3 (PTX3), an acute phase protein produced by several cell types under inflammatory stimuli, among them, endothelial cells were first identified (13), were measured by inhouse sandwich enzyme-linked immunosorbent assay. High-sensitivity C-reactive protein (hsCRP), which is mostly produced in the liver stimulated by inflammation, was measured with an automatic immunoturbidimetric method (Beckman-Coulter, Galway, Ireland).
CCTA analysis
All CCTA scans were transferred to the CCTA Core Lab (Centro Cardiologico Monzino, Milano) for a central blinded analysis of the coronary angiograms using the CardioQ3 Package-GE Healthcare as post-processing software. Coronary plaques were defined as structures of at least 1 mm2 area within and/or adjacent to artery lumen, clearly distinguishable from the vessel lumen, and surrounded by pericardial tissue; tissue with signal intensity below -40HU was considered a pericardial fat and excluded from the analysis. High-risk plaque features (HPFs) were evaluated as follow: arterial remodeling index (RI) assessed using vessel area (RIa=lesion plaque area/reference area), plaque burden (PB=[lesion plaque arealesion lumen area]/lesion plaque area), napkin ring sign (NRS) defined as the presence of a semicircular thin enhancement around the plaque along the outer contour of the vessel and small spotty calcifications (SC) as any discrete calcification ≤3 mm in length and occupying ≤90° arc when viewed in short axis. Plaque consistency was assessed using Hounsfield Unit (HU) and lowattenuation plaque (LAP) was defined as the presence of any voxel <30 HU(14). Total plaque volume (TPV) was evaluated and reported in mm3(15). Low attenuation plaque volume (LAP-Vol) and non-calcified plaque volume (NCP-Vol) have been expressed as the amount of plaque with <30 HU and <150 HU, respectively. All plaque volumes have also been evaluated as qualitative dichotomic variables using the highest quartile (HQ) as cut-off on a per-patient basis. In all patients segment involvement score (SIS) and segment stenosis score (SSS) were calculated, as previously described (8) and evaluated as qualitative dichotomic variables using the HQ as cut-off on a perpatient basis.
Statistical analysis
Continuous variables were presented as mean with SD or as median with interquartile range (IQR: 25°-75°) if more appropriate (non-normal distribution). Continuous variables normally distributed were compared using the Student t-test for independent samples. When the variable distribution was not normal, Mann-Whitney U tests for independent samples were used. The proportion of the categorical variables was compared using a χ2 analysis or Fisher exact test, as appropriate. A pvalue <0.05 was considered statistically significant.
Estimated odds ratios (OR) and accompanying 95% confidence intervals (CIs) are presented for each class of the variables that became significant at the univariate analysis. Logistic regression analysis was used in order to evaluate the relationship between clinical and biological variables with CCTA advanced coronary atherosclerosis characteristics. Significant predictors of CCTA plaque characteristics were evaluated at multivariable analysis and different models were used in order to avoid overfitting. A separate analysis will be performed to evaluate the prevalence of coronary plaque features according to age and sex. Statistical significance of the contribution of each added variable to different models was assessed by Harrell C-index (area under the receiver operating characteristic curve), and differences in predictive performance between models were tested using a nonparametric method(16)
Statistical analysis and graphics were produced with MedCalc (version 11.6.1.0, Med-Calc Software; 1993–2011).
RESULTS
As previously reported, 528 patients were definitively enrolled in CAPIRE study(14). Mean age was 59.9±8.3, and male prevalence was 58.3% (308/528) (Table 1). Risk factors prevalence was significantly higher among high vs low risk group patients (Table 2), as expected. CCTA showed no coronary artery disease in 348 (65.9%) patients. Among the entire population enrolled in the study SIS and SSS were 2.5 ± 3.7 and 3.9 ± 6.4, respectively. When only CAD patients were considered SIS and SSS were 7.6 ± 1.9 and 11.6 ± 5.8, respectively. Highest quartile for definition of elevated low-attenuation, non-calcified and total plaque volume were 10 mm3, 80 mm3 and 300 mm3 respectively.
Clinical characteristics, laboratory data in patients with and without CAD.
Subjects with CAD at CCTA were older and male sex was more prevalent in CAD group when compared to patients without CAD (81.1%vs.46.5%, respectively; p<0.001). Similarly, the prevalence of traditional RFs was higher in CAD vs. no CAD subjects (Table 1). Total cholesterol serum values were higher in patients without CAD compared with those with CAD while serum level of triglycerides, HbA1c, high sensitive TnT (hs-TnT), high sensitive C-reactive protein (hsCRP) was higher in patients with CAD. No differences in pentraxin-3 (PTX3) levels were detected between patients with CAD and those without CAD (Table 1). As expected, the prevalence of patients at higher clinical risk was higher in patients with CAD (52.8% vs.34.5%; p<0.001).
Univariate analysis of clinical and biomarker predictors of high-risk atherosclerosis
Among clinical characteristics, male gender appeared to be the most important predictor of qualitative high-risk plaque features (PR, LAP, NRS, and SC) and resulted to be associated with the presence of elevated total, non-calcified and low-attenuation plaque volume (eTable 1 in the Supplement). Similarly, both older age and traditional risk factors are significantly associated with high-risk coronary atherosclerosis, while positive family history for CAD was associated neither to qualitative high-risk plaque features nor to elevated coronary plaque volume. Of note, being included in the high clinical risk group was found to be significantly associated with qualitative high-risk plaque features and to elevated low-attenuation plaque volume (OR[95%CI]:2.08 [1.08-4.01]; p=0.027), but not to total plaque volume (OR[95%CI]:1.29 [0.69-2.41]; p=0.421).
Among circulation biomarkers, high HDL-cholesterol was significative associated with the absence of high-risk plaque features (OR[95%CI]:0.95 [0.93-0.97]; p<0.001 for LAP; OR[95%CI]:0.95 [0.94-0.97]; p<0.0001 for RI>1.1; OR[95%CI]:0.92 [0.89-0.95], p<0.001 for NRS and OR[95%CI]:0.96 [0.95-0.98], p<0.001 for SC) and to the absence of elevated total, non-calcified and low-attenuation plaque volume (OR[95%CI]:0.94 [0.92-0.97], p<0.001; OR[95%CI]:0.94 [0.92-0.97], p<0.001; OR[95%CI]:0.94 [0.92-0.97], p<0.001 respectively), while LDL cholesterol was not associated with any high-risk plaque features and total cholesterol values were positively associated with LAP and PR only. Elevated HbA1c and hs-CRP were found to be positively associated with each high-risk plaque features (eTable 1 in the Supplement), while hs-TnT was associated with LAP and RI>1.1 and to both total and low attenuation plaque volume (eTable 2 in the Supplement). Of note, PTX-3, whose serum values were similar among patients with and without CAD, was found to be associated with elevated low-attenuation plaque volume but not with total coronary plaque volume (eTables 1 and 2 in the Supplement).
Prevalence of coronary plaque features according to age and sex eTable 3 shows the prevalence of qualitative and quantitative plaque features at CCTA according to age and sex. The prevalence of each plaque feature was significantly higher in males than in female subjects (Figure 1). Of note, among female patients younger than 65 years old, only one subject was found to have elevated low-attenuation plaque volume, and no female subjects had elevated noncalcified plaque volume or elevated total plaque volume (eTable 3 in the Supplement).
Multivariable analysis of clinical and biomarker predictors of high-risk atherosclerosis
After adjusting for most relevant clinical predictors of high risk atherosclerosis [male sex, older age and high risk clinical profile (3 or more RFs)] and for coronary atherosclerosis burden (SSS and SIS) hs-CRP was found to be associated with qualitative high-risk plaque features (OR[95%CI]: 2.02 [1.24-3.27], p=0.004; 2.02 [1.25-3.27], p=0.012 for LAP and, RI>1.1 respectively] On the contrary, neither HbA1c nor hs-TnT and PTX-3 was found to be associated with qualitative highrisk plaque features (Table 3).
For what concern quantitative plaque analysis, elevated total plaque volume was not associated with any biomarkers. On the contrary, after adjusting for age, sex, clinical risk profile and atherosclerosis burden, only HbA1c and PTX-3 were significantly associated with elevated low attenuation plaque volume (OR[95%CI]: 1.71 [1.91-2.44], p=0.003; 1.04 [1.02-1.08], p=0.002, respectively) (Table 3). HbA1c was found to be associated with elevated non-calcified plaque volume as well (OR[95%CI]: 1.48 [1.02-2.15], p=0.036). However, even if a significant independent association with high-risk plaque feature was found, a multivariable model including biomarkers did not improve prediction of high risk plaque features at CCTA when compared to a clinical model including traditional risk factors and male sex (Model 3, Model 4 and Model 5 vs Model 2 in eTable 4 in the Supplement); conversely, older age and male sex appeared to significantly improve the prediction of high-risk atherosclerosis over cardiovascular risk factors (Model 2 vs. Model 1 in eTable 4 in the Supplement).
DISCUSSION
The findings of the present study suggest that inflammatory biomarkers, more specifically hs-CRP and PTX-3, are associated with high-risk features of coronary atherosclerosis. Moreover, a positive association between high-risk atherosclerosis and HbA1c was evident. Of note, these associations were confirmed at multivariate analysis after correction for age, sex, traditional risk factors and atherosclerosis burden (Table 3).
Notably, only PTX-3 and HbA1c were significantly higher among subjects with elevated lowattenuation plaque volume (OR 1.04 [1.02-1.08], p=0.002 for PTX-3 and OR 1.71 [1.91-2.44], p=0.003 for HbA1c) (Table 3), while hs-CRP was correlated with the presence of qualitative highrisk plaque features (LAP, RI>1.1) and only a borderline association (p=0.076) was identified with elevated low-attenuation plaque volume at multivariable analysis. On the contrary, no significant association was found between inflammatory biomarkers and elevated total coronary plaque volume (Figure 2). These data support the role of inflammation in atherosclerosis pathophysiology, suggesting that active inflammation may be linked to high-risk atherosclerosis presence and progression (Figure 3 and 4). Noteworthy, the absence of the association between inflammatory biomarkers and total coronary plaque volume may suggest that inflammation selectively enhance high-risk atherosclerosis development, but it is not directly involved in the global coronary atherosclerosis burden. This hypothesis is consistent with previous studies linking hs-CRP with non-calcified plaque (17) and suggesting hs-CRP as a prognostic marker (18); moreover, recent studies demonstrated the association between high-risk atherosclerosis subtypes and cardiovascular events at follow-up, even in the absence of diffuse and/or obstructive disease (6,7,19).
Besides the role of inflammatory biomarkers, we found a positive association between high-risk atherosclerosis and HbA1c; specifically, HbA1c level was associated with both low-attenuation plaque volume and non-calcified plaque volume at multivariate analysis. Otherwise, no significant relationship between HbA1c and total coronary plaque volume was evidenced. Even if patients with diabetes a single risk factors were excluded from the study, these findings supported the role of diabetes and glycemic metabolism abnormalities as strong independent risk factor for atherosclerosis and coronary events and appeared to be consistent with previous data reporting a reduction in HbA1c serum level after anti-inflammatory therapy, suggesting a link between diabetes and vascular inflammation (20). Of interest, we found a positive association between HbA1c and high risk atherosclerosis even if mean values of HbA1c were low (3.5% in the entire population enrolled and 3.6% in patients with CAD as reported in Table 1); this may suggest that even a very mild diabetic metabolic state may increase the risk of high risk atherosclerosis development, that has been previously associated with higher risk of future cardiovascular events (6-8).
However, it should be underlined that patients included in the high clinical risk group resulted in having an increased global coronary atherosclerosis burden and high-risk plaque features were more prevalent when compared to patients included in the low clinical risk group. Moreover, older age and male sex appeared to be strongly related to the presence of atherosclerosis and the prevalence of high-risk atherosclerosis was very low in younger female (Figure 1) (eTable 3 in the Supplement), as previously reported (21).
Sequential model for the prediction of high risk atherosclerosis showed a significative improvement when older age and male gender were added on top of clinical risk (Model 2 vs Model 1 in eTable 4 in the Supplement), confirming that older age and male gender are significantly associated with atherosclerosis development and progression, while only mild and not significative improvement was demonstrated when inflammatory biomarkers were added (Model 3-4-5-6 vs. Model 2 in eTable 4 in the Supplement).
Some limitations of this study should be acknowledged. First, patients enrolled in the study had a clinical indication for CCTA; thus the application of study results to patients without suspected CAD is limited. Second, patients with 1
12. Latini R, Maggioni AP, Peri G et al. Prognostic significance of the long pentraxin PTX3 in acute myocardial infarction. Circulation.2004;16:2349-54.
13. M Introna, V V Alles, M Castellano, G Picardi, L De Gioia, B Bottazzai, G Peri, F Breviario, M Salmona, L De Gregorio, T A Dragani, N Srinivasan, T L Blundell, T A Hamilton, A Mantovani. Cloning of Mouse ptx3, a New Member of the Pentraxin Gene Family Expressed at Extrahepatic Sites. Blood 1996, 87 (5), 1862-72
14. Motoyama S, Ito H, Sarai M 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;66:337-46
15. Conte E, Mushtaq S, Pontone G et al. Plaque quantification by coronary computed tomography angiography using intravascular ultrasound as a reference standard a comparison between standard and last generation computed tomography scanners. European Heart Journal – Cardiovascular Imaging (2019) doi:10.1093/ehjci/jez089-In Press
16. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44:837–845
17. Rubin J, Chang HJ, Nasir K, et al. Association between high-sensitivity C-reactive protein and coronary plaque subtypes assessed by 64-slice coronary computed tomography angiography in an asymptomatic population. Circ Cardiovasc Imaging. 2011;4(3):201-9.
18. Yousuf O, Mohanty BD, Martin SS, et al. High-sensitivity C-reactive protein and cardiovascular disease: a resolute belief or an elusive link? J Am Coll Cardiol.30;62(5):397408
19. Ferencik M, Mayrhofer T, Bittner DO et al. Use of High-Risk Coronary Atherosclerotic Plaque Detection for Risk Stratification of Patients With Stable Chest Pain: A Secondary Analysis of the PROMISE Randomized Clinical Trial. JAMA Cardiol. 2018;3(2):144-152
20. Larsen CM, Faulenbach M, Vaag A, et al. Interleukin-1-receptor antagonist in type 2 diabetes mellitus. N Engl J Med.2007;356:1517–1526;
21. Merz AA, Cheng S. Sex differences in cardiovascular ageing. Heart. 2016;102(11):825-31