Perivascular Fat Attenuation For Predicting Adverse Cardiac Events In Stable Patients Undergoing Invasive Coronary Angiography

Abstract

Background

Inflammation surrounding the coronary arteries can be non-invasively assessed using pericoronary adipose tissue attenuation (PCAT). While PCAT holds promise for further risk stratification of patients with low coronary artery disease (CAD) prevalence, its value in higher risk populations remains unknown.

Methods

CORE320 enrolled patients referred for invasive coronary angiography with known or suspected CAD. Coronary computed tomography angiography (CCTA) images were collected for 381 patients for whom clinical outcomes were assessed 5 years after enrollment. Using semi-automated image analysis software, PCAT was obtained and normalized for the right coronary (RCA), left anterior descending (LAD), and left circumflex arteries (LCx). The association between PCAT and MACE during follow up was assessed using Cox regression models.

Results

Thirty-seven patients were excluded due to technical failure. For the remaining 344 patients, median age was 62 (interquartile range, 55–68) with 59% having ≥1 coronary artery stenosis of ≥50% by quantitative coronary angiography. Mean attenuation values for PCAT in RCA, LAD, and LCx were −74.9, −74.2, and −71.2, respectively. Hazard ratios and 95% confidence intervals (CI) for normalized PCAT in the RCA, LAD, and LCx for MACE were 0.96 (CI: 0.75–1.22, p=0.71), 1.31 (95% CI: 0.96–1.78, p=0.09), and 0.98 (95%CI: 0.78–1.22, p=0.84), respectively. For death, stroke, or myocardial infarction only, hazard ratios were 0.68 (0.44–1.07), 0.85 (0.56–1.29, and 0.57 (0.41–0.80), respectively.

Conclusions

In patients referred for invasive coronary angiography with suspected CAD, PCAT did not predict MACE during long term follow up. Further studies are needed to understand the relationship of PCAT with CAD risk.

TOC SUMMARY

Increased pericoronary adipose tissue attenuation (PCAT) has been introduced as a marker of perivascular inflammation and as a predictor of major adverse cardiovascular events (MACE) in patients with suspected coronary artery disease undergoing CT angiography. It remains unknown if PCAT is predictive of MACE in higher risk populations. In our study of patients with known or suspected coronary artery disease referred for invasive coronary angiography, raw or normalized PCAT assessed in the three coronary arteries did not predict MACE. Our results suggest that population specific characteristics may influence the performance of PCAT for predicting patient outcome.

Introduction

Vascular inflammation is a driver of coronary atherosclerotic plaque formation, rupture vulnerability, and a prothrombotic state.¹ Imaging biomarkers which noninvasively assess coronary artery inflammation hold potential for improved risk stratification beyond traditional makers in primary and secondary prevention of coronary artery disease (CAD).²

Pericoronary adipose tissue attenuation (PCAT) has recently emerged as a marker of coronary artery inflammation which is fairly easily obtained from standard coronary computed tomography angiograms (CCTA), a widely used first-line, non-invasive imaging modality for CAD evaluation.^2,3 Higher PCAT values (reflecting tissue with greater X-ray attenuation) is associated with perivascular inflammation, which, in turn, is linked with increased risk of adverse cardiovascular events (MACE).⁴ Studies in populations of low CAD risk revealed higher PCAT values to be independently associated with all-cause mortality, cardiac death, and non-fatal myocardial infarction (MI).^4,5 Importantly, PCAT has been shown to add incremental prognostic value beyond clinical reporting and quantitative plaque measurements.⁶ The direct mechanistic relationship between perivascular inflammation and plaque formation remains unclear and PCAT is not consistently associated with quantitative plaque features.^6,7

Though PCAT’s prognostic utility has been documented in low-risk populations, scarce data are available on the predictive value of PCAT in higher risk cohorts.⁸ The present study investigates the predictive value of PCAT in an intermediate-to-high risk, symptomatic population referred for invasive coronary angiography (ICA) for suspected or known CAD.

Methods

Study Design

The CORE320 (Coronary Artery Evaluation Using 320-Row Multidetector Computed Tomography Angiography and Myocardial Perfusion) is a prospective, multicenter, multinational clinical study (www.clincialtrials.gov, ClinicalTrials.gov Identifier: NCT00934037). The study spanned 16 hospitals in 8 countries. All patients signed individual informed consent and all centers obtained approval from local institutional review boards.

Patient Population

The study design and population of CORE320 have been described in detail in prior reports.⁹ The CORE320 cohort included patients between 45 and 85 years of age, referred for ICA with suspected or known CAD between November 2009 and July 2011. Patients were excluded for: history of allergies to iodinated contrast media, history of contrast-induced nephropathy, elevated serum creatinine (>1.5 mg/dL) or calculated creatinine clearance of <60 mL/min, atrial fibrillation, uncontrolled tachyarrhythmia, second or third degree atrio-ventricular block, previous cardiac surgery or coronary intervention within the past 6 months, evidence of acute coronary syndrome with thrombolysis, myocardial infarction risk score ≥5, or elevated cardiac enzymes in the past 72 hours, body mass index > 40 kg/m2, high radiation exposure (≥ 5.0 rem or ≥ 2 nuclear or CT studies) in the 18 months before consent among others. The population did not exclude participants with high coronary calcium score or those with stents. The study included and completed imaging from 381 consenting participants.

CTA Image Acquisition

CTA methods for CORE320 have been described in detail in prior reports.⁹ In short, coronary calcium scores (CCS) and coronary CTA images were collected within 60 days prior to ICA. All images were conducted with a 320 × 0.5 mm row detector CT scanner (Aquilion One, Canon Medical Systems, Otawara, Japan). Patients with BMI of ≤ 30 kg/m² and heart rate ≥ 60 beats/min received 75 mg of oral metoprolol. Patients with BMI of ≥ 30 kg/m² and heart rate ≥ 60 beats/min received 150 mg of oral metoprolol. Patients with heart rates ≥ 60 beats/min received IV (≤15 mg) metoprolol. Patients with systolic blood pressure ≥ 110 mmHg received sublingual fast-acting nitrates. CT angiography was performed using iodinated contrast (Iopamidol 370 mg), with doses ranging from 50–70 mL based on patient weight. Contrast was injected intravenously at 4.0–5.0 mL/sec with real-time bolus tracking in the descending aorta. Tube voltage was held constant at 120 kilovolts for all patients.

Pericoronary Adipose Tissue Attenuation (PCAT)

Pericoronary adipose tissue attenuation (PCAT), reported in Hounsfield units (HU), is an imaging biomarker of the composition of perivascular fat around the coronary arteries.² Inflammation of perivascular fat shifts its attenuation from the lipid (more negative HU values) to the aqueous phase (less negative HU values).

Analysis of PCAT was based on a commercially available image processing software (Aquilion One V6.0) with minor custom modifications, including configuring the desktop to be run with limited computing power. No modifications to the image analysis algorithms were performed. Image segmentation and analysis was performed by a single investigator who was blinded to population demographics and outcomes. To assess inter-user variability of PCAT measurements, we presented correlations and Bland-Altman analyses using a second independent investigator on a randomly selected sample of 20 patients.

Perivascular fat was defined as the adipose tissue (−190 to −31 HU) within a 5 millimeter (mm) radial distance from the outer vessel wall (Figure 1). The most pertinent information on perivascular fat attenuation was found within such radius around the artery.² We determined PCAT for all three major coronary arteries. Straightened multi-planar reformations were used for vessel display with automated lumen and vessel contour detection (Figure 1). Using the software provided ruler, a 40 mm segment was selected for the proximal right coronary artery (RCA), starting 10 mm distal to the ostium to avoid proximity to the aortic wall (2). For the left anterior descending (LAD) and the left circumflex (LCx) arteries, a 40 mm segment was selected starting at the bifurcation of the left main coronary artery. Fat attenuation was measured in concentric cylindrical shells of 1 mm thick layers around the outer vessel wall to a radial distance of 5 mm beyond the outer vessel wall, resulting in a distance gradient. All 5 shells were averaged on a per-vessel basis to yield an overall PCAT value. Examples of vessels with relatively high and low PCAT values are shown in Figure 2.

Figure 1.

An example of computer-aided segmentation of a right coronary artery with traced lumen and vessel contours and pericoronary adipose attenuation. The left panel (1A) shows a three-dimensional reconstruction of CT images with the proximal right coronary artery highlighted (red rectangle) which is shown below in a straightened curved multi-planar reformatted CT image along with lumen and vessel border tracing. The right upper panel shows a cross-sectional view of the red rectangle with an 1mm perivascular shell closest to vessel wall analyzed for X-ray attenuation. The right lower panel depicts all 5 shells out to 5mm beyond the outer vessel wall. Panels 1B and C depict PCAT analysis out to 1- and 5-mm shells, respectively, with 1D and 1E being the grayscale equivalents. Blue highlights the perivascular region within −30 to 337 Hounsfield units [HU], while red highlights perivascular fat within −190 to −31 HU. Green highlights the lumen, which was excluded from PCAT analysis. Green circle indicates inner vessel wall; yellow circle presents outer vessel wall; purple circle represents region of PCAT analysis.

Figure 2.

Exemplar images of patients with (3A) low and (3B) high PCAT values out to 2mm beyond the vessel wall, with panels 3C and 3D being the corresponding grayscale images. Blue highlights the perivascular region within −30 to 337 Hounsfield units [HU], while red highlights perivascular fat within −190 to −31 HU. In panels 2B and 2D, yellow circles present the inner and outer vessel wall, while the green line represents the region of PCAT analysis (in this example, 2mm from outer vessel wall).

On a per-vessel basis, PCAT was averaged along the 40 mm vessel segment and then normalized to the average lumen attenuation adjacent to the analyzed section. Normalization was performed to account for attenuation variation based on acquisition settings, including tube current, contrast flow, and patient characteristics (2). Normalization to lumen density as opposed to aortic attenuation was performed to account for minor attenuation variability within vessels (data not shown). We computed the normalized PCAT by taking the mean HU across all slices (1 mm-5 mm) and then divided by the corresponding vessel-specific proximal segment lumen mean intensity. Except where indicated, all reported associations with PCAT refer to this lumen-attenuation-normalized value. Reporting of PCAT values without normalization (i.e., unadjusted) are termed “raw PCAT.

Outcomes

The CORE320 cohort was followed for the occurrence of major adverse cardiovascular events (MACE) at 30 days, 6 months, 1 year, 2 years, and 5 years after enrollment. These events were further characterized by all MACE vs. hard MACE defined as stroke, myocardial infarction, or death. All MACE included revascularization >30 days after ICA, all-cause death, myocardial infarction, stroke, hospitalization for chest pain or congestive heart failure, or outpatient presentations for chest pain or arrhythmia. All events were adjudicated by the CORE320 adjudication committee. Hazard ratios (HR) were adjusted for age, sex, hypertension, dyslipidemia, diabetes mellitus, current or former smoking history, and previous myocardial infarction.

Statistical Analysis

Continuous variables are presented as median (interquartile range [IQR]) and categorical variables are presented as absolute numbers with percentages. Patient demographics, baseline clinical variables, and imaging characteristics were compared using Wilcoxon rank-sum test for continuous variables and Pearson’s chi-square test for binary or categorical variables. PCAT, being normally distributed, is presented as mean ± standard deviation and compared using t-tests. Hazard ratios (HR) with 95% confidence intervals (CIs) were calculated using Cox proportional hazards models. In adjusted models, covariates included age, sex, hypertension, dyslipidemia, diabetes, smoking and previous MI. A p-value < 0.05 was considered statistically significant. Statistical analyses were performed using SAS 9.4 (SAS Foundation, Cary, NC, USA) or R 4.1.0 (R Foundation, The Netherlands).

Results

Patient Population

Of the 381 patients in the CORE320 cohort who underwent CTA, 37 were excluded because of software failure to load images, typically related to poor image quality or other technical issues. Of the remaining 344 patients, 229 (67%) were male and the median age was 62 (IQR, 55–68) (Table 1). Among 344 patients, 922 of 1032 (89%) vessels met the 40 mm length requirement. The final analysis included 922 vessels, with 294 RCA, 318 LAD, and 310 LCx vessels. Disease prevalence was high with 202/344 (59%) of patients revealing ≥1 coronary artery stenosis of ≥50% by quantitative coronary angiography. Of 344 participants, 213 (62%) had no prior history of CAD, and 187/344 (54%) were on a statin therapy at baseline.

Table 1.

Patient demographics and baseline clinical characteristics by presence or absence of major adverse cardiovascular event (MACE). Data are reported as number of observations (%) or median [IQR].

	Overall (n=344)	No MACE Event (n=262)	MACE Event (n=82)	P value
Demographics
Age, y	62 [55, 68]	62 [55, 68]	64 [58, 69]	0.03
Male	229 (67%)	165 (63%)	64 (78%)	0.01
BMI, kg/m2	26 [24, 30]	27 [24, 30]	26 [24, 30]	0.54
Ethnicity
Hispanic	28 (8%)	24 (9%)	4 (5%)	0.004
Non-Hispanic	295 (86%)	228 (87%)	67 (82%)
Other or Unknown	21 (6%)	10 (4%)	11 (13%)
Race
White	194 (56%)	139 (53%)	55 (67%)	0.03
Black	36 (10%)	31 (12%)	5 (6%)
Asian	111 (32%)	91 (35%)	20 (24%)
Other or Unknown	3 (1%)	1 (<1%)	2 (2%)
Baseline Risk Factors
Hypertension	264 (77%)	196 (75%)	68 (83%)	0.16
Diabetes Mellitus	112 (33%)	81 (31%)	31 (38%)	0.25
Dyslipidemia	226 (67%)	171 (67%)	55 (70%)	0.61
Prior history of myocardial infarction	91 (26%)	63 (24%)	28 (34%)	0.07
Family history of CAD	143 (44%)	104 (42%)	39 (51%)	0.19
Smoking
Current	55 (17%)	41 (16%)	14 (18%)	0.42
Past	120 (36%)	87 (35%)	33 (42%)
Never	154 (47%)	122 (49%)	32 (41%)
Medication
Statin	187 (54%)	141 (54%)	46 (56%)	0.72
Aspirin/salicylates	219 (64%)	164 (63%)	55 (67%)	0.46
β-blocker	188 (55%)	139 (53%)	49 (60%)	0.29
ACE-inhibitor or ARB	154 (45%)	116 (44%)	38 (46%)	0.74
Ca channel blocker	6 (6%)	2 (3%)	4 (17%)	0.01
Prior history of revascularization
PCI	105 (31%)	76 (29%)	29 (35%)	0.28
Type of Chest Pain
Typical angina	123 (45%)	89 (43%)	34 (52%)	0.15
Atypical angina	138 (51%)	109 (53%)	29 (45%)
Nonspecific chest pain	5 (2%)	5 (2%)	0 (0%)
Absence of chest pain	3 (1%)	1 (<1%)	2 (3%)
Unknown	2 (1%)	2 (1%)	0 (0%)
None/asymptomatic	24 (7%)	18 (7%)	6 (7%)	0.89
Imaging characteristics
Calcium Score	154 [5, 511]	109 [2, 434]	349 [77, 850]	<0.0001
Duke CAD Index	48 [0, 56]	37 [0, 56]	56 [37, 74]	0.0001
Clinically reported:
Obstructive CAD	221 (64%)	154 (59%)	67 (82%)	0.0002
High-risk plaque	153 (44%)	106 (40%)	47 (57%)	0.007
CTA initial heart rate	60 [55, 67]	60 [55, 68]	60 [56, 66]	0.92
Contrast amount, mL
50	47 (14%)	35 (13%)	12 (15%)	0.81
60	280 (81%)	215 (82%)	65 (79%)
70	17 (5%)	12 (5%)	5 (6%)
β-blocker (oral)
75	166 (48%)	122 (47%)	44 (54%)	0.43
150	55 (16%)	45 (17%)	10 (12%)
NONE	123 (36%)	95 (36%)	28 (34%)
β-blocker (intravenous), dose, mg	10 [5, 15]	10 [5, 15]	9 [5, 15]	0.52
Nitroglycerin during CTA	296 (86%)	228 (87%)	68 (83%)	0.35
Radiation dose, mSv	3.2 [2.8, 3.6]	3.1 [2.8, 3.6]	3.2 [2.9, 3.6]	0.14
Tube Amperage, mA	450 [400, 450]	445 [400, 450]	450 [400, 450]	0.65

Patients were followed for at least 5 years after enrollment. Overall, there were 19 deaths, 13 strokes, 6 myocardial infarctions, and 139 revascularization procedures (percutaneous intervention or coronary artery bypass grafting) at any time after enrollment. After excluding repeated events, 82 first events were considered for analysis (Supplementary Table 1).

Perivascular Fat Attenuation Index

Perivascular fat attenuation (PCAT) for all 3 vessels was normally distributed, with mean ± SD values for raw PCAT in the RCA, LAD, and LCx at −74.9 ± 9.8 HU, −74.2 ± 7.8, and −71.2 ± 7.5 respectively (Table 2, Figure 3). Mean normalized PCAT for the three coronary arteries were −0.24, −0.22, and −0.23, respectively (Table 2). By Bland-Altman analysis, inter-user reproducibility of PCAT measurements were excellent in all three vessels (Supplemental Figure 1).

Table 2.

Pericoronary adipose attenuation for each of the 3 main coronary arteries.

	Overall	No Hard Event	Hard Event	P value
RCA
Vessels analyzed	294	273	21
Raw PCAT (HU)	−74.9 ± 9.8	−74.8 ± 9.8	−76.9 ± 10.3	0.35
Normalized PCAT	−0.24 ± 0.05	−0.24 ± 0.05	−0.25 ± 0.06	0.30
LAD
Vessels analyzed	318	292	26
Raw PCAT (HU)	−74.2 ± 7.9	−74.2 ± 7.8	−74.7 ± 8.7	0.76
Normalized PCAT	−0.22 ± 0.05	−0.22 ± 0.05	−0.22 ± 0.06	0.86
LCx
Vessels analyzed	310	284	26
Raw PCAT (HU)	−71.2 ± 7.5	−71.2 ± 7.4	−72.1 ± 8.5	0.55
Normalized PCAT	−0.23 ± 0.05	−0.22 ± 0.05	−0.24 ± 0.06	0.07
	Overall	No MACE Event	MACE Event	P value
RCA
Vessels analyzed	294	225	69
Raw PCAT (HU)	−74.9 ± 9.8	−74.3 ± 9.8	−76.9 ± 9.7	0.06
Normalized PCAT	−0.24 ± 0.05	−0.24 ± 0.05	−0.24 ± 0.06	0.83
LAD
Vessels analyzed	318	243	75
Raw PCAT (HU)	−74.2 ± 7.9	−74.1 ± 7.9	−74.4 ± 7.9	0.77
Normalized PCAT	−0.22 ± 0.05	−0.22 ± 0.05	−0.21 ± 0.04	0.02
LCx
Vessels analyzed	310	233	77
Raw PCAT (HU)	−71.2 ± 7.5	−71.3 ± 7.5	−71.1 ± 7.6	0.87
Normalized PCAT	−0.23 ± 0.05	−0.23 ± 0.05	−0.22 ± 0.06	0.66

Data are reported as mean ± SD. HU, Hounsfield units.

Figure 3.

Distribution of pericoronary adipose tissue attenuation values for those with and without MACE (3A) and hard events (3B). RCA: right coronary artery; LAD: left anterior descending artery; LCx: left circumflex artery.

There was no positive association of raw PCAT values for any of the three coronary arteries with MACE (Table 2). Normalized PCAT values were not associated with MACE except for the LAD (Table 2). Median coronary lumen HU, utilized for normalization, was different between the RCA, LAD, and LCx (325 vs. 339 vs. 317 HU, respectively, p<0.001) justifying individual normalization per vessel (Supplementary Figure 2). The average diameter (mm) for the lumen of the RCA, LAD, LCx was 2.81 [2.34, 3.35], 2.41 [2.06, 2.79], and 2.41 [2.15, 2.82]. Hazard ratios were not statistically significant except for the LCx, where higher PCAT counterintuitively reduced the hazard of hard events (Table 3). There were weak correlations between PCAT and Agatston calcium scores (p<0.01 for all vessels) (Supplementary Figure 3). There was no association between vessel-specific PCAT and severe lumen stenosis (≥50% by quantitative coronary angiography).

Table 3.

Pericoronary adipose attenuation and its relationship with MACE and hard events for the three major coronary arteries.

	MACE		Hard events
	HR (95% CI)	p-value	HR (95% CI)	p-value
RCA	0.955 (0.748, 1.221)	0.71	0.682 (0.435, 1.070)	0.10
LAD	1.308 (0.960, 1.781)	0.09	0.848 (0.558, 1.288)	0.44
LCx	0.977 (0.780, 1.224)	0.84	0.573 (0.410, 0.800)	0.001

Lumen-attenuation normalized PCAT, expressed as per 0.05-unit increase.

HR, hazard ratio; CI, confidence interval.

In subgroup analyses, PCAT was not different between statin users and non-users (Supplementary Figure 4). For those taking statins at baseline, greater PCAT of the LCx was associated with a lower hazard ratio for hard events (HR: 0.587, 95% CI: 0.384–0.896, P=0.01). Additionally, PCAT did not differ between those with or without known CAD, and it was not associated with (hard) MACE in those with known CAD. Finally, PCAT was not different between those who had early revascularization (<30 days after imaging) compared to those without early revascularization. In patients who did undergo early revascularization, PCAT was not predictive of MACE or hard events.

In multiple regression models, adding PCAT of the RCA as a covariate did not improve prediction of MACE (ΔAUC=0, p=0.70, Figure 4A) or hard events (ΔAUC=2, p=0.18, Figure 4B) beyond the previously established risk factors of age, male sex, hypertension, dyslipidemia, diabetes, smoking status, and previous MI. In adjusted models, male sex was the only significant predictor of MACE (HR=1.90, 95% CI: 1.04–3.46, p=0.04). Age (HR=1.08, 95% CI: 1.01–1.15), dyslipidemia (HR=0.18, 95% CI: 0.07–0.46, p=0.0005), diabetes (HR=2.96, 95% CI: 1.08–8.07, p=0.03) and current smoking (HR=3.81, 95% CI: 1.22–11.91, p=0.02) were independently associated with hard events. PCAT of the LAD and LCx also did not add discriminative value to the multivariable model (Supplementary Figure 5).

Figure 4.

Receiver operator characteristics (ROC) curves for multivariable models predictive of MACE (4A, P=0.70 for ΔAUC) or hard events (4B, P=0.18 for ΔAUC). Red line represents model with variables of age, male sex, hypertension, dyslipidemia, diabetes, smoking status, and previous MI, while green line represents a model with the above variables in addition to PCAT of the right coronary artery (RCA).

To address a possible effect of averaging PCAT values in the measurement area, we assessed the distribution of PCAT within the shells around the vessel lumen. As expected, PCAT was highest in the innermost shell, closest to the lumen, compared to the shell furthest away from the lumen (p<0.0001 for all 3 vessels) (Supplementary Figure 6). Nonetheless, the PCAT value from each of the layers and the overall gradient were not associated with MACE.

Discussion

Our study investigated the prognostic value of coronary inflammation as assessed by PCAT in a cohort of patients referred for invasive coronary angiography with known or suspected obstructive CAD. In contrast to other studies,^4,10 we did not find PCAT to be a significant predictor of MACE in our cohort.

Active inflammation surrounding coronary vessels is associated with inhibited adipogenesis and increased X-ray attenuation.¹¹ Over time, sustained vascular and perivascular inflammation may lead to atherosclerotic plaque formation and eventual rupture or calcification.¹² Previous studies examining perivascular fat as a marker of local inflammation included low-intermediate risk patients referred for coronary CT angiography.^4,6 Patients referred for a conventional invasive coronary angiography are a select group with greater disease severity. Expectedly, our cohort had higher coronary calcium scores, obstructive CAD prevalence, and Duke CAD Index scores compared to previous studies.^4,6 Although disease severity is greater in these patients, prior translational studies suggest that active inflammation may slow down as these plaques mature.² As a maker of dynamic inflammation, PCAT may not reflect lower or altogether extinguished signals of chronic inflammation. Radiomic texture analysis of the perivascular region is a better tool to detect chronic changes in perivascular fat and represents a natural future step in analysis of the CORE320 population.¹³

Given the disease etiology of these intermediate to high-risk patients, it may be unsurprising that PCAT fails as a prognostic marker for cardiac-related events. Indeed, multivariable modeling did not reveal PCAT to aid prediction of hard events or MACE, with standard clinical risk factors being most predictive. Features of the plaques themselves, rather than ongoing perivascular inflammation, may be better suited to add prognostic value in this cohort.

There are a few possible explanations for the association of PCAT in the LCx with a decreased risk of hard MACE. First, this finding may be stochastic in nature and not representative of biological processes. For example, the CRISP-CT study did not observe a relationship between PCAT in the LCx with cardiac-related mortality.⁴ Similarly, van Diemen et al. reported that only RCA PCAT is of additional prognostic value on top of quantitative plaque features.⁶ Second, as hypothesized previously, our result could be due to differential biological factors between the coronaries. The LCx has many small, marginal branches, more surrounding myocardium, and less fat around it compared to the RCA, which has been the vessel of choice for prior PCAT studies.¹⁴ These structures may affect the semi-automated image analysis and fat density calculations. The RCA may be the only vessel in which PCAT from CCTA is a reliably measured biomarker. In any case, the inter-vessel discrepancy highlights the importance of vessel-specific calculations for PCAT.

Considering technical factors, van Diemen et al. observed significant differences in mean pericoronary fat attenuation based on the CT scanner used.⁶ The mean PCATs using 64- and 256-slice CT scanner were −72.7 and −80.2 HU, respectively, with different scanner-specific thresholds for clustering patients. In our population, we observed a mean RCA PCAT of −74.8 HU with a range consistent with these prior reports.

Our study is the first to investigate lumen-normalized PCAT to account for scan related differences. Controlling for differences in contrast would, in theory, minimize technical differences in interpretation of perivascular fat, since its density could differ depending on luminal contrast density. Normalization attempts to adjust for factors which may influence tissue attenuation and thus may affect PCAT assessment. Normalization addressed differences in X-ray tube settings during acquisition and may therefore allow more adequate comparison among CT systems and scan conditions.

We acknowledge several limitations of our study. CORE320 was neither designed nor powered with the intent to assess the prognostic utility of PCAT. As such, it is conceivable that lack of statistical significance observed in our analyses is due to insufficient power. Additionally, patient outcomes ideally focus only on hard events in this context instead of the broader definition of MACE (which includes revascularization). The number of hard events, however, was fairly small despite the high population risk. On the other hand, the total number of adverse events (87) was considerable, and we did not detect consistent trends among our results which warrant concern. Lastly, while we made meticulous attempts to consider factors influencing our PCAT analysis, we cannot exclude the possibility of unaccounted variables.

Conclusions

In our cohort of intermediate-high risk patients referred for invasive coronary angiography with suspected CAD, the perivascular fat attenuation index was not predictive of MACE. Further investigations are needed to understand the association of PCAT with atherogenesis and cardiac events.

Abbreviations

AUC

area under receiver operator characteristics curve

CABG

coronary artery bypass grafting

CAD

coronary artery disease

CCTA

coronary computed tomography angiogram

CI

confidence interval

HR

hazard ratio

HU

Hounsfield units

ICA

invasive coronary angiography

LAD

left anterior descending artery

LCx

left circumflex artery

MACE

major adverse cardiac event

MI

myocardial infarction

PCI

percutaneous coronary intervention

PCAT

pericoronary adipose tissue attenuation

RCA

right coronary artery