Impact of Pericoronary Adipose Tissue Attenuation on Periprocedural Myocardial Injury in Patients With Chronic Coronary Syndrome

Abstract

Background

Perivascular inflammation contributes to the development of atherosclerosis and microcirculatory dysfunction. Pericoronary adipose tissue (PCAT) attenuation, measured by coronary computed tomography angiography, is a potential indicator of coronary inflammation. However, the relationship between PCAT attenuation, microcirculatory dysfunction, and periprocedural myocardial injury (PMI) remains unclear.

Methods and Results

Patients with chronic coronary syndrome who underwent coronary computed tomography angiography before percutaneous coronary intervention were retrospectively identified. PCAT attenuation and adverse plaque characteristics were assessed using coronary computed tomography angiography. The extent of microcirculatory dysfunction was evaluated using the angio‐based index of microcirculatory resistance before and after percutaneous coronary intervention. Overall, 125 consecutive patients were included, with 50 experiencing PMI (PMI group) and 75 without PMI (non‐PMI group). Multivariable analysis showed that older age, higher angio‐based index of microcirculatory resistance, presence of adverse plaque characteristics, and higher lesion‐based PCAT attenuation were independently associated with PMI occurrence (odds ratio [OR], 1.07 [95% CI, 1.01–1.13]; P=0.02; OR, 1.06 [95% CI, 1.00–1.12]; P=0.04; OR, 6.62 [95% CI, 2.13–20.6]; P=0.001; and OR, 2.89 [95% CI, 1.63–5.11]; P<0.001, respectively). High PCAT attenuation was correlated with microcirculatory dysfunction before and after percutaneous coronary intervention and its exacerbation during percutaneous coronary intervention. Adding lesion‐based PCAT attenuation to the presence of adverse plaque characteristics improved the discriminatory and reclassification ability in predicting PMI.

Conclusions

Adding PCAT attenuation at the culprit lesion level to coronary computed tomography angiography‐derived adverse plaque characteristics may provide incremental benefit in identifying patients at risk of PMI. Our results highlight the importance of microcirculatory dysfunction in PMI development, particularly in the presence of lesions with high PCAT attenuation.

Registration

URL: https://center6.umin.ac.jp/cgi‐open‐bin/ctr/ctr_view.cgi?recptno=R000057722; Unique identifier: UMIN000050662.

Nonstandard Abbreviations and Acronyms

APC

adverse plaque characteristic

CCS

chronic coronary syndrome

IMR

index of microcirculatory resistance

PCAT

pericoronary adipose tissue

PMI

periprocedural myocardial injury

QFR

quantitative flow ratio

Clinical Perspective.

What Is New?

• High baseline pericoronary adipose tissue attenuation is independently associated with periprocedural myocardial infarction.

• Microvascular dysfunction contributes to the development of periprocedural myocardial infarction, particularly in the context of lesions characterized by high pericoronary adipose tissue attenuation.

What Are the Clinical Implications?

• Preoperative coronary computed tomography angiography evaluation, including pericoronary adipose tissue attenuation, has the potential to assist in determining the treatment strategy.

• It may be reasonable to defer invasive procedures in favor of statin treatment and plaque stabilization, or to guide optimal initiation of appropriate antiplatelet therapy, for chronic coronary syndrome patients with high pericoronary adipose tissue attenuation.

Percutaneous coronary intervention (PCI) effectively relieves symptoms and reduces the incidence of spontaneous myocardial infarction in patients with chronic coronary syndrome (CCS). However, periprocedural myocardial injury (PMI) occurs constantly, related to worse clinical outcomes after PCI in these patients. ¹ Previous studies have repeatedly reported that target plaque characteristics assessed using various imaging modalities can predict PMI after PCI. Nonetheless, adverse plaque characteristics (APCs) observed using these modalities at a single time point may not necessarily reflect the activity of the plaque during a long period of plaque development. ² Therefore, accurate PMI prediction based solely on image‐based plaque characterization remains challenging. ³ , ⁴ , ⁵

The cause of PMI is currently considered multifactorial. Distal coronary embolization of atheromatous material and intracoronary thrombus can result in PMI through no‐reflow/slow‐flow phenomena during PCI. In addition, coronary vasospasm may be involved due to thrombosis, neuro‐hormonal activation, and coronary microcirculatory dysfunction. Among several potential contributors, local inflammation is a key driver contributing to most of these underlying mechanisms. Inflammation promotes plaque progression, increases plaque vulnerability and instability, and accelerates atheromatous embolization, thrombogenicity, and microcirculatory dysfunction. ⁶ , ⁷ , ⁸ Therefore, we hypothesized that assessing local vascular inflammation before PCI might enhance the predictive accuracy for PMI over the traditional plaque characterization‐based approach using APCs.

Recent advancements in coronary computed tomography angiography (cCTA) have allowed for noninvasive evaluation of the local inflammatory status of coronary arteries. Notably, several previous studies have shown that cCTA has the feasibility of measuring epicardial adipose tissue that closely overlays coronary arteries, known as pericoronary adipose tissue (PCAT). PCAT is closely opposed to the coronary arteries; therefore, PCAT attenuation on cCTA indicates inflammatory activity within the underlying vascular wall. Previous studies reported a significant association between increased PCAT attenuation and future adverse events in patients with coronary artery disease. ⁹ , ¹⁰ Therefore, we aimed to clarify the relationship between PCAT attenuation before PCI and the occurrence of PMI. In addition, we evaluated the extent of microcirculatory dysfunction before and after PCI using a single‐view Murray's law‐based quantitative flow ratio (μQFR) analysis. The objective was to elucidate the interplay between APC, local inflammation, and microcirculatory dysfunction in developing PMI.

METHODS

Study Design

This retrospective, single‐center, observational study included consecutive patients with CCS who underwent cCTA before PCI. The study period was between January 2014 and December 2020. The inclusion criteria were patients with CCS who (1) underwent PCI with a drug‐eluting stent, (2) underwent cCTA within 90 days before PCI, (3) underwent serial measurements of high‐sensitivity cardiac troponin I (cTnI) before PCI and within 48 hours after PCI, and (4) were aged ≥20 years. The exclusion criteria are described in Data S1. The cTnI measurements were not explicitly ordered based on symptoms, ECG changes, or other specific indications. However, they were routinely obtained as part of our institution's standard protocol for PCI procedures. This study protocol complied with the Declaration of Helsinki and was approved by the Ethics Committee of Kobe University Hospital. Informed consent was obtained in the form of an opt‐out on the website of the Division of Cardiovascular Medicine, Kobe University Graduate School of Medicine. This study was registered in the University Hospital Medical Information Network Clinical Trial Registry (UMIN000050662). The data that support the findings of this study are available from the corresponding author on reasonable request.

Angiogram Analysis

In this study, μQFR analysis was performed using the QFR software (AngioPlus Core, version V3; Shanghai Pulse Medical Technology, Shanghai, China). The research was conducted by 2 experienced and certified analysts blinded to patients’ clinical characteristics pre‐ and post‐PCI. The detailed methodology for single‐view μQFR computation has been previously described. ¹¹ , ¹² , ¹³

Angiographic microvascular resistance was computed as P _d divided by the hyperemic flow velocity (Velocity_hyp).

$$\begin{array}{c}\text{Angiographic microvascular resistance}\left(\mathrm{mm}\mathrm{Hg}\times \mathrm{s}/\mathrm{cm}\right)\hfill \\ =P_{\mathrm{d}}/{\text{Velocity}}_{\mathrm{hyp}}=P_{\mathrm{a}}\times \text{μQFR}/{\text{Velocity}}_{\mathrm{hyp}}.\hfill \end{array}$$

The index of microcirculatory resistance (IMR) can be calculated as IMR [mm Hg×s/cm]=[angiographic microvascular resistance−0.9]/0.07, as previously reported. ¹⁴ In this computation, an average aorta pressure of 86 mm Hg during maximum hyperemia is assumed. ¹⁵ Δ IMR value was calculated as post‐PCI IMR minus pre‐PCI IMR.

In addition to the μQFR analysis, quantitative coronary angiography data were also available from the software. Quantitative coronary angiography included reference vessel diameter, minimum lumen diameter, percent diameter stenosis, and acute gain. The acute gain was defined as the difference between the minimum lumen diameter pre‐ and post‐PCI. Transient no‐reflow was described as an angiogram showing a deterioration of coronary flow of Thrombolysis In Myocardial Infarction (TIMI) grade 0, 1, or 2 during the procedure, regardless of the timing, and TIMI grade 3 at the final angiogram. The final slow flow was defined as an angiogram showing a deterioration of coronary flow of TIMI grade 1 or 2 at the final angiogram. ⁵

cCTA Imaging Acquisition

Initially, a noncontrast cardiac computed tomography (CT) was performed to assess the extent of coronary calcification through the Agatston coronary calcium score. cCTA images were obtained following the Society of Cardiovascular Computed Tomography guidelines on cCTA. ¹⁶ Oral β‐blockers were administered in subjects with heart rates ≥65 beats per minute. Immediately before cCTA acquisition, 0.3 mg sublingual nitroglycerine was administered to all patients. cCTA was performed using retrospective ECG‐gated spiral acquisition. All examinations were performed with Somatom Definition Flash (128×0.6 mm section collimation; Siemens, Forchheim, Germany). The scan parameters included 70 to 120 kilovolt peak (kVp) tube voltage and 260 to 1150 mA tube current (depending on body habitus). All images were reconstructed using thin slices (0.5–0.75 mm) and medium smooth reconstruction filters in different phases.

cCTA Image Analysis of Plaque Characteristics

The cCTA image analysis was performed using a semi‐automated plaque analysis workstation (SYNAPSE VINCENT; Fujifilm, Tokyo, Japan) with an additional minor manual optimization to obtain qualitative plaque features. cCTA images were analyzed by 2 independent investigators blinded to patients' clinical characteristics except for the information on the culprit lesion location. The plaque measurements included absolute volumes (in cubic millimeters) and the corresponding burden (plaque volume×100%/vessel volume) of calcified plaques (>350 Hounsfield units [HU]) and noncalcified plaques (≤350 HU), as well as the remodeling index, lesion length, and diameter stenosis. Noncalcified plaques were further divided into their components: low‐attenuation plaque (−30 to 30 HU), intermediate‐attenuation plaque (31–130 HU), and high‐attenuation plaque (131–350 HU) volumes, and the corresponding plaque burden. ¹⁷ The presence of APCs (low‐attenuation plaque, positive remodeling, spotty calcification, and napkin‐ring sign) was assessed in each lesion following the definitions from previous studies. ² , ¹⁸ A detailed methodology of cCTA image analysis is described in the Supplemental Material.

PCAT Analysis

PCAT analysis was performed using a workstation (SYNAPSE VINCENT; Fujifilm, Tokyo, Japan). To measure PCAT attenuation, 3‐dimensional layers within radial distance from the outer coronary wall equal in thickness to the average diameter of the vessel were constructed automatically from the CTA. Within the predefined volume of interest, voxels with tissue attenuation ranging from −190 to −30 HU were considered adipose tissue. The automated analysis did not include the myocardial tissue adjacent to the vessel wall and the coronary branches originating from the vessel of interest. PCAT attenuation was defined as the mean attenuation within such contamination‐free volumes of interest. These measurements were performed in each patient around the specific culprit lesions (PCAT_Lesion), the proximal culprit vessels (PCAT_Vessel), and the proximal right coronary artery (RCA) (PCAT_RCA). PCAT_Vessel and PCAT_RCA were measured to compare the diagnostic ability for the occurrence of PMI among lesion, vessel, and patient‐level PCAT attenuation. PCAT_Lesion was measured around target lesions, with the proximal and distal borders of the analysis region defined as the proximal and distal ends of the lesion. ¹⁹ , ²⁰ The proximal 40‐mm segments of the left anterior descending coronary artery and left circumflex coronary artery, and the proximal 10‐ to 50‐mm segment of the RCA were traced, as previously described. ⁹ PCAT_RCA represented PCAT measurements at the patient level. ⁹ , ¹⁰ Inter‐ and intraobserver variabilities in assessing PCAT attenuation for all cases were quantified using κ concordance analysis.

Outcomes

The study's primary outcome was the occurrence of PMI. PMI was defined as >5×99th percentile upper reference limit increase of high‐sensitivity cTnI values within 48 hours after PCI in patients with CCS with normal baseline cTnI values ¹ (normal reference range: <24 pg/mL).

Statistical Analysis

All statistical analyses were performed using the Microsoft R open software version 3.4.1 (R Development Core Team, Vienna, Austria). Continuous variables are presented as mean±SD or median (25th, 75th percentiles) and compared using the Student t test for normally distributed data or the Mann‐Whitney U test for nonnormally distributed data. Data at different time points were analyzed using the paired Student t test or Wilcoxon test, as appropriate. Categorical variables are presented as frequencies with percentages and were compared using the χ² or Fisher exact test, as appropriate. Statistical significance was set at a 2‐sided P value <0.05. The variables included in the multivariable logistic regression analysis encompassed baseline clinical characteristics (age, hypertension, dyslipidemia, diabetes, current smoking), cCTA findings (lesion length, percent diameter stenosis, total plaque volume, low‐attenuation plaque volume, presence of any APCs), PCAT variables, and pre‐PCI IMR. Within the multivariable logistic regression analysis, significant correlations emerged among specific covariates. The presence of any APCs, which encompasses individual components of APCs such as low‐attenuation plaque, positive remodeling, spotty calcification, and the napkin‐ring sign, were identified as related factors. Additionally, significant correlations were observed among PCAT_Lesion, PCAT_Vessel, and PCAT_RCA. To comprehensively investigate the relationship between each variable and the occurrence of PMI, we used multiple models, which are as follows: Model A (comprising low‐attenuation plaque, positive remodeling, spotty calcification, napkin‐ring sign, and PCAT_Lesion), Model B (involving the presence of any APCs and PCAT_Lesion), Model C (involving the presence of any APCs and PCAT_Vessel), and Model D (involving the presence of any APCs and PCAT_RCA). The area under the curve (AUC) was calculated from receiver operating characteristic (ROC) curve analysis for PCAT_Lesion, PCAT_Vessel, and PCAT_RCA, and compared using the DeLong method. Three prediction models for PMI were constructed to determine the incremental discriminatory and reclassification performance of PCAT attenuation. As a baseline, clinical Model 1 was derived from cardiovascular risk factors (age, hypertension, diabetes, dyslipidemia, and current smoking) and quantitative cCTA analysis (lesion length, percent diameter stenosis, and total plaque volume). Clinical Model 2 was constructed using clinical Model 1 + cCTA plaque quality analysis (the presence of APCs and low‐attenuation plaque volume). Clinical Model 3 was derived from clinical Model 2 + PCAT attenuation at the culprit lesion level. The discriminatory ability of Models 2 and 3 was assessed by ROC analysis, and the reclassification performance of each model was compared using the relative integrated discrimination improvement and category‐free net reclassification index. ROC analysis was used to determine the optimal cutoff value of PCAT_Lesion associated with PMI according to tube voltage as follows: 70 or 100 to 120 kVp.

RESULTS

Study Population

In total, 253 patients with CCS underwent PCI with prior cCTA within 90 days before PCI during the study period. Among them, 201 patients had serial cTnI measurements performed before PCI and within 48 hours after PCI and met the inclusion criteria. After excluding patients based on various criteria, including patients with a positive cTnI before PCI (n=9), patients with coronary artery bypass graft lesions (n=7), patients with chronic total occlusion (n=11), patients with in‐stent restenosis (n=3), patients with severely calcified lesions (n=16), patients with left main trunk lesions (n=10), patients with >1 main coronary artery intervention during single PCI procedure (n=9), patients with side branch occlusion during PCI procedure (n=3), patients with cardiogenic shock during PCI (n=2), and patients with insufficient CT data quality (n=6), 125 patients were finally included in the analysis. Among these patients, 50 (40.0%) experienced PMI (PMI group), whereas 75 (60.0%) did not experience PMI (non‐PMI group) (Figure S1).

Comparison of Baseline Characteristics Between PMI and Non‐PMI Groups

The baseline patient characteristics are shown in Table 1. Patients with PMI were significantly older than those with non‐PMI. The prevalence of coronary risk factors and medications on admission were similar between the 2 groups. CK‐MB and cTnI levels before PCI were comparable between the 2 groups; however, those after PCI in the PMI group were significantly higher than those in the non‐PMI group (7.5 [5–11] U/L versus 4 [4–4.5] U/L, P<0.001; 920 [482.5–1890] ng/mL versus 70 [45–80] ng/mL, P<0.001; respectively).

Table 1.

Baseline Patient Characteristics

Variables	All lesions (n=125)	PMI (n=50)	Non‐PMI (n=75)	P value
Baseline patient characteristics
Age, y	69.0±9.2	73.0±8.6	69.1±11.4	0.046
Sex, male, n (%)	95 (76.0%)	36 (72.0%)	59 (78.7%)	0.40
Hypertension, n (%)	89 (71.2%)	36 (72.0%)	53 (70.1%)	0.87
Dyslipidemia, n (%)	87 (69.6%)	36 (72.0%)	51 (68.0%)	0.64
Diabetes, n (%)	54 (43.2%)	22 (44.0%)	32 (42.3%)	0.88
Current smoker, n (%)	58 (46.4%)	24 (48.0%)	34 (45.3%)	0.77
Prior PCI, n (%)	20 (16.0%)	11 (22.0%)	9 (12.0%)	0.14
Chronic kidney disease, n (%)	48 (38.4%)	20 (40.0%)	28 (37.3%)	0.77
Hemodialysis, n (%)	6 (4.8%)	4 (8.0%)	2 (2.7%)	0.18
LVEF, %	59.7±8.7	58.4±10.8	60.5±6.9	0.18
Laboratory data
BNP, pg/mL	40.5 (18.8, 96.1)	43.4 (21.4, 136.9)	36.1 (18.8, 88.9)	0.23
Estimated GFR, mL/min per 1.73 m2	60.6±18.2	59.9±21.1	61.2±15.9	0.70
Low‐density lipoprotein cholesterol, mg/dL	105.8±30.3	104.9±29.5	106.4±30.8	0.78
High‐density lipoprotein cholesterol, mg/dL	46.6±12.7	46.5±11.9	46.7±13.1	0.94
HbA1c, %	6.4±1.1	6.3±1.0	6.5±1.2	0.32
WBC count, per μL	6212±1764	6281±1667	6110±1895	0.60
hs‐CRP, mg/L	0.50±1.56	0.52±1.46	0.49±1.63	0.91
Pre‐CK‐MB, U/L	4 (4, 5)	4 (4, 5)	4 (4, 5)	0.68
Pre‐cTnI, pg/mL	10 (10, 20)	10 (10, 20)	10 (10, 20)	0.71
Post‐CK‐MB, U/L	4 (4, 7)	7.5 (5, 11)	4 (4, 4.5)	<0.001
Post‐cTnI, pg/mL	90 (60, 700)	920 (482.5, 1890)	70 (45, 80)	<0.001
Medications
Statins, n (%)	72 (57.6%)	29 (58.0%)	43 (57.3%)	0.94
β‐Blockers, n (%)	67 (53.6%)	27 (54.0%)	40 (53.3%)	0.94
RAS inhibitors, n (%)	72 (57.6%)	26 (52.0%)	46 (61.3%)	0.30
Calcium channel blockers, n (%)	54 (43.2%)	22 (44.0%)	32 (42.7%)	0.88

Values are expressed as mean±SD, median (25th, 75th percentiles), or n (%). BNP indicates brain‐type natriuretic peptide; CK‐MB, creatine kinase‐MB; cTnI, cardiac troponin I; GFR, glomerular filtration rate; HbA1c, hemoglobin A1c; LVEF, left ventricular ejection fraction; hs‐CRP, high‐sensitivity C‐reactive protein; PCI, percutaneous coronary intervention; PMI, periprocedural myocardial injury; RAS, renin‐angiotensin system; and WBC, white blood cell.

The baseline lesion, procedural, and angiographic characteristics are shown in Table 2. The 2 groups had no significant differences in target vessels, lesion location, number of implanted stents, and stent diameter. Stent length in the PMI group was significantly longer than that in the non‐PMI group. The occurrence of transient no‐reflow and final slow flow was significantly higher in the PMI group than in the non‐PMI group. Pre‐PCI reference vessel diameter, minimum lumen diameter, percent diameter stenosis, and μQFR were comparable; however, pre‐PCI IMR was significantly higher in the PMI group than in the non‐PMI group. In addition, post‐PCI IMR and Δ IMR were significantly higher in the PMI group than in the non‐PMI group. IMR levels were significantly increased during PCI in the PMI group but not in the non‐PMI group (Table 2). If patients with transient or final no‐reflow phenomenon were excluded, post‐PCI IMR values remained higher in the PMI group (n=39) than those in the non‐PMI group (n=74) (42.9±10.4 mm Hg×s/cm versus 27.0±6.7 mm Hg×s/cm, P<0.001).

Table 2.

Lesion, Procedural, and Angiographic Characteristics

Variables	All lesions (n=125)	PMI (n=50)	Non‐PMI (n=75)	P value
Lesion characteristics
Target vessel: LAD/LCX/RCA, %	53.6/16.0/30.4	56.0/20.0/24.0	52.0/13.3/34.7	0.39
Lesion location: proximal/mid/distal, %	36.8/56.8/6.4	30.0/66.0/4.0	41.3/50.7/8.0	0.24
Multivessel disease	73 (58.4%)	30 (60%)	43 (57.3%)	0.77
Procedural characteristics
No. of stents	1.21±0.43	1.28±0.45	1.16±0.40	0.12
Stent diameter, mm	3.10±0.43	3.03±0.35	3.15±0.48	0.14
Stent length, mm	30.0±14.5	34.2±14.9	27.1±13.5	0.007
Distal protection device, n (%)	13 (10.4%)	7 (14.0%)	6 (8.0%)	0.29
Angiographic analysis
Pre‐PCI
Reference vessel diameter, mm	2.68±0.49	2.58±0.44	2.75±0.51	0.17
Minimum lumen diameter, mm	1.08±0.40	0.98±0.31	1.04±0.25	0.20
Diameter stenosis, %	57.6±11.9	59.3±10.0	57.2±11.1	0.30
Post‐PCI
Minimum lumen diameter, mm	2.39±0.39	2.34±0.27	2.42±0.45	0.29
Diameter stenosis, %	14.6±11.1	15.3±10.9	14.1±11.2	0.57
Transient no‐reflow, n (%)	12 (9.6%)	11 (22.0%)	1 (1.3%)	<0.01
Final slow flow, n (%)	5 (4.0%)	5 (10.0%)	0 (0.0%)	0.005
Acute gain, mm	1.36±0.58	1.43±0.48	1.31±0.63	0.27
QFR analysis
μQFR pre‐PCI	0.68±0.16	0.69±0.13	0.68±0.17	0.96
μQFR post‐PCI	0.89±0.09	0.92±0.06	0.87±0.10	0.01
IMR analysis
IMR pre PCI, mm Hg×s/cm	29.8±9.7	34.5±8.1	26.6±9.5	<0.001
IMR post‐PCI, mm Hg×s/cm	33.4±11.5	43.0±10.5*	27.0±6.6	<0.001
ΔIMR, mm Hg×s/cm	3.6±9.3	8.6±9.4	0.4±7.6	<0.001

Values are expressed as mean±SD or n (%). μQFR indicates Murray's law‐based quantitative flow ratio; IMR, index of microvascular resistance; LAD, left anterior descending artery; LCX, left circumflex artery; PCI, percutaneous coronary intervention; PMI, periprocedural myocardial injury; QFR, quantitative flow ratio; and RCA, right coronary artery.

^*Indicates P<0.05 vs IMR pre‐PCI.

Comparison of cCTA Findings Between PMI and Non‐PMI Groups

Table 3 summarizes cCTA findings. The 2 groups had no statistically significant differences in percent diameter stenosis, lesion external elastic membrane CSA, minimum lumen area, and total plaque volume. However, lesion length and remodeling index were significantly greater, low‐attenuation plaque volume was numerically larger, and the CT attenuation value of the culprit plaque was significantly lower in the PMI group than in the non‐PMI group. Spotty calcification and plaques with napkin‐ring signs were significantly more frequent in the PMI group than in the non‐PMI group. In addition, the prevalence of any APCs was significantly higher in the PMI group than in the non‐PMI group (72.0% versus 50.7%, P=0.017). As for PCAT attenuation, PCAT_Lesion, PCAT_Vessel, and PCAT_RCA were significantly higher in the PMI group than in the non‐PMI group (all P values <0.01). Inter‐ and intraobserver agreements for measuring PCAT attenuation were within the acceptable range (PCAT_Lesion: intraobserver, 0.975; interobserver, 0.918; PCAT_Vessel: intraobserver, 0.980; interobserver, 0.930; PCAT_RCA: intraobserver, 0.988; interobserver, 0.955).

Table 3.

cCTA Findings

Variables	All lesions (n=125)	PMI (n=50)	Non‐PMI (n=75)	P value
cCTA settings
Tube voltage, n (%)				0.46
70 kVp	61 (48.8%)	21 (42%)	40 (53.3%)
100–120 kVp	64 (51.2%)	29 (58%)	35 (46.7%)
Quantitative cCTA analysis
Lesion length, mm	24.6±11.0	27.6±11.9	22.6±9.8	0.012
Diameter stenosis, %	72.0±19.0	72.6±21.2	71.5±17.4	0.77
Lesion EEM CSA, m2	11.5±5.2	11.8±5.1	11.3±5.2	0.65
Lesion MLA, m2	1.05±0.72	0.94±0.49	1.12±0.83	0.18
Remodeling index	1.13±0.35	1.20±0.35	1.08±0.34	0.048
CT attenuation value, HU	58.1±29.1	44.4±26.2	67.2±27.3	<0.001
Coronary artery calcium score	341 (51, 808)	341 (61, 707)	327 (48, 954)	0.82
Total plaque volume, mm3	261.9±144.9	272.1±140.3	255.0±147.6	0.52
cCTA plaque quality analysis
Plaque composition volume, mm3
Low‐attenuation plaque	53.9±16.0	57.3.±15.3	51.6±16.1	0.052
Intermediate‐attenuation plaque	80.2±72.1	83.2±59.2	78.2±79.3	0.71
High‐attenuation plaque	109.1±84.9	113.5±83.1	106.2±86.0	0.64
Calcified plaque	18.7±28.5	18.2±26.7	19.1±29.6	0.87
Plaque composition burden, %
Low‐attenuation plaque	26.4±17.2	27.6±21.2	25.6±13.8	0.52
Intermediate‐attenuation plaque	28.9±17.1	29.9±17.4	28.3±16.9	0.60
High‐attenuation plaque	37.9±13.7	36.5±15.8	39.0±12.0	0.32
Calcified plaque	6.7±9.5	6.0±8.0	7.2±10.3	0.49
Presence of APCs	74 (59.2%)	36 (72.0%)	38 (50.7%)	0.017
Positive remodeling, n (%)	49 (39.2%)	24 (48.0%)	25 (33.3%)	0.10
Low attenuation plaque, n (%)	27 (21.6%)	15 (30.0%)	12 (16.0%)	0.063
Spotty calcification, n (%)	41 (32.8%)	24 (52.2%)	17 (22.7%)	0.011
Napkin‐ring sign, n (%)	20 (16.0%)	12 (24.0%)	8 (10.7%)	0.047
PCAT attenuation analysis
PCATLesion, HU	−81.3±11.4	−75.9±10.9	−84.9±10.3	<0.001
PCATVessel, HU	−82.8±11.7	−79.1±11.7	−85.3±11.1	0.004
2003PCATRCA, HU	−83.1±11.1	−78.9±10.1	−85.7±11.0	<0.001

Values are expressed as mean±SD, median (25th, 75th percentiles), or n (%). APCs indicates adverse plaque characteristics; cCTA, coronary computed tomography angiography; CSA, cross sectional area; EEM, external elastic membrane; HU, Hounsfield units; kVp, kilovolt peak; MLA, minimum lumen area; PCAT, pericardial adipose tissue; PMI, periprocedural myocardial injury; and RCA, right coronary artery.

Factors Associated With PMI

The results of the univariate and multivariable logistic regression analysis examining clinical factors and cCTA findings associated with PMI are summarized in Table 4. The multivariable Model A showed that age, pre‐PCI IMR, and PCAT_Lesion were independently associated with PMI. The multivariable Models B, C, and D showed that in addition to older age, the presence of any APCs, pre‐PCI IMR, and PCAT‐related variables such as PCAT_Lesion (Model B; odds ratio [OR], 2.89 [95% CI, 1.63–5.11]; P<0.001), PCAT_Vessel (Model C; OR, 1.96 [95% CI, 1.22–3.16]; P=0.006), and PCAT_RCA (Model D; OR, 1.69 [95% CI, 1.14–2.53]; P=0.010) were independently associated with the occurrence of PMI. To improve the validity and interpretability of the results, we have reduced the variables in the multivariable models in Table S1. After reducing the variables, the multivariable model consistently showed that PCAT_Lesion (OR, 2.50 [95% CI, 1.48–4.21]; P<0.001) was independently associated with the occurrence of PMI (Table S1). The AUC of the presence of any APCs for identifying PMI was 0.61 (95% CI, 0.52–0.69), and the sensitivity, specificity, positive predictive value, and negative predictive value of the presence of any APCs alone were 72.0%, 49.3%, 48.6%, and 72.5%, respectively. The AUC of PCAT_Lesion for identifying PMI was significantly higher than that of PCAT_Vessel and PCAT_RCA in diagnostic performance (Figure 1).

Table 4.

Uni‐ and Multivariable Logistic Regression Analysis of Baseline Patient Characteristics, cCTA Findings, and Angiogram Analysis Associated With PMI

Variables	Univariate analysis		Multivariable Model A		Multivariable Model B		Multivariable Model C		Multivariable Model D
	OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value	OR (95% CI)	P value
Baseline patient characteristics
Age	1.04 (1.00–1.08)	0.049	1.09 (1.02–1.16)	0.01	1.07 (1.01–1.13)	0.02	1.05 (1.00–1.11)	0.04	1.04 (0.99–1.09)	0.12
Hypertension	1.07 (0.48–2.36)	0.87	0.95 (0.23–3.11)	0.93	1.03 (0.33–3.18)	0.96	0.97 (0.33–2.82)	0.96	1.02 (0.35–2.97)	0.97
Dyslipidemia	1.21 (0.55–2.65)	0.63	1.00 (0.28–3.52)	1.00	1.33 (0.45–3.91)	0.60	1.29 (0.46–3.66)	0.63	1.33 (0.46–3.86)	0.59
Diabetes	1.06 (0.51–2.17)	0.88	1.11 (0.38–3.25)	0.85	1.37 (0.50–3.74)	0.54	1.50 (0.56–4.02)	0.42	1.49 (0.55–4.02)	0.43
Current smoking	1.11 (0.54–2.28)	0.77	1.10 (0.36–3.39)	0.87	0.85 (0.32–2.30)	0.75	0.92 (0.36–2.33)	0.86	1.00 (0.39–2.57)	0.99
cCTA findings
Lesion length (per 5‐mm increase)	1.24 (1.01–1.46)	0.015	1.69 (0.87–2.43)	0.07	1.36 (0.99–1.85)	0.06	1.32 (0.99–1.76)	0.06	1.35 (0.99–1.80)	0.06
Diameter stenosis, % (per 1% increase)	1.00 (0.98–1.02)	0.76	0.99 (0.96–1.02)	0.46	0.99 (0.97–1.02)	0.56	0.99 (0.97–1.02)	0.54	0.99 (0.96–1.02)	0.42
Total plaque volume (per 1‐m3 increase)	1.00 (0.99–1.00)	0.52	0.99 (0.99–1.02)	0.12	0.99 (0.99–1.00)	0.15	1.00 (0.99–1.00)	0.18	1.00 (0.99–1.00)	0.31
Low‐attenuation plaque volume (per 1‐m3 increase)	1.02 (0.99–1.05)	0.06	1.02 (0.98–1.05)	0.34	1.01 (0.98–1.04)	0.48	1.02 (0.99–1.05)	0.29	1.02 (0.99–1.05)	0.24
Presence of any APCs	2.50 (1.16–5.38)	0.019			6.62 (2.13–20.6)	0.001	5.49 (1.90–15.9)	0.002	5.69 (1.96–16.5)	0.001
Low‐attenuation plaque	2.25 (0.95–5.34)	0.066	3.71 (0.99–13.8)	0.06
Positive remodeling	1.85 (0.89–3.85)	0.10	3.65 (0.90–12.8)	0.07
Spotty calcification	2.68 (1.23–5.83)	0.013	1.63 (0.49–5.47)	0.43
Napkin‐ring sign	2.64 (0.99–7.04)	0.052	5.37 (0.98–26.9)	0.06
PCAT analysis
PCATLesion (per 10‐HU increase)	2.30 (1.53–3.45)	<0.001	2.98 (1.64–5.42)	<0.001	2.89 (1.63–5.11)	<0.001
PCATVessel (per 10‐HU increase)	1.63 (1.16–2.29)	0.005					1.96 (1.22–3.16)	0.006
PCATRCA (per 10‐HU increase)	1.49 (1.11–2.0)	0.008							1.69 (1.14–2.53)	0.010
Pre‐PCI IMR (per 5‐mm Hg×s/cm increase)	1.61 (1.28–2.02)	<0.001	1.06 (1.02–1.12)	0.03	1.06 (1.00–1.12)	0.04	1.08 (1.02–1.14)	0.005	1.09 (1.03–1.15)	0.003

APC indicates adverse plaque characteristics; cCTA, coronary computed tomography angiography; HU, Hounsfield units; IMR, index of microcirculatory resistance; OR, odds ratio; PCAT, pericoronary adipose tissue; PCI, percutaneous coronary intervention; PMI, periprocedural myocardial injury; and RCA, right coronary artery.

Figure 1. Diagnostic ability of PCAT attenuation at the culprit lesion, culprit vessel, and patient levels to identify PMI.

Comparison of PCAT_Lesion, PCAT_Vessel, and PCAT_RCA receiver operating characteristic curves and the AUCs are shown. AUC indicates area under the curve; PCAT, pericoronary adipose tissue; PMI, periprocedural myocardial injury; and RCA, right coronary artery.

Discriminatory Diagnostic Ability by the Addition of Factors for PMI

Three prediction models were constructed to determine the incremental discriminatory and reclassification performance of cCTA findings and PCAT attenuation. Figure 2 shows ROC curves for the 3 models and their predictive performance. Compared with Model 1 (cardiovascular risk factors and quantitative cCTA analysis), Model 2 (Model 1+ cCTA plaque quality analysis) showed a numerically higher discriminatory ability (AUC: 0.74 versus 0.67; P=0.06) and a significantly higher reclassification ability (net reclassification index: 0.49; P=0.005; relative integrated discrimination improvement: 0.08; P<0.001) for identifying the occurrence of PMI. In addition, Model 3 (Model 2 + PCAT_Lesion) showed a significant increase in discriminatory ability (AUC: 0.85 versus 0.74; P=0.01) and incremental reclassification ability (net reclassification index: 0.95; P<0.001; relative integrated discrimination improvement: 0.19; P<0.001) compared with Model 2 (Figure 2). When limiting the predictive model to the most important variables, PCAT continued to demonstrate incremental predictive value (Figure S2).

Figure 2. Comparison between discriminant and reclassification ability of predictive models for the presence of PMI.

To determine incremental discriminatory and reclassification capacities of PCAT attenuation at the culprit lesion level, we constructed 3 analytical models as follows: Model 1: cardiovascular risk factors and quantitative cCTA analysis (black line); Model 2: Model 1 + cCTA plaque quality analysis (blue line); Model 3: Model 2+PCAT attenuation at culprit lesion level (red line). %DS indicates percent diameter stenosis; APCs, adverse plaque characteristics; AUC, area under the curve; cCTA, coronary computed tomography angiography; DL, dyslipidemia; DM, diabetes mellitus; HT, hypertension; IDI, integrated discrimination improvement; LAP, low‐attenuation plaque; NRI, net reclassification index; PCAT, pericoronary adipose tissue; PMI, periprocedural myocardial injury; and TPV, total plaque volume.

Incidence of PMI and the Risk for PMI According to the Presence of APC and PCAT Attenuation Grade

ROC analysis of PCAT_Lesion showed that the cutoff value for identifying patients with PMI was −80.7 HU (sensitivity, 66.7%; specificity, 82.5%; AUC, 0.75 [95% CI, 0.62–0.88]) on 70‐kVp images and −72.9 HU (sensitivity, 55.2%; specificity, 85.7%; AUC, 0.72 [95% CI, 0.59–0.85]) on 100‐ to 120‐kVp images. Based on these cutoff values, the lesions were categorized into high PCAT (PCAT_Lesion ≥−80.7 HU on 70‐kVp images or PCAT_Lesion ≥−72.9 HU on 100‐ to 120‐kVp images) or low PCAT (PCAT_Lesion <−80.7 HU on 70‐kVp images or PCAT_Lesion <−72.9 HU on 100‐ to 120‐kVp images). When lesions were classified based on the presence of APCs and PCAT attenuation grade, the incidence of PMI was the highest in lesions with both APCs (+) and high PCAT (Figure 3A), which showed a higher risk for PMI than those with APCs (−) and low PCAT (OR, 24.7 [95% CI, 5.86–105.0]; P<0.001) (Figure 3B). Representative cases are described in Figure S3.

Figure 3. Incidence of PMI and the risk for PMI according to the presence of APCs and PCAT attenuation grade.

A, Incidence of PMI. All lesions were classified according to the presence of APCs and PCAT attenuation grade. The incidence of PMI was significantly higher in lesions with APCs and high PCAT. B, Odds ratio for PMI. Lesions with APCs and high PCAT attenuation showed significantly higher risk than those with no APCs and low PCAT. APC indicates adverse plaque characteristic; PCAT, pericoronary adipose tissue; and PMI, periprocedural myocardial injury.

Relationship Between PCAT Attenuation, Microcirculatory Dysfunction, and PMI

A significant correlation was found between PCAT attenuation and IMR, both pre‐PCI (r=0.324, P<0.001) and post‐PCI (r=0.429, P<0.001), and there was a weak correlation between PCAT attenuation and Δ IMR (r=0.19, P=0.03). Pre‐ and post‐PCI IMR and Δ IMR were significantly higher in the lesions with high PCAT attenuation than in those with low PCAT attenuation (Figure 4). Moreover, a significant correlation was observed between post‐PCI cTnI levels and PCAT attenuation (r=0.29, P<0.001). There were also significant correlations between post‐PCI cTnI levels and post‐PCI IMR (r=0.43, P<0.001) or Δ IMR (r=0.30, P<0.001) (Figure 5).

Figure 4. Correlations between IMR‐related variables and PCAT attenuation, and comparison of IMR‐related variables according to PCAT attenuation grade.

A, Pre‐PCI IMR. A significant correlation was found between PCAT attenuation and pre‐PCI IMR (r=0.324, P<0.001). Pre‐PCI IMR in the lesions with high PCAT was significantly higher than in those with low PCAT (34.9 mm Hg×s/cm vs 26.2 mm Hg×s/cm, P<0.001). B, Post‐PCI IMR. A significant correlation was found between PCAT attenuation and post‐PCI IMR (r=0.429, P<0.001). Post‐PCI IMR in the lesions with high PCAT was significantly higher than in those with low PCAT (42.4 mm Hg×s/cm vs 27.2 mm Hg×s/cm, P<0.001). C, Δ IMR. There was a weak correlation between PCAT attenuation and Δ IMR (r=0.19, P=0.03). Δ IMR in the lesions with high PCAT was significantly higher than in those with low PCAT (6.8 mm Hg×s/cm vs 1.0 mm Hg×s/cm, P<0.001). IMR indicates index of microvascular resistance; PCAT, pericoronary adipose tissue; and PCI, percutaneous coronary intervention.

Figure 5. Correlations between post‐PCI cTnI level and PCAT attenuation or IMR‐related variables.

A, PCAT attenuation. A significant correlation was found between PCAT attenuation and post‐PCI cTnI (r=0.29, P<0.001). B, Pre‐PCI IMR. A weak correlation was found between pre‐PCI IMR and post‐PCI cTnI (r=0.22, P=0.015). C, Post‐PCI IMR. A significant correlation was found between post‐PCI IMR and post‐PCI cTnI (r=0.43, P<0.001). D, Δ IMR. A significant correlation was found between Δ IMR and post‐PCI cTnI (r=0.30, P<0.001). cTnT indicates cardiac troponin I; IMR, index of microvascular resistance; PCAT, pericoronary adipose tissue; and PCI, percutaneous coronary intervention.

DISCUSSION

To our knowledge, this is the first study investigating the relationship between PCAT attenuation measured using pre‐PCI cCTA imaging and the occurrence of PMI among patients with CCS. This study's main findings can be summarized as follows: (1) Patients who experienced PMI after PCI had a higher level of vascular inflammation as indicated by increased PCAT attenuation at the culprit lesion, culprit vessel, and patient level. (2) In addition to older age and cCTA‐derived APC, increased pre‐PCI PCAT attenuation measured using cCTA and microcirculatory resistance assessed using angio‐derived IMR were independently associated with the occurrence of PMI. (3) PCAT attenuation at the culprit lesion level demonstrated superior predictive ability in identifying PMI compared with PCAT attenuation at the culprit vessel and patient level. (4) PCAT attenuation at the culprit lesion level provided incremental predictive value over the model incorporating clinical risk factors and traditional cCTA findings, including APCs, for identifying patients at risk of PMI. (5) Microcirculatory resistance significantly increased after PCI in patients with PMI but not in those without PMI. Moreover, high PCAT attenuation was significantly associated with microcirculatory dysfunction before and after PCI and its exacerbation during PCI. These findings highlight the importance of PCAT attenuation as a potential marker of vascular inflammation and its association with PMI risk. In addition, the study underscores the role of microcirculatory dysfunction in developing PMI, particularly in the context of lesions characterized by high PCAT attenuation.

Incremental Value of Inflammation to APC

Distal coronary embolization of intracoronary thrombus and atheromatous material has been recognized as a major cause of PMI. ¹ APCs observed on cCTA, such as positive remodeling, low attenuation plaque, spotty calcification, and napkin‐ring sign, indicate plaque vulnerability with higher lipid content. ²¹ , ²² Consequently, plaques with APCs may be susceptible to fragmentation into numerous atherosclerotic particles after stent expansion, resulting in PMI. A prior cCTA study demonstrated a significant association between APCs on cCTA and post‐PCI cTnT elevation. ⁵ However, a previous serial cCTA study ² showed that many plaques with APCs remained stable and event‐free >1 year, whereas some developed instability, ultimately resulting in acute coronary syndrome. This finding suggests that the presence of APCs alone may not necessarily predict the likelihood of plaque instability or its activity. ²³ The present study indicated low positive predictive values for the presence of APCs alone in predicting PMI (sensitivity: 72.0%, specificity: 49.3%, positive predictive value: 48.6%, negative predictive value: 72.5%). Therefore, we currently consider that assessing APCs alone using cCTA may be insufficient; other factors should be evaluated with cCTA to better predict PMI.

Inflammation plays a vital role in collagen degradation in the fibrous cap, which precedes fibrous cap rupture, by releasing collagen‐degrading enzymes. In addition, inflammation significantly contributes to plaque thrombogenicity by providing a source of tissue factor. ²⁴ Therefore, it is conceivable that inflammation increases plaque instability and thrombogenicity, predisposing to intraprocedural microembolization with subsequent microvascular injury or obstruction after PCI. ²⁵ Recently, cCTA has emerged as a noninvasive method to evaluate PCAT attenuation, potentially representing adjacent vessels' inflammatory status. In a landmark study, Antonopoulos et al demonstrated a relationship between coronary perivascular fat attenuation and inflammation in and around the coronary artery. ¹⁰ In the present study, we evaluated PCAT attenuation as a measure of local coronary inflammation to obtain a more accurate risk stratification of PMI.

Consequently, we found that increased PCAT attenuation was independently associated with the occurrence of PMI. Previous studies have clarified the additive value of PCAT attenuation over APC features. Kwiecinski et al showed a correlation between PCAT attenuation and 18F‐sodium fluoride (18F‐NaF) activity, which provided information on plaque activity. ¹⁹ In this study, not all plaques with APCs on cCTA manifested plaque activity as assessed by 18F‐NaF uptake or altered PCAT attenuation, suggesting that the presence of APCs alone may not necessarily represent plaque activity. Moreover, Oikonomou et al ²⁶ demonstrated that APC features with low inflammation assessed by PCAT attenuation were not associated with increased cardiac risk. However, when inflammation is present, APC features can identify a particularly high‐risk group of patients. Considering these findings, assessment of APCs alone may be insufficient for plaque risk assessment; evaluating plaque activity by measuring increased PCAT attenuation could be important. Therefore, we speculate that inflammation assessed by PCAT attenuation could provide incremental value to the evaluation of APC, providing a more accurate risk stratification of subsequent PMI after PCI. Our study demonstrated that PCAT attenuation at the culprit lesion level provided incremental predictive value over models incorporating clinical risk factors and traditional cCTA findings, including APCs, in identifying patients at risk of PMI. Furthermore, lesions with both the presence of APCs and high PCAT attenuation had a 24.7‐fold higher risk of PMI than other lesions. These findings suggest that assessing inflammation through PCAT attenuation on cCTA may help detect lesions susceptible to PMI, offering advantages over a simple evaluation based solely on APCs.

Relationship Between Vascular Inflammation, Microcirculatory Dysfunction, and PMI

Coronary microcirculatory dysfunction represents another mechanism contributing to the occurrence of PMI. A previous study demonstrated a significant association between baseline and post‐PCI IMR values and PMI. Periprocedural IMR changes were significantly higher in patients who developed PMI, with post‐PCI IMR emerging as a stronger and independent predictor of PMI occurrence. ²⁷ Consistent with these findings, the present study also revealed a significant association between microcirculatory dysfunction, as assessed using angio‐based pre‐ and post‐PCI IMR, and periprocedural IMR changes with PMI. Moreover, our study clarified the relationship between local inflammatory status, microcirculatory dysfunction, and the occurrence of PMI. Lesions with high PCAT attenuation had significantly higher pre‐ and post‐PCI IMR and Δ IMR than those with low PCAT attenuation. In addition, there were weak but statistically significant correlations between PCAT attenuation and IMR‐related variables. Furthermore, this study showed that both pre‐PCI PCAT levels and microcirculatory deterioration during PCI significantly correlated with post‐PCI cTnI levels, representing the extent of myocardial necrosis due to PCI. These findings suggest a substantial association between local vessel inflammation, preexisting microcirculatory dysfunction, and likelihood of exacerbating microcirculatory dysfunction due to PCI, ultimately contributing to PMI occurrence. The detailed mechanisms remain speculative; however, we believe that local vessel inflammation may indicate increased plaque vulnerability and instability, accelerating atheromatous embolization, thrombogenicity, and microcirculatory dysfunction. Notably, even after excluding cases with transient or final no‐reflow phenomenon, post‐PCI IMR values remained higher in the PMI group when compared with those in the non‐PMI group. These findings emphasize that, even in cases where there is no angiographically apparent deterioration of coronary flow, PMI can still occur due to microcirculatory dysfunction. This reinforces our hypothesis that microcirculatory dysfunction plays a significant role in the development of PMI, particularly in the context of lesions characterized by high PCAT attenuation.

It is well established that highly vulnerable plaques contain many macrophages that secrete several inflammatory cytokines and substantial amounts of tissue factor. ²⁸ A previous experimental study has demonstrated that the release of active tissue factors alone can lead to the occurrence of the no‐reflow phenomenon. ²⁹ Based on these findings, we speculate that during PCI, the mechanical disruption of highly inflammatory lipid‐rich plaques, particularly those with APCs, may release such active tissue factors into the coronary circulation. Consequently, these circulating active tissue factors and cytokines have the potential to induce microcirculatory spasm and/or dysfunction, contributing to PMI occurrence. Moreover, baseline microcirculatory dysfunction resulting from inflammation may predispose individuals to a reduced tolerance toward further microcirculatory dysfunction induced by the mechanical damage of the epicardial coronary artery by PCI.

In addition, inflamed plaques may be prone to embolization of plaque debris and thrombotic aggregates, further exacerbating microcirculatory dysfunction. The combined effects of inflammation on plaque instability and microcirculatory dysfunction play a significant role in PMI occurrence. Therefore, assessing inflammation through PCAT attenuation on cCTA may provide additive information on the activity of plaques with APCs and microcirculatory dysfunction in the coronary vessel. Further studies are necessary to confirm our findings.

Different Predictive Abilities in Identifying PMI Among Lesion‐, Vessel‐, and Patient‐Level PCAT Attenuation

In the present study, PCAT attenuation at the culprit lesion level demonstrated superior predictive ability in identifying PMI compared with PCAT attenuation at the culprit vessel and patient level. This suggests that lesion‐specific local inflammation may have a greater impact on PMI occurrence than inflammation at the vessel or patient level. The level at which PCAT attenuation should be measured to assess inflammation remains controversial. Recently, several studies have investigated relationships between lesion‐specific PCAT attenuation and culprit lesion plaque vulnerability. Goeller et al ¹⁷ showed that PCAT attenuation was higher around culprit lesions than nonculprit lesions in patients with acute coronary syndromes or in control subjects with stable coronary artery disease. In addition, the authors reported that low attenuation plaque was independently associated with high PCAT attenuation at the culprit lesion level. These findings highlighted the potential importance of a lesion‐specific evaluation of PCAT attenuation to assess local plaque characteristics. Furthermore, by evaluating plaques in all coronary segments, Kwiecinski et al ¹⁹ found that those with 18F‐NaF uptake had higher PCAT attenuation than those without increased tracer activity. This relationship was not observed when only the proximal RCA was assessed for PCAT. These findings suggest that a lesion‐specific assessment of PCAT may provide greater insight into atherosclerosis activity than RCA measurements or evaluation of the proximal segments of the major arteries alone. Considering that vulnerability and instability in local plaques contribute to PMI occurrence, it appears appropriate to evaluate PCAT attenuation at the culprit lesion level for predicting the occurrence of PMI, as shown in the present study.

Limitations

First, we acknowledge that the small sample size of the present study may increase the potential for selection bias. Additionally, our study focused on a specific, high‐risk patient population with a retrospective design that required recent prior CT scans. Furthermore, the observed rate of PMI in our research may appear relatively high. It is also important to emphasize that in this single‐center study, there remains a possibility of bias due to unmeasured confounding. Although we have made diligent efforts to account for known confounders and have used rigorous statistical methods, the possibility of unmeasured confounding cannot be completely eliminated in any observational study. Moreover, the observational nature of the data limits the ability to draw definitive conclusions. Therefore, caution is advised when generalizing the clinical implications of our findings. It is also important to highlight that our results should undergo confirmation through external validation. A large‐scale prospective study is warranted to validate the usefulness of PCAT attenuation in treatment decision making. Second, we did not directly measure coronary inflammation; however, recent studies have shown an association between PCAT attenuation and biopsy‐proven vascular inflammation in patients undergoing cardiac surgery. ¹⁰ This supports the potential of PCAT attenuation as a surrogate marker for coronary inflammation. Third, it should be noted that various CT tube voltages influenced PCAT attenuation. However, even when applying different optimal cutoff values of PCAT_Lesion according to tube voltage, the incidence of PMI was high in lesions with high PCAT attenuation, regardless of the tube voltage used (70‐ or 100‐ to 120‐kVp) (Figure 4A). Fourth, we excluded severely calcified lesions in the present study. There is a lack of data on lesion‐specific PCAT attenuation in the presence of severe calcification at the exact lesion site. In addition, the spatial resolution of CT may limit PCAT assessment in areas adjacent to heavy coronary calcification, affecting the evaluation of luminal stenosis and APCs. Furthermore, using rotational atherectomy in 9 of 16 severely calcified lesions during the index PCI could introduce debris and potentially influence microcirculatory dysfunction, necessitating differentiation between PMI caused by rotational atherectomy and PMI from other factors. Fifth, the PCAT attenuation at the culprit lesion level was the best predictor in this retrospective study. However, the limitations of measuring lesion‐specific PCAT attenuation when the exact lesion or location undergoing PCI is uncertain. Therefore, as a noninvasive risk marker, PCAT at the vessel or RCA level might be a more feasible risk marker evaluated before PCI. Sixth, in regard to the selection of confounding variables, we acknowledge that it is a complex process with inherent limitations. In our study, we carefully considered variables such as age, hypertension, diabetes, smoking, plaque measurements such as lesion length, and other pertinent factors consistently associated with PMI in prior studies. ³⁰ , ³¹ In addition, when constructing the multivariable model, we aimed to include clinically significant variables and potential confounders based on established literature and expert consensus. Despite accounting for these clinically relevant variables, we found that PCAT attenuation demonstrated incremental predictive value for PMI beyond their influence. Conversely, it is worth noting that our models included 11 predictors, which may seem extensive, particularly given the relatively small sample size and the number of events. Therefore, we also presented the results of the multivariable logistic regression analysis with a reduced set of covariates. Even when restricting the predictive model to the most critical variables, PCAT continued to demonstrate incremental predictive value (as shown in Data S1). Seventh, given the relatively small sample size and constraints on internal cross‐validation, there is a possibility of overfitting in the AUC results. Finally, IMR‐related variables were measured using an angio‐based rather than a wire‐based approach. Wire‐based IMR evaluation is the current gold standard; a previous validation study has demonstrated that IMR derived solely from a single angiographic view can be a feasible computational alternative to pressure wire‐based IMR, with good diagnostic accuracy in assessing microcirculatory dysfunction. ¹⁴

CONCLUSIONS

The noninvasively determined PCAT attenuation at the culprit lesion level in patients with CCS was independently associated with PMI occurrence. Adding PCAT attenuation at the culprit lesion level to cCTA‐derived APCs may provide incremental benefit for identifying PMI. Our results highlight the importance of microcirculatory dysfunction in PMI development, particularly in the presence of lesions characterized by high PCAT attenuation.

Sources of Funding

None.

Disclosures

None.

Supporting information

Data S1

Table S1

Figures S1–S3