Integrated Coronary CT Angiography Assessment of Plaque Vulnerability and Clinical Outcomes

Abstract

Background:

Invasive intracoronary imaging represents the gold standard for identifying vulnerable coronary plaques, but it is not suitable for widespread clinical use. Coronary computed tomography angiography (CCTA) may offer a non-invasive alternative.

Objectives:

To integrate CCTA-derived plaque morphology, pericoronary inflammation, and plaque burden into a unified Morphology–Inflammation–Burden (MIB) score, and to evaluate its association with plaque vulnerability and clinical outcomes.

Methods:

Patients undergoing CCTA followed by optical coherence tomography (OCT) and intravascular ultrasound (IVUS) were followed for a median of 31 months. High-risk plaque (HRP), pericoronary adipose tissue attenuation (PCATa), and total plaque burden (TPB) were quantified and compared with invasive imaging. A vulnerable lesion was defined as ≥2 OCT vulnerability features.

Results:

A total of 438 patients (67 years) and 1,038 plaques were included; 45.4% presented with non–ST-segment elevation acute coronary syndrome (NSTE-ACS). HRP, elevated PCATa, and high TPB were independently associated with OCT-defined vulnerability (all p < 0.05). TPB correlated with IVUS percent atheroma volume (Pearson’s r=0.69; p<0.001). The MIB score demonstrated a stepwise increase in vulnerability, exceeding a predicted risk of 90% in the highest category. Vulnerable patients, defined by the presence of ≥1 untreated lesion with a high MIB score, had a significantly higher rate of cardiac death, ACS, or revascularization (15.3% vs. 4.4%, p<0.001).

Conclusions:

A CCTA-derived MIB score correlates with plaque vulnerability by intracoronary imaging and identifies patients at increased risk of adverse events. These findings support the value of CCTA for non-invasive risk stratification in clinical practice.

Unstructured Abstract:

An integrated CCTA framework incorporating plaque morphology, vascular inflammation, and total atherosclerotic burden was combined into the Morphology–Inflammation–Burden (MIB) score. This score accurately identified OCT-defined coronary vulnerability and was associated with adverse events over 31 months. These findings support CCTA as a practical non-invasive tool for detecting vulnerable patients and improving cardiovascular risk stratification.

Central Illustration. Integration of CCTA Metrics for Vulnerability Assessment and Clinical Implications

HRP, PCATa, and TPB were predictors of vulnerable lesions assessed with OCT (Panel A). When integrated into a risk score (MIB score), these metrics were associated with a higher prevalence of vulnerability features across categories (Panel B). Vulnerable patients (≥1 untreated lesion with a high MIB score) had a higher risk of adverse cardiovascular events during follow-up (Panel C).

CCTA, coronary computed tomography angiography; HRP, high-risk plaque; MIB, Morphology–Inflammation–Burden; OCT, optical coherence tomography; OR, odds ratio; PCATa, pericoronary adipose tissue attenuation; TCFA, thin-cap fibroatheroma; TPB, total plaque burden.

INTRODUCTION

Acute coronary syndromes (ACS) usually result from the rupture or erosion of a coronary plaque.¹ Early identification of lesions prone to disruption is essential to prevent adverse events.² Although optical coherence tomography (OCT) and intravascular ultrasound (IVUS) represent the reference standard for plaque characterization and plaque burden quantification, their invasive nature limits routine use.³

Coronary computed tomography angiography (CCTA) enables non-invasive assessment of features related to plaque vulnerability, including high-risk plaque (HRP) features, pericoronary adipose tissue attenuation (PCATa), and total plaque burden (TPB).^4–6 However, these metrics reflect overlapping biological domains of atherosclerotic disease,^7–9 and evaluating them individually may overestimate their effect.

To address this gap, we developed an integrated CCTA-based framework, the Morphology–Inflammation–Burden (MIB) score. By combining HRP, PCATa, and TPB, this score provides a non-invasive profile of vulnerability.

In this study, we aimed to: i) correlate HRP, PCATa, and TPB with features of plaque vulnerability using invasive imaging as the reference standard; ii) evaluate whether integrating these metrics into the MIB score improves the identification of vulnerable lesions; and iii) examine the association between the MIB score and clinical outcomes.

METHODS

Study Population and Design

This retrospective study used data from the Massachusetts General Hospital and Tsuchiura Kyodo General Hospital Coronary Imaging Collaboration (NCT04523194). Clinical and imaging data were collected at Tsuchiura Kyodo General Hospital (Ibaraki, Japan) and transferred to Massachusetts General Hospital (Boston, USA) for centralized analysis. The study was approved or deemed “Not Human Subjects Research” by local ethics committees and conducted in accordance with the Declaration of Helsinki.

Between January 2011 and July 2020, consecutive patients who underwent CCTA before percutaneous coronary intervention (PCI), with subsequent OCT, were considered eligible. Patients with ST-segment elevation myocardial infarction (STEMI), prior PCI in the culprit vessel, history of coronary artery bypass grafting (CABG), vasospastic angina, chronic total occlusion, unclear culprit identification, poor CCTA or OCT quality, or CCTA–OCT interval >180 days were excluded. A detailed study flowchart is provided in Supplemental Figure 1.

Study Procedures

CCTA preceded PCI, while intracoronary imaging during the index procedure was performed with OCT, supplemented by IVUS when available. Invasive imaging was conducted only in the culprit vessel, where both the culprit lesion and additional non-culprit lesions were analyzed. The culprit vessel was identified based on clinical and angiographic criteria. Further details, including culprit vessel distribution and lesion characteristics, are provided in the Supplemental Methods.

CCTA Analysis

CCTA was performed using a 320-slice scanner (Aquilion ONE, Canon Medical Systems Corporation) and analyzed by trained research personnel blinded to clinical data. Discrepancies were resolved by consensus.

Coronary plaques were evaluated for HRP features, including positive remodeling (PR), low-attenuation plaque (LAP), napkin-ring sign (NRS), and spotty calcification (SC). HRP was defined as the presence of ≥2 of these features. PCATa measurements were obtained from the three major coronary arteries, averaged, and applied to all lesions from the same patient. High PCATa was defined using a previously published threshold of −70.1 HU.¹⁰ Volumetric plaque measurements, including TPB, were quantified at the lesion level within vessel segments matched to plaques identified by OCT and IVUS. Full technical definitions and measurement protocols are available in the Supplemental Methods.

Intracoronary Imaging Analysis

OCT was performed using frequency-domain (C7/C8 OCT Intravascular Imaging System, St. Jude Medical) or time-domain systems (M2/M3 Cardiology Imaging System, LightLab Imaging). Thrombus aspiration was allowed when necessary to improve visualization of the underlying plaque. IVUS was performed with a 40-MHz catheter (OptiCross, Boston Scientific). All images were analyzed offline by independent investigators blinded to clinical and CCTA findings; discrepancies were resolved by consensus.

OCT vulnerability features, including thin-cap fibroatheroma (TCFA), lipid-rich plaque (LRP), macrophages, and microvessels, were defined according to established criteria.¹¹ A “vulnerable lesion” was defined as the presence of at least two OCT features within the same plaque. IVUS analysis consisted of calculating percent atheroma volume (PAV) within coronary segments corresponding to the same plaques identified by CCTA and OCT. Detailed acquisition protocols are provided in the Supplemental Methods.

Clinical Endpoints and Follow-Up

Patients were followed after discharge through telephone interviews or outpatient visits. Recorded clinical events included all-cause mortality, cardiac death, ACS, ischemia-driven revascularization, and stroke or transient ischemic attack (TIA). Deaths were adjudicated as cardiac unless a clear non-cardiac cause was established. The primary endpoint was a composite of cardiac death, ACS, or ischemia-driven myocardial revascularization.

For prognostic analyses, patients were classified as “vulnerable patients” if they had ≥1 residual untreated lesion with a high MIB score after PCI. Lesions treated during the index procedure were not considered for patient risk classification. All clinical outcomes were assessed at the patient level. Detailed definitions of the endpoints are available in the Supplemental Methods.

Statistical Analysis

Normality of continuous variables was assessed using histograms and the Shapiro–Wilk test. Normally distributed continuous variables are reported as mean ± SD and non-normally distributed variables as median (Q1-Q3). Comparisons were performed using Student’s t-test or the Mann–Whitney U test, as appropriate. Correlations between continuous variables were assessed using Pearson correlation coefficients. Categorical variables are reported as counts (%) and compared using χ² or Fisher’s exact test, as appropriate. Analyses were structured to (i) examine lesion-level associations between CCTA parameters and invasive vulnerability, (ii) derive an integrated risk score, and (iii) assess its prognostic value.

Because multiple lesions could be present within the same patient and PCATa was defined at the patient level, all lesion-level analyses accounted for within-patient clustering using generalized linear models (GLMs) with generalized estimating equations (GEE). A logit link was applied for binary outcomes and an identity link for continuous outcomes. Wald’s chi-square test was used for post hoc comparisons, and multiple testing adjustment was performed using Tukey’s method. Optimal TPB thresholds to detect vulnerability were determined using the Youden index on ROC analysis.

Univariable and multivariable logistic regression models with GEE were used to assess the association of HRP, elevated PCATa, and elevated TPB with OCT-defined plaque vulnerability. A weighted MIB score was derived from the multivariable GEE model coefficients. The association between vulnerable patient status and the primary endpoint was assessed using Cox proportional hazards regression. Cumulative incidence functions were compared using Gray’s test to account for the competing risk of non-cardiac death, and Fine–Gray competing-risk regression models were performed as sensitivity analyses. A pre-specified subgroup analysis by baseline statin therapy was conducted to evaluate potential effect modification in Cox regression analyses.

A two-sided p-value < 0.05 was considered statistically significant.

All analyses were performed using R version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Full statistical details are provided in the Supplemental Methods.

RESULTS

Baseline Clinical Characteristics

Among 517 screened patients, 438 (84.7%) were included in the analysis (Supplemental Figure 1). Of these, 306 (69.9%) underwent IVUS assessment. The median age was 67 years (59–74); 350 patients (79.9%) were male, and 199 (45.4%) had non–ST-segment elevation acute coronary syndrome (NSTE-ACS). The overall median interval between CCTA and invasive imaging was 8 days (0-38). When stratified by clinical presentation, the interval was 0 days (0-2) in patients with NSTE-ACS and 35 days (18.5-60.75) in those with stable angina pectoris (SAP). Detailed clinical and laboratory characteristics are presented in Table 1.

Table 1.

Baseline and clinical characteristics

	Overall (n = 438)
Age, y	67 (59-74)
Male, (%)	350 (79.9)
Clinical Presentation
SAP, (%)	239 (54.3)
UAP, (%)	47 (10.7)
NSTEMI, (%)	152 (34.7)
Creatinine, mg/dL	0.8 (0.7 - 0.9)
Risk Factors
Hypertension, (%)	261 (59.6)
Hyperlipidemia, (%)	256 (58.4)
Diabetes mellitus, (%)	177 (40.4)
Chronic kidney disease, (%)	130 (29.7)
Current smoker, (%)	130 (29.7)
Previous PCI, (%)	74 (16.9)
History of MI, (%)	56 (12.8)
Laboratory Data
LDL–C, mg/dL	104 (83 - 129)
HDL–C, mg/dL	46 (40 - 56)
Triglycerides, mg/dL	122 (84.2 - 181.8)
Glucose, mg/dL	118 (101 - 147)
eGFR, mL/min/1.73 m2	72.5 (62.3 – 83)
HbA1c, %	6 (5.6 - 6.7)
Medication Before PCI
Aspirin, (%)	180 (41.1)
ARB/ACEi, (%)	278 (63.4)
Beta-blocker, (%)	157 (35.8)
Statin, (%)	224 (51.1)

Data are given as n (%) and median (Q1-Q3).

ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; eGFR, estimated glomerular filtration rate; HbA1c, glycosylated hemoglobin; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; PCI, percutaneous coronary intervention; SAP, stable angina pectoris; UAP, unstable angina pectoris.

High-Risk Plaque Features and Pericoronary Inflammation

A total of 1,038 coronary plaques (441 culprit and 597 non-culprit) were analyzed in the culprit vessels. Vulnerability features assessed by OCT were significantly more frequent in plaques with HRP compared with those without: TCFA (39.9% vs. 16.6%), LRP (93.4% vs. 70.0%), macrophages (77.6% vs. 58.5%), and microvessels (55.3% vs. 43.0%) (all p<0.001; Figure 1A). Lesions in patients with elevated PCATa demonstrated a similar pattern. These lesions exhibited a higher prevalence of TCFA (32.4% vs. 24.3%), LRP (86.2% vs. 76.5%), macrophages (72.9% vs. 62.7%), and microvessels (56.5% vs. 41.8%) compared with lesions in patients with lower PCATa (all p≤0.05; Figure 1B).

Figure 1. OCT features according to different CCTA characteristics.

Lesions with HRP (A), those occurring in patients with elevated PCATa (B), or with high TPB (C) showed a higher prevalence of TCFA, LRP, macrophages, and microvessels, indicating an association between CCTA findings and invasive markers of plaque vulnerability.

CCTA, coronary computed tomography angiography; HRP, high-risk plaque; LRP, lipid-rich plaque; OCT, optical coherence tomography; PCATa, pericoronary adipose tissue attenuation; TCFA, thin-cap fibroatheroma; TPB, total plaque burden.

Atherosclerotic Plaque Burden

To assess plaque burden by invasive and non-invasive modalities, IVUS-derived PAV was compared with CCTA-derived TPB. These two measurements showed a strong correlation (Pearson’s r=0.69; p<0.001) across 515 coronary segments corresponding to individual plaques (Supplemental Figure 2). A TPB ≥60% emerged as the optimal threshold for identifying vulnerable plaques (Supplemental Results). Plaques located in coronary segments with high TPB demonstrated a significantly higher prevalence of TCFA (31.8% vs. 22.2%), LRP (86.1% vs. 74.5%), macrophages (72.2% vs. 61.4%), and microvessels (53.4% vs. 42.2%) compared with those in segments with lower TPB (all p≤0.05; Figure 1C).

Reproducibility analyses for CCTA measurements are reported in Supplemental Tables 1 and 2.

Integration of Morphology, Inflammation, and Plaque Burden: The MIB Score

Among 1,038 coronary plaques, CCTA-derived parameters were evaluated to identify vulnerable lesions, defined as the presence of ≥2 independent OCT features of vulnerability (Supplemental Table 3). Vulnerable lesions had a higher prevalence of HRP (57.7% vs. 24.6%), were more common in patients with elevated PCATa (55.0% vs. 34.7%), and were more frequently observed in coronary segments with high TPB (66.1% vs. 47.4%) (all p<0.001; Figure 2). Multivariable analysis confirmed HRP (OR 3.16, 95% CI 2.24 - 4.47, p<0.001), PCATa (OR 1.85, 95% CI 1.29 – 2.65, p<0.001), and TPB (OR 1.43, 95% CI 1.11 – 1.99, p=0.032) as independent predictors of vulnerable lesions (Table 2, Central Illustration). The prevalence of vulnerable plaques increased stepwise with the number of CCTA features, with each additional feature associated with greater vulnerability (p for trend<0.001; Supplemental Figure 3).

Figure 2. Association between CCTA parameters and vulnerable lesions.

Vulnerable lesions (≥2 OCT features of vulnerability) showed a higher prevalence of HRP, high PCATa, and high TPB compared with low-vulnerability lesions (all p<0.001).

CCTA, coronary computed tomography angiography; HRP, high-risk plaque; OCT, optical coherence tomography; PCATa, pericoronary adipose tissue attenuation; TPB, total plaque burden.

Table 2.

Logistic regressions for CCTA-derived predictors of vulnerable lesions

Predictor	OR	Univariable		aOR	Multivariable
		95% CI	p value		95% CI	p value
High risk plaque (HRP)	4.08	2.89 - 5.76	< 0.001	3.16	2.24 – 4.47	< 0.001
High PCATa	2.34	1.60 - 3.41	< 0.001	1.85	1.29 – 2.65	< 0.001
High TPB	2.17	1.58 - 2.98	< 0.001	1.43	1.11 – 1.99	0.032
Age	0.99	0.77 – 1.57	0.456
Male	1.66	1.08 – 2.54	0.020	1.40	0.92 – 2.13	0.119
NSTE-ACS	1.1	0.77 – 1.57	0.601
Hypertension	1.15	0.78 – 1.7	0.464
Diabetes mellitus	1.25	0.86 – 1.8	0.240
Current smoker	1.07	0.73 – 1.56	0.735
LDL–C, mg/dL	1	0.99 – 1.01	0.146
Previous PCI	0.98	0.52 – 1.84	0.954
History of MI	0.9	0.49 – 1.68	0.754
Previous statin	0.74	0.52 – 1.06	0.090	0.86	0.61 – 1.2	0.363

Variables with p<0.10 in the univariable analysis were included in the multivariable model.

aOR, adjusted odds ratio; HRP, high-risk plaque; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; NSTE-ACS, non–ST-segment elevation acute coronary syndrome; OR, odds ratio; PCATa, pericoronary adipose tissue attenuation; PCI, percutaneous coronary intervention; TPB, total plaque burden.

Based on regression coefficients, we developed the MIB score, a composite index ranging from 0 to 6. The predicted probability of vulnerability increased stepwise with higher MIB values, reaching 91.4% when HRP, elevated PCATa, and high TPB were all present (Figure 3). The MIB score demonstrated good discrimination and acceptable calibration in internal cross-validation (Supplemental Results; Supplemental Table 4). Score distribution showed separation into three categories: score 0, intermediate values (1–3), and high values (4–6), with the predicted probability of vulnerability approaching a plateau above 80% in the highest category. The MIB score was therefore combined into three clinical risk strata: low (0), intermediate (1–3), and high (4–6). A stepwise gradient was observed across MIB categories, with increasing prevalence of TCFA (11.2%, 22.2%, 39.8%), lipid-rich plaque (66.5%, 72.5%, 94.6%), macrophages (55.3%, 61.0%, 78.5%), and microvessels (33.5%, 47.1%, 56.6%) (all p<0.001; Figure 4; Central Illustration). Pairwise comparisons are detailed in Supplemental Table 5.

Figure 3. Predicted probabilities of vulnerable lesions across MIB score categories.

The predicted risk of an OCT-defined vulnerable lesion increased stepwise with higher MIB scores, reaching >90% when all three features (HRP, PCATa, and TPB) were present.

HRP, high-risk plaque; MIB, Morphology–Inflammation–Burden; OCT, optical coherence tomography; PCATa, pericoronary adipose tissue attenuation; TPB, total plaque burden.

Figure 4. Prevalence of OCT features across MIB score categories.

The prevalence of TCFA, LRP, macrophages, and microvessels increased progressively with higher MIB score categories.

LRP, lipid-rich plaque; MIB, Morphology–Inflammation–Burden; OCT, optical coherence tomography; TCFA, thin-cap fibroatheroma.

When compared with established anatomic classifications, the MIB score demonstrated improved discrimination for detection of vulnerable lesions, outperforming CAD-RADS (AUC 0.71 vs. 0.61; ΔAUC +0.10; p<0.001; NRI 0.46; IDI 0.075) and coronary artery calcium score (CACS) (AUC 0.71 vs. 0.59; ΔAUC +0.12; p<0.001; NRI 0.71; IDI 0.093).

Prognostic Impact of the MIB Score

Patients with ≥1 residual untreated lesion with a high MIB score were considered “vulnerable patients”. Clinical follow-up was available in 437 patients (99.8%). Over a median of 31 months, cardiac death occurred in 2 vulnerable patients (1.4%) compared with 2 non-vulnerable patients (0.7%) (p=0.601), while ACS occurred in 2 (1.4%) versus 3 patients (1.0%) (p=0.666). Myocardial revascularization was more frequent among vulnerable patients (21 patients, 14.6%) compared with non-vulnerable patients (12 patients, 4.1%, p<0.001). Overall, the primary composite endpoint of cardiac death, ACS, or revascularization occurred in 22 vulnerable patients (15.3%) compared with 13 non-vulnerable patients (4.4%) (p<0.001). The incidence of individual adverse events is shown in Table 3.

Table 3.

Incidence of clinical outcomes according to CCTA-based risk profile.

	Overall (n=437)	Vulnerable Patients (n=144)	Non-Vulnerable Patients (n=293)	p value
Death, n (%)	12 (2.7)	3 (2.1)	9 (3.1)	0.758
Cardiac death, n (%)	4 (0.9)	2 (1.4)	2 (0.7)	0.601
ACS, n (%)	5 (1.1)	2 (1.4)	3 (1.0)	0.666
TVR, n (%)	18 (4.1)	11 (7.6)	7 (2.4)	0.019
TVTLR, n (%)	9 (2)	3 (2.1)	6 (2.0)	1.000
TVNLR, n (%)	9 (2)	8 (5.6)	1 (0.3)	< 0.001
NTVR, n (%)	21 (4.8)	14 (9.7)	7 (2.4)	0.002
Stroke / TIA, n (%)	3 (0.7)	2 (1.4)	1 (0.3)	0.254
Cardiac death, ACS, or myocardial revascularization, n (%)	35 (8)	22 (15.3)	13 (4.4)	< 0.001

Data are presented as n (%).

Vulnerable patients were defined by the presence of ≥1 untreated lesion with a high MIB score.

ACS, acute coronary syndromes; NTVR, non–target vessel revascularization; TIA, transient ischemic attack; TVNLR, target vessel non–target lesion revascularization; TVR, target vessel revascularization; TVTLR, target vessel target lesion revascularization.

Cumulative incidence curves showed a higher occurrence of the primary endpoint among vulnerable patients (p<0.001) (Figure 5, Central Illustration). In multivariable Cox regression analysis, the vulnerable profile remained independently associated with the primary endpoint (adjusted HR, 3.24; 95% CI, 1.62–6.47; p<0.001) (Table 4). Results from Fine–Gray competing risk models were consistent with the Cox analysis (Supplemental Table 6). The MIB score demonstrated superior performance compared with its individual components, with risk increasing progressively with the accumulation of adverse CCTA domains (Supplemental Table 7). Moreover, it provided incremental prognostic value over CAD-RADS and CACS (Supplemental Results). In pre-specified subgroup analyses, patients not receiving statin therapy were more often classified as vulnerable (37.7% vs 28.6%; p=0.042). Among patients not receiving statins, the incidence of the primary endpoint was significantly higher in vulnerable patients (18.8% vs. 3.8%; p<0.001), whereas no significant difference was observed among statin users (10.9% vs. 5.0%; p=0.20; Supplemental Figure 4). Consistently, the vulnerable profile was associated with the primary endpoint among patients not on statins (HR 5.50, 95% CI 2.00–15.1; p=0.001), but not in statin users (HR 1.99, 95% CI 0.72–5.49; p=0.18; p for interaction=0.24).

Figure 5. Cumulative incidence curves of the primary endpoint according to risk profile.

Vulnerable patients (at least one untreated lesion with a high MIB score) had significantly higher event rates than non-vulnerable patients (P<0.001). Between-group differences were assessed using Gray’s test, accounting for the competing risk of non-cardiac death. These findings support a shift from the vulnerable plaque to the concept of a vulnerable patient, emphasizing the value of comprehensive risk stratification.

MIB, Morphology–Inflammation–Burden.

Table 4.

Cox proportional hazards regression analyses for the primary composite endpoint.

Predictor	Univariable			aHR	Multivariable
	HR	95% CI	p value		95% CI	p value
Vulnerable Patients	3.48	1.75 – 6.91	< 0.001	3.24	1.62 - 6.47	< 0.001
Age	0.98	0.95 – 1.01	0.189
Male	0.82	0.37 – 1.81	0.624
NSTE-ACS	2.46	1.22 – 4.95	0.011	1.97	0.95 – 4.12	0.069
Hypertension	0.74	0.37 – 1.48	0.398
Diabetes mellitus	1.3	0.67 – 2.53	0.437
Current smoker	0.78	0.37 – 1.63	0.507
LDL–C, mg/dL	1.01	1 – 1.02	0.059	1.06	0.96 – 1.17	0.223
Previous PCI	1.04	0.43 – 2.51	0.929
History of MI	1.17	0.45 – 3.01	0.746
Previous statin	0.72	0.37 – 1.4	0.335

The primary composite endpoint included cardiac death, ACS, or myocardial revascularization.

aHR, adjusted hazard ratio; HR, hazard ratio; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; NSTE-ACS, non–ST-segment elevation acute coronary syndrome; PCI, percutaneous coronary intervention.

DISCUSSION

This study investigated the role of CCTA in assessing plaque vulnerability and explored its prognostic value for future adverse cardiac events. The main findings were: (i) HRP, PCATa, and TPB were independently associated with OCT-defined plaque vulnerability; (ii) combining these complementary domains into the MIB score enabled a single, integrated measure of plaque vulnerability; and (iii) the MIB score was associated with an increased risk of adverse cardiovascular events.

To the best of our knowledge, this is the first study to integrate plaque morphology, perivascular inflammation, and atherosclerotic burden into a single framework, using OCT and IVUS as reference standards.

CCTA Metrics and Invasive Features of Plaque Vulnerability

In our analysis, HRP lesions were significantly associated with invasive vulnerability. Each HRP component reflects a distinct biological process linked to specific OCT features.^4,12 PR and LAP predominantly correspond to lipid accumulation, whereas SC and NRS are more often associated with cholesterol crystals and microvessels, hallmarks of advanced fibroatheroma.^13,14 To integrate these signals, we adopted a composite HRP definition requiring at least two features within the same lesion. Our findings extend prior evidence by showing that ≥2 HRP characteristics reliably identify a vulnerable plaque phenotype, supporting the use of CCTA for non-invasive plaque characterization.^15,16

PCATa demonstrated an independent association with vulnerability features, underscoring inflammation as a key determinant of plaque instability. Perivascular inflammation contributes to plaque destabilization through cytokine pathways, promoting both subclinical and acute clinical events.^10,17 The correlation between PCATa and 18F-NaF uptake further supports its role as a surrogate marker of active coronary inflammation.¹⁸ These observations support PCATa as a marker of pancoronary inflammatory activity, consistent with prior evidence of diffuse vulnerability in patients with plaque rupture.¹⁹

Finally, plaque burden, traditionally assessed with IVUS, has emerged as a major determinant of both plaque vulnerability and future events.^20,21 In the present cohort, TPB showed a strong correlation with IVUS-derived PAV and was associated with OCT-defined vulnerability. These findings reinforce CCTA as a reliable tool for quantifying both structural and biological disease burden.

Taken together, these findings support CCTA as a multidimensional, non-invasive imaging modality for comprehensive characterization of coronary plaque vulnerability.

MIB Score for Vulnerable Plaque Detection

In our study, HRP, PCATa, and TPB were independent predictors of plaque vulnerability. While previous studies have typically evaluated these markers individually,^5,16,22 emerging evidence indicates substantial biological overlap across these categories. Lesions with high-risk morphology frequently show higher perivascular inflammation, HRP features cluster in segments with greater plaque burden, and elevated PCATa correlates with larger plaque. ^23–25 Thus, these metrics likely represent interconnected dimensions of the same pathological process rather than independent signals. Evaluating them separately may therefore overestimate their relative contribution.

To address this gap, we developed the MIB score integrating HRP, PCATa, and TPB. This scoring system demonstrated incremental predictive value over individual parameters, with the probability of vulnerability exceeding 90% when all three features were present. Moreover, increasing score categories were consistently associated with a higher prevalence of TCFA, LRP, macrophages, and microvessels.

The MIB score also outperformed CACS and CAD-RADS for identifying vulnerable plaques. CACS only quantifies the calcified component, which is a late and typically stabilized stage of atherosclerosis. Thus, it fails to identify non-calcified or inflamed plaques.²⁶ Similarly, CAD-RADS prioritizes stenosis severity and neglects biological instability.²⁷ In contrast, MIB integrates complementary biological domains: HRP reflects structural fragility and predisposition to rupture, PCATa captures the inflammatory activity, and TPB quantifies the overall atherosclerotic burden. Their combination, therefore, provides a comprehensive representation of coronary vulnerability (Table 5). This approach aligns with a recent position statement emphasizing that the convergence of morphology, burden, and inflammation identifies the highest risk.²⁸

Table 5.

Components of the Morphology–Inflammation–Burden (MIB) Framework

Domain	maging Feature	Role in Coronary Risk	Interpretation
Morphology	HRP (≥2 Features: LAP, PR, NRS, SC)	Reflects structural plaque vulnerability	High rupture potential
Inflammation	PCATa > –70.1 HU	Marker of active vascular inflammation	Biologically active disease
Burden	TPB ≥ 60 %	Represents the global atherosclerotic burden	Advanced atherosclerotic disease

The MIB framework integrates three complementary CCTA-derived domains providing a multidimensional assessment of coronary vulnerability.

CCTA, coronary computed tomographic angiography; HRP, high-risk plaque; LAP, low-attenuation plaque; MIB, Morphology–Inflammation–Burden; NRS, napkin-ring sign; PCATa, pericoronary adipose tissue attenuation; PR, positive remodeling; SC, spotty calcification; TPB, total plaque burden.

Clinical Implications of the MIB Score

Whether identifying and treating vulnerable plaques improves outcomes remains uncertain. Although conceptually intriguing, preventive interventions would be relevant only if they demonstrate improved clinical outcomes, particularly a reduction of cardiac death and/or acute myocardial infarction.²⁸

Current interventional studies suggest potential benefit, but their impact on hard outcomes is still limited.^20,21,29 Indeed, pathology and OCT studies have shown that plaque disruption is often clinically silent, with subsequent healing promoting plaque growth.³⁰ Moreover, vulnerable features often reflect overall atherosclerotic burden rather than isolated targets, and plaque phenotype dynamically evolves, with most TCFAs regressing over time.^31,32 Additionally, up to 40% of ACS are caused by plaque erosion, whose precursor cannot be detected by current imaging.³³

Collectively, these observations not only challenge the strategy of targeting vulnerable plaques but also question the definition of vulnerability, which should describe lesions causing acute MI or sudden death. Instead, such events result from a ‘perfect storm’, where plaque disruption coincides with a prothrombotic and inflammatory milieu that promotes the formation of an occlusive thrombus.³⁴

This evolving understanding supports a transition from the concept of a single vulnerable plaque to the broader idea of a vulnerable patient, where risk is determined by overall atherosclerotic burden, metabolic activity, and systemic response.³⁵ Accordingly, an integrated approach capturing morphology, inflammation, and plaque burden may offer a more accurate representation of individual coronary risk.

In our study, the MIB score was associated with meaningful differences in event rates. The composite endpoint occurred in 15.3% of vulnerable patients, compared with 4.4% of patients with lower scores.

The score also outperformed CACS and CAD-RADS, likely because these classifications focus primarily on anatomic stenosis or calcification rather than biological activity. ^36,37

Together, these findings underscore the potential of the MIB score, which shifts the focus from the “vulnerable plaque” to the “vulnerable patient.” This integrative approach bridges the distinct dimensions of plaque biology that collectively determine individual coronary risk, revealing potential opportunities for clinical translation. Indeed, a high MIB score captures residual risk beyond revascularization, highlighting a possible role in refined patient stratification.

In our study, patients not receiving lipid-lowering therapy were more frequently vulnerable patients. Conversely, the prognostic association between the vulnerable patient profile and adverse outcomes was attenuated among individuals on statin treatment. This attenuation likely reflects pharmacological modulation of disease activity, resulting in a more stable coronary phenotype with lower biological activity.³⁸ These exploratory observations suggest that the MIB score may help identify patients with modifiable, treatment-responsive risk.

Intensive lipid-lowering strategies and emerging anti-inflammatory therapies targeting perivascular activity may therefore be particularly relevant in vulnerable patients.^39,40 Finally, the high reproducibility of HRP, PCATa, and TPB, and the compatibility of these measures with existing CCTA software, suggest that the MIB framework may be readily implemented in clinical practice. Future automation may simplify workflow and enable broader use and prospective validation.

LIMITATIONS

The present study has several limitations.

First, its retrospective, single-center design may introduce selection bias, as patients without significant stenosis on CCTA were less likely to undergo invasive imaging. Second, since the precursors of plaque erosion remain unknown, this study focused only on features related to plaque rupture. Third, patients with STEMI were excluded, which limits the applicability of the findings to this clinical presentation. Fourth, the cohort consisted exclusively of Japanese patients, which may affect generalizability to other populations. Fifth, plaque vulnerability may also be influenced by coronary blood flow and vessel physiology, which were not evaluated in this study. Finally, although internal cross-validation was performed, external validation is lacking, and prospective multicenter studies are required to confirm these findings. Accordingly, the MIB score should be interpreted as an integrative imaging framework rather than as a standalone clinical risk score.

CONCLUSIONS

HRP, PCATa, and TPB were strongly associated with OCT-defined features of plaque vulnerability. Integrating these parameters into the MIB score enabled a more comprehensive assessment of vulnerable lesions and was associated with adverse cardiovascular events. These findings support CCTA as a practical non-invasive strategy for vulnerability assessment and risk stratification, helping identify patients who may benefit from targeted preventive therapy.

PERSPECTIVES

Competency in Medical Knowledge

CCTA provides a noninvasive assessment of plaque morphology, perivascular inflammation, and atherosclerotic plaque burden. Integrating these domains into a unified framework helps identify vulnerable plaques and is associated with subsequent clinical events.

Translational Outlook

Incorporating plaque morphology, vascular inflammation, and plaque burden into the MIB score enables identification of patients with residual risk after PCI. Individuals with high-MIB lesions represent vulnerable patients who may benefit from more intensive pharmacologic therapy or closer clinical surveillance.

Abbreviations and acronyms:

ACS

Acute coronary syndromes

CCTA

Coronary computed tomography angiography

HRP

High-risk plaque

IVUS

Intravascular ultrasound

MIB

Morphology–Inflammation–Burden

OCT

Optical coherence tomography

PCATa

Pericoronary adipose tissue attenuation

SAP

Stable Angina Pectoris

TCFA

Thin-cap fibroatheroma

TPB

Total plaque burden