Computed tomography-based assessment of pericoronary adipose tissue in cardiovascular diseases: Diagnostic and prognostic implications

Abstract

Pericoronary adipose tissue (PCAT) plays an important role in the pathogenesis and progression of cardiovascular diseases due to its bidirectional communication with the coronary artery wall. In recent years, PCAT parameters measured using coronary computed tomography have emerged as potential noninvasive imaging biomarkers for quantifying coronary artery inflammation, with significant clinical value in the early detection, disease progression assessment, treatment efficacy evaluation, and prognosis prediction of cardiovascular diseases. Furthermore, new technologies such as PCAT radiomics analysis have broadened its potential applications in evaluating coronary plaque vulnerability, predicting cardiovascular events, and improving risk stratification. This review discusses recent advances in PCAT research, focusing on its role in coronary artery disease risk identification and inflammation monitoring, and aims to offer imaging-based insights to support its future clinical use in cardiovascular disease management.

Key Words: Pericoronary adipose tissue; Radiomics; Coronary computed tomography angiography; Vascular inflammation; Cardiovascular diseases

Core Tip: Pericoronary adipose tissue (PCAT) is a key biomarker for coronary inflammation, assessed via coronary computed tomography. Advances in PCAT attenuation and radiomics improve prediction of cardiovascular events, plaque vulnerability, and risk stratification, offering imaging-based insights for early detection, monitoring, and treatment evaluation in cardiovascular disease management.

INTRODUCTION

Cardiovascular diseases continue to be the leading cause of morbidity and mortality worldwide[1], with research indicating that vascular inflammation plays a key role in their onset and progression. As a result, biomarkers that indicate vascular inflammatory activity may become valuable tools for the diagnosis and risk prediction of cardiovascular diseases. Previously, circulating biomarkers such as C-reactive protein and interleukin-6 were used to evaluate systemic inflammation[2,3], but these indicators are easily influenced by other systemic or local inflammatory diseases. In recent years, changes in perivascular adipose tissue (PVAT) have been explored as indicators of vascular inflammation. PVAT refers to adipose tissue with endocrine and paracrine functions surrounding major vessels such as the abdominal aorta, coronary arteries, and carotid arteries. However, the definitions of measurement regions vary considerably: PVAT around the abdominal aorta is typically assessed using a cylindrical volume extending 5 mm or 10 mm from the aortic wall[4,5], while for the carotid artery, fat tissue within a radial distance equal to the vessel diameter is analyzed[6]. This review focuses on pericoronary adipose tissue (PCAT), a distinct form of PVAT located around the coronary arteries within a radial distance equal to the vessel diameter, which functions as a novel imaging biomarker of local coronary inflammation[7,8]. As a dynamic and metabolically active tissue, PCAT indicates potential inflammatory activity adjacent to the coronary artery walls, playing an important role in the development and progression of coronary artery disease (CAD) by forming a local inflammatory microenvironment owing to its proximity and biochemical communication with the artery wall. When PCAT becomes dysfunctional, it secretes bioactive factors that cause endothelial dysfunction, worsen inflammation, and atherosclerosis. However, pathological changes in the coronary artery wall can affect PCAT via paracrine mechanisms, resulting in abnormal lipid metabolism, reduced lipid accumulation, and changes in PCAT parameters[7]. Coronary computed tomography (CT) imaging allows for the noninvasive detection and quantitative assessment of this biological process. Such imaging-based evidence supports PCAT’s potential as a reliable biomarker for monitoring coronary inflammation, assessing plaque vulnerability, and predicting cardiovascular disease risk. As a result, this review summarizes the use of PCAT as a novel risk biomarker in the early detection and prognosis of cardiovascular diseases, with a focus on its potential in coronary artery inflammation monitoring, plaque vulnerability assessment, treatment efficacy evaluation, and cardiovascular event prediction.

PCAT PARAMETERS

PCAT thickness

PCAT thickness is a simple and quick measurement method that is typically performed using cross-sectional CT images to calculate the maximum fat width around the proximal coronary artery. Studies have shown that the PCAT thickness, as a measurement parameter, is highly reproducible[3]. However, when compared to other PCAT indicators, PCAT thickness provides limited information, and it is difficult to visually distinguish PCAT from epicardial fat thickness on imaging[7]. Furthermore, this method does not adequately reflect changes in PCAT density, limiting its usefulness in clinical research and applications.

PCAT volume

Currently, there are three main methods to measure PCAT volume: (1) The original segmentation method based on PCAT mean attenuation, which determines PCAT volume by counting the number of voxels within the PCAT attenuation range[8]; (2) Manual tracing and summation, which involves manually outlining the PCAT region on axial images perpendicular to the coronary artery centerline and then calculating the volume[9]; and (3) Region-of-interest definition along the coronary artery trajectory, which involves manually defining the analysis area along the coronary artery to get a more accurate PCAT volume measurement[10]. According to studies, PCAT volume offers more benefits than fat attenuation index (FAI) measurements on non-contrast coronary CT images and provides more three-dimensional structural information than PCAT thickness[8,10]. However, distinctions in PCAT volume measurement methods and measurement ranges limit data comparability across studies.

PCAT attenuation

PCAT density is typically assessed using CT by calculating the mean Hounsfield units (HU) value of the epicardial adipose tissue, primarily to determine the overall fat tissue density[11]. This method is easy and straightforward to use; however, because it only provides an average HU value, it may overlook dynamic variations in fat tissue characteristics across anatomical locations. To overcome these limitations, the FAI was developed, which corrects PCAT’s mean CT attenuation by accounting for attenuation gradients, resulting in a more precise quantification of pericoronary inflammation levels[12]. FAI, a noninvasive imaging biomarker for coronary inflammation, has been widely used to detect inflammatory changes in the left anterior descending artery (LAD), left circumflex artery (LCX), right coronary artery (RCA), and lesion-specific regions (Figure 1). Depending on the contrast used, PCAT attenuation can be measured using non-contrast CT or contrast-enhanced coronary computed tomography angiography (CCTA), both of which have a high correlation. However, non-contrast CT typically produces lower PCAT attenuation values, possibly due to capillary leakage of iodine contrast caused by local inflammation, partial volume effects, and beam-hardening artifacts[8,13]. While threshold adjustments can help to correct these discrepancies, non-contrast CT measurements are highly prone to central line errors and lack standardized measurement protocols and recognition thresholds. Further research is required to optimize its clinical applicability.

Figure 1 Pericoronary adipose tissue attenuation measurement in proximal 40 mm segments and lesion-specific regions. A-C: Schematic of pericoronary adipose tissue (PCAT) attenuation measurement in the proximal 40 mm segments of the three coronary arteries (left anterior descending artery, left circumflex artery, right coronary artery); D and E: Schematic of lesion-specific PCAT attenuation measurement at the site of coronary lesions; F: Cross-sectional PCAT measurement. The measurement radial range is defined as the PCAT within a distance equal to the vessel diameter. LAD: Left anterior descending; CX: Circumflex; RCA: Right coronary artery.

FAI measurements are affected by several factors, including imaging technical parameters and individual characteristics. Studies have shown that tube voltage, scanning parameters, and image reconstruction algorithms all have a direct impact on FAI values, with higher tube voltage increasing PCAT attenuation, emphasizing the importance of optimized scanning parameters for improving measurement accuracy in clinical applications[13]. In addition, several technical factors such as reconstruction methods and scanner types have been shown to influence PCAT attenuation. Studies have shown that images reconstructed using iterative reconstruction yield higher attenuation values than those using filtered back projection[14]. Other reconstruction parameters-such as adaptive statistical iterative reconstruction-V percentage, reconstruction kernel, and slice thickness, also significantly affect PCAT measurements[15]. To minimize bias and improve reproducibility, these parameters should be standardized during scanning or appropriately adjusted in cross-study comparisons. While some studies have reported that photon-counting CT provides more stable PCAT values by reducing image noise and beam-hardening artifacts[16], others have found no statistically significant differences among scanner models[14]. Given the limited data on inter-scanner variability, further research is needed to determine the impact of scanner type on the consistency of PCAT assessment.

Furthermore, individual factors such as gender, age, smoking history, and obesity have a significant impact on FAI, with studies indicating that men have higher PCAT attenuation than women, while aging, chronic smoking, and obesity are associated with increased PCAT attenuation, implying a role in promoting pericoronary inflammation[11,17]. Furthermore, the influencing factors for PCAT attenuation differ between coronary arteries. For example, LAD PCAT attenuation is linked to body mass index, smoking duration, alcohol consumption, and hyperlipidemia, whereas RCA PCAT attenuation is more strongly related to hyperlipidemia and statin use[18]. Anatomical differences also influence PCAT attenuation, with studies indicating that distal coronary segments have higher PCAT attenuation than proximal segments, most likely due to differences in the local vascular environment and surrounding tissue characteristics[10]. As a result, when analyzing FAI measurements, these influencing factors should be carefully considered to improve the assessment’s accuracy and clinical applicability.

PCAT radiomics features

Artificial intelligence-based PCAT radiomics has opened up new avenues for imaging evaluation. This method uses high-throughput data characterization algorithms to automatically extract and quantify imaging features, decreasing manual processing time and operator variability while providing more accurate and rapid assessments[19]. Compared to traditional PCAT attenuation analysis, radiomics captures a broader range of texture, morphological, and spatial features and employs machine learning algorithms to create disease prediction models, demonstrating major advantages in cardiovascular risk assessment[20]. Studies have demonstrated that PCAT radiomic features significantly expand the information provided by PCAT attenuation, making them even more effective than attenuation analysis in identifying high-risk plaques and more accurately predicting the advancement of CAD[21,22]. However, its clinical application remains limited due to a lack of standardized imaging processing protocols, complex region-of-interest delineation, and the requirement for extensive imaging analysis expertise[23]. In the future, advancements in automated segmentation technologies and large-scale research studies are expected to improve PCAT radiomics’ role in the early identification and personalized risk prediction of CAD. Figure 2 illustrates the workflow of PCAT radiomics, while Table 1 summarizes the characteristics of each parameter.

Figure 2 Workflow diagram of pericoronary adipose tissue radiomics. LAD: Left anterior descending; CX: Circumflex; RCA: Right coronary artery; CAD: Coronary artery disease; MACE: Major adverse cardiac event; ROC: Receiver operating characteristic; AUC: Area under the curve.

Table 1 Comparison of pericoronary adipose tissue measurement parameters.

Measurement parameter	Definition/method	Advantages	Limitations	Clinical applications
PCAT thickness	Measured as the maximum width of proximal pericoronary fat	(1) Simple, quick, and highly reproducible; and (2) Suitable for preliminary assessment	(1) Provides only 2D information; and (2) Does not reflect PCAT density changes	Primarily used for initial evaluation with limited independent value in clinical decision-making
PCAT volume	(1) Attenuation-based segmentation: Calculates voxel volume within the defined range; (2) Manual delineation: Outlines PCAT on axial images; and (3) ROI-based tracing: Manually defines analysis regions along the coronary trajectory	Provides 3D structural information	(1) Lack of standardized measurement methods; and (2) Poor comparability across studies	Useful for studying PCAT’s 3D structure, but limited clinical application due to lack of standardization
PCAT density	Directly measures the mean HU value of PCAT	(1) Simple method, applicable to routine CT scans; and (2) Easily standardized with high reproducibility	(1) Ignores dynamic changes in PCAT; and (2) Influenced by CT imaging parameters	A potential biomarker for inflammation and atherosclerosis, useful for screening high-risk cardiovascular patients
PCAT attenuation/FAI	Quantifies pericoronary inflammation by correcting the mean CT attenuation of PCAT	(1) Precisely quantifies inflammation; and (2) Highly automated with AI integration	(1) Requires specialized software; and (2) Inconsistent across devices, needing standardization	A noninvasive inflammatory biomarker widely used for coronary inflammation detection
PCAT radiomic features	Leverages AI and deep learning to analyze PCAT texture, morphology, and spatial distribution	(1) Provides more detailed lesion characteristics; and (2) Minimizes errors and improves consistency	(1) Requires standardized imaging protocols and clinical validation; (2) Complex ROI delineation; and (3) High computational cost requiring specialized software	Holds great potential in precision medicine for individualized cardiovascular risk prediction and treatment optimization

PCAT: Pericoronary adipose tissue; FAI: Fat attenuation index; HU: Hounsfield unit; ROI: Region of interest; AI: Artificial intelligence; CT: Computed tomography; 3D: Three dimensional; 2D: Two dimensional.

CLINICAL APPLICATIONS

PCAT in CAD

Assessing coronary plaque progression: CCTA-based PCAT parameters provide a noninvasive method for quantifying coronary artery inflammation and plaque vulnerability. Research conducted by Goeller et al[24] found that the FAI was considerably higher around culprit lesions. FAI ≥ -68.2 HU was proposed as a potential threshold for distinguishing culprits from non-culprit lesions. Furthermore, combining high-risk plaque characteristics with PCAT attenuation improved the ability to identify vulnerable plaques, which is consistent with the findings of Yuvaraj et al[25]. Jiang et al[26] found that combining PCAT radiomic features with clinical indicators in CT scan images accurately diagnoses non-calcified vulnerable plaques [area under the curve (AUC) = 0.752, sensitivity = 75.0%, specificity = 77.8%]. Plaque progression is an intermediate step between subclinical atherosclerosis and coronary events. Studies have looked into the relationship between coronary FAI and plaque progression[27,28], and they found that an increase in non-calcified plaque burden correlates with higher FAI levels. Specifically, RCA FAI ≥ -75 HU proximal is an independent predictor of higher non-calcified plaque burden [odds ratio (OR) = 3.07, P < 0.008]. This finding was confirmed by Suzuki et al[29] and Fujimoto et al[30], who found that patients with higher non-calcified plaque burdens had significantly higher PCAT attenuation values, implying that larger non-calcified plaques are more vulnerable to vascular inflammation. Dai et al[31] conducted follow-up tests on patients given statin therapy and found a significant reduction in FAI levels around non-calcified and mixed plaques post-treatment (P < 0.001), but there was no significant change in calcified plaques, indicating that the medication primarily affects inflammation around vulnerable plaques. Cheng et al[32] also found that both per-lesion and RCA PCAT attenuation were considerably reduced after statin therapy compared to the untreated group, with statin intensity independently associated with changes in lesion PCAT attenuation after accounting for risk factors. These findings suggest that PCAT attenuation could be used as an imaging biomarker to monitor coronary artery response to drug therapy, though more large-scale studies are needed to confirm these results.

Coronary perivascular inflammation promotes the rupture of high-risk plaques, resulting in acute coronary syndromes. Nakajima et al[33] studied 198 patients with plaque rupture and found considerably greater PCAT attenuation in culprit plaques and vessels compared to plaque erosion cases (-67.9 ± 5.7 HU vs -69.9 ± 6.8 HU, P < 0.030), which is consistent with the findings of Araki et al[34]. In addition, Chen et al[35] discovered that PCAT attenuation is an independent predictor of coronary artery plaque progression in 500 patients with suspected or known CAD (OR = 1.037, P = 0.007). The PCAT radiomic features had a higher predictive power than the traditional quantitative plaque characteristics (training set AUC = 0.814 vs 0.615, P < 0.001; validation set AUC = 0.736 vs 0.594, P = 0.007). Another multicenter study also confirmed this view, with PCAT radiomic models achieving AUC values of 0.85 (training set), 0.84 (internal validation), and 0.81 (external validation) for predicting plaque progression[36]. Furthermore, recent research indicates that PCAT attenuation around plaques may be more useful in assessing plaque vulnerability than proximal vessel PCAT attenuation[37]. This could establish PCAT attenuation as a standard biomarker for determining the severity of coronary atherosclerosis, but more large-scale studies are required for validation.

Identification of the clinical subtypes of CAD: PCAT attenuation is a quantitative marker of coronary inflammation that has been used to differentiate between clinical subtypes of CAD. Lin et al[38] found that RCA FAI was significantly higher in myocardial infarction patients than in those with stable CAD and non-CAD individuals. Qi et al[39] discovered that PCAT volume and attenuation gradually increased in normal controls, stable angina pectoris (SAP) patients, and acute myocardial infarction (AMI) patients. AMI patients had significantly higher lesion-related vessel PCAT attenuation than those with SAP (P < 0.05). Similarly, Jing et al[40] found that FAI values in acute coronary syndrome patients were considerably higher than in chronic coronary syndrome and non-CAD patients (P < 0.05), suggesting that PCAT attenuation reflects coronary inflammation levels. Their study also found that combining PCAT radiomic features with FAI and clinical parameters resulted in the best diagnostic performance for distinguishing acute coronary syndrome, chronic coronary syndrome, and non-CAD patients (training set AUC = 0.87, validation set AUC = 0.74).

With the growing importance of PCAT radiomics as a noninvasive tool for detecting high-risk CAD identification, Zhan et al[41] demonstrated that CCTA-based PCAT radiomics could effectively distinguish unstable angina (UA) from SAP. Their comprehensive model outperformed FAI models (AUC = 0.68 and 0.51), clinical feature models (AUC = 0.84 and 0.67), and radiomics models (AUC = 0.85 and 0.73), with an overall AUC of 0.87 in the training and 0.74 in the validation sets. Another study found that PCAT radiomic features could accurately distinguish non-ST-segment elevation myocardial infarction from UA. When the radiomic features of the three coronary arteries were combined, the model showed the best discriminatory ability (AUC = 0.889, sensitivity = 81%, specificity = 81%)[42]. These findings highlight the potential of PCAT attenuation and radiomic features as noninvasive tools for early CAD detection, risk stratification, and accurate therapy decision-making in clinical settings.

Prediction of coronary vascular dysfunction: The gold standard for evaluating coronary artery function is fractional flow reserve (FFR), with a value of ≤ 0.80 indicating coronary ischemia. However, recent research has shown that PCAT attenuation can also be used as a noninvasive biomarker to identify coronary hemodynamic abnormalities, potentially providing an alternative method for detecting ischemic lesions. Wang et al[43] found that higher PCAT attenuation and lower PCAT volume were significantly linked to severe coronary stenosis (P < 0.05). Similarly, Duncker et al[44] discovered that patients with myocardial ischemia had significantly higher RCA PCAT attenuation than non-ischemic individuals (-75.1 ± 10.8 HU vs -81.1 ± 10.6 HU, P = 0.011), indicating that increased RCA PCAT attenuation could be a major predictor of myocardial ischemia (OR = 1.06, P = 0.014). In a study of 167 SAP patients, Yu et al[45] found that pericoronary FAI was a strong predictor of lesion-specific ischemia (OR = 1.028, P = 0.01). Furthermore, a model that included FAI, diameter stenosis, and total plaque volume performed similarly to CT-derived FFR (CT-FFR) in predicting ischemia-inducing coronary stenosis (AUC = 0.821 vs 0.850, P = 0.426), which is consistent with the findings of Ma et al[46]. Another study revealed that integrating FAI ≥ -71.9 HU into plaque assessment models greatly enhanced ischemia detection accuracy, achieving diagnostic performance similar to that of standalone CT-FFR (AUC = 0.772 vs 0.762, P = 0.771)[47]. More recently, combining PCAT radiomics with CT-FFR models has shown significant value in distinguishing flow-limiting from non-flow-limiting lesions, with diagnostic efficacy highly superior to CT-FFR or PCAT radiomics alone (AUC = 0.900 vs 0.803 vs 0.776, P < 0.05)[48]. Furthermore, new information suggests a positive correlation between PCAT parameters and hyperemic microvascular resistance (r = 0.5, P = 0.37), lending support to the hypothesis that upstream coronary inflammation influences downstream microvascular function[49]. This finding sheds new light on PCAT’s potential diagnostic role in patients with microvascular dysfunction, opening up new avenues for noninvasive coronary microvascular disease evaluation.

Prognostic value of PCAT in CAD: Pericoronary FAI has the potential to determine patients with elevated coronary artery inflammation and hemodynamic disturbances, and it could be a useful imaging biomarker for predicting adverse cardiac events. Wen et al[50] studied 1313 patients with acute chest pain over three years, and 142 (10.81%) of them had major adverse cardiac events (MACE), such as UA, coronary artery revascularization, nonfatal myocardial infarction, and all-cause mortality. Their study discovered that RCA PCAT attenuation was an independent predictor of MACE after adjusting for clinical risk factors [hazard ratio (HR) = 1.033, P = 0.006], which was also confirmed by Nishihara et al[51]. Furthermore, PCAT radiomic features had some predictive value for MACE (AUC = 0.703, sensitivity = 0.682, and specificity = 0.644)[52]. According to recent research, lesion-specific PCAT attenuation is a better predictor of MACE than proximal coronary PCAT attenuation. Chen et al[53] found that lesion-specific PCAT attenuation significantly improved the prediction of MACE risk in CAD patients, outperforming traditional RCA and LAD proximal 40 mm PCAT attenuation. In a study of 608 CAD patients, PCAT radiomic analysis showed that the lesion-specific radiomics score (Rad-score) had a strong connection with MACE (HR = 2.528, P < 0.01), and a combined model of the Rad-score and clinical features (C-index = 0.718) outperformed both the clinical-only (C-index = 0.639) and radiomics-only (C-index = 0.653) models (P < 0.05)[54]. Huang et al[55] confirmed this by demonstrating that a model combining lesion-specific PCAT radiomic features and clinical characteristics performed exceedingly well in predicting the 1-year MACE risk in patients with suspected or confirmed CAD, with AUCs of 0.970, 0.957, and 0.960 in the training and two external validation cohorts, respectively. However, another study found that while there was a substantial positive association between lesion-specific and proximal RCA PCAT attenuation, only RCA PCAT attenuation was an independent risk factor for MACE in patients with non-obstructive CAD[56]. This disparity could be attributed to the presence of multiple plaques, as lesion-specific PCAT attenuation can sometimes misrepresent the characteristics of the primary culprit plaque. As a result, in clinical practice, the importance of proximal RCA PCAT attenuation as a biomarker for general coronary inflammation in predicting adverse cardiac events should not be ignored.

Despite several studies indicating that PCAT attenuation and radiomics may predict MACE, their long-term prognostic value remains debatable. Chatterjee et al[57] performed a 5-year follow-up study on 381 patients with suspected CAD and discovered that PCAT attenuation did not efficiently predict long-term MACE. A separate study of 483 patients with a median follow-up of 9.5 years also found that neither PCAT attenuation at the three coronary artery levels or lesion-specific PCAT attenuation could predict long-term MACE[58]. This discrepancy may be due to dynamic changes in PCAT attenuation, with increases occurring shortly before cardiovascular events, making it challenging for CCTA to determine the true culprit lesions during long-term follow-up[58]. As a result, when using PCAT attenuation and radiomics as predictive tools, it is important to consider both short- and long-term prognostic effectiveness, and these should be combined with other clinical indicators and imaging features to improve MACE prediction accuracy.

Prognostic value in predicting post-percutaneous coronary intervention complications: Furthermore, PCAT parameters can detect abnormal coronary changes following the coronary intervention, especially in conditions like stent failure, periprocedural myocardial injury (PMI), and chronic total occlusion (CTO). Nogic et al[59] examined 151 patients after coronary stent implantation and discovered that 17.2% (26 patients) had stent failure (in-stent restenosis > 50% or in-stent thrombosis). The PCAT attenuation in these patients was considerably higher (-79.0 ± 12.6 HU vs -85.9 ± 10.3 HU, P = 0.035), indicating that it may be an independent predictor of stent failure (OR = 1.06, P = 0.035). Another study also found that PCAT radiomics features could efficiently detect in-stent restenosis after percutaneous coronary intervention (PCI), with an AUC of 0.82, a sensitivity of 76.9%, and a specificity of 70.8%[60], implying that vascular inflammation may be a key mechanism in-stent restenosis. Furthermore, Yamamoto et al[61] discovered that in patients with PMI after PCI, both the culprit lesions and RCA PCAT attenuation were significantly increased (P < 0.001). They proposed that integrating culprit lesion PCAT attenuation into the analysis of adverse plaque features might enhance the identification of patients with PMI (OR = 2.89, P < 0.001). Le et al[62] further assessed the role of FAI as an independent predictor of PCI treatment outcomes for CTO, and based on lesion length ≥ 15 mm, severe calcification, and FAI < 69.5 HU, they developed the PCATA-CTO scoring system (AUC = 0.72). This scoring system outperformed other predictive scoring models in terms of interobserver consistency (kappa = 0.74) and intraobserver consistency (kappa = 0.90, both P < 0.01), resulting in significantly improved prediction model reliability. Conclusively, PCAT attenuation and its radiomic features have significant clinical potential as adjunctive indicators for optimizing risk evaluation and control in patients undergoing PCI by predicting postintervention complications.

PCAT in non-ischemic heart disease

Atrial fibrillation: Previous studies have found a strong link between epicardial fat and the risk of atrial fibrillation (AF), but recent findings suggest that PCAT may also be useful in predicting AF recurrence. Gerculy et al[63] utilized the FAI score (personalized assessment adjusted for age and gender) to determine the inflammatory levels in the three main coronary arteries, showing that AF patients had considerably greater average FAI scores than non-AF patients (15.53 ± 10.29 vs 11.09 ± 6.70, P < 0.05), with the LAD region exhibiting the most noticeable inflammation (13.17 ± 7.91 vs 8.80 ± 4.75, P = 0.008), potentially reflecting increased perivascular inflammation in the left coronary circulation of AF patients. To predict AF recurrence after catheter ablation, Nogami et al[64] carried out a follow-up study on 364 patients who underwent their first cryoballoon ablation (CBA) for AF. The study found that 90 patients (24.7%) had recurrence. The average PCAT attenuation of the three main coronary arteries was considerably greater in the recurrence group than in the non-recurrence group (-67.09 ± 6.60 HU vs -69.45 ± 6.86 HU, P = 0.005). Furthermore, PCAT attenuation was found to be an independent predictor of recurrence following CBA (HR = 1.03, P = 0.046). Ma et al[65] investigated 189 patients with persistent AF who received their first radiofrequency ablation treatment. The researchers discovered that LCX FAI was a significant independent risk factor for AF recurrence (OR = 1.254, P < 0.01). Specifically, LCX FAI > -81.5 HU had an AUC of 0.722 for predicting AF recurrence, with 87.2% sensitivity and 47.2% specificity, outperforming epicardial fat-related parameters. Another study discovered that RCA PCAT attenuation was an independent risk factor for AF recurrence (HR = 1.04, P = 0.001) and that including RCA PCAT attenuation in clinical models significantly improved AF recurrence prediction (AUC = 0.724 vs 0.686, P = 0.024)[66]. The potential mechanism for these findings is that, following AF catheter ablation, PCAT may alter the phenotype of perivascular fat by secreting inflammatory factors, influencing the risk of AF recurrence. However, the coronary artery regions affected vary between studies, and the underlying mechanisms of this phenomenon need to be investigated further.

Heart failure: PCAT has been closely linked to heart failure with preserved ejection fraction (HFpEF), most likely because it is associated with coronary microvascular dysfunction and chronic perivascular inflammation[67]. A retrospective study by Nishihara et al[68] revealed that PCAT attenuation in the three main coronary arteries was significantly higher in HFpEF patients than in healthy controls. Each of these parameters was found to be an independent predictor of HFpEF (LAD: OR = 1.427, P < 0.001; LCX: OR = 1.619, P < 0.001; RCA: OR = 1.372, P = 0.007), indicating that PCAT inflammation may be a key factor in the development and progression of HFpEF, possibly through local microvascular dysfunction and the effects of inflammation on the left ventricle. Furthermore, PCAT attenuation has shown significant value in predicting the risk of readmission for heart failure. Liu et al[69] tested 107 non-ischemic heart failure patients and 129 healthy controls and discovered that PCAT attenuation in three coronary arteries was higher in heart failure patients and was closely linked to the risk of readmission (LAD: HR = 1.3, P = 0.002; LCX: HR = 1.33, P < 0.001; RCA: HR = 1.1, P = 0.002). These findings suggest that PCAT attenuation may aid in the early detection of HFpEF and independently predict the risk of heart failure readmission. However, research into the role of PCAT in HFpEF and heart failure is still limited, and its underlying mechanisms and clinical applications require further investigation.

PCAT in metabolic diseases

Type 2 diabetes mellitus: Patients with type 2 diabetes mellitus (T2DM) are at a higher risk of developing CAD, which is the leading cause of T2DM-related death and is closely linked to microvascular dysfunction caused by long-standing chronic inflammation[70]. As a result, assessing coronary artery inflammation and vascular dysfunction using PCAT parameters in the early, asymptomatic stages may aid in the identification of high-risk T2DM patients. A study found that notwithstanding the presence of obstructive stenosis or high-risk plaques on CCTA, RCA PCAT attenuation in diabetic patients was consistently and significantly higher than that in non-diabetic patients (-83.60 ± 9.51 HU vs -88.58 ± 9.37 HU, P < 0.001). After accounting for tube voltage, the PCAT attenuation differed significantly between the two groups[71]. Furthermore, Dong et al[72] discovered that RCA PCAT attenuation was an independent predictor of CAD in T2DM patients (OR = 1.151, P < 0.001), outperforming the LAD and LCX arteries. A prediction model incorporating clinical features and CT imaging markers showed the highest diagnostic performance (AUC = 0.93, sensitivity = 88.5%, specificity = 77.3%). Furthermore, PCAT radiomics has shown significant value in detecting CAD. Miao et al[73] discovered that the Rad-score based on PCAT radiomic features was an independent predictor of obstructive CAD (OR = 1.297, P < 0.001), outperforming PCAT attenuation alone (AUC = 0.835 vs 0.550)[74]. However, Dong et al[72] discovered that PCAT radiomic features did not significantly improve the performance of the model based on clinical and CT imaging markers in diagnosing T2DM with CAD (training group AUC = 0.961 vs 0.960, validation group AUC = 0.929 vs 0.930, P > 0.05), perhaps due to the inclusion of T2DM patients with pre-existing coronary artery inflammation, which may have interfered with the spatial distribution of PCAT voxel intensities. While PCAT parameters have the potential for assessing coronary artery inflammation and CAD risk in T2DM patients, the study of PCAT radiomics is hampered by inflammation interference and research heterogeneity, necessitating additional optimization and validation to enhance clinical applicability.

PCAT attenuation is critical in early risk assessment and has the potential to predict long-term cardiovascular outcomes in T2DM patients. According to a study by Ichikawa et al[75] patients who experienced cardiovascular events, such as cardiac death, acute coronary syndrome hospitalization, late coronary revascularization, or heart failure hospitalization had considerably higher LAD PCAT attenuation than those who did not (-68.5 ± 6.5 HU vs -70.8 ± 6.1 HU, P = 0.045), with LAD PCAT attenuation > -70.7 HU being strongly linked to cardiovascular events (HR = 2.69, P = 0.020). Furthermore, Liu et al[76] followed 304 T2DM patients for 3 years and discovered that patients with MACE had considerably greater lesion-specific PCAT attenuation than those without MACE (-84.87 ± 11.36 HU vs -88.65 ± 11.89 HU, P = 0.016), and PCAT attenuation > -83.5 HU was independently associated with an elevated risk of MACE (HR = 2.40, 95% confidence interval: 1.399-4.120, P = 0.001). However, another study found that a combined model based on traditional clinical factors and high-risk anatomical features from CCTA performed better in predicting MACE in T2DM patients, while PCAT radiomic features did not add value to risk stratification[77]. These findings suggest that PCAT attenuation holds potential as a prognostic tool for T2DM patients; however, the practical application of PCAT radiomics in clinical decision-making requires further investigation.

Recent research has shown that PCAT attenuation can be used to assess the improvement of inflammation following medication therapy in T2DM patients. Biesenbach et al[78] discovered that medication therapy significantly reduced LAD PCAT attenuation (β coefficient = -2.4, P = 0.029). Another study found that the medication treatment group had significantly lower RCA PCAT density than the control group [-85.0 (-89.0, -82.0) HU vs -79.0 (-83.5, -74.0) HU, P < 0.001][79]. Furthermore, Liu et al[80] found that, compared to T2DM patients not receiving glycemic-lowering interventions, those who received medication had significantly lower PCAT attenuation in all three coronary arteries (LAD: -78.11 ± 8.01 HU vs -75.04 ± 8.26 HU, P = 0.022; LCX: -71.10 ± 8.13 HU vs -68.31 ± 7.90 HU, P = 0.037; RCA: -78.17 ± 8.64 HU vs -73.35 ± 9.32 HU, P = 0.001). These findings suggest that medication may improve coronary artery inflammation, which is consistent with the findings of Li et al[81]. Their study further revealed substantial variations in PCAT attenuation between T2DM patients with poor glycemic control and non-T2DM patients (LCX: -68.75 ± 7.59 HU vs -71.93 ± 7.25 HU, P = 0.008; RCA: -74.37 ± 8.44 HU vs -77.2 ± 7.42 HU, P = 0.026)[80]. As a result, PCAT attenuation may be a more sensitive and useful biomarker than traditional indicators for monitoring T2DM treatment effectiveness and assessing cardiovascular risk in patients with T2DM.

Non-alcoholic fatty liver disease: Non-alcoholic fatty liver disease (NAFLD) is an independent risk factor for cardiovascular mortality, and it may increase the risk of coronary heart disease and other cardiovascular events by causing coronary artery inflammation. Recent research has shown that PCAT attenuation can be used as an imaging-based predictor of cardiovascular risk in NAFLD patients, with potential clinical applications for early identification of high-risk patients and risk stratification. Ichikawa et al[82] discovered that the average PCAT attenuation in NAFLD patients was significantly higher than that in non-NAFLD patients (OR = 2.912, P = 0.005), implying that elevated PCAT attenuation may be an independent predictor of cardiovascular disease in NAFLD patients. Furthermore, Li et al[83] found that NAFLD could cause increased inflammation in three major coronary arteries, with LAD FAI identified as a risk factor for stable CAD in NAFLD patients. Their study used least absolute shrinkage and selection operator regression to select nine radiomic features to create a Rad-score, which, when combined with clinical risk factors and imaging markers, significantly improved diagnostic performance for stable angina (AUC = 0.914, sensitivity = 0.814, specificity = 0.941), emphasizing the importance of PCAT attenuation and radiomics in the early detection of NAFLD-related cardiovascular diseases.

PCAT attenuation measures coronary artery inflammation and predicts the risk of MACE in NAFLD patients. Yang et al[84] examined 514 patients with acute chest pain and discovered that NAFLD (HR = 2.599, P = 0.015) and RCA PCAT attenuation (HR = 1.026, P = 0.038) were independent predictors of MACE, implying that PCAT attenuation can be used to assess cardiovascular risk in NAFLD patients. Another study discovered that LAD PCAT attenuation was an independent predictor of cardiovascular death, nonfatal acute coronary syndrome, and hospitalization for heart failure (HR = 3.321, P = 0.014) and that combining it with high-risk plaque features could improve cardiovascular risk stratification in NAFLD patients, allowing for more precise management strategies[85]. They also discovered that reducing PCAT attenuation could help prevent cardiovascular events in NAFLD patients. However, there is currently no direct evidence linking pharmaceutical interventions that reduce PCAT attenuation to a better long-term prognosis in NAFLD patients. Future research should look into the dynamic changes in PCAT attenuation and its role in drug therapy to optimize cardiovascular risk management strategies for this population.

PCAT in inflammatory diseases

Coronavirus disease 2019-related myocardial injury: Recent research indicates that PCAT may play a role in the pathophysiology of coronavirus disease 2019 (COVID-19), particularly in disease severity and cardiovascular complications[86]. Turker Duyuler et al[87] studied the relationship between PCAT thickness and COVID-19 severity in 504 patients and discovered a significant difference in PCAT thickness between intensive care unit (ICU) patients and non-ICU patients [11.2 (10.3-13.2) mm vs 9.3 (7.4-11.5) mm, P < 0.001]. Furthermore, PCAT thickness was identified as an independent predictor of ICU admission (OR = 1.111, P = 0.031), with a progressive increase in PCAT thickness correlating with disease severity. These findings suggest that PCAT thickness could be used as a marker to identify COVID-19 patients who are at higher risk of disease progression. Aside from its role in acute COVID-19 severity, PCAT may also help with cardiovascular risk assessment in post-COVID-19 patients. Mátyás et al[88] used artificial intelligence to analyze coronary perivascular FAI, revealing that COVID-19 survivors had significantly higher levels of coronary perivascular inflammation than non-infected individuals, implying that COVID-19 infection may increase the risk of coronary plaque instability and supporting FAI as a biomarker for identifying high-risk individuals and guiding targeted cardiovascular prevention strategies. Further evidence from Tuttolomondo et al[89] found that PCAT attenuation was significantly associated with higher mortality in severe COVID-19 cases (P < 0.001). Even after accounting for covariates, PCAT attenuation remained an independent predictor of in-hospital mortality in COVID-19 patients. These findings highlight PCAT’s clinical potential in predicting COVID-19 severity and cardiovascular complications; however, the effect of coronary perivascular FAI on localized plaque vulnerability in COVID-19 patients remains unknown. Additional large-scale studies and clinical data are needed to determine the clinical utility of FAI in COVID-19-related cardiovascular risk stratification.

Autoimmune-mediated chronic inflammatory diseases: Takayasu arteritis (TAK), a chronic inflammatory vascular disease that affects the aorta and its major branches, has been linked to PCAT attenuation, which is an imaging biomarker for disease activity and has the potential to be used for risk evaluation and therapy monitoring in chronic inflammatory diseases. Zhao et al[90] discovered that PCAT attenuation is an independent predictor of active TAK (OR = 1.57, P < 0.001), with high diagnostic performance (AUC = 0.911, sensitivity = 93.9%, specificity = 74.4%, P < 0.001), which is consistent with Wall et al[91]. Zhao et al[90] emphasized that PCAT parameters can detect TAK disease activity and distinguish between active coronary arteritis and stable coronary atherosclerosis in TAK patients. However, relevant studies are limited, and the specific clinical application of PCAT in patients with TAK requires additional research and validation. Beyond TAK, PCAT attenuation has been studied in other chronic inflammatory diseases, including psoriasis. Elnabawi et al[92] found a significant reduction in PCAT attenuation in psoriasis patients receiving biologic therapy, implying that such treatments may reduce coronary artery inflammation. In contrast, Bao et al[93] discovered lower PCAT attenuation in psoriasis patients compared to healthy controls, which could be attributed to psoriasis’ chronic low-grade inflammatory state, which is insufficient to induce significant changes in PCAT attenuation, highlighting the need for further research into its dynamic changes across populations. Although existing research suggests that PCAT attenuation has the potential to improve risk assessment and disease monitoring in chronic inflammatory diseases, large-scale, long-term follow-up studies are still needed.

CONCLUSION

With the widespread use of CCTA, PCAT has emerged as a promising noninvasive imaging biomarker for measuring coronary artery inflammation and plaque vulnerability. Its clinical applications include assessing coronary hemodynamics, predicting MACE, and providing an imaging-based foundation for early screening, precise risk stratification, treatment efficacy assessment, and prognostic analysis in cardiovascular diseases. However, there are some limitations to the current research on PCAT based on CCTA. First, the optimal measurement site and range for PCAT attenuation have yet to be clearly stated, with current studies primarily focusing on proximal coronary artery PCAT attenuation. However, research suggests that lesion-specific PCAT attenuation may better reflect local inflammation and plaque vulnerability, potentially providing superior predictive value for coronary events. Second, PCAT measurements are affected by individual characteristics and imaging parameters, which limit data comparability across studies and highlight the need for future multicenter research to optimize measurement techniques and establish standardized protocols. Furthermore, recent studies have shown that PCAT parameters derived from non-contrast coronary CT may be useful for early cardiovascular disease screening, particularly in patients with iodine contrast allergies or chronic kidney disease, emphasizing the need for more research into their clinical utility for high-risk patient screening and early disease detection. In the future, artificial intelligence-based radiomic analysis of PCAT is expected to serve as a noninvasive tool for improving diagnostic accuracy, predicting cardiovascular risk, and guiding personalized treatment strategies for CAD. Thus, while PCAT has significant clinical potential in the management of cardiovascular diseases, more research is needed to improve measurement methods and investigate its use in precision medicine.