Coronary computed tomography angiography adipose tissue attenuation in diabetic patients with myocardial ischemia and non-obstructive coronary arteries

Abstract

BACKGROUND

Type 2 diabetes mellitus (T2DM) substantially increases the risk of cardiovascular disease, including ischemia with non-obstructive coronary artery disease (INOCA). Coronary computed tomography angiography (CCTA) enables early detection of coronary abnormalities; however, it may fail to identify INOCA due to the absence of overt stenosis. Pericoronary adipose tissue attenuation (PCATa) values derived from CCTA in the proximal segments of the left anterior descending (LAD), left circumflex (LCX), and right coronary artery (RCA) serve as effective imaging biomarkers for coronary inflammation. However, its clinical use for identifying INOCA in patients with T2DM remains poorly defined.

AIM

To investigate PCATa differences and their diagnostic value for identifying INOCA in patients with T2DM.

METHODS

This retrospective study involved 228 T2DM patients underwent CCTA and 120 healthy individuals. The mean PCATa values within the proximal segments of the three major coronary arteries were compared between groups. Further subgroup analysis was performed to assess the differences in PCTAa and clinical characteristics between T2DM patients with and without INOCA. Logistic regression analysis was conducted to identify the independent risk factors for INOCA, and the receiver operating characteristic curves were generated to evaluate the diagnostic performance of each indicator.

RESULTS

Compared with controls, T2DM patients exhibited significantly higher PCATa values in all three major coronary arteries. Among them, those with concomitant INOCA showed further increases compared to those without INOCA (all P < 0.05). Multivariate logistic regression identified age; female sex; elevated glycated hemoglobin; and increased PCATa in the LAD, LCX, and RCA as independent risk factors for INOCA. Receiver operating characteristic analysis showed good diagnostic performance for PCATa [LAD area under the curve (AUC) = 0.809; LCX AUC = 0.777; RCA AUC = 0.758], outperforming traditional clinical indicators (AUC = 0.731). Combining PCATa with clinical parameters yielded the highest diagnostic accuracy (LAD AUC = 0.851; LCX AUC = 0.842; RCA AUC = 0.841).

CONCLUSION

Elevated proximal PCATa is an independent risk factor for INOCA in T2DM. Combining PCATa with clinical data improves diagnostic performance in this population.

Key Words: Type 2 diabetes mellitus; Ischemia with non-obstructive coronary arteries; Pericoronary adipose tissue attenuation index; Coronary computed tomography angiography; Coronary artery disease

Core Tip: This study reveals significantly increased pericoronary adipose tissue attenuation (PCATa) in the proximal segments of the left anterior descending, left circumflex, and right coronary arteries in patients with type 2 diabetes mellitus, particularly those with ischemia with non-obstructive coronary arteries (INOCA). PCATa serves as an independent imaging biomarker for identifying INOCA and demonstrated superior diagnostic performance compared with conventional clinical indicators. The combination of PCATa and clinical parameters further improves diagnostic accuracy. These results highlight the potential value of coronary computed tomography angiography-derived PCATa in early risk stratification and noninvasive identification of INOCA in patients with type 2 diabetes mellitus.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder characterized by disturbances in glucose and lipid metabolism. Long-standing metabolic abnormalities contribute to multisystem damage, with cardiovascular disease being the leading cause of mortality among affected individuals[1]. In individuals with T2DM, sustained hyperglycemia contributes to oxidative stress and inflammatory response[2]. These alterations promote the development of atherosclerotic plaque and play a key role in the pathogenesis of cardiovascular disease[3]. In light of these risks, the current clinical guidelines recommend the use of standardized coronary computed tomography angiography (CCTA) in patients with diabetes mellitus to facilitate the early detection of cardiovascular risk[4]. Although CCTA serves as an effective modality for early screening and surveillance of coronary artery disease, its diagnostic utility is primarily limited to identifying anatomically significant stenotic lesions. Consequently, it may fail to elucidate ischemic symptoms in patients without obstructive coronary stenosis. Recent evidence indicates that obstructive coronary artery disease is identified in < 40% of T2DM patients undergoing CCTA for chest pain evaluation. Notably, more than half of these patients exhibit no significant coronary obstruction, despite experiencing symptoms suggestive of myocardial ischemia. Subsequent examinations confirm that a substantial proportion of this subgroup meets the diagnostic criteria for ischemia with non-obstructive coronary artery disease (INOCA)[5]. The pathophysiology of INOCA involves multiple interrelated mechanisms, including chronic inflammation and persistent oxidative stress. Owing to their subtle and localized nature, these inflammatory changes are often difficult to detect in the early stages[6]. Furthermore, accumulating evidence indicates that INOCA is a significant contributor to major adverse cardiovascular events and cardiovascular mortality[7]. However, the limited clinical awareness of INOCA often leads to the misattribution of myocardial ischemia symptoms to non-cardiac chest pain when CCTA reveals no significant coronary stenosis. This diagnostic oversight may result in a failure to initiate appropriate management, thereby increasing the risk of adverse cardiovascular outcomes[8]. Therefore, early detection of INOCA in T2DM patients is of significant clinical importance, as it enables timely intervention to improve the quality of life and reduce the cardiovascular mortality risk of patients.

Currently, the diagnosis of INOCA in clinical practice primarily relies on a combination of CCTA/digital subtraction angiography, electrocardiographic findings, and laboratory indicators. However, this approach remains complex, and certain parameters may be confounded by recent pharmacologic interventions[9]. Therefore, there is a growing need for a stable and reliable biomarker capable of identifying INOCA. Moreover, increasing evidence suggests that inflammation in the pericoronary region plays a central role in the pathogenesis of INOCA, in the absence of significant coronary stenosis. In this regard, the pericoronary adipose tissue attenuation (PCATa) index, derived from CCTA, has gained attention as a novel noninvasive imaging biomarker for assessing coronary inflammation in clinical and research settings[10]. When coronary artery disease is present, vascular endothelial cells release proinflammatory cytokines that initiate paracrine signaling to the surrounding pericoronary adipose tissue, leading to lipid metabolic alterations detectable as measurable changes in PCATa. Consequently, the comprehensive evaluation of pericoronary adipose tissue characteristics using CCTA may help identify patients with INOCA in the context of T2DM.

Elevated PCATa is an independent predictor of cardiovascular risk in individuals with nonalcoholic fatty liver disease (NAFLD)[11]. Given the shared state of chronic low-grade systemic inflammation in both NAFLD and T2DM, evidence derived from NAFLD populations may have translational relevance to individuals with T2DM. However, the current data regarding the application of PCATa in diabetes mellitus populations remain limited, particularly in the context of evaluating the INOCA risk. Although existing research supports the potential utility of PCATa in assessing coronary inflammation[12], its diagnostic accuracy for identifying INOCA in T2DM patients has yet to be validated. The present study aimed to evaluate the predictive value of PCATa for INOCA risk in T2DM by integrating clinical indicators and CCTA imaging parameters. The study findings may provide imaging-based evidence to support the early identification of high-risk individuals and inform timely clinical decision-making.

MATERIALS AND METHODS

Study population

The present study was conducted in accordance with the guidelines stipulated in Declaration of Helsinki and approved by the Ethics Committee of the Affiliated Hospital of North Sichuan Medical College (Approval No. 2024ER648-1). The requirement for obtaining informed consent from the patients was waived due to the retrospective nature of the present study. Our retrospective analysis included patients with clinically diagnosed T2DM who underwent CCTA at the Affiliated Hospital of North Sichuan Medical College between January 2020 and August 2024.

T2DM was diagnosed in accordance with the 2023 European Society of Cardiology Guidelines for diabetes and cardiovascular disease[13]. The T2DM diagnosis was established if any of the following criteria were met: (1) Presence of classic hyperglycemic symptoms accompanied by a random plasma glucose level of ≥ 11.1 mmol/L on at least two occasions, a fasting plasma glucose level of ≥ 7.0 mmol/L on at least two occasions, or a 2-hour plasma glucose level of ≥ 11.1 mmol/L following an oral glucose tolerance test; (2) Glycated hemoglobin (HbA1c) level of ≥ 6.5%; and (3) A documented history of physician-diagnosed diabetes.

The diagnosis of INOCA was independently confirmed by two cardiologists according to established criteria[14,15]. Patients were diagnosed when the degree of coronary artery stenosis is < 50% coexisting with at least one objective indicator of myocardial ischemia: (1) Typical angina pectoris symptoms; (2) Ischemic changes on 12-lead electrocardiogram (ECG) during symptomatic episodes; (3) Evidence of stress-induced myocardial perfusion defects or regional ventricular wall motion abnormalities; and (4) Impaired coronary flow reserve.

The inclusion criteria were clinically diagnosed T2DM with completed CCTA examination. The exclusion criteria were as follows: (1) History of coronary stenting or coronary artery bypass grafting within the previous 3 months; (2) Presence of acute or chronic inflammatory disease; (3) Incomplete clinical datasets; (4) Poor-quality CCTA images (defined as an unmeasurable Agatston score due to motion artifacts); (5) Severe cardiac, pulmonary, or renal dysfunction, or allergy to iodinated contrast agents; (6) Congenital coronary artery anomalies (including origin, course, or developmental variants); and (7) ≥ 50% stenosis in any major coronary artery branch [i.e., left anterior descending artery (LAD), left circumflex artery (LCX), or right coronary artery (RCA)]. Ultimately, 228 T2DM patients were recruited in the present study; of these, 110 were diagnosed with INOCA, whereas 118 exhibited no evidence of coronary artery disease. Additionally, 120 healthy individuals undergoing routine health examinations at our institution who received CCTA were included as controls.

CCTA acquisition

All patients underwent CCTA using a third-generation dual-source computed tomography (CT) scanner (SOMATOM Force, Siemens Healthineers, Germany). Prior to the examination, the heart rates of all participants were controlled, and respiratory training was conducted. For patients with baseline heart rates of > 80 beats per minute, intravenous metoprolol (25-100 mg) was administered to reduce the heart rate to ≤ 80 beats per minute. Patients also received respiratory training to ensure breath-hold at end-expiration. During image acquisition, patients were positioned supine with a head-first orientation. ECG electrodes were placed on the left anterior chest wall to optimize R-wave signal detection. After identifying the optimal scan center, prospective ECG-triggered CCTA was performed during end-expiratory breath-hold. The scanning parameters were as follows: Rotation time of 280 milliseconds, tube current of 400 mA, detector collimation of 256 mm × 0.625 mm, and ECG gating during the 35% to 75% phase of the RR interval. Tube voltage was automatically modulated between 70 kV and 120 kV to optimize the image quality and minimize radiation exposure. All CCTA images were reconstructed using standard kernel with 50% adaptive statistical iterative reconstruction-V (ASiR-V, GE Healthcare, IL, United States). Diastolic phase images were selected for subsequent analysis. To minimize the influence of varying tube voltage on PCATa measurements, we applied standardized correction factors derived from previously validated studies[16], adjusting all PCATa values to an equivalent 120 kV peak. Specifically, PCATa values obtained at lower tube voltages (e.g., 70 kV peak, 100 kV peak) were divided by the corresponding voltage-specific correction coefficients to ensure consistency across scans.

CCTA imaging parameters

CCTA image analysis was performed using dedicated cardiovascular software (Shukun Technology Co., Ltd, Beijing, China), and coronary artery segmentation was conducted using internally developed proprietary software (TIMESlice, version 4.0.19, Shenyang, China). PCATa was quantified as the mean CT value of adipose voxels (-190 HU to -30 HU) within a radial distance equivalent to the vessel diameter surrounding the outer coronary wall[17]. Standardized measurements were acquired in the proximal RCA segment (10-50 mm from the ostium, avoiding the aortic wall effects), proximal LAD, and proximal LCX, with all segments analyzed using an established methodology, summarized in Figure 1. Inter-observer and intra-observer variability analyses of PCATa were performed based on the coefficient of variation (CV) to evaluate the measurement reproducibility of PCATa. Variability was calculated using the formula: CV (%) = (s/X) × 100, where s denotes the SD and X represents the mean of the cross-sectional areas. Precision was expressed as an average%CV. The initial measurements were considered as the final PCATa when the CV was below 15%.

Figure 1 The methodology of the pericoronary adipose tissue analysis. A: Left anterior descending artery; B: Left circumflex artery proximal 40-mm segments; C: Right coronary artery proximal 10-50 mm segment. Pericoronary adipose tissue attenuation was quantified within the adipose voxels (-190 HU to -30 HU), extending radially from the vessel wall at distances equivalent to the coronary artery diameters.

The coronary artery calcium score (CACS) was calculated from non-contrast-enhanced CT scans using uAI Sphere software (United Imaging Intelligence Co., Ltd, Shanghai, China) according to the Agatston method[18]. CACS was determined by multiplying the area of each calcified lesion (mm²) by a weighting factor based on the peak CT attenuation. The density scores were assigned as follows: 1 for 130-199 HU, 2 for 200-299 HU, 3 for 300-399 HU, and 4 for ≥ 400 HU. The total CACS for each coronary artery was obtained by summing all individual calcified plaque scores within that vessel. Additionally, patients were stratified into three groups based on their total CACS values: CACS = 0, CACS 1-99, and CACS ≥ 100[19].

All coronary segments were subjected to semi-automated plaque quantification using dedicated analysis software. The segments with a diameter of < 2 mm, those affected by severe motion artifacts, or those with insufficient contrast opacification were excluded from the analysis. The software automatically identified scan-specific thresholds to quantify the plaque components within the manually defined regions of interest. The total plaque volume was calculated for each patient. All automated outputs were reviewed and manually adjusted when necessary to ensure measurement accuracy.

All measurements were independently performed by two cardiovascular radiologists who were blinded to the patients’ clinical and prognostic information. After each observer conducted separate assessments of the lesions, the average of their measurements was calculated and adopted as the final quantitative value.

Statistical analysis

All statistical analyses were conducted using SPSS (version 27.0, SPSS Inc., Chicago, IL, United States) and R studio (version 4.3.3). The Shapiro-Wilk test was applied to assess the normality of continuous variables. Normally distributed data were compared using independent t and presented as mean ± SD, while non-normally distributed data were expressed as median (interquartile range) and compared using the Mann-Whitney U test. A continuous variable that contains a substantial proportion of zero values may be converted into a binary variable (e.g., presence vs absence) to minimize the effect of zero inflation on statistical comparisons and preserve analytical validity. Categorical variables were compared using the χ² test and summarized as frequencies and percentages. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors associated with INOCA in T2DM patients. A sensitivity analysis was conducted to assess the robustness of PCATa as a predictor by including statins, antiplatelet agents, and insulin (commonly prescribed drugs with known anti-inflammatory and metabolic effects) as additional covariates in the multivariate models. Diagnostic performance was evaluated using receiver operating characteristic (ROC) curves, with area under the curve (AUC) comparisons conducted via DeLong’s test. Net reclassification improvement (NRI) and integrated discrimination improvement (IDI) analyses were conducted to further assess the incremental value of PCATa over conventional clinical factors. P < 0.05 was considered statistically significant.

RESULTS

Baseline characteristics

The present study included 228 participants (mean age 66.71 ± 10.44 years; 58.33% female), of whom 110 (48.24%) were diagnosed with INOCA. The baseline clinical characteristics of the control, non-INOCA, and INOCA groups are summarized in Table 1. Compared with the healthy control group, the T2DM group were significantly older and had significantly higher proportion of females, lower body mass index (BMI), and higher systolic blood pressure, triglyceride, total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol levels (all P < 0.05). The T2DM group was further divided into two subgroups according to the presence of INOCA. Compared with the non-INOCA group, the INOCA group were older and had a higher proportion of females and smokers. This group also exhibited lower BMI and TC, LDL-C, and high-density lipoprotein cholesterol levels. Additionally, the fasting plasma glucose and HbA1c levels were significantly higher and the platelet and lymphocyte counts were significantly lower in the INOCA group. The INOCA group also demonstrated higher usage rates of antiplatelet agents, statins, and insulin therapy (all P < 0.05).

Table 1 Comparison of the patients’ baseline characteristics among groups, median (interquartile range)/n (%)/mean ± SD.

	Healthy control group (n = 120)	Non-INOCA group (n = 118)	INOCA group (n = 110)
Clinical characteristics
Age (years)	56 (50-62)	66 (58-74)a	69 (60.75-76)a,b
Gender (female)	66 (55.0)	61 (51.7)	72 (62.5)a,b
BMI, kg/m2	22.82 (20.81-24.21)	24.98 (23.42-27.07)a	24.22 (21.74-26.31)a,b
Smoking	29 (24.16)	34 (28.8)	46 (41.8)a,b
Drinking	32 (26.67)	36 (30.5)	35 (31.8)
SBP, mmHg	122 (112-130)	133 (124-141.25)a	137 (123.75-148.5)a
DBP, mmHg	77.63 ± 12.21	82.36 ± 10.85a	80.40 ± 13.98
Laboratory tests
TG, mmol/L	1.19 (0.92-1.60)	1.49 (1.01-2.31)a	1.35 (1.05-1.94)a
LDL-C, mmol/L	2.60 (2.19-3.08)	2.55 (1.90-3.14)	2.12 (1.50-2.65)a,b
TC, mmol/L	4.64 ± 1.06	4.71 ± 1.22	4.10 ± 1.22a,b
HDL-C, mmol/L	1.27 (1.04-1.48)	1.12 (0.96-1.35)a	1.08 (0.89-1.25)a,b
Fast glucose, mmol/L	5.09 (4.47-5.49)	7.26 (6.07-8.42)a	7.91 (6.75-10.43)a,b
HbA1c (%)	5.87 (5.64-6.12)	7.14 (6.68-7.85)a	7.76 (7.02-9.46)a,b
WBC, × 10 9/L	5.55 (4.81-7.14)	6.43 (5.26-8.10)	6.62 (5.28-8.01)
NEU, × 10 9/L	3.44 (2.67-4.38)	4.06 (3.16-5.27)	4.17 (3.17-5.60)
LYM, × 10 9/L	1.67 (1.34-2.03)	1.75 (1.35-2.14)	1.54 (1.18-1.84)b
MONO, × 10 9/L	0.32 (0.26-0.42)	0.38 (0.29-0.49)	0.38 (0.31-0.51)
PLT, × 10 9/L	196.18 ± 63.42	192.8 ± 57.88	175.95 ± 58.87b
Medication situation
Antiplatelet agents		44 (37.3)	74 (67.3)b
Diuretics		10 (8.5)	15 (13.6)
β-blockers		23 (19.5)	25 (22.7)
ACEI/ARB		31 (26.3)	30 (27.3)
Statins		65 (55.1)	86 (78.2)b
CCBs		39 (33.1)	46 (41.8)
Hypoglycemic agents		72 (61.0)	78 (70.9)
Insulin therapy		19 (16.1)	43 (39.1)b

^aP < 0.05 vs control group.

^bP < 0.05 vs non-ischemia with non-obstructive coronary artery disease group.

BMI: Body mass index; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; TG: Triglycerides; LDL-C: Low-density lipoprotein cholesterol; TC: Total cholesterol; HDL-C: High-density lipoprotein cholesterol; HbA1c: Glycated hemoglobin; WBC: White blood cell; NEU: Neutrophil; LYM: Lymphocyte; MONO: Monocyte; PLT: Platelet; ACEI: Angiotensin-converting enzyme inhibitor; ARB: Angiotensin receptor blocker; CCB: Calcium channel blocker; INOCA: Ischemia with non-obstructive coronary arteries.

Image characteristics

As summarized in Table 2, PCATa was significantly higher in T2DM patients (LAD: -89.81 ± 19.77 vs -77.31 ± 7.77; LCX: -83.52 ± 8.09vs -71.08 ± 7.11; RCA: -92.69 ± 8.89 vs -77.90 ± 8.19; all P < 0.05) than in the control group (P < 0.05). In the subgroup analysis, the INOCA group showed further increases in PCATa as compared to the non-INOCA group (LAD: -73.03 ± 7.28vs -81.29 ± 5.87; LCX: -67.46 ± 6.68 vs -74.45 ± 5.71; RCA: -74.02 ± 7.71 vs -81.51 ± 6.90; all P < 0.05). Notably, although the INOCA group demonstrated numerically higher plaque volumes and CACS than the non-INOCA group, these differences did not reach statistical significance.

Table 2 Comparison of the image characteristics among groups, mean ± SD/n (%).

	Healthy control group (n = 120)	Non-INOCA group (n = 118)	INOCA group (n = 110)
PCATa (HU)
LAD	-89.81 ± 19.77	-81.29 ± 5.87a	-73.03 ± 7.28a,b
LCX	-83.52 ± 8.09	-74.45 ± 5.71a	-67.46 ± 6.68a,b
RCA	-92.69 ± 8.89	-81.51 ± 6.90a	-74.02 ± 7.71a,b
Presence of coronary plaque
LAD		46 (39.0)	49 (44.54)
LCX		41 (34.7)	47 (42.7)
RCA		38 (32.2)	40 (37.0)
CACS
0		63 (53.38)	51 (45.46)
1-99		55 (46.61)	53 (49.09)
≥ 100		0	6 (5.4)

^aP < 0.05 vs control group.

^bP < 0.05 vs non-ischemia with non-obstructive coronary artery disease group.

PCATa: Pericoronary adipose tissue attenuation; LAD: Left anterior descending artery; LCX: Left circumflex artery; RCA: Right coronary artery; CACS: Coronary artery calcium score; INOCA: Ischemia with non-obstructive coronary artery disease.

Repeatability of measurement of PCATa

To assess the reproducibility of PCATa measurements, we calculated intra- and inter-observer %CV. The intra-observer %CVs were 3.46% for LAD, 3.67% for LCX, and 3.45% for RCA, while the corresponding inter-observer %CVs were 5.54%, 5.30%, and 5.27%. These findings indicate that PCATa measurements are stable and reproducible across both repeated and independent assessments.

Identification of independent predictors of INOCA in T2DM patients

To identify the independent risk factors for INOCA in T2DM patients, logistic regression analysis was conducted. In the univariate analysis, advanced age, female sex, lower BMI, history of smoking, elevated levels of LDL-C, TC, fasting plasma glucose, and HbA1c, diabetes duration of > 10 years, reduced platelet count, and higher diastolic PCATa values across all three major coronary branches were associated with the presence of INOCA (all P < 0.05). Subsequent multivariate logistic regression analysis identified smoking history, lower BMI, increased TC and HbA1c levels, and higher diastolic PCATa in the proximal segments of the LAD, LCX, and RCA as independent predictors of INOCA in T2DM patients. A sensitivity analysis was conducted to assess the robustness of the association between PCATa and INOCA in the presence of potential pharmacological confounders by including statins, antiplatelet agents, and insulin as covariates in the multivariate logistic regression models. After adjustment, PCATa values from all three major coronary arteries remained significantly associated with INOCA, with only minimal alterations in odds ratios [LAD: Adjusted odds ratio (OR) for 1.246, 95% confidence interval (CI): 1.168-1.329; LCX: Adjusted OR = 1.207, 95%CI: 1.138-1.281; RCA: Adjusted OR = 1.164, 95%CI: 1.105-1.227; all P < 0.001], supporting the robustness of the results (Table 3).

Table 3 Logistic regression analysis of the factors associated with ischemia with non-obstructive coronary artery disease.

	Univariable			Multivariable
	OR	95%CI	P value	OR	95%CI	P value
Age	1.031	1.005-1.058	0.021	1.037	1.008-1.067	0.013
Gender (female)	1.844	1.079-3.150	0.025	2.242	1.230-4.086	0.008
Smoking	1.769	1.024-3.056	0.041
Drinking	1.063	0.607-1.862	0.831
BMI	0.892	0.823-0.966	0.005	0.888	0.816-0.968	0.007
SBP	1.010	0.996-1.024	0.169
DBP	0.987	0.967-1.008	0.235
TG	0.872	0.734-1.037	0.122
LDL-C	0.717	0.544-0.946	0.019
TC	0.655	0.519-0.828	< 0.001
HDL-C	0.377	0.160-0.889	0.026
Fast glucose	1.217	1.092-1.357	< 0.001
HbA1c	1.589	1.293-1.953	< 0.001	1.587	1.283-1.963	< 0.001
WBC	1.026	0.907-1.160	0.688
NEU	1.075	0.942-1.227	0.282
LYM	0.724	0.489-1.074	0.108
MONO	1.726	0.324-9.180	0.522
PLT	0.995	0.990-1.000	0.033
Time since diagnose T2DM
< 5			0.010
5-10	1.987	0.850-4.643	0.113
> 10	2.363	1.338-4.172	0.003
LAD-PCATa	1.230	1.159-1.305	< 0.001	1.246	1.168-1.329	< 0.001
LCX-PCATa	1.205	1.140-1.274	< 0.001	1.207	1.138-1.281	< 0.001
RCA-PCATa	1.172	1.113-1.234	< 0.001	1.164	1.105-1.227	< 0.001

BMI: Body mass index; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; TG: Triglycerides; LDL-C: Low-density lipoprotein cholesterol; HDL-C: High-density lipoprotein cholesterol; HbA1c: Glycated hemoglobin; WBC: White blood cell; NEU: Neutrophil; LYM: Lymphocyte; MONO: Monocyte; PLT: Platelet; LAD: Left anterior descending artery; PCATa: Pericoronary adipose tissue attenuation; LCX: Left circumflex artery; RCA: Right coronary artery; OR: Odds ratio; CI: Confidence interval.

ROC curve analysis and diagnostic performance evaluation

Following the ROC curve analysis in LAD, LCX, and RCA, the cutoff values of LAD (-79.38 HU), LCX (-71.27 HU), and RCA (-81.52 HU) assisted in discriminating between the INOCA and non-INOCA cases (Figure 2). The AUCs for PCATa in the LAD (0.809), LCX (0.777), and RCA (0.758) were higher than the individual clinical risk factors (0.731), although the differences did not reach statistical significance. When the clinical risk factors were combined with the PCATa values from these three vessels, the diagnostic performance was further improved (clinical + LAD: 0.851; clinical + LCX: 0.842; clinical + RCA: 0.841, all P < 0.05). This result indicates that combining the imaging and clinical parameters enhances the ability to identify INOCA in T2DM patients, as compared to either indicator alone. Table 4 summarizes the AUC, sensitivity, and specificity values of each parameter. We conducted NRI and IDI analyses to assess the added diagnostic value of PCATa. The inclusion of PCATa substantially improved risk classification for INOCA in patients with T2DM: LAD-PCATa: NRI = 0.59 (95%CI: 0.38-0.8, P < 0.001) and IDI = 0.14 (95%CI: 0.10-0.18, P < 0.001); LCX-PCATa: NRI = 0.41 (95%CI: 0.20-0.62, P < 0.001) and IDI = 0.07 (95%CI: 0.04-0.09, P < 0.001); RCA-PCATa: NRI = 0.34 (95%CI: 0.16-0.56, P < 0.001) and IDI = 0.07 (95%CI: 0.04-0.10, P < 0.001). These findings indicate that PCATa provides incremental discriminative power for identifying INOCA beyond conventional clinical risk factors (Table 5).

Figure 2 Receiver operating characteristic curves for identifying ischemia with non-obstructive coronary arteries in patients with type 2 diabetes mellitus. A: Clinical variables vs left anterior descending artery-pericoronary adipose tissue attenuation and their combination; B: Clinical variables vs left circumflex-pericoronary adipose tissue attenuation and their combination; C: Clinical variables vs right coronary artery-pericoronary adipose tissue attenuation and their combination. The diagnostic performance was assessed by comparing the area under the curve for each parameter. Combined models demonstrated improved diagnostic accuracy compared with individual indicators. LAD: Left anterior descending artery; PCATa: Pericoronary adipose tissue attenuation; LCX: Left circumflex; RCA: Right coronary artery.

Table 4 Diagnostic performance of the clinical and imaging parameters in distinguishing ischemia with non-obstructive coronary artery disease among type 2 diabetes mellitus patients based on the receiver operating characteristic curve analysis.

	Cutoff	AUC	95%CI	SEN	SEP
Clinical factors		0.731	0.666-0.795	0.745	0.644
LAD PCATa	-79.38	0.809	0.755-0.864	0.836	0.610
LCX PCATa	-71.27	0.777	0.717-0.836	0.736	0.720
RCA PCATa	-81.52	0.758	0.698-0.819	0.900	0.500
Clinical factors + LAD PCATa		0.851	0.802-0.900	0.755	0.805
Clinical factors + LCX PCATa		0.842	0.792-0.909	0.691	0.839
Clinical factors + RCA PCATa		0.841	0.792-0.890	0.655	0.839

LAD: Left anterior descending artery; PCATa: Pericoronary adipose tissue attenuation; LCX: Left circumflex artery; RCA: Right coronary artery; AUC: Area under the curve; CI: Confidence interval; SEN: Sensitivity; SEP: Specificity.

Table 5 Improvement in risk classification by pericoronary adipose tissue attenuation for identifying ischemia with non-obstructive coronary arteries in patients with type 2 diabetes mellitus.

	NRI (95%CI)	P value	IDI (95%CI)	P value
LAD PCATa	0.59 (0.38-0.80)	< 0.001	0.14 (0.10-0.18)	< 0.001
LCX PCATa	0.41 (0.20-0.62)	< 0.001	0.07 (0.04-0.09)	< 0.001
RCA PCATa	0.34 (0.16-0.56)	< 0.001	0.07 (0.04-0.10)	< 0.001

LAD: Left anterior descending artery; PCATa: Pericoronary adipose tissue attenuation; LCX: Left circumflex artery; RCA: Right coronary artery; NRI: Net reclassification improvement; CI: Confidence interval; IDI: Integrated discrimination improvement.

DISCUSSION

The present study that enrolled 228 participants evaluated the differences in PCATa at the proximal segments of the three major coronary branches using CCTA between T2DM patients and healthy controls. Furthermore, T2DM patients were stratified based on the presence or absence of concomitant INOCA to investigate the potential of PCATa and clinical indicators for early risk identification. Our study results demonstrated that PCATa was significantly higher in T2DM patients than in healthy controls and was further elevated in those with concomitant INOCA, as compared to those without INOCA. The subsequent diagnostic performance analysis indicated that PCATa outperformed the conventional clinical risk factors in distinguishing INOCA. Additionally, the integration of imaging and clinical parameters led to further improvements in diagnostic accuracy.

Given the high incidence of cardiovascular disease in T2DM patients, routine CCTA screening has become a clinical consensus, especially for those with an elevated cardiovascular risk[20]. The current guidelines recommend an active intervention in patients with coronary artery stenosis of > 50%, as identified by CCTA[21]. However, the cases with < 50% stenosis are frequently overlooked in clinical settings. Despite the absence of significant luminal obstruction, these patients remain at an increased risk of developing cardiovascular complications[22]. In such cases, conventional imaging markers (such as fractional flow reserve or plaque volume) may exhibit limited sensitivity in detecting early-stage or inflammation-driven disease processes, thereby impeding timely risk identification and management. The present study highlights the central role of inflammation in the pathogenesis of INOCA. In particular, inflammation-induced hypercontraction of the vascular smooth muscle cells and coronary microvascular spasms are recognized as key contributors to its development[23,24]. As an imaging biomarker of vascular inflammation, pericoronary adipose tissue offers a dynamic reflection of pathological changes within the coronary arteries. An elevated PCATa has been identified as a sensitive indicator of localized inflammatory activity[25], and its increase may precede the formation of plaques or onset of luminal stenosis detectable by conventional imaging techniques.

Our study demonstrated that the PCATa values of the three major coronary branches were significantly elevated in T2DM patients, as compared to normal controls, indicating a persistent inflammatory response in the pericoronary region. Previous studies have confirmed that T2DM is characterized not only by systemic chronic inflammation but also by localized inflammatory responses within the coronary arteries[26]. Notably, the severity of coronary endothelial dysfunction and inflammation has been shown to correlate positively with blood glucose levels[26], which aligns with our study findings. Furthermore, our subgroup analysis revealed that the PCATa values of the three major coronary branches were further elevated in T2DM patients with concomitant INOCA. This finding indicates the presence of a more pronounced pericoronary inflammatory response in this subgroup, which is possibly associated with inflammation-mediated mechanisms underlying the development of INOCA. PCAT, functioning as a dynamic endocrine and paracrine tissue[27], releases various proinflammatory mediators in response to inflammatory stimuli. These mediators infiltrate the adjacent vascular wall and contribute to endothelial dysfunction[28], subsequently altering the water-lipid composition of PCAT. Such compositional changes are reflected as increased CT attenuation on imaging[29]. Building upon this, the hyperglycemic metabolic state in T2DM patients leads to the accumulation of advanced glycation end-products, which bind to their receptors and chronically activate inflammatory signaling pathways within PCAT[30]. Under these inflammatory conditions, pericoronary adipose tissue secretes interleukin-1β, which elevates the intracellular calcium levels in the vascular smooth muscle cells, thereby enhancing their contractile responsiveness to vasoconstrictive stimuli. Simultaneously, diminished nitric oxide bioavailability impairs endothelial vasodilation. Together, these pathological alterations disrupt the coronary vasomotor balance and ultimately lead to the onset of INOCA[24,31]. Therefore, our findings suggest that elevated PCATa values not only reflects the degree of coronary artery inflammation in T2DM patients, but also provides an imaging basis for early screening of INOCA in T2DM patients.

Notably, recent pharmacological treatments may affect PCATa values; however, we conducted a sensitivity analysis to assess the extent of this influence. Specifically, statins, antiplatelet agents, and insulin, which are commonly prescribed drugs with known anti-inflammatory and metabolic effects, were included as covariates in the multivariate regression models. After adjustment, the association between PCATa and INOCA remained statistically significant across all three major coronary arteries. These findings indicate that pharmacological treatment did not substantially affect the diagnostic utility of PCATa in this patient population. This supports the robustness of our results and highlights the independent contribution of perivascular adipose tissue inflammation to the pathophysiology of INOCA in individuals with T2DM.

Additionally, the present study further screened relevant clinical risk factors and imaging parameters. Subsequently, the ROC curve analysis was conducted to assess the diagnostic value of these indicators for the early identification of INOCA in T2DM patients. Our results demonstrated that the ROC curve based on the PCATa values of the three major coronary branches yielded a higher AUC value than that based on clinical risk factors alone, although the difference was not statistically significant. Notably, the ROC curve generated by combining the PCATa values and clinical risk factors showed significantly greater diagnostic performance as compared to either parameter used independently. These findings suggest that PCATa, as a noninvasive indicator of coronary inflammation, when combined with conventional clinical parameters, may offer a more comprehensive assessment of INOCA risk in T2DM patients. The combined use of multiple high-risk indicators appears to enhance the diagnostic accuracy. Interestingly, during data analysis, we observed that the INOCA group exhibited significantly lower lipid-related indices and BMI than the non-INOCA group (P < 0.05). Further analysis revealed a negative association between elevated lipid levels and lower BMI with the occurrence of INOCA. One possible explanation for this phenomenon is the higher prevalence of statin use in the INOCA group, which may have influenced the lipid metabolism and contributed to the observed findings[32].

Although the present study demonstrates the potential value of PCATa in identifying INOCA among patients with T2DM, several limitations should be acknowledged. Firstly, this was a single-center retrospective study without external validation. Despite rigorous design, inherent limitations remain, and future studies are warranted to include multicenter cohorts with external validation to strengthen generalizability. Second, PCATa was measured only in the proximal segments of the three main coronary arteries, and correction was applied only for tube voltage without accounting for the potential influence of reconstruction algorithms, which may not fully capture regional heterogeneity in perivascular inflammation. Third, this study primarily focused on total plaque volume without detailed analysis of plaque characteristics, such as peri-plaque PCATa, plaque composition, or vulnerability features. Although prior studies suggest that peri-plaque PCATa may be more sensitive for detecting focal coronary inflammation, most INOCA patients in our cohort had no visible plaques or only a low plaque burden, which limited the feasibility and reliability of plaque-specific analyses. Fourth, our ROC analysis revealed good diagnostic performance of PCATa; however, the derived cutoff values remain unvalidated in an independent external cohort, which limits their current clinical applicability. Future studies should validate these cutoff values in independent external cohorts to improve their clinical applicability. Fifth, assessing the incremental value of PCATa over traditional biomarkers, such as high-sensitivity C-reactive protein and coronary flow reserve, would further improve translational relevance; however, such data were not routinely available in our study due to limited laboratory and echocardiographic testing.

CONCLUSION

Our study found that PCATa has a significant value in identifying the risk of INOCA among T2DM patients. The integration of PCATa with conventional clinical indicators further enhanced the diagnostic performance. These findings may assist clinicians in the early identification of high-risk INOCA patients in the T2DM population and support more timely and targeted clinical management decisions.