Abstract
Background
Coronary microvascular dysfunction (CMD) contributes to myocardial ischemia in patients with angina and non-obstructive coronary artery disease. Quantitative positron emission tomography (PET) allows noninvasive assessment of myocardial blood flow and identification of CMD.
Objectives
To investigate the association between PET-assessed coronary microvascular dysfunction (CMD), coronary plaque burden, and vascular remodeling, integrating microvascular resistance reserve (MRR).
Methods
This post hoc analysis of the PACIFIC-1 trial included coronary arteries from symptomatic patients with suspected stable CAD who underwent [15O]water PET, coronary CT angiography (CTA), and invasive fractional flow reserve (FFR). Coronary arteries were categorized into three groups: non-ischemic vessels (normal FFR and hyperemic myocardial blood flow [hMBF], n = 352), CMD vessels (normal FFR but impaired hMBF, n = 88), and epicardial flow-limiting vessels (abnormal FFR, n = 159). MRR was calculated by integrating PET-derived coronary flow reserve with invasive FFR. AI-based CTA quantified plaque burden and remodeling.
Results
Diameter stenosis and plaque burden increased progressively from non-ischemic to CMD to epicardial flow-limiting vessels (all P < 0.05). CMD vessels demonstrated the largest lumen volume (583.75 mm3 [IQR 337.02–754.30]) and lumen area (4.63 mm2 [IQR 3.77–6.67]) among the three groups (both P < 0.05). In non-ischemic vessels, noncalcified plaque burden was associated with worse microvascular function detected by MRR (B = -0.12, 95%CI [-0.21, -0.04], P = 0.006). In CMD vessels, remodeling index was independently associated with higher MRR after adjustment (B = 0.10, 95%CI [0.05, 0.15], P < 0.001).
Conclusions
CMD is associated with an intermediate atherosclerotic burden and paradoxical lumen enlargement, representing a distinct functional-structural phenotype defined by PET-derived microvascular dysfunction and CT-based plaque characteristics.
Clinical Trial Registration:
URL: https://www.clinicaltrials.gov; Unique identifier: NCT01521468.
Graphical Abstract
Abbreviations: CMD, coronary microvascular dysfunction; FFR, fractional flow reserve; MBF, myocardial blood flow; MRR, microvascular resistance reserve; PET, positron emission tomography-computed tomography.
Graphical abstract. Abbreviations: CMD, coronary microvascular dysfunction; FFR, fractional flow reserve; MBF, myocardial blood flow; MRR, microvascular resistance reserve; PET, positron emission tomography-computed tomography
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1007/s00259-026-08054-3.
Keywords: Coronary microvascular dysfunction, Coronary artery disease, Microvascular resistance reserve, Vascular remodeling, Plaque quantification
Introduction
It is now well established that chronic coronary syndromes (CCS) with ischemia frequently occur in the absence of flow-limiting epicardial stenosis. Approximately 50% of patients referred for coronary angiography with angina do not exhibit obstructive coronary artery disease (CAD) [1–3]. This clinical scenario, now increasingly recognized as angina with non-obstructive coronary arteries (ANOCA), has received growing attention in the recent European Society of Cardiology guideline on CCS, which emphasize the need for its proper diagnosis and management [4].
Coronary microvascular disease (CMD) represents one of the key mechanisms underlying ANOCA, contributing to myocardial ischemia and adverse cardiovascular outcomes despite angiographically normal or near-normal coronary arteries [5, 6]. Although CMD has traditionally been viewed as a distinct pathological entity separate from epicardial coronary artery atherosclerosis, emerging evidence suggests that the boundary between these two processes may be more complex than previously considered [7]. Recent studies indicate a potential interplay between early epicardial plaque formation and microvascular dysfunction, pointing to a pro-atherogenic environment within the epicardial vessels of patients with CMD [8].
Advances in quantitative PET imaging, combined with invasive physiological indices, have enabled comprehensive assessment of both macrovascular and microvascular coronary function [10–12]. In this framework, PET provides the quantitative basis for myocardial blood flow assessment, while microvascular resistance reserve (MRR) extends this approach by integrating invasive pressure measurements into an epicardial flow-independent index of microvascular function [13].
In this study, we sought to investigate the relationship between epicardial atherosclerotic characteristics by quantitative coronary CT and CMD by integrating PET and invasive fractional flow reserve (FFR). Our aim was to improve understanding of the epicardial vascular responses associated with microvascular dysfunction and to explore their potential implications for the early development of CAD.
Methods
Study population
The current study is a post hoc analysis of the Prospective Comparison of Cardiac PET/CT, SPECT/CT Perfusion Imaging and CT Coronary Angiography With Invasive Coronary Angiography (PACIFIC-1) trial. The PACIFIC-1 trial was a prospective controlled clinical single-center trial conducted from January, 2012, to October, 2014, at the Free University Medical Center, Amsterdam, the Netherlands (NCT01521468) [14]. All participants (n = 208), with suspected obstructive CAD, completed a 2-week protocol including [15O]water PET and coronary computed tomography angiography (CCTA) prior to invasive coronary angiography (ICA) in combination with routine three-vessel invasive FFR examination. All index tests were interpreted blindly by core laboratories. The study protocol was approved by the Medical Ethics Committee of the Free University Medical Center and all participants provided written informed consent.
PET procedure
PET scans were performed using a hybrid PET/CT equipment (Philips Gemini TF 64, Philips Healthcare, Best, the Netherlands). The scan procedure involved a dynamic 6 min scanning protocol that initiated simultaneously with an injection of 370 MBq [15O]water PET during resting and adenosine-induced hyperemic conditions (140 µg/kg/min). Low-dose CT scans were performed for attenuation correction. Parametric MBF images were created and quantitatively analyzed using Carimas software (Turku PET Centre, University of Turku and Turku University Hospital, Turku, Finland). MBF was expressed in mL/min/g of perfusable tissue and analyzed according to the 17 segments of the left ventricle as per the American Heart Association model with standardized allocation of segments to the three vascular territories [15]. Regional hyperemic myocardial blood flow (hMBF) was defined as mean hMBF of the entire vascular territory in the absence of a perfusion defect or as the mean hMBF of the perfusion defect (≥ 2 adjacent segments with a hMBF ≤ 2.3 mL/min/g) when present [16]. Coronary flow reserve (CFR) was defined as the ratio of hMBF to resting MBF (rMBF). Coronary arteries were categorized into three groups: non-ischemic vessels (normal FFR and normal hMBF), CMD vessels (normal FFR but impaired hMBF), and epicardial flow-limiting vessels (abnormal FFR). In this framework, CMD vessels were defined as non-flow-limiting epicardial vessels subtending myocardial territories with PET evidence of impaired hyperemic perfusion.
ICA and physiological assessments
ICA was performed according to standard clinical protocols [14]. All major coronary arteries (> 2 mm) were routinely interrogated by FFR, irrespective of stenosis severity and imaging results. To induce maximal coronary hyperemia, adenosine was administered intracoronary as a 150 μg bolus or intravenously (140 μg/kg/min). FFR was calculated as the ratio of mean distal intracoronary to aortic guiding pressure during hyperemia [17]. An FFR ≤ 0.80 was graded as abnormal. Vessels in which FFR was not required due to stenosis severity were considered as ischemic as well. MRR was derived based on the framework by De Bruyne et al. by combining invasive FFR measurements and noninvasive PET flow measurements [13]. The formula used in the current analysis is a quotient of CFR and FFR with the correction for the impact of hemodynamics, as follows:
 |
where CFR indicates PET-derived CFR, FFR indicates pressure-wire derived FFR, and Pa,rest and Pa,hyper indicate mean aortic pressure during non-hyperemic and maximal hyperemic PET, respectively. Based on prior research [18, 19], we adopted a prognostically significant threshold of MRR ≤ 3.0 as a second marker for CMD.
Coronary CTA acquisition
CT scans were performed using a 256-slice CT scanner (Philips Brilliance iCT, Philips Healthcare, Best, the Netherlands) with a tube current between 200 and 360 mAs at 120 kV. Respective CT parameters for the 256-slice entailed a section collimation of 128 × 0.625 mm and a gantry rotation time of 270 ms. Each patient received 800 mcg of sublingual nitroglycerine immediately prior to CCTA. If necessary, metoprolol was given prior to CCTA by oral or intravenous administration. For visualization of the coronary artery lumen, a bolus of iobitridol (Xenetix 350) was injected intravenously followed immediately by a saline chaser. Prospective electrocardiogram gating between 72 and 78% of the R-R interval was performed to reduce radiation dose [14, 20].
Artificial intelligence-guided analysis of quantitative computed tomography (AI-QCT) from coronary CTA
Coronary CTA scans were analyzed using the previously described AI-QCT algorithm [21, 22]. The quantitative coronary CTA evaluation was performed blinded to the clinical and outcomes data. This Food and Drug Administration (FDA)-cleared software uses a series of validated convolutional neural networks (3-dimensional U-Net and Visual Geometry Group [VGG] network variants) for image-quality assessment, coronary artery segmentation and labeling, lumen wall evaluation and vessel contour determination, and plaque characterization.
Total plaque volume (TPV) (mm3) was defined as the sum of all plaque volumes calculated for each coronary artery lesion among all vessels. Noncalcified plaque (NCP) was defined as a lesion with a CT density between 30 and 350 HU, and calcified plaque (CP) above 350 HU [23]. The percentage of atheroma volume (PAV) was calculated by dividing plaque volume (PV)/vessel volume × 100%. Positive remodeling was defined by a remodeling index of ≥ 1.1, which was calculated by the vessel size at the site of the maximal remodeling divided by the vessel size of the reference site [24].
Statistical analysis
Continuous variables were described as mean ± standard deviation or median (interquartile range), and compared between the groups using either 2 sample t-test or nonparametric Wilcoxon rank sum test. Categorical variables were summarized as counts (percentages) and compared using the Pearson chi-square test or Fisher exact test if cell frequencies were insufficient. Bonferroni adjustment was applied to account for multiple testing.
All analyses were performed at the vessel level. For the primary analyses, 599 vessels with complete hMBF data were included. A subset of 508 vessels, with complete data for FFR, hMBF, and MRR, was used for analyses involving MRR. A detailed flowchart outlining vessel inclusion and exclusion is presented in Fig. 1. To correct for cardiovascular risk factors and for multiple vessels within a patient, generalized estimating equations (GEEs) were used to assess the relation between risk factors and vessel-specific MRR. A two-sided P < 0.05 was considered statistically significant. Significant predictors of vessel-specific MRR in the univariable analysis were included into the multivariable model. Sensitivity analyses were performed adjusting for alternative plaque composition parameters and for coronary vessel type. Statistical analysis was performed using SPSS version 22.0 (SPSS, IL, USA), GraphPad Prism 8.0 (GraphPad Software, USA), MedCalc 19.0.4 (MedCalc Software, Belgium) and R version 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria).
Fig. 1.
Flow chart. Abbreviations: FFR, fractional flow reserve; MBF, myocardial blood flow; MRR, microvascular resistance reserve; PET, positron emission tomography
Results
Clinical characteristics
As shown in Table 1, a total of 599 vessels with complete FFR and hMBF data were included, comprising 352 (58.8%) non-ischemic vessels, 88 (14.7%) with CMD, and 159 (26.5%) with epicardial flow-limiting disease. A higher prevalence of male sex, diabetes, hyperlipidemia and smoking history was found in CMD vessels and epicardial flow-limiting vessels compared with non-ischemic vessels (all P < 0.05). Medication use was generally comparable across groups, with the exception of significantly higher use of aspirin (epicardial flow-limiting: 93.1% [148/159] vs. non-ischemic: 84.7% [298/352]; P = 0.009) and statin (epicardial flow-limiting: 86.2% [137/159] vs. non-ischemic: 73.0% [257/352]; P = 0.001) in the epicardial flow-limiting group.
Table 1.
Vessel-level clinical characteristics
| Characteristics | Overall (n = 599) |
Non-ischemic (n = 352) |
CMD (n = 88) |
Epicardial flow-limiting (n = 159) |
P-value |
|---|---|---|---|---|---|
| Age, years | 59.00 [52.00, 64.00] | 58.00 [51.00, 64.00] | 60.00 [52.00, 66.00] | 60.00 [53.00, 65.00] | 0.580 |
| Male, % | 376 (62.8) | 164 (46.6) | 74 (84.1) | 138 (86.8) | < 0.001ac |
| BMI, kg/m2 | 26.84 [24.36, 29.34] | 26.54 [23.74, 29.05] | 26.70 [24.62, 29.05] | 27.17 [24.90, 29.64] | 0.134a |
| Diabetes mellitus, % | 94 (15.7) | 44 (12.5) | 17 (19.3) | 33 (20.8) | 0.036 |
| Hypertension, % | 281 (46.9) | 166 (47.2) | 41 (46.6) | 74 (46.5) | 0.990 |
| Hyperlipidemia, % | 237 (39.6) | 121 (34.4) | 36 (40.9) | 80 (50.3) | 0.003c |
| Smoking history, % | 294 (49.1) | 152 (43.2) | 52 (59.1) | 90 (56.6) | 0.002ac |
| Family history of CAD, % | 305 (50.9) | 183 (52.0) | 42 (47.7) | 80 (50.3) | 0.762 |
| Symptoms, % | < 0.001ac | ||||
| Typical angina | 210 (35.1) | 99 (28.1) | 46 (52.3) | 65 (40.9) | |
| Atypical angina | 222 (37.1) | 145 (41.2) | 27 (30.7) | 50 (31.4) | |
| Aspecific chest pain | 167 (27.9) | 108 (30.7) | 15 (17.0) | 44 (27.7) | |
| Medication | |||||
| Aspirin, % | 525 (87.6) | 298 (84.7) | 79 (89.8) | 148 (93.1) | 0.022c |
| Statin, % | 465 (77.6) | 257 (73.0) | 71 (80.7) | 137 (86.2) | 0.003c |
| ACEI/ARB, % | 226 (37.7) | 134 (38.1) | 35 (39.8) | 57 (35.8) | 0.813 |
| Beta blocker, % | 394 (65.8) | 219 (62.2) | 62 (70.5) | 113 (71.1) | 0.090 |
| CCB, % | 177 (29.5) | 109 (31.0) | 18 (20.5) | 50 (31.4) | 0.128 |
| Long-acting nitrates, % | 64 (10.7) | 44 (12.5) | 9 (10.3) | 11 (7.0) | 0.180 |
Values are n (%) or median [IQR]
a indicates a significant difference between non-ischemic vessels and CMD vessels (Bonferroni-adjusted P < 0.05)
b indicates a significant difference between CMD vessels and epicardial flow-limiting vessels (Bonferroni-adjusted P < 0.05)
c indicates a significant difference between non-ischemic vessels and epicardial flow-limiting vessels (Bonferroni-adjusted P < 0.05)
ACEI/ARB angiotensin converting enzyme inhibitor/angiotensin receptor blocker; BMI body mass index; CAD coronary artery disease; CCB calcium channel blocker; CMD coronary microvascular dysfunction
Imaging characteristics
Atherosclerotic burden increased progressively across non-ischemic, CMD, and epicardial flow-limiting vessels. This pattern was reflected in diameter stenosis (non-ischemic: 8.00%, CMD: 18.50%, epicardial flow-limiting: 67.00%; P < 0.001), plaque volume (non-ischemic: 12.60 mm3 [IQR 2.88–59.30], CMD: 50.40 mm3 [IQR 12.20–139.85], epicardial flow-limiting: 167.10 mm3 [IQR 80.05–269.65]; P < 0.001), and PAV (non-ischemic: 3.01% [IQR 0.77–11.89], CMD: 12.52% [IQR 3.06–20.52], epicardial flow-limiting: 28.17% [IQR 15.35–41.36]; P < 0.001) (Table 2 and Fig. 2).
Table 2.
Vessel-level imaging characteristics
| Characteristics | Overall (n = 599) |
Non-ischemic (n = 352) |
CMD (n = 88) |
Epicardial flow-limiting (n = 159) |
P-value |
|---|---|---|---|---|---|
| Baseline MBF | 1.00 [0.90, 1.30] | 1.10 [0.90, 1.30] | 0.90 [0.80, 1.10] | 0.90 [0.80, 1.10] | < 0.001ac |
| Hyperemic MBF | 3.10 [2.00, 4.00] | 3.80 [3.20, 4.40] | 1.90 [1.67, 2.10] | 1.80 [1.30, 2.10] | < 0.001ac |
| Diameter stenosis, % | 17.00 [4.00, 50.00] | 8.00 [0.00, 21.00] | 18.50 [7.00, 41.00] | 67.00 [38.50, 78.00] | < 0.001abc |
| Vessel volume, mm3 | 531.60 [359.00, 763.65] | 486.40 [305.20, 685.45] | 733.95 [379.25, 912.60] | 599.50 [447.25, 775.85] | < 0.001ac |
| Lumen volume, mm3 | 436.90 [282.40, 626.55] | 439.60 [279.78, 637.85] | 583.75 [337.02, 754.30] | 400.00 [273.90, 523.45] | < 0.001ab |
| Average vessel area, mm2 | 4.77 [3.70, 6.36] | 4.45 [3.47, 6.15] | 5.28 [4.28, 7.69] | 5.02 [4.15, 6.16] | < 0.001ac |
| Average lumen area, mm2 | 3.77 [2.95, 5.43] | 4.00 [3.12, 5.68] | 4.63 [3.77, 6.67] | 3.19 [2.71, 3.80] | < 0.001abc |
| Remodeling index | 1.30 [1.20, 1.50] | 1.20 [1.10, 1.40] | 1.30 [1.20, 1.42] | 1.50 [1.30, 1.60] | < 0.001bc |
| Positive remodeling, % | 162 (27.0) | 56 (15.9) | 22 (25.0) | 84 (52.8) | < 0.001bc |
| Fractional flow reserve | 0.92 [0.78, 0.97] | 0.95 [0.91, 0.99] | 0.94 [0.89, 0.97] | 0.70 [0.53, 0.76] | < 0.001abc |
| Plaque volume, mm3 | |||||
| Total plaque | 38.60 [6.45, 139.65] | 12.60 [2.88, 59.30] | 50.40 [12.20, 139.85] | 167.10 [80.05, 269.65] | < 0.001abc |
| Calcified plaque | 10.60 [0.00, 56.15] | 0.45 [0.00, 22.25] | 11.25 [0.00, 54.27] | 57.70 [21.15, 128.90] | < 0.001abc |
| Non-calcified plaque | 22.90 [5.20, 70.40] | 8.90 [2.80, 32.28] | 36.60 [11.30, 81.23] | 81.80 [50.20, 122.75] | < 0.001abc |
| Percentage atheroma volume, % | |||||
| Total plaque | 8.77 [1.46, 22.60] | 3.01 [0.77, 11.89] | 12.52 [3.06, 20.52] | 28.17 [15.35, 41.36] | < 0.001abc |
| Calcified plaque | 2.21 [0.00, 9.26] | 0.07 [0.00, 4.56] | 2.27 [0.00, 9.56] | 10.17 [3.80, 22.70] | < 0.001abc |
| Non-calcified plaque | 4.58 [1.35, 11.44] | 2.24 [0.76, 6.30] | 7.01 [2.17, 11.09] | 13.53 [8.76, 20.76] | < 0.001abc |
Values are n (%) or median [IQR]
a indicates a significant difference between non-ischemic vessels and CMD vessels (Bonferroni-adjusted P < 0.05)
b indicates a significant difference between CMD vessels and epicardial flow-limiting vessels (Bonferroni-adjusted P < 0.05)
c indicates a significant difference between non-ischemic vessels and epicardial flow-limiting vessels (Bonferroni-adjusted P < 0.05)
CMD coronary microvascular dysfunction
Fig. 2.
The comparison of stenosis severity, plaque burden and vessel size between the groups. The box indicates the interquartile range (25th to 75th percentiles), and the central line marks the median. The whiskers extend to the minimum and maximum values. Abbreviations: CMD, coronary microvascular dysfunction; PAV, percentage atheroma volume
CMD vessels had the largest lumen volume (583.75 mm3 [IQR 337.02–754.30]) and average lumen area (4.63 mm2 [IQR 3.77–6.67]) among the three groups (both P < 0.05). In contrast, vessel volume (CMD: 733.95 mm3 [IQR 379.25–912.60] vs. epicardial flow-limiting: 599.50 mm3 [IQR 447.25–775.85]; P = 0.263) and average vessel area (CMD: 5.28 mm2 [IQR 4.28–7.69] vs. epicardial flow-limiting: 5.02 mm2 [IQR 4.15–6.16]; P = 0.141) were similar between CMD and epicardial flow-limiting vessels, but both were significantly larger than in non-ischemic vessels (all P < 0.05) (Fig. 2). In accordance, the prevalence and extent of positive remodeling showed a stepwise increase across the three groups (Table 2).
Distribution of MRR
A total of 508 vessels with complete FFR, hMBF, and MRR data were included in the MRR-related analysis, comprising 329 (64.8%) non-ischemic vessels, 83 (16.3%) CMD vessels, and 96 (18.9%) epicardial flow-limiting vessels. As shown in Fig. 3, CMD vessels demonstrated a significantly lower MRR (2.22 [IQR 1.71–2.85]) compared with both non-ischemic vessels (3.86 [IQR 3.26–4.67]) and epicardial flow-limiting vessels (3.40 [IQR 2.62–4.43]) (P < 0.001 for both comparisons), with 79.5% of CMD vessels exhibiting an MRR ≤ 3.00.
Fig. 3.
The comparison of MRR between the groups. The results were calculated in the 508 coronary vessels with full FFR data. The box indicates the interquartile range (25th to 75th percentiles), and the central line marks the median. The whiskers extend to the minimum and maximum values. The grey dashed line indicates an MRR of 3. Abbreviations: CMD, coronary microvascular dysfunction; MRR, microvascular resistance reserve
Predictors of MRR
In the CMD group (Table 3), univariable analysis showed that remodeling index (B = 0.12, 95% CI [0.07, 0.17], P < 0.001), total PAV, PAV-CP, and PAV-NCP were associated with MRR. After adjustment for PAV-NCP, higher remodeling index (per 10% increase) remained significantly associated with higher MRR (B = 0.10, 95% CI [0.05, 0.15], P < 0.001). Additional multivariable analyses adjusting separately for total PAV, PAV-CP, and vessel type are presented in Tables S3 and S4, and consistently demonstrated an independent association between higher remodeling index and higher MRR (all P < 0.05).
Table 3.
Prediction of MRR in coronary microvascular dysfunction vessels
| Parameters | Univariable | Multivariable | ||
|---|---|---|---|---|
| B (95%CI) | P-value | B (95%CI) | P-value | |
| Age, per 1 year increase | −0.02 (−0.04, 0.00) | 0.114 | ||
| Male | 0.33 (−0.09, 0.74) | 0.121 | ||
| BMI, per 1 kg/m2 increase | 0.02 (−0.03, 0.08) | 0.426 | ||
| Diabetes mellitus | −0.13 (−0.62, 0.35) | 0.593 | ||
| Hypertension | −0.17 (−0.58, 0.25) | 0.424 | ||
| Hyperlipidemia | 0.22 (−0.19, 0.62) | 0.293 | ||
| Smoking history | 0.04 (−0.36, 0.44) | 0.851 | ||
| Family history of CAD | 0.07 (−0.35, 0.49) | 0.734 | ||
| Symptoms | ||||
| Typical angina | Reference | |||
| Atypical angina | 0.39 (−0.16, 0.94) | 0.162 | ||
| Aspecific chest pain | −0.01 (−0.46, 0.45) | 0.982 | ||
| Imaging characteristics | ||||
| Diameter stenosis, per 5% increase | 0.02 (−0.01, 0.05) | 0.119 | ||
| Diameter stenosis ≥ 50% | 0.16 (−0.14, 0.45) | 0.301 | ||
| Remodeling index, per 10% increase | 0.12 (0.07, 0.17) | < 0.001 | 0.10 (0.05, 0.15) | < 0.001 |
| Average lumen area, per 1 mm2 increase | −0.03 (−0.07, 0.00) | 0.062 | ||
| Percentage atheroma volume, per 5% increase | ||||
| Total plaque | 0.09 (0.03, 0.15) | 0.002 | ||
| Calcified plaque | 0.11 (0.02, 0.20) | 0.016 | ||
| Non-calcified plaque | 0.18 (0.03, 0.33) | 0.021 | 0.10 (−0.04, 0.24) | 0.151 |
BMI body mass index; CAD coronary artery disease; CI confidence interval
In the non-ischemic vessel group, only total PAV and PAV-NCP were significantly associated with MRR, both showing a negative correlation (Table S1). In the epicardial flow-limiting group, the presence of hypertension (B = −0.80, 95% CI [−1.36, −0.23], P = 0.006), remodeling index, and PAV-NCP were associated with lower MRR in univariable analysis. After adjustment, only hypertension remained independently associated with reduced MRR (B = −0.73, 95% CI [−1.26, −0.21], P = 0.007) (Table S2).
Discussion
This study investigated coronary physiology and morphology across three groups of vessels stratified by [15O]water PET-derived hyperemic myocardial blood flow and invasive FFR. We found that vessels with coronary microvascular dysfunction exhibited an intermediate degree of atherosclerosis, greater than non-ischemic vessels but lower than flow-limiting epicardial disease. At the same time, CMD vessels demonstrated increased vessel and lumen size, consistent with positive remodeling. In contrast, vessels with flow-limiting disease showed comparable vessel size but the smallest lumen area, reflecting luminal encroachment. Within this framework, CMD may represent a distinct structural–functional phenotype characterized by the coexistence of microvascular dysfunction, plaque accumulation, and adaptive vascular remodeling.
Potential mechanisms
Our data extend these observations by demonstrating that CMD vessels on CT exhibit a distinctive structural–functional profile. Among the three groups, CMD vessels had the largest lumen volume and area indexed by vessel length, indicating pronounced positive remodeling, and significantly greater atherosclerotic burden (higher diameter stenosis, plaque volume, and plaque area volume) compared to non-ischemic vessels [25–27]. This pattern suggests that CMD is not merely an indistinct intermediate but may reflect a phenotype in which microvascular dysfunction, epicardial plaque accumulation and compensatory arterial changes coexist.
At an earlier stage, even vessels classified as “non-ischemic” already showed an association between higher plaque burden and lower MRR (Table S1), consistent with previous studies linking early atherosclerosis to microvascular impairment [28, 29]. Interestingly, remodeling index was found related to higher MRR in CMD vessel, which might possibly indicate an adaptive stage during which the epicardial vessels can still positively remodel to maintain perfusion (around 12% PAV). We hypothesize that positive remodeling may temporarily mitigate microvascular impairment while simultaneously creating low endothelial shear stress (ESS), a hemodynamic environment known to promote plaque progression [30, 31]. This hypothesis is partially supported by Siasos et al. [8], who found that coronary segments from patients with non-obstructive CAD and abnormal microvascular function exhibited lower ESS than those with normal microvascular function. Although our definition of CMD differs, their results provide a relevant physiological context.
Importantly, the nearly normal or even slightly enlarged lumen size of CMD vessels observed in our study, a feature reported previously [25, 26, 32] but often overlooked, emphasizes that epicardial arterial compensation may already be present in vessels with microvascular dysfunction, even in the absence of flow-limiting stenosis. By analyzing a larger number of vessels, we were able to highlight this phenomenon, which has traditionally been attributed only to epicardial atherosclerosis [33]. In support of this, a previous study in patients with angina and non-obstructive CAD demonstrated a significant inverse correlation between lumen volume and minimal microvascular resistance [26], further reinforcing the idea that positive remodeling may occur in response to microvascular impairment.
Microvascular resistance reserve
MRR, a robust metric reflecting microvascular vasodilatory capacity, was incorporated in our analysis as a key index of microvascular function. Previous studies have demonstrated that MRR is independently associated with major adverse cardiac events over a 5-year follow-up, regardless of the presence/absence of flow-limiting epicardial disease [18], and may serve as a valuable reference measure of coronary microcirculation [34]. Prior work from our team also established the utility of MRR in patients with suspected CAD and revealed systemic associations between plaque burden and microvascular function [35, 36]. However, the current study uncovered an interesting pattern of MRR change across different vessel types. In non-ischemic vessels, non-calcified plaque burden was associated with a decrease in MRR, whereas in CMD vessels, it was associated with increased MRR prior to adjustment for positive remodeling. This seemingly paradoxical observation once again highlights the potential compensatory role of positive remodeling, which may initially buffer microvascular dysfunction while inadvertently facilitating plaque progression. Nevertheless, given the observational nature of this study, the directionality and causality of these associations cannot be definitively established. Longitudinal studies will be required to clarify the temporal relationship between early atherosclerosis, positive remodeling, and microvascular dysfunction.
Clinical implications
Patients with anginal symptoms but no angiographic evidence of obstructive coronary disease are often underdiagnosed and discharged without targeted treatment or follow-up, under the assumption of “normal” coronary arteries [28]. However, individuals with CMD, despite having predominantly non-obstructive CAD, already exhibit a median plaque volume of 50.40 mm3 or 12.5% plaque burden, corresponding to approximately the 60th percentile in men and the 80th percentile in women [37], levels that are themselves associated with an increased risk of future coronary events [28, 38, 39]. These patients therefore deserve closer clinical attention. In particular, non-obstructive vessels with a normal or even enlarged lumen but relatively high plaque burden, features suggestive of CMD and positive remodeling, may represent a critical window during which timely intervention could potentially slow or reverse atherosclerosis; however, this possibility requires confirmation in longitudinal studies. Our study highlights this potential interplay between myocardial microvascular dysfunction and early epicardial atherosclerosis, suggesting a potentially under-recognized association between early epicardial atherosclerotic changes and microvascular dysfunction that may have implications for earlier disease detection. This finding may help raise awareness among clinicians and patients regarding the importance of CMD as an early contributor to disease progression. Moreover, the integration of MRR in our study demonstrates its potential utility in future research. Combining PET-based microvascular assessments with noninvasive epicardial hemodynamic indices such as CT-derived FFR could offer a promising, fully noninvasive approach to CMD evaluation, paving the way for broader clinical application.
Limitations
First, this study was a post hoc analysis of the single center PACIFIC study, which may introduce selection bias and limit generalizability. In addition, the use of [15O]water PET requires specialized infrastructure and expertise for handling ultra-short half-life tracers, potentially limiting broader clinical applicability of the findings. Furthermore, because PACIFIC-1 was originally designed to evaluate patients with suspected obstructive CAD rather than CMD or ANOCA specifically, women and patients with classical microvascular angina phenotypes were underrepresented in the current cohort, which may have introduced selection bias and reduced applicability to broader CMD populations. Second, endothelial function testing and invasive assessment of microvascular resistance were not available, restricting direct mechanistic interpretation. Third, as an observational study, the current analysis does not include clinical outcome data, limiting prognostic inferences. Fourth, due to the cross-sectional nature of the data, causal relationships between CMD, positive remodeling, and plaque progression could not be established; the current findings remain hypothesis-generating. Future serial studies are warranted to further confirm the hypothesis. Fifth, 55 coronary arteries with total or subtotal occlusion were excluded from the MRR analysis, which may have introduced selection bias. Sixth, myocardial perfusion territories were assigned according to the standardized 17-segment AHA model without systematic adjustment for individual coronary dominance patterns, which may have introduced some degree of vessel-level misclassification. Finally, while the sample size was larger than prior studies, subgroup analyses, particularly those involving vessel-level imaging and physiology, may remain underpowered for certain stratifications.
Conclusions
This study identified distinct coronary phenotypes across vessels with normal function, CMD, and epicardial flow-limiting disease. CMD vessels were characterized by intermediate plaque burden, and pronounced positive remodeling, representing a unique structural–functional profile. The association between positive remodeling and preserved MRR underscores a potential link between epicardial morphology and microvascular function. Together, these findings highlight the complex interplay between microvascular and epicardial pathology and support further longitudinal studies to clarify their temporal relationships and prognostic implications.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Full list of the PACIFIC-1 investigators: Roel S. Driessen, MD, PhDc, Pepijn van Diemen, MDc, Ruurt Jukema, MDc, Lilian Meijboom, MD, PhDf, Juhani Knuuti, MD, PhDh, Antti Saraste, MD, PhDh, Pieter G. Raijmakers, MD, PhDf, Paul Knaapen, MD, PhDc, Ibrahim Danad, MD, PhDc
Author contributions
All authors contributed to the study conception and design. Data collection was performed by Roel Hoek, Roel S. Driessen, Pepijn van Diemen, Ruurt Jukema, Juhani Knuuti, Pieter G. Raijmakers, Paul Knaapen, and Ibrahim Danad. The first draft of the manuscript was written by Yipu Ding, review and editing was performed by Nick S. Nurmohamed, Jeroen J. Bax, and Alexander van Rosendael. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
Data availability
The data underlying this article are not publicly available due to study restrictions and data privacy considerations.
Declarations
Ethics approval
The study protocol was approved by the Medical Ethics Committee of the Free University Medical Center.
Consent to participate
All participants provided written informed consent.
Competing interests
Dr. Knuuti received consultancy fees from GE Healthcare and Synektik and speaker fees from Siemens, outside of the submitted work. Dr. Nurmohamed reports grants from the Dutch Heart Foundation (Dekker 03–007-2023–0068) and European Atherosclerosis Society (2023), research funding/speaker fees from Cleerly, Daiichi Sankyo, Novartis and Ultragenyx, and is co-founder of Lipid Tools.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yipu Ding and Roel Hoek contributed equally to this work.
Contributor Information
Yipu Ding, Email: dingyipu123@126.com.
on the behalf of the PACIFIC-1 Investigators.:
Roel S. Driessen, Pepijn van Diemen, Ruurt Jukema, Lilian Meijboom, Juhani Knuuti, Antti Saraste, Pieter G. Raijmakers, Paul Knaapen, and Ibrahim Danad
References
- 1.Patel MR, Peterson ED, Dai D, Brennan JM, Redberg RF, Anderson HV, et al. Low diagnostic yield of elective coronary angiography. N Engl J Med. 2010;362:886–95. 10.1056/NEJMoa0907272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zimmerli A, Salihu A, Antiochos P, Lu H, Pitta Gros B, Berger A, et al. Evolution of coronary microvascular dysfunction prevalence over time and across diagnostic modalities in patients with ANOCA: a systematic review. J Clin Med. 2025;14:829. 10.3390/jcm14030829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stenstrom I, Maaniitty T, Uusitalo V, Pietila M, Ukkonen H, Kajander S, et al. Frequency and angiographic characteristics of coronary microvascular dysfunction in stable angina: a hybrid imaging study. Eur Heart J Cardiovasc Imaging. 2017;18:1206–13. 10.1093/ehjci/jex193. [DOI] [PubMed] [Google Scholar]
- 4.Vrints C, Andreotti F, Koskinas KC, Rossello X, Adamo M, Ainslie J, et al. 2024 ESC Guidelines for the management of chronic coronary syndromes. Eur Heart J. 2024;45:3415–537. 10.1093/eurheartj/ehae177. [DOI] [PubMed] [Google Scholar]
- 5.Hong D, Shin D, Lee SH, Joh HS, Choi KH, Kim HK, et al. Prognostic impact of coronary microvascular dysfunction according to different patterns by invasive physiologic indexes in symptomatic patients with intermediate coronary stenosis. Circ Cardiovasc Interv. 2023;16:e012621. 10.1161/CIRCINTERVENTIONS.122.012621. [DOI] [PubMed] [Google Scholar]
- 6.Lee JM, Jung JH, Hwang D, Park J, Fan Y, Na SH, et al. Coronary flow reserve and microcirculatory resistance in patients with intermediate coronary stenosis. J Am Coll Cardiol. 2016;67:1158–69. 10.1016/j.jacc.2015.12.053. [DOI] [PubMed] [Google Scholar]
- 7.Reynolds HR, Smilowitz NR. Ischaemia with non-obstructive coronary arteries in the 2024 european society of cardiology guidelines for the management of chronic coronary syndromes. Eur Heart J Acute Cardiovasc Care. 2025;14:173–7. 10.1093/ehjacc/zuaf005. [DOI] [PubMed] [Google Scholar]
- 8.Siasos G, Sara JD, Zaromytidou M, Park KH, Coskun AU, Lerman LO, et al. Local low shear stress and endothelial dysfunction in patients with nonobstructive coronary atherosclerosis. J Am Coll Cardiol. 2018;71:2092–102. 10.1016/j.jacc.2018.02.073. [DOI] [PubMed] [Google Scholar]
- 9.Travieso A, Jeronimo-Baza A, Faria D, Shabbir A, Mejia-Renteria H, Escaned J. Invasive evaluation of coronary microvascular dysfunction. J Nucl Cardiol. 2022;29:2474–86. 10.1007/s12350-022-02997-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Schindler TH, Fearon WF, Pelletier-Galarneau M, Ambrosio G, Sechtem U, Ruddy TD, et al. Myocardial perfusion PET for the detection and reporting of coronary microvascular dysfunction: a JACC: cardiovascular imaging expert panel statement. JACC Cardiovasc Imaging. 2023;16:536–48. 10.1016/j.jcmg.2022.12.015. [DOI] [PubMed] [Google Scholar]
- 11.Hoek R, van Diemen PA, Somsen YBO, de Winter RW, Jukema RA, Dahdal JE, et al. Myocardial perfusion imaging in advanced coronary artery disease. Eur J Clin Invest. 2025;55:e70024. 10.1111/eci.70024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pijls NH, De Bruyne B, Peels K, Van Der Voort PH, Bonnier HJ, Bartunek JKJJ, et al. Measurement of fractional flow reserve to assess the functional severity of coronary-artery stenoses. N Engl J Med. 1996;334:1703–8. 10.1056/NEJM199606273342604. [DOI] [PubMed] [Google Scholar]
- 13.De Bruyne B, Pijls NHJ, Gallinoro E, Candreva A, Fournier S, Keulards DCJ, et al. Microvascular resistance reserve for assessment of coronary microvascular function: JACC technology corner. J Am Coll Cardiol. 2021;78:1541–9. 10.1016/j.jacc.2021.08.017. [DOI] [PubMed] [Google Scholar]
- 14.Danad I, Raijmakers PG, Driessen RS, Leipsic J, Raju R, Naoum C, et al. Comparison of coronary CT angiography, SPECT, PET, and hybrid imaging for diagnosis of ischemic heart disease determined by fractional flow reserve. JAMA Cardiol. 2017;2:1100–7. 10.1001/jamacardio.2017.2471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation. 2002;105:539–42. 10.1161/hc0402.102975. [DOI] [PubMed] [Google Scholar]
- 16.Danad I, Uusitalo V, Kero T, Saraste A, Raijmakers PG, Lammertsma AA, et al. Quantitative assessment of myocardial perfusion in the detection of significant coronary artery disease: cutoff values and diagnostic accuracy of quantitative [(15)O]H2O PET imaging. J Am Coll Cardiol. 2014;64:1464–75. 10.1016/j.jacc.2014.05.069. [DOI] [PubMed] [Google Scholar]
- 17.Nurmohamed NS, Danad I, Jukema RA, de Winter RW, de Groot RJ, Driessen RS, et al. Development and validation of a quantitative coronary CT angiography model for diagnosis of vessel-specific coronary ischemia. JACC Cardiovasc Imaging. 2024;17:894–906. 10.1016/j.jcmg.2024.01.007. [DOI] [PubMed] [Google Scholar]
- 18.Boerhout CKM, Lee JM, de Waard GA, Mejia-Renteria H, Lee SH, Jung JH, et al. Microvascular resistance reserve: diagnostic and prognostic performance in the ILIAS registry. Eur Heart J. 2023;44:2862–9. 10.1093/eurheartj/ehad378. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lee SH, Choi KH, Hong D, Shin D, Joh HS, Kim HK, et al. Prognostic implications of microvascular resistance reserve in symptomatic patients with intermediate coronary stenosis. JACC Cardiovasc Interv. 2024;17:786–97. 10.1016/j.jcin.2024.01.008. [DOI] [PubMed] [Google Scholar]
- 20.Leipsic J, Abbara S, Achenbach S, Cury R, Earls JP, Mancini GJ, et al. SCCT guidelines for the interpretation and reporting of coronary CT angiography: a report of the Society of cardiovascular computed tomography guidelines committee. J Cardiovasc Comput Tomogr. 2014;8:342–58. 10.1016/j.jcct.2014.07.003. [DOI] [PubMed] [Google Scholar]
- 21.Choi AD, Marques H, Kumar V, Griffin WF, Rahban H, Karlsberg RP, et al. CT Evaluation by artificial intelligence for atherosclerosis, stenosis and vascular morphology (CLARIFY): a multi-center, international study. J Cardiovasc Comput Tomogr. 2021;15:470–6. 10.1016/j.jcct.2021.05.004. [DOI] [PubMed] [Google Scholar]
- 22.Griffin WF, Choi AD, Riess JS, Marques H, Chang HJ, Choi JH, et al. AI evaluation of stenosis on coronary CTA, comparison with quantitative coronary angiography and fractional flow reserve: a CREDENCE Trial Substudy. JACC Cardiovasc Imaging. 2023;16:193–205. 10.1016/j.jcmg.2021.10.020. [DOI] [PubMed] [Google Scholar]
- 23.Boogers MJ, Broersen A, van Velzen JE, de Graaf FR, El-Naggar HM, Kitslaar PH, et al. Automated quantification of coronary plaque with computed tomography: comparison with intravascular ultrasound using a dedicated registration algorithm for fusion-based quantification. Eur Heart J. 2012;33:1007–16. 10.1093/eurheartj/ehr465. [DOI] [PubMed] [Google Scholar]
- 24.Motoyama S, Sarai M, Harigaya H, Anno H, Inoue K, Hara T, et al. Computed tomographic angiography characteristics of atherosclerotic plaques subsequently resulting in acute coronary syndrome. J Am Coll Cardiol. 2009;54:49–57. 10.1016/j.jacc.2009.02.068. [DOI] [PubMed] [Google Scholar]
- 25.AlBadri A, Eshtehardi P, Hung OY, Bouchi Y, Khawaja S, Mercado K, et al. Coronary microvascular dysfunction is associated with significant plaque burden and diffuse epicardial atherosclerotic disease. JACC Cardiovasc Interv. 2019;12:1519–20. 10.1016/j.jcin.2019.05.003. [DOI] [PubMed] [Google Scholar]
- 26.Collet C, Sakai K, Mizukami T, Ohashi H, Bouisset F, Caglioni S, et al. Vascular remodeling in coronary microvascular dysfunction. JACC Cardiovasc Imaging. 2024;17:1463–76. 10.1016/j.jcmg.2024.07.018. [DOI] [PubMed] [Google Scholar]
- 27.Ang DTY, Carberry J, Ford TJ, Kamdar A, Sykes R, Sidik NP, et al. Coronary microvascular function and atherosclerotic plaque burden in ischaemia and no obstructive coronary arteries: a secondary analysis of the CorMicA trial. Heart. 2025;111:117–24. 10.1136/heartjnl-2024-324677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Godo S, Corban MT, Toya T, Gulati R, Lerman LO, Lerman A. Association of coronary microvascular endothelial dysfunction with vulnerable plaque characteristics in early coronary atherosclerosis. EuroIntervention. 2020;16:387–94. 10.4244/EIJ-D-19-00265. [DOI] [PubMed] [Google Scholar]
- 29.Zeiher AM, Drexler H, Wollschlager H, Just H. Endothelial dysfunction of the coronary microvasculature is associated with coronary blood flow regulation in patients with early atherosclerosis. Circulation. 1991;84:1984–92. 10.1161/01.cir.84.5.1984. [DOI] [PubMed] [Google Scholar]
- 30.Hartman EMJ, De Nisco G, Kok AM, Tomaniak M, Nous FMA, Korteland SA, et al. Wall shear stress-related plaque growth of lipid-rich plaques in human coronary arteries: an near-infrared spectroscopy and optical coherence tomography study. Cardiovasc Res. 2023;119:1021–9. 10.1093/cvr/cvac178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kok AM, Molony DS, Timmins LH, Ko YA, Boersma E, Eshtehardi P, et al. The influence of multidirectional shear stress on plaque progression and composition changes in human coronary arteries. EuroIntervention. 2019;15:692–9. 10.4244/EIJ-D-18-00529. [DOI] [PubMed] [Google Scholar]
- 32.Ahn SG, Suh J, Hung OY, Lee HS, Bouchi YH, Zeng W, et al. Discordance between fractional flow reserve and coronary flow reserve: insights from intracoronary imaging and physiological assessment. JACC Cardiovasc Interv. 2017;10:999–1007. 10.1016/j.jcin.2017.03.006. [DOI] [PubMed] [Google Scholar]
- 33.Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ. Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med. 1987;316:1371–5. 10.1056/NEJM198705283162204. [DOI] [PubMed] [Google Scholar]
- 34.Rigattieri S, Barbato E, Berry C. Microvascular resistance reserve: a reference test of the coronary microcirculation? Eur Heart J. 2023;44:2870–2. 10.1093/eurheartj/ehad291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Hoshino M, Jukema RA, Hoek R, Dahdal J, Raijmakers P, Driessen R, et al. Microvascular resistance reserve in relation to total and vessel-specific atherosclerotic burden. Eur Heart J Cardiovasc Imaging. 2025;26:481–8. 10.1093/ehjci/jeae293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Hoshino M, Hoek R, Jukema RA, Dahdal J, van Diemen P, Raijmakers P, et al. Homogeneity of the coronary microcirculation in angina with non-obstructive coronary artery disease. Eur Heart J Cardiovasc Imaging. 2025;26:1120–7. 10.1093/ehjci/jeaf101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Ichikawa K, Ronen S, Bishay R, Krishnan S, Benzing T, Kianoush S, et al. Coronary plaque volume in an asymptomatic population: miami heart study at baptist health South Florida. JACC Cardiovasc Imaging. 2025. 10.1016/j.jcmg.2025.08.001. [DOI] [PubMed] [Google Scholar]
- 38.Smilowitz NR, Toleva O, Chieffo A, Perera D, Berry C. Coronary microvascular disease in contemporary clinical practice. Circ Cardiovasc Interv. 2023;16:e012568. 10.1161/CIRCINTERVENTIONS.122.012568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Taqueti VR, Solomon SD, Shah AM, Desai AS, Groarke JD, Osborne MT, et al. Coronary microvascular dysfunction and future risk of heart failure with preserved ejection fraction. Eur Heart J. 2018;39:840–9. 10.1093/eurheartj/ehx721. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data underlying this article are not publicly available due to study restrictions and data privacy considerations.