Prognostic Value of Endogenous-Type Coronary Microvascular Dysfunction After Elective Percutaneous Coronary Intervention

Abstract

Background: Global coronary flow reserve (G-CFR) impairment represents coronary microvascular dysfunction (CMD) and correlates with poor prognosis. Hyperemic coronary flow is reduced in conventional CMD, but normal or mildly reduced with elevated resting flow in endogenous-type CMD (E-CMD). This retrospective study assessed the prognostic value of post-percutaneous coronary intervention (PCI) CMD, focusing on E-CMD.

Methods and Results: We included 320 chronic coronary syndrome (CCS) patients undergoing PCI and post-PCI phase contrast cine-cardiac magnetic resonance imaging (CMR). Major adverse cardiac and cerebrovascular events (MACCE) were evaluated, considering the presence of post-PCI CMD and E-CMD based on G-CFR and resting myocardial flow assessed by coronary sinus flow using CMR. CMD was defined as G-CFR <2.0 and classified as E-CMD or non-E-CMD. Post-PCI CMD was observed in 43.4% of patients, 63.3% exhibiting E-CMD. During a median 2.5-year follow-up, MACCE occurred in 26 (8.1%) patients, more often in those with CMD (11.5% vs. 5.5%; P=0.063). MACCE incidence was higher in E-CMD than non-E-CMD and non-CMD (14.8% vs. 5.9% and 5.5%, respectively; P=0.027). Kaplan-Meier analysis revealed worse prognosis in E-CMD (P=0.025). Cox proportional hazards modeling revealed that E-CMD independently predicted MACCE (hazard ratio 3.24; 95% confidence interval 1.47–7.14; P=0.004).

Conclusions: Post-PCI CMD, particularly E-CMD, was significantly associated with worse outcomes in CCS patients. Post-PCI CMD evaluation could guide therapeutic strategies for CCS patients.

Conventionally, the therapeutic management of coronary artery disease, supported by objective evidence of ischemia, involves the use of percutaneous coronary intervention (PCI) to relieve epicardial coronary stenosis. However, this approach is based on the assumption that revascularization can improve myocardial ischemia and potentially enhance clinical outcomes. However, recent trials have failed to demonstrate significant superiority of elective PCI over guideline-directed medical therapy in terms of its impact on clinical outcomes in patients with chronic coronary syndrome (CCS).¹

Coronary microvascular dysfunction (CMD), evaluated by indices such as coronary flow reserve (CFR), is a condition affecting the structure and/or function of the coronary microcirculation. In addition to epicardial coronary stenosis, CMD has been shown to contribute incrementally to poor clinical outcomes.^2–4 Of the non-invasive myocardial perfusion imaging techniques, cardiac magnetic resonance (CMR) imaging using phase contrast CMR (PC-CMR) to measure absolute coronary sinus flow (CSF) is an established method for assessing global CFR (G-CFR),⁵ with cardiac positron emission tomography (PET) considered the gold standard.⁶ This CMR approach offers several advantages, including broad applicability, no radiation exposure, and no requirement for a radioactive tracer. Impaired G-CFR has been associated with adverse prognosis.⁷ Traditionally, impaired CFR has been primarily considered to be associated with a reduction in hyperemic myocardial blood flow (MBF), referred to as classical-type CMD.⁸ Conversely, a recent study has suggested the clinical significance of an endogenous-type of CMD (E-CMD), characterized by an elevation in resting MBF alongside normal or mildly reduced hyperemic MBF.⁸ Whereas classical-type CMD (or “non-E-CMD”) is related to impaired hyperemic MBF, E-CMD reflects impaired vasodilator capacity due to increased myocardial workload and oxygen demands at rest.⁹ Although previous studies have reported an association between higher resting MBF and subsequent worse prognosis,^10,11 the specific prognostic impact of E-CMD compared with non-E-CMD remains to be determined. Furthermore, the prevalence and clinical significance of global E-CMD and non-E-CMD after elective PCI remain unknown.

Therefore, the aims of this retrospective observational study of patients with CCS undergoing elective PCI were to: (1) investigate, using PC-CMR, the prevalence of CMD after complete revascularization by elective and uncomplicated PCI; (2) assess the predictors of post-PCI CMD; (3) examine the relationship between post-PCI CMD and clinical outcomes; and (4) evaluate the prevalence and prognostic significance of post-PCI E-CMD and its predictors.

Methods

Study Design and Patient Population

Between November 2016 and September 2022, patients with CCS who underwent elective PCI with post-PCI CMR examinations at Tsuchiura Kyodo General Hospital were identified from the institutional PCI database. Upon identifying a lesion suitable for PCI by diagnostic catheterization, patients were informed of the study and their consent was obtained for magnetic resonance imaging examination after the PCI procedure. Of the 892 patients who underwent PCI for CCS, 48 with left main lesions, 210 with contraindications such as renal dysfunction, claustrophobia, or the presence of metallic devices, 13 with a history of coronary bypass grafting, and 96 identified as restenosis were excluded from the study. Of the remaining 525 patients, 349 consented to undergo CMR before and after PCI.

A total of 349 patients with CCS who had undergone post-PCI CMR evaluations were initially identified and retrospectively investigated in this study. Clinical indications for CMR were determined by the attending physicians and included evaluation of perfusion impairment and left ventricular (LV) function. These patients were selected based on the following inclusion criteria: age >20 years; detection of an identifiable, de novo single culprit lesion located in the proximal segment of a native coronary artery; manifestation of angina symptoms; and PC-CMR examination within 3 months after PCI. In this study, CCS was defined as the presence of angina symptoms with consistent frequency, duration, or intensity within 6 weeks before PCI. The target lesion was identified based on a combination of diagnostic coronary angiograms, angiographic lesion morphology, electrocardiogram results, PCI suitability, non-invasive stress testing, and fractional flow reserve (FFR) values. Exclusion criteria included angiographically significant left main disease, a history of myocardial infarction (MI) or coronary artery bypass surgery, renal insufficiency with a baseline serum creatinine level >1.5 mg/dL, and the presence of cardiomyopathy, acute coronary syndrome, and active congestive heart failure. Following PCI, we examined cardiac troponin I and excluded patients with periprocedural MI, as delineated by the Fourth Universal Definition of Myocardial Infarction.¹² The presence of periprocedural MI determination was based on blood samples taken approximately 20–24 h after PCI, symptoms, and other post-PCI objective findings. The exclusion of periprocedural MI was due to its potential impact on clinical outcomes. Patients exhibiting minor elevations in cardiac troponin levels without other manifestations of periprocedural MI described above were included in the final analysis. All patients received timely initiation of guideline-directed medical therapy before undergoing PCI.¹³

This study was approved by and conducted in accordance with the guidelines of the Ethics Committee of Tsuchiura Kyodo General Hospital (Approval no. TKGH-IRB 2021FY16). This study adhered to the principles of the Declaration of Helsinki for human investigations. The requirement for written informed consent from participating patients was waived and patients were notified of the option to opt-out of the study via the hospital bulletin board.

Invasive Coronary Angiography and PCI

Each CCS patient initially underwent routine selective coronary angiography via the radial artery, using a 5-Fr system to assess the coronary anatomy. Intracoronary bolus injections of nitroglycerin (0.2 mg) were administered at the start of the procedure and were subsequently repeated at 30-min intervals. Angiographically intermediate lesions were assessed using FFR at the discretion of the interventionalist. PCI was performed using a 6-Fr system via the radial artery according to recent guidelines.¹⁴ Quantitative coronary angiographic analyses were performed using the CMS-MEDIS system (Medis Medical Imaging Systems, Leiden, Netherlands). The choice of stent (newer than a second-generation drug-eluting stent) and PCI strategy were at the discretion of the interventionalist. Online quantitative coronary angiography was performed to determine appropriate stent size. Successful PCI was defined as <20% residual stenosis, Thrombolysis in Myocardial Infarction (TIMI) Grade 3 flow, no side branch occlusion (>2 mm in diameter) and distal embolization, and the absence of periprocedural MI (Type 4A).¹² All patients in this study underwent post-PCI CMR imaging a median of 32 days (interquartile range [IQR] 18–40 days) after PCI.

CMR Examination

CMR Acquisition and Cine CMR Post-PCI CMR images were obtained using a 1.5-T scanner (Philips Achieva; Philips Medical Systems, Best, Netherlands) equipped with 32-channel cardiac coils. Cardiac gating and heart rate recording were achieved using the vectorcardiogram device. The cine CMR parameters used were as follows: repetition time/echo time, 4.1 ms/1.4 ms; slice thickness, 6 mm; flip angle, 55°; field of view, 350×350 mm²; matrix size, 128×128; and 20 phases per cardiac cycle. Ventricular volumes, ejection fraction (EF), and LV masses were determined by outlining the endo- and epicardial borders on short-axis cine images. LV mass and volume were computed using Simpson’s rule and CMR data. Cine CMR imaging was performed using a retrospectively gated steady state free-precession sequence. Overall, 12 short-axis slices of the left ventricle were acquired from the apex to the base.

CSF and G-CFR Measurement Using CMR The coronary sinus (CS) was located in the atrioventricular groove by using basal slices from the short-axis stack. The plane for flow measurement using phase contrast CMR was positioned perpendicular to the CS at 1–2 cm from the ostium. Velocity-encoded images were captured using retrospective electrocardiographic gating during 15-s breath holds. The imaging parameters were configured as follows: repetition time/echo time, 7.3 ms/4.4 ms; flip angle, 10°; field of view, 250×250 mm²; acquisition matrix, 128×128; 20 phases per cardiac cycle; encoding, 50 cm/s; and slice thickness, 6 mm. Maximal stable hyperemia was induced via intravenous administration of adenosine (140 μg/kg/min through a central vein). All patients were instructed to strictly refrain from ingesting caffeinated beverages for >24 h before CMR examination. The duration between the end of hyperemia and resting image acquisition was 10 min.

Quantitative CSF analyses were performed in a blinded manner by 2 expert investigators (K.N. and Y.K.) using proprietary software (Philips View Forum; Philips Medical Systems). The CS contour was traced on the magnified images spanning the entire cardiac cycle. CSF was quantified by integrating the flow rates from each cardiac cycle and multiplying them by the mean heart rate during the acquisition period. The resting CSF value was corrected using rate-pressure products (RPP) as follows:^5,15

RPP = systolic blood pressure (mmHg) × heart rate

Corrected CSF (mL/min) = (CSF / RPP) × 10,000

Corrected CSF (mL/min/g) = corrected CSF (mL/min) / LV mass (g)

G-CFR was evaluated using the CSF reserve, determined by dividing the CSF during maximal hyperemia by the resting CSF as follows:

G-CFR = hyperemic CSF (mL/min) / corrected resting CSF (mL/min)

Figure 1 shows representative cases. Post-PCI CMD was defined as impaired G-CFR <2.0.⁵ There is no established cut-off value of elevated resting CSF, so we determined this value based on the same percentile matching method used in previous reports, particularly with reference to impaired G-CFR.¹⁶ Because G-CFR <2.0 corresponds to the 43rd percentile of the total cohort, elevated resting CSF was defined as 1.06 mL/min/g, which corresponds to 57th percentile of resting CSF measurements. Among patients with CMD, those with normal resting CSF (<1.06 mL/min/g) were categorized into the non-E-CMD group and those with elevated resting CSF showing impaired G-CFR were categorized into the E-CMD group.

Figure 1.

Cardiac magnetic resonance (CMR) images and angiography in representative cases. (A) Magnitude image and (B) phase contrast image of coronary sinus (red circle). In the magnitude image (A), the coronary sinus was identified at the atrioventricular groove, and a region of interest (ROI) was drawn. The phase contrast image (B) was synchronized with the magnitude image, and the ROI was automatically positioned in the corresponding location. (C,D) Representative patterns of coronary sinus flow (CSF) curves in patients without coronary microvascular dysfunction (Non-CMD; resting CSF 0.81 mL/min/g, hyperemic CSF 3.56 mL/min/g, global coronary flow reserve [G-CFR] 4.37; C) and with endogenous-type CMD (E-CMD; resting CSF 1.67 mL/min/g, hyperemic CSF 2.27 mL/min/g, G-CFR 1.36; D).

Definition of Clinical Outcomes and Data Collection

The primary clinical outcome of this study was major adverse cardiac and cerebrovascular events (MACCE), defined as a composite of cardiac death, non-fatal MI, hospitalization due to congestive heart failure, and ischemic stroke over a median follow-up period of 2.5 years (IQR 1.3–4.0 years). Clinical endpoints were determined through a blinded assessment of hospital records or telephone interviews by cardiologists blinded to the clinical, angiographic, and CMR data (K.N., E.U.). The time to event was calculated as the interval from the PCI to the first occurrence of MACCE. Patients without MACCE were censored at their final follow-up visit.

Statistical Analysis

Statistical analyses were performed using R version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria). Categorical data are expressed as numbers and percentages and were compared using Chi-squared tests. The normality of the distributed values was assessed using Shapiro-Wilk statistics. Continuous data that were not normally distributed are expressed as the median with IQR and were analyzed using the Mann-Whitney U test or Kruskal-Wallis test. Data that were normally distributed are presented as the mean±SD and were analyzed using unpaired Student’s t-test or 1-way analysis of variance (ANOVA). ANOVA was used to assess the significance of differences between groups for variables with and without normal distributions.

Patient characteristics and CMR findings were compared between patients with and without post-PCI CMD, as well as among non-CMD, E-CMD, and non-E-CMD groups. Furthermore, patients were categorized into 2 groups based on the presence or absence of MACCE to explore prognostic factors. Kaplan-Meier analysis was used to assess the occurrence of MACCE in the presence or absence of post-PCI CMD, as well as in the non-CMD, E-CMD, and non-E-CMD groups. Logistic regression analysis was used to predict post-PCI CMD using the pre-PCI parameters with P<0.100 in Table 1. Multivariable analysis was conducted using the factors that were significant in the univariable logistic regression analysis, namely creatinine, N-terminal pro B-type natriuretic peptide (NT-proBNP), and SYNTAX score. G-CFR and the components used for calculation, such as rest and hyperemic CSF, were considered to be confounding factors with CMD and were therefore excluded from the multivariable analysis. To avoid overfitting, we selected variables based on Akaike’s information criterion. A Cox proportional hazards regression model was used to identify MACCE predictors of post-PCI. Two-sided P<0.05 was considered statistically significant.

Table 1.

Patient Characteristics According to the Presence or Absence of CMD After PCI

	All patients (n=320)	Patients without post-PCI CMD (n=181)	Patients with post-PCI CMD (n=139)	P value
Demographics
Age (years)	70 [61–75]	69 [60–75]	70 [64–77]	0.054
Sex				0.637
Male	272 (85.0)	152 (84.0)	120 (86.3)
Female	48 (15.0)	29 (16.0)	19 (13.7)
Body surface area (m2)	1.71 [1.61–1.81]	1.71 [1.61–1.82]	1.72 [1.60–1.81]	0.674
Prior PCI history	117 (36.6)	62 (34.3)	55 (39.6)	0.350
Smoking history	218 (68.1)	120 (66.3)	98 (70.5)	0.468
Hypertension	246 (76.9)	142 (78.5)	104 (74.8)	0.504
Diabetes	137 (42.8)	77 (42.5)	60 (43.2)	1.000
Dyslipidemia	175 (54.7)	101 (55.8)	74 (53.2)	0.652
Family history of CAD	39 (12.2)	19 (10.5)	20 (14.4)	0.306
Medication
Aspirin	318 (99.4)	179 (98.9)	139 (100.0)	0.507
Statin	297 (92.8)	165 (91.2)	132 (95.0)	0.275
β-blocker	233 (72.8)	127 (70.2)	106 (76.3)	0.255
ACEi/ARB	241 (75.3)	135 (74.6)	106 (76.3)	0.794
Calcium channel blocker	125 (39.1)	74 (40.9)	51 (36.7)	0.489
Nitrate	49 (15.3)	31 (17.1)	18 (12.9)	0.349
Presentation
PCI target, RCA/LAD/LCx	75/191/54 (23.4/59.7/16.9)	46/106/29 (25.4/58.6/16.0)	29/85/25 (20.9/61.2/18.0)	0.616
SYNTAX score	9 [6–14]	9 [6–13]	10 [6–16]	0.066
MACCE	26 (8.1)	10 (5.5)	16 (11.5)	0.063
Laboratory data
CRP (mg/dL)	0.07 [0.03–0.20]	0.05 [0.03–0.18]	0.09 [0.04–0.22]	0.044
Creatinine (mg/dL)	0.85 [0.73–0.97]	0.82 [0.72–0.94]	0.89 [0.74–1.01]	0.013
Total cholesterol (mg/dL)	157 [140–186]	156 [140–186]	159 [140–186]	0.995
LDL-C (mg/dL)	86 [68–106]	86 [68–111]	85 [66–104]	0.492
HDL-C (mg/dL)	48 [41–58]	48 [42–59]	47 [41–57]	0.564
Triglyceride (mg/dL)	127 [88–166]	125 [89–162]	128 [88–175]	0.707
NT-proBNP (ng/L)	176 [66–453]	119 [47–413]	207 [87–470]	0.002
HbA1c (%)	6.3 [5.7–6.9]	6.2 [5.7–6.9]	6.3 [5.7–7.0]	0.762
Hemoglobin (g/dL)	13.5 [12.4–14.4]	13.5 [12.5–14.3]	13.4 [12.3–14.5]	0.782
Post-PCI CMR parameters
End-diastolic LV volume (mL)	122 [103–142]	121 [101–138]	125 [105–148]	0.078
End-systolic LV volume (mL)	47 [32–64]	46 [33–60]	47 [32–69]	0.183
EF (%)	61.4 [51.7–70.8]	61.9 [54.2–70.7]	61.3 [49.0–71.2]	0.262
LV mass (mL)	124 [103–148]	122 [99–144]	128 [105–152]	0.116
Resting CSF (mL/min)	108.9 [73.6–153.6]	89.2 [57.0–125.4]	135.9 [100.9–175.9]	<0.001
Corrected resting CSF (mL/min)	123.4 [86.6–170.9]	98.0 [68.2–132.3]	163.3 [122.8–208.2]	<0.001
Corrected resting CSF (mL/min/g)	1.00 [0.67–1.41]	0.80 [0.55–1.09]	1.26 [0.88–1.71]	<0.001
Hyperemic CSF (mL/min)	272.0 [204.6–372.4]	317.7 [236.8–409.1]	231.8 [168.8–300.4]	<0.001
Hyperemic CSF (mL/min/g)	2.25 [1.53–3.11]	2.62 [1.79–3.55]	1.79 [1.29–2.46]	<0.001
G-CFR	2.17 [1.62–3.04]	2.95 [2.44–4.18]	1.53 [1.22–1.76]	<0.001

Unless indicated otherwise, data are given as median [interquartile range] or n (%). ACEi, angiotensin-converting enzyme inhibition; ARB, angiotensin receptor blocker; CAD, coronary artery disease; CMD, coronary microvascular dysfunction; CMR, cardiac magnetic resonance imaging; CRP, C-reactive protein; CSF, coronary sinus blood flow; EF, ejection fraction; G-CFR, global coronary flow reserve; HDL-C, high-density lipoprotein cholesterol; LAD, left anterior descending artery; LCx, left circumflex artery; LDL-C, low-density lipoprotein cholesterol; LV, left ventricular; MACCE, major adverse cardiac and cerebrovascular events (defined as cardiac death, non-fatal myocardial infarction, hospitalization for congestive heart failure, and ischemic stroke); NT-proBNP, N-terminal pro-B-type natriuretic peptide; PCI, percutaneous coronary intervention; RCA, right coronary artery.

Results

Baseline Patient Characteristics and CMR Findings

After excluding 22 patients from the final analysis because of unsatisfactory CMR imaging quality and another 7 patients due to periprocedural MI (Type 4A MI), attributed to side-branch occlusion or distal embolization,¹² the final analysis included 320 patients who underwent uncomplicated PCI with complete post-PCI CMR data. Post-PCI median G-CFR and resting CSF were 2.17 (IQR 1.62–3.04) and 1.00 mL/min/g (IQR 0.67–1.41 mL/min/g), respectively. The cut-off value of the 43rd percentile from the top corresponding to the elevated post-PCI resting CSF for E-CMD was 1.06 mL/min/g.

Prevalence of Post-PCI CMD and Predictive Factors

Post-PCI CMD was observed in 139 (43.4%) patients. Table 1 presents baseline characteristics and CMR data of patients according to the presence or absence of post-PCI CMD. Patients with CMD had significantly higher resting CSF, lower hyperemic CSF, higher C-reactive protein (CRP), higher creatinine, and NT-proBNP than those without CMD (Table 1). After adjusting for risk factors, which included pre-PCI factors with P<0.10 in Univariable analysis, creatinine, NT-proBNP (per100-ng/L increase), and SYNTAX score were significantly associated with post-PCI CMD post-PCI, with odds ratios of OR 3.72 (95% confidence interval [CI] 1.22–11.30; P=0.021), 1.03 (95% CI 1.00–1.05; P=0.048), and 1.04 (95% CI 1.00–1.08; P=0.027), respectively (Table 2). Table 3 presents baseline characteristics and CMR data for patients in the non-CMD, non-E-CMD and E-CMD groups. Of patients with CMD, 88 (63.3%) had E-CMD and 51 (36.7%) had non-E-CMD. There was a tendency for a higher proportion of women in the E-CMD group than in the non-E-CMD and non-CMD groups, although the difference was not statistically significant (17.0% vs. 7.8% and 16.0%, respectively; P=0.289).

Table 2.

Predictors of CMD After PCI (Logistic Regression Analysis)

	Univariable analysis		Multivariable analysis
	OR (95% CI)	P value	OR (95% CI)	P value
Age	1.02 (1.00–1.04)	0.078
CRP	1.11 (0.90–1.38)	0.324
Creatinine	4.24 (1.42–12.7)	0.010	3.72 (1.22–11.30)	0.021
NT-proBNP (per100-ng/L increase)	1.03 (1.00–1.60)	0.023	1.03 (1.00–1.05)	0.048
End-diastolic LV volume	1.01 (1.00–1.01)	0.092
SYNTAX score	1.04 (1.01–1.08)	0.012	1.04 (1.00–1.08)	0.027

CI, confidence interval; OR, odds ratio. Other abbreviations sa in Table 1.

Table 3.

Patient Characteristics According to CMD Subtype

	No CMD (n=181)	Non-E-CMD (n=51)	E-CMD (n=88)	P value
Demographics
Age (years)	69 [60–75]	70 [58–78]	70 [66–77]	0.093
Sex				0.289
Male	152 (84.0)	47 (92.2)	73 (83.0)
Female	29 (16.0)	4 (7.8)	15 (17.0)
Body surface area (m2)	1.71 [1.61–1.82]	1.75 [1.65–1.81]	1.67 [1.58–1.81]	0.361
Prior PCI history	62 (34.3)	19 (37.3)	36 (40.9)	0.565
Smoking history	120 (66.3)	43 (84.3)	55 (62.5)	0.021
Hypertension	142 (78.5)	37 (72.5)	67 (76.1)	0.665
Diabetes	77 (42.5)	25 (49.0)	35 (39.8)	0.565
Dyslipidemia	101 (55.8)	25 (49.0)	49 (55.7)	0.675
Family history of CAD	19 (10.5)	7 (13.7)	13 (15.3)	0.564
Medication
Aspirin	179 (98.9)	54 (100.0)	85 (100.0)	0.462
Statin	165 (91.2)	48 (94.1)	84 (95.5)	0.408
β-blocker	127 (70.2)	39 (76.5)	67 (76.1)	0.478
ACEi/ARB	135 (74.6)	40 (78.4)	66 (75.0)	0.851
Calcium channel blocker	74 (40.9)	22 (43.1)	29 (33.0)	0.370
Nitrate	31 (17.1)	3 (5.9)	15 (17.0)	0.125
Presentation
PCI target, RCA/LAD/LCx	46/106/29 (25.4/58.6/16.0)	8/32/11 (15.7/62.7/21.6)	21/56/14 (23.9/60.2/15.9)	0.639
Total SYNTAX score	9 [6–13]	9 [6–15]	10 [7–17]	0.148
MACCE	10 (5.5)	3 (5.9)	13 (14.8)	0.027
Laboratory data
CRP (mg/dL)	0.05 [0.03–0.18]	0.13 [0.04–0.26]	0.08 [0.03–0.18]	0.017
Creatinine (mg/dL)	0.82 [0.72–0.94]	0.91 [0.75–1.00]	0.88 [0.75–1.01]	0.038
Total cholesterol (mg/dL)	156 [140–186]	153 [135–177]	161 [143–193]	0.400
LDL-C (mg/dL)	86 [68–111]	84 [65–105]	86 [69–102]	0.642
HDL-C (mg/dL)	48 [42–59]	47 [41–57]	48 [41–56]	0.799
Triglyceride (mg/dL)	125 [89–162]	125 [94–148]	129 [87–181]	0.707
NT-proBNP (ng/L)	119 [47–413]	215 [87–520]	205 [89–447]	0.008
HbA1c (%)	6.2 [5.7–6.9]	6.4 [5.8–7.2]	6.3 [5.7–6.9]	0.515
Hemoglobin (g/dL)	13.5 [12.5–14.3]	13.6 [12.2–14.3]	13.3 [12.2–14.3]	0.497
CMR parameters
End-diastolic LV volume (mL)	121 [101–138]	127 [104–153]	123 [107–146]	0.186
End-systolic LV volume (mL)	46 [33–60]	47 [27–70]	47 [38–67]	0.270
EF (%)	61.9 [54.2–70.7]	63.9 [48.7–75.7]	59.5 [49.2–69.2]	0.094
LV mass (mL)	122 [99–144]	141 [119–165]	117 [100–137]	<0.001
Resting CSF (mL/min)	89.2 [57.0–125.4]	99.1 [85.1–135.1]	158.9 [121.6–193.0]	<0.001
Corrected resting CSF (mL/min)	98.0 [68.2–132.3]	115.9 [95.3–138.5]	189.0 [161.9–243.6]	<0.001
Corrected resting CSF (mL/min/g)	0.80 [0.55–1.09]	0.80 [0.69–0.90]	1.64 [1.30–2.02]	<0.001
Hyperemic CSF (mL/min)	317.7 [236.8–409.1]	172.0 [151.6–214.3]	271.1 [225.2–345.9]	<0.001
Hyperemic CSF (mL/min/g)	2.62 [1.79–3.55]	1.20 [1.05–1.48]	2.29 [1.80–2.98]	<0.001
G-CFR	2.95 [2.44–4.18]	1.59 [1.35–1.75]	1.47 [1.15–1.76]	<0.001
Prognosis
MACCE	10 (5.5)	3 (5.9)	13 (14.8)	0.019
Cardiac death	5 (2.8)	2 (3.9)	7 (8.0)
Myocardial infarction	6 (3.3)	3 (5.9)	3 (3.4)
Heart failure	0 (0.0)	0 (0.0)	2 (2.3)
Cerebral infarction	2 (1.1)	0 (0.0)	2 (2.3)

Unless indicated otherwise, data are given as median [interquartile range] or n (%). E-CMD, endogenous-type coronary microvascular dysfunction; eGFR, estimated glomerular filtration rate. Other abbreviations sa in Table 1.

CMR Parameters and Prognosis

During the follow-up period (median 2.5 years; IQR 1.3–4.0 years), MACCE occurred in 26 (8.1%) patients. Patients’ baseline characteristics and CMR data according to the presence or absence of MACCE are presented in Table 4. Compared with patients without MACCE, those with MACCE had a lower post-PCI G-CFR (median 2.22 [IQR 1.64–3.11] vs. 1.71 [IQR 1.35–2.42]; P=0.012), higher post-PCI resting CSF (median 0.98 [IQR 0.66–1.39] vs. 1.19 [IQR 0.90–1.65] mL/min/g; P=0.047), higher HbA1c, higher NT-proBNP, higher end-systolic LV volume, and lower EF. Hyperemic CSF did not differ significantly between the 2 groups. Patients with post-PCI CMD had a higher prevalence of MACCE (11.5% vs. 5.5%; P=0.063). In addition, the prevalence of MACCE was significantly higher in the E-CMD than non-CMD and non-E-CMD groups, and this difference was statistically among the groups. (14.8% vs. 5.5% and 5.9%, respectively; P=0.027; Table 3).

Table 4.

Patient Characteristics According to the Occurrence of MACCE

	All patients (n=320)	Patients without MACCE (n=294)	Patients with MACCE (n=26)	P value
Demographics
Age (years)	70 [61–75]	70 [61–75]	73 [67–76]	0.189
Sex				0.565
Male	272 (85.0)	251 (85.4)	21 (80.8)
Female	48 (15.0)	43 (15.6)	5 (19.2)
Body surface area (m2)	1.71 [1.61–1.82]	1.71 [1.61–1.81]	1.70 [1.56–1.82]	0.637
Hypertension	246 (76.9)	224 (76.2)	22 (84.6)	0.467
Diabetes	137 (42.8)	121 (41.2)	16 (61.5)	0.061
Dyslipidemia	175 (54.7)	158 (53.7)	17 (65.4)	0.307
Family history of CAD	39 (12.2)	37 (12.6)	2 (7.7)	0.754
Smoking history	218 (68.1)	198 (67.3)	20 (76.9)	0.384
Prior PCI history	117 (36.6)	109 (37.1)	8 (30.8)	0.672
Target vessel RCA/LAD/LCx	75/191/54 (23.4/59.7/16.9)	67/177/50 (22.8/60.2/17.0)	7/15/4 (26.9/57.7/15.4)	0.881
Medication
Aspirin	318 (99.4)	292 (99.3)	26 (100.0)	1.000
Statin	297 (92.8)	275 (93.5)	22 (84.6)	0.104
β-blocker	233 (72.8)	214 (72.8)	19 (73.1)	1.000
ACEi/ARB	241 (75.3)	221 (75.2)	20 (76.9)	1.000
Calcium channel blocker	125 (39.1)	116 (39.5)	9 (34.6)	0.680
Nitrate	49 (15.3)	46 (15.6)	3 (11.5)	0.779
Laboratory data
CRP (mg/dL)	0.07 [0.03–0.20]	0.07 [0.03–0.19]	0.12 [0.04–0.30]	0.114
Creatinine (mg/dL)	0.85 [0.73–0.97]	0.85 [0.72–0.97]	0.82 [0.74–1.01]	0.887
eGFR	67.6 [56.9–80.1]	67.5 [57.2–80.1]	67.7 [51.9–77.8]	0.481
Total cholesterol (mg/dL)	157 [140–186]	157 [139–186]	166 [151–181]	0.361
LDL-C (mg/dL)	86 [68–106]	84 [68–106]	95 [82–110]	0.129
HDL-C (mg/dL)	48 [41–58]	48 [42–59]	46 [37–53]	0.138
Triglyceride (mg/dL)	127 [88–166]	127 [91–169]	123 [79–156]	0.286
NT-proBNP (ng/L)	176 [66–453]	165 [62–412]	528 [136–1,162]	0.002
HbA1c (%)	6.3 [5.7–6.9]	6.2 [5.7–6.9]	6.9 [6.1–7.6]	0.008
Hemoglobin (g/dL)	13.5 [12.4–14.4]	13.5 [12.4–14.5]	13.2 [11.4–14.2]	0.226
CMR parameters
End-diastolic LV volume (mL)	122 [103–142]	122 [103–140]	144 [109–182]	0.074
End-systolic LV volume (mL)	47 [32–64]	46 [32–62]	68 [41–102]	0.002
EF (%)	61.4 [51.7–70.8]	61.9 [52.7–71.2]	53.2 [39.6–61.4]	0.001
LV mass (mL)	124.0 [102.8–148.0]	123.0 [101.0–147.8]	132.0 [108.5–151.0]	0.238
Resting CSF (mL/min)	108.9 [73.6–153.6]	105.7 [72.3–148.7]	152.2 [105.9–171.8]	0.007
Corrected resting CSF (mL/min)	123.4 [86.6–170.9]	118.9 [84.7–168.0]	163.2 [130.0–181.0]	0.005
Corrected resting CSF (mL/min/g)	1.00 [0.67–1.41]	0.98 [0.66–1.39]	1.19 [0.90–1.65]	0.047
Hyperemic CSF (mL/min)	272.0 [204.6–372.4]	271.4 [204.5–370.8]	295.2 [214.0–385.4]	0.702
Hyperemic CSF (mL/min/g)	2.25 [1.53–3.11]	2.26 [1.52–3.13]	2.10 [1.56–2.97]	0.631
G-CFR	2.17 [1.62–3.04]	2.22 [1.64–3.11]	1.71 [1.35–2.42]	0.012
Post-PCI CMD	139 (43.4)	123 (41.8)	16 (61.5)	0.063
Non-E-CMD	51 (15.9)	48 (16.3)	3 (11.5)	0.780
E-CMD	88 (27.5)	75 (25.5)	13 (50.0)	0.011

Unless indicated otherwise, data are given as the mean±SD, median [interquartile range], or n (%). Abbreviations sa in Tables 1,3.

Univariable analysis using the Cox proportional hazard model investigating factors predicting MACCE revealed that the presence of post-PCI CMD, the presence of E-CMD, higher post-PCI resting CSF, lower post-PCI G-CFR, higher HbA1c, higher NT-proBNP, higher end-diastolic LV volume, higher end-systolic LV volume, and lower EF were significantly associated with the occurrence of MACCE (Table 5). In multivariable analysis, the presence of E-CMD (hazard ratio [HR] 3.24; 95% CI 1.47–7.14; P=0.004), HbA1c (HR 1.36; 95% CI 1.10–1.68; P=0.005), and end-systolic LV volume (HR 1.01; 95% CI 1.01–1.02; P<0.001) were independent significant factors for predicting MACCE. Kaplan-Meier analysis revealed that post-PCI CMD was significantly associated with the occurrence of MACCE (log-rank test, P=0.035; Figure 2). When CMD patients were further stratified into non-E-CMD and E-CMD groups, patients with E-CMD showed worse prognosis (P=0.025; Figure 3).

Table 5.

Predictors of MACCE (Cox Proportional Hazard Model)

	Univariable analysis		Multivariable analysis
	HR (95% CI)	P value	HR (95% CI)	P value
Diabetes	2.14 (0.97–4.73)	0.059
HbA1c	1.38 (1.14–1.68)	0.001	1.36 (1.10–1.68)	0.005
NT-proBNP	1.00 (1.00–1.00)	0.004
End-diastolic LV volume	1.01 (1.00–1.02)	0.002
End-systolic LV volume	1.01 (1.01–1.02)	<0.001	1.01 (1.01–1.02)	<0.001
EF	0.95 (0.93–0.98)	<0.001
Resting CSF (mL/min)	1.00 (1.00–1.01)	0.014
Corrected resting CSF (mL/min)	1.00 (1.00–1.00)	0.076
Corrected resting CSF (mL/min/g)	1.14 (0.83–1.57)	0.428
G-CFR	0.65 (0.44–0.96)	0.032
Presence of post-PCI CMD	2.28 (1.04–5.04)	0.041
E-CMD	2.77 (1.28–5.97)	0.010	3.24 (1.47–7.14)	0.004

HR, hazard ratio. Other abbreviations sa in Tables 1–3.

Figure 2.

Kaplan-Meier analysis of major adverse cardiac and cerebrovascular events based on the presence or absence of coronary microvascular dysfunction (CMD) after percutaneous coronary intervention (PCI). The cut-off value for post-PCI global coronary flow reserve (G-CFR) was 2.0. Patients with post-PCI CMD had a worse prognosis.

Figure 3.

Kaplan-Meier analysis of major adverse cardiac and cerebrovascular events based on the presence or absence of coronary microvascular dysfunction (CMD) after percutaneous coronary intervention (PCI) and CMD subtype. The cut-off value for elevated resting coronary sinus blood flow (CSF) was 1.06, derived from the 57th percentile of resting CSF measurements. Endogenous-type CMD (E-CMD) was defined as low (<2.0) global coronary flow reserve (G-CFR) and high (>1.06) resting CSF; non-E-CMD was defined as low (<2.0) G-CFR and low (<1.06) resting CSF. Patients with E-CMD had a significantly worse prognosis.

Discussion

The number of studies reporting adverse prognosis associated with CMD with decreased G-CFR is increasing.^17–19 However, the prevalence of CMD after elective PCI and its clinical significance remain unknown. In addition, recent attention has been drawn to E-CMD, characterized by decreased CFR due to elevated resting MBF. The prevalence of E-CMD and its link to prognosis after elective PCI have not been elucidated. The aim of this study was to elucidate the prevalence of post-PCI CMD, particularly focusing on E-CMD and non-E-CMD, to explore the prognostic value of post-PCI CMD and E-CMD. The primary findings of this study are that, in patients with CCS undergoing elective and uncomplicated PCI: (1) the presence of post-PCI CMD was significantly associated with pre-PCI creatinine, NT-proBNP levels, and SYNTAX scores; (2) post-PCI CMD was observed in 43.4% of patients, 63.3% of whom had E-CMD; (3) post-PCI CMD was significantly associated with the occurrence of MACCE; and (4) E-CMD provided more prognostic information than non-E-CMD. To the best of our knowledge, the present study is the first to demonstrate an association between worse clinical outcomes and CMD following elective PCI assessed using PC-CMR-derived G-CFR and resting absolute MBF, particularly suggesting a significant association between E-CMD and worse outcomes.

Relationship Between Post-PCI CMD and Prognosis

Recently, CMD has been increasingly recognized as a significant factor for the symptoms of myocardial ischemia and prognosis.²⁰ G-CFR serves as an indicator of CMD, and there is a relationship between significant stenosis of the epicardial coronary artery and G-CFR values,²¹ although G-CFR as determined by PET has been reported to provide prognostication regardless of the presence or absence of epicardial coronary stenosis.²² The risk factors for CMD share much in common with those for epicardial coronary atherosclerosis,²³ suggesting the coexistence of epicardial coronary artery disease and CMD. Low G-CFR values after elective PCI without residual significant epicardial stenosis suggest the presence of CMD, indicating the possibility that despite the successful PCI by alleviating stenosis in the epicardial coronary artery, pre-existing CMD may prevent adequate coronary flow and myocardial perfusion restoration, potentially leading to a residual post-PCI risk. Post-PCI CMD may preclude patients from benefitting from PCI, highlighting the importance of predicting coexisting microvascular dysfunction in advance for appropriate assessment of PCI indications and its prognostication after PCI. The results of our study indicate that creatinine, NT-proBNP levels, and SYNTAX scores are significant predictors of post-PCI CMD. However, due to the small sample size, our study falls short of adequately predicting post-PCI CMD, particularly E-CMD.

Our group previously reported that a worse prognosis was significantly associated with lower hyperemic CSF and G-CFR after acute MI (AMI).²⁴ Lower G-CFR and lower hyperemic CSF may indicate the presence of structural CMD, although myocardial damage due to MI could potentially induce structural CMD and reflect its severity, at least at the AMI territory. In contrast, the present study focused on successful PCI for CCS, which may explain the differences in coronary flow characteristics compared with post-AMI studies. Nevertheless, Kanaji et al. also observed a trend for higher resting CSF in the event group.²⁴ This may suggest common features of CMD that affect prognosis in both the previous and present study populations, although further large and comprehensive studies including CCS and MI cohorts are needed to confirm these hypotheses.

E-CMD and Prognosis

Hyperemic MBF serves as an assessment of the total integrated function of the coronary circulation, encompassing both the microvasculature and the epicardial conduit vessels.⁸ Traditionally, focus has been placed on conventional CMD, characterized primarily by a reduction in hyperemic MBF.⁷ However, recent attention has also focused on E-CMD, which is characterized by a decrease in CFR primarily attributable to elevated resting MBF.^8,25 Elevated resting coronary flow is associated with an increased myocardial workload required to meet oxygen and metabolic demands, which can be influenced by factors such as arterial hypertension, hyperlipidemia, diabetes, anemia, increased myocardial contractility, LV wall stress, and sympathetic activation.^9,26,27 In the present study, no significant association was observed between some of these systemic factors and E-CMD; however, a potential relationship cannot be ruled out.

Previous studies showed women tend to have higher resting MBF.^26,28 In the present study, the prevalence of women in the E-CMD group was relatively higher than that in the non-E-CMD group, although the difference was not statistically significant (P=0.199). However, it is considered that this study was statistically underpowered due to the relatively small proportion of women included.

Although the association between elevated resting MBF and poor clinical outcomes was reported previously,¹¹ the prognosis of E-CMD after PCI has not been reported and remains unknown. Our study suggests a worse prognosis of post-PCI CMD, particularly noting that post-PCI E-CMD was associated with adverse clinical outcomes, whereas the prognosis of non-E-CMD did not differ from that of the group with normal microvascular function (i.e., no CMD). This finding suggests that more aggressive treatment or new interventions may be needed, particularly for patients with E-CMD after PCI. Furthermore, our results may indicate the importance of CMD, including E-CMD, in consideration of outcomes after uncomplicated PCI. Further studies with large sample sizes are needed to test our hypothesis-generating results and clarify specific therapeutic strategies targeting E-CMD post-PCI.

Study Limitations

This study has inherent limitations associated with it being a single-center retrospective analysis of an observational nature. The rigorous exclusion criteria and CMR protocol limited the number of patients included and may have resulted in a certain level of selection bias. Patients were enrolled considering contraindications to CMR and the importance of electrocardiogram gating, which led to further selection bias because there were no patients with metallic device implants, bronchospasm, claustrophobia, or atrioventricular block. Although we aimed to perform CMR post-PCI when possible, due to economic constraints and patient preferences, CMR was not performed in all PCI patients during the study period. This study also included patients with physiologically significant epicardial coronary stenosis treated with PCI. This characteristic, including a low proportion of women and the prevalence of high-risk patients, suggests that our population may differ from population in previous studies reporting on resting coronary flow. It must be taken into consideration that the small number of MACCE, which precluded extensive subgroup analyses, is an important limitation of this study. In addition, some cases in this study had an undetermined cause of death, and detailed post-mortem examinations were not conducted, which is a notable limitation of this study. E-CMD may be associated with systemic and non-cardiac disorders, resulting in worse prognosis. Further studies are required to investigate the association between E-CMD and subsequent cardiac and non-cardiac events. However, we limited MACCE only to hard endpoints in this study, indicating the potentially important clinical impact. Finally, there is no established cut-off value of resting CSF to define E-CMD, and the cut-off value we used in this study was arbitrary.

Conclusions

CMD was observed in 43.4% of patients after elective PCI and E-CMD, characterized by increased resting coronary flow, was observed in 63.3% of patients with post-PCI CMD. Post-PCI CMD was associated with worse prognoses, particularly indicating the association between E-CMD and worse clinical outcomes. The findings of this study suggest that it may be necessary to consider the possible coexistence of CMD with epicardial coronary disease and its prognostication after PCI in CCS patients. Further large studies appear to be warranted to develop personalized therapeutic strategies targeting post-PCI CMD, particularly E-CMD.

Acknowledgments

None.

Sources of Funding

This study did not receive any specific funding.

Disclosures

None.

IRB Information

This study was approved by the Ethics Committee of Tsuchiura Kyodo General Hospital (Approval no. TKGH-IRB 2021FY16).

References

• 1.   Maron DJ, Hochman JS, Reynolds HR, Bangalore S, O’Brien SM, Boden WE, et al. Initial invasive or conservative strategy for stable coronary disease. N Engl J Med 2020; 382: 1395–1407.
• 2.   Boerhout CKM, de Waard GA, Lee JM, Mejia-Renteria H, Lee SH, Jung JH, et al. Prognostic value of structural and functional coronary microvascular dysfunction in patients with non-obstructive coronary artery disease; from the multicentre international ILIAS registry. EuroIntervention 2022; 18: 719–728.
• 3.   Kim J, Shin D, Lee JM, Lee SH, Hong D, Choi KH, et al. Differential prognostic value of revascularization for coronary stenosis with intermediate FFR by coronary flow reserve. JACC Cardiovasc Interv 2022; 15: 1033–1043.
• 4.   Lee JM, Choi KH, Hwang D, Park J, Jung JH, Kim HY, et al. Prognostic implication of thermodilution coronary flow reserve in patients undergoing fractional flow reserve measurement. JACC Cardiovasc Interv 2018; 11: 1423–1433.
• 5.   Kato S, Saito N, Nakachi T, Fukui K, Iwasawa T, Taguri M, et al. Stress perfusion coronary flow reserve versus cardiac magnetic resonance for known or suspected CAD. J Am Coll Cardiol 2017; 70: 869–879.
• 6.   Kikuchi Y, Naya M, Oyama-Manabe N, Manabe O, Sugimori H, Kudo K, et al. Assessment of coronary flow velocity reserve in the left main trunk using phase-contrast MR imaging at 3T: Comparison with ¹⁵O-labeled water positron emission tomography. Magn Reson Med Sci 2019; 18: 134–141.
• 7.   Gupta A, Taqueti VR, van de Hoef TP, Bajaj NS, Bravo PE, Murthy VL, et al. Integrated noninvasive physiological assessment of coronary circulatory function and impact on cardiovascular mortality in patients with stable coronary artery disease. Circulation 2017; 136: 2325–2336, doi:10.1161/CIRCULATIONAHA.117.029992.
• 8.   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: 536–548.
• 9.   Schindler TH, Zhang XL, Vincenti G, Mhiri L, Lerch R, Schelbert HR. Role of PET in the evaluation and understanding of coronary physiology. J Nucl Cardiol 2007; 14: 589–603.
• 10.   de Winter RW, Jukema RA, van Diemen PA, Schumacher SP, Somsen YBO, van de Hoef TP, et al. Prognostic value of modified coronary flow capacity derived from [¹⁵O]H₂O positron emission tomography perfusion imaging. Circ Cardiovasc Imaging 2023; 16: e014845.
• 11.   Guerraty MA, Rao HS, Anjan VY, Szapary H, Mankoff DA, Pryma DA, et al. The role of resting myocardial blood flow and myocardial blood flow reserve as a predictor of major adverse cardiovascular outcomes. PLoS One 2020; 15: e0228931.
• 12.   Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, et al. Fourth universal definition of myocardial infarction (2018). Circulation 2018; 138: e618–e651.
• 13.   Nakano S, Kohsaka S, Chikamori T, Fukushima K, Kobayashi Y, Kozuma K, et al. JCS 2022 guideline focused update on diagnosis and treatment in patients with stable coronary artery disease. Circ J 2022; 86: 882–915.
• 14.   Nakamura M, Yaku H, Ako J, Arai H, Asai T, Chikamori T, et al. JCS/JSCVS 2018 guideline on revascularization of stable coronary artery disease. Circ J 2022; 86: 477–588.
• 15.   Kanaji Y, Murai T, Yonetsu T, Usui E, Araki M, Matsuda J, et al. Effect of elective percutaneous coronary intervention on hyperemic absolute coronary blood flow volume and microvascular resistance. Circ Cardiovasc Interv 2017; 10: e005073.
• 16.   Hamaya R, van de Hoef TP, Lee JM, Hoshino M, Kanaji Y, Murai T, et al. Differential impact of coronary revascularization on long-term clinical outcome according to coronary flow characteristics: Analysis of the international ILIAS registry. Circ Cardiovasc Interv 2022; 15: e011948.
• 17.   Indorkar R, Kwong RY, Romano S, White BE, Chia RC, Trybula M, et al. Global coronary flow reserve measured during stress cardiac magnetic resonance imaging is an independent predictor of adverse cardiovascular events. JACC Cardiovasc Imaging 2019; 12: 1686–1695, doi:10.1016/j.jcmg.2018.08.018.
• 18.   Herzog BA, Husmann L, Valenta I, Gaemperli O, Siegrist PT, Tay FM, et al. Long-term prognostic value of ¹³N-ammonia myocardial perfusion positron emission tomography. Added value of coronary flow reserve. J Am Coll Cardiol 2009; 54: 150–156, doi:10.1016/j.jacc.2009.02.069.
• 19.   Bom MJ, Van Diemen PA, Driessen RS, Everaars H, Schumacher SP, Wijmenga JT, et al. Prognostic value of [¹⁵O]H₂O positron emission tomography-derived global and regional myocardial perfusion. Eur Heart J Cardiovasc Imaging 2020; 21: 777–786, doi:10.1093/ehjci/jez258.
• 20.   Crea F, Camici PG, Merz CNB. Coronary microvascular dysfunction: An update. Eur Heart J 2014; 35: 1101–1111.
• 21.   Sampson UK, Dorbala S, Limaye A, Kwong R, Di Carli MF. Diagnostic accuracy of rubidium-82 myocardial perfusion imaging with hybrid positron emission tomography/computed tomography in the detection of coronary artery disease. J Am Coll Cardiol 2007; 49: 1052–1058.
• 22.   Taqueti VR, Hachamovitch R, Murthy VL, Naya M, Foster CR, Hainer J, et al. Global coronary flow reserve is associated with adverse cardiovascular events independently of luminal angiographic severity and modifies the effect of early revascularization. Circulation 2015; 131: 19–27, doi:10.1161/CIRCULATIONAHA.114.011939.
• 23.   Upadhyaya A, Bhandiwad A, Lang J, Sadhu JS, Barrs C, Jain S, et al. Coronary circulatory function with increasing obesity: A complex U-turn. Eur J Clin Invest 2022; 52: e13755.
• 24.   Kanaji Y, Sugiyama T, Hoshino M, Yasui Y, Nogami K, Ueno H, et al. Prognostic value of coronary sinus flow quantification by cardiac magnetic resonance imaging in patients with acute myocardial infarction. J Am Heart Assoc 2022; 11: e023519.
• 25.   Bhandiwad AR, Valenta I, Jain S, Schindler TH. PET-determined prevalence of coronary microvascular dysfunction and different types in a cardio-metabolic risk population. Int J Cardiol Heart Vasc 2023; 46: 101206.
• 26.   Chareonthaitawee P, Kaufmann PA, Rimoldi O, Camici PG. Heterogeneity of resting and hyperemic myocardial blood flow in healthy humans. Cardiovasc Res 2001; 50: 151–161.
• 27.   Murai T, Hikita H, Yamaguchi M, Ito A, Warisawa T, Ikeda H, et al. Basal coronary microvascular resistance predicting death and heart failure in patients without functional coronary stenosis. Circ J 2024; 88: 1788–1797, doi:10.1253/circj.CJ-24-0022.
• 28.   Kobayashi Y, Fearon WF, Honda Y, Tanaka S, Pargaonkar V, Fitzgerald PJ, et al. Effect of sex differences on invasive measures of coronary microvascular dysfunction in patients with angina in the absence of obstructive coronary artery disease. JACC Cardiovasc Interv 2015; 8: 1433–1441.