Prognostic Impact of Infarct Size and Coronary Microvascular Function Via Cardiovascular Magnetic Resonance and Coronary Angiography‐Derived Index of Microcirculatory Resistance in Myocardial Infarction With Nonobstructive Coronary Arteries

Abstract

Background

Infarct size (IS) and microvascular obstruction (MVO), measured by cardiovascular magnetic resonance, are key prognostic markers after myocardial infarction (MI). The index of microcirculatory resistance (IMR) predicts MVO and IS; however, its role in MI with nonobstructive coronary arteries (MINOCA) is unclear. This study explored the relationship between coronary angiography‐derived IMR, MVO, and IS in true MINOCA and assessed the prognostic value of coronary microvascular dysfunction (CMD), IS, and MVO.

Methods

Patients with suspected MINOCA who completed cardiovascular magnetic resonance and coronary angiography‐derived IMR were included. CMD was defined as coronary angiography‐derived IMR >25, and IS as a percentage of left ventricular mass (IS%LV) measured by late gadolinium enhancement. True MINOCA was identified by cardiovascular magnetic resonance evidence of ischemia or infarction. Individuals with true MINOCA were divided into 3 groups: non‐CMD and IS%LV <16.78%, CMD or IS%LV ≥16.78%, and CMD with IS%LV ≥16.78%. Follow‐up for major adverse cardiovascular events was conducted.

Results

Of 317 patients, 102 had true MINOCA (91.2% non–ST‐segment–elevation MI). CMD was similarly distributed across all suspected MINOCA. Coronary angiography‐derived IMR did not predict MVO or IS among true MINOCA. Over a median 27 months, 33.3% of patients with true MINOCA experienced major adverse cardiovascular events, with a higher risk in patients with CMD or IS%LV ≥16.78%. Multivariate Cox analysis revealed that CMD combined with IS%LV ≥16.78% had a hazard ratio (HR) of 7.40 (95% CI, 1.94–28.23), and either CMD or IS%LV ≥16.78% had an HR of 8.77 (95% CI, 2.57–29.94), but MVO did not. Sensitivity analyses excluding ST‐segment–elevation MI supported these findings.

Conclusions

CMD with larger IS strongly predicts major adverse cardiovascular events in MINOCA, underscoring their utility in risk stratification, especially for non–ST‐segment–elevation MI.

REGISTRATION

URL: https://clinicaltrials.gov. Unique identifier: NCT06502899.

Nonstandard Abbreviations and Acronyms

caIMR

coronary angiography‐derived index of microcirculatory resistance

CMD

coronary microvascular dysfunction

IS

infarct size

LGE

late gadolinium enhancement

MACE

major adverse cardiovascular events

MINOCA

myocardial infarction with nonobstructive coronary arteries

MVO

microvascular obstruction

Clinical Perspective.

What Is New?

• This study is the first to demonstrate that, among patients with true myocardial infarction with nonobstructive coronary arteries, defined by cardiac magnetic resonance evidence of ischemia or infarction, the combination of coronary microvascular dysfunction and large infarct size (percentage of left ventricular mass) is strongly associated with an increased risk of major adverse cardiovascular events, particularly among those with non–ST‐segment–elevation myocardial infarction; in contrast, coronary angiography‐derived index of microcirculatory resistance was not predictive of infarct characteristics.

What Are the Clinical Implications?

• Cardiac magnetic resonance plays a critical role in diagnosing true myocardial infarction with nonobstructive coronary arteries by identifying ischemia or infarction and quantifying infarct size, while also helping to distinguish it from other conditions such as myocarditis, Takotsubo syndrome, or structural cardiomyopathies; in this study, coronary microvascular dysfunction and a large infarct size provided important insights into the underlying pathophysiological burden in true myocardial infarction with nonobstructive coronary arteries.

• Coronary microvascular dysfunction and infarct size, as measured by coronary angiography‐derived index of microcirculatory resistance and cardiac magnetic resonance, were independently associated with a higher risk of major adverse cardiovascular events; these findings highlight the potential value of incorporating coronary microvascular dysfunction assessment and infarct quantification into routine evaluation to improve risk stratification and guide clinical management in patients with true myocardial infarction with nonobstructive coronary arteries.

Myocardial infarction (MI) with nonobstructive coronary arteries (MINOCA) presents a unique clinical challenge, accounting for ∼6% to 15% of all MI cases. ¹ , ² Unlike typical MI, MINOCA is characterized by the absence of significant coronary artery stenosis (<50%) on coronary angiography. ² , ³ MINOCA is considered a working diagnosis that includes a heterogeneous group of conditions with ischemic causes, such as plaque disruption, coronary artery spasm, coronary embolism, spontaneous coronary artery dissection, and coronary microvascular dysfunction (CMD), as well as nonischemic causes like myocarditis, Takotsubo syndrome (TTS), and nonischemic cardiomyopathy, requiring a specific diagnostic algorithm to differentiate it from other diagnoses. ¹ Importantly, suspected MINOCA includes both ischemic and nonischemic myocardial mimics, whereas true MINOCA is confirmed when an ischemic cause is identified by cardiovascular magnetic resonance (CMR). MINOCA is not a benign condition, with studies showing increased risk of death and recurrent MI. ⁴ , ⁵ Although some reports noted comparable rates of stroke‐related rehospitalization between MINOCA and obstructive coronary artery disease, ⁶ overall major adverse cardiovascular events (MACE) ⁶ and mortality ⁶ , ⁷ are generally lower in MINOCA. Therefore, understanding the mechanisms behind this syndrome is crucial, as it could guide therapies that reduce adverse outcomes for patients with MINOCA.

CMR is recommended by both the American Heart Association ⁸ and the European Society of Cardiology ⁹ as an essential imaging tool for diagnosing the underlying causes of MINOCA. CMR is valuable for clarifying the underlying cause, assessing the extent of late gadolinium enhancement (LGE) and microvascular obstruction (MVO), and guiding treatment to improve outcomes. CMR can detect MVO and infarct size (IS), which are strong predictors of adverse outcomes not only in ST‐segment–elevation MI (STEMI), ¹⁰ , ¹¹ , ¹² but also in non‐ST‐segment elevation MI (NSTEMI), where MVO, though less frequent, ¹³ , ¹⁴ remains associated with adverse prognosis. ¹⁵ However, the role of MVO and IS in MINOCA remains unclear. Meanwhile, CMD, commonly associated with MINOCA and considered a hallmark of its subtypes, ¹⁶ , ¹⁷ presents a challenge in determining whether it is a cause or a consequence of acute MI (AMI). ¹ Although CMD can be invasively evaluated through the index of microcirculatory resistance (IMR), which predicts MVO, IS, and adverse events in patients with AMI, ¹⁰ , ¹¹ , ¹⁸ , ¹⁹ its complexity limits its routine use. A noninvasive alternative, coronary angiography‐derived IMR (caIMR), which strongly correlates with invasive IMR, has emerged as a valuable tool with strong prognostic value in various cardiovascular diseases. ²⁰ , ²¹ , ²² Our recent study demonstrated that caIMR could be used for risk stratification in MINOCA, offering promise for clinical decision‐making. ²³ Moreover, caIMR has shown significant potential in predicting MVO and IS in STEMI without requiring invasive measures. ²⁴ , ²⁵ Nonetheless, the relationship of caIMR with MVO and IS in patients with true MINOCA remains unexplored.

The primary objective of this study was to evaluate the prognostic value of CMD, IS, and MVO using CMR and caIMR in patients with true MINOCA. The secondary objective was to explore the distribution of CMD across different causes of suspected MINOCA and to examine the relationship between caIMR, MVO, and IS in patients with true MINOCA, enhancing the understanding of MINOCA pathophysiology.

METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study Design and Population

This single‐center retrospective study, registered on ClinicalTrials.gov (ID: NCT06502899), enrolled patients with suspected AMI who underwent coronary angiography at the Affiliated Hospital of Xuzhou Medical University between January 2019 and June 2024. The inclusion criteria were as follows: (1) criteria for AMI included hs‐TnT (high‐sensitivity troponin T) levels exceeding the 99th percentile, along with supporting clinical evidence of AMI as outlined in the Fourth Universal Definition of MI ²⁶ ; (2) coronary angiography results showing <50% coronary artery stenosis; (3) completion of a CMR including LGE imaging examination during hospitalization; and (4) patients proceeded with caIMR assessments to evaluate microvascular function. Exclusion criteria involved missing or low‐quality CMR images and absent caIMR data. After exclusion, this study enrolled a total of 317 participants, including 102 cases of true MINOCA, 57 cases with normal findings, and 158 cases of nonischemic cause (43 cases of TTS, 79 cases of myocarditis, and 36 cases of cardiomyopathy) (Figure 1).

Figure 1. Flow chart of eligible patients.

AMI indicates acute myocardial infarction; caIMR, coronary angiography‐derived index of microvascular resistance; CMD, coronary microvascular dysfunction; CMR, cardiovascular magnetic resonance; IS%LV, infarct size left ventricular; MINOCA, myocardial infarction with nonobstructive coronary arteries; and TTS, Takotsubo syndrome.

This study received approval from the Institutional Review Board of the Affiliated Hospital of Xuzhou Medical University. Given its retrospective nature and minimal risk to patients, the board waived the need for informed consent.

Baseline clinical data for all patients, which included demographic information such as age and sex, as well as cardiovascular risk factors, including smoking, hypertension, diabetes, and a history of stroke, were collected from electronic medical records. Discharge medications were also documented, covering antiplatelet agents, statins, renin‐angiotensin‐aldosterone system inhibitors, and β‐blockers. During hospitalization, blood samples were taken for standard analyses, including NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide), hsTnT, fasting blood glucose, triglycerides, total cholesterol, and low‐ and high‐density lipoprotein cholesterol. The peak levels of NT‐proBNP and hsTnT were recorded during the hospital stay.

CMR Protocol and Reclassification

CMR image analysis was independently performed by 2 experienced, blinded physicians, with each patient’s images evaluated twice at separate intervals. CMR scans were performed using a 3.0 T system (Ingenia, Philips, The Netherlands), following previously published protocols that detail the CMR methodology and related parameters. ²⁷ A balanced turbo field echo sequence was used, with the following scan parameters: slice thickness = 7 mm, no interlayer gap; echo time = 1.47 ms, repetition time = 2.94 ms; flip angle = 60°; field of view = 300 × 300 mm; matrix = 280 × 240 mm; and voxel size = 1.22 × 1.22 × 8.0 mm³. Cine imaging, LGE, and T2 or T1 imaging were acquired during the scan. The images were analyzed using CVI42 software (cvi42 version 5.13.5, Circle Cardiovascular Imaging, Canada).

Left ventricular (LV) measurements, including end‐diastolic volume, end‐systolic volume, and ejection fraction (LVEF), were automatically calculated and adjusted for body surface area. The left atrial (LA) structure was analyzed by automatic tracing of the LA border, with atrial volume determined using the biplane method: LA volume = (0.85 × 4‐chamber LA area × 2‐chamber LA area)/shortest length of the LA vertical axis on 2 or 4 chambers, corrected for body surface area. LA EF was calculated as [(LAmax − LAmin)/LAmax] × 100%. Feature tracking using CMR‐feature tracking was applied to measure myocardial deformation.

The extent of LGE% was calculated as a percentage of the total LV myocardial mass, and IS was expressed as a percentage of the LV mass (IS%LV) based on LGE measurements. In this study, IS%LV was determined based on the cutoff value of LGE%, which was obtained by optimizing the specificity and sensitivity percentages using the Youden index. Figure 2A shows the presence of LGE among patients with MINOCA, highlighting areas of myocardial scarring detected through CMR. MVO was identified on LGE images as regions of hypoenhancement within areas of hyperenhancement ²⁸ (Figure 2B).

Figure 2. Cardiac imaging and functional assessment in MINOCA.

A, LGE demonstrating areas of myocardial scar. B, MVO appearing as hypoenhanced regions within hyperenhanced LGE areas. C, caIMR illustrating microvascular function assessment. caIMR indicates coronary angiography‐derived index of microcirculatory resistance; LGE, late gadolinium enhancement; MINOCA, myocardial infarction with nonobstructive coronary arteries; and MVO, microvascular obstruction.

The reclassification process was carried out by assessing wall motion abnormalities identified using CMR, LGE, T2‐weighted, or T1 imaging techniques. Patients were then grouped into the following categories: “true” MINOCA, myocarditis, TTS, cardiomyopathy (including subtypes like dilated, hypertrophic, and arrhythmogenic cardiomyopathies), and normal findings. These categories were defined as follows: true MINOCA was identified by the presence of LGE or myocardial edema localized in the subendocardial or transmural regions within a coronary artery territory. ²⁹ Myocarditis was diagnosed according to the 2018 Lake Louise criteria. ³⁰ TTS was characterized by a focal wall motion abnormality outside a specific vascular region, typically seen in the mid‐ or apical myocardium, with associated myocardial edema but without LGE. Cardiomyopathies, including dilated, hypertrophic, and arrhythmogenic types, were classified according to relevant clinical guidelines. ³¹ A “normal” classification was assigned when no abnormalities were found in T2‐weighted, T1 imaging, or LGE.

caIMR Measurement

The caIMR was employed to assess coronary microvascular function using the FlashAngio system, which includes the FlashAngio console, software, and Flash Pressure transducer (Rainmed Ltd., Suzhou, China). Certified analysts conducted the caIMR analysis, and the measurement procedures have been thoroughly described by Yong Huo’s team ²¹ and in our recent study. ²³ The caIMR was calculated as follows:

$$\text{caIMR}={\left(\mathit{Pd}\right)}_{\mathrm{hyp}}\frac{L}{K·V_{\text{diastole}}}$$

(Pd)_hyp refers to the mean pressure at the distal position during maximal hyperemia, L is a constant representing the length from the inlet to the distal position, V _diastole is the average flow velocity at the distal position during diastole, K is a constant (K = 2.1), and V _hyp = K·V _diastole represents the mean flow velocity at the distal position during maximal hyperemia. Figure 2C illustrates the measurement of caIMR among patients with suspected MINOCA, providing insights into the microvascular function.

Two cardiologists, blinded to baseline data and outcome information, interpreted the results of the caIMR analysis. In total, caIMR was measured in at least 1 coronary artery of 317 patients, assessing 848 coronary arteries, including the left anterior descending artery (n = 266), left circumflex artery (n = 289), and right coronary artery (n = 293). The epicardial vessel showing the highest caIMR value among the 3 major coronary arteries was selected for further evaluation. In the present study, due to the fact that most of the patients presented with NSTEMI, we adopted a caIMR threshold of >25 U to define CMD in this cohort study. ²¹ , ³² , ³³ , ³⁴

End Points

Patients were followed up after discharge through outpatient visits and telephone interviews using a standardized questionnaire. The primary follow‐up end point of this study was MACE, which included all‐cause mortality, recurrent MI, stroke, or hospital readmission for cardiovascular conditions such as heart failure, arrhythmia, or angina pectoris. Recurrent MI was diagnosed in line with the current guidelines, ²⁶ and stroke was defined as neurological dysfunction and cerebrovascular damage due to cerebral ischemia or hemorrhage. In cases where a patient had >1 event, the initial event was chosen as the combined clinical end point. For patients lost to follow‐up, their deaths were confirmed through the local official household registry.

Statistical Analysis

Statistical analysis was conducted using SPSS version 24 and GraphPad version 8.0.1. The normality of the data was assessed using the Kolmogorov–Smirnov test. Continuous variables were presented as mean ± SD for normally distributed data or as median with interquartile range for nonnormally distributed data. Normally distributed continuous variables were compared using Student’s t test, and the Mann–Whitney U test or Kruskal–Wallis test was employed for nonnormally distributed variables. Categorical variables were expressed as frequencies and percentages, and the chi‐square test was used for their analysis. The predictive value of LGE% for MACE was evaluated using the area under the curve from receiver operating characteristic curves, with the optimal cutoff point determined by the Youden index. Additionally, receiver operating characteristic curves were also used to evaluate the ability of caIMR to predict MVO and IS among true MINOCA. Restricted cubic spline regression was performed to investigate potential linear relationships between LGE% and MACE, adjusting for variables such as age, sex, body mass index, smoking, hypertension, diabetes, stroke, and LVEF level. Kaplan–Meier curves illustrated the cumulative incidence of adverse outcomes, and log‐rank tests were employed to compare group differences. Cox proportional hazard models, providing hazard ratios (HR) and 95% CI, were used to examine the association between IS and CMD with clinical outcomes. The variables considered clinically relevant or potentially associated with MACE among MINOCA were initially evaluated using univariate Cox proportional hazards analysis. Variables that demonstrated a P value of <0.1 in the univariate analysis were then included in the multivariable model using a stepwise forward selection approach to identify independent predictors. Each model included the following variables, with Model 1 including diabetes, LVEF, and IS% >16.78; Model 2 including diabetes, LVEF, LGE, and CMD; and Model 3 including diabetes, LVEF, IS%LV, and CMD. To verify the stability of our results, a sensitivity analysis was performed by excluding patients with STEMI to assess the impact of excluding these patients on the primary outcomes. Statistical significance was set at P < 0.05, with all P values derived from 2‐sided tests.

RESULTS

A total of 317 participants were enrolled in the study (mean age 53.44 ± 11.97 years, 166 [52.4%] men), with a median of 4 (interquartile range, 3–5) days between admission and CMR. Among them, there were 102 cases of true MINOCA, 43 cases of TTS, 79 cases of myocarditis, 57 cases with normal findings, and 36 cases of cardiomyopathy.

CMR Findings and Coronary Microvascular Function Among Different Suspected MINOCA Causes

Table 1 presents the CMR findings and coronary microvascular function among different suspected MINOCA causes, highlighting key differences in IS, caIMR, and CMD across the groups. Significant differences were noted in IS, as measured by LGE%, with cardiomyopathy (15.8%; interquartile range, 0–31.4) showing the largest area of myocardial fibrosis, whereas TTS and normal findings had no detectable infarct. CMD was prevalent in approximately one third of patients with suspected MINOCA and was similarly distributed across all suspected MINOCA causes, with being notably higher in cardiomyopathy cases (52.8%) compared with other causes.

Table 1.

CMR Findings and Coronary Microvascular Function Among Suspected MINOCA Causes

	True MINOCA (n = 102)	TTS (n = 43)	Myocarditis (n = 79)	Normal (n = 57)	Cardiomyopathy (n = 36)	P value
CMR findings
Global longitudinal strain, %	17.0 (15.8–19.0)	14.0 (13.0–17.0)	15.0 (13.0–19.0)	20.0 (17.0–23.5)	12.5 (10.3–15.0)	<0.001
LV‐end‐diastolic volume index, mL/m2	74.5 ± 18.1	73.8 ± 15.6	81.3 ± 22.5	72.0 ± 17.9	80.3 ± 19.2	0.02
LV‐end‐systolic volume index, mL/m2	39.1 ± 13.0	34.9 ± 8.8	47.2 ± 20.3	33.3 ± 14.0	47.0 ± 15.2	<0.001
LV ejection fraction, %	56.3 ± 9.1	53.5 ± 6.5	54.9 ± 9.9	62.9 ± 7.5	48.9 ± 9.7	<0.001
LGE%	14.0 (8.9–22.6)	0 (0–0)	15.4 (10.0–24.0)	0 (0–0)	15.8 (0–31.4)	<0.001
LGE, n (%)	100 (98.0)	0 (0.0)	75 (94.9)	0 (0.0)	19 (52.8)	<0.001
Microvascular obstruction, n (%)	7 (6.9)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0.003
Microvascular parameters
Coronary angiography‐derived index of microvascular resistance	22.0 (18.0–28.8)	23.0 (18.0–28.0)	21.0 (16.0–27.0)	21.0 (18.3–27.0)	26.2 (20.2–30.8)	0.27
Coronary microvascular dysfunction	36 (35.3)	16 (37.2)	28 (35.4)	20 (35.1)	19 (52.8)	0.40
Coronary angiography‐derived fractional flow reserve	0.93 (0.89–0.94)	0.92 (0.89–0.94)	0.92 (0.89–0.94)	0.92 (0.88–0.94)	0.92 (0.89–0.94)	0.24

Values are median (IQR), n (%), or mean ± SD.

CMR indicates cardiac magnetic resonance; LGE, late gadolinium enhancement; LV, left ventricular; MINOCA, myocardial infarction with nonobstructive coronary arteries; and TTS, Takotsubo syndrome.

Baseline Characteristics of True MINOCA According to IS and CMD

In the present study, a total of 102 patients were diagnosed with true MINOCA by CMR, with 93 (91.2%) presenting with NSTEMI, highlighting the predominance of NSTEMI in this cohort. Table 2 summarizes the baseline characteristics, medication use, CMR findings, and microvascular parameters across 3 groups: non‐CMD with IS%LV <16.78% (n = 40), CMD or IS%LV ≥16.78% (n = 47), and CMD with IS%LV ≥16.78% (n = 15). The results indicate that smoking and hypertension were less common in the group with CMD and IS%LV ≥16.78%. Laboratory results showed no significant differences in lipid profiles or other biomarkers, although NT‐proBNP, a heart strain marker, tended to be higher in the groups CMD or IS%LV ≥16.78%. Aspirin and P2Y12 inhibitors were used similarly across all groups, with no significant differences in the use of angiotensin‐converting enzyme inhibitors/angiotensin II receptor blockers/sacubitril‐valsartan. However, patients with both CMD and large IS had lower statin use compared with those without CMD and low IS. Additionally, CMR revealed worse global longitudinal strain and larger end‐systolic volume i in the groups with CMD or IS%LV ≥16.78%, with MVO more frequently observed in the group with CMD or IS%LV ≥16.78%.

Table 2.

Characteristics of True MINOCA According to Infarct Size and CMD

	Non‐CMD and IS%LV <16.78% (n = 40)	CMD or IS%LV ≥16.78% (n = 47)	CMD and IS%LV ≥16.78% (n = 15)	P value
General characteristics
Age, y	54.6 ± 10.0	55.8 ± 12.0	58.2 ± 11.1	0.56
Male sex, n (%)	14 (35.0)	22 (46.8)	7 (46.7)	0.50
Body mass index, kg/m2	25.5 ± 3.2	25.3 ± 2.5	25.5 ± 2.9	0.89
Non‐ST‐segment–elevation myocardial infarction, n (%)	40 (100.0)	39 (83.0)	14 (93.3)	0.02
Comorbidities
Smoking, n (%)	17 (42.5)	23 (48.9)	2 (13.3)	0.03
Hypertension, n (%)	18 (45.0)	24 (51.1)	2 (13.3)	0.02
Diabetes, n (%)	7 (17.5)	4 (8.5)	3 (20.0)	0.35
Stroke, n (%)	5 (12.5)	5 (10.6)	1 (6.7)	0.81
Laboratory findings
Triglycerides, mmol/L	1.6 (1.1, 2.2)	1.3 (0.9, 1.7)	1.8 (1.2, 2.2)	0.07
Total cholesterol, mmol/L	4.5 ± 1.2	4.2 ± 0.9	4.6 ± 1.2	0.26
Low‐density lipoprotein cholesterol, mmol/L	2.6 ± 0.8	2.5 ± 0.8	2.8 ± 1.2	0.46
High‐density lipoprotein cholesterol, mmol/L	0.96 (0.86–1.15)	1.03 (0.93–1.12)	1.08 (0.80–1.16)	0.69
Fasting blood glucose, mmol/L	5.5 (4.7–6.8)	5.6 (5.0–6.3)	6.3 (5.4–8.3)	0.34
Estimated glomerular filtration rate, mL/min/1.73 m2	119.9 (101.9–120.0)	120.0 (109.3–120.0)	115.6 (109.1–120.0)	0.59
High‐sensitivity troponin T, ng/L	264.7 (75.1–581.3)	347.0 (154.7–881.3)	308.8 (120.9–522.0)	0.18
N‐terminal pro‐B‐type natriuretic peptide, pg/mL	428.2 (154.5–790.0)	643.7 (390.3–1314.0)	760.0 (300.5–1059.0)	0.06
Medications
P2Y12, n (%)	25 (62.5)	27 (57.4)	8 (53.3)	0.80
Aspirin, n (%)	38 (95.0)	46 (97.9)	14 (93.3)	0.66
Statins, n (%)	40 (100.0)	45 (95.7)	12 (80.0)	0.01
Angiotensin‐converting enzyme inhibitors/angiotensin II receptor blockers/sacubitril/valsartan, n (%)	30 (75.0)	36 (76.6)	11 (73.3)	0.96
β‐blocker, n (%)	26 (65.0)	32 (68.1)	11 (73.3)	0.84
Cardiac magnetic resonance findings
Global longitudinal strain, %	18.5 (17.1–21.6)	16.9 (14.4–18.2)	16.1 (14.6–17.4)	<0.001
LV‐end‐diastolic volume index, mL/m2	72.9 ± 18.8	73.0 ± 13.4	83.4 ± 26.2	0.12
LV‐end‐systolic volume index, mL/m2	36.7 ± 10.9	38.4 ± 11.6	47.6 ± 18.9	0.02
LV ejection fraction, %	58.9 ± 8.1	54.9 ± 9.2	53.5 ± 10.2	0.05
Late gadolinium enhancement%	9.8 (6.6–13.0)	17.8 (9.9–27.7)	24.7 (22.4–27.7)	<0.001
Microvascular obstruction, n (%)	0 (0.0)	6 (12.8)	1 (6.7)	0.02
Microvascular parameters
Coronary angiography‐derived index of microvascular resistance	19.0 (16.3–21.0)	25.0 (20.0–30.0)	30.0 (26.0–33.0)	<0.001
Coronary angiography‐derived fractional flow reserve	0.92 (0.87–0.93)	0.93 (0.91–0.94)	0.94 (0.93–0.95)	0.002

Values are median (interquartile range), n (%), or mean ± SD.

CMD indicates coronary microvascular dysfunction; IS%LV, ; LV, left ventricular; and MINOCA, myocardial infarction with nonobstructive coronary arteries.

Relationships between caIMR With IS and MVO Among True MINOCA

Our study found that the MVO and IS% LV ≥16.78% rates were similar among patients with true MINOCA with and without CMD (Figure S1). Specifically, IS% ≥16.78% was 41.7% in patients with CMD and 39.4% in patients without CMD (P = 0.82), whereas MVO rates were 2.8% and 9.1%, respectively (P = 0.43). Furthermore, we explored the predictive value of the caIMR for MVO and IS. The results indicated that caIMR was not a reliable predictor, with an area under the curve of 0.54 (95% CI, 0.43–0.65; P = 0.50) for IS and 0.53 (95% CI, 0.34–0.73; P = 0.77) for MVO (Figure S2).

Association Between CMD, IS and MVO With Clinical Outcomes Among True MINOCA

Over a median follow‐up of 27.0 (interquartile range, 16.8–36.0) months, 33.3% of patients with true MINOCA experienced MACE. The incidence of MACE was higher in patients with CMD or IS%LV ≥16.78% (48.9% in the group with CMD or IS%LV ≥16.78% and 53.3% in the group with CMD and IS%LV ≥16.78%), compared with 7.5% in the group without CMD and IS%LV <16.78% (P < 0.001) (Table 3).

Table 3.

Clinical Outcomes of True MINOCA According to Infarct Size and CMD

	Non‐CMD and IS%LV <16.78% (n = 40)	CMD or IS%LV ≥16.78% (n = 47)	CMD and IS%LV ≥16.78% (n = 15)	P value
Major adverse cardiovascular events, n (%)	3 (7.5)	23 (48.9)	8 (53.3)	<0.001
All‐cause death, n (%)	0 (0.0)	5 (10.6)	1 (6.7)	0.04
Recurrent myocardial infarction, n (%)	2 (5.0)	4 (8.5)	0 (0.0)	0.30
Stroke, n (%)	0 (0.0)	2 (4.3)	1 (6.7)	0.20
Cardiovascular readmission, n (%)	1 (2.5)	12 (25.5)	6 (40.0)	0.001
Heart failure, n (%)	0 (0.0)	2 (4.3)	2 (13.3)
Arrhythmia, n (%)	0 (0.0)	2 (4.3)	2 (13.3)
Angina pectoris, n (%)	1 (2.5)	8 (17.0)	2 (13.3)

Values are n (%).

The receiver operating characteristic analysis showed an area under the curve of 0.71 (95% CI, 0.60–0.82; P = 0.001), suggesting a moderate discriminative ability of LGE% for predicting MACE, and the optimal cutoff value of LGE% for predicting MACE was 16.78% (Figure 3). The restricted cubic spline analysis revealed a significant linear relationship between LGE% and MACE risk (P = 0.03), with no indication of nonlinearity (nonlinear P = 0.63) (Figure 4).

Figure 3. Receiver operating characteristic illustrating the associations of LGE% with MACE among true MINOCA.

AUC indicates area under the curve; LGE, late gadolinium enhancement; MACE, major adverse cardiovascular events; MINOCA, myocardial infarction with nonobstructive coronary arteries; and ROC, receiver operating characteristic.

Figure 4. Spline curves analysis illustrating the associations of LGE% with MACE among true MINOCA.

HR indicates hazard ratio; LGE, late gadolinium enhancement; MACE, major adverse cardiovascular events; and MINOCA, myocardial infarction with nonobstructive coronary arteries.

The Kaplan–Meier analysis illustrates that patients with IS%LV ≥16.78% exhibit a significantly higher cumulative incidence of MACE compared with those with IS%LV <16.78% (log‐rank P < 0.001) (Figure 5A). In Figure 5B, the analysis indicates a trend toward an increased incidence of MACE in patients with CMD compared with those without CMD, although this difference is not statistically significant (log‐rank P = 0.08). Notably, Figure 5C shows that the highest cumulative incidence of MACE is seen in patients with both CMD and IS%LV ≥16.78%, followed by those with either CMD or IS%LV ≥16.78% (log‐rank P < 0.001), highlighting the combined effect of CMD and IS on MACE risk. Additionally, our results showed that patients with MVO exhibited a slightly higher rate of MACE (57.1%) compared with those without MVO (31.6%), though the difference was not statistically significant (P = 0.33) (Figure S3).

Figure 5. Cumulative incidence of MACE in patients with true MINOCA stratified according to IS (A), CMD (B), and combination of IS and CMD (C).

CMD indicates coronary microvascular dysfunction; HR, hazard ratio; IS%LV, infarct size left ventricular; and MACE, major adverse cardiovascular events.

Cox Proportional Hazard Analysis

Tables 4 and 5 present the results of Cox proportional hazards models. In Table 4, the univariate analysis shows that both CMD (HR, 1.82 [95% CI, 0.92–3.58], P = 0.09) and IS%LV ≥16.78% (HR, 3.57 [95% CI, 1.73–7.34], P = 0.001) are associated with increased risk of MACE. The combination of CMD or IS%LV ≥16.78% significantly raises the risk (HR, 7.97 [95% CI, 2.39–26.55], P = 0.001), and CMD and IS%LV ≥16.78% further increase the risk (HR, 8.42 [95% CI, 2.23–31.87], P = 0.002), whereas MVO did not predict MACE (HR, 1.69 [95% CI, 0.60–4.80], P = 0.33). Table 5 reinforces these findings, showing that the combination of CMD or IS%LV ≥16.78% continues to exhibit a strong association with MACE (HR, 8.77 [95% CI, 2.57–29.94], P = 0.001), and the combination of CMD and IS%LV ≥16.78% shows a high risk (HR, 7.40 [95% CI, 1.94–28.23], P = 0.003), adjusting for factors like diabetes and LVEF.

Table 4.

Univariate Cox Proportional Hazards Model for MACE

	Univariate analysis HR (95% CI)	P value
Age	1.00 (0.97–1.04)	0.78
Male sex	1.34 (0.68–2.62)	0.40
Hypertension	0.73 (0.36–1.47)	0.38
Diabetes	2.07 (0.90–4.78)	0.09
Stroke	1.29 (0.50–3.33)	0.60
Smoking	1.16 (0.59–2.30)	0.66
Total cholesterol	1.09 (0.80–1.50)	0.59
Estimated glomerular filtration rate	0.99 (0.97–1.02)	0.66
Peak high‐sensitivity troponin T	1.00 (1.00–1.00)	0.12
Peak N‐terminal pro‐B‐type natriuretic peptide	1.00 (1.00–1.00)	0.95
LV ejection fraction	0.97 (0.94–1.00)	0.07
Statins	0.63 (0.15–2.66)	0.53
Late gadolinium enhancement%	1.04 (1.02–1.07)	0.001
IS%LV ≥16.78%	3.57 (1.73–7.34)	0.001
Coronary angiography‐derived index of microvascular resistance	1.03 (1.00–1.06)	0.04
CMD	1.82 (0.92–3.58)	0.09
IS%LV + CMD
Non‐CMD and IS%LV <16.78%		Reference
CMD or IS%LV ≥16.78%	7.97 (2.39–26.55)	0.001
CMD and IS%LV ≥16.78%	8.42 (2.23–31.87)	0.002
Microvascular obstruction	1.69 (0.60–4.80)	0.33

CMD indicates coronary microvascular dysfunction; HR, hazard ratio; IS, infarct size; and LV, left ventricular.

Table 5.

Multivariable Cox Proportional Hazards Model for MACE

	Multivariate
	Model 1		Model 2		Model 3
	HR (95%CI)	P value	HR (95%CI)	P value	HR (95%CI)	P value
Diabetes	2.43 (1.04–5.69)	0.04	1.99 (0.86–4.59)	0.12	2.79 (1.17–6.67)	0.02
LV ejection fraction	0.97 (0.94–1.00)	0.09	0.98 (0.94–1.01)	0.19	0.99 (0.95–1.02)	0.46
IS%LV ≥16.78%	3.63 (1.76–7.49)	<0.001	Not in the model
CMD	Not in the model		1.55 (0.76–3.16)	0.23
IS%LV + CMD	Not in the model		Not in the model
Non‐CMD and IS <16.78%						Reference
CMD or IS%LV ≥16.78%					8.77 (2.57–29.94)	0.001
CMD and IS%LV ≥16.78%					7.40 (1.94–28.23)	0.003

Model 1: All variables with a univariate P < 0.1 were included in the adjustment, with the exception of CMD. Model 2: All variables with a univariate P < 0.1 were included in the adjustment, with the exception of IS%LV. Model 3: Adjusting for all covariates showing a P < 0.1 in the univariate analysis.

CMD, coronary microvascular dysfunction; HR, hazard ratio; IS, infarct size; and LV, left ventricular.

To confirm the robustness of our findings, a sensitivity analysis was performed by excluding patients with STEMI. Notably, in the fully adjusted Cox regression models, the combination of CMD and IS%LV ≥16.78% demonstrated a significant association with increased risk, with HR values of 9.39 (95% CI, 2.62–33.62, P = 0.001) for CMD or IS%LV ≥16.78%, and 12.11 (95% CI, 3.08–47.50, P < 0.001) for CMD and IS%LV ≥16.78%, further reinforcing the robustness of our results in patients with NSTEMI (Table 6).

Table 6.

Sensitivity Analysis of Cox Regression Models Excluding Patients With STEMI

	Multivariate
	Model 1		Model 2		Model 3
	HR (95%CI)	P value	HR (95%CI)	P value	HR (95%CI)	P value
Diabetes	3.61 (1.36–9.55)	0.01	3.03 (1.15–7.97)	0.02	4.53 (1.71–11.97)	0.002
LV ejection fraction	0.97 (0.93–1.02)	0.33	0.99 (0.95–1.04)	0.97	1.01 (0.96–1.05)	0.74
IS%LV ≥16.78%	4.10 (1.85–9.06)	<0.001	Not in the model
CMD	Not in the model		1.55 (0.76–3.16)	0.02
IS%LV + CMD	Not in the model		Not in the model
Non‐CMD and IS <16.78%						Reference
CMD or IS%LV ≥16.78%					9.39 (2.62–33.62)	0.001
CMD and IS%LV ≥16.78%					12.11 (3.08–47.50)	<0.001

Model 1: All variables with a univariate P < 0.1 were included in the adjustment, with the exception of CMD. Model 2: All variables with a univariate P < 0.1 were included in the adjustment, with the exception of IS%LV. Model 3: Adjusting for all covariates showing a P < 0.1 in the univariate analysis.

CMD indicates coronary microvascular dysfunction; HR, hazard ratio; IS, infarct size; LV, left ventricular; and STEMI, ST‐segment–elevation myocardial infarction.

DISCUSSION

To the best of our knowledge, this study is the first to examine the relationship between caIMR, MVO, and IS in patients with true MINOCA, while evaluating the prognostic significance of CMD, IS, and MVO using both CMR and caIMR. Our study population primarily consisted of patients with NSTEMI (92%), reflecting the typical distribution of MINOCA in clinical practice. Key findings include, first, CMD was prevalent in about one third of patients with suspected MINOCA, particularly in those with cardiomyopathies. Second, caIMR did not predict MVO or IS in true MINOCA. Third, the combination of CMD and IS%LV ≥16.78% was a significant independent predictor of MACE, whereas MVO was not associated with MACE among true MINOCA. These findings highlight the importance of combining CMD and IS assessments for improved risk stratification, particularly among patients with NSTEMI with true MINOCA.

MINOCA is a complex syndrome with both ischemic and nonischemic causes, which complicates diagnosis and treatment. ¹ The role of CMD in MINOCA is essential, as it is increasingly seen as a major factor contributing to ischemia or infarction in nonobstructive coronary artery disease. Various mechanisms, including microvascular spasm and endothelial dysfunction, can disrupt coronary microvascular function, leading to poor outcomes in patients with MINOCA. Previous research reports a wide prevalence of CMD and microvascular spasm in MINOCA, ranging from 35% to 54%, mainly due to differences in diagnostic criteria and thresholds. ¹⁶ , ³⁵ , ³⁶ , ³⁷ The high rate of CMD in our cohort (about one third of patients) highlights its key role in the underlying mechanisms of MINOCA. In our study, CMD was defined as caIMR >25 U, consistent with prior NSTEMI studies ³² , ³³ , ³⁴ ; however, higher thresholds (IMR >40 U) are usually used in STEMI cases, ³⁸ where more extensive myocardial damage results in higher IMR values. The complexity of CMD in MINOCA is further highlighted by observed differences in traditional risk factors within our cohort. We found that smoking and hypertension were less common in the group with CMD and large IS. Although this contrasts with existing research, it underscores the diverse nature of the population with MINOCA, where traditional risk factors may not always match clinical outcomes. Previous studies show that most patients with nonobstructive coronary artery disease have traditional risk factors, with only 8% to 10% lacking these risk factors. ³⁹ , ⁴⁰ Smoking and hypertension impair microvascular function through endothelial dysfunction and structural changes. ⁴¹ However, the lack of a significant association in our cohort may reflect unique mechanisms in MINOCA, such as coronary vasospasm or supply–demand mismatch (type II MI). Further research is necessary to determine CMD thresholds and diagnostic criteria for MINOCA, while also clarifying the role of traditional risk factors to enhance diagnosis and treatment for both STEMI and NSTEMI subtypes.

IS, a well‐established marker of myocardial injury, plays a crucial role in the prognosis of AMI, with larger infarcts typically correlating with worse clinical outcomes. The extent of myocardial damage is commonly assessed through imaging techniques such as LGE on CMR, providing valuable prognostic information. ²⁸ , ⁴² , ⁴³ , ⁴⁴ In the context of MINOCA, however, the relationship between IS and outcomes remains less clear due to the absence of significant epicardial coronary obstruction. In our study, we observed that patients with true MINOCA generally have smaller infarcts, which is consistent with CMR‐based studies that have demonstrated limited IS in this population. ⁴⁵ , ⁴⁶ With current LGE techniques, a minimum of ∼0.2 g of necrotic myocardium is required for detection, ⁴⁷ , ⁴⁸ which may explain why some patients with MINOCA do not exhibit LGE‐defined infarction despite clinical evidence of myocardial injury. Despite these smaller infarcts, we found that IS%LV ≥16.78% significantly increased the risk of MACE, which highlights that even modest infarcts can lead to adverse clinical outcomes. Additionally, the use of LGE% to assess myocardial injury and scarring revealed that larger infarcts were associated with worse clinical outcomes. These findings are consistent with recent studies that also identified LGE% as a predictor of poor prognosis in MINOCA. ⁴⁹ , ⁵⁰ Our study differs from previous research by determining the optimal cutoff value of LGE% and evaluating the extent of IS%LV, highlighting that patients with larger infarcts experienced worse clinical outcomes among true MINOCA. These findings highlight the importance of LGE%, especially in assessing IS%LV, as it enables more accurate risk assessment and the development of personalized therapeutic strategies for patients with MINOCA. Furthermore, sensitivity analysis excluding the subgroup with STEMI confirmed that our main findings remained consistent, reinforcing the robustness of our conclusions. This consistency highlights the predictive value of CMD and IS for MACE in the population with predominantly NSTEMI MINOCA =.

On the other hand, MVO is a recognized prognostic factor in MI, where it is linked to adverse outcomes. ¹⁹ , ²⁸ However, in the context of MINOCA, the role of MVO remains incompletely defined. In our cohort, the prevalence of MVO was 6.7%. This figure is comparable to the 1.5% reported by Bergamaschi et al., ⁴⁹ although their study did not use high‐resolution LGE imaging and excluded patients with incomplete or suboptimal CMR, which may have led to an underestimation of the actual rate. In contrast, Gerbaud et al. ⁴⁵ reported a substantially higher MVO rate (20%), likely reflecting both the use of advanced CMR techniques, including high‐resolution LGE, and differences in patient selection. These discrepancies highlight that reported MVO rates in MINOCA are influenced by variations in imaging protocols, completeness of CMR assessment, and diagnostic criteria. Although patients with MVO showed a slightly higher rate of MACE, the difference was not statistically significant, probably due to the low prevalence of MVO and the small number of patients with STEMI in our cohort. The limited impact of MVO in our study may reflect the different pathophysiology of NSTEMI‐MINOCA, where myocardial damage and microvascular injury are generally less extensive than in STEMI. We also found no association between caIMR and larger IS or MVO in patients with true MINOCA. The absence of a correlation with caIMR may reflect differences in the mechanisms of microvascular dysfunction between MINOCA and STEMI, with STEMI typically involving more extensive microvascular injury. Additionally, our cohort predominantly consisted of patients with NSTEMI, which may explain why caIMR did not exhibit the expected association with MVO or IS. Further research is needed to explore the association between IMR and IS/MVO in MINOCA, especially among patients with NSTEMI.

Given the retrospective nature of this study and the lack of standardized treatment protocols for MINOCA, medical therapy was administered at the discretion of the treating physicians, leading to variability in pharmacologic management. Statin use was lower among patients with both CMD and elevated IS, likely reflecting clinical judgment or uncertainty about the benefits of intensive secondary prevention. Although no significant differences were observed in the use of aspirin, P2Y12 inhibitors, angiotensin‐converting enzyme inhibitors, angiotensin II receptor blockers, or β‐blockers, the potential role of cardioprotective agents targeting microvascular and endothelial function warrants further exploration. Previous studies suggest that therapies such as statins, angiotensin‐converting enzyme inhibitors, β‐blockers, and calcium channel blockers may improve endothelial function and microvascular integrity. ⁵¹ , ⁵² These findings highlight the need for prospective studies to evaluate whether tailored pharmacologic strategies can improve outcomes in patients with MINOCA with CMD, particularly given the observed variability in treatment approaches.

LIMITATIONS

Several limitations should be acknowledged in this study. First, the study was a single‐center retrospective observational analysis; the findings may not be generalizable to broader populations. Second, the follow‐up period was relatively short, which could have influenced the assessment of long‐term outcomes. Third, we lacked the use of optical coherence tomography to assess the exact mechanisms alongside CMR. This may have led to incomplete evaluations, as optical coherence tomography could have provided additional insights into coronary artery conditions. Fourth, a complete CMR examination was not conducted for all patients with suspected MINOCA, which may lead to misclassification in some instances. Furthermore, the incomplete use of CMR across the cohort may have introduced selection bias, as patients who underwent CMR might systematically differ in clinical characteristics or underlying pathophysiology from those who did not. This selective use of imaging may have affected the derivation and interpretation of caIMR thresholds, potentially limiting the generalizability of the findings to the broader MINOCA population, especially in settings with restricted access to CMR. Fifth, because nearly 91% of patients in this true MINOCA cohort presented with NSTEMI, we selected a caIMR threshold of >25 units to define CMD, which may restrict the generalizability of this threshold to other acute coronary syndrome subtypes, especially in populations with a more diverse mix of NSTEMI and STEMI presentations. Notably, the small number of patients with STEMI also limited our ability to draw statistically significant conclusions regarding the association between MVO and MACE in this subgroup. Additionally, CMD was assessed using noninvasive caIMR rather than invasive measurements such as IMR or coronary flow reserve, which may limit the precision of the results. Finally, treatment decisions were influenced by individual physician judgment in the absence of standardized guidelines for MINOCA, introducing potential variability and confounding that could affect the study results. Future studies should incorporate a multicenter approach and extend the follow‐up period to validate these findings over time.

CONCLUSIONS

Our study demonstrated that CMD is prevalent in approximately one third of patients with suspected MINOCA. The combination of CMD and IS%LV ≥16.78% serves as a strong predictor of MACE in true MINOCA cases, particularly among those with NSTEMI. These findings highlight the importance of evaluating both CMD and IS when assessing the risk in patients with true MINOCA.

Sources of Funding

This work was supported in part by National Natural Science Foundation of China (W2433190), Chinese National Natural Science Foundation (82170521), Shanghai Natural Science Foundation of China (21ZR1449500), Foundation of Shanghai Municipal Health Commission (202140263), Tibet Natural Science Foundation of China (XZ2022ZR‐ZY27(Z), XZ202301ZR0032G), Foundation of Chongming (CKY2021‐21, CKY2020‐29), Clinical Research Plan of Shanghai Tenth People’s Hospital (YNCR2A001), Clinical Research Plan of SHDC (SHDC2020CR4065), and Foundation of the Science and Technology Commission of Shanghai Municipality (Grant No. 20dz1207200).

Disclosures

None.

Supporting information

Figures S1‐S3.