Research indicates that ischemic heart disease (IHD) is a critical cause of cardiac death [1, 2]. The incidence of myocardial infarction in women is closely associated with the progressive rise in fatal IHD rates with advancing age [3]. As of 2018, the IHD mortality rate among non-Hispanic white women reached 64.9% per 100,000 population [3]. A U.S. study revealed that since 2000, there has been minimal improvement in IHD mortality among young women [4], which may be linked to insufficient risk communication by physicians. A survey showed that only 21% of women were informed about the potential adverse outcomes of IHD [5]. Although prior studies have exhaustively analyzed the poor prognosis of IHD, it remains a syndrome encompassing complex pathophysiology [6, 7, 8], with its conceptual scope extending beyond myocardial ischemia caused by atherosclerosis. Notably, the ORBITA (Optimal Medical Therapy of Angioplasty in Stable Angina) [9] and ISCHEMIA (International Study of Comparative Health Effectiveness with Medical and Invasive Approaches) [10] trials have challenged the traditional coronary stenosis-centric therapeutic paradigm for IHD. Furthermore, the SCOT-HEART (Scottish Computed Tomography of the Heart) study [11] demonstrated that most patients with coronary heart disease (CHD) lack epicardial stenosis, suggesting that angina symptoms and ischemic manifestations in non-obstructive CHD may stem from impaired microvascular regulation. Consequently, IHD caused by coronary microvascular dysfunction (CMD) has emerged as a pivotal factor in evaluating coronary revascularization and prognosis, making coronary microvascular disease a pressing public health challenge.

CMD is characterized by functional and structural abnormalities in the microvasculature (non-atherosclerotic stenosis) that disrupt coronary blood flow regulation, manifesting as enhanced microvascular constriction, impaired endothelium-dependent/independent vasodilation, and elevated microcirculation resistance [12, 13]. Diagnosis of CMD relies on imaging and functional assessments, with the pressure wire-derived index of microcirculation resistance (IMR) currently serving as a key tool for evaluating coronary microvascular disease (CMVD) [14, 15, 16, 17]. However, this invasive approach requires pharmacologically induced maximal coronary hyperemia to obtain measurements, often causing adverse effects such as chest tightness and bradycardia. Recent research has focused on angiography-based IMR (Angio-IMR), a non-invasive method that indirectly assesses coronary functional parameters without additional procedures. This study aims to evaluate the diagnostic efficiency of Angio-IMR for identifying CMD in populations suffering from angina pectoris.

Our study demonstrates that Angio-IMR exhibits high diagnostic concordance with invasive pressure wire-derived IMR for detecting CMD. Pooled estimates revealed clinically robust performance indicators, involving sensitivity, specificity, PPV, and NPV. The superior discriminative capacity of Angio-IMR was further evidenced by AUC $>$0.9, supporting its utility as a non-invasive alternative to guide CMD diagnosis in clinical practice.

Patients with angina attributable to CMD constitute a clinically significant subset of those with chronic coronary syndrome (CCS). Although CMD is traditionally classified as a non-atherosclerotic disorder, the Women’s Ischemia Syndrome Evaluation (WISE) substudy employing intravascular ultrasound (IVUS) identified female patients with non-obstructive CAD and revealed that CMD-associated cardiovascular risk correlates strongly with atherosclerosis-related risk factors [31, 32, 33]. Current epidemiological data estimate the prevalence of coronary microvascular disease (CMVD) at 40–60% [34, 35], though heterogeneity persists due to inconsistent diagnostic criteria for CMD. Furthermore, the reliance on invasive pressure wire-derived IMR has limited the detection rate of CMD in clinical practice. These limitations have driven the development of Angio-IMR, a non-invasive computational framework rooted in the hemodynamic principles of invasive IMR. Specifically, invasive IMR is characterized as the product of distal coronary pressure (Pd) and mean transit time (Tmn) when reaching the maximal hyperemic state [36]. Current Angio-IMR algorithms primarily derive from this foundational formula [2, 17, 20]:

(1)$$AMR=\frac{Pd}{Velocit{y}_{hyp}}=\frac{Pa\times QFR}{Velocit{y}_{hyp}}$$

(2)$$AccuIMR=\frac{Pa\times AccuFFRangio\times L}{Velocit{y}_{hyp}}$$

(3)$$caIMR=\frac{Pd\times L}{K*Velocit{y}_{hyp}}$$

Pd: distal coronary pressure, QFR: quantitative flow ratio, Pa: proximal coronary pressure, L: length of blood vessels, Velocity_hyp: flow velocity in the hyperemic status.

The derivation of angiography-based IMR fundamentally involves simulating distal coronary pressure (Pd) via FFR calculations [37], followed by multiplying Pd by aortic pressure (Pa) and contrast transit time (Tmn). This approach hinges on angiography-derived FFR computations, with microvascular resistance (MR) subsequently quantified through computational fluid dynamics (CFD). Large-scale trials have demonstrated non-inferiority of angiography-based FFR-guided percutaneous coronary intervention (PCI) compared to wire-based FFR strategies [38], validating its equivalence in assessing stenotic epicardial vessel function. However, there is limited evidence regarding the correlation between pressure wire-based IMR and Angio-IMR. Furthermore, prior meta-analyses exhibited significant methodological bias due to heterogeneous cohorts (e.g., combining CCS and acute coronary syndrome populations) [39] and inconsistent diagnostic thresholds for the gold-standard IMR across studies. To address these limitations, this meta-analysis exclusively focused on angina populations (excluding myocardial infarction) to minimize confounding variables.

Based on angiographic stenosis severity, enrolled patients were stratified into two cohorts: obstructive CAD (stenosis $\ge$50%) and non-obstructive CAD. Subgroup analysis revealed comparable diagnostic efficiency of Angio-IMR in both groups, with insignificant differences in sensitivity or specificity, suggesting that coronary stenosis severity does not compromise Angio-IMR’s ability to identify CMD. This finding aligns with evidence from Scarsini et al. [28], who demonstrated strong correlation between Angio-IMR and invasive IMR across infarct-related arteries (IRA), non-IRA, and diverse clinical presentations (STEMI, NSTEMI, and CCS). Furthermore, subgroup comparisons of angiography-derived FFR computational platforms (quantitative FFR [QFR] vs. AccuFFR) indicated superior diagnostic accuracy for Accu-IMR derived from AccuFFR.

Given the inclusion of four studies [22, 24, 25, 26] utilizing the AccuFFRangio system to quantify Angio-IMR, we performed additional subgroup analyses. These revealed that hyperemia-induced Angio-IMR measurements (achieved via adenosine-mediated vasodilation) exhibit superior reliability compared to resting-state assessments. This is consistent with the findings of Scarsini et al. [28], who revealed no significant correlation between non-hyperemic Angio-IMR and invasive IMR in CCS cohorts. Collectively, these observations underscore the necessity of hyperemic conditions for optimizing Angio-IMR’s diagnostic utility in CMD.

While the limited number of included studies and heterogeneity in computational platforms introduce potential confounding biases, subgroup analyses reinforced the robustness of the primary findings. The meta-analysis conclusively demonstrates that Angio-IMR achieves high diagnostic performance (AUC: 0.91, 95% CI: 0.89–0.94), indicating its clinical applicability as a non-invasive alternative.

This comprehensive meta-analytical synthesis establishes angio-IMR as a diagnostically robust modality, demonstrating superior discriminative capacity for detecting CMD in angina pectoris cohorts. The concordance with invasive wire-based IMR measurements collectively confirms its clinical validity, thereby positioning this modality as a viable non-invasive surrogate for traditional intracoronary physiological assessment.

The original data for this study is available from the corresponding author.

WW, YC, and YQZ collected the data and wrote the manuscript; MWL, BLX, MJG and LLJ analyzed the data; FHZ and KJC reviewed the manuscript, designed this work, and guided the methodology. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

This work was supported by Hospital capability enhancement project of Xiyuan Hospital, CACMS (No.CI2021A00901), Beijing Clinical Research Ward (BCRW202108) and Beijing Traditional Chinese Medicine Technology Development Fund Project (BJZYZD-2023-11).

The authors declare no conflict of interest.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/RCM25764.