Value of high-density lipoprotein cholesterol, myocardial perfusion index, and global longitudinal strain derived from cardiac magnetic resonance imaging in predicting coronary slow flow in patients with nonobstructive coronary artery disease

Introduction

Coronary slow flow (CSF) is an angiographic phenomenon characterized by delayed contrast opacification. It is defined by a corrected thrombolysis in myocardial infarction frame count (CTFC) >27 frames and stenosis of the epicardial arterial lumen of less than 40% but excludes secondary factors such as coronary ectasia and emboli (1). Prior studies have indicated a CSF incidence of 4.86–18.82% in patients without obstructive epicardial arterial disease (2,3). Moreover, studies on patients with CSF have reported recurrent angina attacks, arrhythmias, and high readmission rates among this population (4-6). Notably, a growing body of evidence supports the presence of increased myocardial mass, an enlarged left ventricle, and impaired left ventricular (LV) function in patients with CSF (2,7). Although the pathogenesis of CSF remains unclear, inflammation, endothelial dysfunction, and coronary microvascular dysfunction (CMD) may be involved (7). CTFC is typically determined by coronary angiography (CAG), which assesses myocardial perfusion in patients with CSF. However, CTFC is influenced by several factors, including heart rate, contrast injection rate, catheter size, and the definition of the distal bifurcation landmark (8). Furthermore, CAG is an invasive test, and its efficiency highly depends on the operator. Recent research suggests that the triglyceride (TG), glucose, body mass index (BMI), dyslipidemia, and vascular disease may serve as predictors of CSF (3,9). A few previous studies have demonstrated that myocardial strain, as assessed by echocardiography, is a significant predictor of CSF, and is moderately correlated with CTFC (10,11). However, these predictors do not offer further insight into the blood flow of patients with CSF at rest. Therefore, it is critical to develop noninvasive tests and examine their ability to quantitatively assess myocardial perfusion and comprehensively evaluate cardiac function in patients with CSF.

Cardiac magnetic resonance (CMR) first-pass perfusion can accurately and noninvasively assess myocardial flow across various cardiac diseases (12,13). In contrast to CTFC, the CMR resting first-pass perfusion technique allows for the objective quantification of myocardial perfusion in patients with CSF through the calculation of the perfusion index (PI). One study found that decreased PI is associated with type 2 diabetes mellitus (12). Moreover, Zhu et al. reported that PI is correlated with ventricular dysfunction in patients with coronary artery disease (CAD) (14). In addition to myocardial perfusion, CMR is capable of simultaneously evaluating cardiac function via CMR feature tracking (CMR-FT). It has been demonstrated that CMR-FT can accurately and repeatably evaluate myocardial strain, an effective and noninvasive index for assessing cardiac function and deformation (15,16). Furthermore, reduced LV strain is associated with early impaired ventricular dysfunction and adverse cardiovascular events in patients with heart failure with preserved ejection fraction (17). However, research on myocardial perfusion and cardiac function derived from CMR in patients with CSF is lacking.

This study employed CMR first-pass perfusion imaging and CMR-FT to assess myocardial perfusion and characterize myocardial strain and further determined the diagnostic value of CMR parameters in predicting CSF. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1668/rc).

Methods

Study population

In this case-control study, we retrospectively screened all patients (N=4,450) who underwent CAG in the catheterization laboratory of the Second Affiliated Hospital of Kunming Medical University from January 2018 to January 2024. All participants underwent CAG due to the presence of typical angina symptoms and evidence of myocardial ischemia on noninvasive tests. The inclusion criteria were (I) coronary artery stenosis ≤40% on CAG and (II) CMR imaging performed for perfusion and strain analysis. Meanwhile, the exclusion criteria were as follows: (I) ST-segment elevation myocardial infarction (STEMI), non-STEMI, and coronary artery stenosis >40%; (II) coronary artery spasm or aneurysmal dilatation, coronary artery thrombosis or embolism, myocardial bridge, left atrial enlargement, cardiomyopathy, heart valve disease, or New York Heart Association (NYHA) cardiac function classification ≥ grade III (18); (III) coronary artery stenting or angioplasty, thrombolytic therapy, or history of other treatment modalities; (IV) connective tissue disease, cancer, severe liver or kidney injury, infection, or hematologic disease; and (V) poor image quality. In total, 146 patients without obstructive epicardial arterial disease who underwent CMR within 1 week of admission to the Second Affiliated Hospital of Kunming Medical University were recruited. These participants were further divided into CSF and control groups according to the diagnostic standard of CSF based on the guidelines delineated by Beltrame, which are the most widely recognized diagnostic criteria (1). Ultimately, 73 patients diagnosed with CSF were included in the CSF group, and 73 participants with normal coronary flow from the same period were included as the control group. The retrospective clinical and imaging data of all participants were collected. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Second Affiliated Hospital of Kunming Medical University Ethics Committee (approval No. PJ2022125). The requirement for individual consent was waived due to the retrospective nature of the analysis. This study was registered in the Chinese Clinical Trial Registry (ChiCTR2200066005).

CAG

CAG was performed via the radial or femoral angiography via the standard Judkins approach. Angiographic images were obtained with cranial and caudal angulation and with right and left views. The thrombolysis in myocardial infarction frame count (TFC) method, as established by Gibson et al. (8), was employed to assess coronary artery flow. TFC was calculated as the number of frames present from the onset of opacification at the ostial vessel to the moment in which the contrast dye reached the distal landmark branch. For the left anterior descending coronary artery (LAD), distal bifurcations, such as the “mustache”, “whale tail”, or “pitchfork”, were defined as the distal landmark branch. For the left circumflex artery (LCx), the segment with the greatest total distance was selected as the distal landmark branch. The first branch of the posterolateral artery in the right coronary artery (RCA) was used as the distal landmark branch for the RCA. To account for the longer length of the LAD, the CTFC was derived by dividing the LAD frame count by 1.7. The mean TFC for each participant was calculated by summing the TFCs of the three coronary arteries and dividing the total by 3. CSF was defined as a TFC (CTFC for LAD) >27 frames (30 frames/s) in at least one coronary artery and coronary artery stenosis ≤40% as per Beltrame’s guidelines (1). CAG was processed by cardiologists blinded to the results of CMR.

Image acquisition

All participants were examined with a 3.0T MR scanner (Achieva 3.0 T, Philips Healthcare, Best, the Netherlands) under electrocardiographic and respiratory procedures. Furthermore, all participants were instructed to withhold antianginal medications, including calcium-channel blockers, beta-blockers, and nitrates, for 48 hours prior to the CMR examination. Cine imaging was carried out with a balanced steady-state free precession sequence (time frames per cardiac cycle =30, repetition time =3.1 ms, echo time =1.54 ms, flip angle =45°, field of view =350×350 mm², and voxel resolution =1.8×1.4×8.0 mm³). Resting first-pass perfusion imaging was acquired via turbo field echo under the following parameters: repetition time =2.5 ms, echo time =1.1 ms, flip angle =20°, field of view =350 × 350 mm², and voxel resolution =1.8×1.4×8.0 mm³. A 0.2-mL/kg bolus of gadoteridol (Gd-HP-DO3A; ProHance, BIPSO GmbH, Singen, Germany) was injected automatically at a rate of 2–3 mL/s with a 20-mL saline flush. The perfusion images were acquired in apical, middle, and basal short-axis slices. All cine images were obtained in continuous slices of short-axis planes covering the entire left ventricle from the apex to base.

Image analysis

CMR data analysis was completed by two experienced radiologists using CVI42 postprocessing software (Circle Cardiovascular Imaging, Inc., Calgary, Canada).

The endo- and epicardial contours were traced from the end-diastolic and end-systolic images via the semiautomatic outlining method. The relevant structural and functional parameters of the left ventricle were calculated by the software, including LV ejection fraction (LVEF), LV end-systolic volume (ESV), LV end-diastolic volume (EDV), and myocardial mass. The time-signal intensity curve of each myocardial segment was acquired for myocardium perfusion analysis, as displayed in Figure 1. The semiquantitative PI derived from CMR images was obtained from the myocardial perfusion curve and assessed for myocardial perfusion. PI was defined as an upslope ratio of myocardial segments to the blood pool that affected the concentration of the gadobutrol contrast agent (13). The LV strain parameters, including global radial strain (GRS), global circumferential strain (GCS), and global longitudinal strain (GLS), were obtained at the LV short-axis level, LV long-axis level, and at the two-, three-, and four chamber heart levels (Figure 2).

Figure 1 CMR first perfusion and time-signal intensity curve. (A,D) Short-axis CMR images of resting first-pass perfusion acquired at end diastole in a patient with CSF (a 56-year-old male) (A) and in a control participant (a 21-year-old female) (D). (B,E) The same images with manual delineation of the endocardial (red) and epicardial (green) contours in the patient with CSF (B) and the control participant (E). (C) A 56-year-old patient with CSF presented with a global LV PI of 4.9. (F) A 21-year-old control with a global LV PI of 12.53. CMR, cardiac magnetic resonance; CSF, coronary slow flow; LV, left ventricular; PI, perfusion index; SI, signal intensity.

Figure 2 CMR-derived myocardial strain. (A-C) A 35-year-old patient with CSF and a GLS of 12.80% (A), a GCS of –17.46% (B), and a GRS of 32.90 (C). (D-F) A 60-year-old control with a GLS of –16.90% (D), a GCS of –22.21% (E), and a GRS of 32.52% (F). GCS, global circumferential strain; GLS, global longitudinal strain; CMR, cardiac magnetic resonance; GRS, global radial strain; CSF, coronary slow flow.

Clinical assessment

Clinical data were retrospectively collected and included gender, age, BMI, low-density lipoprotein cholesterol (LDL-H), high-density lipoprotein cholesterol (HDL-C), TG, serum creatinine, hypertension (defined as the ongoing therapy with any antihypertensive medications or systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg), diabetes (defined as therapy consisting of antidiabetic drugs or fasting serum glucose ≥7.0 mmol/L), hyperuricemia (defined as a plasma uric acid level exceeding 420 µmol/L in males or exceeding 360 µmol/L in females), peripheral arterial disease (defined as an ankle brachial index value ≤0.9), stroke (defined as significantly corresponding clinical presentation and positive signs on head computed tomography or magnetic resonance imaging), and smoking status (with positive smoking status defined as current smoking or a previous smoking habit). The risk factors for cardiovascular disease (hypertension, diabetes, hyperuricemia, peripheral arterial disease, stroke, and smoking) were collected based on previous medical records in our hospital. Heart rate, systolic blood pressure, and diastolic blood pressure were measured immediately prior to the CMR examination, following a rest period of 10 minutes.

Reproducibility

The CMR observers were blinded to the CAG results and the clinical data. Thirty participants were randomly selected for the assessment of interobserver reproducibility. The image measurement and postprocessing were performed by two experienced physicians in CMR diagnosis (≥3 years of experience). To validate intra-observer reproducibility, one of the physicians reanalyzed the same samples after 2 weeks.

Statistical analysis

All analyses were conducted with SPSS version 26.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism version 7.0 (Dotmatics, Boston, MA, USA). Measures conforming to a normal distribution are presented as the mean ± standard deviation, and an independent samples t-test was used to compare variables of the CSF and control groups. Variables with a nonnormal distribution are expressed as the median and interquartile range, and the Mann-Whitney test was performed to compare differences between the two groups. The retrospectively collected clinical data and imaging parameters were compared between the CSF and the control groups via univariate logistic regression, and variables with statistical significance were then included in multivariate logistic regression in order to determine the independent predictors of CSF and build the predictive model. The performance of the selected diagnostic indices and combined model in predicting CSF was assessed via receiver operating characteristic (ROC) curves, with diagnostic accuracy evaluated by the area under the curve (AUC). The optimal cutoff value was determined according to the maximum Youden index, as were the related sensitivity and specificity. The AUC was considered to be unacceptable at a value of 0.5, acceptable between 0.7 and 0.8, excellent between 0.8 and 0.9, and outstanding above exceeding 0.9 (19). The DeLong test was applied to compare the AUCs of the different models. Furthermore, the intraobserver and interobserver variability for PI and myocardial strain were assessed via the intraclass correlation coefficient (ICC) in order to validate the reproducibility of the CMR indices. A two-tailed P value <0.05 was considered statistically significant.

Results

Baseline characteristics

The study included 73 patients with CSF and 73 controls (Figure 3). Compared with controls, the CSF group had a significantly higher proportion of males (78.1% vs. 57.5%), patients with hyperuricemia (37.0% vs. 20.5%), patients with peripheral arterial disease (30.1% vs. 15.1%), and smokers (45.2% vs. 19.2%). There were no significant differences in age, hypertension, or stroke between the two groups (P>0.05). As illustrated in Table 1, the CTFC of the LAD, LCx, and RCA and the mean TFC among patients with CSF were 28.73±1.51, 31.29±2.37, 33.64±3.03, and 31.22±2.11, respectively.

Figure 3 Flowchart of enrolled controls and patients with CSF. CAG, coronary angiography; CMR, cardiac magnetic resonance; CSF, coronary slow flow; NYHA, New York Heart Association; STEMI, ST-segment elevation myocardial infarction.

Compared to controls, patients with CSF had a higher BMI, TG level, ESV, EDV, and myocardial mass index and a decreased HDL-C level (Figure 4A). There was no obvious difference in LVEF between patients with CSF and controls.

Figure 4 Scatter plots illustrating the comparisons of HDL-C, GLS, and LV PI between patients with CSF and controls. (A) HDL-C was significantly lower in patients with CSF than in controls (P<0.001). (B) GLS was significantly reduced in patients with CSF (P<0.001). (C) LV PI was also significantly decreased in the CSF group relative to the control group (P<0.001). CSF, coronary slow flow; GLS, global longitudinal strain; HDL-C, high-density lipoprotein cholesterol; LV, left ventricular; PI, perfusion index.

Myocardial strain and perfusion in patients with CSF

Patients with CSF had lower GLS compared with controls (–12.09%±2.69% vs. –14.38%±2.36%; P<0.001) (Figure 4B), GCS (–18.70%±3.24% vs. –19.80%±2.21%; P=0.018), and global LV PI (11.34%±4.24% vs. 15.25%±8.50%; P<0.001) (Figure 4C). GRS was not significantly different between the groups (Table 2). Additionally, no perfusion defects were observed in any participants during the resting perfusion scan.

Univariable and multivariable logistic analysis of CSF incidence

As illustrated in Table 3, male gender, higher BMI, the presence of hyperuricemia and peripheral arterial disease, smoking status, lower levels of HDL-C, elevated TG levels, increased myocardial mass index, reduced GLS and GCS, and lower global LV PI were significantly associated with CSF in the univariable logistic analysis. After multivariable analysis and adjustment for confounders, the independent predictors of CSF were found to be HDL-C level [odds ratio (OR) 0.119; 95% CI: 0.016–0.897; P=0.039], GLS (OR 1.339; 95% CI: 1.112–1.613; P=0.002), and global LV PI (OR 0.456; 95% CI: 0.209–0.994; P=0.048).

The diagnostic value of indices

The findings of ROC curves analysis of HDL-C level, GLS, global LV PI, and a combined model (HDL-C + GLS + global LV PI) are provided in Figure 5. The combined model had the best diagnostic performance for predicting CSF (AUC =0.783; 95% CI: 0.710–0.856; P<0.001), followed by GLS (AUC =0.742; 95% CI: 0.660–0.823), HDL-C level (AUC =0.654; 95% CI: 0.565–0.743), and global LV PI (AUC =0.616; 95% CI: 0.524–0.707) (Table 4). HDL-C, GLS, and global LV PI at cutoff values of 1.18 mmol/L, –13.59%, and 12.15%, respectively, were all predictors of CSF. Moreover, the DeLong test demonstrated that the AUC of the combined model was significantly superior to that of HDL-C alone (P=0.003) and global LV PI alone (P<0.001). However, there was no significant difference between the combined model and GLS alone (P=0.130) (Table S1).

Figure 5 ROC curves to evaluate the diagnostic efficiency of CMR parameters and HDL-C in predicting patients with CSF. AUC, area under the curve; CMR, cardiac magnetic resonance; CSF, coronary slow flow; GLS, global longitudinal strain; HDL-C, high-density lipoprotein cholesterol; LV, left ventricular; PI, perfusion index; ROC, receiver operating characteristic.

Reproducibility

The ICC of the myocardial strain and PI are summarized in Table S2. For the measurement of LV myocardium strain, there was effective intraobserver (ICC 0.900–0.926) and interobserver (ICC 0.887–0.913) agreement. Meanwhile, for PI, the ICCs for intraobserver and interobserver variability were 0.958 and 0.911, respectively.

Discussion

This study sought to determine the ability of noninvasive indices derived from CMR first-pass perfusion and feature tracking to predict CSF in patients without obstructive epicardial arterial disease. This study was the first of its kind to use CMR parameters to evaluate CSF. The principal findings of this study are as follows: (I) compared to controls, patients with CSF exhibited reduced global LV PI and impaired LV GLS; (II) HDL-C, GLS, and global LV PI were independent predictors of CSF, and the combined model (HDL-C + GLS + global LV PI) showed acceptable diagnostic efficacy in predicting CSF.

CSF was first described as an angiographic finding in 1972 (20), and it has been linked to angina in patients without obstructive epicardial arterial disease. The etiology and pathogenesis of CSF are unknown. Recent studies have suggested that CSF is associated with impaired endothelial function and elevated coronary microvascular resistance (7). In our study, patients with CSF were more likely to be observed in males and associated with increased BMI and peripheral arterial disease, which is in line with the results of a recent study (21). Li et al. reported that patients with CSF often have hyperlipidemia and increased BMI, consistent with our findings (3). These results suggest that patients with CSF have potential risk factors for coronary atherosclerosis. Furthermore, a recent study found that patients with CSF were more likely to experience cardiovascular-specific and all-cause mortality (22). Thus, it is essential to identify CSF in patients without obstructive epicardial arterial disease. CAG is the gold standard for diagnosis CSF and can do so with accuracy, but its operation relies highly on experienced operators, which limits the detection of CSF. Recent research suggest that noninvasive techniques may also be used to identify CSF. Shereef et al. reported that GLS derived from transthoracic echocardiography was a reliable predictor of CSF in patients without significant stenosis observed during CAG (11). He et al. found that coronary flow reserve was associated with thrombolysis in the myocardial infarction frame count evaluated via single-photon emission computed tomography (23). However, echocardiography is an operator-dependent technique and computed tomography involves radiation. Therefore, identifying CMR indices with which to establish predictive models for CSF may considerably benefit clinical practice.

The emergence of the CMR first-pass perfusion technique has provided clinicians with a more convenient means to quantify myocardial perfusion in those with CSF. To the best of our knowledge, this study is the first to investigate the use of CMR resting perfusion in the diagnosis of CSF. We found that patients with CSF exhibited reduced myocardial PI, and PI was independently associated with CSF in the multivariable logistic analysis (OR =0.456; P=0.048). Another recent study reported that semiquantitative parameters were associated with poor outcomes in patients with cardiac masses (24). These findings collectively suggest that a reduced PI may have prognostic value in CSF, although further longitudinal studies are required to validate this conclusion. Other research has identified association between decreased resting first-pass PI and CMD (25). Although the pathophysiological mechanism of CSF remains unknown, a large body of literature suggests that patients with CSF have multiple risk factors for coronary atherosclerosis, including smoking, increased BMI, dyslipidemia, and peripheral vascular disease (9,11). Thus, these risk factors may be involved in the altered endothelial function and microvascular dysfunction in patients with CSF (26) and may lead to reduced myocardial perfusion. However, no perfusion defects were observed in this study, which is consistent with findings from other work related to patients with angina and nonobstructive CAD (27). We speculated that visual assessment in identifying perfusion defects in patients with CSF may be limited due to reduced global myocardial perfusion.

Few studies have evaluated cardiac function in patients with CSF (11,28), and in our study, we identified impaired LV function in these patients. Furthermore, there was a decrease in GLS and GCS in patients with CSF, which is in line with a previous study (11). Myocardial strain is more efficient and sensitive to LV systolic dysfunction in patients with cardiac diseases. Lenell et al. found that GLS was associated with poor outcomes in patients with acute coronary syndrome (29). Another recent study reported that reduced GLS could efficiently predict the adverse events of patients with preserved LVEF (30). Similarly, in our study found, decreased GLS was independently correlated with CSF occurrence (OR =1.339; P=0.002). This study also found that GCS was mildly decreased in patients with CSF, which is consistent with previous work (10). In addition, Korosoglou et al. reported that circumferential myocardial dysfunction may be linked to increased wall stress and myocardial heterogeneity, such as myocardial fibrosis or ischemic changes due to coronary macrovascular or microvascular disease (31). In our study, GRS and LVEF were not significantly different between the two groups, and these results indicate that patients with CSF may have early-stage ventricular dysfunction (32).

We further found that the presence of a lipid metabolism disorder was associated with CSF. A previous study (33) indicated that higher TG and lower HDL-C levels were present in patients with CSF, which is in line with our findings. Indeed, patients with CSF frequently experience typical angina symptoms and present with common atherosclerotic risk factors. Another study reported that CSF was associated with diffuse atherosclerosis as detected by intravascular ultrasound (34). Thus, we speculate that CSF may be accompanied by subclinical atherosclerosis (26). Moreover, Aciksari et al. reported that a lower level of HDL-C was a predictor of CSF (35), which was corroborated by the multivariate logistic regression in our study (OR 0.119; 95% CI: 0.016–0.897; P=0.039). Furthermore, Aksoy et al. revealed that decreased HDL-C was associated with poor prognosis in patients with CSF after a 10-year follow-up (22). These results suggests that HDL-C exerts a critical effect in the pathogenesis of CSF. It is also worth noting that LDL-C, a well-established risk factor for atherosclerosis, did not differ significantly between the CSF group and the control group in our study. Indeed, whether LDL-C has predictive value for CSF remains controversial, which may be attributed to variations in diagnostic criteria and patient selection across studies (3,7). Therefore, HDL-C appears to be a more reliable predictor of CSF in patients without obstructive epicardial arterial disease.

The combined model (HDL-C + GLS + global LV PI) in our study exhibited acceptable diagnostic performance for predicting CSF, with an AUC of 0.783, a sensitivity of 84.9%, and a specificity of 60.3%. Compared with models described elsewhere (9,36), this model showed comparable diagnostic efficiency. Genç et al. reported an AUC of 0.759 for risk scores in identifying CSF in patients with nonobstructive coronary arterial disease (9), while Akkaya et al. reported that a pan-immune inflammation value yielded an AUC of 0.675 (36). Overall, our combined model demonstrated high sensitivity and may serve as a noninvasive means to predicting CSF in patients with nonobstructive CAD. Accordingly, this model may aid clinicians in risk stratification and inform personalized management decisions for these patients. However, the combined model exhibited relatively low specificity. This may be attributed to the difference in enrollment criteria compared with other studies: we included participants undergoing CAG due to typical angina symptoms and evidence of myocardial ischemia but excluded those without CMR perfusion examination. Additionally, impaired PI, increased GLS, and abnormal HDL-C may not be specific to CSF, as they have also been reported in patients with CMD or angina with nonobstructive CAD (27,37). Hence, we speculate that the combined model may generate false-positive results, leading to decreased specificity.

Limitations

Although all participants were strictly enrolled by cardiologists and CMR imaging was analyzed by radiologists blinded to the CAG results, our study consisted of a retrospective, single-center design with a relatively small sample size and lacked external validity assessment. The generalizability of the findings may thus be limited. Our conclusions need to be further confirmed by multicenter, randomized controlled trials with larger samples. Additionally, the analysis of semiquantitative resting first-pass perfusion imaging relies on visual assessment, rendering it vulnerable to variability in global myocardial perfusion. Therefore, interpreting LV PI in conjunction with myocardial functional parameters and laboratory findings is essential for the accurate prediction of CSF. Moreover, patients with CSF were not followed up to further assess their prognosis. Thus, future prospective studies are warranted to systematically evaluate the incidence of major adverse cardiovascular events, recurrence of chest pain, hospitalization rates, and the frequency of repeat CAG among patients with CSF. Finally, data on inflammatory factors, including C-reactive protein and interleukin, were not included due to the retrospective nature of the analysis.

Conclusions

Patients with CSF exhibited a reduced HDL-C level, decreased myocardial perfusion, and impaired LV systolic function as compared to the controls. Furthermore, HDL-C level, global LV PI, and GLS were independent predictors of CSF in patients without obstructive epicardial arterial disease. However, our findings require confirmation and external validation through further prospective studies with larger samples.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-1668/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-24-1668/dss

Funding: This study was supported by Yunnan Provincial Science and Technology Platform Talent Project (Academician Expert Workstation; No. 202305AF150033) and Yunnan Medical and Health Personnel Special “Xing Dian Talent” Plan (No. XDYC-YLWS-2023-0022).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1668/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Second Affiliated Hospital of Kunming Medical University Ethics Committee (No. PJ2022125). The requirement for individual consent was waived due to the retrospective nature of the analysis. This study was registered on the Chinese Clinical Trial Registry (ChiCTR2200066005).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Beltrame JF. Defining the coronary slow flow phenomenon. Circ J 2012;76:818-20. [Crossref] [PubMed]
2. Xing Y, Zhang Y, Zhao R, Shi J, Chen Y, Chen L, Pan C. Changes of left atrial morphology and function evaluated with four-dimensional automated left atrial quantification echocardiography in patients with coronary slow flow phenomenon and preserved left ventricular ejection fraction. Int J Cardiol 2023;393:131351. [Crossref] [PubMed]
3. Li ZP, Chen J, Xin Q, Pei XY, Wu HL, Tan ZX. Triglyceride glucose-body mass index as a novel predictor of slow coronary flow phenomenon in patients with ischemia and nonobstructive coronary arteries (INOCA). BMC Cardiovasc Disord 2024;24:60. [Crossref] [PubMed]
4. Shim A, Seecheran R, Seecheran V, Persad S, Sreenivasan S, Henry R, Seecheran NA. Suspected Brugada Phenocopy Secondary to Coronary Slow Flow. Case Rep Cardiol 2019;2019:9027029. [Crossref] [PubMed]
5. Montone RA, Galiuto L, Meucci MC, Del Buono MG, Vergni F, Camilli M, Sanna T, Pedicino D, Buffon A, D'Amario D, Giraldi L, Trani C, Liuzzo G, Rebuzzi AG, Niccoli G, Crea F. Coronary slow flow is associated with a worse clinical outcome in patients with Takotsubo syndrome. Heart 2020;106:923-30. [Crossref] [PubMed]
6. Chalikias G, Tziakas D. Slow Coronary Flow: Pathophysiology, Clinical Implications, and Therapeutic Management. Angiology 2021;72:808-18. [Crossref] [PubMed]
7. Li M, Su H, Jiang M, Zuo Z, Su Z, Hao L, Yang J, Zhang Z, Wang H, Kong X. Predictive value of thrombolysis in myocardial infarction frame count for coronary microvascular dysfunction evaluated with an angiography-derived index of microcirculatory resistance in patients with coronary slow flow. Quant Imaging Med Surg 2022;12:4942-52. [Crossref] [PubMed]
8. Gibson CM, Cannon CP, Daley WL, Dodge JT Jr, Alexander B Jr, Marble SJ, McCabe CH, Raymond L, Fortin T, Poole WK, Braunwald E. TIMI frame count: a quantitative method of assessing coronary artery flow. Circulation 1996;93:879-88. [Crossref] [PubMed]
9. Genç Ö, Yıldırım A, Alıcı G, Harbalıoğlu H, Quisi A, Erdoğan A, İbişoğlu E, Bilen MN, Çetin İ, Güler Y, Şeker T, Güler A. Thromboembolic risk scores in patients with non-obstructive coronary architecture with and without coronary slow flow: A case-control study. Int J Cardiol 2023;384:1-9. [Crossref] [PubMed]
10. Liu S, Wang Y, Li J, Li G, Kong F, Mu L, Jia D, Li Y, Yang J, Ma C. Incremental Value of Three-dimensional Speckle-tracking Echocardiography for Evaluating Left Ventricular Systolic Function in Patients with Coronary Slow Flow. Curr Probl Cardiol 2022;47:100928. [Crossref] [PubMed]
11. Shereef AS, Abdelmajeed MG, Alshair MH, El-Dosouky II, Khalil WA, Wageeh S, Shehata IE. Coronary slow flow and its correlation with reduced left ventricle global longitudinal strain: a case-control study. Echo Res Pract 2024;11:2. [Crossref] [PubMed]
12. Huang S, Li Y, Jiang L, Ren Y, Wang J, Shi K, Yan WF, Qian WL, Yang ZG. Impact of Type 2 Diabetes Mellitus on Epicardial Adipose Tissue and Myocardial Microcirculation by MRI in Postmenopausal Women. J Magn Reson Imaging 2022;56:1404-13. [Crossref] [PubMed]
13. Yang MX, Li QL, Wang DQ, Ye L, Li KM, Lin XJ, Li XS, Fu C, Ma XM, Guo YK, Yin RT, Yang ZG. Myocardial microvascular function assessed by CMR first-pass perfusion in patients treated with chemotherapy for gynecologic malignancies. Eur Radiol 2022;32:6850-8. [Crossref] [PubMed]
14. Zhu J, Xie Z, Huang H, Li W, Zhuo K, Bai Z, Huang R. Association of Epicardial Adipose Tissue With Left Ventricular Strain and MR Myocardial Perfusion in Patients With Known Coronary Artery Disease. J Magn Reson Imaging 2023;58:1490-8. [Crossref] [PubMed]
15. Zhu X, Shi Y, Lian J, Shen H, Li L, Wu H, Tian Y, Liu P. Left atrial and left ventricular strain in feature-tracking cardiac magnetic resonance for predicting patients at high risk of sudden cardiac death in hypertrophic cardiomyopathy. Quant Imaging Med Surg 2024;14:3544-56. [Crossref] [PubMed]
16. Gao Y, Jiang YN, Shi R, Guo YK, Xu HY, Min CY, Yang ZG, Li Y. Effects of diabetes mellitus on left ventricular function and deformation in patients with restrictive cardiomyopathies: a 3.0T CMR feature tracking study. Cardiovasc Diabetol 2023;22:317. [Crossref] [PubMed]
17. Kerstens TP, Weerts J, van Dijk APJ, Weijers G, Knackstedt C, Eijsvogels TMH, Oxborough D, van Empel VPM, Thijssen DHJ. Association of left ventricular strain-volume loop characteristics with adverse events in patients with heart failure with preserved ejection fraction. Eur Heart J Cardiovasc Imaging 2023;24:1168-76. [Crossref] [PubMed]
18. McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Böhm M, et al. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J 2021;42:3599-726. [Crossref] [PubMed]
19. Mandrekar JN. Receiver operating characteristic curve in diagnostic test assessment. J Thorac Oncol 2010;5:1315-6. [Crossref] [PubMed]
20. Tambe AA, Demany MA, Zimmerman HA, Mascarenhas E. Angina pectoris and slow flow velocity of dye in coronary arteries--a new angiographic finding. Am Heart J 1972;84:66-71. [Crossref] [PubMed]
21. Harkin KL, Loftspring E, Beaty W, Joa A, Serrano-Gomez C, Farid A, Hausvater A, Reynolds HR, Smilowitz NR. Visual Estimates of Coronary Slow Flow Are Not Associated With Invasive Wire-Based Diagnoses of Coronary Microvascular Dysfunction. Circ Cardiovasc Interv 2024;17:e013902. [Crossref] [PubMed]
22. Aksoy S, Öz D, Öz M, Agirbasli M. Predictors of Long-Term Mortality in Patients with Stable Angina Pectoris and Coronary Slow Flow. Medicina (Kaunas) 2023;59:763. [Crossref] [PubMed]
23. He M, Han W, Shi C, Wang M, Li J, He W, Xu X, Gan Q, Guan S, Zhang L, Chen Y, Chang X, Li T, Qu X. A Comparison of Dynamic SPECT Coronary Flow Reserve with TIMI Frame Count in the Treatment of Non-Obstructive Epicardial Coronary Patients. Clin Interv Aging 2023;18:1831-9. [Crossref] [PubMed]
24. Chan AT, Maya TR, Park C, Tak K, Liberman N, Jain RH, Park MJ, Park RY, Grizzard J, Kim G, Tap WD, Jessurun J, Liu J, Kim J, Steingart RM, Weinsaft JW. Incremental Utility of First-Pass Perfusion CMR for Prognostic Risk Stratification of Cancer-Associated Cardiac Masses. JACC Cardiovasc Imaging 2024;17:128-45. [Crossref] [PubMed]
25. Wang J, Yang ZG, Guo YK, Jiang Y, Yan WF, Qian WL, Fang H, Min CY, Li Y. Incremental effect of coronary obstruction on myocardial microvascular dysfunction in type 2 diabetes mellitus patients evaluated by first-pass perfusion CMR study. Cardiovasc Diabetol 2023;22:154. [Crossref] [PubMed]
26. Zhu Q, Wang S, Huang X, Zhao C, Wang Y, Li X, Jia D, Ma C. Understanding the pathogenesis of coronary slow flow: Recent advances. Trends Cardiovasc Med 2024;34:137-44. [Crossref] [PubMed]
27. Vink CEM, Borodzicz-Jazdzyk S, de Jong EAM, Woudstra J, van de Hoef TP, Chamuleau SAJ, Eringa EC, Götte MJW, Appelman Y. Quantitative perfusion by cardiac magnetic resonance imaging reveals compromised myocardial perfusion in patients with angina with non-obstructive coronary artery disease. Clin Res Cardiol 2025; Epub ahead of print. [Crossref]
28. Javadi DH, Sotudeh DS, Javadi DA, Rezaee DM, Hajikarimi DM. Echocardiographic Evaluation of Left and Right Ventricular Function in Patients with Coronary Slow Flow Syndrome: A Comparative Study. Curr Probl Cardiol 2022;47:100925. [Crossref] [PubMed]
29. Lenell J, Lindahl B, Erlinge D, Jernberg T, Spaak J, Baron T. Global longitudinal strain in long-term risk prediction after acute coronary syndrome: an investigation of added prognostic value to ejection fraction. Clin Res Cardiol 2025;114:709-18. [Crossref] [PubMed]
30. Janwetchasil P, Yindeengam A, Krittayaphong R. Prognostic value of global longitudinal strain in patients with preserved left ventricular systolic function: A cardiac magnetic resonance real-world study. J Cardiovasc Magn Reson 2024;26:101057. [Crossref] [PubMed]
31. Korosoglou G, Sagris M, André F, Steen H, Montenbruck M, Frey N, Kelle S. Systematic review and meta-analysis for the value of cardiac magnetic resonance strain to predict cardiac outcomes. Sci Rep 2024;14:1094. [Crossref] [PubMed]
32. Slivnick JA, Singulane C, Sun D, Eshun D, Narang A, Mazzone S, Addetia K, Patel AR, Zareba KM, Smart S, Kwon JW, Husain A, Cody B, Scheetz S, Asch FM, Goyal A, Sarswat N, Mor-Avi V, Lang RM. Preservation of Circumferential and Radial Left Ventricular Function as a Mitigating Mechanism for Impaired Longitudinal Strain in Early Cardiac Amyloidosis. J Am Soc Echocardiogr 2023;36:1290-301. [Crossref] [PubMed]
33. Kaplangoray M, Toprak K, Deveci E, Caglayan C, Şahin E. Could Pan-Immune-Inflammation Value be a Marker for the Diagnosis of Coronary Slow Flow Phenomenon? Cardiovasc Toxicol 2024;24:519-26. [Crossref] [PubMed]
34. Pekdemir H, Cin VG, Ciçek D, Camsari A, Akkus N, Döven O, Parmaksiz HT. Slow coronary flow may be a sign of diffuse atherosclerosis. Contribution of FFR and IVUS. Acta Cardiol 2004;59:127-33. [Crossref] [PubMed]
35. Aciksari G, Cetinkal G, Kocak M, Atici A, Celik FB, Caliskan M. The relationship between triglyceride/high-density lipoprotein cholesterol ratio and coronary slow-flow phenomenon. Int J Cardiovasc Imaging 2022;38:5-13. [Crossref] [PubMed]
36. Akkaya S, Cakmak U. Association between pan-immune-inflammation value and coronary slow flow phenomenon in patients with angiographically normal coronary arteries. Int J Cardiol 2024;398:131631. [Crossref] [PubMed]
37. Chen F, Weng W, Yang D, Wang X, Zhou Y. Myocardial contrast echocardiography evaluation of coronary microvascular dysfunction to Predict MACEs in patients with heart failure with preserved ejection fraction follow-up. BMC Cardiovasc Disord 2024;24:496. [Crossref] [PubMed]