Between May 2020 and July 2022, a total of 426 patients who presented with acute coronary syndrome (ACS) and underwent OCT and were subsequently examined with FMD. These patients were recruited from the Second Affiliated Hospital of Harbin Medical University in Harbin, China. STEMI, NSTEMI, and unstable angina were all identified as ACS. The criteria for the diagnosis of ACS have been described previously [5, 14]. The patients provided written informed consent, and the present study was approved by the Ethics Committee of the Second Hospital of Harbin Medical University (Harbin, China).

B-mode ultrasound images (UNEX EF; Unex Co., Ltd., Nagoya, Japan) were used to measure vasodilator responses in brachial arteries, as described in previous studies [8, 9]. Patients were required to fast for at least 6 h prior to vascular scans. The measurement of FMD was required before coronary intervention, unless it conflicted the guideline-recommended therapy strategy. In these cases, FMD measurements were permitted within 2 weeks of hospitalization. The standard FMD measurement algorithm was based on expert consensus guideline for reducing variations in the process of FMD measurement [13]. The ultrasound probe was placed between 1 and 5 cm above the brachial artery to obtain optimal FMD images for all patients. Vessel diameter and blood flow responses to reactive hyperemia and nitroglycerin were expressed as percentage increases in from their respective baseline values. Impaired FMD was defined as 7.0% (calculated as the mean minus one standard deviation of FMD).

The Cardiovascular Angiography Analysis System (CAAS), version 5.10 (Pie Medical Imaging B.V., Maastricht, Netherlands) was used to perform quantitative coronary angiography (QCA) analysis. The QCA parameters, including reference vessel diameter, minimal lumen diameter, diameter stenosis, and lesion length, were measured as described in a previous study [15]. The culprit artery was determined based on the severity of the angiographic atherosclerosis, ECG changes, and OCT findings.

OCT imaging was performed using the commercially available C7-XR/ILUMIEN OCT system (Abbott Vascular, Santa Clara, CA, USA). The decision to perform OCT imaging was based on the operator’s discretion without prespecified angiographic or FMD demands. OCT imaging was routinely performed in most ACS patients except those with renal dysfunction, or unstable hemodynamics. OCT analyses were independently performed by two investigators (B.Z. and K.Y.) who were blinded to the clinical, angiographic, laboratory, and FMD data using an offline review workstation (Abbott Vascular). Any discordance was resolved by consensus with a third reviewer (W.M.). Quantitative and qualitative analyses of all lesions were performed as previously described [15]. To identify the culprit lesions, angiography, electrogram changes and/or left ventricular wall motion abnormalities were collectively evaluated. Quantitative analysis was performed using 1-mm intervals of cross-sectional OCT images. PE were identified by the presence of attached thrombi overlying an intact and visible plaque, an irregular luminal surface without thrombi, superficial lipid, or calcification immediately accompanied by attenuation of the underlying plaque by a thrombus. PR was characterized by a discontinuous fibrous cap with an intraplaque cavity [4, 15, 16].

All patients were followed for 1, 3, 6 and 12 months and subsequently annually by phone or hospital visits. Major adverse cardiovascular events (MACEs) were defined as a composite of cardiac death, nonfatal myocardial infarction, clinical-driven revascularization, and rehospitalization for unstable or progressive angina. All events were adjudicated by the independent Clinical Events Committee (CEC) of the Second Affiliated Hospital of Harbin Medical University.

Statistical analysis was performed using the SPSS software (SPSS version 23.0, IBM, Armonk, New York, USA). Data distribution was assessed using the Kolmogorov–Smirnov test. Normally distributed continuous variables are presented as mean $\pm$ standard deviation and examined using Student’s t-test. Non-normally distributed continuous variables are presented as medians (interquartile ranges) and examined using the Mann–Whitney U test. Categorical data are presented as counts (proportions) and were compared using the chi-square test or Fisher’s exact test. The association of demographic and traditional risk factors, plaque characteristics, FMD, and culprit mechanisms (PE/PR) was analyzed using a multivariable logistic regression model with stepwise selection of the variable (p 0.1 in the univariate analysis). The predictability of PR or thin cap fibroatheroma (TCFA) with FMD was determined by receiver operating characteristics curves analysis. Kaplan-Meier analysis was used to present time-to-event data and compared by log-rank test. The predictor of MACEs was identified by multivariable Cox regression model. A two-sided p-value 0.05 was considered statistically significant.

Patients undergoing both OCT and FMD testing (n = 426), were recruited between May 2020 to July 2022 and subsequently included in the final analysis. The detailed inclusion and exclusion criteria are shown in Fig. 1. Of these patients, 326 (76.5%) presented with impaired FMD, while 100 (23.5%) patients presented with normal FMD, as summarized in Table 1. The baseline clinical characteristics showed no significant demographic differences between the impaired and normal FMD groups, except for hypertension (69.6% vs. 53.0%, p = 0.002). Patients with impaired FMD exhibited non-significant trends towards both older age (60.0 years vs. 56.4 years, p = 0.053) and higher LDL-C levels (2.2 mmol/L vs. 2.0 mmol/L, respectively; p = 0.052) than those with normal FMD.

Fig. 1.

Inclusion and exclusion criteria study flow-chart. Between May 2020 and July 2022, 426 patients who underwent both FMD and OCT were included in the final analysis. Notably, over half of patients with impaired FMD (FMD 7.0%) exhibited plaque rupture. ACS, acute coronary syndrome; OCT, optical coherence tomography; FMD, flow-mediated vasodilation.

The majority of affected vessels (53.1%) were found in the left anterior descending artery. The culprit artery locations evenly distributed between the left anterior descending (53.4% vs. 52.0%, respectively), left circumflex (20.9% vs. 23.0%, respectively), and right coronary arteries (25.8% vs. 25.0%, respectively). There were no significant differences in quantitative coronary analysis in terms of reference vessel diameter, minimal lumen diameter, diameter stenosis, and lesion length (Table 2).

The baseline brachial artery diameter was 4.2 $\pm$ 0.6 mm, with an average FMD of 4.0% (interquartile: 2.6–7.0%). Analysis of FMD indicated that 76.5% of the patients exhibited impaired FMD, defined as FMD 7.0%. Conversely, normal FMD (FMD $\ge$7.0%) was observed in 23.5% of patients. The distribution of the different spectra of the FMD is presented in Fig. 2.

Fig. 2.

Distribution of FMD spectra and their associated OCT-mechanism. (A) Bar graph depicting the distribution of impaired FMD and non-impaired FMD in patients with ACS. Red indicates impaired FMD (7.0%) and blue indicates non-impaired FMD ($\ge$7.0%). (B) The left sector chart indicates the distribution of OCT mechanisms in patients with impaired FMD. The right-sector chart shows the distribution of OCT mechanisms in patients with normal FMD. ACS, acute coronary syndrome; OCT, optical coherence tomography; FMD, flow-mediated vasodilation; CN, calcified nodules; PR, plaque rupture; PE, plaque erosion.

According to the established OCT criteria, within the study population, 225 (52.8%) experienced a PR, 161 (37.8%) experienced a PE, 25 (5.9%) experienced a calcified nodule, and 15 (3.5%) experienced other complications including 3 (0.7%) spasm, 6 (1.4%) SCAD and 6 (1.4%) tight stenosis. The details of the other mechanisms are presented in Supplementary Table 1. Patients with impaired FMD presented with PRs more frequently than those with normal FMD (58.9% vs. 33.0%, respectively), but this group also presented fewer incidences of PE (30.1% vs. 63.0%, respectively) (Fig. 2 and Table 3). Multivariable analysis revealed that patients with impaired FMD had a 4.2-fold higher risk of PR (odds ratio 4.22, 95% CI: 2.07–6.72; p = 0.012) than those with normal FMD, after adjusting for potential confounders (Table 4). The receiver operates characteristics (ROC) analysis demonstrated that impaired FMD could accurately predict PR (area under curve [AUC] = 0.878, 95% CI: 0.826–0.934, p 0.001) (Fig. 3). Additionally, patients with impaired FMD were more likely to present with red thrombi compared to those with normal FMD (61.0% vs. 47.0%, respectively; p = 0.013). Representative cases illustrating the measurement of impaired FMD with OCT-based PR and normal FMD with OCT-based PE are presented in Supplementary Figs. 1,2.

Fig. 3.

Receiver operating characteristics curve analysis. Receiver operating characteristics curve analysis to predict PR from FMD. The area under curve = 0.878, 95% confidential interval = 0.826–0.934, p 0.001. PR, plaque rupture; FMD, flow-mediated vasodilation.

Plaques in patients with impaired FMD were found to be more vulnerable, as defined by the presence of a fibro cap thickness (FCT) 65 µm (28.2% vs. 13.0%, respectively; p = 0.002) and minimal lumen area (MLA) 3.5 mm${}^2$ (46.0% vs. 23.0%, respectively; p = 0.007), when compared to patients with normal FMD (Table 3 and Fig. 4). Additionally, the presence of lipid-rich plaques was significantly higher in the impaired FMD group compared to the normal FMD group (36.5% vs. 19.0%, respectively; p = 0.001). Notably, the impaired FMD score was moderately predictive of thin cap fibroatheroma (TCFA, AUC = 0.766, 95% CI: 0.691–0.840, p 0.001) (Supplementary Fig. 3). No differences were observed in other plaque features, including cholesterol crystals, microchannels, calcification, and macrophages, and the proportion of lipid plaques were similar between the two groups (55.2% vs. 45.0%, respectively; p = 0.073).

Fig. 4.

Illustration of OCT findings, stratified by normal FMD versus impaired FMD. OCT findings comparing impaired and normal FMD are presented: (A) culprit mechanisms and plaque characteristics, including PE, PR, MLA 3.5 mm${}^2$, TCFA, LRP, and macrophage; (B) MLA stratified by normal and impaired FMD; (C) other plaque characteristics, including cholesterol crystal, calcification, microchannel, lipid plaque, and red/white thrombus; (D) maximal lipid arc stratified by normal and impaired FMD. PR, plaque rupture; PE, plaque erosion; MLA, minimal lumen area; TCFA, thin-cap fibroatheroma; LRP, lipid-rich plaque; FMD, flow-mediated vasodilatation; OCT, optical coherence tomography.

All patients completed their scheduled one-year follow-up. The composite endpoint outcomes and their components are detailed in Table 5. The Kaplan-Meier curve shows the cumulative incidence of major adverse cardiac events (MACE) over time for the patients with impaired and normal FMD (Fig. 5). Incidences of MACEs occurred in 9.5% of patients with impaired FMD and 3.0% of patients with normal FMD (hazard ratio [HR] = 3.23, 95% CI: 1.47–7.12, p = 0.039). A multivariable Cox regression model revealed that impaired FMD was an independent predictor of adverse events (HR = 3.12, 95% CI: 1.27–6.58, p = 0.039) after controlling for potential confounding factors.

Fig. 5.

Kaplan-Meier curves comparing MACE incidence based on FMD status. There was a significant difference in MACE between patients with impaired FMD and normal FMD. MACE incidents included cardiac death, nonfatal myocardial infarction, revascularization, and rehospitalization for angina. MACE, major adverse cardiac event; FMD, flow-mediated vasodilatation; HR, hazard ratio.

Microcirculatory dysfunction is linked to the development and progression of atherosclerosis and thrombosis. Diminished FMD may indicate systemic atherosclerotic risk, which consequently predicts adverse cardiovascular events. Endothelial vasodilation is largely mediated by nitro-oxide (NO); impairment of NO availability leads to endothelial dysfunction [8]. The response of vascular smooth cells is vital for FMD. Overexpression of peroxisome proliferator–activated receptor $\gamma$ coactivator 1-$\alpha$ (PGC-1$\alpha$) enhances NO and hydrogen peroxide during FMD [17]. Additionally, prostacyclin and NO are the main mediators of FMD in younger and older patients, respectively [18]. Therefore, understanding the diverse mechanisms regulating FMD, including the roles of NO and PGC-1$\alpha$, is crucial for identifying potential therapeutic targets to mitigate systemic atherosclerotic risk and improve cardiovascular outcomes.

In a previous study, the FMD percentage was 7.6 $\pm$ 2.5 in patients with ACS, results that are higher than the data we have presented [19]. No significant difference in FMD was observed between patients with ACS and those with stable CAD [19]. However, Kitta et al. [20] reported that the baseline FMD was 3.0 $\pm$ 1.5% in patients with coronary artery disease (CAD). In another study, the baseline percentage of FMD was 2.1 $\pm$ 1.2% in patients with non-ST-elevated ACS [21]. Three-quarters of patients presenting with acute coronary syndrome after PCI were diagnosed with endothelial dysfunction, defined as FMD 7.0% (Figs. 1,2). These findings highlight the variability in FMD measurements across different patient populations and underscore the need for more standardized approaches to assess endothelial function, particularly in the context of ACS and CAD.

Pathological PR and PE are the primary causes of ACS, having been reported in approximately 75% and 25% of ACS cases, respectively, aligning with our results [6, 22]. Jia et al. [3] first established OCT as an in vivo diagnostic algorithm for PE. Due to its high resolution (10–15 µm), OCT currently provides the best diagnostic imaging for PE [5, 23, 24, 25, 26]. Endothelial dysfunction, assessed by OCT-quantified FMD, may precede the asymptomatic vasculature atherosclerosis, potentially predicting future MACE events [19]. However, there is limited evidence of advanced atherosclerosis and endothelial dysfunction in patients with ACS.

This study is the first to highlight the increased risk of PR in patients with impaired FMD. The pro-inflammatory effects of endothelial dysfunction may be a contributing factor to the higher incidence of PR in these patients [27, 28]. This is supported by a previous study showing that impaired FMD was associated with severe coronary stenosis [29]. This suggests that patients with impaired FMD are more likely to experience PR, as severe atherosclerosis is more frequent in patients with PR than PE [4].

Because FMD impairment of can serve as an independent predictor of future adverse cardiovascular events, FMD screening may be an ideal tool for clinicians to develop both long-term and short-term risk management strategies. Emerging evidence suggests that high-risk plaque characteristics, such as TCFA, lipid-rich plaque, MLA 3.5 mm${}^2$, and a large plaque burden, can elevate the risk of major adverse events [30, 31, 32]. In patients with ACS and impaired FMD, the vascular structure exhibited increased plaque vulnerability, more TCFAs, and smaller MLA compared with those with unimpaired FMD. This increased vulnerability at the site of the culprit lesion may have systemic effects on pan-vascular plaque stability [15]. Therefore, FMD impairment is associated with greater plaque vulnerability and may lead to poor clinical outcomes in these at-risk patients.

This study does have several limitations. First, as a retrospective single-center study, it may contain potential confounding factors related to the limited patient population. Second, FMD measurements were not performed routinely for all patients with ACS at the study center. Although no significant differences were observed between patients who underwent FMD measurement and the overall patient group, the non-routine nature of FMD measurements could introduce bias. However, not requiring target patients to undergo this examination might have partially reduced selection bias. Third, the OCT findings in non-culprit plaques were not analyzed due to the non-routine conduction of multivessel OCT for all patients. Finally, the lack of a uniform standard FMD measurement algorithm may have led to variations in the measurement. However, the FMD measurement protocol in this present study was based on updated consensus guidelines [13], bolstering confidence in our results and their applicability to clinical practice and future clinical trials.