Prognostic significance of vascular endothelial dysfunction in patients with vasospastic angina

Abstract

BACKGROUND

Vascular endothelial dysfunction (VED) is thought to contribute to the pathogenesis of vasospastic angina (VSA), but its prognostic significance remains unclear.

AIM

To evaluate whether clinical characteristics and prognosis differ according to the presence or absence of VED in patients with VSA.

METHODS

We retrospectively analyzed 244 patients (mean age 67 years, 107 men) admitted for chest pain screening and diagnosed with VSA by spasm provocation testing. Vascular endothelial function was evaluated using brachial artery flow-mediated dilation (FMD). Patients were divided into a high-FMD group (≥ 3.7%; n = 123) and a low-FMD group (FMD-L, < 3.7%; n = 121). VSA was defined as ≥ 90% coronary vasoconstriction on angiography with chest symptoms and/or electrocardiographic changes during spasm provocation testing. Focal spasm was defined as spasm localized within a single American Heart Association coronary segment. Major adverse cardiovascular events included cardiac death and cardiovascular rehospitalization.

RESULTS

Baseline characteristics, except for age (P = 0.017), did not differ significantly between groups. Focal spasm tended to be more frequent in the FMD-L group (43% vs 32%, P = 0.069). Kaplan-Meier analysis showed a significantly lower major adverse cardiovascular events-free survival rate in the FMD-L group (log-rank P = 0.004). Multivariate Cox regression identified FMD-L (P = 0.025) and older age (P = 0.026) as independent prognostic factors, whereas focal spasm was not significant (P = 0.157).

CONCLUSION

VED is an independent predictor of adverse outcomes in VSA. Evaluation of endothelial function may provide valuable prognostic information in patients with VSA.

Key Words: Coronary spasm; Endothelial function; Flow-mediated dilation; Focal spasm; Vasospastic angina; Vascular endothelial dysfunction

Core Tip: Vasospastic angina usually carries a favorable prognosis with lifestyle modification and vasodilator therapy, but reliable noninvasive prognostic markers remain scarce. This study examined whether vascular endothelial dysfunction, assessed by brachial artery flow-mediated dilation, could serve as a marker. Patients with lower flow-mediated dilation values (< 3.7%) had worse outcomes, indicating that vascular endothelial dysfunction is associated with poorer prognosis in vasospastic angina. These findings highlight the potential role of endothelial function assessment in risk stratification, warranting further validation in larger studies.

INTRODUCTION

Vasospastic angina (VSA) is characterized by transient epicardial coronary artery contraction that induces myocardial ischemia[1]. It has recently been recognized as an important cause of angina pectoris and myocardial infarction in the absence of significant coronary stenosis, such as angina with non-obstructive coronary artery disease (CAD) and myocardial infarction with non-obstructive coronary arteries[2,3]. Clinically, VSA often presents as nocturnal resting chest pain and, less frequently, as exertional angina[1]. Although VSA generally responds well to vasodilator therapy and lifestyle modification, suggesting a favorable overall prognosis[4,5], severe complications such as cardiac arrest have also been reported[5,6]. Several prognostic risk factors have been identified, including variant angina (VA), left anterior descending artery (LAD) spasm, beta-blocker use, history of out-of-hospital cardiac arrest, multivessel spasm (MVS), focal spasm, organic coronary stenosis, and the less frequent use of statins or calcium-channel blockers[4,5,7-11].

VSA is thought to arise from multiple mechanisms, including vascular endothelial dysfunction (VED) and hypercontraction of coronary smooth muscle[3,12]. Several techniques have been developed to assess endothelial function[13-16], but brachial artery flow-mediated dilation (FMD) is widely accepted as a reliable, noninvasive clinical method. Numerous studies have evaluated FMD in patients with VSA[17-19]. Despite evidence that VED contributes to the pathogenesis of VSA, its prognostic significance remains insufficiently explored. Therefore, the present study aimed to clarify the clinical characteristics and prognostic impact of VED in patients with VSA diagnosed at our institution.

MATERIALS AND METHODS

Study patients

This retrospective observational study included 284 patients admitted to our hospital from January 2012 to March 2020 who were diagnosed with VSA by coronary angiography (CAG) with spasm provocation testing (SPT) and underwent vascular endothelial function testing at admission. Of these, 40 patients were excluded for the following reasons: Significant coronary stenosis ≥ 50% on CAG or lesions with fractional flow reserve ≤ 0.8 assessed by pressure wire (n = 15), history of percutaneous coronary intervention (n = 14), history of heart failure or hypertrophic cardiomyopathy (n = 8), or unreliable vascular endothelial function test results (n = 3) (Figure 1). The final study cohort comprised 244 patients (107 men, 137 women; mean age 67 years). The study protocol was approved by the institutional ethics committee (approval No. 2024-38). Written informed consent was obtained from all participants for SPT, and an opt-out method (http://www.jrhh.sakura.ne.jp/annnai/torikumi.html) was used to confirm final participation.

Figure 1 Flowchart of the present study. VSA: Vasospastic angina; SPT: Spasm provocation test; PCI: Percutaneous coronary intervention; FMD: Flow-mediated dilation; FMD-H: High-flow-mediated dilation; FMD-L: Low-flow-mediated dilation.

CAG and SPT

The methodologies for CAG and SPT have been described previously[20]. At our institution, the standard procedure was as follows. For the right coronary artery (RCA), acetylcholine (ACh) was administered at doses of 20 μg, 50 μg, and 80 μg over 20 seconds at 3-minute intervals. For the left coronary artery (LCA), ACh was administered at doses of 50 μg, 100 μg, and 200 μg under the same conditions. If spastic activity was pronounced, a reduced dose was used. If ACh did not induce spasm, methylergometrine maleate or additional ACh was administered as needed[21]. After each provocation, CAG was performed. Nitroglycerin (NTG, 0.3 mg) was administered at the end of testing if coronary spasm was not induced or if spasm resolved promptly; a final CAG was performed immediately afterward. If chest pain persisted or hemodynamic instability occurred during provoked spasm, NTG was administered immediately (“unavoidable NTG use”). Following NTG administration to one coronary artery, SPT was performed in the contralateral artery. VSA was diagnosed when coronary spasm was provoked. If no spasm was induced, the diagnosis was recorded as not assessed. Until April 2016[20], SPT was initiated in the RCA; thereafter, testing began in the LCA. This study also documented which coronary artery was used to initiate SPT.

Factors assessed on CAG and SPT

Quantitative CAG was performed as previously described[20]. An atherosclerotic lesion was defined as coronary artery stenosis of > 20% but < 50%. Significant stenosis was defined as ≥ 50% stenosis with a fractional flow reserve ≤ 0.8, when available; such patients were excluded from the study, as noted above. Myocardial bridging was defined as a reduction in systolic coronary artery diameter of > 20% compared with diastolic diameter[20].

In this study, a positive coronary spasm on SPT was defined as transient epicardial coronary artery constriction ≥ 90% during drug provocation, accompanied by electrocardiographic ST-T changes and/or typical chest pain[3]. A focal spasm was defined as spasm confined to a single American Heart Association coronary segment. Patients were categorized as having focal spasm if it was observed in at least one coronary artery (RCA or LCA). MVS was defined as spasm involving two or more major coronary arteries. However, MVS could not be confirmed if testing was terminated after unavoidable NTG administration[20]. A low-dose ACh response was defined as spasm provoked by ≤ 20 μg ACh in the RCA or ≤ 50 μg ACh in the LCA. Additional spasm-related findings included the following: Transient occlusion, temporary occlusion of a coronary artery induced by spasm; ST-segment elevation, during SPT; and unavoidable NTG use, required injection of NTG due to persistent chest pain or hemodynamic instability. Finally, the incidence of coronary spasms was evaluated separately in the LAD, left circumflex (LCX), and RCA.

Chest symptoms provoked by coronary spasm were recorded as occurring at rest, during exertion, or both. The frequency of anginal attacks was calculated as a monthly average. VA was defined as ST-segment elevation on electrocardiography during spontaneous episodes[22]. Electronic medical records and medication charts were reviewed to document drug use at the time of admission. Coronary dilators were discontinued 48 hours before SPT, but any medications taken prior to that period were also recorded. The number of coronary dilators prescribed at discharge was documented. Major adverse cardiovascular events (MACE) were defined as cardiac death or hospital readmission for cardiovascular disease.

Assessment of endothelial function and group classification

Vascular endothelial function was assessed using the conventionally reported method[23]. Testing was performed in the fasting state on the day before CAG. Endothelium-dependent vasodilation was evaluated using FMD of the brachial artery, while the endothelium-independent vasodilation was assessed using NTG-induced dilation (NID). In this study, the median FMD was 3.7% [interquartile range (IQR): 2.1%-5.2%]. Patients were classified into two groups based on this median value: The high-FMD group (FMD-H, n = 123) included patients with FMD ≥ 3.7%, and the low-FMD group (FMD-L, n = 121) included patients with FMD < 3.7%. In addition, the median NID was 14.0% (IQR: 9.7%-19.1%). Patients (n = 238) in whom NID was assessed were classified into a high-NID group (NID-H, n = 119) and a low-NID group (NID-L, n = 119) based on the median value. The NID-H group included patients with NID ≥ 14%, while the NID-L group included patients with NID < 14%. This parameter was used to assess the degree of overlap with the FMD group and to separately assess patient prognosis within the NID group, as FMD and NID factors may influence outcomes[24].

Other clinical characteristics measured

Smoking and alcohol consumption, as well as family history of CAD, were obtained from patients, family members, and electronic medical records. Fasting blood samples were collected on the day of CAG to measure lipids, glucose, C-reactive protein, and brain natriuretic peptide (BNP). Hypertension and dyslipidemia were defined according to established guidelines[25,26]. Diabetes mellitus was defined as fasting blood sugar ≥ 126 mg/dL, hemoglobin A1c ≥ 6.5%, or current use of antidiabetic medications[27]. The estimated glomerular filtration rate was calculated to assess renal function, and chronic kidney disease was determined accordingly[28]. Left ventricular ejection fraction was measured by transthoracic echocardiography using the modified Simpson method[29].

Statistical analyses

Continuous variables with a normal distribution are presented as mean ± SD, while nonnormally distributed or discrete variables are expressed as median (IQR). Between-group comparisons were performed using the unpaired t-test, Wilcoxon signed-rank test, or χ² test, as appropriate. Spearman’s rank correlation coefficient was used to examine the correlation between FMD and NID. The incidence of MACE was assessed using Kaplan-Meier survival analysis, with differences assessed by the log-rank test. Factors associated with MACE were analyzed using a Cox proportional hazards model, with hazard ratios (HR) reported. In the Cox proportional hazards model, three factors were included, corresponding to one-tenth of the total number of MACEs (34 in this study)[30]. Specifically, we investigated the following three factors: Age, FMD-L, and focal spasm. Although the number of cases decreased, we also examined the three factors of age, FMD-L, and BNP, as well as the three factors of age, FMD-L, and NID-L. All statistical analyses were conducted using JMP version 17 (SAS Institute, Cary, NC, United States). A two-sided P < 0.05 was considered statistically significant.

RESULTS

Clinical characteristics and clinical parameters

Of the 244 study participants, 123 (50%) were classified into the FMD-H group and 121 (50%) into the FMD-L group. Patient characteristics are summarized in Table 1. No differences were observed between the two groups in clinical background, except for age (P = 0.017). Blood biochemistry and echocardiographic parameters are presented in Table 2. Lipid and diabetes-related indices were similar between groups. Levels of C-reactive protein and estimated glomerular filtration rate did not differ. BNP was significantly higher in the FMD-L group compared with the FMD-H group. Echocardiographic indices, including left ventricular ejection fraction, were comparable between groups. Brachial artery ultrasonography revealed a larger baseline brachial artery diameter in the FMD-L group. As expected, both FMD and NID were significantly lower in the FMD-L group than in the FMD-H group. FMD was positively correlated with NID (n = 238, r = 0.288, P < 0.001). Regarding the overlap between the FMD and NID groups, there were 72 cases (30%) in the FMD-H/NID-H group, 46 cases (19%) in the FMD-H/NID-L group, 47 cases (20%) in the FMD-L/NID-H group, and 73 cases (31%) in the FMD-L/NID-L group.

Table 1 Patients’ characteristics, n (%)/mean ± SD.

	FMD-H group	FMD-L group	P value
n	123 (50)	121 (50)
Male/female	48/75	59/62	0.125
Age (years)	65 ± 11	69 ± 11	0.017
Body mass index	24.0 ± 4.4	24.3 ± 4.3	0.576
Coronary risk factor
Smoking (active)	18 (15)	27 (22)	0.122
Hypertension	82 (67)	85 (70)	0.547
Dyslipidemia	73 (59)	66 (55)	0.449
Diabetes mellitus	18 (15)	21 (17)	0.562
Family history of CAD	25 (20)	22 (18)	0.671
Alcohol drinker	44 (36)	49 (41)	0.417
CKD	33 (27)	37 (31)	0.517

FMD-H: High-flow-mediated dilation; FMD-L: Low-flow-mediated dilation; CAD: Coronary artery disease; CKD: Chronic kidney disease.

Table 2 Blood biochemical and echocardiographic parameters, mean ± SD.

	FMD-H group	FMD-L group	P value
Blood biochemical parameters
Total cholesterol (mg/dL)	199 ± 29	197 ± 36	0.630
Triglyceride (mg/dL)	136 ± 76	139 ± 69	0.710
HDL-cholesterol (mg/dL)	60 ± 17	58 ± 17	0.359
LDL-cholesterol (mg/dL)	112 ± 26	111 ± 32	0.731
FBS (mg/dL)	105 ± 27	104 ± 18	0.831
HbA1c (%)	5.9 ± 0.7	5.9 ± 0.6	0.959
CRP (mg/dL), median (IQR)	0.06 (0.02, 0.18)	0.07 (0.03, 0.17)	0.268
BNP (pg/mL), median (IQR)	18 (9, 35)	23 (15, 50)	0.005
	n = 120	n = 119
eGFR (mL/minute/1.73 m2)	71.4 ± 15.9	68.7 ± 15.3	0.184
Echocardiographic parameters
LVEF (%)	67 ± 8	68 ± 9	0.513
Brachial ultrasound
Baseline brachial artery diameter (mm)	3.8 ± 0.6	4.0 ± 0.6	0.027
FMD (%)	5.9 ± 4.2	1.7 ± 1.6	< 0.001
NID (%)	16.4 ± 6.8	13.2 ± 6.3	0.002
	n = 120	n = 118

FMD-H: High-flow-mediated dilation; FMD-L: Low-flow-mediated dilation; HDL: High-density lipoprotein; LDL: Low-density lipoprotein; FBS: Fasting blood sugar; HbA1c: Hemoglobin A1c; CRP: C-reactive protein; IQR: Interquartile range; BNP: Brain natriuretic peptide; eGFR: Estimated glomerular filtration rate; LVEF: Left ventricular ejection fraction; FMD: Flow-mediated dilation; NID: Nitroglycerin-induced dilation.

Medications and factors associated with VSA are summarized in Table 3. There were no significant differences between the FMD-H and FMD-L groups in medications at admission. Similarly, the number of coronary dilators used did not differ between groups at either admission or discharge. The frequency of chest symptoms - occurring at rest, during exertion, or as VA - was comparable between groups. The median number of anginal attacks per month at admission also showed no significant difference.

Table 3 Medications and vasospastic angina-related parameters at admission, n (%)/median (IQR).

	FMD-H group	FMD-L group	P value
Medications
Calcium-channel blockers	42 (34)	46 (38)	0.529
Long-acting nitrate	8 (7)	15 (12)	0.115
Nicorandil	9 (7)	8 (7)	0.829
RAS inhibitor	25 (20)	38 (31)	0.103
Beta-receptor blocker	7 (6)	14 (12)	0.102
Statins	52 (43)	37 (31)	0.058
Antiplatelet drug	24 (20)	24 (20)	0.950
Anti-DM drugs	11 (9)	11 (9)	0.968
Number of coronary vasodilators
At admission	0 (0, 1)	0 (0, 1)	0.444
At discharge	1 (1, 1)	1 (1, 1)	0.953
VSA-related parameters
Chest symptoms at rest, on exertion, and both	109/10/4	97/14/10	0.269
VA	1 (1)	2 (2)	0.552
Number of anginal attacks per month	4 (2, 10)	5 (2, 12)	0.524

FMD-H: High-flow-mediated dilation; FMD-L: Low-flow-mediated dilation; RAS: Renin-angiotensin system; DM: Diabetes mellitus; VSA: Vasospastic angina; VA: Variant angina.

Results of CAG and SPT

Findings from CAG and SPT are summarized in Table 4. The prevalence of atherosclerotic lesions and myocardial bridging did not differ between the FMD-H and FMD-L groups. Similarly, there were no significant differences in the coronary artery in which SPT was initiated, the incidence of coronary spasm requiring unavoidable NTG use, or the occurrence of MVS. In contrast, transient occlusion and ST-segment elevation during SPT, as well as the incidence of focal spasm, tended to be higher in the FMD-L group compared with the FMD-H group. The frequency of coronary spasm in the LAD and RCA was similar between groups, whereas spasm in the LCX artery was less frequent in the FMD-L group than in the FMD-H group.

Table 4 Results of coronary angiography and spasm provocation test, n (%).

	FMD-H group	FMD-L group	P value
CAG
Atherosclerotic lesion	59 (48)	60 (50)	0.800
MB	17 (14)	19 (16)	0.679
SPT
Starting from LCA or RCA	35/88	45/76	0.146
Provocation induced by LDA	28 (23)	26 (21)	0.810
TOC during SPT	4 (3)	10 (8)	0.092
ST elevation during SPT	10 (8)	18 (15)	0.098
Unavoidable use of NTG	35 (30)	35 (31)	0.828
	n = 118	n = 113
Focal spasm	39 (32)	52 (43)	0.069
MVS	73 (67)	69 (66)	0.923
	n = 109	n = 104
Location of coronary spasm
LAD	108 (95)	109 (96)	0.527
	n = 114	n = 113
LCX	33 (29)	18 (16)	0.017
	n = 114	n = 113
RCA	73 (66)	70 (67)	0.889
	n = 111	n = 105

FMD-H: High-flow-mediated dilation; FMD-L: Low-flow-mediated dilation; CAG: Coronary angiography; MB: Myocardial bridging; SPT: Spasm provocation test; LCA: Left coronary artery; RCA: Right coronary artery; LDA: Low-dose acetylcholine; TOC: Transient occlusion; NTG: Nitroglycerin; MVS: Multivessel spasm; LAD: Left anterior descending coronary artery; LCX: Left circumflex coronary artery.

Prognosis

The median follow-up was 68 months (IQR: 19-96) in the FMD-H group and 59 months (IQR: 24-94) in the FMD-L group, with no significant difference (P = 0.945). Of the 205 patients (84%) with complete follow-up data, 34 (17%) experienced MACE, including 2 cardiac deaths, 1 nonfatal myocardial infarction, 1 percutaneous coronary intervention for worsening angina, 5 heart failures, 20 rehospitalizations due to angina, 3 other nonpharmacological cardiac interventions, 1 lower extremity revascularization, and 1 cerebral infarction. MACE was higher in the FMD-L group (25/105, 24%) compared with the FMD-H group (9/100, 9%; P = 0.004). Noncardiac death occurred in 11 patients (5% of 216), with no significant difference between groups (P = 0.130). Kaplan-Meier analysis demonstrated a significantly lower MACE-free survival rate in the FMD-L group (log-rank P = 0.004) (Figure 2A), while no significant difference was observed when using the NID group classification (log-rank P = 0.259) (Figure 2B). Cox proportional hazards analysis identified FMD-L [HR: 2.419, 95% confidence interval (CI): 1.115-5.247, P = 0.025] and older age (HR: 1.048, 95%CI: 1.008-1.094, P = 0.026) as independent predictors of MACE, whereas focal spasm was not significant (HR: 1.642, 95%CI: 0.827-3.262, P = 0.157). In a model including BNP (n = 239), FMD-L (HR: 2.440, 95%CI: 1.133-5.257, P = 0.023), as well as age (HR: 1.051, 95%CI: 1.012-1.097, P = 0.016) and BNP (HR: 1.005, 95%CI: 1.001-1.008, P = 0.002), were all significant predictors of MACE. In a model that included FMD and NID groups (n = 238), FMD-L (HR: 2.513, 95%CI: 1.148-5.504, P = 0.021) and age (HR: 1.054, 95%CI: 1.013-1.100, P = 0.012) were significant predictors of MACE. In contrast, NID-L was not significant (HR: 0.990, 95%CI: 0.486-2.015, P = 0.978).

Figure 2 Kaplan-Meier curve for major adverse cardiovascular event-free survival during the follow-up period. A: The high-flow-mediated dilation and low-flow-mediated dilation groups; B: The high-nitroglycerin-induced dilation and low-nitroglycerin-induced dilation groups. MACE: Major adverse cardiovascular event; FMD-H: High-flow-mediated dilation; FMD-L: Low-flow-mediated dilation; NID-H: High-nitroglycerin-induced dilation; NID-L: Low-nitroglycerin-induced dilation.

DISCUSSION

In this retrospective study, we investigated the clinical characteristics of patients with VSA with VED and assessed its prognostic impact. VED was defined as FMD below the median value (< 3.7%). Patients in the FMD-L group were older and had significantly higher BNP levels, and the frequency of coronary spasm in the LCX was significantly lower compared with the FMD-H group. Kaplan-Meier analysis demonstrated a lower MACE-free survival rate in the FMD-L group. In multivariate Cox proportional hazards analysis, FMD-L and older age were significant predictors of MACE. These findings suggest that VED may serve as an independent prognostic factor in patients with VSA.

Although the overall prognosis of VSA is generally favorable with risk factor management and treatment with coronary dilators[4,5], it is important to recognize that coronary spasm can lead to serious cardiovascular events, including sudden death[5,6]. Previously reported prognostic risk factors include VA, spasm of the LAD, beta-blocker use, history of out-of-hospital cardiac arrest, MVS, focal spasm, organic coronary artery stenosis, and the less frequent use of statins or calcium-channel blockers[4,5,7-11]. These traditional prognostic indicators are clinically useful, but some - such as the presence of VA or MVS - are not always readily obtained from routine clinical assessment or CAG and SPT. Moreover, in patients with angina with non-obstructive CAD, clinical characteristics may vary slightly between institutions or regions[31,32]. Consequently, these risk factors may not be applicable to all patient populations. In our institutional data (not shown), factors such as VA, MVS, and LAD spasm were not predictive of outcomes, with only focal spasm showing a significant difference in MACE-free survival. These findings underscore the importance of identifying and reporting reliable prognostic indicators for VSA. In this study, using the Cox proportional hazards model, we noted that VED and advanced age were factors influencing MACE in patients with VSA. Furthermore, despite the limited number of cases, high BNP levels may be a factor influencing MACE in these patients.

Genetic abnormalities suggestive of VED have been reported in patients with VSA[33,34], and VED has been observed in both coronary and peripheral arteries in this population[17,18,35,36]. Thus, VED is considered one of the underlying mechanisms of VSA. Although many studies and meta-analyses have demonstrated that coronary or peripheral VED can serve as a prognostic marker[16,37-39], its prognostic significance specifically in patients with VSA has not been previously investigated. In the present study, VED assessed by FMD was shown to be a useful prognostic indicator in VSA.

Cox proportional hazards analysis suggested that FMD may provide superior prognostic information compared with focal spasm, a conventional risk factor[7,11]. In our cohort, differences in outcomes began to emerge around 1 year and appeared to widen over time. Regarding FMD cutoff values for prognostic assessment, these may vary depending on the patient population. A recent guideline[3] recommends a threshold of 4.0% for clinical use, while this study used the median value of 3.7%, which is similar[3] and considered acceptable. Nevertheless, whether VED is universally valid as a prognostic marker in VSA - including the optimal cutoff value - requires further investigation in multicenter registries and prospective studies. Finally, studies using FMD have reported that baseline brachial artery diameter and reduced NID, which reflect vascular remodeling or vascular smooth muscle dysfunction, correlate with FMD and are considered important indicators of vascular health[15,24]. In our cohort, similar findings were observed in the FMD-L group, which exhibited larger brachial artery diameters and reduced NID, consistent with previous reports by Maruhashi et al[15]. Furthermore, previous studies have reported that reduced NID also influences cardiovascular events[24]. Although FMD and NID showed a weak positive correlation (r = 0.288) in our analysis, the Kaplan-Meier estimates demonstrated that FMD-based grouping is a useful means of distinguishing the occurrence of MACE. Conversely, NID-based grouping did not provide meaningful prognostic differentiation. In the Cox proportional hazards model, which included FMD and NID factors, reduced FMD emerged as a significant predictor of MACE, whereas reduced NID was not a significant factor. These results suggest that FMD has greater utility than NID for prognostic prediction in patients with VSA. However, it should be noted that our findings may have been influenced by the limited sample size and number of MACE events. Future studies are required to further clarify the prognostic impact of FMD and/or NID in patients with VSA.

This study revealed that advanced age is also a factor influencing the prognosis of patients with VSA. Although aging has been demonstrated to be a significant prognostic factor in CAD[40,41], this finding had not been clearly established for VSA to date. In this study, the finding that advanced age is a prognostic factor for VSA has significant clinical implications. Furthermore, in this study, the subgroup analysis, including BNP - performed owing to the relatively small sample size - suggested that an elevated BNP level can serve as a prognostic marker. Notably, although data are not shown, FMD was not correlated with BNP in our cohort. BNP has been reported as a useful prognostic marker in the general population[42] and in patients with CAD[43]. Based on our findings, BNP may also function as a prognostic indicator in patients with VSA. Further validation of whether advanced age and/or elevated BNP levels are prognostic markers in patients with VSA in multicenter registries and prospective studies is warranted.

This study has important clinical significance. First, regarding the FMD measured in this study, FMD is an established test with over 30 years of reported use[44]. FMD measurements may vary depending on institutional experience, equipment, and operator skill[23]. While NTG is commonly employed, FMD can also be evaluated independently when NTG administration is limited[23], and the test itself is considered noninvasive. Therefore, FMD data are considered one of the more clinically accessible indicators. Close monitoring, including increasing the dose of oral coronary vasodilator medication, implementing more rigorous management of lifestyle-related diseases, and shortening the interval between medical visits, is necessary when a marked decrease in FMD is observed, particularly in older adult patients or in patients with elevated BNP levels. In this study, the frequency of LCX coronary spasm was significantly lower in the FMD-L group. LCX spasm is generally less common than spasm in other coronary arteries and is often observed as part of MVS[45]. However, in the present study, no significant difference in MVS incidence was observed between the groups, and the reason for the reduced frequency of LCX spasm in the FMD-L group remains unclear. Further studies are needed to clarify this association.

This study has several limitations. First, it was a retrospective, single-center study with a relatively small sample size, which may limit the generalizability of the findings. Second, although endothelial function testing is typically performed after discontinuation of coronary dilators, it is unclear whether the effects of these medications were fully eliminated. Third, while endothelial function can improve with pharmacologic therapy or lifestyle modification, it was assessed only once during hospitalization in this study. Fourth, several studies have reported an association between the slow-flow phenomenon and VED[46]. Furthermore, studies have suggested that patients with coronary spasm combined with coronary microvascular dysfunction exhibit a poor prognosis[32]. The present study did not investigate the existence of the slow-flow phenomenon. Whether FMD is reduced in patients with VSA who also demonstrate the slow-flow phenomenon and whether such patients exhibit a poorer prognosis are topics worth addressing in future research. Finally, the frequency of chest pain was recorded as the number of episodes per month, and more objective evaluation methods were not employed[31].

CONCLUSION

In this retrospective study, we evaluated the clinical characteristics and prognostic significance of peripheral VED, assessed by FMD, in patients with VSA. Patients were classified into two groups based on the median FMD value of 3.7%. The FMD-L group was older and exhibited higher BNP levels, as well as a lower frequency of LCX coronary spasm. The incidence of MACE was higher in the FMD-L group, and Kaplan-Meier analysis demonstrated significantly reduced MACE-free survival. In multivariate Cox proportional hazards analysis, FMD-L and older age were identified as independent risk factors for MACE. These findings suggest that FMD may serve as a useful prognostic marker in patients with VSA. However, whether these results, including the optimal FMD cutoff value, are universally applicable requires confirmation in future multicenter registries and prospective studies.

ACKNOWLEDGEMENTS

We thank Ms. Akemi Seno for her secretarial assistance. We also extend our gratitude to the staff of the catheterization laboratory, cardiovascular ward, and cardiovascular outpatient clinic at our institution for their support.