ANGPTL3 and Cardiovascular Outcomes in Patients With Acute Coronary Syndrome and Obstructive Sleep Apnea

Abstract

Background

The aim of this prospective study was to determine the impact of elevated ANGPTL3 (angiopoietin‐like protein 3) on cardiovascular events (CVEs) following acute coronary syndrome (ACS) in patients with or without obstructive sleep apnea (OSA).

Methods and Results

A total of 1174 patients with ACS underwent successful percutaneous coronary intervention were included in this prospective cohort study (NCT03362385). Patients were categorized according to the apnea–hypopnea index (≥15 events/h, OSA) and further classified by ANGPTL3 levels. We analyzed the incidence of CVEs in patients with ACS according to the status of OSA and ANGPTL3. During a median of 3.1 years of follow‐up, 217 (18.48%) CVEs occurred. The patients with ACS with OSA had higher ANGPTL3 levels than those without OSA (30.4 [20.9–43.2] versus 27.80 [19.1–41.5] ng/mL; P<0.001). In all patients with ACS, 29≤ANGPTL3<42 mg/dL and ANGPTL3≥42 mg/dL were associated with an increased risk of CVEs with hazard ratios (HRs) of 1.555 (95% CI, 1.010–2.498) and 2.489 (95% CI 1.613–3.840), respectively. When the status of OSA or not was incorporated in stratifying factors, 29≤ANGPTL3<42 mg/dL and ANGPTL3≥42 mg/dL were associated with a significantly higher risk of CVEs in patients with ACS with OSA (HR, 1.916 [95% CI, 1.019–3.601] and HR, 2.692 [95% CI, 1.379–4.503]) but not without OSA. Moreover, adding ANGPTL3 to the Cox model increased C‐statistic values by 0.035 and 0.029 in the OSA group and all patients with ACS, respectively, but was not statistically improved in patients with ACS without OSA.

Conclusions

In conclusion, our study demonstrates a predictive impact of plasma ANGPTL3 on cardiovascular risk in patients with ACS, especially in patients with ACS with OSA. It might be of clinical value in refining risk stratification and tailoring treatment of patients with ACS and OSA.

Registration

URL: https://www.clinicaltrials.gov; Unique identifier: NCT03362385.

Nonstandard Abbreviations and Acronyms

AHI

apnea–hypopnea index

CVE

cardiovascular event

Q3

quartile 3

Q4

quartile 4

Clinical Perspective.

What Is New?

• This prospective study found that the ANGPTL3 (angiopoietin‐like protein 3) level was independently associated with the increased risk of cardiovascular events in patients with acute coronary syndrome, especially in patients with acute coronary syndrome with obstructive sleep apnea.

What Are the Clinical Implications?

• Plasma ANGPTL3 might be of clinical value in refining risk stratification and tailoring treatment of patients with acute coronary syndrome and obstructive sleep apnea.

ANGPTL3 (angiopoietin‐like protein 3) is an important regulator of lipoprotein metabolism and has emerged as a pharmacological target for managing dyslipidemia and reducing coronary artery disease (CAD). ¹ , ² , ³ ANGPTL3 is produced exclusively in the liver, and its primary role is to inhibit lipoprotein lipase, the rate‐limiting enzyme that removes circulating triglycerides, although it also inhibits endothelial lipase, an enzyme more specialized for hydrolysis of lipoprotein phospholipids, particularly in high‐density lipoprotein particles. ⁴ In humans, there is strong evidence that the inactivation of ANGPTL3 is associated with significant reductions in plasma levels of triglyceride, low‐density lipoprotein cholesterol (LDL‐C), very‐low‐density lipoprotein cholesterol, and high‐density lipoprotein cholesterol (HDL‐C) as well as a lower risk of CAD. ¹ , ²

Obstructive sleep apnea (OSA) is a complex and common chronic disease defined by the presence of repetitive episodes of upper airway collapse with the specific features such as intermittent hypoxia and increased arousals, affecting 20% to 30% of the general population. ⁵ Correspondingly, OSA is a common condition in patients with CAD, affecting approximately 40% to 60% of these patients. ⁶ Indeed, basic, epidemiological, and clinical studies have indicated that OSA is associated with an increased risk of CAD and may aggravate the severity and prognosis of CAD in patients. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ The current data suggest that OSA can cause proatherogenic dyslipidemia and selectively increase triglyceride‐rich lipoproteins. ¹² , ¹³ , ¹⁴ However, the mechanism of OSA on serum lipids remains poorly characterized. The use of continuous positive airway pressure for OSA is the first‐line treatment strategy. Recently, 3 randomized controlled trials reported a neutral effect of continuous positive airway pressure treatment on secondary cardiovascular prevention, whereas randomized controlled trials assessing the effects of continuous positive airway pressure on lipid metabolism presented inconclusive results. ¹² , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ In consideration of the clinical issue and the character of ANGPTL3 to antagonize triglyceride hydrolysis and as a potential pharmacological target for managing dyslipidemia, it is crucial to clarify the role of ANGPTL3 in patients with ACS and OSA, which is an important guide for clinical practice.

Although strong evidence from epidemiological, genetic, and prospective cohort studies has verified that circulating ANGPTL3 levels are associated with the presence of CAD, ¹⁹ , ²⁰ it remains unclear whether the circulating ANGPTL3 levels could predict the risk of recurrent cardiovascular events (CVEs) in patients with ACS, particularly in those with OSA. Therefore, we performed this study to determine the relationship between plasma ANGPTL3 levels and CVEs in patients with ACS with and without OSA.

Methods

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

Study Design and Participants

The OSA‐ACS project (NCT03362385) is a large‐scale, single‐center, prospective cohort study designed to assess the relationship between OSA and cardiovascular outcomes in patients with ACS. As described in the flowchart (Figure 1), 1505 patients aged 18 to 85 years and admitted for ACS who underwent coronary angiography and were successfully treated with percutaneous coronary intervention in the Emergency & Critical Care Center of Beijing An Zhen Hospital, Capital Medical University (Beijing, China) between June 2017 and May 2019 were eligible for inclusion, and the median period of follow‐up was 3.1 years.

Figure 1. Flowchart of the study.

ACS indicates acute coronary syndrome; AHI, apnea–hypopnea index; CPAP, continuous positive airway pressure; OSA, obstructive sleep apnea; and PCI, percutaneous coronary intervention.

The ACS definition and exclusion criteria are provided in Data S1.

This study complied with the Strengthening the Reporting of Observational studies in Epidemiology guidelines and was conducted in accordance with the Declaration of Helsinki. The Ethics Committee of Beijing An Zhen Hospital, Capital Medical University approved the study (2013025), and all patients provided written informed consent.

Patients were followed up at 6‐month intervals for at least 1 year. Trained nurses or physicians who were blinded to the clinical data completed the interviews. Our primary end points (CVEs) were cardiovascular mortality, nonfatal myocardial infarction, and stroke. Our secondary outcome was all‐cause mortality. Survival time was measured from the time of enrollment to censoring (event, loss to follow‐up, or latest date of follow‐up).

Study Procedures and Management

All patients underwent an overnight sleep study using full‐night, laboratory‐based polysomnography (Grael, Compumedics, Australia) according to the recommendations of the American Academy of Sleep Medicine. The sleep studies were performed after clinical stabilization during hospitalization (within 2 weeks after admission). Respiratory polysomnography studies were performed without supplemental oxygen. Nasal airflow, thoracoabdominal movements, snoring episodes, electrocardiography, and pulse oximetry were recorded. An apnea event was defined as the absence of airflow lasting ≥10 seconds (obstructive if thoracoabdominal movement was present and central if thoracoabdominal movement was not). The hypopnea event was defined as a >30% reduction in airflow lasting for ≥10 seconds with oxygen desaturation. Oxygen desaturation was defined as a decrease in arterial oxygen saturation of ≥4%. The apnea–hypopnea index (AHI) was defined as the number of apneas and hypopneas per hour of total recording time. OSA was defined as the AHI ≥15 events per hour and severe OSA as the AHI ≥30 events per hour. A minimum of 3 hours of satisfactory polygraphy signal recording was considered as a valid test. All sleep studies were manually double scored by independent sleep technologists. In cases of discrepancy, further scoring was performed by a senior consultant in sleep medicine.

All patients received standard care during ACS hospitalization according to current guidelines. ²¹ , ²² The specific treatment was described in Data S1. For patients with moderate‐to‐severe sleep apnea (AHI ≥15 events per hour), particularly those with excessive daytime sleepiness, we referred them to a sleep center for further evaluation and consideration of continuous positive airway pressure therapy.

Additional details of the method of the data collection are described in Data S1.

Statistical Analysis

The continuous variables were compared by independent Student t test and nonparametric Mann–Whitney U test and by Pearson χ² test or Fisher exact test for categorical variables, where appropriate. Spearman correlation coefficient was used for analyzing correlations of continuous variables. The cumulative event rates among groups were estimated by the Kaplan–Meier method and compared by the log‐rank test. Univariate and multivariate Cox regression analyses were performed to calculate the hazard ratios (HRs) of CVEs with OSA and ANGPTL3 as a categorical variable divided into quartiles. HRs were expressed relative to the lowest quartile group of ANGPTL3. The multivariate Cox regression was adjusted for traditional cardiac risk factors, including age, sex, body mass index (BMI), HDL‐C, LDL‐C, smoking, presence or absence of diabetes, and hypertension. The effect of ANGPTL3 levels on outcomes in relation to OSA status was evaluated using a Cox proportional hazards model that included ANGPTL3 and the OSA status by ANGPTL3 interaction. To assess the discriminatory ability of the models, Harrell's C‐index was estimated. Subgroup analysis of the effect of OSA on CVEs was performed by using important characteristics in a multivariable‐adjusted Cox regression model.

A statistical significance level of 0.05 was used. Statistical analysis was performed with SPSS 20.0 (IBM Corp., Armonk, NY) and R language, version 3.6.2 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Baseline Characteristics

As presented in Figure 1, from a total of 1505 patients with ACS included in the OSA‐ACS cohort, we included 1174 patients with available data for the variables explored in the present study. The anthropometric variables, biological determinants, pharmacological treatment, clinical background and lifestyle habits, and main polysomnography variables of the study participants are shown in Tables 1 and 2. The mean±SD age of the patients was 50±1.6 years, and 85% were men. The mean±SD BMI of the patients was 27.0±5.2 kg/m², and the median (interquartile range) AHI was 17.7 (8.6–32.9) events per hour. Of the patients, 630 (53.66%) were diagnosed as OSA with AHI ≥15 events per hour (312 with AHI ≥30 events per hour). The patients in the OSA group were more likely to experience excessive daytime sleepiness than those in the non‐OSA group (P<0.001). The BMI, neck circumference, diastolic blood pressure, hemoglobin A_1c, and hsCRP (high‐sensitivity C‐reactive protein) were positively associated with the status of OSA. Patients in the OSA group were more likely to be men and to have a higher prevalence of hypertension and dyslipidemia (P<0.001). Patients with OSA had higher levels of triglyceride (P<0.001) and free fatty acid (P=0.008) than the non‐OSA group. Meanwhile, patients with OSA had significantly lower levels of HDL‐C (P=0.001) and left ventricle ejection fraction (P=0.003) but higher levels of creatine kinase isoenzyme, myoglobin, troponin I (all P<0.001), and BNP (B‐type natriuretic peptide) (P=0.004), and Gensini score (P=0.018) than the non‐OSA group.

Table 1.

Baseline Characteristics and Clinical Presentations of the Participants

Characteristic	All	Non‐OSA	OSA	P value
	N=1174	n=544	n=630
Anthropometric measures
Sex, male	998 (85.0%)	443 (79.8%)	555 (88.1%)	0.001*
Age, y	50±1.6	50±1.5	50±1.6	0.626
Body mass index, kg/m2	27.0±5.2	26.0±5.0	27.8±4.6	<0.001**
Waist–hip ratio	0.99 (0.95–1.02)	0.98 (0.94–1.02)	0.99 (0.96–1.02)	0.114
Neck circumference, cm	41 (38–43)	40 (38–42)	42 (39–44)	<0.001**
Heart rate, beats/min	70 (65–78)	70 (65–76)	70 (65–80)	0.901
Systolic BP, mm Hg	126 (116–139)	125 (116–137)	126 (117–140)	0.278
Diastolic BP, mm Hg	77 (70–85)	75 (69–83)	79 (70–87)	0.007*
Lifestyle risk factors
Smoking				0.068
Current	592 (50.5%)	256 (47.1%)	336 (53.3%)
Former	187 (15.9%)	87 (16.0%)	102 (15.9%)
Never	394 (33.6%)	200 (36.8%)	192 (30.5%)
Drinking				0.266
Current	433 (36.9%)	188 (34.6%)	245 (38.9%)
Former	61 (5.2%)	27 (5.0%)	34 (5.4%)
Never	679 (57.9%)	328 (60.4%)	351 (55.7%)
Medical history
Hypertension	825 (70.2%)	353 (64.9%)	472 (74.9%)	<0.001**
Diabetes	367 (31.3%)	160 (29.5%)	207 (32.9%)	0.212
Dyslipidemia	1032 (88.0%)	465 (85.6%)	567 (90.0%)	<0.001**
Previous CVD	595 (50.7%)	271 (49.8%)	324 (51.4%)	0.615
Previous PCI	284 (24.2%)	127 (23.3%)	157 (24.9%)	0.619
Variables related to ACS severity
ACS type				0.051
Unstable angina	866 (73.8%)	436 (80.1%)	430 (68.3%)
NSTEMI	167 (14.2%)	61 (11.2%)	106 (16.8%)
STEMI	141 (12.0%)	53 (9.7%)	88 (14.0%)
LVEF (%)	63 (58–66)	63 (60–66)	62 (57–66)	0.003*
Gensini score	28 (12–54)	26 (10–54)	32 (12–54)	0.018*
In‐hospital management
Lipid‐lowering agent	490 (41.8%)	228 (42.0%)	262 (41.7%)	0.187
Antidiabetic oral medication or insulin	353 (30.2%)	167 (30.8%)	186 (29.6%)	0.884
Antiplatelet and antithrombotic agents	516 (44.1%)	242 (44.6%)	274 (43.7%)	0.405

Continuous variables are presented as mean±SD or median and interquartile range (25th to 75th percentile). ACS indicates acute coronary syndrome; BMI, body mass index; BP, blood pressure; CABG, coronary artery bypass grafting; CVD, cardiovascular disease; LVEF, left ventricular ejection fraction; NSTEMI, non–ST‐segment–elevation myocardial infarction; OSA, obstructive sleep apnea; PCI, percutaneous coronary intervention; and STEMI, ST‐segment–elevation myocardial infarction.

^* P<0.050.

^** P<0.001.

Table 2.

Sleep Study Results and Laboratory Data of the Participants

Characteristic	All	Non‐OSA	OSA	P value
	n=1174	n=544	n=630
OSA measures
Apnea–hypopnea index, events/h	17.7 (8.6–32.9)	8.2 (4.8–11.6)	29.6 (21.7–42.8)	<0.001**
Oxygen desaturation index>4%, /h	16.9 (9.1–29.9)	9.05 (5.0–12.4)	27.7 (19.5–39.8)	<0.001**
Mean SaO2	94 (93–95)	94 (93–95)	93 (92–94)	<0.001**
Minimum SaO2	85 (81–88)	87 (84–90)	83 (78–86)	<0.001**
Time with SaO2<90%, %	3 (0.1–12.0)	1 (0–3.0)	6 (2.0–17.7)	<0.001**
Longest apnea time, s	33 (21–48)	22 (14–34)	40 (29–59.8)	<0.001**
Longest hypopnea time, s	86 (63–97)	78 (54–93)	91 (71–98)	<0.001**
Epworth Sleepiness Scale	7 (4–11)	6 (3–10)	8 (4–12)	<0.001**
Laboratory data
Glucose, mmol/L	6.1 (5.3–7.6)	5.9 (5.3–7.4)	6.1 (5.4–7.7)	0.080
Hemoglobin A1c, %	6.0 (5.7–7.0)	5.8 (5.4–6.8)	6.1 (5.7–7.1)	0.008*
Leptin, ng/mL	5.4 (3.0–9.4)	5.1 (2.4–8.5)	5.7 (3.2–10.0)	0.014*
hsCRP, mg/L	1.81 (0.68–5.82)	1.45 (0.59–3.97)	2.15 (0.77–7.3)	<0.001**
Interleukin 1β, pg/mL	7.37 (5.97–9.61)	6.44 (5.23–9.42)	7.97 (6.11–9.93)	<0.001**
Creatinine (μmol/L)	67.8 (57.8–77.3)	66.8 (56.9–77.4)	68.4 (58.4–77.2)	0.637
Total cholesterol, mmol/L	4.05 (3.42–4.92)	4.06 (3.36–4.99)	4.07 (3.46–4.82)	0.633
HDL‐C, mmol/L	1.02 (0.88–1.21)	1.04 (0.91–1.23)	1.00 (0.86–1.17)	0.001*
LDL‐C, mmol/L	2.36 (1.82–3.11)	2.37 (1.77–3.13)	2.36 (1.88–3.03)	0.960
Triglyceride, mmol/L	1.51 (1.07–2.13)	1.40 (0.96–1.98)	1.57 (1.13–2.24)	<0.001**
Lipoprotein (a), nmol/L	0.12 (0.05–0.29)	0.13 (0.05–0.32)	0.11 (0.04–0.25)	0.050
Free fatty acid, mmol/L	0.47 (0.33–0.63)	0.43 (0.32–0.61)	0.50 (0.36–0.64)	0.008*
CKMB, ng/mL	2.0 (1.2–3.6)	1.7 (1.2–2.7)	2.0 (1.4–4.8)	<0.001**
Myoglobin, ng/mL	22.7 (16.5–32.7)	22.2 (15.9–30.2)	23.4 (17.05–37.0)	<0.001**
Troponin I, ng/mL	0.10 (0.00–0.40)	0.10 (0.00–0.25)	0.10 (0.01–0.61)	<0.001**
BNP, pg/mL	44 (21–109)	37 (19–86)	49 (24–137)	0.004*
ANGPTL3, ng/mL	29.2 (19.0–42.3)	27.80 (19.1–41.5)	30.4 (20.9–43.2)	<0.001**

Continuous variables were presented as mean±SD or median and interquartile range (25th to 75th percentile). ANGPTL3 indicates angiopoietin‐like protein 3; BNP, B‐type natriuretic peptide; CKMB, creatine kinase isoenzyme; HDL‐C, high‐density lipoprotein cholesterol; hsCRP, high‐sensitivity C‐reactive protein; LDL‐C, low‐density lipoprotein cholesterol; and OSA, obstructive sleep apnea.

^* P<0.050.

^** P<0.001.

Of note, the ANGPTL3 level was higher in the OSA group than in the non‐OSA group (30.4 [20.9–43.2] versus 27.80 [19.1–41.5] ng/mL [P<0.001]; Table 2, Figures S1 and S2). Furthermore, the level of ANGPTL3 was significantly correlated with sex, BMI, Epworth Sleepiness Scale, glucose, leptin, total cholesterol, triglyceride, hsCRP, interleukin 1β, myoglobin, troponin I, and Gensini score. Also, higher ANGPTL3 levels were significantly associated with increasing AHI and oxygen desaturation index and decreasing minimum SaO₂ (Table S1).

OSA, ANGPTL3 Levels, and Cardiovascular Outcomes

During a median follow‐up time of 3.1 years (25th–75th percentile, 2.4–4.1), 217 CVEs occurred (4‐year cumulative incidence estimate, 21.6%), and 23 instances of all‐cause mortality occurred. In Kaplan–Meier analyses, no significant difference was found in the incidence of all‐cause mortality in the higher level ANGPTL3 group than in the lower level ANGPTL3 group (log‐rank P=0.76). The Kaplan–Meier curves for CVEs are shown in Figure 2A. The crude incidence of a CVE was higher in the OSA group than in the non‐OSA group (4‐year estimate, 24.6% versus 16.4%; P=0.0026). OSA predicted the incidence of CVEs in unadjusted Cox regression analysis (HR, 1.517 [95% CI, 1.152–1.997]), and the statistically significant association between them was retained after adjustments for age, sex, BMI, HDL‐C, LDL‐C, smoking, presence or absence of diabetes, and hypertension (HR, 1.426 [95% CI, 1.063–1.913]). Subgroup analyses were performed based on important baseline characteristics (Figure S3). OSA was associated with an increased risk of CVEs in most subgroups; however, no statistically significant effect of OSA was found for the previous‐CVD phenotype (adjusted HR, 1.121 [95% CI, 0.747–1.682]).

Figure 2. Kaplan–Meier analysis for cardiovascular events according to OSA status (A) and different ANGPTL3 levels (B) (with 95% confidence limits, shaded area).

ANGPTL3 indicates angiopoietin‐like protein 3; OSA, obstructive sleep apnea; Q1, quartile 1 (ANGPTL3<19 ng/mL); Q2, quartile 2 (19≤ANGPTL3<29 ng/mL); Q3, quartile 3 (29≤ANGPTL3<42 ng/mL); and Q4, quartile 4 (ANGPTL3≥42 ng/mL).

In the current study, ANGPTL3 levels of the population had skewed distribution (Figure S1). By ANGPTL3 status (quartile 1, ANGPTL3<19 ng/mL; quartile 2, 19≤ANGPTL3<29 ng/mL; quartile 3 [Q3], 29≤ANGPTL3<42 ng/mL; quartile 4 [Q4], ANGPTL3≥42 ng/mL), patients in Q4 were the most likely to have CVEs (4‐year cumulative incidence estimate, 29.0%; P<0.001) (Figure 2B). However, when the patients were evaluated according to both OSA status and ANGPTL3 status (Figure S4), those in the higher quartiles (Q3, Q4) had a significantly higher risk of CVEs than the reference group (patients without in quartile 1) in patients with OSA (all P<0.050); conversely, patients with OSA in quartile 1 and quartile 2 had a similar prognosis as patients without OSA (regardless of the level of ANGPTL3). As shown in Figure 3, among patients with OSA, the higher level ANGPTL3 was strongly associated with a higher rate of CVEs at 4 years compared with patients with lower level ANGPTL3 (Q4 30.2% and Q3 28.4% versus quartile 2 17.4% and quartile 1 15%; P<0.050). A similar trend was also observed in patients with ACS without OSA, but the difference was not significant.

Figure 3. Kaplan–Meier analysis for cardiovascular events according to different ANGPTL3 levels in patients with ACS with OSA and without OSA.

ANGPTL3 indicates angiopoietin‐like protein 3; OSA, obstructive sleep apnea; Q1, quartile 1 (ANGPTL3<19 ng/mL); Q2, quartile 2 (19≤ANGPTL3<29 ng/mL); Q3, quartile 3 (29≤ANGPTL3<42 ng/mL); and Q4, quartile 4 (ANGPTL3≥42 ng/mL).

As presented in Table 3, multivariate Cox regression analyses according to both OSA status and ANGPTL3 status in patients with ACS were performed. In all patients with ACS, higher quartiles of ANGPTL3 (Q3, Q4) were associated with a higher risk of CVEs with HRs of 1.555 (95% CI, 1.010–2.498) and 2.489 (95% CI, 1.613–3.840) in Q3 and Q4 in adjustment model 1 (adjusted for traditional cardiac risk factors including age, sex, BMI, HDL‐C, LDL‐C, smoking, presence or absence of diabetes, and hypertension) and with HRs of 1.332 (95% CI, 1.009–2.997) and 1.632 (95% CI, 1.069–3.997) in Q3 and Q4 in adjustment Model 2 (adjusted for factors in model 1 and troponin I) compared with the lowest quartile. Furthermore, significant interaction between ANGPTL3 level by the OSA status (OSA versus non‐OSA) was observed for the CVEs in the unadjusted and adjusted model (all P for interaction <0.001). In addition, a statistically significant effect of higher levels of ANGPTL3 was found for the patients with OSA with adjusted HRs of 1.916 (95% CI, 1.019–3.601) and 2.692 (95% CI, 1.379–4.503) in Q3 and Q4 in model 1 and adjusted HRs of 1.822 (95% CI, 1.021–2.933) and 2.446 (95% CI, 1.206–3.113) in Q3 and Q4 in model 2 compared with the lowest quartile. Conversely, there was no significant association between the level of ANGPTL3 and the occurrence of CVEs in patients without OSA (Figure 3, Table 3).

Table 3.

Cox Regression Model of the Effect of ANGPTL3 Status on the Risk of Cardiovascular Events by the Status of OSA

Model	ANGPTL3 quartile (Q1–Q4), hazard ratio (95% CI)				P value
	Q1 (<19 ng/mL)	Q2 (19–29 ng/mL)	Q3 (29–42 ng/mL)	Q4 (≥42 ng/mL)
All patients
Unadjusted	Referent	1.264 (0.897–2.056)	1.612 (1.012–2.566)	2.617 (1.704–4.018)	<0.001**
Adjusted	Referent	1.115 (0.968–2.498)	1.555 (1.010–2.498)	2.489 (1.613–3.840)	<0.001**
OSA
Unadjusted	Referent	1.250 (0.943–2.430)	2.000 (1.082–3.697)	2.701 (1.390–4.495)	<0.001**
Adjusted	Referent	1.101 (0.974–1.960)	1.916 (1.019–3.601)	2.692 (1.379–4.503)	<0.001**
Non‐OSA
Unadjusted	Referent	1.345 (0.659–2.744)	1.315 (0.644–2.684)	1.605 (0.704–2.688)	0.522
Adjusted	Referent	1.348 (0.656–2.770)	1.315 (0.636–2.717)	1.711 (0.852–2.844)	0.731
OSA–ANGPTL3 interaction
Unadjusted	Referent	1.111 (0.875–1.830)	1.773 (1.155–2.724)	2.220 (1.506–3.274)	<0.001**
Adjusted	Referent	1.007 (0.907–1.569)	1.747 (1.103–2.766)	2.281 (1.534–3.392)	<0.001**

Adjusted for age, sex, body mass index, high‐density lipoprotein cholesterol, low‐density lipoprotein cholesterol, smoking, presence or absence of diabetes, and hypertension. ANGPTL3 indicates angiopoietin‐like protein 3; OSA, obstructive sleep apnea; Q1, quartile 1; Q2, quartile 2; Q3, quartile 3; and Q4, quartile 4.

^* P<0.050.

^** P<0.001.

Risk Prediction for CVEs in Patients With ACS With or Without OSA

In Cox prediction models to predict the CVEs, the C‐statistic value was 0.636 (95% CI, 0.586–0.686) with traditional cardiac risk factors, and the C‐statistic value was 0.642 (95% CI, 0.577–0.700) with traditional cardiac risk factors and troponin I. The addition of ANGPTL3 to the traditional model resulted in a significant improvement in C‐statistic value for patients with ACS (C‐statistic value 0.665 [95% CI, 0.511–0.719]), and the addition of ANGPTL3 to the traditional cardiac risk factors and troponin I model also resulted in a significant improvement in C‐statistic value for patients with ACS (C‐statistic value 0.671 [95% CI, 0.492–0.743]). Furthermore, the C‐statistic value improved for the prediction of CVEs when adding ANGPTL3 to the traditional model, with the statistic value improving from 0.664 (95% CI, 0.556–0.672) to 0.699 (95% CI, 0.692–0.706) for patients with ACS with OSA, but not for those without OSA.

Discussion

In the present analysis from the prospective OSA‐ACS study of 1174 patients with ACS, the prevalence of OSA was 53.66%. We verified the increased risk of long‐term CVEs in patients with ACS with OSA compared with those without OSA, namely, a 1.5‐fold increased risk of CVEs. Moreover, we found that ANGPTL3 level was higher in the OSA group than in the non‐OSA group. Finally, we found that the ANGPTL3 level is independently and strongly associated with an increased risk of CVEs in all patients with ACS, especially in patients with ACS with OSA but not without OSA.

Some research and our previous studies have suggested a role for OSA in the initiation, progression, or adverse cardiovascular outcomes of CAD. ⁷ , ¹⁰ , ¹¹ , ²³ , ²⁴ , ²⁵ In our study, we also verified that OSA had a deleterious effect in increasing recurrent CVEs, and the effect is significant in patients mainly characterized by no previous CVD and admission for a first ACS, in accordance with previous literature. ⁷ In the present study, patients with OSA had a 24.6% 4‐year CVEs rate. In addition, we found OSA was related to higher levels of creatine kinase isoenzyme, myoglobin, and troponin I, which was also reported in other studies. ²⁶ , ²⁷ , ²⁸ OSA may result in oxidative stress, systemic inflammation, dyslipidemia, and insulin resistance, which may contribute to coronary atherosclerosis and ACS events. Current data from animal models to clinical evidence suggest that OSA can cause dyslipidemia that selectively increases triglycerides rich in lipoproteins, leading to CAD. ¹² , ²⁹ In our study, OSA was also associated with higher triglyceride, total cholesterol, free fatty acid, leptin, hemoglobin A_1c, and hsCRP levels, which may attribute to the deleterious effect of OSA in patients with ACS.

A number of novel biomarkers have been proposed in the risk stratification of patients with ACS. The biomarker ANGPTL3 is most interesting in this context as it is involved in lipid metabolism, and several studies have revealed that plasma ANGPTL3 levels are positively associated with circulating LDL‐C and HDL‐C. ³⁰ In our study, we found a striking relationship between ANGPTL3 and CVEs in patients with ACS with the presence or absence of OSA for the first time. Indeed, a meta‐analysis of 19 studies involving 21 980 patients with CAD and 158 200 control participants reported a 34% lower risk of CAD and notably lower levels of total cholesterol, LDL‐C, and triglyceride among carriers of the ANGPTL3 loss‐of‐function mutation compared with noncarriers. ¹⁹ Several studies have investigated the possible relationship between CAD and plasma ANGPTL3 levels. In a population‐based study, patients in the lowest tercile of plasma ANGPTL3 levels showed a reduced myocardial infarction risk by 29% compared with those in the highest tercile. ¹⁹ Similarly, our previous study has demonstrated that elevated levels of ANGPTL3 were independently associated with a higher likelihood of CAD in patients with OSA. ³¹ Our study extends the prior analysis of a small sample ³² of patients with CAD by focusing on the role of the ANGPTL3 in patients with ACS with OSA and without OSA. We found that the ANGPTL3 level was associated with the severity of ACS (Gensini score, myoglobin, troponin I) and was independently and strongly associated with an increased risk of CVEs in all patients with ACS. This study, for the first time, showed that the ANGPTL3 level was increased in patients with OSA, which highlighted the potential role of ANGPTL3 in OSA. Of note, the higher level of ANGPTL3 in patients with ACS with OSA was independently associated with a higher CVE rate but not without OSA.

The higher level of ANGPTL3 is more prevalent in patients with ACS with OSA, and may reflect dyslipidemia, insulin resistance, and inflammation predisposing to future CVEs in these patients, which may relate with hypoxia. ³³ , ³⁴ , ³⁵ , ³⁶ Specifically, in humans who have homozygous loss‐of‐function mutations in ANGPTL3, levels of plasma glucose and insulin are markedly lower, as are measures of insulin resistance when compared with heterozygotes and noncarriers. ³⁷ The recent reports have shown that OSA reduced the clearance of triglyceride‐rich lipoproteins from human plasma through the inhibition of lipoprotein lipase. ³⁸ , ³⁹ ANGPTL4 (angiopoietin‐like protein 4; a member of the ANGPTL [angiopoietin‐like protein] family) has been shown to be increased in patients with OSA, and chronic intermittent hypoxia has also been shown to play an important role in inhibiting lipoprotein lipase activity and increasing ANGPTL4 levels, leading to dyslipidemia. ⁴⁰ , ⁴¹ , ⁴² The study reported that ANGPTL8 (angiopoietin‐like protein 8) levels were increased in patients with OSA, suggesting that the upregulation of these lipid metabolism regulators might play a role in lipid dysregulation observed in people with OSA. ⁴⁰ However, the association of OSA with ANGPTL3 was not clear. Importantly, we found that ANGPTL3 level was most strikingly associated with OSA and related indicators, such as AHI, minimum SaO₂, and oxygen desaturation index, which has not been reported. In our analysis, ANGPTL3 level was not only associated with prognosis but also with a wide range of CVE risk factors, such as BMI, total cholesterol, triglyceride, glucose, and leptin, which are related to dyslipidemia and insulin resistance. There was no significant relationship between the ANGPTL3 level and LDL‐C as previously reported, probably because of the use of statins in most patients with ACS. In terms of inflammation, there was a significant positive relationship between ANGPTL3 and CRP (C‐reactive protein) and interleukin 1β in our present study.

The increased C‐index for CVEs indicate that the addition of plasma ANGPTL3 to information from established cardiovascular risk factors might represent an important improvement in risk stratification for outcomes in patients with ACS with OSA. Specifically, evinacumab (a human monoclonal antibody of ANGPTL3) was approved by the US Food and Drug Administration for patients with homozygous familial hypercholesterolemia. ⁴³ Considering the strong causal relationship between plasma ANGPTL3 and CAD risk, treatment modalities targeting ANGPTL3 are recently emerging in preclinical and clinical trials, so the plasma ANGPTL3 may play the important prediction value independent of plasma lipid levels in the future.

The current study has some limitations. First, this study was among Chinese patients with ACS recruited at a single center and thus may not apply to other populations. Second, we measured ANGPTL3 levels only at baseline, and the follow‐up levels of ANGPTL3 may also be clinically significant. Third, other family members of ANGPTL3 (such as ANGPTL4 and ANGPTL8) were also associated with the dyslipidemia and the risk with CVEs, and further study is needed to figure out their role in patients with ACS with OSA. Fourth, although several published articles have used the same ELISA to test plasma ANGPTL3, the proper validation has not been done. Fifth, we will go further study to expand the sample size to validate the role of ANGPTL3 in patients with ACS with or without OSA.

Conclusions

In conclusion, our study demonstrates a predictive impact of plasma ANGPTL3 on cardiovascular risk in patients with ACS, especially in patients with ACS with OSA. It might be of clinical value in refining risk stratification and tailoring treatment of patients with ACS and OSA.

Sources of Funding

This study was supported by the National Key Research and Development Program of China (Grant Nos. 2021YFC2500600 and 2021YFC2500603), the National Natural Science Foundation of China (Grant No. 81970224, 81670331) and the Beijing Natural Science Foundation (Grant No. 7192030).

Disclosures

None.

Supporting information

Data S1

Table S1

Figures S1–S4

References 44, 45, 46