Circulating lipid and lipoprotein profiles and their correlation to cardiac function and cardiovascular outcomes in patients with acute myocardial infarction.

Recent studies showed that lipoproteins represent major risk factors, both positive and negative, for atherosclerotic cardiovascular disease. The aim of the present study was to describe the relationship between plasma lipid profile and cardiac function and cardiovascular outcomes in patients with acute myocardial infarction (AMI) after percutaneous coronary intervention (PCI). Two independent groups of subjects including a total of 797 patients diagnosed of AMI undergoing PCI admitted to the First Affiliated Hospital of Xi'an Jiaotong University were included in the present study. We performed a cross-sectional study for the correlation between plasma lipid profile and cardiac function based on the first group, including 503 patients with AMI. We further validated the correlation and did the follow-up of 2.4 years of major cardiovascular outcomes on the second group, including 294 patients with AMI. Our results showed that apolipoprotein A-I (ApoA-I) level was significantly reduced, and the high-density lipoprotein cholesterol (HDL-C):ApoA-I ratio was increased in the patients with lower LVEF or higher N-terminal pro-B-type natriuretic peptide levels compared with the control; there was a positive correlation between cardiac function and ApoA-I, and a negative correlation between cardiac function and the HDL-C:ApoA-I ratio. Meanwhile, multivariate Cox analysis showed that ApoA-I was independent predictors of major adverse cardiovascular events (MACEs). Kaplan-Meier survival analysis showed the ApoA-I levels exhibited a significant effect on predicting the incidence of MACEs. In sum, plasma ApoA-I level is positively associated with the cardiac function of patients with AMI after PCI, and ApoA-I is an independent indicator to predict the incidence of MACEs. © American Federation for Medical Research 2021. Re-use permitted under CC BY-NC. No commercial re-use. Published by BMJ.

The attack of acute myocardial infarction (AMI) is life-threatening, and long-term chronic ischemia of the myocardium will cause adverse clinical outcomes, such as ischemic heart failure as well as fatal arrythmia.

The abnormal lipid metabolism took a critical part in plaque formation and atherosclerosis in the development of AMI.

However, few studies were concerned about the relationship between whole lipid types and cardiac function, and the correlation of prognosis with lipid profile remains controversial due to the lack of long-term follow-up results.

In this study, we found that apolipoprotein A-I (ApoA-I) levels were significantly reduced, and the high-density lipoprotein cholesterol (HDL-C):apolipoprotein A-I (ApoA-I) ratio were increased in the patients with lower LVEF or higher N-terminal pro-B-type natriuretic peptide level.

The Pearson correlation analysis showed positive correlations between cardiac function and ApoA-I, and negative correlations between cardiac function and the HDL-C:ApoA-I ratio.

Moreover, ApoA-I levels exhibited a significant effect on predicting the incidence of major adverse cardiovascular events (MACEs).

The present study provides broad and straightforward support that ApoA-I should be introduced into clinical practice for the assessment of the cardiac function in patients with AMI undergoing percutaneous coronary intervention, and also predicts the incidence of MACEs.

Introduction

Coronary artery disease (CAD) is the major cause of mortality and morbidity in China and worldwide. Despite the technological advancement and the increasing level of awareness,1 2 acute myocardial infarction (AMI) is still a life-threatening emergency, and long-term chronic ischemia of the myocardium will cause adverse clinical outcomes, such as ischemic heart failure as well as fatal arrythmia.3 Actually, the condition of patients with AMI who may have poor prognosis could be greatly improved by timely and appropriate interventions. Based on this, finding an efficient predictor related with cardiac function and cardiovascular outcomes is urgently needed.

Lipid abnormalities have been widely documented to be associated with higher cardiovascular disease (CVD) risk.4 Widely used clinical CVD risk calculators frequently include classical biochemistry measures of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), or a combination of these.5–7 However, the predictive value of non-traditional lipid risk factors has also gained increasing attention from researchers, including high-density lipoprotein cholesterol (HDL-C), non-HDL, apolipoprotein A-I (ApoA-I), etc.8 HDL-C was ascribed as ‘good’ cholesterol and negatively correlates to the risk of CVDs as proven by several clinical and animal studies.9–12 Non-HDL-C has been suggested as a pragmatic and cost-effective alternative to direct LDL-C measurement, also proven to be associated with increased CVD risk.13 ApoA-I is the principal protein component of HDL particles and is also of interest for its potential value for predicting CVD risks.14–16 Specifically, ApoA-I level is a consistent discriminator of atherosclerotic burden among patients with stable CAD.17 However, the correlation between ApoA-I and cardiac function, as well as long-term outcomes in patients with AMI undergoing percutaneous coronary intervention (PCI), remain underexplored.

With these considerations, our work was conducted to evaluate the lipid profile of patients with AMI after PCI and the relationship among dyslipidemia, cardiac function and long-term cardiovascular outcomes.

Materials and methods

Study population

This study enrolled two independent groups of subjects including a total of 797 patients diagnosed with AMI undergoing PCI, admitted to the First Affiliated Hospital of Xi’an Jiaotong University. We performed a cross-sectional study of the correlation between plasma lipid profile and cardiac function based on the first group including 503 patients with AMI admitted between January 2013 and December 2015. We further validated the correlation and did the follow-up of 2.4 years on the second group including 294 patients with AMI between January 2016 and December 2016; 28 (9.52%) patients were lost follow-up. AMI was defined based on the universal definition criteria by the joint European Society of Cardiology (ESC)/American College of Cardiology Foundation/American Heart Association/World Heart Federation Task Force.18 The exclusion criteria were (1) age<18 years, (2) pregnancy, (3) renal dysfunction (serum creatinine>133 μmol/L) or liver dysfunction (serum alanine transaminase>3 times the upper normal limit), and (4) malignant tumors. All patients received guideline-recommended therapy for AMI. The detailed demographic, clinical, drug, hematological, echocardiography and angiographic data were obtained from the hospital documentation.

The estimation of sample size was performed using G*Power software V.3.1.9.6.19 A sample size of 252 achieves 95% power to detect an effect size of 0.25 using F tests with a significance level of 0.05.

Lipid profile measures

Venous blood samples for lipidomic analyses were collected before coronary catheterization. The following laboratory assays were performed in the clinical laboratory department: TC and triglyceride (TG) were detected using detection kit from FUJIFILM via N-(3-sulfopropyl)- 3-methoxy-5-methylaniline (HMMPS) method; HDL-C and LDL-C were detected using detection kit via direct measurement method from FUJIFILM; ApoA-I, Apo B and Apo E were measured using a detection kit from SEKISUI by turbidimetric inhibition immunoassay. All laboratory assays were performed in duplicate and the results were averaged.

Other blood biochemical measures

Standard clinical biochemical and hematological measures were made by the local laboratory of the First Affiliated Hospital of Xi’an Jiaotong University. Serum was collected for analysis including liver, kidney and electrolytes (HITACHI 7180; HITACHI, Tokyo, Japan). Full blood samples were used to test the hematological parameters (KX 21 n analyzers; Sysmex, Kobe, Japan). After these tests, all samples were stored at −80°C for future analysis. The serum N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels were detected as a batch analysis in a central laboratory by electrochemiluminescence immunoassay (Roche Diagnostics, Rotkreuz, Switzerland).

Evaluation of the echocardiography

Echocardiography was performed using Philips iE33 ultrasound system (Philips, Amsterdam, Netherlands) by experienced cardiologists of the First Affiliated Hospital of Xi’an Jiaotong University. The left ventricular ejection fraction (LVEF) value was uniformly measured by biplane Simpson rule.20

Assessment of outcomes

Patients in the follow-up cohort were followed up semiannually by clinic visits or by telephone interviews conducted by trained nurses or doctors. Major adverse cardiovascular event (MACE) is defined as the end point of this study, which referred to the composite of all-cause death, heart failure, non-fatal myocardial infarction (MI), and symptom-driven revascularization. The follow-up ended on June 31, 2018 or patient death.

Statistical analysis

All statistical analyses were performed by SPSS V.22.0 for Windows. Data were presented as frequencies and percentages for categorical variables, as mean±SD for normally distributed continuous variables and median (with 25th and 75th percentiles) for non-normally distributed continuous variables. The Kolmogorov-Smirnov test was used to assess normal distribution of quantitative variables. Simple t-test was used to compare continuous variables which are in normal distribution. Kruskal-Wallis test was used to compare continuous variables which do not conform to the normal distribution. χ2 test was used to compare categorical variables. To describe the relationship between plasma lipid profile and cardiac function, patients were divided into subgroups according to the baseline LVEF levels and the NT-proBNP level. The cut-off point of LVEF was defined based on the 2016 ESC guidelines for the diagnosis and treatment of acute and chronic heart failures, which represent varying degrees of cardiac function (<40%, 40%–50%, and >50%). To limit the influence of extreme observations, NT-proBNP was natural logarithmically transformed to obtain log NT-proBNP. Then patients were grouped according to the tertiles of baseline log NT-proBNP level. Simple linear analysis was used to calculate the correlation between plasma lipid profile and cardiac function. Univariate and multivariate survival analyses involving Cox regression analysis were constructed to calculate HRs and 95% CIs for MACEs. The multivariable Cox models were adjusted by age, sex, height, weight, creatinine, LVEF, and NT-proBNP. To assess the prognostic value of the ApoA-I level, Kaplan-Meier survival curves were used and compared by log-rank test. All probability values were two-tailed. A p value of <0.05 was considered statistically significant.

Results

Baseline characteristics of the first group

A total of 503 patients with a diagnosis of AMI after PCI were enrolled in the first group. The patients were grouped according to the baseline LVEF levels (<40%, 40%–50%, and >50%). Baseline characteristics of patients in different LVEF subgroups are shown in table 1. Compared with the patients in the group with higher levels of LVEF, the patients with lower levels of LVEF showed a higher level of heart rate (84.43±19.92 vs 72.86±12.26, p<0.001). In addition, the patients with lower levels of LVEF had significantly lower levels of ApoA-I level (1.02±0.24 vs 1.11±0.19, p<0.05) in the plasma. Furthermore, no differences were found in other lipids levels among different LVEF groups, such as TG, TC, LDL-C, HDL-C, non-HDL, the HDL-C:LDL- C ratio and the HDL-C:ApoA-I ratio. No differences in other risk factors were found between different LVEF level groups, such as age, gender, alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatine, and blood urea nitrogen (BUN).

Table 1

Baseline characteristics of the first group

Characteristics	Grouped by LVEF (%)			P value	Grouped by log NT-proBNP (pg/mL)			P value
	<40 (n=47)	40–50 (n=79)	>50 (n=377)		<6.82 (n=140)	6.82–9.06 (n=142)	>9.06 (n=139)
Age (years)	66.21±12.71	66.58±12.88	66.46±11.94	ns	68.07±12.23	66.13±11.23	65.09±13.33	ns
Male, n (%)	36 (76.60)	61 (77.22)	245 (64.99)	ns	86 (61.42)	99 (69.71)	104 (74.82)	ns
Height (cm)	171.39±5.74	169.13±5.94	168.88±7.40	ns	168.41±7.461	168.46±7.59	170.58±5.59	ns
Weight (kg)	74.30±9.74	71.22±10.44	69.20±10.75	ns	67.32±10.76	69.59±11.19	73.10±9.52	<0.05
Heart rate (beats/min)	84.43±19.92	77.19±16.06	72.86±12.26	<0.001	71.37±9.54	73.14±13.91	78.26±17.36	<0.001
ALT (U/L)	49.61 (18.85–37.50)	53.25 (24.00–75.20)	29.94 (18.00–36.00)	ns	27.41 (17.25–32.00)	29.27 (19.00–34.00)	42.72 (21.00–46.00)	<0.05
AST (U/L)	113.48 (23.45–83.25)	165.91 (37.00–293.60)	54.24 (21.00–50.50)	ns	37.54 (20.50–27.50)	76.35 (21.50–71.75)	89.42 (24.00–113.00)	<0.05
BUN (mmol/L)	11.19 (5.46–15.58)	5.20 (4.10–6.45)	5.44 (4.15–6.71)	ns	4.98 (3.83–6.21)	5.75 (4.43–6.79)	6.03 (4.44–6.79)	ns
Creatine (µmol/L)	167.71 (81.70–163.50)	82.57 (63.00–97.00)	82.75 (69.90–92.75)	ns	77.80 (67.00–89.20)	87.89 (73.75–94.95)	96.12 (75.60–101.00)	<0.001
UA (µmol/L)	411.47 (337.75–397.03)	322.89 (264.48–403.97)	323.92 (258.64–386.10)	ns	305.34 (251.27–367.80)	326.44 (245.53–403.67)	329.70 (272.30–369.20)	ns
Serum lipid profile
TG (mmol/L)	1.37±0.71	1.78±1.29	1.78±1.17	ns	1.90±1.34	1.81±1.18	1.45±0.81	<0.05
TC (mmol/L)	3.80±1.14	4.02±0.87	3.93±0.92	ns	3.97±0.87	3.91±0.92	3.97±1.01	ns
HDL-C (mmol/L)	0.90±0.27	0.93±0.22	0.95±0.54	ns	0.96±0.23	0.93±0.25	0.95±0.24	ns
LDL-C (mmol/L)	2.73±3.26	2.43±0.71	2.35±0.77	ns	2.36±0.70	2.35±0.78	2.40±0.85	ns
non-HDL (mmol/L)	3.75±5.76	4.37±11.37	2.94±0.98	ns	3.00±0.79	2.96±0.88	3.02±0.94	ns
LDL/HDL	2.82±1.76	2.72±0.98	2.56±1.02	ns	2.52±0.79	2.62±1.01	2.63±1.07	ns
ApoA-I (g/L)	1.02±0.24	1.04±0.17	1.11±0.19	<0.05	1.11±0.18	1.10±0.19	1.04±0.20	<0.05
HDL/ApoA-I (mmol/g)	0.88±0.13	0.89±0.13	0.85±0.15	ns	0.86±0.19	0.84±0.12	0.90±0.12	<0.05

Data are mean±SD and number (%).

ALT, alanine aminotransferase; ApoA-I, apolipoprotein A-I; ApoA-I, apolipoprotein A-I; AST, aspartate aminotransferase; BUN, blood urea nitrogen; HDL, high-density lipoprotein; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; ns, not significant; NT-proBNP, N-terminal pro-B-type natriuretic peptide; TC, total cholesterol; TG, triglyceride; UA, uric acid.

As shown in table 1, the patients were grouped according to the baseline NT-proBNP level, which was transformed by natural logarithm (<6.82, 6.82–9.06, and >9.06). We found that the patients with higher levels of log NT-proBNP also had significantly higher levels of weight, heart rate, ALT, AST, creatine and the HDL-C:LDL- C ratio but lower TG and ApoA-I level in the plasma.

Correlation between lipid and lipoprotein profiles and cardiac function in patients with AMI in the first group

As shown in table 2, Pearson correlation analysis also showed a significantly positive correlation between ApoA-I level and LVEF (r=0.165, p<0.001) but a significantly negative correlation between plasma ApoA-I level and log NT-proBNP (r=−0.181, p<0.001). Interestingly, the HDL-C:ApoA-I ratio was positively correlated with the log NT-proBNP level (r=0.14, p<0.05), and the TG level was negatively correlated with the log NT-proBNP level (r=−0.171, p<0.05).

Table 2

Correlation between serum lipid profile and cardiac function in patients with acute myocardial infarction in the first group

Characteristics	LVEF			Log NT-proBNP
	Pearson correlation	Sig. (two-tailed)	N	Pearson correlation	Sig. (two-tailed)	N
TG (mmol/L)	0.088	ns	471	−0.171	<0.05	402
TC (mmol/L)	−0.023	ns	469	−0.02	ns	399
HDL-C (mmol/L)	0.081	ns	470	−0.037	ns	401
LDL-C (mmol/L)	−0.031	ns	470	0.008	ns	401
Non-HDL (mmol/L)	−0.051	ns	467	−0.009	ns	398
LDL/HDL	−0.07	ns	470	0.04	ns	401
ApoA-I (g/L)	0.165	<0.001	469	−0.181	<0.001	400
HDL-C/ApoA-I (mmol/g)	0.039	ns	471	0.14	<0.05	400

ApoA-I, apolipoprotein A-I; ApoA-I, apolipoprotein A-I; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; ns, not significant; NT-proBNP, N-terminal pro-B-type natriuretic peptide; TC, total cholesterol; TG, triglyceride.

Baseline characteristics of the second group

A total of 294 patients with a diagnosis of AMI after PCI were enrolled in the validation cohort. The patients were also grouped according to the baseline LVEF level (<40%, 40%–50%, and >50%) and the log NT-proBNP level (<7.51, 7.51–9.58, and >9.58). Compared with patients in the group with the highest level of LVEF, the patients in the group with lower levels of LVEF had a higher level of heart rate (80.13±17.96 vs 69.05±11.45, p<0.001) and lower levels of ApoA-I level (1.00±0.23 vs 1.08±0.18, p=0.175) (table 3). In addition, the patients with lower levels of LVEF had significantly higher levels of AST, BUN, creatine and UA. We found the same results in the different log NT-proBNP level group analysis. Interestingly, the patients with higher levels of log NT-proBNP also had significantly lower levels of ApoA-I but higher levels of HDL-C:ApoA-I ratio (table 3).

Table 3

Baseline characteristics of the second group

Characteristics	Grouped by LVEF (%)			P value	Grouped by log NT-proBNP (pg/mL)			P value
	<40 (n=23)	40–50 (n=58)	>50 (n=193)		<7.51 (n=95)	7.51–9.58 (n=91)	>9.58 (n=93)
Age (years)	59.52±12.14	62.48±11.86	59.93±10.18	ns	57.84±10.46	60.47±9.73	63.34±11.70	ns
Male, n (%)	22 (95.65)	49 (84.48)	141 (73.05)	ns	68 (71.57)	67 (73.62)	80 (86.02)	ns
Height (cm)	171.44±6.60	169.10±7.63	167.80±7.16	ns	168.43±7.67	168.33±7.09	168.40±7.22	ns
Weight (kg)	74.03±11.25	72.54±10.47	70.39±11.45	ns	71.37±11.27	71.36±12.23	70.06±10.70	ns
Heart rate (beats/min)	80.13±17.96	71.45±13.36	69.05±11.45	<0.001	68.49±11.90	68.67±10.81	74.62±14.60	<0.001
ALT (U/L)	38.71 (18.65–52.50)	43.99 (20.05–51.50)	36.52 (20.05–51.50)	ns	36.35 (17.00–38.80)	37.71 (17.94–50.30)	39.14 (16.90–52.50)	ns
AST (U/L)	77.09 (18.30–61.62)	90.32 (20.07–123.05)	48.69 (18.57–57.28)	<0.05	41.65 (17.34–36.61)	54.35 (20.54–62.33)	77.36 (19.89–103.46)	<0.001
BUN (mmol/L)	5.42 (4.30–6.37)	5.43 (4.45–6.05)	4.64 (3.69–5.34)	<0.001	4.69 (3.90–5.36)	4.83 (3.80–5.49)	5.24 (4.05–5.93)	<0.05
Creatine (µmol/L)	76.68 (63.17–81.05)	78.29 (64.05–85.73)	66.43 (57.16–74.18)	<0.05	64.80 (57.94–71.36)	67.49 (56.92–76.32)	76.81 (62.00–82.07)	<0.05
UA (µmol/L)	351.34 (277.55–414.69)	318.94 (261.01–380.00)	311.31 (255.64–349.03)	<0.05	315.48 (252.16–352.48)	306.18 (248.97–348.78)	325.11 (266.83–375.34)	ns
Serum lipid profile
TG (mmol/L)	1.60±1.45	1.46±0.87	1.66±0.89	ns	1.77±0.90	1.74±1.16	1.36±0.69	ns
TC (mmol/L)	4.02±1.04	3.51±0.89	3.82±0.80	ns	3.77±0.81	3.71±0.80	3.88±1.00	ns
HDL-C (mmol/L)	0.88±0.26	0.91±0.24	0.92±0.21	ns	0.93±0.24	0.92±0.20	0.88±0.23	ns
LDL-C (mmol/L)	2.23±0.95	2.02±0.73	2.30±0.81	ns	2.25±0.77	2.30±0.84	2.13±0.84	ns
Non-HDL (mmol/L)	3.10±0.98	2.63±0.82	2.90±0.76	ns	2.82±0.79	2.80±0.76	2.97±0.93	ns
LDL/HDL	2.60±0.90	2.31±0.89	2.60±1.02	ns	2.56±1.07	2.55±0.89	2.51±1.03	ns
ApoA-I (g/L)	1.00±0.23	1.06±0.20	1.08±0.18	ns	1.11±0.20	1.08±0.16	1.00±0.19	<0.001
HDL/ApoA-I (mmol/g)	0.87±0.12	0.86±0.11	0.84±0.10	ns	0.83±0.96	0.84±0.10	0.87±0.11	<0.05

Data are mean±SD and number (%).

ALT, alanine aminotransferase; ApoA-I, apolipoprotein A-I; ApoA-I, apolipoprotein A-I; AST, aspartate aminotransferase; BUN, blood urea nitrogen; HDL, high-density lipoprotein; HDL-C, high-density lipoprotein cholesterol; LDL, low-density lipoprotein; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; ns, not significant; NT-proBNP, N-terminal pro-B-type natriuretic peptide; TC, total cholesterol; TG, triglyceride; UA, uric acid.

Correlation between plasma lipid profile and cardiac function in patients with AMI in the second group

As shown in table 4, Pearson correlation analysis also showed a significantly positive correlation between ApoA-I level and LVEF (r=0.165, p<0.05) but a significantly negative correlation between ApoA-I level and log NT-proBNP (r=−0.23, p<0.001). Interestingly, the HDL-C:ApoA-I ratio was positively correlated to the log NT-proBNP level (r=0.14, p<0.05), and the TG level was negatively correlated to the log NT-proBNP level (r=−0.175, p<0.05).

Table 4

Correlation between serum lipid profile and cardiac function in patients with acute myocardial infarction in the second group

Characteristics	LVEF			Log NT-proBNP
	Pearson correlation	Sig. (two-tailed)	N	Pearson correlation	Sig. (two-tailed)	N
TG (mmol/L)	0.035	ns	271	−0.175	<0.001	277
TC (mmol/L)	0.093	ns	169	−0.01	ns	165
HDL-C (mmol/L)	0.111	ns	271	−0.085	ns	277
LDL-C (mmol/L)	0.100	ns	271	−0.08	ns	277
Non-HDL (mmol/L)	0.076	ns	169	0.01	ns	165
LDL/HDL	0.036	ns	271	−0.03	ns	277
ApoA-I (g/L)	0.165	<0.05	271	−0.23	<0.001	277
HDL-C/ApoA-I (mmol/g)	−0.028	ns	271	0.181	<0.05	277

ApoA-I, apolipoprotein A-I; ApoA-I, apolipoprotein A-I; HDL, high-density lipoprotein; HDL-C, high-density lipoprotein cholesterol; LDL, low-density lipoprotein; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; ns, not significant; NT-proBNP, N-terminal pro-B-type natriuretic peptide; TC, total cholesterol; TG, triglyceride.

ApoA-I level as an independent predictor of MACE occurrence

During the median of 28.57 months of follow-up period, 76 (25.90%) patients experienced MACEs. Kaplan-Meier curves were used to illustrate the survival free from adverse events in different ApoA-I level groups in patients with AMI undergoing PCI, as shown in figure 1. Overall, patients with lower ApoA-I levels had a significantly worse outcome of survival free from MACEs during the follow-up period. Kaplan-Meier survival analysis demonstrated that lower admission ApoA-I level was significantly associated with MACE occurrence (p<0.001, log-rank test).

Figure 1

Kaplan-Meier analysis of MACEs based on the ApoA-I levels. The 294 patients were divided by tertiles of the ApoA-I levels: 0.99, 0.99–1.14, and >1.14 g/L. Risk of a MACE increased with decreasing tertile of the ApoA-I levels (log-rank test 33.354, p<0.001). ApoA-I, apolipoprotein A-I; MACE, major adverse cardiovascular event.

We then used Cox regression model for further analysis as shown in table 5. In univariate Cox analysis, we found that the lower ApoA-I level was significantly associated with an increased risk of MACEs in patients with AMI undergoing PCI (HR 2.294, 95% CI 1.239 to 4.248; p=0.008) over a median of 2.4 years of follow-up. This relationship remained significant in multivariate Cox analysis (HR 3.411, 95% CI 1.373 to 8.665; p=0.008) after adjustment for age, sex, height, weight, creatinine, LVEF, and NT-proBNP.

Table 5

Univariate and multivariate Cox analysis for MACEs in the second group

Variable	Univariate analysis			Multivariate analysis
	HR	95% CI	P value	HR	95% CI	P value
ApoA-I levels
High (>1.14 g/L)	Reference			Reference
Middle (0.99–1.14 g/L)	1.484	0.713 to 3.091	0.291	1.734	0.649 to 4.631	0.272
Low (<0.99 g/L)	2.294	1.239 to 4.248	0.008	3.411	1.373 to 8.665	0.008

Adjusted for age, sex, height, weight, creatinine, left ventricular ejection fraction, and N-terminal pro-B-type natriuretic peptide.

ApoA-I, apolipoprotein A-I; 95 % CI, 95 % confidence interval; HR, hazard ratio; MACE, major adverse cardiovascular event.

Discussion

CVD, especially AMI, is still the leading cause of death in China and worldwide, and its morbidity and mortality have continued to increase in recent years.1 2 Despite advances in therapeutic strategies for AMI, patients remain at a high risk of MACEs, particularly in the immediate weeks to months after the event.21 Dyslipoproteinemia is common in patients with AMI and usually predicts recurrent cardiovascular events.22 23 In the present study, we assessed the relationship between circulating lipid and lipoprotein profiles to the cardiac function and cardiology outcomes in patients with AMI undergoing PCI. Our results showed that ApoA-I levels were significantly reduced, and the HDL-C:ApoA-I ratios were increased in the patients with lower LVEF or higher NT-proBNP levels; there were positive correlations between cardiac function and ApoA-I, and negative correlations between cardiac function and the HDL-C:ApoA-I ratio. The ApoA-I levels exhibited a significant effect on predicting the incidence of MACEs.

The major novelty is that in the present study, we have demonstrated the utility of ApoA-I for predicting future adverse cardiovascular events in patients with AMI undergoing PCI from two clinical groups. Findings from the Incremental Decrease in End Points Through Aggressive Lipid Lowering (IDEAL) study concur that levels of ApoA-I were higher compared with their CAD population without heart failure among patients with new-onset heart failure.24 A large cohort of Controlled Rosuvastatin Multinational Trial in Heart Failure (CORONA) study proved that higher baseline HDL and ApoA-I were associated with a better prognosis; particularly, ApoA-I was more predictive than LDL or HDL.25 Moreover, a cross-sectional study of 199 patients with stable CAD also found that ApoA-I levels increased with increasing NYHA class.24 In our study, we have identified the correlation between ApoA-I and the cardiac function of patients with AMI after PCI. The cardiac function is strongly associated with the changes in lipid profile, with positive correlations between cardiac function and ApoA-I, and negative correlations between cardiac function and the HDL-C:ApoA-I ratio.

Furthermore, ApoA-I is an independent indicator to predict the incidence of MACEs. It is identified that ApoA-I is a key functional apolipoprotein component of HDL particles and plays a central role in cholesterol efflux, and has also been of interest for predicting CVD risks.26 27 In agreement with our results, the Apo-I Event Reducing in Ischemic Syndromes-I (AEGIS-I) trial, a phase IIb trial of patients with recent MI, found that a reconstituted ApoA-I (CSL112) was developed to enhance cholesterol efflux capacity. Notably, CSL112 was safe and confirmed its potential to remove cholesterol from atherosclerotic plaques.28–30 A meta-analysis proved that incident CVD events occurred more frequently in those subjects with lower ApoA-I, and ApoA-I had the strongest (inverse) associations with risk of fatal CVD.31 32 A case–control study found that ApoA-I was inversely related to mortality: for each 1 SD increase of ApoA-I, 31% and 33% decreases in all-cause and cardiovascular mortality were recorded.33 So far, ApoA-I has been little used in epidemiological studies. Furthermore, ApoA-I measurement is much less influenced than HDL-C by intravascular enzymes and lipid transfer proteins, which participate in HDL remodeling. Thus, ApoA-I measurement may improve assessment of cardiovascular risk.34

In this study, we evaluated the circulating levels of TG, TC, HDL, LDL, non-HDL, ApoA-I, etc. It is interesting that only ApoA-I showed its correlation to cardiac function. Theoretically, there are several reasons. First, HDL protects against atherosclerosis through multiple mechanisms, including amelioration of endothelial dysfunction, removal of excess cholesterol from macrophages, as well as antioxidative, anti-inflammatory, and antiapoptotic effects. ApoA-I is the primary functional apolipoprotein component of HDL, which plays pivotal roles in the reverse cholesterol transport pathways by modulating HDL-C formation, stabilization, binding to the hepatic scavenger receptors, and activating lecithin cholesterol acyl transferase. Therefore, the oxidation of particular residues on ApoA-I creates a dysfunctional HDL particle that is associated with an increased incidence of cardiovascular events.9 10 35 36 Our data provide evidence that ApoA-I could be introduced into clinical practice for assessing the cardiac function of patients with AMI undergoing PCI and for predicting the incidence of MACEs.

Limitations

The present study had several limitations. First, this was a single-center study restricted to Chinese patients with AMI after PCI. As mentioned previously, caution should be exercised when generalizing our findings to other ethnic groups, and further studies involving different ethnic groups are needed to support our findings. Moreover, we did not collect any data whether the patients had taken any medication (particularly lipid-lowering medication) during the 2.4 years of follow-up and, if they were (which is most likely), which medication and in what doses.

Conclusion

In summary, our results demonstrate that ApoA-I levels were significantly reduced, and the HDL-C:ApoA-I ratios were increased in the patients with lower LVEF or higher NT-proBNP level compared with the control. Pearson correlation analysis further showed positive correlations between cardiac function and ApoA-I and negative correlations between cardiac function and the HDL-C:ApoA-I ratio. In addition, the ApoA-I levels exhibited a significant effect on predicting the incidence of MACEs. Therefore, the ApoA-I level was positively associated with the cardiac function of patients with AMI after PCI, and ApoA-I is an independent indicator to predict the incidence of MACEs.