Proprotein convertase subtilisin-kexin type 9 as a biomarker for the severity of coronary artery disease

Figures & data

Figure 1. Flow chart of the selection of the present study population including inclusion and exclusion criteria.

Table I. Baseline characteristics of the study population. Bold values in the Table indicate statistical significance.

Variables	CAD	Non-CAD	P value
Demographics parameters
Age (years)	57.82 ± 9.72	52.29 ± 11.50	< 0.001
Gender, men % (n)	71.6 (395)	47.4 (227)	< 0.001
Coronary risk factors
BMI (kg/m^2)	25.77 ± 3.41	24.86 ± 3.41	< 0.001
Hypertension, % (n)	65.6 (362)	43.2 (207)	< 0.001
Diabetes mellitus, % (n)	26.8 (148)	11.9 (57)	< 0.001
Hypercholesterolemia, % (n)	33.9 (187)	33.8 (162)	0.947
Hypertriglyceridemia, % (n)	46.4 (256)	39.0 (187)	0.031
Smoking, % (n)	47.3 (261)	26.5 (127)	< 0.001
Family history of CAD, % (n)	14.7 (81)	7.31 (35)	< 0.001
Laboratory parameters
Triglyceride (mmol/L)	1.65 (1.19–2.35)	1.49 (1.03–2.10)	< 0.001
Total cholesterol (mmol/L)	4.92 ± 1.03	4.94 ± 1.06	0.771
Non-HDL cholesterol (mmol/L)	3.84 ± 1.00	3.75 ± 1.02	0.171
LDL cholesterol (mmol/L)	3.33 ± 0.97	3.23 ± 1.01	0.099
Apolipoprotein B (g/L)	1.04 ± 0.28	0.99 ± 0.28	0.016
WBC count (× 10^9/L)	6.29 ± 1.77	6.19 ± 1.61	0.394
Hs-CRP (mg/L)	1.53 (0.70–2.91)	1.03 (0.44–2.22)	< 0.001
Fibrinogen (g/L)	3.17 ± 0.80	2.99 ± 0.62	0.002
PCSK9 (ng/mL)	224.75 (183.90–274.72)	224.64 (184.54–281.09)	0.928
Prior drug treatment
Statins	0 (0)	0 (0)
Aspirin, % (n)	45.6 (252)	24.6 (118)	< 0.001
Beta-blocker, % (n)	23.7 (131)	17.1 (82)	0.027
ACEI/ARB, % (n)	18.1 (100)	11.3 (54)	< 0.001
CCB, % (n)	21.9 (121)	15.2 (73)	0.011

Data shown are mean ± SD, median (Q1–Q3 quartiles), or n (%).

ACEI/ARB = angiotensin-converting enzyme inhibitors/angiotensin receptor blocker; BMI = body mass index; CAD = coronary artery disease; CCB = calcium channel blocker; HDL-C = high-density lipoprotein cholesterol; hs-CRP = high-sensitivity C-reactive protein; LDL-C = low-density lipoprotein cholesterol; PCSK9 = proprotein convertase subtilisin/kexin type 9; TC = total cholesterol; TG = triglyceride; WBC = white blood cell count.

Table II. Adjusted associations of PCSK9, atherogenic lipids, and inflammation with CAD. Bold values in the Table indicate statistical significance.

Variables		CAD (n = 552)	Non-CAD (n = 479)	P value	Low GS (n = 183)	Intermediate GS (n = 184)	High GS (n = 185)	P for trend
TG (mmol/L) ^a	Model 1	1.69 ± 1.02	1.54 ± 1.03	0.007	1.59 ± 1.04	1.71 ± 1.04	1.81 ± 1.04	0.032
	Model 2	1.66 ± 1.02	1.58 ± 1.03	0.085	1.58 ± 1.04	1.71 ± 1.04	1.82 ± 1.03	0.032
	Model 3	1.65 ± 1.02	1.58 ± 1.02	0.103	1.57 ± 1.04	1.72 ± 1.04	1.82 ± 1.03	0.017
TC(mmol/L)	Model 1	4.98 ± 0.05	4.86 ± 0.05	0.068	4.89 ± 0.08	4.98 ± 0.08	5.05 ± 0.07	0.006
	Model 2	4.97 ± 0.05	4.89 ± 0.05	0.121	4.80 ± 0.08	4.96 ± 0.08	5.05 ± 0.08	0.004
	Model 3	4.97 ± 0.05	4.89 ± 0.05	0.126	4.80 ± 0.08	4.96 ± 0.08	5.05 ± 0.08	0.004
Non-HDL-C (mmol/L)	Model 1	3.87 ± 0.04	3.71 ± 0.05	0.021	3.72 ± 0.08	3.90 ± 0.08	3.96 ± 0.06	0.003
	Model 2	3.86 ± 0.04	3.73 ± 0.05	0.034	3.67 ± 0.07	3.89 ± 0.07	3.96 ± 0.06	0.002
	Model 3	3.85 ± 0.04	3.73 ± 0.05	0.047	3.67 ± 0.07	3.89 ± 0.07	3.96 ± 0.07	0.002
LDL-C (mmol/L)	Model 1	3.34 ± 0.04	3.19 ± 0.05	0.019	3.18 ± 0.08	3.37 ± 0.08	3.49 ± 0.06	< 0.001
	Model 2	3.36 ± 0.04	3.20 ± 0.05	0.018	3.21 ± 0.07	3.37 ± 0.08	3.50 ± 0.06	< 0.001
	Model 3	3.36 ± 0.04	3.21 ± 0.05	0.029	3.21 ± 0.07	3.37 ± 0.08	3.50 ± 0.06	< 0.001
ApoB (g/L)	Model 1	1.04 ± 0.01	0.985 ± 0.01	0.002	1.01 ± 0.02	1.04 ± 0.02	1.07 ± 0.02	0.036
	Model 2	1.04 ± 0.01	0.991 ± 0.01	0.012	1.01 ± 0.02	1.04 ± 0.02	1.06 ± 0.02	0.047
	Model 3	1.04 ± 0.01	0.991 ± 0.01	0.013	1.01 ± 0.02	1.04 ± 0.02	1.06 ± 0.02	0.047
WBC count (× 10^9/L)	Model 1	6.28 ± 0.07	6.24 ± 0.08	0.221	5.91 ± 0.14	6.29 ± 0.14	6.55 ± 0.11	0.002
	Model 2	6.29 ± 0.08	6.22 ± 0.07	0.194	5.92 ± 0.13	6.37 ± 0.13	6.60 ± 0.13	0.001
	Model 3	6.30 ± 0.08	6.22 ± 0.07	0.097	5.89 ± 0.13	6.39 ± 0.13	6.61 ± 0.13	< 0.001
Hs-CRP (mg/L) ^a	Model 1	1.32 ± 1.06	0.99 ± 1.06	0.001	1.12 ± 1.11	1.40 ± 1.11	1.61 ± 1.09	0.028
	Model 2	1.27 ± 1.06	1.03 ± 1.07	0.019	1.13 ± 1.11	1.41 ± 1.11	1.58 ± 1.09	0.037
	Model 3	1.27 ± 1.06	1.03 ± 1.07	0.025	1.14 ± 1.11	1.40 ± 1.11	1.58 ± 1.09	0.041
Fibrinogen (g/L)	Model 1	3.17 ± 0.03	3.00 ± 0.05	0.005	3.08 ± 0.07	3.21 ± 0.07	3.20 ± 0.05	0.096
	Model 2	3.17 ± 0.03	3.00 ± 0.05	0.007	3.06 ± 0.07	3.27 ± 0.06	3.19 ± 0.06	0.059
	Model 3	3.17 ± 0.03	3.00 ± 0.05	0.008	3.05 ± 0.06	3.28 ± 0.06	3.19 ± 0.06	0.040
PCSK9 (ng/mL) ^a	Model 1	228.56 ± 1.01	218.27 ± 1.02	0.035	212.81 ± 1.01	233.88 ± 1.01	229.99 ± 1.01	0.013
	Model 2	228.56 ± 1.01	218.77 ± 1.02	0.031	213.30 ± 1.01	233.35 ± 1.01	229.99 ± 1.01	0.019
	Model 3	228.03 ± 1.01	219.28 ± 1.02	0.046	213.30 ± 1.01	233.35 ± 1.01	229.99 ± 1.01	0.019

Values were obtained from general linear models. Model 1 adjusted for age and gender. Model 2 adjusted for variables in Model 1 + BMI, hypertension, and diabetes. Model 3 adjusted for variables in Model 2 + smoking and family history of CAD.

^aLog-transformed data.

ApoB = apolipoprotein B; CAD = coronary artery disease; GS = Gensini score; HDL-C = high-density lipoprotein cholesterol; hs-CRP = high-sensitivity C-reactive protein; LDL-C = low-density lipoprotein cholesterol; PCSK9 = proprotein convertase subtilisin/kexin type 9; TC = total cholesterol; TG = triglyceride; WBC = white blood cell count.

Figure 2. Adjusted odd ratios for incident CAD (A) and severity of CAD (high GS versus low GS) (B) according to the individual markers of lipid and inflammation, PCSK9 levels (log-transformed data), and the multi-marker indexes. Logistic regression analysis was performed, and models were adjusted for age and gender.

Figure 3. Correlations of lipid and inflammation markers with plasma PCSK9 levels in overall study population (A–H). Regression lines were calculated by Pearson correlation analysis. Coefficients and P value are presented.

Table III. Relationships of PCSK9 with multi-marker indexes of atherogenic lipids and inflammation. Bold values in the Table indicate statistical significance.

Models	Levels of PCSK9 (ng/mL)^a				P for trend
	Multi-marker index (quartiles)
Atherogenic lipids	Q1	Q2	Q3	Q4
Crude	209.80 ± 1.39	212.52 ± 1.33	228.45 ± 1.33	249.40 ± 1.35	< 0.001
Model 1	211.35 ± 1.02	212.81 ± 1.02	228.03 ± 1.02	247.74 ± 1.02	< 0.001
Model 2	211.35 ± 1.02	212.32 ± 1.02	228.03 ± 1.02	247.17 ± 1.02	< 0.001
Model 3	207.97 ± 1.02	215.28 ± 1.02	220.80 ± 1.02	240.99 ± 1.02	< 0.001
Inflammation	Q1	Q2	Q3	Q4
Crude	201.72 ± 1.36	224.13 ± 1.36	211.40 ± 1.36	240.27 ± 1.33	< 0.001
Model 1	211.84 ± 1.02	223.36 ± 1.02	211.84 ± 1.02	240.99 ± 1.02	< 0.001
Model 2	211.84 ± 1.02	223.36 ± 1.02	211.84 ± 1.02	242.66 ± 1.02	< 0.001
Model 3	215.28 ± 1.02	220.29 ± 1.02	220.80 ± 1.02	229.88 ± 1.02	< 0.001

Values were obtained from general linear models. Crude model was unadjusted. Model 1 adjusted for age and gender. Model 2 adjusted for variables in Model 1 + BMI, hypertension, diabetes, smoking, and family history of CAD. Model 3 adjusted for variables in Model 2 + multi-marker index.

^aLog-transformed data.

Table IV. Mediator analysis corroborating the association of PCSK9 with CAD. Bold values in the Table indicate statistical significance.

Mediator	Parameters	Effect size
		Incidence of CAD	Severity of CAD
Multi-marker of atherogenic lipids	Total effect of PCSK9	0.160; P < 0.05	0.214; P < 0.05
	Effect not mediated by lipids	0.133; P < 0.05	0.178; P < 0.05
	Effect mediated by lipids	0.027; P < 0.05	0.036; P < 0.05
	Proportion of PCSK9 effect mediated by lipids	0.17	0.17
Multi-marker of inflammation	Total effect of PCSK9	0.160; P < 0.05	0.214; P < 0.05
	Effect not mediated by inflammation	0.144; P < 0.05	0.176; P < 0.05
	Effect mediated by inflammation	0.016; P < 0.05	0.038; P < 0.05
	Proportion of PCSK9 effect mediated by inflammation	0.10	0.18

Mediation analysis with lipids and inflammation as mediators for the associations of PCSK9 levels with CAD was performed.

CAD = coronary artery disease; GS = Gensini score; PCSK9 = proprotein convertase subtilisin/kexin type 9.