If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Apolipoprotein B Displays Superior Predictive Value Than Other Lipids for Long-Term Prognosis in Coronary Atherosclerosis Patients and Particular Subpopulations: A Retrospective Study
Accumulating evidence that apolipoprotein B (apoB) plays a critical role in predicting coronary heart disease (CHD) and future outcomes. The 2019 European Society of Cardiology/European Atherosclerosis Society guidelines suggest that apoB can be an alternative to non–HDL-C or LDL-C in patients with high triglyceride levels, diabetes, obesity, metabolic syndrome, or very low LDL-C levels. This study explores whether apoB can also serve as predictive value for long-term major adverse cardiovascular events (MACEs) in normal people and specific coronary atherosclerosis patients.
Methods
A total of 826 patients were followed up over 10 years, and the risk factors for MACEs were retrospectively analyzed in patients with CHD and particular subpopulations. All statistical analyses were performed in R software. Cox regressions were performed to assess independent risk factors of long-term MACEs in the atherosclerosis group and CHD subgroups. Kaplan-Meier survival curves were used to evaluate the survival rate for patients in different apoB quartiles, and receiver-operating characteristic curves were used to compare apoB and other lipids in predicting the presence of long-term MACE.
Findings
apoB could be a “risk-enhancing factor” in patients with coronary atherosclerosis disease, whereas in the Normal population, LDL-C still acted as a major risk factor for predicting MACEs. apoB was a good risk predictor for long-term cardiovascular events in coronary atherosclerosis (AS) patients, including the AS group and CHD subpopulations (including CHD + triglyceride ≥2.3 mmol/L, CHD + diabetes mellitus, CHD + body mass index ≥25 kg/m2, or CHD + metabolic syndrome). In patients with CHD whose condition was complicated with diabetes, obesity, and metabolic syndrome, apoB performed better than other lipids in predicting the presence of myocardial infarction, hospitalization due to angina, and cardiac death. Despite achieving optimal LDL-C or non–HDL-C levels, patients with CHD are still at risk of worse survival if they are unable to reach a low apoB level (lower cut points such as 65 mg/dL).
Apolipoprotein B (apoB) has been reported to play a critical role in predicting coronary heart disease (CHD) and future outcomes, performing even better than LDL-C and non–HDL-C.
In physiological state, each VLDL, intermediate-density lipoprotein, and LDL remnant and lipoprotein a (Lp[a]) particles contain one molecule of apoB. Hence, apoB represents the sum of the atherogenic particle in the plasma,
In some pathologic situations, there may be normal levels of LDL-C but high apoB, which reflects increased amounts of small and dense LDL in the blood. When apoB disaccords with non–HDL-C or LDL-C (cholesterol-enriched or cholesterol-depleted apoB particles), overestimation or underestimation of the atherogenic risk should be cautioned, and it would be unwise to only focus on LDL-C or non–HDL-C levels.
Secondary prevention studies also reported that the baseline apoB level is significantly better at predicting the recurrence of cardiovascular events, especially those associated with recurrent coronary events in patients with metabolic syndrome.
AACC Lipoproteins and Vascular Diseases Division Working Group on Best Practices. Apolipoprotein B and Cardiovascular Disease Risk: Position Statement from the AACC Lipoproteins and Vascular Diseases Division Working Group on Best Practices.
In patients with mild to moderate hypertriglyceridemia, elevated baseline apoB levels help identify ASCVD risk that is not reflected by non–HDL-C or LDL-C levels.
European Atherosclerosis Society Consensus Panel. The Polygenic Nature of Hypertriglyceridaemia: Implications for Definition, Diagnosis, and Management.
These studies reinforce the idea that more attention should be paid to apoB levels for specific populations at heightened risk of cardiovascular events.
The purpose of the present study was to identify the lipid factors that yield advantages over other lipid factors for long-term cardiovascular events in patients with coronary atherosclerosis (AS) and in particular subpopulations.
Patients and Methods
Study Population
This study enrolled 826 patients (520 men and 306 women; aged 33–94 years) admitted to the Department of Cardiology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, for coronary angiography from 2008 to 2012. Coronary angiography was performed by the chief or associate physicians. Patients with significant vascular disease, cardiomyopathies, myocarditis, pulmonary heart disease, ongoing systemic inflammatory diseases, systemic autoimmune disease, renal or hepatic dysfunction, malignancy, and stress disorders such as cerebral hemorrhage, cerebral infarction, and acute abdominal disease were excluded from the study.
To establish the overall baseline data, patients were divided into 3 groups according to the angiography findings. In general, 532 patients were classified as the CHD group (lesion stenosis ≥50% of least one major branch), 85 as the AS group (lesion stenosis <50% of least one major branch), and 209 as the Normal group (normal coronary arteries). For research purposes, we further divided the 532 patients with CHD into 5 groups according to other additional risk factors: 105 patients with triglyceride (TG) levels ≥2.3 mmol/L, 249 patients with diabetes, 197 obese patients with a body mass index (BMI) ≥25 kg/m2, 327 patients with metabolic syndrome according to the Adult Treatment Panel III criteria,
and 70 patients with LDL-C levels <1.8 mmol/L. Furthermore, all 826 patients and those classified into the 5 groups were respectively sectioned into 4 groups, referencing quartiles of apoB.
Baseline Investigation
To better present the general characteristics of patients, we summarized the general and detailed baseline characteristics, including sex, age, BMI, systolic blood pressure, diastolic blood pressure, heart rate, smoking, hypertension history, hypercholesterol history, diabetes history, Gensini score, and laboratory parameters such as serum creatinine, uric acid, fasting plasma glucose, hemoglobin A1c (HbA1c), international normalized ratio, total cholesterol (TC), TG, HDL-C, LDL-C, non–HDL-C, apoB, apolipoprotein A-I (apoA-I), Lp(a), C-reactive protein, creatine kinase isoenzyme (CK-MB), troponin I, and pro–brain natriuretic peptide.
Follow-up on End Point Events
Follow-up interviews through telephone calls were conducted to obtain information on out-of-hospital major adverse cardiovascular events (MACEs); median follow-up time was 11.06 (±0.96) years after discharge. Of the 826 patients, 106 were lost to follow-up (12.83%). We evaluated MACEs as combined end points including myocardial infarction, coronary revascularization, new-onset heart failure, hospitalization due to angina, presumed ischemic stroke, and cardiac death.
Statistical Analysis
Continuous parametric variables are presented as mean (SD). Parametric variables were compared between groups with the Student t test when data were normally distributed and homogeneous from variance; otherwise, the Wilcoxon rank-sum test was used. Categorical variables are presented in the form of number and proportion (no. [%]), and differences between groups were tested by using the χ 2 test. Receiver-operating characteristic (ROC) curves were used to analyze the diagnostic value of different lipid markers for long-term MACEs. Kaplan-Meier plots described the survival of different groups divided by quartile of apoB, and the differences between groups were calculated by using the log-rank test. Cox regression was used to analyze the effects of various factors on patient survival, and the results are shown in a forest plot. All statistical analyses were conducted in R software version 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria). The main R packages used included “dplyr,” “mice,” “epiDisplay,” “pROC,” “survival,” and “survminer.” Missing values were addressed through the multiple imputation method.
This retrospective study was approved by the College Committee for clinical trials of Shanghai Jiao Tong University (Shanghai, China), and the clinical trial number is “[2021] Re-view No. [314]”.
Results
Baseline Characteristics
A total of 826 participants were included in this cohort study. All follow-up telephone interviews were completed before October 2020, and the median follow-up time was 11.06 (±0.96) years. At enrollment time, patients were divided into 3 groups according to the angiography findings: 532 patients into the CHD group, 85 patients into the AS group, and 209 patients into the Normal group. Table I Table I demonstrates baseline characteristics of the CHD group and the AS group, compared to the Normal group. The quartiles of apoB levels are shown in Table II. As shown in Table III, patients with CHD and TG ≥2.3 mmol/L compared vs those with TG <2.3 mmol/L had higher levels of BMI, lipid profiles (ie, TG, apoB), and CK-MB; lower levels of HDL-C; more prevalent hypercholesterol history; and were younger (P < 0.05). There are more female subjects and smokers in the group CHD with diabetes than those CHD patients without diabetes. CHD patients with diabetes compared vs those without diabetes presented higher levels of BMI, systolic blood pressure, glucose, HbA1c, international normalized ratio, lipid profiles (ie, TG, LDL-C), C-reactive protein, CK-MB, and troponin I; lower levels of apoA-I; and more hypertension and history of hypercholesterol (P < 0.05). CHD patients with BMI ≥25 kg/m2 compared with those with BMI <25 kg/m2 possessed higher levels of uric acid, glucose, HbA1c, and lipid profiles (ie, TG, LDL-C, apoB); less prevalent diabetes history; and lower levels of apoA-I (P < 0.05). CHD patients with metabolic syndrome vs those without metabolic syndrome had higher levels of BMI, systolic blood pressure, uric acid, glucose, HbA1c, lipids (ie, TG, apoB), CK-MB, and troponin I; lower levels of HDL-C and apoA-I; and included a higher percentage of male subjects and smokers (P < 0.05). CHD patients with a low LDL-C level (<1.8 mmol/L) shared similar baseline characteristics with CHD patients whose LDL-C levels were ≥1.8 mmol/L.
Table Icharacteristics of all 826 patients divided into 3 groups according to angiography findings.
Table IIQuartile interval values of apoB in different groups.
Group
Q1
Q2
Q3
Q4
All 826 patients
(0.45, 0.76]
(0.76, 0.91]
(0.91, 1.06]
(1.06, 3.4]
Normal group
(0.56, 0.72]
(0.72, 0.89]
(0.89, 0.98]
(0.98, 3.4]
AS group
(0.67, 0.89]
(0.89, 1.00]
(1.00, 1.09]
(1.09, 2.45]
CHD with triglycerides ≥2.3 mmol/L
(0.57, 0.89]
(0.89, 1.03]
(1.03, 1.20]
(1.20, 1.71]
CHD with diabetes
(0.45, 0.74]
(0.74, 0.89]
(0.89, 1.09]
(1.09, 2.24]
CHD with BMI ≥25 kg/m2
(0.46, 0.78]
(0.78, 0.99]
(.099, 1.13]
(1.13, 2.24]
CHD with metabolic syndrome
(0.45, 0.76]
(0.76, 0.94]
(0.94, 1.12]
(1.12, 2.24]
CHD with LDL-C <1.8 mmol/L
(0.45, 0.56]
(0.56, 0.63]
(0.63, 0.72]
(0.72, 1.06]
AS = atherosclerosis; BMI = body mass index; CHD = coronary heart disease; Q1 = lower quartile of apoB; Q2 = second quartile of apoB; Q3 = third quartile of apoB; Q4 = upper quartile of apoB.
Risk of Long-term MACEs Significantly Correlated With apoB Level in Patients With Coronary AS Disease
Significant predictive factors of MACE in the Cox regression analysis are shown in forest plots. The data have been adjusted for sex, age, smoking, hypertension, diabetes, and other lipids such as TC, TG, LDL-C, Lp(a), and HDL-C. Except for the Normal group (Figure 1B), apoB remained significantly associated with risks of MACE: hazard ratio [HR] = 2.43; 95% CI, 1.40–4.21; P = 0.002) for all 826 patients (Figure 1A), HR = 5.07(95% CI, 1.14–22.5; P = 0.033) for the AS group (Figure 1C), HR = 2.58 (95% CI, 0.833–8.0; P = 0.1) for CHD patients with TG ≥2.3 mmol/L (Figure 1D), HR = 4.26 (95% CI, 1.40–13.0; P = 0.011) for CHD patients with diabetes (Figure 1E), HR = 3.65 (95% CI, 1.27–10.5; P = 0.017) for CHD patients with BMI ≥25 kg/m2 (Figure 1F), and HR = 3.89 (95% CI, 1.53–9.9; P = 0.004) for CHD patients with metabolic syndrome (Figure 1G). In CHD patients with LDL-C <1.8 mmol/L, apoB was not significantly associated with risk of MACE (Figure 1H).
Figure 1Forest plots of Cox regression analysis for identifying significant predictive factors of long-term major adverse cardiovascular events. (A) Group of all 826 patients. (B) Normal group. (C) Atherosclerosis (AS) group. (D) Coronary heart disease (CHD) + triglyceride (TG) ≥2.3 mmol/L group. (E) CHD + diabetes group. (F) CHD + body mass index (BMI) ≥25 kg/m2 group; (G) CHD + metabolic syndrome group. (H) CHD + LDL-C <1.8 mmol/L group. AIC = Akaike information criterion; apoB = apolipoprotein B; Lp(a) = lipoprotein a; TC = total cholesterol.Compared with reference: * p<0.05; ** p<0.01; *** p<0.001.
According to Cox regression results, we can predict that apoB was an independent risk factor for long-term MACEs in the patients with coronary AS, including the AS group and CHD subpopulations (CHD patients with diabetes, BMI ≥25 kg/m2, or metabolic syndrome). In the Normal group, the risk of MACEs was mainly attributed to LDL-C. In the CHD + TG ≥2.3 mmol/L and CHD + LDL-C <1.8 mmol/L groups, OR values corresponding to apoB were >1, indicating that apoB is a risk factor, but the P values were not statistically significant.
Higher apoB Levels Translate to Lower Survival Rate in the Populations With Coronary Artery Disease
In all 826 patients (Figure 2A), Kaplan-Meier survival curves showed worse survival for the patients with apoB in the upper quartile (apoB = (1.06, 3.4]; log-rank test, P < 0.0001), event-free survival proportion was 77.83%. In the Normal group (Figure 2B), although the survival rate of the fourth quartile apoB was not the lowest, the survival rate of the AS group strictly followed apoB levels inversely (log-rank test, P = 0.003) (Figure 2C). In CHD patients with TG ≥2.3 mmol/L, diabetes, BMI ≥25 kg/m2, and metabolic syndrome (Figure 2D-2G), Kaplan-Meier curves indicated worse survival for the patients with apoB in the fourth quartile (log-rank test, P = 0.02, P = 0.00018, P = 0.00091, and P < 0.0001, respectively), and event-free survival proportions were 60.87%, 58.82%, 55.00%, and 53.73%. In CHD patients with LDL <1.8 mmol/L (Figure 2H), the survival of the upper quartile group was not the worst among the quartile groups, and the P value of the log-rank test was >0.05, indicating no significant differences between groups. This could be due to the low number of people in this group (n = 4) experiencing MACE, as the number of CHD patients with LDL-C <1.8 mmol/L was too small to be statistically significant. in each group.
Figure 2Event-free survival curve for long-term major adverse cardiovascular events in the group of all 826 patients (A), the Normal group (B), the atherosclerosis (AS) group (C), the coronary heart disease (CHD) + triglyceride (TG) ≥2.3 mmol/L group (D), the CHD + diabetes group (E), the CHD + body mass index (BMI) ≥25 kg/m2 group (F), the CHD + metabolic syndrome group (G), and the CHD + LDL-C <1.8 mmol/L group (H) according to the quartile of apolipoprotein B (apoB) as described in Table III. Median follow-up time is 11.06 (0.96) years. Purple line represents top quartile of apoB group, and the log-rank test was used for the differences between groups.
In the logistic regression analysis, we set whether MACEs have occurred as the dependent variable. Quartiles of apoB were set as independent variables, with the first quartile of apoB defined as the reference group, adjusted for sex, age, smoking, hypertension, hypercholesterol, and diabetes. As shown in Table IV, ORs were calculated for MACEs in all 826 patients, the AS group, and CHD + diabetes, CHD + BMI ≥25 kg/m2, and CHD + metabolic syndrome groups compared with the first quartile of apoB. The adjusted OR for MACEs of patients in the fourth quartile of apoB were 12.4, 20.03, 3.31, 1.9, and 4.24, respectively (P < 0.001, P = 0.009, P = 0.005, P = 0.007, and P = 0.008). The ORs of apoB for prediction of MACEs in the CHD + TG ≥2.3 mmol/L group was 3.31 but was without significance (P = 0.101). Compared with patients in the first quartile of the Normal and CHD + LDL-C <1.8 mmol/L groups, patients in the other 3 groups were not at higher risk (P > 0.05). These results indicate that in the all patients group, the AS group, and the CHD with TG ≥2.3 mmol/L, diabetes, BMI ≥25 kg/m2, and metabolic syndrome groups, those in the fourth quartile of apoB experienced a higher incidence of MACEs that was not observed in the Normal group and low LDL-C group.
Table IVLogistic regression analysis for the association between the apolipoprotein B (apoB) quartile and long-term major adverse cardiovascular events. Each group was compared using the lower quartile group of apoB as a reference.
All 826 Patients
CHD + TG ≥GD mmol/L
Crude OR (95% CI)
Adjusted OR (95% CI)
P (Wald's Test)
Crude OR (95% CI)
Adjusted OR (95% CI)
P (Wald's Test)
Group apoB ref. = Q1
Q2
1.58 (0.6–4.15)
1.54 (0.58–4.08)
0.383
0.63 (0.09–4.23)
0.62 (0.08–4.13)
0.636
Q3
3.21 (1.32–7.82)
3.29 (1.34–8.1)
0.01
1.64 (0.36–7.48)
1.61 (0.34–7.45)
0.525
Q4
19.78 (8.86–44.18)
19.59 (8.66–44.35)
< 0.001
3.3 (0.79–13.74)
3.31 (0.76–13.71)
0.101
CHD + Diabetes
CHD + BMI ≥MI kg/m2
Crude OR (95% CI)
Adjusted OR (95% CI)
P (Wald's test)
Crude OR (95% CI)
Adjusted OR (95% CI)
P (Wald's test)
Group apoB ref. = Q1
Q2
1.11 (0.26–4.72)
1.47 (0.27–7.91)
0.653
1.96 (0.68–5.71)
1.84 (0.62–5.51)
0.273
Q3
0.61 (0.13–2.79)
0.66 (0.11–4.04)
0.651
3.39 (1.24–9.32)
3.12 (1.12–8.72)
0.03
Q4
1.76 (0.18–3.27)
1.9 (0.16–4.96)
0.005
4.49 (1.62–12.41)
4.24 (1.49–12.07)
0.007
CHD + Metabolic Syndrome
CHD + LDL <1.8 mmol/L
Crude OR (95% CI)
Adjusted OR (95% CI)
P (Wald's Test)
Crude OR (95% CI)
Adjusted OR (95% CI)
P (Wald's Test)
Group apoB ref. = Q1
Q2
1.7 (0.82–3.52)
1.69 (0.81–3.54)
0.161
1.11 (0.26–4.72)
1.47 (0.27–7.91)
0.653
Q3
1.7 (0.83–3.52)
1.67 (0.8–3.49)
0.171
0.61 (0.13–2.79)
0.66 (0.11–4.04)
0.651
Q4
2.74 (1.32–5.7)
2.73 (1.3–5.76)
0.008
0.76 (0.18–3.27)
0.9 (0.16–4.96)
0.905
BMI = body mass index; CHD = coronary heart disease; TG = triglyceride; Q1 = lower quartile of apoB; Q2 = second quartile of apoB; Q3 = third quartile of apoB; Q4 = upper quartile of apoB.
During the 11.06 (0.96) years of follow-up, we recorded 119 patients who experienced MACEs. The total number of MACEs was 342, including 90 myocardial infarction, 19 coronary revascularization, 44 new-onset heart failure, 114 hospitalizations due to angina, 13 presumed ischemic strokes, and 62 cardiac deaths. Quantitative numbers of MACEs occurring in different groups sorted according to apoB quartile are presented in Figure 3. Except for the Normal group and the CHD + LDL-C <1.8 group, risks of MACEs in the fourth quartile of apoB (yellow column) are much higher than that of the lower 3 quartiles.
Figure 3Histograms of concrete out-of-hospital major adverse cardiac events according to quartiles of apolipoprotein B (apoB) in the group of all 826 patients (A), the Normal group (B), the atherosclerosis (AS) group (C), the coronary heart disease (CHD) + triglyceride (TG) ≥2.3 mmol/L group (D), the CHD + diabetes group (E), the CHD + body mass index (BMI) ≥25 kg/m2 group (F), the CHD + metabolic syndrome group (G), and the CHD + LDL-C <1.8 mmol/L group (H) during the 11.06 (0.96) years of follow-up. Q = quartile.
Based on the quantitative results in Figure 3, we can roughly assert that high levels of apoB were most closely associated with hospitalization due to angina in all 826 patients, AS group, and the CHD patients. Logistic regression analysis was performed to further clarify the correlation between apoB and adverse events (Table V). In all patients, apoB indicated that new-onset heart failure was significantly more likely than other adverse events (OR, 21.85; 95% CI, 5.18–92.15; P < 0.001). ORs for prediction of myocardial infarction, coronary revascularization, hospitalization due to angina, and cardiac death were 2.65 (P < 0.001), 7.18 (P = 0.01), 3.87 (P < 0.001), and 11.58 (P < 0.001), respectively. In the Normal group, the OR for prediction of hospitalization due to angina was 18.49 (P = 0.01), but the ORs of the other groups were not statistically significant.
Table VLogistic regression analysis for the association between the apolipoprotein B (apoB) quartile and detailed major adverse cardiovascular events for all 826 patients, the Normal group, atherosclerosis (AS) group, and coronary heart disease (CHD) patients with triglycerides (TG) ≥2.3 mmol/L, diabetes, body mass index (BMI) ≥25 kg/m2, or metabolic syndrome.
All 826 patients
Myocardial Infarction
Coronary Revascularization
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
0.98 (0.38–2.51)
0.960
Q2
0.98 (0.14–7)
0.981
Q3
1.63 (0.68–3.90)
0.273
Q3
1.1 (0.15–7.88)
0.925
Q4
9.85 (4.74–20.46)
<0.001
Q4
7.15 (1.59–32.08)
0.010
New-Onset Heart Failure
Hospitalization Due to Angina
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
2.48 (0.48–12.91)
0.282
Q2
1.2 (0.49–2.97)
0.686
Q3
1.1 (0.15–7.88)
0.925
Q3
2.32 (1.02–5.30)
0.046
Q4
21.75 (5.16–91.71)
<0.001
Q4
13.96 (6.76–28.82)
<0.001
Presumed Ischemic Stroke
Cardiac Death
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
3956182.6 (0–infinity)
0.99
Q2
3.04 (0.97–9.59)
0.057
Q3
8948015.1 (0–infinity)
0.99
Q3
4.07 (1.32–12.58)
0.015
Q4
44048218.65 (0–infinity)
0.988
Q4
11.52 (4.03–32.97)
<0.001
CHD patients with TG ≥2.3 mmol/L
Myocardial Infarction
Coronary Revascularization
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
0.43 (0.02–7.63)
0.564
Q2
1 (0–infinity)
1
Q3
1.64 (0.16–16.73)
0.678
Q3
1 (0–infinity)
1
Q4
6.11 (0.71–52.25)
0.098
Q4
101730409.71 (0–infinity)
0.997
New Onset-Heart Failure
Hospitalization due to Angina
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
1 (0–infinity)
1
Q2
0.19 (0.02–2.4)
0.200
Q3
1 (0–infinity)
1
Q3
0.95 (0.15–5.94)
0.958
Q4
202390022.79 (0–infinity)
0.997
Q4
3.83 (0.73–19.99)
0.111
Presumed Ischemic Stroke
Cardiac Death
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
1 (0–infinity)
1
Q2
2025943.47 (0–infinity)
0.991
Q3
1 (0–infinity)
1
Q3
10129717.35 (0– infinity)
0.990
Q4
103238582.11 (0–infinity)
0.998
Q4
13371226.91 (0–infinity)
0.990
CHD patients with diabetes
Myocardial Infarction
Coronary Revascularization
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
1.2 (0.19–7.44)
0.842
Q2
1 (0–infinity)
1
Q3
4.01 (0.74–21.73)
0.107
Q3
1 (0–infinity)
1
Q4
17.11 (3.82–76.54)
<0.001
Q4
240296731.77 (0–infinity)
0.996
New-Onset Heart Failure
Hospitalization due to Angina
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
10955584 (0– infinity)
0.994
Q2
2.51 (0.49–12.87)
0.271
Q3
1 (0–infinity)
1
Q3
4.95 (0.95–25.79)
0.058
Q4
309087327.45 (0–infinity)
0.993
Q4
28.65 (6.45–127.3)
<0.001
Presumed Ischemic Stroke
Cardiac Death
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
0.79 (0.11–5.79)
0.818
Q2
4.19 (0.48–36.8)
0.196
Q3
0 (0–infinity)
0.992
Q3
12.06 (1.43–101.92)
0.022
Q4
1.5 (0.24–9.3)
0.663
Q4
17.36 (2.21–136.5)
0.007
CHD patients with BMI ≥25 kg/m2
Myocardial Infarction
Coronary Revascularization
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
1.36 (0.36–5.05)
0.650
Q2
0 (0–infinity)
0.995
Q3
1.36 (0.36–5.05)
0.650
Q3
0 (0–infinity)
0.995
Q4
4.51 (1.42–14.26)
0.010
Q4
3.29 (0.38–28.47)
0.279
New-Onset Heart Failure
Hospitalization due to Angina
Group apoB ref = Q1
OR (95% CI)
P (Wald's test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's test)
Q2
0.74 (0.04–12.24)
0.833
Q2
0.8 (0.27–2.34)
0.682
Q3
0 (0–infinity)
0.992
Q3
0.59 (0.19–1.82)
0.359
Q4
8.81 (1.11–70.11)
0.040
Q4
3 (1.19–7.54)
0.019
Presumed Ischemic Stroke
Cardiac Death
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's Test)
Q2
0.36 (0.03–4.12)
0.411
Q2
2.41 (0.46–12.76)
0.299
Q3
0.36 (0.03–4.12)
0.411
Q3
2.89 (0.56–14.85)
0.204
Q4
0.76 (0.12–4.78)
0.771
Q4
3.9 (0.83–18.37)
0.085
CHD patients with metabolic syndrome
Myocardial Infarction
Coronary Revascularization
Group apoB ref = Q1
OR (95% CI)
P (Wald's) Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's) Test)
Q2
0.49 (0.13–1.91)
0.307
Q2
0 (0–infinity)
0.993
Q3
1.4 (0.42–4.67)
0.584
Q3
0 (0–infinity)
0.994
Q4
6.71 (2.47–18.21)
< 0.001
Q4
6.07 (0.75–49.08)
0.091
New-Onset Heart Failure
Hospitalization due to Angina
Group apoB ref = Q1
OR (95% CI)
P (Wald's) Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's) Test)
Q2
0.64 (0.04–10.41)
0.753
Q2
0.51 (0.15–1.75)
0.285
Q3
0 (0–infinity)
0.991
Q3
1.33 (0.43–4.09)
0.615
Q4
17.27 (2.26–131.65)
0.006
Q4
7.95 (3.14–20.1)
< 0.001
Presumed Ischemic Stroke
Cardiac Death
Group apoB ref = Q1
OR (95% CI)
P (Wald's) Test)
Group apoB ref = Q1
OR (95% CI)
P (Wald's) Test)
Q2
1.96 (0.2–19.24)
0.564
Q2
2.35 (0.47–11.67)
0.298
Q3
0.97 (0.06–15.82)
0.982
Q3
4.28 (0.87–21.01)
0.073
Q4
2.56 (0.28–23.4)
0.406
Q4
7.53 (1.7–33.45)
0.008
BMI body mass index; CHD = coronary heart disease; ref = reference; TG triglyceride; Q1 = lower quartile of apoB; Q2 = second quartile of apoB; Q3 = third quartile of apoB; Q4 = upper quartile of apoB.
In the AS group, patients in the fourth quartile group of apoB had 21.85 times and 9.05 times higher risk of new-onset heart failure and hospitalization due to angina than the first quartile group, respectively (P < 0.001 and P = 0.044). According to the baseline table, the prevalence of hypertension was higher in the AS group than in the Normal group, which may account for the higher incidence of heart failure. However, there were no positive results, and results were nonsignificant (P > 0.05) in the CHD patients with TG ≥2.3 mmol/L. Such contradictory results may be due to average HDL-C levels <1 mmol/L in these patients with high TG levels (0.96 [0.32] vs 1.13 [0.31]; P < 0.001) (Table III). In CHD patients with diabetes, it is striking that the ORs for myocardial infarction, hospitalization due to angina, and cardiac death were significant (17.11, 28.65, and 17.36; P < 0.001, P < 0.001, and P = 0.007). In CHD patients with BMI ≥25 kg/m2, ORs for prediction of myocardial infarction, new-onset heart failure, and hospitalization due to angina were 4.51 (P = 0.01), 8.81 (P = 0.04), and 3 (P = 0.019). In CHD patients with metabolic syndrome, aside from presumed ischemic stroke, apoB was associated with all remaining 5 risk events (P ≤ 0.05 for all). The results in Table V indicate that there was a significant positive correlation between apoB and myocardial infarction, hospitalization due to angina, and cardiac death, especially in patients with obesity, diabetes, and metabolic syndrome.
apoB Performed Better Than Other Lipids in Predicting the Presence of Long-Term MACEs in CHD Patients With Diabetes, Obesity, and Metabolic Syndrome
Analysis was performed to assess the predictive power of lipids for the long-term MACEs. ROC curves were used to illustrate the predictive value of apoB, TC, LDL-C, TG, non–HDL-C, Lp(a), HDL-C, and apoA-I for the presence of MACEs in different groups. In Figure 4, apoB is highlighted with a red line, and the light yellow region represents the AUC of apoB. Table VI summarizes the AUC values of apoB, LDL-C, and non–HDL-C for MACEs in different groups. In the Normal group and the AS group, the AUC of LDL-C was much larger than apoB (0.912 vs 0.774 [P = 0.042]: 0.906 vs 0.707 [P = 0.101]). This suggests that, for normal populations or patients with only mild AS, the traditional risk factor LDL-C still has a higher predictive value for long-term MACEs than apoB. One striking finding from the ROC results indicated that CHD patients with obesity, diabetes, or metabolic syndrome had curves of all other lipids below the apoB curve, which suggests that in these specific populations, apoB can better predict long-term MACEs than other lipids. This predictive advantage of apoB was not seen in the low LDL-C group. Taken together, these results suggest that apoB exhibits superior risk prediction values for long-term MACEs than other lipids, especially in CHD patients with diabetes, obesity, or metabolic syndrome.
Figure 4Receiver-operating characteristic curves of apolipoprotein B (apoB), total cholesterol (TC), LDL-C, triglyceride (TG), non–HDL-C, lipoprotein a (Lp[a]), HDL-C, and apolipoprotein A-I (apoA-I) for predicting the presence of major adverse cardiac events in different groups. (A) Group of all 826 patients. (B) Normal group. (C) Atherosclerosis (AS) group. (D) Coronary heart disease (CHD) + TG ≥2.3 mmol/L group. (E) CHD + diabetes group. (F) CHD + body mass index (BMI) ≥25 kg/m2 group; (G) CHD + metabolic syndrome group. (H) CHD + LDL-C <1.8 mmol/L group.Compared with apoB: * p<0.05; ** p<0.01; *** p<0.001.
Despite Achieving Optimal LDL-C or Non–HDL-C Levels, CHD Patients May Still Be at Risk of Worse Survival if They Are Unable to Achieve Tighter apoB Control (Lower Cut Points Such as 65 mg/dL)
Through deep analysis of ROC curves, we found that except for apoB, commonly used lipid indicators such as LDL-C and non–HDL-C also showed satisfactory AUC values in the 4 groups (CHD + TG ≥2.3 mmol/L, CHD + diabetes, CHD + BMI ≥25 kg/m2, and CHD + metabolic syndrome) (Table V). However, in the CHD patients, plotting the Kaplan-Meier curves for optimal LDL-C or non–HDL-C levels showed that strict control of apoB is required for better long-term survival.
The 2019 European Society of Cardiology (ESC)/European Atherosclerosis Society (EAS) guidelines recommend that the non–HDL-C secondary goal is <2.2 mmol/L for very-high-risk people and the apoB secondary goal is <65 mg/dL. Results indicate that subjects with apoB >65 mg/dL yielded a worse survival probability (70% vs 97.56%; P = 0.0078) even when reductions in non–HDL-C to <2.2 mmol/L are achieved (Figure 5A). As shown in Figure 5B, when LDL-C levels are <1.8 mmol/L, the survival curve of the apoB >65 mg/dL group is below that of the apoB ≤65 mg/dL group, with a survival probability of 79.31% vs 97.67% (P = 0.055). Although the Kaplan-Meier curve P value was >0.05, individual and quantitative results of MACEs indicated that more events of myocardial infarction (n = 8 [22.9%]; P < 0.0001), hospitalization due to angina (n = 9 [25.7%]; P < 0.001), presumed ischemic stroke (n = 1 [0.03%]; P < 0.05), and cardiac death (n = 6 [17.1%]; P < 0.0001) occurred for the apoB ≥poBp g/L group than in the apoB <0.65 g/L group even when LDL-C levels are maintained at <1.8 mmol/L) (Table VI). Because there is no specific value defined for “very low LDL-C” in 2019 ESC/EAS, we opted for LDL-C <1.8 mmol/L, not LDL-C <1.4 mmol/L or 0.7 mmol/L, as a cutoff value following the consensus of lipid control from 10 years ago. This was used to analyze whether apoB is associated with worse survival when LDL-C has been lowered to an acceptable level. Unfortunately, no statistical results were acquired.
Figure 5Kaplan-Meier curve of the apolipoprotein B (apoB) <0.65 g/L group and the apoB ≥0.65 g/L group when non–HDL-C was <2.2 mmol/L or LDL-C was <1.8 mmol/L in patients with coronary heart disease (CHD).
Risk assessment of ASCVD is critical for disease prevention and treatment strategies. In addition to traditional risk factors, such as smoking, LDL-C, blood pressure, diabetes, BMI, family history of ASCVD, age, and sex, more and more nontraditional risk factors have begun to be an area of concern. There is growing supportive evidence for the idea that apoB directly reflects the circulating numbers of atherogenic LDL particles, making it a more reliable indicator of risk compared with LDL-C or non–HDL-C. Calculated apoB helps to achieve incremental prognostic information of cardiovascular events in patients with ST-segment elevation myocardial infarction, superior to both non–HDL-C and LDL-C.
found that high apoB (>0.65 mg/L) was associated with MACEs and particularly subsequent nonfatal myocardial infarction after acute coronary syndrome. A study in the Journal of the American College of Cardiology
From the American Association of Neurological Surgeons (AANS), American Society of Neuroradiology (ASNR), Cardiovascular and Interventional Radiology Society of Europe (CIRSE), Canadian Interventional Radiology Association (CIRA), Congress of Neurological Surgeons (CNS), European Society of Minimally Invasive Neurological Therapy (ESMINT), European Society of Neuroradiology (ESNR), European Stroke Organization (ESO), Society for Cardiovascular Angiography and Interventions (SCAI), Society of Interventional Radiology (SIR), Society of NeuroInterventional Surgery (SNIS), and World Stroke Organization (WSO) Multisociety Consensus Quality Improvement Revised Consensus Statement for Endovascular Therapy of Acute Ischemic Stroke.
included 13,015 statin-treated patients and conducted follow-up with a median of 8 years. The study found discordant apoB levels above the median with LDL-C or nonHDL-C levels below labeled apoB as a more accurate marker of all-cause mortality risk in statin-treated patients than LDL-C or nonHDL-C, and a more accurate marker of risk for myocardial infarction than LDL-C, compared to concordant values below medians. The analysis concluded that there is a J-shaped association between apoB and all-cause mortality, with an increased risk of all-cause mortality at both high and low apoB concentrations, with the lowest risk at 73 mg/dL. A meta-analysis of 7 placebo-controlled statin trials
From the American Association of Neurological Surgeons (AANS), American Society of Neuroradiology (ASNR), Cardiovascular and Interventional Radiology Society of Europe (CIRSE), Canadian Interventional Radiology Association (CIRA), Congress of Neurological Surgeons (CNS), European Society of Minimally Invasive Neurological Therapy (ESMINT), European Society of Neuroradiology (ESNR), European Stroke Organization (ESO), Society for Cardiovascular Angiography and Interventions (SCAI), Society of Interventional Radiology (SIR), Society of NeuroInterventional Surgery (SNIS), and World Stroke Organization (WSO) Multisociety Consensus Quality Improvement Revised Consensus Statement for Endovascular Therapy of Acute Ischemic Stroke.
for the Management of Dyslipidemia recommend non–HDL-C and apoB as alternative indicators for LDL-C. The 2018 American Heart Association/American College of Cardiology cholesterol management guidelines
2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.
advised that in some circumstances, especially hypertriglyceridemia, measurement of apoB has an advantage over LDL-C, and detection of apoB levels when TG ≥200 mg/dL is recommended. The 2019 ESC/EAS guidelines
suggest that apoB can be an alternative to non–HDL-C or LDL-C as the primary measurement in subjects with high TG levels, diabetes, obesity, metabolic syndrome, or very low LDL-C levels.
Thus far, no relevant clinical studies have systematically explored the predictive value of apoB for the long-term MACEs of these specific CHD patients (CHD + TG ≥2.3 mmol/L, CHD + diabetes, CHD + BMI ≥25 kg/m2, CHD + metabolic syndrome, and CHD+ very low LDL-C). We divided CHD patients into 5 groups for whom the 2019 ESC guidelines recommend that apoB levels should be of particular concern. In our cohort, we found that the baseline apoB level was positively associated with the hazard of MACEs over 10 years after adjusting for sex, age, smoking, hypertension, diabetes, and other lipids such as TC, TG, LDL-C, Lp(a), and HDL-C. Further statistical analysis indicated that apoB was superior to other lipid components in predicting long-term cardiovascular events in coronary AS, including the AS group and CHD subpopulations (group CHD + diabetes, CHD + BMI ≥25, and CHD + metabolic syndrome); in the Normal population, LDL-C remains the major risk factor for predicting MACEs. There was a significant positive correlation between apoB and myocardial infarction, hospitalization due to angina, and cardiac death, particularly in CHD patients whose condition is complicated with diabetes, obesity, and metabolic syndrome. Our further analysis revealed that despite achieving optimal LDL-C or non–HDL-C levels, patients with CHD are still at risk of worse survival if they did not reach the tighter apoB control level (lower cut points such as 65 mg/dL). apoB is currently not a routine part of the standard lipid profiles and ASCVD risk assessment models (traditional lipid profile evaluating ASCVD risk includes TC, TG, HDLC, and LDL-C).
Instead, as the main target of drug intervention, LDL-C is widely used and supported by much high-level clinical research evidence. In patients with TG >4.5 mmol/L, however, the Friedewald formula is not applicable, which may lead to inaccurate measurements of LDL-C. An increasing number of studies and guidelines recommend that apoB measurements be used if available.
American Association of Clinical Endocrinologists and American College of Endocrinology Guidelines for Management of Dyslipidemia and Prevention of Cardiovascular Disease.
European Atherosclerosis Society (EAS) and the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Joint Consensus Initiative. Quantifying Atherogenic Lipoproteins for Lipid-Lowering Strategies: Consensus-Based Recommendations from EAS and EFLM.
Although apoB is challenging the position of LDL-C, apoB has not been completely and widely validated, and it is still viewed as an “alternate target” due to the incomplete evidence of clinical performance and clinical effectiveness.
European Atherosclerosis Society (EAS) and the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Joint Consensus Initiative. Quantifying Atherogenic Lipoproteins for Lipid-Lowering Strategies: Consensus-Based Recommendations from EAS and EFLM.
Meta-Analysis of Comparison of Effectiveness of Lowering Apolipoprotein B versus Low-Density Lipoprotein Cholesterol and Nonhigh-Density Lipoprotein Cholesterol for Cardiovascular Risk Reduction in Randomized Trials.
Our research provides evidence that a high baseline apoB level is associated with poor prognosis in coronary AS patients, including the AS group and specific CHD subpopulations. Focusing solely on LDL-C in these patients would ignore the more accurate risk represented by apoB. ApoB could be a “risk-enhancing factor,” earlier intervention of which would contribute to the risk management of long-term MACEs.
The present study had certain limitations. Because this study was a retrospective analysis, biases in data analysis were unavoidable. Limitations of our study include small sample size and selection bias, as all the study patients came from one hospital rather than from multiple centers. For CHD patients complicated with very low LDL-C, this was especially true, where only a few patients with LDL-C <1.8 mmol/L were available for statistical analysis. In addition, as an observational study, we only evaluated the association of baseline apoB level and long-term MACEs; it is thus unclear whether the emergence of MACEs would be influenced by the dynamic changes of apoB. Thus, the study results provide some evidence for clinical reference but cannot be used to generalize implications for broader patient populations, and more large-scale clinical trials and prospective studies are required.
Conclusions
We found that apoB could be a “risk-enhancing factor” in patients with coronary AS disease, whereas in the Normal population, LDL-C still acts as a major risk factor for predicting MACEs. apoB is a good predictor for long-term cardiovascular events in those with coronary AS disease, including the AS group and CHD subpopulations (including CHD + DM, CHD + BMI ≥25, and CHD + metabolic syndrome). apoB is superior to other lipids in predicting the presence of myocardial infarction, hospitalization due to angina, and cardiac death, particularly in CHD patients whose condition is complicated with diabetes, obesity, and metabolic syndrome. Despite achieving optimal LDL-C or non–HDL-C levels, patients with CHD are still at risk of worse survival if they are unable to reach the tighter apoB control levels (lower cut points such as 65 mg/dL). More attention should be paid to special populations with residual elevations of atherogenic particle numbers, and a focus on lowering baseline apoB may help achieve more long-term benefits.
Table VIIQuantitative numbers of MACEs in the apoB<0.65gg/L group and apoB ≥0.65 gg/L group when non–LDL-C or LDL-C were controlled under target goals.
Variable
Myocardial Infarction
Coronary Revascularization
New-Onset Heart Failure
Hospitalization due to Angina
Presumed Ischemic Stroke
Cardiac Death
Non–HDL-C <2.2 mmol/L
apoB <0.65g/L (n = 41)
2 (4.9%)
0
0
2 (4.9%)
0
2 (2.4%)
apoB ≥0.65g/L (n = 20)
6 (30%) *
0
0
6 (30.0%)*
1 (5.0%)
6 (30.0%)**
LDL-C <1.8 mmol/L
apoB <0.65g/L (n = 35)
0
0
0
0
0
0
apoB ≥0.65g/L (n = 35)
8 (22.9%)***
0
0
9 (25.7%)***
1 (0.03%)*
6 (17.1%)***
Compared with events that did not occur (χ 2 test): *P < 0.05; **P < 0.01; ***P < 0.001.
This work was supported by the National Natural Science Foundation of China (81370331) to Dr Chen.
Dr Zhang was responsible for resources, formal analysis, software, writing/original draft preparation; Dr Ni was responsible for data curation and supervision; and Dr Chen was responsible for conceptualization and writing/reviewing and editing.
The authors thank all the patients and their families for their cooperation during follow-up information.
Declarations of Interest
None declared.
Data Availability
All the data were collected by the graduate students of the research group through the electronic medical records system of the hospital. Results of out-of-hospital events were obtained by follow-up telephone call. All code/data will be provided upon reasonable request.
AACC Lipoproteins and Vascular Diseases Division Working Group on Best Practices. Apolipoprotein B and Cardiovascular Disease Risk: Position Statement from the AACC Lipoproteins and Vascular Diseases Division Working Group on Best Practices.
European Atherosclerosis Society Consensus Panel. The Polygenic Nature of Hypertriglyceridaemia: Implications for Definition, Diagnosis, and Management.
From the American Association of Neurological Surgeons (AANS), American Society of Neuroradiology (ASNR), Cardiovascular and Interventional Radiology Society of Europe (CIRSE), Canadian Interventional Radiology Association (CIRA), Congress of Neurological Surgeons (CNS), European Society of Minimally Invasive Neurological Therapy (ESMINT), European Society of Neuroradiology (ESNR), European Stroke Organization (ESO), Society for Cardiovascular Angiography and Interventions (SCAI), Society of Interventional Radiology (SIR), Society of NeuroInterventional Surgery (SNIS), and World Stroke Organization (WSO)
Multisociety Consensus Quality Improvement Revised Consensus Statement for Endovascular Therapy of Acute Ischemic Stroke.
2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.
American Association of Clinical Endocrinologists and American College of Endocrinology Guidelines for Management of Dyslipidemia and Prevention of Cardiovascular Disease.
European Atherosclerosis Society (EAS) and the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Joint Consensus Initiative. Quantifying Atherogenic Lipoproteins for Lipid-Lowering Strategies: Consensus-Based Recommendations from EAS and EFLM.
Meta-Analysis of Comparison of Effectiveness of Lowering Apolipoprotein B versus Low-Density Lipoprotein Cholesterol and Nonhigh-Density Lipoprotein Cholesterol for Cardiovascular Risk Reduction in Randomized Trials.
00222-3&id=gr1.jpg){kind=link}
00222-3&id=gr2.jpg){kind=link}
00222-3&id=gr3.jpg){kind=link}
00222-3&id=gr4.jpg){kind=link}
00222-3&id=gr5.jpg){kind=link}