Liver Fibrosis Scoring Systems as Novel Tools for Predicting Recurrent Cardiovascular Events in Patients with a Prior Cardiovascular Event : Cardiology Discovery

Journal Logo

Original Articles

Liver Fibrosis Scoring Systems as Novel Tools for Predicting Recurrent Cardiovascular Events in Patients with a Prior Cardiovascular Event

Liu, Huihui1; Cao, Yexuan1; Jin, Jinglu1; Guo, Yuanlin1; Zhu, Chenggang1; Wu, Naqiong1; Hua, Qi2; Li, Yanfang3; Hong, Lifeng4; Dong, Qian1; Li, Jianjun1,∗

Editor(s): Xu, Tianyu; Fu, Xiaoxia

Author Information
Cardiology Discovery 1(4):p 214-222, December 2021. | DOI: 10.1097/CD9.0000000000000033

Abstract

CLINICAL PERSPECTIVE

WHAT IS NEW?

  • Among patients with coronary artery disease (CAD) who had experienced a prior cardiovascular event (CVE), those with high liver fibrosis scores (LFSs) had a significantly higher incidence of recurrent CVEs (RCVEs) compared with those with low LFSs. Higher levels of LFSs were independently associated with an increased risk of RCVEs.
  • The association between LFSs and RCVEs was not modified by sex, age, body mass index, diabetes, and hypertension status.

WHAT ARE THE CLINICAL IMPLICATIONS?

  • Considering the close relationship between LFSs and RCVEs, LFSs may be used as novel tools for risk stratification in CAD patients with a prior CVE.
  • CAD patients with a prior CVE who have high LFSs should be given strict secondary prevention management.

Introduction

The risk of recurrent cardiovascular events (RCVEs) in individuals with coronary artery disease (CAD) is high, especially in those who have experienced a prior cardiovascular event (CVE).[1,2] Efficient secondary prevention strategies in those who have experienced a prior CVE are important, as RCVEs are associated with increased hospital visits, physician time, testing, other procedures, medications, and ultimately increased costs.[1] Hence, the need to decrease the risk of RCVEs is urgent and should be considered when making clinical decisions. However, there is substantial variation in the risk of RCVEs among patients with established CAD.[3–5] Although well-validated instruments have been developed to predict the risk of initial CVEs,[6] there are fewer validated tools to predict long-term prognosis and guide decision-making regarding patients with a prior CVE.[7] There is great interest among cardiovascular researchers in the improvement of risk stratification and identification of high-risk patients in patients who have experienced a prior CVE.

Liver fibrosis (LF) in patients with non-alcoholic fatty liver disease (NAFLD) has been suggested to be a strong predictor of mortality,[8,9] and cardiovascular disease (CVD) and CVEs.[10,11] Liver biopsy and elastography are effective methods for evaluating the severity of LF, but they are invasive and/or expensive. For practical and economic reasons, several non-invasive scoring systems have been developed to evaluate LF.[12] These LF scores (LFSs) incorporate demographic, clinical, and simple laboratory parameters. The most commonly used are the NAFLD fibrosis score (NFS),[13] fibrosis-4 (FIB-4) index,[14] Forns score,[15] and body mass index (BMI), aspartate aminotransferase (AST)/alanine aminotransferase (ALT) ratio, and diabetes score (BARD),[16] gamma-glutamyltransferase/platelet ratio (GPR),[17] AST/ALT ratio,[18] and AST/platelet ratio index (APRI),[19] which were originally developed in patients with various liver diseases. Besides their reasonable accuracy at detecting advanced LF, these LFSs have recently been shown to be associated with adverse outcomes (liver-related complications, all-cause death, or mortality mainly from cardiovascular causes) in large studies, not just limited to NAFLD patients,[20–22] but also the general population[23] and patients with different CVDs.[24–26]

However, there have been no studies examining the predictive value of non-invasive LFSs for RCVE in patients with a prior CVE, except for a recent study evaluating the association between NFS and the risk of RCVE in individuals with the post-acute coronary syndrome (ACS).[3] Thus, this study aimed to comprehensively assess the associations of LFSs with the risk of RCVE in a large, real-world cohort of CAD patients with a prior CVE (ACS, stroke, percutaneous coronary intervention, or coronary artery bypass grafting) to test the hypothesis that LFSs are useful tools for predicting RCVEs in these patients.

Methods

All procedures involving humans were approved by the hospital's ethical review board (at FuWai Hospital & National Center for Cardiovascular Diseases, Beijing, China) and follow the 1964 Declaration of Helsinki and its later amendments. Informed consent was obtained from all participants.

Study design and population

As shown in Figure 1, from March 2011 to December 2016, 11,906 Chinese patients scheduled for coronary angiography because of angina-like chest pain, positive treadmill exercise test, and/or significant stenosis (indicated by coronary computed tomography angiography) were consecutively recruited from 4 medical centers according to the same protocol. Based on medical histories, assistant examinations, and angiographic results, 7846 patients were determined to have CAD along with a prior CVE (ACS, stroke, percutaneous coronary intervention, or coronary artery bypass grafting) 2 to 12 months before enrollment. Next, 1296 patients were excluded based on the following exclusion criteria: missing detailed laboratory data, hepatitis B/C virus infection, autoimmune hepatitis, hereditary liver disease, excessive alcohol consumption (>21 drinks/week in men and >14 drinks/week in women), secondary causes of fatty liver, drug-induced liver disease, active infections, severe liver and/or renal insufficiency, and malignant disease. Subsequently, 23 patients were lost to follow-up. The resulting sample consisted of 6527 CAD patients with a prior CVE. All patients were prescribed standard secondary prevention therapy (antiplatelet drugs, statins, beta-blockers, angiotensin-converting enzyme inhibitors, or angiotensin receptor blockers) following enrollment and were regularly followed up for the occurrence of major RCVEs, comprising cardiovascular death, non-fatal myocardial infarction (MI), and stroke.

F1
Figure 1:
Selection flowchart of the study sample. CAD: Coronary artery disease; CVEs: Cardiovascular events.

Clinical assessment

For each participant, baseline information, including lifestyle factors, demographic factors, medical history, and medication use was collected by cardiologists. According to our previous studies,[27–29] the traditional risk factors were defined as follows. Diabetes mellitus (DM) was defined as fasting plasma glucose (FPG) ≥7.0 mmol/L, 2-hour plasma glucose ≥11.1 mmol/L in an oral glucose tolerance test, or currently taking hypoglycemic drugs or insulin. Hypertension was defined as self-reported hypertension, currently taking antihypertensive drugs, or recorded systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg three or more consecutive times. Current smoking was defined as regular smoking over the previous 12 months. Baseline medication use was defined as taking the medication continuously for ≥3 months before enrollment.

Biochemical analysis

Blood samples were collected from each patient after ≥12 hours of fasting in the morning. Serum liver enzymes (ALT, AST, GGT, and alkaline phosphatase (ALP)), lipid profiles (total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C)), albumin, creatinine, platelet count, FPG, glycosylated hemoglobin, and high-sensitivity C-reactive protein were measured using the same standard laboratory methods in each hospital, as described in our previous studies.[29,30]

Liver fibrosis scores

LFSs were computed according to the following published formulas [Supplementary Table 1, https://links.lww.com/CD9/A15].[13–19] NFS = −1.675 + (0.037 × age [years]) + (0.094 × BMI [kg/m2]) + (1.13 × impaired fasting glucose/DM [yes = 1, no = 0]) + (0.99 × AST/ALT ratio) − (0.013 × platelet count [109/L]) − (0.66 × albumin [g/dL]). FIB-4 = (age [years] × AST [U/L])/(platelet count [109/L] × ALT[U/L]). Forns score = 7.811 − 3.131 × log (platelet count [109/L]) + 0.781 × log (GGT [U/L]) + 3.467 × log (age [years]) − 0.014 × TC (mg/dL). BARD = (BMI ≥28 kg/m2 [yes = 1, no = 0]) + (AST/ALT ratio ≥0.8 [yes = 2, no = 0]) + (DM [yes = 1, no = 0]). GPR = GGT/platelet count (109/L). AST/ALT ratio = AST/ALT. APRI = AST (U/L)/upper limit of normal for AST × 100/platelet count (109/L). The originally developed cutoffs were used to determine the low-, intermediate-, and high-score subgroups for each LFS. Additionally, to better clarify the associations of LFSs with the risk of RCVE, age-adjusted cut-offs for NFS (−1.455 and 0.120) and FIB-4 (1.3 and 2.0) were also used to determine NFS and FIB-4 subgroups.[31]

Follow-up

Patients were followed up at 6-month intervals via face-to-face and/or telephone interviews until the occurrence of a major RCVE (cardiovascular death, non-fatal MI, or stroke) or December 2018 by well-trained cardiologists and nurses who were blinded to the study aims. The definitions of RCVEs were as follows. Cardiovascular death was defined as death primarily caused by acute MI, congestive heart failure, malignant arrhythmia, or other structural or functional cardiac diseases. Non-fatal MI was defined as positive cardiac troponins along with typical chest pain or typical electrocardiogram serial changes. Stroke was defined as persistent neurological dysfunction with documentation of acute cerebral infarction on computed tomography and/or magnetic resonance imaging. The events were then independently considered by 3 experienced cardiologists.

Statistical analysis

Continuous variables are expressed as mean ± SD or median (Q1, Q3), as appropriate. The differences between groups were determined using analysis of variance or the Kruskal–Wallis H test, as appropriate. Categorical variables are presented as frequency (percentage) and were analyzed by the χ2 test or Fisher exact test, as appropriate. The event-free survival rates (RCVE, and cardiovascular death, non-fatal MI, and stroke separately) among LFS-based risk subgroups were calculated using the Kaplan-Meier method. Cox proportional hazard regression and Poisson regression analyses were performed to evaluate the relationships between LFSs and RCVEs. A restricted cubic spline was used to assess the linearity assumptions of the association of continuous LFSs with RCVEs. To further evaluate the consistency of the results, potential effect modification was assessed by stratification by sex, age (<65 vs. ≥65 years), BMI (<24 vs. ≥24 kg/m2), DM, and hypertension status. For all analyses, two-tailed P values < 0.05 were considered statistically significant. The statistical analyses were performed with SPSS version 24.0 software (SPSS Inc., Chicago, Illinois, USA).

Results

Baseline characteristics

The baseline characteristics of the participants were categorized as low-, intermediate-, and high-score subgroups according to the published LFS cutoffs [Table 1] [Supplementary Tables 2–4, https://links.lww.com/CD9/A15]. The mean age of the participants was (57.8 ± 10.9) years and 70.4% were men. Participants with high LFSs were more likely to be older, with higher SBP and lower albumin, TC, LDL-C, TG, platelet count, and current smoking rate compared to those with low LFSs.

Table 1 - Baseline characteristics of participants by NFS and FIB-4 levels.
NFS FIB-4


Variable Low (n = 2368) Intermediate (n = 3630) High (n = 529) P Low (n = 4249) Intermediate (n = 2067) High (n = 211) P
Age (years), mean ± SD 49.7 ± 11.8 60.2 ± 8.9 67.8 ± 8.8 <0.001 54.4 ± 10.3 64.1 ± 8.5 65.1 ± 11.0 <0.001
Men, n (%) 1796 (75.8) 2439 (67.2) 362 (68.4) <0.001 3068 (72.2) 1372 (66.4) 157 (74.3) <0.001
BMI (kg/m2), mean ± SD 25.49 ± 3.20 25.96 ± 3.14 26.92 ± 3.46 <0.001 26.11 ± 3.18 25.38 ± 3.15 25.26 ± 3.56 <0.001
Hypertension, n (%) 1345 (56.8) 2351 (64.8) 380 (71.8) <0.001 2622 (61.7) 1327 (64.2) 127 (60.2) 0.135
Diabetes, n (%) 492 (20.8) 1196 (32.9) 186 (35.2) <0.001 1232 (29.0) 575 (27.8) 67 (31.8) 0.376
Current smokers, n (%) 1200 (50.7) 1251 (34.5) 136 (25.7) <0.001 1886 (44.4) 637 (30.8) 64 (30.3) <0.001
Current drinkers, n (%) 345 (14.6) 353 (9.7) 55 (10.4) <0.001 514 (12.1) 221 (10.7) 18 (8.5) 0.099
Revascularization, n (%) 1814 (76.6) 2834 (78.1) 400 (75.6) 0.696 3293 (77.5) 1604 (77.6) 151 (71.6) 0.124
SBP (mmHg), mean ± SD 126 ± 18 129 ± 18 130 ± 17 <0.001 126 ± 17 128 ± 17 128 ± 18 <0.001
DBP (mmHg), mean ± SD 79 ± 11 78 ± 11 75 ± 10 <0.001 78 ± 11 77 ± 10 76 ± 9 <0.001
LVEF (%), mean ± SD 64.47 ± 8.39 63.12 ± 7.69 62.18 ± 8.41 0.040 63.06 ± 8.07 63.25 ± 8.63 61.13 ± 9.52 0.002
Biochemical parameters
 ALT (U/L), median (Q1, Q3) 30 (21, 46) 22 (17, 31) 16 (12, 22) <0.001 25 (18, 37) 20 (15, 30) 23 (15, 39) <0.001
 AST (U/L), median (Q1, Q3) 20 (16, 26) 18 (15, 23) 18 (15, 21) <0.001 18 (15, 22) 20 (17, 25) 28 (20, 54) <0.001
 GGT (U/L), median (Q1, Q3) 34 (23, 54) 26 (19, 39) 21 (16, 32) <0.001 29 (21, 44) 24 (18, 36) 28 (19, 44) <0.001
 ALP (U/L), median (Q1, Q3) 68 (56, 83) 66 (55, 78) 59 (48, 74) <0.001 66 (55, 79) 64 (54, 76) 64 (54, 80) 0.002
 Albumin (g/L), mean ± SD 45.26 ± 4.74 42.59 ± 4.16 39.59 ± 4.62 <0.001 43.68 ± 4.71 42.72 ± 4.47 41.72 ± 6.28 <0.001
 FPG (mmol/L), mean ± SD 5.86 ± 1.84 6.17 ± 1.90 5.93 ± 1.69 <0.001 5.91 ± 1.84 5.78 ± 1.65 5.91 ± 1.72 0.014
 HbA1c (%), mean ± SD 6.13 ± 1.15 6.52 ± 1.14 6.48 ± 0.87 <0.001 6.37 ± 1.16 6.31 ± 1.00 6.27 ± 1.05 0.053
 TC (mmol/L), mean ± SD 4.21 ± 1.28 4.09 ± 1.18 3.91 ± 1.01 <0.001 4.15 ± 1.20 4.01 ± 1.12 3.92 ± 1.06 <0.001
 HDL-C (mmol/L), mean ± SD 1.04 ± 0.31 1.06 ± 0.30 1.05 ± 0.28 0.152 1.03 ± 0.28 1.09 ± 0.30 1.07 ± 0.30 <0.001
 LDL-C (mmol/L), mean ± SD 2.58 ± 1.13 2.46 ± 0.98 2.33 ± 0.91 <0.001 2.53 ± 1.03 2.39 ± 0.99 2.32 ± 0.90 <0.001
 TG (mmol/L), median (Q1, Q3) 1.56 (1.14, 2.24) 1.44 (1.06, 2.01) 1.33 (0.97, 1.78) <0.001 1.55 (1.17, 2.14) 1.37 (1.00, 1.90) 1.25 (0.92, 1.77) <0.001
 HsCRP (mg/L), median (Q1, Q3) 1.41 (0.81, 3.06) 1.34 (0.77, 2.70) 1.41 (0.79, 3.06) <0.001 1.44 (0.78, 2.90) 1.26 (0.70, 2.72) 1.53 (0.85, 3.47) <0.001
 Platelet count (109/L), mean ± SD 246.19 ± 61.55 194.46 ± 42.57 150.51 ± 39.59 <0.001 230.77 ± 56.40 174.53 ± 38.06 133.28 ± 59.27 <0.001
 Creatinine (μmol/L), mean ± SD 77.47 ± 15.57 77.07 ± 18.15 79.56 ± 18.11 0.087 77.16 ± 18.09 79.48 ± 19.96 81.19 ± 20.23 <0.001
Baseline medications, n (%)
 Aspirin 1745 (73.7) 2721 (74.9) 405 (76.5) 0.660 3178 (74.8) 1540 (74.5) 153 (72.6) 0.797
 Statins 1900 (80.2) 2841 (78.3) 379 (71.6) 0.023 3344 (78.7) 1627 (78.7) 149 (70.6) 0.045
 ACEI/ARB 553 (23.3) 819 (22.6) 123 (23.3) 0.829 1016 (23.9) 438 (21.2) 41 (19.4) 0.054
 β-blockers 1053 (44.5) 1665 (45.9) 208 (39.3) 0.150 1937 (45.6) 901 (43.6) 88 (41.7) 0.300
 CCB 402 (17.0) 788 (21.7) 92 (17.4) 0.013 841 (19.8) 403 (19.5) 38 (18.2) 0.887
ACEI: Angiotensin-converting enzyme inhibitors; ALP: Alkaline phosphatase; ALT: Alanine aminotransferase; ARB: Angiotensin receptor blockers; AST: Aspartate aminotransferase; BMI: Body mass index; CCB: Calcium channel blockers; DBP: Diastolic blood pressure; FIB-4: Fibrosis-4; FPG: Fasting plasma glucose; GGT: Gamma-glutamyl transferase; HbA1c: Glycosylated hemoglobin; HDL-C: High-density lipoprotein cholesterol; HsCRP: High-sensitivity C-reactive protein; LDL-C: Low-density lipoprotein cholesterol; LVEF: Left ventricular ejection fraction; NFS: Non-alcoholic fatty liver disease fibrosis score; SBP: Systolic blood pressure; TC: Total cholesterol; TG: Triglyceride.

LFSs and RCVEs

Over a mean follow-up of (54.67 ± 18.80) months, 532 RCVEs (225 cardiovascular deaths, 113 non-fatal MIs, and 194 strokes) were recorded. The Kaplan-Meier curves showed that the individuals with high NFS, FIB-4, Forns, BARD, and AST/ALT scores had significantly lower total event-free survival rates than those with low scores (all P < 0.05) [Figure 2]. Additionally, individuals with intermediate FIB-4, BARD, and AST/ALT scores also had an increased risk of RCVE compared to those with low scores (all P < 0.05). However, there were no significant differences in the risk of RCVE among the 3 subgroups based on GPR or APRI (both P > 0.05) [Figure 2] [Supplementary Figure 1, https://links.lww.com/CD9/A15]. When cardiovascular death and stroke were analyzed separately, we found that individuals with high NFS, FIB-4, Forns, BARD, and AST/ALT scores, and also individuals with intermediate FIB-4, BARD, and AST/ALT scores, had significantly lower cardiovascular death-free survival rates than those with low scores (all P < 0.05). Meanwhile, individuals with high NFS, FIB-4, and AST/ALT scores had an increased risk of stroke compared to those with low scores (all P < 0.05).

F2
Figure 2:
Cumulative event-free survival analysis of RCVEs by LFS level. (A) NFS; (B) FIB-4; (C) Forns score; (D) body mass index, AST/ALT ratio, and BARD; (E) GPR; and (F) AST/ALT ratio. AST/ALT: Aspartate aminotransferase/alanine aminotransferas; BARD: Diabetes score; FIB-4: Fibrosis-4; GPR: Gamma-glutamyltransferase/platelet ratio; LFS: Liver fibrosis score; NFS: Non-alcoholic fatty liver disease fibrosis score; RCVEs: Recurrent cardiovascular events.

The adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) of RCVEs according to LFS-based subgroups are shown in Table 2. After adjusting for potential confounders, the risk of RCVE increased in individuals with intermediate NFS scores (total RCVEs: HR = 1.63, 95% CI: 1.07–2.47; cardiovascular death: HR = 2.45, 95% CI: 1.36–4.43; stroke: HR = 1.68, 95% CI: 1.04–2.72) or high NFS scores (total RCVEs: HR = 2.52, 95% CI: 1.45–4.39; cardiovascular death: HR = 6.46, 95% CI: 3.31–12.61) compared to those with low NFS scores. Similar increases in the risk of RCVE were also observed for individuals with intermediate FIB-4 scores (total RCVEs: HR = 1.75, 95% CI: 1.45–2.11; cardiovascular death: HR = 2.43, 95% CI: 1.81–3.27; stroke: HR = 1.77, 95% CI: 1.30–2.41) or high FIB-4 scores (total RCVEs: HR = 1.62, 95% CI: 1.03–2.27; cardiovascular death: HR = 2.38, 95% CI: 1.26–4.49). Regarding BARD and Forns scores, the multivariable-adjusted HR (95% CI) were 1.94 (1.06–3.54) and 1.35 (1.06–1.73), respectively, for total RCVEs, and 3.13 (1.11–8.78) and 2.08 (1.42–3.04), respectively for cardiovascular death in individuals with high compared to low scores. Regarding BARD scores, the multivariable-adjusted HR (95% CI) were 1.42 (1.15–1.76) for total RCVEs and 1.96 (1.39–2.76) for cardiovascular death in individuals with intermediate compared to low scores. Regarding the AST/ALT ratio, only individuals with intermediate (not high) scores had an increased risk of cardiovascular death (HR = 1.51, 95% CI: 1.03–2.23) compared to those with low scores.

Table 2 - Multivariate Cox regression analyses of RCVEs by LFS levels among patients with coronary artery disease and a prior CVE.
Total RCVEs Cardiovascular death Non-fatal MI Stroke




Score Events, n/total AHR (95%CI) P AHR (95%CI) P AHR (95%CI) P AHR (95%CI) P
NFS
 Low 110/2368 1.00 1.00 1.00 1.00
 Intermediate 336/3630 1.63 (1.07–2.47) 0.023 2.45 (1.36–4.43) 0.003 1.12 (0.60–2.11) 0.725 1.68 (1.04–2.72) 0.033
 High 86/529 2.52 (1.45–4.39) 0.001 6.46 (3.31–12.61) <0.001 0.99 (0.32–3.03) 0.982 1.46 (0.67–3.18) 0.348
 Per 1-SD 532/6527 1.45 (1.21–1.73) <0.001 1.93 (1.49–2.50) <0.001 1.16 (0.85–1.58) 0.363 1.29 (1.04–1.60) 0.020
FIB-4
 Low 266/4249 1.00 1.00 1.00 1.00
 Intermediate 243/2067 1.75 (1.45–2.11) <0.001 2.43 (1.81–3.27) <0.001 0.87 (0.56–1.35) 0.527 1.77 (1.30–2.41) <0.001
 High 23/211 1.62 (1.03–2.27) 0.039 2.38 (1.26–4.49) 0.008 0.92 (0.29–2.96) 0.893 1.40 (0.61–3.21) 0.433
 Per 1-SD 532/6527 1.09 (1.03–1.16) 0.005 1.14 (1.07–1.22) <0.001 0.83 (0.59–1.16) 0.262 1.06 (0.93–1.21) 0.360
Forns score
 Low 24/336 1.00 1.00 1.00 1.00
 Intermediate 470/5872 1.26 (0.79–2.01) 0.339 1.93 (0.81–4.61) 0.141 1.14 (0.45–2.87) 0.782 0.77 (0.37–1.62) 0.491
 High 38/319 1.94 (1.06–3.54) 0.031 3.13 (1.11–8.78) 0.031 1.56 (0.45–5.39) 0.482 1.04 (0.38–2.85) 0.945
 Per 1-SD 532/6527 1.16 (1.03–1.31) 0.016 1.26 (1.05–1.50) 0.011 0.99 (0.77–1.28) 0.944 1.15 (0.92–1.43) 0.215
BARD
 Low 199/3100 1.00 1.00 1.00 1.00
 Intermediate 206/2143 1.42 (1.15–1.76) 0.001 1.96 (1.39–2.76) <0.001 1.29 (0.83–2.01) 0.257 1.08 (0.76–1.53) 0.688
 High 127/1284 1.35 (1.06–1.73) 0.016 2.08 (1.42–3.04) <0.001 0.89 (0.50–1.59) 0.704 1.04 (0.69–1.57) 0.839
GPR
 Low 154/2014 1.00 1.00 1.00 1.00
 Intermediate 190/2332 1.00 (0.76–1.33) 0.994 0.84 (0.54–1.31) 0.437 0.78 (0.42–1.42) 0.409 1.41 (0.89–2.24) 0.138
 High 188/2181 1.05 (0.79–1.39) 0.758 1.17 (0.79–1.64) 0.475 0.93 (0.51–1.69) 0.802 0.96 (0.58–1.60) 0.880
 Per 1-SD 532/6527 0.99 (0.88–1.11) 0.838 1.01 (0.91–1.12) 0.815 0.87 (0.56–1.34) 0.520 0.92 (0.67–1.26) 0.586
AST/ALT
 Low 238/3429 1.00 1.00 1.00 1.00
 Intermediate 242/2651 1.21 (0.94–1.55) 0.143 1.51 (1.03–2.23) 0.037 0.91 (0.52–1.57) 0.725 1.08 (0.72–1.64) 0.706
 High 52/447 1.37 (0.89–2.10) 0.151 1.69 (0.89–3.22) 0.109 1.24 (0.47–3.30) 0.669 1.12 (0.55–2.29) 0.759
 Per 1-SD 532/6527 1.09 (1.00–1.19) 0.064 1.13 (1.00–1.27) 0.053 0.98 (0.74–1.29) 0.861 1.08 (0.94–1.25) 0.290
APRI
 Low 504/6198 1.00 1.00 1.00 1.00
 Intermediate 28/288 0.91 (0.51–1.62) 0.741 0.63 (0.23–1.72) 0.371 1.09 (0.34–3.47) 0.891 1.19 (0.48–2.92) 0.711
 High 0/41
 Per 1-SD 532/6527 0.92 (0.77–1.10) 0.363 0.90 (0.68–1.19) 0.460 0.79 (0.47–1.33) 0.373 0.98 (0.78–1.25) 0.893
As there were only 41 individuals with high APRI scores and no RCVEs were recorded, the calculated hazard ratio was 0, so “–“ is used in the table, which means not available. “–“ in other cells also indicates the data are not available. Adjusted model included age, sex, body mass index, current smoking, alcohol consumption, diabetes mellitus, hypertension, left ventricular ejection fraction, revascularization (percutaneous coronary intervention/coronary artery bypass grafting), creatinine, triglyceride, low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, and baseline statin use, in addition to the variables included in the score formula.AHR: Adjusted hazard ratio; ALT: Alanine aminotransferase; APRI: AST/platelet ratio index; AST: Aspartate aminotransferase; BARD: Body mass index, AST/ALT ratio, and diabetes score; CI: Confidence interval; CVE: Cardiovascular event; FIB-4: Fibrosis-4; GPR: Gamma-glutamyl transferase/platelet ratio; LFS: Liver fibrosis score; MI: Myocardial infarction; NFS: Non-alcoholic fatty liver disease fibrosis score; RCVE: Recurrent cardiovascular event.

Furthermore, a 1-SD increase in NFS, FIB-4, and Forns was associated with a 9%–45% increase in the risk of total RCVEs and a 14%–93% increase in the risk of cardiovascular death, while a 1-SD increase in NFS was associated with a 29% increase in the risk of stroke. However, LFSs were not associated with non-fatal MI, and GPR and APRI were not associated with RCVE. As there were only 41 patients with high APRI scores and no RCVEs were recorded, the calculated hazard ratio for RCVEs was 0, which was presented as “–“ in Table 2.

When age-adjusted cutoffs of NFS and FIB-4 were applied rather than unadjusted cutoffs, more patients were at high risk of LF, and the predictive value for RCVE of high scores was slightly higher (NFS: HR = 2.56, 95% CI: 1.78–3.69; FIB-4: HR = 1.70, 95% CI: 1.21–2.37) [Table 3].

Table 3 - Multivariate Cox regression analyses of RCVEs by age-adjusted NFS and FIB-4 levels
Total RCVEs Cardiovascular death Non-fatal MI Stroke




Score Events, n/total AHR (95%CI) P AHR (95%CI) P AHR (95%CI) P AHR (95%CI) P
Age-adjusted NFS
 Low 111/2368 1.00 1.00 1.00 1.00
 Intermediate 250/3037 1.61 (1.16–2.24) 0.004 2.49 (1.37–4.54) 0.003 1.12 (0.58–2.14) 0.740 1.38 (0.83–2.28) 0.215
 High 171/1122 2.56 (1.78–3.69) <0.001 4.33 (2.28–8.21) <0.001 1.07 (0.46–2.50) 0.881 2.48 (1.42–4.31) 0.001
Age-adjusted FIB-4
 Low 266/4249 1.00 1.00 1.00 1.00
 Intermediate 185/1619 1.66 (1.28–2.14) 2.17 (1.46–3.23) <0.001 1.05 (0.60–1.86) 0.856 1.59 (1.04–2.41) 0.031
 High 81/659 1.70 (1.21–2.37) 0.002 2.43 (1.50–3.94) <0.001 0.74 (0.29–1.90) 0.535 1.61 (0.91–2.83) 0.101
“–“ indicates the data are not available. Adjusted model included sex, body mass index, current smoking, alcohol consumption, diabetes mellitus, hypertension, left ventricular ejection fraction, revascularization (percutaneous coronary intervention/coronary artery bypass grafting), creatinine, triglyceride, low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, and baseline statin use, in addition to the variables included in the score formula.AHR: Adjusted hazard ratio; CI: Confidence interval; FIB-4: Fibrosis-4; MI: Myocardial infarction; NFS: Non-alcoholic fatty liver disease fibrosis score; RCVE: Recurrent cardiovascular event.

In the subgroup analyses of sex, age, BMI, DM, and hypertension status, the increased risk of RCVEs with high LFSs (NFS, FIB-4, Forns, and BARD) was maintained across the different subgroups [Table 4].

Table 4 - Multivariate Cox regression analyses of RCVEs by LFS levels in various subgroups (adjusted HR (95%CI)).
NFS FIB-4 Forns score BARD




Subgroup HR (95%CI) P HR (95%CI) P HR (95%CI) P HR (95%CI) P
Sex
 Men 2.99 (1.81–4.95) 0.011 1.75 (1.40–2.21) <0.001 1.83 (0.84–4.01) 0.131 1.46 (1.16–1.83) 0.001
 Women 1.81 (0.82–3.99) 0.142 1.67 (1.21–2.31) 0.002 1.97 (0.54–7.18) 0.303 1.25 (0.87–1.80) 0.226
Age (years)
 <65 3.79 (1.41–10.17) 0.008 1.51 (1.05–2.15) 0.024 1.73 (0.54–5.51) 0.353 1.34 (0.98–1.84) 0.064
 ≥65 1.51 (0.90–2.54) 0.119 1.61 (1.27–2.05) <0.001 3.90 (0.87–17.57) 0.076 1.49 (1.16–1.91) 0.002
BMI (kg/m2)
 <24 2.81 (1.21–6.52) 0.016 1.64 (1.15–2.33) 0.006 1.28 (0.47–3.49) 0.635 1.43 (0.99–2.08) 0.058
 ≥24 2.66 (1.57–4.52) <0.001 1.80 (1.45–2.24) <0.001 2.32 (1.01–5.35) 0.048 1.38 (1.09–1.73) 0.007
DM
 No 2.52 (1.49–4.25) 0.001 1.81 (1.44–2.28) <0.001 1.71 (0.79–3.72) 0.176 1.65 (1.29–2.12) <0.001
 Yes 2.79 (1.29–6.04) 0.009 1.60 (1.17–2.19) 0.003 1.70 (0.61–4.79) 0.312 0.96 (0.70–1.33) 0.818
Hypertension
 No 2.79 (1.41–5.52) 0.003 1.82 (1.31–2.53) <0.001 2.95 (0.97–9.00) 0.057 1.76 (1.23–2.53) 0.003
 Yes 2.60 (1.51–4.49) 0.001 1.72 (1.37–2.15) <0.001 1.45 (0.67–3.13) 0.345 1.27 (1.00–1.59) 0.047
Adjusted for sex, body mass index, current smoking, alcohol consumption, diabetes mellitus, hypertension, left ventricular ejection fraction, revascularization (percutaneous coronary intervention/coronary artery bypass grafting), creatinine, triglyceride, low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, and baseline statin use, in addition to the variables included in the score formula and variables used for stratification.
High vs. low NFS and Forns scores.
Intermediate + high vs. low FIB-4 and BARD scores.BARD: Body mass index, aspartate aminotransferase/alanine aminotransferase ratio, and diabetes score; BMI: Body mass index; CI: Confidence interval; DM: Diabetes mellitus; FIB-4: Fibrosis-4; HR: Hazard ratio; LFS: Liver fibrosis score; NFS: Non-alcoholic fatty liver disease fibrosis score; RCVE: Recurrent cardiovascular event.

Discussion

Compared to previous studies on the prognostic value of LFSs in different CVDs,[24–26] less is known about whether LFSs can predict RCVEs. This study evaluated the association of non-invasive LFSs with the risk of RCVE among CAD patients who had experienced a prior CVE, and it had a large sample size and long-term follow-up. We observed that high NFS, FIB-4, Forns, and BARD scores were significantly and independently associated with an increased risk of RCVE. Clinically, our data indicate that LFSs can predict adverse outcomes in CAD patients with a prior CVE, providing alternative tools for decision-making regarding secondary prevention of CAD.

Given the high prevalence and mortality regarding CAD, prevention of RCVEs in patients with established CAD is a major public health goal. Moreover, the risk of RCVE in patients with CAD has increased in recent years.[1,32] Notably, the prevalence of RCVE is even higher in CAD patients with a prior CVE. In non-hospitalized individuals who had been diagnosed with ACS, the risk of cardiovascular death is 3%–5% at 30 days and 5%–8% at 6 months, while major CVEs occur in 15%–20% of this patient population by 6 months.[33] In a study of survivors of a prior MI, the incidences of MI recurrence were 5.6% for men and 7.2% for women at 1 year, and 13.9% and 16.2%, respectively at 7 years.[34] However, patients with a prior CVE may show a substantial variation in residual risk of RCVE.[4,35] Thus, risk assessment in this population is essential for effective medical decision-making.

Several established risk factors, including hypertension, DM, hyperlipidemia, smoking, physical inactivity, and left ventricular hypertrophy, are associated with increased risk of RCVE.[36,37] In addition, abnormal findings on cardiac stress test post-MI, decreased left ventricular ejection fraction, persistent advanced heart block, or a new intraventricular conduction abnormality, albumin, and creatinine can predict RCVE.[36] However, even with optimal treatment based on the known risk factors, many patients with CVD have >20% and even >30% 10-year risk of RCVE.[35] Thus, the identification of novel biomarkers or risk factors for better risk stratification and aggressive intervention is needed.

Recent evidence suggested that non-invasive LFSs have prognostic value in various clinical settings,[12,20,22,23] not just in patients with diagnosed NAFLD.[20,22] For example, an analysis of the US National Health and Nutrition Examination Survey population demonstrated that NFS, FIB-4, APRI, and Forns scores predicted increased liver-related and all-cause mortality in the general population (without confirmed NAFLD).[23] Additionally, a registry study with a 3-year follow-up showed that the FIB-4 was independently associated with the risk of CVE and all-cause mortality in patients with atrial fibrillation,[26] while another study reported that NFS was a significant predictor of CVE in 516 patients with chronic heart failure.[25] Moreover, a recent study indicated that higher LFSs, including NFS, FIB-4, Forns, GPR, and APRI, were associated with increased risk of all-cause and cardiovascular mortality in patients with established CAD.[24] The above evidence strongly suggests that LFSs might be useful indexes for predicting worse outcomes even in the general population, whether they are diagnosed with NAFLD or not. Nonetheless, the relationships between LFSs and the risk of RCVE have been less investigated. There has only been 1 study assessing the association between NFS and the risk of RCVE in post-ACS patients,[3] indicating a potential new use for NFS. However, the predictive value of other LFSs and the significance of LFSs in other types of CAD have not previously been investigated.

To comprehensively clarify the association between each LFS and the risk of RCVE, we conducted this study in a cohort of CAD patients with a prior CVE with strict inclusion and exclusion criteria. The results showed that among the LFSs assessed, NFS, FIB-4, Forns, and BARD significantly predicted RCVEs. Furthermore, individuals with intermediate NFS and FIB-4 scores also had an increased risk of stroke compared to those with low scores, which is an interesting finding and needs to be further explored. Sometimes, the adjusted HR values were slightly lower in individuals with high compared to intermediate scores, which may be attributable to the lower blood pressure, weight, and cholesterol levels in individuals with severe LF, and/or the small number of individuals with high scores. Additionally, the allocation of fatal MI to cardiovascular death and insufficient follow-up duration may have led to the lack of associations of the above scores with non-fatal MI. As mentioned above, the relatively complex LFSs (NFS, FIB-4 index, Forns, and BARD, and especially the former 2) had better predictive value for RCVE than the others, which was in line with previous studies.[12,38] The reason may be that, compared to the other 3 simple LFSs, these 4 LFSs involve more variables.

Several potential study limitations must be considered. First, as patients with excessive alcohol consumption and any other liver diseases were excluded, we assumed that the increased LFSs were caused by NAFLD. However, we cannot completely rule out the existence of other undiagnosed liver diseases at baseline. Second, we only calculated the baseline LFSs and could not evaluate the changes at follow-up. Some individuals may have developed increased LF, leading to misclassification. However, this misclassification was non-differential among the participants. Third, due to the inherent limitations of observational studies, the causal relationships between baseline LFSs and RCVEs, as well as the underlying mechanisms, were not clearly determined in this study.

Conclusion

This study, with large sample size and long-term follow-up, comprehensively evaluated the prognostic value of non-invasive LFSs for the risk of RCVE. The novel and important finding are that LFSs can predict the risk of RCVE. Thus, the LFSs may be helpful for risk stratification and medical decision-making for secondary prevention in patients with CAD.

Funding

This study was funded by the Capital Health Development Fund (201614035), CAMS Major Collaborative Innovation Project (2016-I2M-1-011), Fundamental Research Funds for the Central Universities (2019-XHQN09), and Youth Research Fund of Peking Union Medical College (2019-F11). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author contributions

Huihui Liu participated in the research design, obtained the funding, and drafted the manuscript; Yexuan Cao and Jinglu Jin participated in the data analysis; Yuanlin Guo, Chenggang Zhu, Naqiong Wu, Qi Hua, Yanfang Li, Lifeng Hong, and Qian Dong participated in the performance of the research and collected the data. Jianjun Li designed and supervised the study, obtained the funding, and made critical revisions to the manuscript. All authors gave approval of the final version to be published and agreed to be accountable for all aspects of the work.

Conflicts of Interest

None.

References

[1]. LaRosa JC, Deedwania PC, Shepherd J, et al. Comparison of 80 versus 10 mg of atorvastatin on occurrence of cardiovascular events after the first event (from the Treating to New Targets [TNT] trial). Am J Cardiol 2010;105(3):283–287. doi: 10.1016/j.amjcard.2009.09.025.
[2]. Verma S, Bain SC, Buse JB, et al. Occurence of first and recurrent major adverse cardiovascular events with liraglutide treatment among patients with type 2 diabetes and high risk of cardiovascular events: a post hoc analysis of a randomized clinical trial. JAMA Cardiol 2019;4(12):1214–1220. doi: 10.1001/jamacardio.2019.3080.
[3]. Simon TG, Corey KE, Cannon CP, et al. The nonalcoholic fatty liver disease (NAFLD) fibrosis score, cardiovascular risk stratification and a strategy for secondary prevention with ezetimibe. Int J Cardiol 2018;270:245–252. doi: 10.1016/j.ijcard.2018.05.087.
[4]. Morrow DA. Cardiovascular risk prediction in patients with stable and unstable coronary heart disease. Circulation 2010;121(24):2681–2691. doi: 10.1161/CIRCULATIONAHA.109.852749.
[5]. Cannon CP, Blazing MA, Giugliano RP, et al. Ezetimibe added to statin therapy after acute coronary syndromes. N Engl J Med 2015;372(25):2387–2397. doi: 10.1056/NEJMoa1410489.
[6]. Eckel RH, Jakicic JM, Ard JD, et al. 2013 AHA/ACC guideline on lifestyle management to reduce cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2014;129(25):S76–S99. doi: 10.1161/01.cir.0000437740.48606.d1.
[7]. Schwartz GG, Ballantyne CM, Barter PJ, et al. Association of lipoprotein(a) with risk of recurrent ischemic events following acute coronary syndrome: analysis of the dal-outcomes randomized clinical trial. JAMA Cardiol 2018;3(2):164–168. doi: 10.1001/jamacardio.2017.3833.
[8]. Ekstedt M, Hagström H, Nasr P, et al. Fibrosis stage is the strongest predictor for disease-specific mortality in NAFLD after up to 33 years of follow-up. Hepatology 2015;61(5):1547–1554. doi: 10.1002/hep.27368.
[9]. Hagström H, Nasr P, Ekstedt M, et al. Fibrosis stage but not NASH predicts mortality and time to development of severe liver disease in biopsy-proven NAFLD. J Hepatol 2017;67(6):1265–1273. doi: 10.1016/j.jhep.2017.07.027.
[10]. Henson JB, Simon TG, Kaplan A, et al. Advanced fibrosis is associated with incident cardiovascular disease in patients with non-alcoholic fatty liver disease. Aliment Pharmacol Ther 2020;51(7):728–736. doi: 10.1111/apt.15660.
[11]. Baratta F, Pastori D, Angelico F, et al. Nonalcoholic fatty liver disease and fibrosis associated with increased risk of cardiovascular events in a prospective study. Clin Gastroenterol Hepatol 2020;18(10):2324–2331. e4. doi:10.1016/j.cgh.2019.12.026.
[12]. Önnerhag K, Hartman H, Nilsson PM, et al. Non-invasive fibrosis scoring systems can predict future metabolic complications and overall mortality in non-alcoholic fatty liver disease (NAFLD). Scand J Gastroenterol 2019;54(3):328–334. doi: 10.1080/00365521.2019.1583366.
[13]. Angulo P, Hui JM, Marchesini G, et al. The NAFLD fibrosis score: a noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatology 2007;45(4):846–854. doi: 10.1002/hep.21496.
[14]. Sterling RK, Lissen E, Clumeck N, et al. Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology 2006;43(6):1317–1325. doi: 10.1002/hep.21178.
[15]. Forns X, Ampurdanès S, Llovet JM, et al. Identification of chronic hepatitis C patients without hepatic fibrosis by a simple predictive model. Hepatology 2002;36(4 Pt 1):986–992. doi: 10.1053/jhep.2002.36128.
[16]. Harrison SA, Oliver D, Arnold HL, et al. Development and validation of a simple NAFLD clinical scoring system for identifying patients without advanced disease. Gut 2008;57(10):1441–1447. doi: 10.1136/gut.2007.146019.
[17]. Lemoine M, Shimakawa Y, Nayagam S, et al. The gamma-glutamyl transpeptidase to platelet ratio (GPR) predicts significant liver fibrosis and cirrhosis in patients with chronic HBV infection in West Africa. Gut 2016;65(8):1369–1376. doi: 10.1136/gutjnl-2015-309260.
[18]. Sheth SG, Flamm SL, Gordon FD, et al. AST/ALT ratio predicts cirrhosis in patients with chronic hepatitis C virus infection. Am J Gastroenterol 1998;93(1):44–48. doi: 10.1111/j.1572-0241.1998.044_c.x.
[19]. Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology 2003;38(2):518–526. doi: 10.1053/jhep.2003.50346.
[20]. Angulo P, Bugianesi E, Bjornsson ES, et al. Simple noninvasive systems predict long-term outcomes of patients with nonalcoholic fatty liver disease. Gastroenterology 2013;145(4):782–789. e4. doi: 10.1053/j.gastro.2013.06.057.
[21]. Treeprasertsuk S, Björnsson E, Enders F, et al. NAFLD fibrosis score: a prognostic predictor for mortality and liver complications among NAFLD patients. World J Gastroenterol 2013;19(8):1219–1229. doi: 10.3748/wjg.v19.i8.1219.
[22]. Kim D, Kim WR, Kim HJ, et al. Association between noninvasive fibrosis markers and mortality among adults with nonalcoholic fatty liver disease in the United States. Hepatology 2013;57(4):1357–1365. doi: 10.1002/hep.26156.
[23]. Unalp-Arida A, Ruhl CE. Liver fibrosis scores predict liver disease mortality in the United States population. Hepatology 2017;66(1):84–95. doi: 10.1002/hep.29113.
[24]. Chen Q, Li Q, Li D, et al. Association between liver fibrosis scores and the risk of mortality among patients with coronary artery disease. Atherosclerosis 2020;299:45–52. doi: 10.1016/j.atherosclerosis.2020.03.010.
[25]. Takahashi T, Watanabe T, Shishido T, et al. The impact of non-alcoholic fatty liver disease fibrosis score on cardiac prognosis in patients with chronic heart failure. Heart Vessels 2018;33(7):733–739. doi: 10.1007/s00380-017-1113-1.
[26]. Saito Y, Okumura Y, Nagashima K, et al. Impact of the Fibrosis-4 Index on risk stratification of cardiovascular events and mortality in patients with atrial fibrillation: findings from a Japanese multicenter registry. J Clin Med 2020;9(2):584. doi:10.3390/jcm9020584.
[27]. Liu HH, Cao YX, Li S, et al. Impacts of prediabetes mellitus alone or plus hypertension on the coronary severity and cardiovascular outcomes. Hypertension 2018;71(6):1039–1046. doi:10.1161/HYPERTENSIONAHA.118.11063.
[28]. Liu HH, Cao YX, Sun D, et al. Impact of non-alcoholic fatty liver disease on cardiovascular outcomes in patients with stable coronary artery disease: a matched case-control study. Clin Transl Gastroenterol 2019;10(2):e00011. doi:10.14309/ctg.0000000000000011.
[29]. Liu HH, Cao YX, Jin JL, et al. Association of lipoprotein(a) levels with recurrent events in patients with coronary artery disease. Heart 2020;106(16):1228–1235. doi: 10.1136/heartjnl-2020-316586.
[30]. Liu HH, Cao YX, Jin JL, et al. Predicting cardiovascular outcomes by baseline lipoprotein(a) concentrations: a large cohort and long-term follow-up study on real-world patients receiving percutaneous coronary intervention. J Am Heart Assoc 2020;9(3):e014581. doi: 10.1161/JAHA.119.014581.
[31]. McPherson S, Hardy T, Dufour JF, et al. Age as a confounding factor for the accurate non-invasive diagnosis of advanced NAFLD fibrosis. Am J Gastroenterol 2017;112(5):740–751. doi:10.1038/ajg.2016.453.
[32]. Murphy SA, Pedersen TR, Gaciong ZA, et al. Effect of the PCSK9 inhibitor evolocumab on total cardiovascular events in patients with cardiovascular disease: a prespecified analysis from the FOURIER trial. JAMA Cardiol 2019;4(7):613–619. doi:10.1001/jamacardio.2019.0886.
[33]. Goldberg RJ, Currie K, White K, et al. Six-month outcomes in a multinational registry of patients hospitalized with an acute coronary syndrome (the Global Registry of Acute Coronary Events [GRACE]). Am J Cardiol 2004;93(3):288–293. doi: 10.1016/j.amjcard.2003.10.006.
[34]. Smolina K, Wright FL, Rayner M, et al. Long-term survival and recurrence after acute myocardial infarction in England, 2004 to 2010. Circ Cardiovasc Qual Outcomes 2012;5(4):532–540. doi: 10.1161/CIRCOUTCOMES.111.964700.
[35]. Kaasenbrood L, Boekholdt SM, van der Graaf Y, et al. Distribution of estimated 10-year risk of recurrent vascular events and residual risk in a secondary prevention population. Circulation 2016;134(19):1419–1429. doi:10.1161/CIRCULATIONAHA.116.021314.
[36]. Wattanakit K, Folsom AR, Chambless LE, et al. Risk factors for cardiovascular event recurrence in the Atherosclerosis Risk in Communities (ARIC) study. Am Heart J 2005;149(4):606–612. doi: 10.1016/j.ahj.2004.07.019.
[37]. Kaplan RC, Heckbert SR, Furberg CD, et al. Predictors of subsequent coronary events, stroke, and death among survivors of first hospitalized myocardial infarction. J Clin Epidemiol 2002;55(7):654–664. doi:10.1016/s0895-4356(02)00405-5.
[38]. De Vincentis A, Costanzo L, Vespasiani-Gentilucci U, et al. Association between non-invasive liver fibrosis scores and occurrence of health adverse outcomes in older people. Dig Liver Dis 2019;51(9):1330–1336. doi: 10.1016/j.dld.2019.01.017.
Keywords:

Liver fibrosis score; Recurrent cardiovascular event; Coronary artery disease; Risk factor; Prognosis

Supplemental Digital Content

Copyright © 2021 The Chinese Medical Association, published by Wolters Kluwer Health, Inc.