Introduction
Coronary artery disease (CAD) and liver cirrhosis are major causes of death worldwide and share common risk factors, such as obesity, diabetes, and metabolic syndrome.1,2 CAD is classified into chronic coronary syndromes and acute coronary syndromes (ACS).2 In 2020, an estimated 244.11 million people globally lived with CAD, and 8.95 million patients died from it, especially from ACS.3 Liver cirrhosis is the end stage of chronic liver disease and leads to lethal complications, including bacterial infection, acute kidney injury, and acute gastrointestinal bleeding.1 In 2017, it was reported that 122.60 million people worldwide lived with liver cirrhosis, with 1.32 million deaths attributed to the disease.4
Liver cirrhosis is often complicated by systemic inflammation, hyperactivity of the sympathetic nervous system, and increased cardiac output, all of which are potentially associated with the development of CAD.5,6 Additionally, patients with liver cirrhosis have a high risk of bleeding due to the coexistence of portal hypertension and thrombocytopenia.1 Consequently, CAD patients with liver cirrhosis are less likely to receive antithrombotic drugs and have a higher risk of adverse outcomes, including mortality, readmission, and gastrointestinal bleeding,7 compared to those without liver cirrhosis. Conversely, the presence of CAD also increases post-transplant mortality in patients with advanced liver cirrhosis.8
Epidemiological data on CAD in patients with cirrhosis are heterogeneous among studies,9,10 probably due to differences in target populations and the definitions and diagnostic approaches of CAD. To the best of our knowledge, only one meta-analysis has investigated the prevalence of CAD in liver cirrhosis, finding a pooled prevalence of 12.6%, though it included only five studies.10 In recent years, the number of studies addressing the epidemiology of CAD in patients with cirrhosis has rapidly increased. However, there remains a lack of studies to estimate the incidence and prevalence of CAD in patients with cirrhosis, assess the association between the two diseases, and identify factors associated with CAD in cirrhosis. Therefore, we conducted this systematic review and meta-analysis to address these gaps.
Methods
This systematic review and meta-analysis was performed according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) and Meta-analysis of Observational Studies in Epidemiology guidelines.11,12
Registration
This study was registered in the International Prospective Register of Systematic Reviews (hereinafter referred to as PROSPERO) with the registration number CRD42022315248. There was no significant deviation from the protocol registered in PROSPERO.
Search strategy
We searched the PubMed, EMBASE, and Cochrane Library databases from inception to May 17, 2023, without language restriction. Reference lists from relevant papers were manually screened to identify eligible studies. The search terms were as follows: ((liver cirrhosis [all fields]) OR (hepatic cirrhosis [all fields])) AND ((coronary disease [all fields]) OR (coronary heart disease [all fields]) OR (coronary artery disease [all fields]) OR (coronary arteriosclerosis [all fields]) OR (myocardial infarction [all fields]) OR (acute coronary syndrome [all fields]) OR (angina [all fields])).
Selection criteria
Selection criteria were established according to the PICO rule. Participants should be cirrhotic patients, regardless of stages and etiologies. Intervention was not restricted. Comparison should be conducted between patients with and without cirrhosis, if any. The outcome should be the incidence and/or prevalence of CAD.
Exclusion criteria were as follows: 1) duplicated articles; 2) comments, notes, or letters; 3) guidelines or consensus statements; 4) reviews and/or meta-analyses; 5) case reports; 6) experimental or animal studies; 7) patients not diagnosed with liver cirrhosis; 8) CAD not evaluated; 9) overlapping relevant data among studies; and 10) relevant data that could not be extracted.
Definitions
CAD, which refers to the development of thrombosis in the coronary vessels, is divided into chronic coronary syndrome and ACS. ACS primarily includes unstable angina, non-ST-segment elevation myocardial infarction (hereinafter referred to as NSTEMI), and ST-segment elevation myocardial infarction (hereinafter referred to as STEMI). NSTEMI and STEMI are collectively defined as myocardial infarction (MI). The incidence of CAD refers to the new onset of CAD events after a diagnosis of cirrhosis based on data from cohort studies. The prevalence of CAD refers to the presence of CAD in cirrhosis based on data from cross-sectional studies. Severity of CAD was categorized as non-obstructive, obstructive, and severe CAD, defined as luminal stenosis of <50%, ≥50%, and ≥70% in one of the three major coronary arteries, respectively.13
Data extraction
Two authors (CG and LD) independently extracted and evaluated the following data from the included studies: first author, publication year, region, enrollment period, study design, type of publication, number of patients with and without liver cirrhosis, number of patients who developed CAD, endpoint events (i.e., CAD, ACS, and MI), and etiology of cirrhosis. To evaluate the differences in baseline characteristics between cirrhotic patients with and without CAD, the following data were further extracted: diabetes mellitus, hypertension, hyperlipidemia, smoking history, family history of CAD, hepatocellular carcinoma, body mass index, and Child-Pugh and Model for End-Stage Liver Disease (MELD) scores. Disagreements between the two authors (CG and LD) were resolved through discussion with a third author (XQ) until a consensus was achieved.
Study quality assessment
Included studies were assessed using the Joanna Briggs Institute Critical Evaluation.14 Assessment was mainly based on the risk of bias, adequate reporting, and statistical analysis. Responses included “yes”, “no”, “unclear”, and “not applicable”. Only “yes” was scored as one, while “no”, “unclear”, or “not applicable” were scored as zero. The maximum score was 10. Studies that scored ≥7, 5–6, and ≤4 were classified as high, moderate, and low quality, respectively.
Statistical analyses
All analyses were conducted using RStudio version 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria) and SPSS Version 20.0 (SPSS Software, Chicago, IL, USA). The incidence, prevalence, and risk factors of CAD were pooled using a random-effects model. The pooled incidence and prevalence of CAD were expressed as percentages with their 95% confidence intervals (CIs). The incidence rate of CAD in cirrhotic patients was calculated by dividing the number of individuals with new-onset CAD by the total number of individuals with liver cirrhosis. The incidence rate per 1,000 person-years was also calculated, when applicable. The prevalence rate of CAD in cirrhotic patients was calculated by dividing the total number of individuals with CAD by the total number of individuals with liver cirrhosis. Odds ratios (ORs), risk ratios (RRs), and mean differences (MD) with their 95% CIs were calculated for the combined estimates of raw data, when appropriate. A p-value of <0.05 was considered statistically significant. When the reported outcome was incomplete for meta-analysis, results were described in narrative form. Statistical heterogeneity was assessed via I2 statistics and the Chi2 test, where I2 values of 25%, 50%, and 75% represented low, moderate, and high degrees of heterogeneity, respectively, and p < 0.10 by the Chi2 test was considered significant for heterogeneity. The Egger test was used to assess publication bias, with p < 0.1 indicating significant publication bias. Meta-regression and subgroup analyses were performed to explore the sources of heterogeneity. The following covariates were used in the meta-regression and subgroup analyses: region (America vs. Asia vs. Europe vs. Africa), publication year (Before 2015 vs. After 2015), study design (Prospective vs. Retrospective), type of publication (Full-texts vs. Abstracts), study quality (High and Moderate vs. Low), sample size (≥4,545 vs. <4,545; ≥243 vs. <243), etiology of liver cirrhosis (Non-alcoholic steatohepatitis [NASH] cirrhosis vs. Hepatitis C virus [HCV] cirrhosis vs. Alcoholic cirrhosis vs. Primary biliary cirrhosis [PBC] vs. Hepatitis B virus [HBV] cirrhosis), sex (Male vs. Female), mean age (≥57 years vs. <57 years; ≥56 years vs. <56 years), diabetes mellitus (Yes vs. No), hypertension (Yes vs. No), smoking history (Yes vs. No), hyperlipidemia (Yes vs. No), family history of CAD (Yes vs. No), and severity of CAD (Non-obstructive vs. Obstructive vs. Severe). The interaction between subgroups was tested, with p < 0.1 considered indicative of a statistically significant interaction.
Discussion
Our study aimed to assess the epidemiology of CAD in patients with liver cirrhosis and evaluate the association between cirrhosis and CAD. We found that CAD is not uncommon in patients with liver cirrhosis, but current evidence does not support a definitive association between liver cirrhosis and CAD. Additionally, traditional cardiovascular risk factors, including advanced age, male sex, diabetes mellitus, hypertension, dyslipidemia, smoking, family history of CAD, and certain etiologies of chronic liver disease—namely NASH and HCV—are associated with the presence of CAD in these patients.
Our study confirms that traditional risk factors for CAD may also predict or promote the development of cardiovascular disease in patients with cirrhosis. Furthermore, liver cirrhosis is characterized by decreased nitric oxide levels, increased oxidative stress, and elevated levels of vasoconstrictor agents (such as thromboxane A2, COX-1-derived prostanoids, and endothelin-1), as well as inflammatory markers (such as tumor necrosis factor-alpha, nuclear factor kappa B, Toll-like receptor, and angiotensin II). These factors play a significant role in endothelial dysfunction, which can contribute to the development of CAD.6,65–67 In cases of infection, encephalopathy, or bleeding, a fragile hemostatic balance may be disrupted, leading to a heightened risk of thrombosis. Jepsen et al. demonstrated an 8.7-fold increased risk of MI in patients with decompensated cirrhosis who had recently undergone variceal ligation/sclerotherapy or ascites puncture/drainage within 90 days of treatment, compared to those with compensated cirrhosis.17 Additionally, decreased peripheral resistance, compensatory hyperdynamic circulation, and increased cardiac output and heart rate may reduce coronary blood flow, thereby increasing the risk of ACS.1
Patients with cirrhosis should be referred for transplantation when they develop severe hepatic dysfunction (i.e., MELD score ≥15) or experience decompensation events (i.e., ascites, variceal bleeding, hepatic encephalopathy, or hepatorenal syndrome).68 However, CAD is a significant predictor of adverse prognosis in liver transplantation candidates.13 The American Heart Association and the American College of Cardiology Foundation have recommended noninvasive stress testing for liver transplantation candidates with multiple risk factors (e.g., diabetes, prior cardiovascular disease, left ventricular hypertrophy, age over 60 years, smoking, hypertension, or dyslipidemia).69 Patients with known cardiac disease and those with abnormal screening tests should undergo further evaluation with coronary computed tomography angiography. The European Association for the Study of the Liver guidelines recommend that all liver transplant candidates undergo electrocardiography and echocardiography and that patients with multiple risk factors or those older than 50 years undergo cardiopulmonary exercise testing to identify asymptomatic CAD.70 Most guidelines focus on evaluating CAD in liver transplant candidates, with less emphasis on patients with advanced cirrhosis. High-risk cirrhotic patients should undergo a careful cardiac evaluation to promptly identify the type of CAD and stratify risk, enabling the formulation of appropriate management strategies that could reduce overall and cardiac-related mortality.
There is a mutual interaction between liver cirrhosis and CAD. Evidence suggests that CAD may be more severe in cirrhotic individuals compared to non-cirrhotic individuals.9,42 Patients with cirrhosis often exhibit significantly more non-obstructive lesions, more extensive involvement of coronary vessels,9 longer atherosclerotic plaques, and higher total volumes of calcified or non-calcified plaques.42 Additionally, increasing levels of liver fibrosis and cirrhosis biomarkers are associated with more severe plaque and CAD. Liver fibrosis (LF) scores, including the Fibrosis-4 score and the non-alcoholic fatty liver disease fibrosis score, have been shown to predict the presence of coronary calcification.71 The non-alcoholic fatty liver disease fibrosis score is positively associated with the degree of coronary stenosis, while the Fibrosis-4 score correlates with the number of diseased coronary vessels.72 LF also negatively impacts the long-term prognosis of CAD patients. A prospective cohort study indicated that higher LF scores are associated with increased risks of all-cause and cardiovascular mortality among CAD patients.73 Taken together, advanced liver fibrosis appears to correlate with the severity of CAD, suggesting that these patients may require closer monitoring and screening for cardiovascular risk factors.
Notably, the association between liver cirrhosis and CAD may depend on the underlying etiology of the cirrhosis. Our study found a positive association between HCV cirrhosis and the presence of CAD. Similarly, previous studies have shown that HCV increases the risk of CAD.74 HCV directly and indirectly influences glucose and lipid metabolism, leading to a high prevalence of insulin resistance, steatosis, and diabetes mellitus.75 Additionally, the virus may have direct effects on the vessel wall.76 Our study also found that NASH was positively associated with CAD presence in liver cirrhosis. NASH is commonly associated with dyslipidemia, insulin resistance, and increased pro-inflammatory cytokines, all of which play important roles in the pathophysiology of atherosclerosis.77 Conversely, we did not find a significant association between alcoholic cirrhosis and CAD presence; however, other studies suggest that alcoholic cirrhosis is associated with both the occurrence and severity of CAD.18,20 Furthermore, coronary arteriosclerosis is particularly extensive in alcoholic cirrhosis. Patients with alcoholic cirrhosis had significantly higher median coronary artery calcium scores, which quantify coronary artery calcification, compared to those with non-alcoholic cirrhosis.78 Alcohol-related liver disease was also significantly associated with a coronary artery calcium score >300, indicating a high risk of cardiovascular events.79 This is likely due to excessive alcohol consumption, which is associated with increased levels of low-density lipoprotein and the expression of adhesion molecules.80
A meta-analysis conducted by Zhao et al. included five studies but only pooled the prevalence of CAD in cirrhosis.10 Another meta-analysis by Ungprasert et al. included four studies and reported an increased risk of CAD in PBC patients,81 a finding not confirmed by our study. In comparison, our study has several strengths. First, to our knowledge, we are the first to systematically report the incidence of CAD, the association between CAD and liver cirrhosis, and the factors related to the occurrence and presence of CAD in cirrhosis. Second, we included all types of cirrhosis rather than focusing solely on a single type, such as PBC. Third, we performed a comprehensive literature search using three major databases without language limitations to maximize the inclusion of epidemiological studies on CAD in liver cirrhosis. Fourth, we categorized CAD into ACS and MI to explore the effects of cirrhosis on different types of CAD.
Our study also has some limitations. First, significant heterogeneity remains in our meta-analyses, which necessitates cautious interpretation of our findings. Most of the included studies were retrospective, some had small sample sizes, and they utilized various definitions and diagnostic approaches for cirrhosis and CAD, potentially introducing bias into the results. Due to substantial differences and publication bias among the included studies, the pooled results may not accurately reflect the true effect size, impacting the reliability of the results. Second, the number of relevant studies was limited, and some were of low quality, which compromised the reliability of our findings. Only five studies reported associations between specific etiologies and CAD, making our conclusions implausible. Large-scale, well-designed prospective cohort studies are necessary to support our findings in the future. Additionally, we were unable to obtain information on the use of antithrombotic drugs and antiviral drugs for HCV, which could affect CAD development. Meanwhile, we could not perform subgroup analyses according to the different stages of cirrhosis and were unable to explore the association of HBV and PBC—with significant effects on lipid metabolism—with CAD in cirrhotic patients. Third, two previous meta-analyses found significant associations between NASH and HCV with cardiovascular diseases,74,77 which appear to contradict our findings. This discrepancy may be attributed to the differences in the etiologies of cirrhosis evaluated. Our included studies featured patients with mixed etiologies of liver cirrhosis, suggesting that liver cirrhosis is related not only to NASH but also to other causes, such as HBV. Indeed, two other previous meta-analyses indicated that HBV infection does not increase the risk of CAD.82,83 Our subgroup analyses based on different etiologies of cirrhosis demonstrated that the prevalence of CAD in NASH-related cirrhosis was higher than in other etiologies. Additionally, we found that the presence of NASH in cirrhosis increased the risk of CAD by 1.59 times. Fourth, the term “primary biliary cirrhosis” has been replaced by “primary biliary cholangitis” in recent years.84 However, earlier studies specifically referred to primary biliary cirrhosis, and some patients without a definitive diagnosis of liver cirrhosis were also attributed to the primary biliary cirrhosis group, which raises the possibility of misclassification. Finally, potential competing events, such as liver-related deaths, may compromise the development of CAD during follow-up, thereby influencing the true estimates of CAD.