Role of Anti-Beta-1-Adrenergic Receptor Antibodies in Cardiac Dysfunction in Patients with Cirrhotic Cardiomyopathy

Cirrhotic cardiomyopathy (CCM) is a recognized complication of cirrhosis and is associated with poor outcomes, especially under challenges such as surgery/liver transplantation. However, the mechanism is not clear, and the treatment is not specific. The present study aimed to evaluate the role of anti-β1-adrenergic receptor antibodies (anti-β1-AR) in CCM. We enrolled 3 groups: healthy controls, cirrhotic patients without CCM, and patients with CCM. We found that the anti-β1-AR levels in the CCM group were significantly higher than that in the non-CCM group; anti-β1-AR was positively correlated to NT-proBNP, negatively correlated to left ventricular ejection fraction, fractional shortening ((r =  − 0.466, P < 0.05), and the ratio of peak early (E wave) and atrial (A wave) flow velocities (E/A (r =  − 0.475, P < 0.05) in CCM patients. Anti-β1-AR is a useful predictive biomarker for the presence of CCM and eventually may also have therapeutic implications. Clinical Trials Registration: Chinese Clinical Trials No. ChiCTR 2,000,037,730


Introduction
Cirrhotic cardiomyopathy (CCM) is a syndrome of cardiac dysfunction in patients with cirrhosis, in the absence of prior heart disease [1][2][3]. Because of the reduced ventricular afterload in cirrhosis due to peripheral vasodilation, thus "auto-treating" the patient, the cardiac dysfunction in patients with CCM is latent [1,4,5]. However, when challenged, the cardiac dysfunction becomes overt. One of the main pathogenic mechanisms is abnormalities of the β-adrenergic receptor (βAR) pathway [6].
It is well known that βAR plays a major role in cardiac contraction [7]. In the heart, βAR are subdivided into either β1-or β2-subtypes. More than 60 years ago, it was noticed that cardiac responsiveness of patients with cirrhosis to exogenous [8] infusions of catecholamines or endogenous production (i.e., caused by exercise) [9] is attenuated. However, the underlying mechanism of hyporesponsiveness was unclear then. Three decades ago, Lee and colleagues measured heart rate responsiveness to isoprenaline and myocardial βAR-binding characteristics in cirrhotic rats and found that compared with sham-operated controls, a significantly higher dose of isoprenaline was needed to raise basal heart rate by 50 beats/min (102 ± 19 vs. 28 ± 11 ng/kg) in cirrhotic rats. They also found that myocardial βAR density was significantly lower in cirrhotic rats and the dissociation constant was higher. Subtype analysis demonstrated that the decreased total βAR density in cirrhotic heart was entirely due to selective downregulation of the β1-AR subtype. Moreover, the βAR affinity for agonist was not altered. They concluded that βAR downregulation is responsible for the myocardial hyporesponsiveness to catecholamines in the cirrhotic rat heart. However, the exact mechanism underlying the βAR downregulation in cirrhosis remains unclear.
Autoantibodies attack self cells/organs and cause diseases. Antibodies to glomerular basement membrane cause immune-complex crescentic nephritis [10]. Anti-β1-AR binds to and constitutively stimulates the β1-AR to cause βAR desensitization and downregulation [11]. However, whether anti-β1-AR plays a role in the decreased density of β1-AR and CCM is not clear. The present pilot study therefore aimed to investigate the hypothesis that β1-AR downregulation is due to the presence of increased activity of β1-AR antibodies in patients with CCM.

Patients
The protocol was approved by the Beijing Youan Hospital ethics committee and conformed to the guidelines of the Helsinki Declaration. All patients provided written informed consent. The study was registered with the Chinese Clinical Trials No. ChiCTR 2,000,037,730.
From January 2016 to December 2018, 352 patients with cirrhosis were consecutively screened in Youan Hospital. The diagnosis of cirrhosis was based on clinical features, and laboratory, ultrasonography, and histology examinations. The inclusion criteria were age 18-65 and presence of cirrhosis. Patients with the following conditions that may impact cardiac function were excluded: fever, systemic inflammatory response syndrome, severe anemia (hemoglobin < 60 g/L), alcoholic liver disease, cardiovascular diseases (hypertension, coronary artery disease, rheumatic heart disease, pericarditis, myocarditis, and valvular heart disease), thyroid diseases, renal dysfunction, acute or chronic respiratory diseases, diabetes mellitus, and malignance. Patients who had taken drugs such as β-blockers, calcium channel blockers, and angiotensin-converting enzyme inhibitors within the past week were also excluded (Fig. 1). The diagnosis of CCM was based on the Montreal 2005 World Congress of Gastroenterology criteria [12]. The patients were divided into 3 groups: CCM group, cirrhotic non-CCM group, and age and sex-matched healthy volunteer group (n = 20).
Amongst the 352 screened cases, 125 patients were considered for enrolment, whereas 227 patients were excluded due to the presence of alcoholic hepatitis, coronary artery disease, hypertension, and hepatocellular carcinoma (Fig. 1). Of the 125 patients, 19 had CCM (CCM group), and 106 had non-CCM. After propensity matching for sex, age, and Child-Pugh score, 76 of the 106 cases were selected for the non-CCM group. The average age in the CCM group was 58.4 ± 9.1 years; 12 were males and 7 were females. The etiologies were HCV-cirrhosis (n = 9), HBV-cirrhosis (n = 7), and primary biliary cholangitis (PBC, n = 3). We also enrolled 20 sex-and age-matched healthy volunteers.
The demographics, physical exams, electrocardiography (ECG), echocardiography, and routine blood tests were recorded. Echocardiography was performed by an experienced cardiologist (ID) on a iE33 ultrasonography (Philips Medical Systems, Best, the Netherlands) with structural and functional parameters measured according to current guidelines. Child-Pugh score, Model for End-Stage Liver Disease (MELD) score, and MELD-Na score were used to evaluate the severity of the cirrhosis. Compensated cirrhosis was defined as liver is scarred but still able to perform most its basic functions, whereas decompensated was characterized by the presence of ascites, jaundice, variceal hemorrhage, or hepatic encephalopathy.
Natriuretic drug history was carefully evaluated to avoid the effect of these agents on clinical variables and ultrasonography. Patients were examined while on a low sodium diet and, where applicable, at least 48 h after paracentesis. The serum was separated for anti-β1-AR, NT-proBNP, and troponin I detection. The blood samples were also sent to relevant laboratories to test routine biochemistry, liver and kidney function, electrolytes, and coagulation function.

Diagnosis of Cirrhotic Cardiomyopathy
According to the Montreal 2005 diagnostic criteria [12], systolic dysfunction was defined as ejection fraction < 55%, and diastolic dysfunction was defined as two of the three indices: the ratio of peak early (E wave) and atrial (A wave) flow velocities (E/A; early/late diastolic filling velocities) ratio (age corrected) < 1.0, prolonged mitral deceleration time (DT; > 200 ms), and prolonged isovolumic relaxation time (> 80 ms). The supportive criteria were corrected QT interval (QTc) > 0.440 s, increased B-type natriuretic peptide (BNP) and prob-type natriuretic peptide (pro BNP), and increased troponin I.

Serum Anti-β1-AR Assay
The serum from all the participants were saved at − 80 °C. The serum level of anti-β1-AR was measured by enzymelinked immunosorbent assay (ELISA). The kit, named human anti β1 adrenergic receptor antibody (Anti-β1AR) ELISA kit, was purchased from Blue Gene for Life Science (E01A0813, Shanghai, China). The samples and reagents in the ELISA kit were recovered to room temperature; each sample was tested in duplicate. Standards or samples (100 µl) were incubated with 50 µl horseradish peroxidase at 37 °C for 1 h. After washing for 5 times, the substrates A and B (50 µl) were added, and the plate was incubated in dark at 37 °C for 15 min. The stop solution was added, and the plate was read at 450-nm wavelength.

Serum NT-proBNP, Troponin I Assay
The serum NT-proBNP and troponin I were measured by enzyme-linked immunosorbent assay (ELISA). The ELISA kits were purchased from Abcam, (ab263877, ab200016, Cambridge, UK.). The samples and reagents in the ELISA kits were recovered to room temperature; each sample was tested in duplicate. All the reagents, samples, and standards were prepared as instructed. First, 50 µL standards or samples were added to appropriate wells; then, 50 µL antibody cocktail were added to all wells and incubated at room temperature for 1 h. After washing for three times with 350 µL 1X wash buffer. 3,3′,5,5′-Tetramethylbenzidine (TMB, 100 µl) was added and incubated for 10 min, the stop solution (100 µl) was added, and the plate was read at 450-nm wavelength.

Statistical Analysis
Statistical analysis was performed using SPSS 16.0 (SPSS Inc, Chicago, IL, USA). To avoid potential confounders, we did propensity scores matching for sex, age, and Child-Pugh score between CCM and non-CCM groups. Normal distribution data are expressed as mean ± standard deviation, and skewed data are expressed as median (inter quartile range, IQR). Comparisons between different groups were analyzed by a Kruskal-Wallis or Mann-Whitney test. Spearman correlations were applied. Logistic regression was performed with the backward method for multivariable analysis (the entry probability for stepwise: P < 0.05). The diagnostic performance of an assay was assessed by the analysis of receiver-operating characteristic (ROC) curve. The assessment of sensitivity and specificity was determined using selected cutoff value. P < 0.05 was considered statistically significant.

Cardiac Function Parameters
NT-proBNP in the CCM group was significantly higher than that in the non-CCM group (213 ± 56 vs 159 ± 112 pg/ mL, P < 0.05). There were no statistically significant differences in other laboratory tests or cirrhosis complications between the two groups (Table 1). In the CCM group, the left ventricular ejection fraction (LVEF), fractional shortening (FS), and E/A ratio were significantly lower, and left ventricular posterior wall minor motion amplitude (LVPWA) 1 3 was significantly higher than those in the non-CCM group (P < 0.05) ( Table 2).

Serum Anti-βAR
The anti-β1-AR levels in the CCM group were significantly higher than that in non-CCM group (0.97 ± 0.37 vs 0.74 ± 0.37 ng/mL, P < 0.05); this value was 0.12 ± 0.06 ng/ mL in healthy controls. We also divided all patients according to the severity of cirrhosis and found that anti-β1-AR levels were not significantly different between compensated and decompensated groups (P > 0.05) (Fig. 2).

Relationship Between Serum Anti-β1-AR and Liver Function, Clinical Presentation, or Portal Hypertension
In the CCM group, serum anti-β1-AR had no signifi-  Table 4).

Diagnostic Value of Serum Anti-β1-AR CCM
The AUROC of serum anti-β1-AR on CCM was 0.678 (95% CI 0.515-0.768). If the cutoff value was set at 0.669 ng/mL, the sensitivity was 89.5%, and the specificity was 57% for diagnosis of CCM.

Fig. 2 Serum anti-β1-AR levels in patients with and without CCM and in compensated or decompensated patients. A
Serum anti-β1-AR levels were higher in CCM than the non-CCM group. B There was no difference of serum anti-β1-AR levels in decompensated as compared to compensated patients 1 3 shortening) and diastolic function (E/A ratio). It is thought that NT-proBNP correlates with left ventricular systolic and diastolic dysfunction in several noncirrhotic cardiac conditions [13]. NT-proBNP also correlates with cardiac dysfunction in cirrhotic patients [14]. Our data revealed that anti-β1-AR was correlated with the serum level of NT-proBNP in CCM patients. All these data suggested that anti-β1-AR is a useful biomarker of CCM. Patients with CCM show overt cardiac dysfunction when challenged [12]. Lee and coworkers demonstrated that the desensitization of the cirrhotic rat heart to βAR stimulation is caused, at least in part, by downregulation of cardiac βAR density. It thus seems reasonable to hypothesize that (1) the long-term activation of the sympathetic nervous system induces downregulation of βAR and (2) alteration in βAR metabolism in the cirrhotic environment contributes to the defect in β-adrenergic system function [15].
In that respect, there is no previous study on the relationship between anti-βAR antibody and βAR status in patients with CCM. However, in patients with dilated cardiomyopathy, Patel et al. demonstrated that the anti-β1-AR binds to and constitutively stimulates the β1-AR to cause βAR desensitization and downregulation [11]. Limas et al. found that the serum from patients with idiopathic dilated cardiomyopathic inhibits the binding of [3H]dihydroalprenolol to rat cardiac membranes [16], and the inhibition could be prevented by preincubating the sera with anti-human IgG.
Using peptide corresponding to the second extracellular loop of the β 1 -AR to immunize rats, the animals developed anti-β 1 -AR. These rats developed dilated cardiomyopathy. Furthermore, after infusing the serum from antibody-positive rats to healthy, antibody-negative rats, the latter also developed dilated cardiomyopathy [17]. These studies strongly suggest a causal role for anti-β 1 -AR in the development of dilated cardiomyopathy.
Autoantibodies attack self cells/organs and play a pathogenic role in several conditions. Myasthenia gravis [18], Graves' disease [19], and insulin-resistant diabetes [20] are well-documented examples. Besides the desensitization of the β 1 -AR, anti-β 1 -AR also affect cardiac systolic function which indicates that anti-β 1 -AR is a cardiodepressant in antibody-positive patients. Stork et al. found that the antiβ 1 -AR increases all-cause and cardiovascular mortality risk in patients with dilated cardiomyopathy [21]. Although β 1 -ARs are the main stimulatory receptors for cardiac contractility, they are not the only type. Inhibitory M2-muscarinic receptors (anti-M2R) are also increased in different cardiac diseases. Serum anti-β1-AR and anti-M2R negative correlate with LVEF and LV Fractional Shortening [22,23].
Autoantibodies are common in patients with cirrhosis. Immuno-regulatory dysfunction contributes to liver diseases such as autoimmune hepatitis, primary biliary cholangitis, and primary sclerosing cholangitis. Furthermore, patients with cirrhosis of other etiologies also have been documented with autoantibody-related conditions. Gilman et al. found that autoantibodies, such as antinuclear antibody, antimitochondrial antibody, antismooth muscle antibody, and/or antiliver kidney microsomal antibody, are positive in some patients with hepatitis C infection [24]. Karpova et al. demonstrated that increases of autoantibodies (anti-intestinal antigens) are associated with degenerative changes in the intestinal mucosa in patients with cirrhosis [25,26]. Papp et al. [27] also demonstrated high prevalence of IgA class antineutrophil cytoplasmic antibodies (ANCA) that is associated with increased risk of bacterial infection in patients with cirrhosis. Another study found that the function of follicular T helper cells is impaired in patients with advanced cirrhosis, which jeopardizes antibody responses against microbial pathogens [28]. In summary, immune dysfunction in cirrhotic patients plays an important role in cirrhosis-related complications.
The mechanism of the anti-β1AR production in our study is not clear. One possibility is that cardiomyocyte injury releases the β1-AR from the cell membrane to the circulation, which acts as an autoantigen stimulating the immune system to form anti-β1-AR. Our previous study demonstrated that myocardial apoptosis is increased in a rat model of CCM [29]. We believe that anti-β1-AR and cardiac injury thus comprise a "vicious circle." Blocking or weakening this circle may be therapeutically beneficial in CCM. Muller and colleagues [26,30] used immunoglobulin adsorption to decrease the amount of anti-β1-AR in dilated cardiomyopathy. They achieved a therapeutic effect with an improvement of New York Heart Association (NYHA) class, decreased markers of oxidative stress, and reduction in hospital stay. Labovsky et al. demonstrated  that immunoglobulin adsorption significantly decreases the serum β1-AR in patients with chronic Chagas heart disease [31]. Another alternative method of reducing autoantibodies is plasma exchange (PEx); Gesinde and colleagues used PEx to decrease the anti-β1-AR levels in a patient with dilated cardiomyopathy, leading to significant improvement of the cardiac function [32]. These studies raise the intriguing possibility that immunoglobulin adsorption or PEx may be a new strategy for the treatment of CCM patients with high levels of anti-β1-AR antibodies.
Besides the correlation between anti-β 1 -AR and cardiac function, we also found that anti-β 1 -AR was negatively correlated with portal hypertension, but not with the severity of liver dysfunction. The negative correlation with portal and splenic vein diameters defies easy explanation, but we speculate that this reflects some type of poorly understood relationship between peripheral blood vessels which contain abundant β-AR and factors that lead to dilatation or structural changes in the vessel wall or lumen.
The exact prevalence of CCM in cirrhotic patients remains unclear, but recent estimates place it around 60% [33,34]. In our study, it was only 15%. The explanation for the lower prevalence in our study is likely patient selection. Most of our patients were Child-Pugh class A, as decompensated class B or C patients generally were excluded due to infection, bleeding, or other complications. Furthermore, it is also reported that immunoglobulin production decreases in decompensated class B or C patients and causes a state of functional immunosuppression, which lead to complications of cirrhosis [28]. Most evidence suggests that the prevalence of CCM correlates with the extent of liver failure in cirrhosis [35].
Since there is no specific biomarker for the diagnosis of CCM, we evaluated the predictive value of anti-β 1 -AR for CCM using ROC curves and found that the level of antiβ 1 AR showed a significant predictive value for CCM. Since this study is a cross-sectional study, we did not follow up the long-term outcomes of patients. It has been reported that anti-β 1 -AR levels are higher in chronic heart failure and serve as an independent predictor for sudden cardiac death [36]. The association between the levels of anti-β 1 -AR and patient survival bears further study.
This study has limitations. Firstly, it is a pilot study with a limited sample size; these results should be validated by larger studies with longer follow-up.
Another limitation is that we used the old 2005 Montreal diagnostic criteria [12]; the new proposed 2019 criteria were published after we finished this study [3], whether the new criteria are superior to the old remains unknown. A preliminary study compared these two criteria in cirrhotic patients waiting for liver transplantation [37]. They found that 8% of their cohort had CCM according to the new criteria, while the percentage was 74% according to the old criteria [37]. A recent paper by Razpotnik and colleagues compared the two diagnostic criteria in 122 patients with cirrhosis [34]. They found that the overall prevalence of cirrhotic cardiomyopathy was significantly higher with the old (67.2%), compared to the new (19.7%). Since approximately 7-21% of deaths after liver transplantation result from cardiovascular complications [12], it is possible that the new CCC criteria may be too restrictive to apply to advancedcirrhotic patients. The diagnostic criteria thus need to be standardized and validated in further studies to determine which better correlate with relevant outcomes such as mortality and cardiovascular morbidity [38]. Until such time as the issue of which criteria are best is resolved, which will take several years, we believe that the present study based on the Montreal criteria remains valuable in the cirrhotic cardiomyopathy literature. The comparisons of anti-β 1 -AR levels in CCM patients diagnosed either with the new or old criteria and the prognostic role of anti-β 1 -AR in patients with CCM merit further investigation.
In conclusion, anti-β 1 -AR was increased in CCM patients. Anti-β 1 -AR may play an important role in the desensitization of β1-AR in patients with CCM. Furthermore, the presence of anti-β 1 AR may increase cardiovascular complications in cirrhotic patients. Anti-β 1 -AR are also valuable as a potential biomarker to diagnose CCM. Using immunological methods to reduce/diminish anti-β 1 -AR levels may represent a promising new therapeutic strategy for patients with cirrhotic cardiomyopathy.