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
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaChristian Doppler Laboratory for Portal Hypertension and Liver Fibrosis, Medical University of Vienna, Vienna, AustriaLudwig Boltzmann Institute for Rare and Undiagnosed Diseases (LBI-RUD), Vienna, AustriaCeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, Vienna, Austria
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, Austria
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, Austria
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, Austria
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, Austria
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaChristian Doppler Laboratory for Portal Hypertension and Liver Fibrosis, Medical University of Vienna, Vienna, Austria
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, Austria
Corresponding author. Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Waehringer Guertel 43 18-20, 1090 Vienna, Austria. P: +43 1 40400 47440, F: +43 1 40400 47350 M.
Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaVienna Hepatic Hemodynamic Lab, Division of Gastroenterology and Hepatology, Department of Medicine III, Medical University of Vienna, Vienna, AustriaChristian Doppler Laboratory for Portal Hypertension and Liver Fibrosis, Medical University of Vienna, Vienna, AustriaLudwig Boltzmann Institute for Rare and Undiagnosed Diseases (LBI-RUD), Vienna, AustriaCeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, Vienna, Austria
Complement factors, immunoglobulins, and acute-phase proteins are dysregulated in patients with advanced chronic liver disease (ACLD), correlate with disease severity, and indicate alterations of innate and adaptive immunity.
-
Low complement C3c levels independently predicted decompensation or liver-related death of patients with ACLD.
-
High IgG-1 levels independently predicted the incidence of infections in patients with ACLD.
Abstract
Background
Cirrhosis-associated immune dysfunction (CAID) affects both innate and adaptive immunity. This study investigated the complement system, immunoglobulins, and acute-phase-proteins and their prognostic relevance in patients with advanced chronic liver disease (ACLD).
Methods
Patients with ACLD (hepatic venous pressure gradient [HVPG] ≥6mmHg) but without acute decompensation/infections were characterized by HVPG and by clinical EASL stages: compensated (cACLD;S0-2) vs. decompensated ACLD (dACLD) with previous variceal bleeding (S3), non-bleeding decompensation (S4), or further decompensation (S5). Complement factors (C3c,C4,CH50), immunoglobulins (IgA,IgM,IgG,IgG1-4), acute-phase-proteins and systemic inflammation biomarkers (WBC,CRP,IL-6,procalcitonin) were measured.
Results
Two-hundred-and-forty-five patients (median MELD:11(9-15), median HVPG: 17(12-21)mmHg) were included with 150(61%) presenting dACLD. Complement levels and activity significantly decreased in dACLD substages S4 and S5 (p<0.001). Total IgA/IgM/IgG and IgG1-4 subtype levels increased in patients with dACLD (all p<0.05). Complement and immunoglobulin levels correlated negatively and positively, respectively, with systemic inflammation (all p<0.05). High IgG-1 (aHR per 100mg/dL: 1.12, 1.04-1.19, p=0.002) and IL-6 (aHR: 1.03, 1.00-1.05, p=0.023) levels predicted the development of infections during follow-up. High IgA (stratified by median; log-rank p<0.001), high IgG1 (log-rank p=0.043) and low C3c (log-rank p=0.003) indicated a higher risk of first/further decompensation or liver-related death (composite endpoint). Next to HVPG and IL-6, low C3c (aHR per mg/dL: 0.99, 0.97-0.99, p=0.040) remained independently associated with the composite endpoint on multivariate Cox regression analysis.
Conclusion
Complement levels and immunoglobulins may serve as surrogates of CAID and associate with cirrhosis severity and systemic inflammation. Low complement C3c predicted decompensation and liver-related death, while high IgG-1 indicated an increased risk for infections.
Lay summary
Patients with cirrhosis are at increased risk for infections, which worsen their prognosis. We found a significant dysregulation of several essential components of the immune system that was linked to disease severity and indicated a risk for infections and other complications. Simple blood tests identify patients at particularly high risk, who may be candidates for preventive measures.
BeSi has received travel support from AbbVie and Gilead and was supported by an International Research scholar by Gilead Sciences awarded to TR.
DB has received travel support from AbbVie and Gilead, speaker fees from AbbVie and Siemens, as well as grant support from Philips, Siemens, and Gilead.
BeSc received travel support from AbbVie, Ipsen and Gilead.
MT received grant support from Albireo, Alnylam, Cymabay, Falk, Gilead, Intercept, MSD, Takeda and Ultragenyx, honoraria for consulting from Albireo, Boehringer Ingelheim, BiomX, Boehringer Ingelheim, Falk, Genfit, Gilead, Intercept, Janssen, Merck, MSD, Novartis, Phenex, Regulus and Shire, speaker fees from BMS, Falk, Gilead, Intercept and MSD, as well as travel support from Abbvie, Falk, Gilead and Intercept.
MM served as a speaker and/or consultant and/or advisory board member for AbbVie, Gilead, Collective Acumen, and W. L. Gore & Associates and received travel support from AbbVie, and Gilead.
TR received grant support from Abbvie, Boehringer-Ingelheim, Gilead, MSD, Philips Healthcare, Gore; speaking honoraria from Abbvie, Gilead, Gore, Intercept, Roche, MSD; consulting/advisory board fee from Abbvie, Bayer, Boehringer-Ingelheim, Gilead, Intercept, MSD, Siemens; and travel support from Abbvie, Boehringer-Ingelheim, Gilead and Roche.
RP, AFS, BH, LH, MJ and RM declare no conflict of interest.
Financial support statement
No funding was received for this study.
Authors’ contributions
Authors contributed either to study concept and design (BeSi, MM, TR) and/or data acquisition (all authors), analysis (BeSi, MM, TR) or interpretation (all authors). BeSi and TR drafted the manuscript, which was critically revised by all other authors.
Data availability statement
Data are available at reasonable request to the corresponding author.
Introduction
Cirrhosis-associated immune dysfunction (CAID) has been attributed a central pathophysiological role in advanced chronic liver disease (ACLD) and is characterized by systemic inflammation and impaired immunocompetence [
]. The manifestation of CAID may range from increased systemic inflammation in both patients without clinical evidence of infection and patients with acute-on-chronic liver failure (ACLF) [
]. In either situation, however, these immunological alterations seem to hold clinical significance, as previous studies from our and other centers demonstrated that systemic inflammation levels, as reflected by inflammatory cytokines and acute-phase proteins (APP), predict disease progression and/or mortality in both clinically stable patients without overt infection [
Von Willebrand factor indicates bacterial translocation, inflammation, and procoagulant imbalance and predicts complications independently of portal hypertension severity.
]. Furthermore, the increased incidence of bacterial infections and their impact on prognosis in patients with ACLD underlines the profound clinical implications of CAID [
]. For example, impairment of the gut-liver axis - involving intestinal dysbiosis and increased intestinal permeability - causes constant exposure of the immune system to bacterial pathogens and pathogen-associated molecular patterns (PAMPs) [
]. Furthermore, liver injury may promote inflammatory reactions in the absence of pathogenic triggers (‘sterile’ inflammation), that are often mediated by danger-associated molecular patterns (DAMPs) [
]. These conditions activate both innate and adaptive immune responses that may, on the one hand, promote systemic inflammation, and on the other hand, lead to exhaustion of the immune system [
As mentioned above, inflammatory cytokines such as interleukin-6 (IL-6) and APPs such as C-reactive protein (CRP) were previously linked to disease severity and prognosis in ACLD [
]. Previous studies have also demonstrated a relationship between complement factors (part of the innate immune system) and disease severity as well as the incidence of infections and mortality in patients with ACLD [
]. Furthermore, circulating immunoglobulin levels (important elements of adaptive immunity) exhibited an association with the presence and/or severity of fibrosis/cirrhosis in patients with chronic liver disease [
]. Nevertheless, the relationship between these immune components and hepatic dysfunction, portal hypertension, systemic inflammation, and complications of ACLD remains incompletely understood.
The present study aimed to simultaneously characterize circulating levels of complement and immunoglobulin levels and assess their dynamics across different ACLD stages, as well as their link to systemic inflammation (reflected by APPs) and disease prognosis in a well-characterized cohort of patients with ACLD undergoing hepatic venous pressure gradient (HVPG) measurement.
Patients and methods
Study design, patient selection, and clinical characterization
Patients undergoing HVPG measurement were recruited into the Vienna Cirrhosis Study (VICIS; NCT03267615) between 02/2019 and 09/2021. The presence of ACLD was confirmed by an HVPG ≥6 mmHg that also defines the presence of sinusoidal portal hypertension. Patients with portal hypertension due to vascular liver disease, active malignancies (including hepatocellular carcinoma), previous liver transplantation, acute hepatic decompensation, or infection at the timepoint of HVPG measurement were excluded by review of prospectively collected clinical data. After exclusion of patients without information on laboratory parameters that are relevant to this study, the final study cohort comprised 245 patients with ACLD (Supplementary figure-1). ACLD was staged according to D’Amico et al. and the European Association for the Study of the Liver (EASL) guidelines [
]. Patients with compensated ACLD (cACLD) were subsumed as stage 0-2 (S0-S2), while patients with decompensated ACLD (dACLD) were divided into patients with previous variceal bleeding (S3), patients with one non-bleeding decompensation event (S4), and patients with two or more decompensation events (S5).
Measurement of hepatic venous pressure gradient
Measurement of HVPG followed a standard operating procedure at the Vienna Hepatic Hemodynamic Lab, Medical University of Vienna, Austria, as described in previous publications [
Reiberger, T., Schwabl, P., Trauner, M., Peck-Radosavljevic, M. and Mandorfer, M., Measurement of the Hepatic Venous Pressure Gradient and Transjugular Liver Biopsy. J Vis Exp, 2020(160).
]. Briefly, after placing a catheter introducer sheath into the right internal jugular vein, the catheter tip was advanced into a large hepatic vein. After affirmation of the correct positioning of the catheter tip, at least three measurements of free and wedged hepatic vein pressures were performed. The mean of the differences between these pressures determined HVPG.
Biomarker measurements
Biomarkers were measured in blood samples taken from the catheter introducer sheath at the timepoint of HVPG measurement. Measurements were performed without prior storage at the Department of Laboratory Medicine, Medical University of Vienna, Austria, according to standardized and ISO-certified procedures. Laboratory personnel was blinded to clinical and hemodynamic data. Quantitative determination of human complement factors C3c (normal range 90-180 mg/dL) and C4 (normal range 10-40 mg/dL), immunoglobulin (Ig) subtypes IgG (normal range 700-1600 mg/dL), IgM (normal range 40-230 mg/dL), IgA (normal range 70-400 mg/dL), and IgG subclasses IgG1 (normal range 405-1011 mg/dL), IgG2 (normal range 169-786 mg/dL), IgG3 (normal range 11.0-85 mg/dL), and IgG4 (normal range 3.0-201 mg/dL), as well as the APPs alpha-1-acid glycoprotein (AGP; normal range 50-120 mg/dL), serum amyloid A (SAA; upper limit of normal 6.4 mg/L), and alpha-2-macroglobulin (A2M; normal range 130-300 mg/dL) were performed by nephelometry on a BNII System using N antisera (Siemens Healthcare Diagnostics, Vienna, Austria). Complement activity was assessed in serum by a CH50 in vitro liposome immunoassay (normal range 31,6-57,6 U/mL). Tissue transglutaminase IgA antibodies (TTG-IgA; upper limit of normal <10 U/mL) levels were determined in serum by an indirect ELISA (Orgentec Diagnostika, Mainz, Germany).
Statistical analysis
Statistical analyses were performed using IBM SPSS Statistics 27 (IBM, Armonk, NY, USA) or GraphPad Prism 9 (GraphPad Software, La Jolla, CA, USA). Continuous variables are presented as mean (± standard error of the mean) or median (interquartile range) depending on the normal distribution of variables, as assessed by D’Agostino & Pearson and Shapiro-Wilk tests. Comparison of continuous variables was performed by Student’s t-test, Mann-Whitney U test, Kruskal-Wallis test, or analysis of variance (ANOVA), as appropriate. Tukey’s or Dunn’s multiple comparisons tests were applied. Categorical variables are presented as absolute and relative frequencies. Comparison of categorical variables was performed using Chi-squared or Fisher’s Exact test. Pearson or Spearman correlation coefficients (95% confidence interval) were determined to assess the correlation between continuous variables. Kaplan-Meier curves were used to illustrate the incidence of the composite endpoint “first/further decompensation or liver-related death”, as well as the incidence of infections, at 24 months in patients stratified by serum biomarker levels. The composite endpoint was defined as the incidence and/or worsening of ascites, variceal bleeding, hepatic encephalopathy, or liver-related death during follow-up (see Supplementary material). Log-rank test was applied to compare the incidence of the respective endpoints between groups. Univariate and multivariate Cox proportional hazard models were used to determine the prognostic value of complement factors and immunoglobulin levels for these clinical events. The threshold for statistical significance was set at a two-sided p-value <0.05 for all analyses.
Compliance with ethical standards
The study was conducted following the principles of the Declaration of Helsinki and its latest amendments and approved by the local ethics committee (EK 1262/2017). Patients gave written informed consent for liver vein catheterization and participation in the VICIS study.
Results
Patient characteristics
Patients included in the study had a median age of 57 (50-66) years and displayed a median HVPG of 17 (12-21) mmHg, a median MELD of 11 (9-15) points. Most patients (n=213, 87%) had clinically significant portal hypertension (CSPH; HVPG ≥10 mmHg). Ninety-five patients (39%) had cACLD (S0-S2) and among patients with dACLD, 8 (3%) patients were in S3, 81 (33%) in S4, and 61 (25%) in S5. Alcohol-related liver disease (ALD) and viral hepatitis represented the main etiologies of ACLD in the study cohort. Consistent with previous results [
], an increase in HVPG, MELD, systemic inflammation markers CRP and IL-6, as well as the prevalence of ALD were observed in patients with dACLD and the respective dACLD subgroups. Twenty-three (9%) patients reported intake of antibiotics as therapy for hepatic encephalopathy (HE) or prophylaxis for spontaneous bacterial peritonitis (SBP) at the timepoint of HVPG measurement: 20 patients were on rifaximin, 2 on both rifaximin and norfloxacin, and 1 on norfloxacin. Patient characteristics are summarized in Table-1 and Supplementary Table-1.
Table-1Patient characteristics.
Compensated ACLD (n=95)
Decompensated ACLD (n=150)
P-value
Stage 0-2
Stage 3 Bleeding (n=8)
Stage 4 Non-bleeding decompensation (n=81)
Stage 5 Further decompensation (n=61)
Age (years)
58 (47-66)
58 (54-61)
56 (49-66)
58 (48-68)
0.863
Sex (M, %)
64 (67)
6 (75)
52 (64)
39 (64)
0.899
Etiology (n, %)
-
ALD
30 (32)
3 (38)
51 (63)
40 (66)
<0.001
-
Viral
24 (25)
4 (50)
4 (5)
3 (5)
-
ALD + Viral
3 (3)
0 (0)
6 (7)
8 (13)
-
NASH
17 (18)
0 (0)
0 (0)
1 (2)
-
Cholestatic
12 (13)
0 (0)
5 (6)
2 (3)
-
Other
9 (10)
1 (12)
15 (19)
7 (12)
HVPG (mmHg)
12 (9-18)
18 (12-23)
19 (15-22)
19 (15-23)
<0.001
MELD (points)
9 (8-11)
10 (9-12)
12 (10-15)
14 (11-17)
<0.001
Bilirubin (mg/dL)
0.87 (0.67-1.35)
1.25 (0.81-1.39)
1.55 (0.97-2.39)
1.59 (0.92-2.34)
<0.001
Albumin (g/L)
39.9 (36.7-42.4)
39.4 (37.8-42.1)
35.2 (31.1-38.4)
34.4 (31.3-37.0)
<0.001
ASAT (U/L)
38 (27-56)
34 (23-49)
44 (31-62)
42 (29-54)
0.275
CRP (mg/dL)
0.16 (0.07-0.36)
0.21 (0.09-0.98)
0.42 (0.17-0.91)
0.45 (0.24-1.16)
<0.001
IL-6 (pg/mL)
5.45 (3.34-9.08)
5.72 (4.82-13.8)
11.2 (6.69-18.2)
15.3 (9.77-27.8)
<0.001
WBC (G/L)
4.82 (3.31-6.17)
3.26 (2.64-4.64)
4.86 (3.57-6.32)
4.69 (3.75-6.10)
0.134
PCT (ng/mL)
0.07 (0.05-0.11)
0.06 (0.04-0.18)
0.11 (0.07-0.17)
0.13 (0.07-0.19)
<0.001
LBP (μg/mL)
6.45 (4.91-8.48)
7.60 (6.18-11.2)
6.66 (4.83-9.39)
6.31 (5.22-8.23)
0.521
C3c (mg/dL)
101 (86.9-114)
103 (88.8-116)
84.0 (67.7-102)
77.5 (59.8-91.6)
<0.001
C4 (mg/dL)
14.8 (10.3-19.9)
16.1 (10.6-17.2)
13.7 (10.3-17.0)
10.3 (8.00-14.3)
<0.001
CH50 (U/mL)
57.9 (48.9-60.0)
58.1 (44.8-60.0)
51.1 (36.5-59.7)
44.5 (27.7-54.4)
<0.001
IgA (mg/dL)
274 (199-424)
284 (242-667)
456 (300-618)
490 (298-674)
<0.001
IgM (mg/dL)
118 (68.2-175)
151 (62.6-196)
159 (92.3-237)
145 (87.9-230)
0.036
IgG (mg/dL)
1340 (1060-1670)
1510 (1360-1705)
1560 (1275-1970)
1510 (1220-1845)
0.006
IgG1 (mg/dL)
882 (684-1100)
1145 (775-1318)
1050 (786-1380)
976 (776-1225)
0.062
IgG2 (mg/dL)
302 (220-412)
282 (172-534)
341 (262-536)
354 (247-478)
0.094
IgG3 (mg/dL)
47.8 (31.7-72.1)
51.2 (35.4-67.8)
56.2 (37.7-86.1)
59.2 (37.2-88.9)
0.066
IgG4 (mg/dL)
44.8 (21.6-94.4)
65.7 (37.2-118)
61.5 (26.4-143)
57.0 (33.1-108)
0.174
AGP (mg/dL)
51.3 (40.7-62.3)
52.3 (39.3-58.5)
47.0 (32.3-62.0)
46.9 (34.1-61.1)
0.349
SAA (mg/L)
4.16 (4.16-7.01)
4.16 (4.16-4.63)
4.16 (4.16-7.27)
4.16 (4.16-8.79)
0.818
A2M (mg/dL)
286 (228-331)
360 (285-379)
231 (198-263)
205 (159-260)
<0.001
Statistical analysis: Kruskal-Wallis test was applied to compare continous variables between groups. P-values <0.05 are indicated in bold. Abbreviations: (A2M) alpha-2-macroglobulin; (AGP) alpha-1-acid glycoprotein; (ALD) alcohol-related liver disease; (c/dACLD) compensated/decompensated advanced chronic liver disease; (C3c) complement C3 component; (CRP) C-reactive protein; (HVPG) hepatic venous pressure gradient; (Ig) immunoglobulin; (IL-6) interleukin-6; (LBP) lipopolysaccharide binding protein; (M) male sex; (MELD) Model of End Stage Liver Disease; (NASH) non-alcoholic steatohepatitis; (PCT) procalcitonin; (SAA) serum amyloid A.
Complement component C3c and C4 levels decreased in disease stages S4 and S5 as compared to patients with S0-S2: C3c levels were 101 (86.9-114) in S0-S2, 84.0 (67.7-102) mg/dL in S4, and 77.5 (59.8-91.6) in S5. C4 levels were 14.8 (10.3-19.9) mg/dL in S0-S2, 13.7 (10.3-17.0) mg/dL in S4, and 10.3 (8.00-14.3) mg/dL in S5 (all p<0.001). Concordantly, CH50 activity – reflecting the haemolytic activity of the complement system in vitro – decreased in S4 and S5 as compared to S0-S2: CH50 was 57.9 (48.9-60.0) mg/dL in S0-S2, 51.1 (36.5-59.7) mg/dL in S4, and 44.5 (27.7-54.4) mg/dL in S5 (p<0.001). The decrease of complement (activity) levels was even more pronounced in S5 as compared to S4, indicating that these markers are closely associated with further decompensation (Figure-1). Concordantly, complement markers exhibited a significant negative association with HVPG and MELD (Supplementary Figure-2).
Figure-1Complement and immunoglobulin serum levels across different stages of advanced chronic liver disease (ACLD).
Circulating levels of the immunoglobulin subtypes IgA, IgM, and IgG, as well as the IgG isoforms IgG 1-4 and TTG-IgA antibody levels, increased in patients with dACLD (S3-S5) as compared to cACLD (S0-S2; all p<0.05; Supplementary Table-1; Supplementary Figure-S3), and correlated with the severity of portal hypertension (reflected by HVPG) and liver dysfunction (reflected by MELD; all p<0.05; Supplementary Figure-S4). When differentiating patients by different decompensated ACLD stages, IgA levels exhibited a stepwise increase from S0-S2 (274 [199-424] mg/dL, p<0.001; Figure-1) to patients with S4 (456 [300-618] mg/dL) and S5 (490 [298-674] mg/dL). The IgA isoform TTG-IgA – a potential marker of mucosal immunity, as described for other mucosa-associated antibodies in the context of cirrhosis [
] – exhibited a similar increase in S4 (2.10 [1.25-3.15] U/mL) and S5 (2.10 [1.20-3.00] U/mL; S0-S2: 1.20 [1.00-2.20] U/mL; p<0.001), indicating that TTG-IgA levels are also linked to disease severity in ACLD. Of note, all patients displayed TTG-IgA levels below the diagnostic threshold for celiac’s disease. IgA and TTG-IgA serum levels exhibited a strong correlation (Spearman’s rs=0.797, 0.75-0.84, p<0.001; Supplementary Figure-S5). As mentioned above, IgG and IgM levels increased in patients with dACLD upon comparison to cACLD. When stratifying patients by different dACLD stages, however, rather weak statistical trends were observed, which might be related to the smaller sample size and adjustment for multiple testing (Table-1; Figure-1; Supplementary Figure-S6). Nevertheless, these results may suggest that the increase of IgA in patients with S4 and S5 represents the most pronounced change among circulating immunoglobulin subtypes in dACLD.
Acute-phase reaction and systemic inflammation
Systemic inflammation parameters CRP, IL-6, and PCT increased across disease stages, while LBP and white blood cells (WBC) did not display significant dynamics across patient groups (Table-1). In order to assess the acute-phase reaction in more detail, we quantified the circulating levels of AGP, SAA, A2M – being regarded as APPs and modulators of inflammation and immunity – across different stages of ACLD severity. AGP and SAA levels were, however, similar across ACLD stages (p=0.349 and p=0.818, respectively; Table-1). Conversely, A2M exhibited a decrease in patients with S4 (231 [198-263] mg/dL) and S5 (205 [159-260], p<0.001) as compared to S0-S2 (286 [228-331] mg/dL; p<0.001). All markers displayed a weak negative association with MELD, indicating that the synthesis of these proteins may be impaired as liver dysfunction progresses (Supplementary Figure-S7/8).
The link between innate and adaptive immunity, acute-phase, and systemic inflammation
To explore whether circulating components of innate and adaptive immunity align with markers of systemic inflammation and bacterial translocation, their correlation with CRP, IL-6, PCT, and LBP levels was assessed (Figure-2). IL-6 (rather than CRP or PCT) exhibited significant associations with complement factors (negative correlation) and immunoglobulin levels (positive correlation) in the study cohort. Among studied parameters, IL-6 displayed the strongest link to IgA (rs=0.501, 0.40-0.59, p<0.001), TTG-IgA (rs=0.382, 0.27-0.49, p<0.001) and A2M (rs= -0.436, -0.53-[-]0.33, p<0.001) levels. LBP levels had no meaningful correlation with immunoglobulin levels in the circulation, however, exhibited positive correlations with complement factors, AGP, and SAA levels. A correlation matrix of all studied parameters (including all immunoglobulin subtypes and complement levels) is displayed in Supplementary Figure-S9.
Figure-2Correlation between biomarkers of systemic inflammation and circulating components of innate and adaptive immunity.
Furthermore, we explored whether the link between systemic inflammation and circulating immune components was restricted to patients with either cACLD or dACLD. The link between IL-6 and IgA, TTG-IgA, and IgG levels was observed in both cACLD and dACLD (Figure-2). Other immunoglobulin subtypes and complement factors exhibited only weak correlations with inflammation parameters in the overall cohort and displayed similar results (i.e., non-significant or weak correlations) in patients with cACLD and dACLD, respectively.
Prediction of decompensation and liver-related mortality
Considering that CAID is believed to impact on the course of ACLD, we assessed whether serum levels of innate and adaptive immunity components in the systemic circulation were indicative of disease progression – as reflected by the composite endpoint of first/further decompensation or liver-related death (LRD) at 24 months. The median transplant-free follow-up was 6.90 (3.30-12.9) months. Eight (3%) patients had no follow-up, and thus, were excluded for outcome analyses. Forty-three (18%) patients developed decompensation or LRD during follow-up.
Patients were stratified into groups above and below the median values of immunoglobulins and complement factors investigated in this study. The incidence of the composite endpoint was significantly higher in patients with IgA >365 mg/dL (log-rank HR: 3.61, 1.98-6.58, p<0.001), total IgG >1470 mg/dL (log-rank HR: 1.96, 1.08-3.57, p=0.032) and IgG1 >973 mg/dL (log-rank HR: 1.89, 1.04-3.44, p=0.043), as well as reduced complement levels, particularly C3c (log-rank HR: 2.58, 1.42-4.70, p=0.003) (Figure-3; Supplementary Figure-S10; Supplementary Table-S2).
Figure-3Incidence of decompensation/liver-related death in patients stratified by complement and immunoglobulin levels.
Statistical analysis: Patients were stratified by median complement and immunoglobulin levels. The incidence of events in different patient groups was compared by log-rank test. Abbreviations: (Ig) immunoglobulin; (C3c) complement C3 component.
To determine parameters that were independent predictors of disease progression, multivariate Cox regression models were performed. Besides established parameters such as HVPG, MELD, and IL6, complement and most immunoglobulin levels were associated with decompensation or LRD on univariate Cox regression. On multivariate analysis that was also adjusted for the presence of dACLD, C3c (aHR per mg/dL: 0.98, 0.97-0.99, p<0.001), and IgG-1 (aHR per 100mg/dL: 1.05, 0.99-1.12, p=0.103) – next to HVPG and IL-6 – displayed an independent association with the composite endpoint (Table-2).
Table-2Cox proportional hazard regression model assessing predictors of decompensation/liver-related death.
Parameter
Univariable
Multivariable (last step)
HR
95% CI
P-value
HR
95% CI
P-value
Age (per year)
1.01
0.99-1.04
0.299
Sex (male)
0.66
0.36-1.21
0.177
MELD (per point)
1.15
1.07-1.24
<0.001
0.98
0.87-1.10
0.724
HVPG (per mmHg)
1.13
1.07-1.18
<0.001
1.09
1.03-1.15
0.002
IL-6 (per pg/mL)
1.03
1.02-1.05
<0.001
1.02
1.01-1.04
0.009
dACLD (yes)
2.46
1.18-5.13
0.017
0.94
0.41-2.15
0.888
C3c (per mg/dL)
0.98
0.97-0.99
<0.001
0.99
0.97-0.99
0.040
IgA (per 25 mg/dL)
1.04
1.02-1.06
<0.001
1.00
0.98-1.03
0.725
IgM (per 100 mg/dL)
1.20
1.05-1.37
0.009
1.07
0.91-1.25
0.405
IgG-1 (per 100 mg/dL)
1.10
1.05-1.17
<0.001
1.05
0.99-1.12
0.092
IgG-2 (per 100 mg/dL)
1.04
0.90-1.21
0.589
IgG-3 (per 25 mg/dL)
1.04
0.96-1.13
0.333
IgG-4 (per 25 mg/dL)
1.05
1.01-1.09
0.023
1.03
0.99-1.07
0.195
Statistical analysis: Multivariable analysis was performed using a backward stepwise Cox proportional hazards regression model. Abbreviations: (C3c) complement C3 component; (HVPG) hepatic venous pressure gradient; (Ig) immunoglobulin; (IL-6) interleukin-6; (MELD) Model of End Stage Liver Disease.
Furthermore, we explored whether the susceptibility to infections – indicating a potential clinical manifestation of CAID – was related to immunoglobulin or complement levels. Infections during follow-up were recorded in twenty-eight (11%) patients at 24 months (cACLD: n=11/95; 12%; dACLD: n=17/150, 11%): n=6 had SBP, n=6 urinary tract infection, n=5 respiratory infection (n=1 COVID-19), n=2 sepsis, n=2 gastrointestinal infection, n=1 cholangitis, n=1 superinfected pancreatitis, n=2 other infections (n=1 incarcerated hernia with secondary peritonitis and sepsis, n=1 rectal abscess), and in n=3 cases antibiotic treatment was recorded due to clinical and laboratory signs of infection/inflammation, although no clear focus was identified.
Again, patients were initially stratified by median immunoglobulin and complement levels. The incidence of infections was significantly higher in patients with low C3c levels (log-rank HR: 2.34, 1.12-4.92, p=0.030), and tended to correlate with high IgG1 levels (log-rank HR: 2.14, 1.02-4.48, p=0.054; Supplementary Figures-S11/12; Supplementary Table-S3). Importantly, IL-6, IgG-1 (all p<0.05), and IgG-4 (p=0.076) were associated with the development of infections on univariate Cox regression analysis. Multivariate analysis indicated that IL-6 (aHR per pg/mL: 1.03, 1.00-1.05, p=0.023) and IgG-1 (aHR per 100 mg/dL: 1.12, 1.04-1.19, p=0.002) were independently linked to the development of infections in patients with ACLD (Table-3).
Table-3Cox proportional hazard regression model assessing predictors of infections during follow-up.
Parameter
Univariable
Multivariable (last step)
HR
95% CI
P-value
HR
95% CI
P-value
Age (per year)
1.02
0.99-1.05
0.246
Sex (male)
0.82
0.38-1.74
0.596
MELD (per point)
1.05
0.95-1.16
0.346
HVPG (per mmHg)
1.05
0.99-1.11
0.116
IL-6 (per pg/mL)
1.03
1.01-1.05
0.007
1.03
1.00-1.05
0.023
dACLD (yes)
0.95
0.44-2.02
0.887
C3c (per mg/dL)
0.99
0.98-1.01
0.191
IgA (per 25 mg/dL)
1.02
0.99-1.05
0.191
IgM (per 100 mg/dL)
1.05
0.83-1.32
0.719
IgG-1 (per 100 mg/dL)
1.13
1.06-1.22
<0.001
1.12
1.04-1.19
0.002
IgG-2 (per 100 mg/dL)
0.91
0.74-1.11
0.354
IgG-3 (per 25 mg/dL)
1.04
0.93-1.15
0.521
IgG-4 (per 25 mg/dL)
1.04
1.00-1.09
0.076
1.03
0.98-1.07
0.282
TTG-IgA (per U/mL)
1.09
0.88-1.35
0.442
Statistical analysis: Multivariable analysis was performed using a backward stepwise Cox proportional hazards regression model. Abbreviations: (C3c) complement C3 component; (HVPG) hepatic venous pressure gradient; (Ig) immunoglobulin; (IL-6) interleukin-6; (MELD) Model of End Stage Liver Disease; (TTG) tissue transglutaminase.
In the present study, we demonstrate a dysregulation of innate and adaptive immunity in 245 well-characterized ACLD patients without infections or acute decompensation. Importantly, systemic C3c and IgG1 levels seem to be valuable biomarkers indicative of CAID[
]. A dysfunctional gut-liver axis and bacterial translocation are believed to promote hepatic and systemic inflammatory responses, and thus, the development of CAID [
Complement proteins are primarily synthesized by hepatocytes and are involved in the removal of pathogens through opsonization, mediation of inflammatory processes, and cytotoxicity [
]. However, decreased complement levels have been reported in patients with cirrhosis presumably caused by hepatic dysfunction, and linked to an increased risk for infections and a worse prognosis [
]. In the present study, we were able to confirm the link between complement levels and liver dysfunction as well as portal hypertension in a large cohort of thoroughly characterized ACLD patients. Importantly, we demonstrate a pronounced decrease in complement levels, as well as the in vitro complement activity assay CH50, particularly in patients with S4 (non-bleeding decompensation) and S5 (further decompensation). The decrease in complement levels/activity in ACLD is likely related to both impaired synthetic function and increased complement consumption (e.g., in response to bacterial translocation). Such et al. observed that intestinal decontamination increased C3 levels in serum and ascitic fluid in a small study on patients with cirrhosis and ascites [
]. We acknowledge that our study cannot provide further mechanistic evidence on the source of reduced complement levels in patients with dACLD. Nevertheless, the observed negative correlation between IL-6 and complement levels and the prognostic value of complement levels towards disease progression underlines the clinical significance of reduced complement levels in ACLD.
Furthermore, the production of immunoglobulins by B-cells is an important part of the adaptive immune system. Previous studies have suggested B-cell dysfunction in cirrhosis [
]. Importantly, immunoglobulin levels (IgA, IgM, IgG) decreased in patients with alcoholic cirrhosis after regeneration of hepatic function following liver transplantation [
]. Our study demonstrates that the immunoglobulin subtypes IgA, IgG, and IgM increased in patients with dACLD and were linked to portal hypertension and liver dysfunction. Interestingly, IgA serum levels exhibited a pronounced increase in patients with S4 and S5 and exhibited a significant correlation with systemic inflammation markers. Considering the impaired intestinal barrier in cirrhosis and the resulting exposure of gut-associated lymphoid tissue to pathogens from the intestinal lumen [
], it is tempting to speculate that the observed elevation of IgA in patients with dACLD (S5>S4) reflect bacterial translocation in ACLD. Concordantly, Massonnet et al. suggested that the upregulation of IgA production in cirrhosis is related to activation of toll-like receptor pathways [
]. Furthermore, the strong correlation between IgA and ‘subclinical’ TTG-IgA levels (i.e., below the threshold indicative of celiac disease) also suggests a potential link to mucosal immunity in cirrhosis. To this end, also anti-gliadin antibodies were previously associated with portal hypertension severity and intestinal permeability in patients with ACLD [
Although IgA levels showed strong dynamics across ACLD stages, only the IgG isoform IgG1 (next to HVPG, IL-6, and C3c) displayed independent predictive value for first/further decompensation or liver-related death on multivariate analysis. The prognostic value of IL-6 and IgG1 likely underline the clinical significance of a proinflammatory state in ACLD. Interestingly, the finding of increased IgG-4 levels in dACLD – which is considered as a mediator of anti-inflammatory effects or even of immunotolerance in other contexts [
CAID is believed to ultimately predispose patients with ACLD to infections and decompensation. When we analysed the association of our biomarkers with the development of infections during follow-up, IgG1 and IL-6 were independently predictive for infections in our cohort of ACLD patients. These results further emphasize the potential role of circulating biomarkers of immune response to bacterial translocation as surrogates for CAID, particularly since patients developing infections were at increased risk of liver-related complications in previous studies [
]. While it can only be speculated why complement factors were not indicative towards developing infections, one could argue that the local defence against the entry of pathogens may not be primarily related to the amount of circulating complement factors and their activity. Furthermore, patients with dACLD exhibited a similar infection rate in our study as compared to patients with cACLD, which is surprising, e.g., when considering the high number of infections reported in patients with acute decompensation included in the PREDICT study [
]. One could speculate that the selection of patients without acute decompensation, infections, or non-elective hospitalization at the timepoint of HVPG measurement may be the key difference to other studies focusing on patients with dACLD. For example, patients allocated to the “stable decompensated cirrhosis” group in the PREDICT study exhibited the most infection events “at” or shortly “before and after” study inclusion (i.e., the timepoint of acute decompensation) [
]. Notably, the relatively short follow-up interval of our cohort needs to be considered when interpretating the predictive value of these biomarkers towards infections, indicating that this question should be addressed in future studies with long-term follow-up of patients with dACLD.
Finally, our study investigated whether certain APPs were related to ACLD stage or to systemic inflammation: AGP exerts numerous immunomodulatory functions that include adhesion, pathogen binding, and regulation of leukocyte functions [
]. Interestingly, A2M decreased in patients with dACLD (S4 and S5), displayed a negative correlation with CRP or IL-6, and was also linked to clinical events during follow-up. These results emphasize that the mechanistic implications caused by dysregulation of certain APPs such as A2M may further contribute to immune dysfunction and thus, to CAID in patients with ACLD [
In summary, our study demonstrates that the dysregulation of complement factors, immunoglobulins, and APPs indicate the presence of CAID and are closely linked to ACLD severity and prognosis. Low C3c was an independent predictor of the composite endpoint “decompensation or liver-related death”, while high IgG1 was an independent predictor of infections. Further mechanistic studies are warranted to decipher the underlying mechanistic effects as potential therapeutic targets and to explore the effects of therapies targeting the gut-liver axis on C3C and IgG1 as biomarkers of CAID.
Appendix A. Supplementary data
The following is the supplementary data to this article:
Von Willebrand factor indicates bacterial translocation, inflammation, and procoagulant imbalance and predicts complications independently of portal hypertension severity.
Reiberger, T., Schwabl, P., Trauner, M., Peck-Radosavljevic, M. and Mandorfer, M., Measurement of the Hepatic Venous Pressure Gradient and Transjugular Liver Biopsy. J Vis Exp, 2020(160).
00043-5&id=ga1.jpg){kind=link}
00043-5&id=gr1.jpg){kind=link}
00043-5&id=gr2.jpg){kind=link}
00043-5&id=gr3.jpg){kind=link}