Complement 3 could be available marker of liver function evaluation in patients with hepatitis B virus related cirrhosis and independent predictor of poor outcome in decompensated

Hepatocytes biosynthesize about 80–90% of plasma complement proteins, including complement 3 (C3) and 4 (C4). The purpose of our study is to research the signicance of complement in hepatitis B virus related cirrhosis (HBV-cirrhosis), discuss whether C3 and C4 could be served as available markers to assess liver function and prognostic in HBV-cirrhosis. HBV-cirrhosis. signicance negative correlation with to platelet spite of C4 level was signicance between three There was no signicant (p=0.293); observed to be signicantly negatively correlated with MELD signicance negative correlation with to platelet ratio. score, End-stage Liver of C3 95%CI: markedly higher than C4 (AUC=0.669, 95%CI: 0.533-0.806, p=0.013); (G) Kaplan-Meier survival analysis revealed that signicantly worse survival rate in low C3 level grade compared with high group; Kaplan-Meier survival analysis revealed that high C4 level group were better in survival rate of decompensated cirrhosis than low group. IF, infection; P,


Introduction
Cirrhosis is end stage of chronic liver diseases caused by variety of etiological factor, the major is Hepatitis B virus (HBV) infection in Asian region, due to repeated in ammation-necrosis-regeneration process (1). Usually, it is classi ed as compensated and decompensated based on portal hypertension and liver function, the decompensated cirrhosis is frequent and dangerous state causes numerous clinical contacts due to its complications, patients often die of multiple organ failure due to complicated ascites, gastrointestinal hemorrhage, hepatic encephalopathy, hepatorenal syndrome and sepsis (2,3).
Prognosis of cirrhosis is in uenced by competent clinical decision and selected clinical approach, precise liver function evaluation is essential for this process. The liver is multiple function organ ful lls the various metabolic requires of body, such as synthesizing proteins, producing digestive enzymes, and detoxifying various metabolites (4)(5)(6). Currently, liver function evaluation is based on the capacity of synthesizing albumin and clotting factors, as well as bilirubin metabolism. Several prognosis score systems, such as Child-Pugh score (C-P score) and End-stage Liver Disease score (MELD score), were broadly applied to clinical practice (4). Emerging views suggest that liver is considered a predominant immune organ and ful ll important immunologic function (7,8). Hepatocytes account for up-to 70% of the liver cells, biosynthesize about 80-90% of plasma complement proteins (9). The complement as part of the immune system performs an important mission in physiological and pathological processes, such as microbial clearance, maintenance homeostasis, immune regulation, tissue regeneration and in ammation (10)(11)(12). The similar as activation of coagulation system, the complement system also through limited proteolytic steps result in generating protein-protein complexes. Three pathways, including the classical, the lectin and the alternative, serve for activation of the complement system, initiated through different molecules, converges at level of the complement 3 (C3) (13).
Number of research recognize that the etiopathogenesis of some liver disorders is contributed by the activation of complement system, involved in liver failure (14)(15)(16), viral hepatitis (17), alcoholic fatty liver disease(18), nonalcoholic fatty liver disease (19,20), and autoimmune liver diseases (21). In nonalcoholic steatohepatitis, it was suggest that serum C3 level is closely correlation with the disease severity (20).
Previous studies discussed the role of C3 in patients with HBV related acute-on-chronic liver failure and indicated that it can be independent risk factor for mortality (14,15). A research in cirrhosis patients revealed that the level of lectin pathway complement components was decreased and closely related to disease severity, but did not investigate the role of C3 (22).
Despite the importance of complement system is currently turning increased attention by pathogenesis research of liver diseases, the signi cance in liver function test for cirrhosis is leaved out. Presently, more and more clinicians in practical decision-making ask for more well understanding and more precisely evaluating liver function to establish treatment strategy and improve outcome for patients with cirrhosis.
Therefore, given the complements as the bridge between innate and adaptive immune systems and are produced by liver (13), we seek to explore its role in HBV-cirrhosis and its relationship with liver function and disease activity.

Patients and Sample Collection
In our study, we enrolled 306 adult patients with HBV related cirrhosis, they were inpatients who admitted to The Liver Disease Center of PLA, The 980th Hospital of PLA Joint Logistics Support Force, from January 2019 to January 2021. All patients were willing to participate. The exclusion criteria were: 1) the patient declined to be recruited in this study; 2) age less than 18 years old; 3) co-infected with hepatitis A, C, D, and E or HIV; 4) co-existence other liver disease, including steatosis hepatis, inherited metabolic liver disease, autoimmune liver diseases, cancer, biliary obstruction; 5) liver injury caused by drug intake and alcohol abuse; 6) presence history of rheumatic, renal, pancreas, endocrine, pulmonary, cardiovascular disease, splenectomy; 7) pregnancy.
The diagnosis of chronic HBV infection is based on Asian-Paci c clinical practice guidelines on the management of hepatitis B(1). The diagnosis of cirrhosis is based on Assessment and management of cirrhosis in people older than 16 years: NICE guidance (23). Cirrhosis was established by imageological diagnosis including computerized tomography scans or magnetic resonance image. We also enrolled 23 chronic CHB patients from our hospital. The CHB was de ned as Hepatitis B surface antigen positive in serum more than 6 months and persistently repeatedly elevated alanine aminotransferase without others reasons(1). The CHB group patients were without cirrhosis and liver failure. The control group was consisting of 12 healthy adults. Morning fasting blood samples were collected after admission, then immediately detected and analyzed. The study was approved by the Ethical Committee of The 980th Hospital of PLA Joint Logistics Support Force (2018-KY-33), the written consent was obtained from each patient.

Patients Characteristics and Parameters
Detailed clinical data were captured including age, gender, decompensated event. The decompensated cirrhosis is de ned as development of ascites, gastrointestinal hemorrhage, hepatic encephalopathy, hepatorenal syndrome and bacterial infection (2). Patient clinical parameters for evaluating disease severity involved routine blood tests, liver and renal function tests, coagulation function tests, cholesterol, Alpha-fetoprotein (AFP), HBV-DNA. Bacterial infection was identi ed by attending doctor according to patient clinical symptoms and infection index, such as routine blood tests, urine analysis, Creactive protein (CRP), procalcitonin (PCT) and cultures (including blood, urine, sputum), in addition, take ascites analysis into account if ascites was present (including ascites culture)(2).

Disease Severity Evaluation
The liver-oriented scores, such as C-P score, MELD score and the aspartate aminotransferase (AST) to platelet ratio, were employed to evaluate severity of liver disease in this study. The C-P score and MELD score is calculated following the previously described criteria (4). The AST to platelet ratio is noninvasive tests for brosis and calculated as the ratio between the serum AST level and the platelet count(6).

C3, C4 Analysis
The serum levels of complement components C3 and C4 were detected by commercial kits following the manufacturer's instruction (Yilikang Biotech Co. Ltd, Zhejiang, P.R. China). Normal ranges of C3 and C4 is referred as 0.9 -1.5 g/L and 0.2 -0.4 g/L, respectively.

Statistical Analysis
The SPSS version 26 and GraphPad Prism 8 software were employed to perform the statistical analysis for all data. Variables were expressed as means (standard deviation, SD) or medians (interquartile range).
Differences in categorical variables were compared by chi-squared test or Fisher's exact test. Differences in continuous variables were compared by non-parametric Mann-Whitney U-test or Kruskal-Wallis Htest. Spearman's rank correlation test was employed to analysis the relationship between two variables. The receiver operating characteristic (ROC) curve was applied for predicting the optimal cut-point of variable on infection and death. Cumulative probability of survival was calculated by Kaplan-Meier analysis and compared by the log-rank test. The correlation between clinical variables and infection or death was evaluated by univariate Logistic-regression or univariate Cox-regression analysis, respectively. The independent factors were assessed by Multivariate Logistic-regression or Cox-regression analysis with backward stepwise and likelihood ratio test. Two-tailed p value < 0.05 was considered to be statistically signi cant.

Patients Characteristics
The baseline characteristics of total of 306 patients with HBV-cirrhosis, 23 CHB and 12 healthy adults were presented in Table 1. Gender and creatinine were not signi cant differences among three groups. Among 306 HBV-cirrhosis patients, 80 developed gastrointestinal hemorrhage, 51 developed hepatic encephalopathy, 122 developed ascites, 33 developed bacterial infection. Unfortunately, 20 patients with cirrhosis death. Thus, 190 patients were de ned as decompensated cirrhosis, 116 patients were de ned as compensated cirrhosis. There were no signi cant differences in gender, age, alanine aminotransferase (ALT), creatinine, AFP and HBV-DNA levels between compensated and decompensated cirrhosis, nevertheless, the International Normalized Ratio (INR), AST, albumin, bilirubin, cholinesterase, cholesterol, urea nitrogen, leukocyte, platelet levels and C-P, MELD score, AST to platelet ratio were signi cant differences.

Complement 3 and 4 levels markedly decreased in HBV-cirrhosis
The C3 and C4 levels were markedly decreased in HBV-cirrhosis group than those in CHB and control group (Fig1A, Fig1E). The median (interquartile range) concentration of C3  We also investigated the serum C3 and C4 levels in cirrhosis sub-group, the result shown that decompensated cirrhosis group were signi cantly lower compared to compensated (Fig1A, Fig1E). The C3 was 0.865 (0.7225-1.0175) g/L and 0.52 (0.42-0.6325) in compensated and decompensated cirrhosis, the C4 was 0.17 (0.13-0.2175) g/L and 0.1 (0.07-0.13), both concentration were signi cant differences in two sub-group (p<0.0001). In spite of the C3 and C4 levels were reduced in compensated cirrhosis than CHB, there was no signi cant difference in C4 (p=0.346), while C3 level was signi cant difference (p=0.042). In addition, the concentration of C3 and C4 were markedly decreased in cirrhosis with gastrointestinal hemorrhage ( There were signi cance differences in all cirrhosis sub-groups (p<0.05). These results suggested that C3 and C4 might be play important role in disease advance of patient with HBVcirrhosis.

C3 and C4 closely correlated with laboratory markers of liver function in HBV-cirrhosis
For investigating whether C3 and C4 could be used to assess liver function in HBV-cirrhosis, we analyzed the relationship between this complement components and laboratory markers of liver function. The correlation was presented in Table 2. The C3 and C4 shown signi cant negative correlation with INR ( Fig.2A, Fig.3A), bilirubin (Fig.2C, Fig.3C), and signi cant positive correlation with albumin (Fig.2B,  Fig.3B), cholinesterase (Fig.2D, Fig.3D), cholesterol (Fig.2E, Fig.3E), platelet (Fig.2F, Fig.3F). In addition, C3 and C4 shown slightly negative correlation with AST and urea nitrogen, slightly positive correlation with leukocyte. None of them were signi cant correlation with age, ALT, creatinine, AFP and HBV-DNA in HBV-cirrhosis.

C3 and C4 closely correlated with liver-oriented scores in HBV-cirrhosis
We also analyzed whether the levels of C3 and C4 were correlated with liver-oriented scores in HBVcirrhosis. Firstly, we observed the levels of C3 and C4 in HBV-cirrhosis with different C-P score. Serum C3 level was signi cant lower in cirrhosis with C-P Score C compared with C-P Score A/B (Fig.4A), it was 0.835 (0.68-0.99) g/L in A, 0.56 (0.47-0.64) g/L in B, 0.42 (0.315-0.51) g/L in C. There was signi cant difference in three groups (p<0.0001), and signi cance negative correlation between C3 and C-P Score (p<0.0001). In terms of C4 (Fig.4D), it was 0.165 (0.13-0.2) g/L in A, 0.1 (0.08-0.13) g/L in B, 0.09 (0.06-0.11) g/L in C. In spite of C4 level was signi cance negative correlation with C-P Score (p<0.0001) and was signi cance difference between three C-P score groups (p<0.0001), there was no signi cant difference between B and C (p=0.293). Secondly, we observed the correlation between levels of C3/C4 and MELD score in HBV-cirrhosis. The median (interquartile range) of MELD score was 7 (6-9) and 12.5 (10)(11)(12)(13)(14)(15)(16) in compensated and decompensated cirrhosis, it was signi cant difference (p<0.0001). Spearman's rank correlation test revealed that both C3 and C4 were signi cance negative correlation with MELD score (Fig.4B, Fig.4E). In addition, we also observed the correlation between levels of C3/C4 and the AST to platelet ratio. The median (interquartile range) of AST to platelet ratio was 0.392 (0.22-0.678) and 0.78 (0.541-1.321) in compensated and decompensated cirrhosis, it was signi cant difference (p<0.0001). Similarity, C3 and C4 were signi cance negative correlation with AST to platelet ratio (Fig.4C Fig.4F). These results indicated that the C3 and C4 levels could be used to assess liver function in HBV-cirrhosis.  (Fig.5C), the area under curve of C3 (AUC=0.817, 95%CI: 0.735-0.9, p<0.0001) was markedly higher than C4 (AUC=0.671, 95%CI: 0.564-0.779, p=0.002). These results suggested that C3 could be used to more well assess disease severity in HBV-cirrhosis compare with C4, the lower of C3 level, the higher bacterial infection risk in decompensated.
Next, we observed whether the C3/C4 could be served as predictor of mortality in decompensated HBVcirrhosis. 190 decompensated HBV-cirrhosis patients were grouped into 2 groups according to the optimal cut-point of C3 level (low grade: C3 ≤ 0.415 g/L, n = 46; high grade: C3 > 0.415 g/L, n = 144). Similarly, based on the optimal cut-point of C4 level, patients with decompensated cirrhosis also were grouped into 2 groups (low grade: C4 ≤ 0.065 g/L, n = 36; high grade: C4 > 0.065 g/L, n = 154). We analyzed the potential impact of different C3 or C4 levels on survival of decompensated HBVcirrhosis, Kaplan-Meier analysis revealed that signi cantly worse survival rate in low C3 level grade than high C3 group (Fig. 5G, p<0.0001), high C4 level group were better in survival rate of decompensated cirrhosis than low C4 group (Fig. 5H, p=0.001).
In addition, assessment by univariate Cox-regression analysis were presented in Table 3. We showed that C3 ≤ 0.415 g/L (HR: 15.194 .001) and C-P Score C (HR: 4.661, 95%CI: 1.021-21.286, P=0.047) were independently hazardous factor for mortality. These results suggest that the C3 level could be served as predictor of mortality in decompensated HBV-cirrhosis.

Discussion
There were lots of researches explored the signi cance of complement in various liver disease, however, the linkage of complement protein to liver function is poorly alluded to in liver disease study let along in cirrhosis. With respect to liver function tests, many of tests is not to evaluate actual liver function but pinpoint damage source where may be taking place, such as aminotransferase and alkaline phosphatase, it is misguided(6, 13). The factual liver function based on a series of physiological function, including synthetic, secretion, excretory and immunity. Complement protein has not been applied as part of liver function tests to become follow-up marker. In our study, of the pivotal components of complement system, the C3 and C4 were systematically evaluated.
In patients with HBV-cirrhosis, we found that C3 and C4 levels were markedly decreased compared with CHB and health control, especially in decompensated stage, such as gastrointestinal hemorrhage, hepatic encephalopathy and ascites. Previous research pointed that the liver (mainly hepatocytes) biosynthesize about 90% of complement proteins, including C3 and C4. One rational explanation for lower C3 and C4 levels in cirrhosis is that decreased biosynthesis of hepatogenous proteins on account of impairing hepatocytes synthesize function. For verifying the hypothesis, we investigated the other laboratory markers which re ect hepatic synthetic function, including albumin and clotting factors (4,5). Results shown that the signi cant correlations were existed between albumin or INR and C3/C4, which was highly consistent with disease severity. This signi cant correlation also were existed in serum cholinesterase (24) or total cholesterol (25), which were hepatocyte synthesize. These results supported that the C3 and C4 could be server as biochemical markers to assess hepatocytes synthesize function in cirrhosis.
Total bilirubin is considered as important biochemical marker of liver function and included in C-P score or MELD score systems(4, 5), our results shown that C3 and C4 levels were signi cant correlation with total bilirubin. The C-P score or MELD score were composite scores systems and widespread applied to assess liver function and severity of liver disease. There were signi cant negative correlations between C3 or C4 levels and C-P score or MELD score. However, unlike serum C3 level were gradually decreased following C-P score A up-to C, C4 level was no signi cant differences between C-P Score B and C. Therefore, we suggested that C3 might be a potential biochemical marker to evaluate severity of HBV-cirrhosis and superior to C4, the lower of C3 level, the more severely of disease and deteriorated liver function.
The decreased platelet count is generally regard as suggestive marker of cirrhosis, which consistent with advancing brosis and cirrhosis(6). As brosis progresses, the AST to platelet ratio is gradually elevated(6). Some previous research pointed that complement system is contributed to pathogenesis of liver brosis (9,26,27). Our study revealed that the declined level of C3 and C4 were consistent with the decreased platelet count in cirrhosis, and there was signi cant correlation with the AST to platelet ratio. Intriguingly, the serum C3 level was signi cant difference between compensated cirrhosis and CHB, this phenomenon cannot be observed in C4, albumin, INR and total bilirubin but in leukocyte and platelet levels. Thus, we hypothesize that C3 might be potential noninvasive suggestive biochemical markers of predict cirrhosis, it is worthy of further exploration because of very important for improving prognosis of patients with chronic HBV infection. There was study reported that HBV could inhibit the synthesis and expression of C3 and C4 in vivo or in vitro(28). However, we did not observe the correlation between HBV-DNA level and C3 or C4 level in cirrhosis, the widespread antiviral therapy for HBV-cirrhosis patient might be reasonable explanation for this inconsistent phenomenon, the further clari cation is necessary.
Emerging views suggest that liver is considered a predominant immune organ and ful ll important immunologic function (7,8). The liver not only enrich a great deal of immune cells but also biosynthesize most of complement components and pathogen-recognition receptors, it has profound implications on innate immunity and adaptive immunity. With progresses of cirrhosis and worsening of liver function, innate and adaptive immune dysfunction are inevitable outcome, enhanced infection risk including systemic bacterial infections and bacterial translocation has signi cant impact on patient consequence (29,30). Complement system play a bridging role of innate and adaptive immunity, cascaded activation through classical, lectin pathway or alternative pathways, the C3 takes center stage inside this process (13). Number of previous researches mentioned that lower complement level could enhance the sensibility to bacterial infection in cirrhosis patient due to weaken microbes recognition and clearance capacity (22,31).
In our decompensated HBV-cirrhosis patient population, the serum C3 levels was remarkably decreased in those subjects with infection, and superior to C4 in predicting infection according to ROC curve. Previous study on patients with alcoholic cirrhosis indicated that acquired C3 de ciency predispose to infection (31), which is line with our research. The role of complement component not only involved in extracellular but also in intracellular or subcellular space (10,12). One study on patients with cirrhosis revealed that decreased lysozyme and C3 levels in alveolar lining components induce de ciency of bactericidal activity is contributed to higher sensitivity to Streptococcus Pneumoniae (32). The other study in gut epithelial cells clari ed that complement play dramatic role in maintaining homeostasis of host/environmental mucosal interface (33,34). The most common infection in cirrhosis is Spontaneous bacterial peritonitis (35,36), the homeostasis in gut epithelial cells is broken by defect of complement component induce bacterial translocation may be a probable reason, it is worth to further explore. Patient with cirrhosis occur infection generally result in higher mortality, early identi cation of high risk factors for bacterial infection episodes are paramount importance. Presently, only liver disease severity and gastrointestinal hemorrhage are considered as important clinical predictors(36), our study suggested that lower C3 level might be another, which could be considered as an available biochemical marker to evaluate liver immunologic function.
In addition, we also revealed that the lower C3 and C4 levels were associated the worse survival rate in decompensated HBV-cirrhosis patient population, and C3 might be an independent prediction factor of the mortality. Deteriorated liver function in advanced cirrhosis might be sound reason for this phenomenon. The serum C3 and C4 levels is maintained by biosynthesis and activation consumption (11). The activation of complement system maintains subtle balance between proin ammatory and anti-in ammatory, it is a double-edged sword in disease progress. On the one hand, activation could result in liver tissue damage and amplify systemic in ammatory. Evidence that membrane attack complex could be found to deposit at liver parenchyma of patient with fulminant hepatitis or acute liver failure(16, 37, 38). In rat model, soluble complement receptor 1 could be administrated to e ciently inhibit C3 activation result in ameliorating hepatic ischemia-reperfusion injury induced liver injury (39,40). In primate sepsis model, inhibiting C3 activation could attenuate Escherichia coli induced liver in ammatory injury (41). More importantly, C3 de ciency mice had been shown the lower systemic in ammation protect against fulminant liver failure compare with wildtype (42). These researches imply that excessive activation of complement might be a part reason in higher mortality of decompensated HBV-cirrhosis patients. Unfortunately, activation markers were not being quanti ed, it is limitation in our study. On the other hand, complement activation is advantageous for liver regeneration and required for hepatocyte survival (43). Number of studies had been shown the diminished ability of liver regeneration and exacerbated pathological changes in C3-de cient or/and C5de cient mice than wild (44)(45)(46). One study on liver regeneration after hepatectomy con rmed that C3 play pivotal role in this process(47). It is might be the other reason of higher mortality in decompensated HBV-cirrhosis patients with lower C3 level, and worth to further explore. Taken together, our study suggested that C3 level might be a potential predictor of mortality in decompensated HBV-cirrhosis in admission.

Conclusions
In conclusion, our study shown that the serum levels of C3 and C4 are markedly decreased in HBVcirrhosis, particularly in decompensated stage, and closely correlation with disease severity. The C3 level could be served as available follow-up marker to evaluate liver function, including synthetic and immunity, superior to C4, the lower of C3 level suggest the higher bacterial infection risk in decompensated cirrhosis. Furthermore, decreased C3 level might be independent predictor of mortality in decompensated HBV-cirrhosis. These results might assist clinician more well understanding and more precisely evaluating liver function to establish better treatment strategy and improve outcome for patients with cirrhosis when practical decision-making. The detailed role of the complement component in etiopathogenesis of HBV-cirrhosis urgently need to be su ciently clari ed.

Availability of data and materials
All data generated or analysed during this study are included in this article. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Identifying/con dential patient data however will not be shared.

Competing interests
The authors declare that they have no competing interests.

Authors' contributions
JR and DXS contributed to the study idea; JR, JL and HFZ contributed to clinical data collection; JR contributed to statistical analysis and manuscript writing. All authors read and approve the nal manuscript.
47. He S, Atkinson C, Qiao F, Cian one K, Chen X, Tomlinson S. A complement-dependent balance between hepatic ischemia/reperfusion injury and liver regeneration in mice. J Clin Invest.    Correlation of C3 and C4 with liver-oriented score systems. (A) Serum C3 level was signi cant lower in cirrhosis with C-P C compared with C-P A/B (p<0.0001); (B) C3 was observed to be signi cantly negatively correlated with MELD score; (C) C3 was signi cance negative correlation with AST to platelet ratio; (D) In spite of C4 level was signi cance difference between three C-P score groups (p<0.0001), There was no signi cant differences between B and C (p=0.293); (E) C4 was observed to be signi cantly negatively correlated with MELD score; (F) C4 was signi cance negative correlation with AST to platelet ratio. C-P score, Child-Pugh score; MELD score, End-stage Liver Disease score. level was signi cantly lower in decompensated cirrhosis with infection compared to without (p=0.002); (C) The receiver operating characteristic (ROC) curve was applied to predict the optimal cut-point of C3 and C4 levels on decompensated HBV-cirrhosis with bacterial infection, the optimal cut-point of C3 level for predicting infection was 0.425 g/L (sensitivity: 72.7%, speci city: 84.7%), the C4 level was 0.095 g/L (sensitivity: 66.7%, speci city: 63.1%), the area under curve of C3 (AUC=0.817, 95%CI: 0.735-0.9, p<0.0001) was markedly higher than C4 (AUC=0.671, 95%CI: 0.564-0.779, p=0.002); (D) C3 level was signi cantly lower in died decompensated cirrhosis compared to survived (p<0.0001); (E) C4 level was signi cantly lower in died decompensated cirrhosis compared to survived (p=0.014); (F) According to ROC curve, the optimal cut-point of C3 level for predicting death was 0.415 g/L (sensitivity: 80%, speci city: 82.4%), the C4 level was 0.065 g/L (sensitivity: 45%, speci city: 84.1%), the area under curve