The role of IL-22 in disease progression of patients with liver cirrhosis is currently uncertain. Lymphoid cells are supposed to be the major source of IL-22 , though cells of the macrophage/dendritic cell type have been likewise reported to be capable of producing the cytokine [14, 29]. In contrast, major liver targets of IL-22 are certainly hepatocytes which express functional IL-22 receptors to a large extent . In the present study, we performed immunohistochemistry to confirm IL-22 expression in cirrhotic livers. Indeed, IL-22 was detectable in non-parenchymal cells in the hepatic lobe and in areas of necroinflammation. These observations are in agreement with previous data from patients and murine models, suggesting that the liver is an important source of serum IL-22 under conditions of hepatic inflammation/injury [20, 26].
The potential role of IL-22 in liver diseases has been intensively studied in murine models for T cell-mediated hepatitis , fulminant hepatic failure , alcoholic liver injury  and regeneration after hepatectomy . In those models, IL-22 attenuated liver injury [20, 22], prevented hepatic failure  and improved hepatic steatosis . On the other hand, blockage of IL-22 bioactivity increased liver injury  and was associated with decreased hepatocyte proliferation following hepatectomy . On the whole, with the exception of experimental hepatitis B virus infection , murine models largely suggest a tissue protective function of IL-22 in hepatic disorders.
The pathophysiological relevance and prognostic potential of serum IL-22 in patients with liver diseases of various etiologies is less clear. The limited data available suggest increased serum IL-22 in patients with acute HBV infection  and in patients with chronic hepatitis , respectively. This latter study also links hepatocarcinogenesis to IL-22 function. IL-22 may be particularly important for outcome of liver cirrhosis. To relate systemic IL-22 to the prognosis of clinical liver cirrhosis, we performed a prospective cohort study in which patients with advanced liver cirrhosis were consecutively enrolled and longitudinally followed. Our data show that, compared to healthy donors, IL-22 serum levels were significantly elevated in patients with liver cirrhosis. Furthermore, elevated IL-22 levels were associated with liver-related complications, such as ascites, hepatorenal syndrome and spontaneous bacterial peritonitis. These observations altogether indicate that high IL-22 serum levels may reflect the severity of liver disease.
IL-22 sera contents in healthy donors were, for the most part, barely detectable and set the basis for calculation of a reference range. This reference range defined levels below 18 pg/ml as being normal (ULN), which agrees with previous reports on IL-22 in sera of healthy donors obtained in the US and Europe [6, 9, 27, 32]. According to this threshold, 47.5% of patients with liver cirrhosis showed elevated IL-22 serum concentrations. Follow-up analyses of serum IL-22 levels in patients with liver cirrhosis suggest that mean IL-22 levels increase during the course of liver disease. The majority of patients with elevated IL-22 baseline levels maintain elevated levels during follow-up, while more than half of patients with normal IL-22 serum levels at baseline develop increased levels during follow-up. These results indicate that IL-22 elevation is not a transient phenomenon in patients with liver cirrhosis. The mechanisms mediating this IL-22 increase in the patients are not yet clear. However, it can be assumed that increasing IL-22 levels are connected with increased cytokine production as well as reduced hepatic or renal elimination.
Up-regulation of IL-22 might be associated with the underlying etiology of liver cirrhosis. In the present study, however, we observed no difference in the IL-22 levels in sera from patients with prevalent etiologies of liver disease, that is, HBV, HCV and alcoholic liver cirrhosis. In the course of this study, hepatocellular carcinoma was diagnosed in two patients with detectable IL-22. This observation is in line with elevated systemic IL-22 in patients with hepatocellular carcinoma .
Advanced liver cirrhosis is associated with poor prognosis. Accordingly, prediction of hepatic deterioration is key to efficient management of the clinical disease. In the present study, approximately one quarter of patients died after one year due to hepatic deterioration. The primary goal of this study was to investigate whether serum IL-22 relates to survival of patients with liver cirrhosis. In fact, serum IL-22 was significantly associated with survival. However, patients with elevated serum IL-22 (>18 pg/ml) levels showed reduced survival compared to those patients with normal IL-22 levels indicating that processes leading to deterioration of liver cirrhosis and its sequelae come along with an increase of serum IL-22. With respect to the hepatoprotective effects of IL-22 in animal models, we assume that elevated IL-22 serum levels in liver cirrhosis still serve protective functions. However, it cannot be excluded that IL-22 beyond a certain threshold may be pathogenic.
The best currently evaluated prognosis score for patients with liver cirrhosis is the MELD score. Systemic IL-22 levels in patients with liver cirrhosis significantly correlated with the MELD score, substantiating that IL-22 is associated with deterioration of liver function and subsequent mortality of cirrhotic patients. The MELD score is a short-term (three- to six-month) predictor of survival in patients with end-stage liver disease, but is a weak predictor of survival in patients with compensated liver cirrhosis in the long term. In multivariate analysis, systemic IL-22 was (independently from age, presence of liver-related complications, elevated creatinine, high CRP and high MELD score) associated with long-term mortality. Taking into account that IL-22 serum levels were stably increased in the majority of patients in the course of liver cirrhosis, the IL-22 serum level may be a valuable prognostic marker for long-term survival of patients with advanced liver cirrhosis.
The MELD score includes three blood surrogate parameters addressing different aspects of liver deterioration. INR and bilirubin reflect liver synthetic capacity and excretory function, while creatinine indicates renal decompensation due to hepatic failure. IL-22 serum levels correlated with two parameters of the MELD score, that is, creatinine and INR. Furthermore, weak inverse correlations were observed between systemic IL-22 and serum albumin and ALT. The weak but significant inverse correlation between the ALT level and IL-22 serum level must be carefully interpreted, but may indicate that IL-22 has a hepatoprotective function in patients with advanced liver cirrhosis.
IL-22 also correlated with CRP, a well-established surrogate marker of hepatic inflammation and prognosis of liver cirrhosis . CRP and creatinine were both associated with serum IL-22, further suggesting that enhanced systemic IL-22 is driven by hepatic inflammation along with renal deterioration. Whether IL-22 bioactivity likewise contributes to renal deterioration is unknown
In contrast to the surrogate parameters currently used for the assessment of the prognosis of liver cirrhosis, IL-22 likely plays an active role in liver inflammation, regeneration and carcinogenesis [20, 22, 30, 31]. Of particular interest is the association among serum IL-22, CRP and spontaneous bacterial peritonitis, indicating that IL-22 production is augmented by the infection. In fact, previous studies showed up-regulation of the systemic IL-22 level in patients with abdominal sepsis .