Original article

J. ZUWALA-JAGIELLO1, K. SIMON2, M. KUKLA3, E. MURAWSKA-CIALOWICZ4,
J. GORKA-DYNYSIEWICZ1, E. GRZEBYK1, M. PAZGAN-SIMON2

INCREASED CIRCULATING ENDOCAN IN PATIENTS WITH CIRRHOSIS:
RELATION TO BACTERIAL INFECTION AND SEVERITY OF DISEASE

1Department of Pharmaceutical Biochemistry, Wroclaw Medical University, Poland; 2Clinic of Infectious Diseases, Liver Diseases and Acquired Immune Deficiency, Wroclaw Medical University, Poland; 3Department of Gastroenterology and Hepatology, Medical University of Silesia, Poland; 4Department of Physiology and Biochemistry, University of Physical Education, Wroclaw, Poland
Life expectancy of patients with liver cirrhosis is closely linked to the degree of liver dysfunction and the occurrence of bacterial infection. An early diagnosis of infection helps to initiate adequate and timely measures and improves outcome of cirrhotic patients. Endocan is a newly recognized biomarker of sepsis. However, there have been no studies of the trends in endocan levels in cirrhotic patients with bacterial infection and their associations with markers of infection and inflammation. This study sought to assess the diagnostic value of serum levels of endocan, procalcitonin (PCT), C-reactive protein (CRP), tumor necrosis factor-α (TNF-α), and interleukin-6 (IL-6) in 126 patients with cirrhosis: 51 with decompensated infected cirrhosis, 56 with decompensated uninfected and 19 with compensated uninfected cirrhosis at inclusion. We analyzed the association of endocan with clinical factors in cirrhosis by comparison with indicators of infection and inflammation. Endocan, PCT, CRP, IL-6 and TNF-α were assayed in serum samples by ELISA analyses. Serum levels of endocan, PCT, CRP and TNF-α were significantly higher in cirrhotic patients with clinically overt infections. Endocan levels were correlated to neither PCT levels nor IL-6 levels in each group of patients with cirrhosis. CRP and TNF-α levels and Child-Pugh score correlated only in the infected group of patients with endocan levels, while in the uninfected groups of cirrhotic patients no significant correlation could be detected. The diagnostic accuracy of endocan increased in advanced stage of the disease. Serum endocan levels ≥ 2.05 ng/ml had a sensitivity of 76.1% and specificity of 85% for the diagnosis bacterial infection in decompensated cirrhotic patients. The endocan measured at admission is a good clinical parameter predicting the occurrence of infection in these patients. Elevated endocan may reflect the degree of endothelial cell injury induced by a systemic inflammatory response, a pathologic process that could modify the course of advanced cirrhosis.
Key words:
bacterial infection, chronic liver disease, cirrhosis, C-reactive protein, endocan, inflammation, interleukin-6, procalcitonin, tumor necrosis factor-α

INTRODUCTION

Patients suffering from liver cirrhosis often die of life-threatening bacterial infections. The course of advanced cirrhosis, regardless of its etiology, is complicated by cirrhosis-associated immune dysfunction and this constitutes the pathophysiological hallmark of an increased susceptibility to bacterial infection, distinctive of the disease (1). An early diagnosis of infection helps to initiate adequate and timely measures and improves outcome of cirrhotic patients (2-5). In clinical practice, a large number of molecules secreted by the endothelial cells have been investigated as potential biomarkers for the early diagnosis of bacterial infection. These have included regulators of endothelial activation, adhesion molecules, as well as mediators of inflammation (6).

Endothelial cell-specific molecule-1 (ESM-1) - so-called endocan - is a soluble 50-kDa proteoglycan that is mainly produced by activated endothelial cells (7). This proteoglycan is a key player in the regulation of cell adhesion, inflammatory disorders, and tumor progression (8). It has been established as a serum biomarker in cancer (9-15), bacteremia (16) and sepsis (17-19). Additionally, the serum endocan levels are altered by antileukemic chemotherapy (20) and leukapheresis procedure (stem cell harvesting) in multiple myeloma patients (21) as well as acute graft-versus-host disease after allogeneic stem cell transplantation (22). Endocan gene expression levels may be due to the inflammatory cytokines as well as the lipopolysaccharide (LPS) of the gram-negative bacterial cell wall, and thus increases (23, 24, 17). On the other hand, the endocan itself has been shown to elicit severe inflammatory responses both in vitro in human umbilical vein endothelial cells (HUVECs) and in vivo in mice model (25).

Bacterial LPS is the main stimulus for the production of endocan and proinflammatory cytokines such as IL-6 and TNF-α. The inflammatory response to infection as estimated by the levels of TNF-α and IL-6 in serum are increased in patients with liver cirrhosis. A few studies have shown that endocan can be acknowledged as a good marker of endothelial dysfunction and multiple organ failure in sepsis (18, 19). In clinical practice, however, the usefulness and characteristics of endocan in bacterial infection in cirrhotic patients, have not been investigated.

In light of increasing evidence that endocan and inflammatory response are closely linked, the purpose of the present study was to investigate the association between endocan, as a specific biomarker of bacterial infection in patients with liver cirrhosis, and markers of infection and inflammation. Furthermore, we evaluated the performance of endocan assay for the early diagnosis of bacterial infection in decompensated cirrhotic patients.

PATIENTS AND METHODS

Patients

This study was performed in 126 patients with liver cirrhosis with and without decompensation admitted to the Clinic of Infectious Diseases, Liver Diseases and Acquired Immune Deficiency for evaluation. Patients were divided into three groups based on morphological ad bacteriological results: patients with decompensated cirrhosis and infection at admission (n = 51), patients with decompensated cirrhosis without infection (n = 56) and compensated patients without infection at admission (n = 19). The control group (n = 25) consisted of healthy blood donors (16 males / 9 females, median age 53 years) with normal aminotransferases, normal blood counts and negative markers for virus hepatitis and HIV. Blood samples were collected in the Department of Physiology and Biochemistry, University of Physical Education in Wroclaw. Clinical and biochemical characteristics of the study group are reported in detail in Table 1.

Inclusion criteria were: histological or clinical diagnosis of cirrhosis, no evidence of metabolic, toxic or autoimmune liver disease and at least 1 year of alcohol abstinence. Diagnosis of cirrhosis was established according histological criteria when liver biopsy was performed, or by the combination of clinical, biochemical and ultrasound imaging data presence of irregular margins on ultrasound, portal hypertension with laboratory evidence of chronic liver disease consistent with such a diagnosis. Patients were grouped according to Child-Pugh classification. Three biochemical variables (serum albumin, bilirubin, and prothrombin time (international normalized ratio, INR)) in addition to the presence or absence of ascites and clinical signs of encephalopathy determine the Child-Pugh score (26). Patients were scored as follows: 5 – 6 as class (group) A, 7 – 9 as class (group) B and 10 – 5 as class (group) C. At the time of the study no Child-Pugh A patients showed clinical features of decompensated liver cirrhosis (ascites or hepatic encephalopathy).

Table 1. Clinical and biochemical characteristics of the study subjects.
Table 1
INR, normalized international ratio; WBC, white blood cells.

The model for end-stage liver disease (MELD) score was also calculated which is the most commonly used alternative prognostic indicator for cirrhotic patients to the Child-Pugh score. At enrollment, variceal bleeding was detected by endoscopy in 29% of patients, ascites and hepatic encephalopathy were present by physical examination in 52 (41%) and 22 (17%) patients, respectively (Table 1). Presence of ascites was assessed by ultrasonography. Bacterial infection was confirmed by clinical history, physical examination, differential and total white blood cell count, analysis and culture of urine, thorax X-ray and by the culture and white blood cell count of ascetic fluid in patients with ascites. Infections of pneumonia, skin and urinary tract (cystitis, pyelonephritis) were diagnosed on the basis of conventional criteria. Spontaneous bacterial peritonitis was defined as an infection of the ascitic fluid in the absence of any intra-abdominal source of infection with an ascitic fluid polymorphonuclear cell (PMN) count higher than 250 cells/mm3 and/or positive culture. The onset of infection was defined as the time of admission.

Exclusion criteria were co-existing diseases like chronic kidney disease, diabetes mellitus, cardiovascular disease, cardiac decompensation, and hepatocellular carcinoma.

Peripheral venous blood from fasted healthy subjects and fasted cirrhotic patients was collected in separate tubes, one containing the anticoagulant EDTA and the other without serum anticoagulant. The blood was allowed to clot for 30 min at 25°C and centrifuged at 2000 × g for 15 min at room temperature, and the serum was then separated and aliquoted into tubes for storage. The tubes were stored frozen at –80°C until they were used to study different parameters. Average storage time for patients was 30 months and for controls 25 months. The consent of the Bioethics Committee of the Wroclaw Medical University was obtained and all patients were informed about the character of analyses made. Studies were conducted in compliance with the ethical standards formulated in the Helsinki Declaration of 1975 (revised in 1983).

Laboratory determinations

Biochemical parameters were measured before the use of antibiotics at admission. Endocan levels were determined by ELISA analyses (JDIEK H1) (Lunginnov SAS, Lille, France). The assay range of the ELISA kit was 0.15 ng/ml to 10 ng/ml. Procalcitonin (PCT) was analyzed using an immunoluminometric assay (LUMI test R PCT; BRAHMS Diagnostica, Berlin, Germany). Detection limit was 0.05 ng/ml. Serum C-reactive protein (CRP) level was determined with a high-sensitivity nephelometric method using the Beckman Image Immunochemistry system (Beckman Instruments, Fullerton, CA, USA), which has a minimum level of detection of 0.2 mg/L. Serum levels of TNF-α and IL-6 were assayed with enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems Inc., Minneapolis, MN, USA). All analyses were performed in duplicate strictly according to the manufacturer’s instructions.

Statistical analysis

Statistical analyses were performed using Statistica 12.5 software. Continuous variables are expressed as mean standard deviation (S.D.) or as median - interquartile range (IQR) and categorical variables as number (percentage). Frequency data were compared using the χ2 test or the Fischer’s exact test when necessary. Because many of the variables analyzed did not have a normal distribution as determined by the Kolmogorov-Smirnov test, nonparametric tests were used for comparison of data. The Mann-Whitney U test and the Kruskal-Wallis ANOVA test were used to analyze differences among two or more groups, respectively. Regression analysis to determine significant correlations among different parameters was performed using the Spearman correlation coefficient. Statistical significance was established at P < 0.05.

To evaluate the diagnostic performance of endocan receiver operator characteristics (ROC) analysis was performed for all significant differences between groups. ROC curves were generated by plotting the sensitivity against 1 - specificity, and the area under the curve (AUC) with 95% confidence intervals (95% CI) was calculated. The empirical non-parametric method according to DeLong (27) was performed to make pairwise comparisons of ROC curves. The optimum cut-off point based on the ROC analysis was established by selecting the value that provides the greatest sum of the sensitivity and specificity, that is, the point closest to the upper left point of the ROC plot. For the optimum cut-off point provided by each ROC analysis, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated using standard formulas.

RESULTS

A total of 126 patients with liver cirrhosis were consecutively analyzed. Table 1 exhibits the characteristics of included patients. The median age was 58 years (range 47 to 70 years) and a male predominance was observed (60%). The causes of liver cirrhosis were HCV infection (n = 28), HBV infection (n = 19) and alcohol abuse (n = 79). Alcohol was the predominant reason of cirrhosis (63%). The most frequent infection were spontaneous bacterial peritonitis (59%) and urinary tract infection (22%).

Characteristics of decompensated cirrhotic patients with infection at admission

The infected patients with liver cirrhosis were compared to the uninfected ones (Table 2). The prevalence of infection particularly tended to increase with the severity of liver disease (21%, 35% and 44%, respectively for Child-Pugh class A, B and C). Moreover, bacterial infection was associated with significantly higher MELD scores (P < 0.001). Two further factors tended to be related to bacterial infection: presence of ascites (59%) and alcoholic cirrhosis (73%) (Table 2). No differences between infected and uninfected cirrhotic patients were noted regarding the presence of variceal bleeding and encephalopathy. Albumin and total bilirubin levels were similar among all groups. On the other hand, bacterial infection was associated with higher median values of white blood cells (WBC) counts, INR and creatinine (P < 0.001, P < 0.01, P < 0.001, respectively) (Table 2).

Table 2. Clinical and biochemical variables associated with bacterial infection at admission.
Table 2
Continuous variables are expressed as median (interquartile range, IQR) and categorical variables as number (percentage). Significance between groups: *P < 0.05; **P < 0.01 versus healthy controls; +P < 0.01; ++P < 0.001 versus decompensated cirrhosis without infection.
INR, normalized international ratio; MELD, model for end stage liver disease; WBC, white blood cells.

Serum concentrations of endocan, procalcitonin, C-reactive protein and inflammatory cytokines in cirrhotic patients according to the presence of bacterial infection

Serum endocan levels have been reported to vary in healthy subjects, and our results (median, 0.95 ng/mL; range, 0.0–1.5 ng/mL) were somewhat higher than those reported in previous studies (median, 0.3 – 0.77 ng/mL; range 0.0 – 1.0 ng/mL) (9, 11, 25). The cut-off value for endocan was established as 2.05 ng/mL. The patients’ results are presented in Figs. 1-5 and the statistical analyses are summarized in Table 3. Serum endocan (Fig. 1), CRP (Fig. 2), PCT (Fig. 3) and levels of TNF-α (Fig. 4) and IL-6 (Fig. 5) were statistically significantly higher in both decompensated cirrhotic patients with and without infection when compared with healthy subjects. Infection at admission was associated with significantly higher median levels of endocan (P < 0.001), TNF-α (P < 0.001), CRP (P < 0.0001) and PCT (P < 0.0001) as compared to the uninfected patients with both compensated and decompensated cirrhosis (Table 3). IL-6 levels in serum were similar between groups. All decompensated patients with a bacterial infection had endocan concentrations > 2.05 ng/mL. Additionally, there was no overlap between the decompensated cirrhosis with bacterial infection and the compensated cirrhosis and healthy controls (Fig. 1). There was only a small overlap between the levels of endocan (Fig. 1) and PCT (Fig. 3) in the infected group of patients and that of endocan and PCT in the uninfected group of decompensated patients. Although serum CRP levels were significantly higher in the decompensated patients with infection than in the uninfected patients, this biomarker was not a good discriminator because of the considerable overlap among all groups (Fig. 2). Meanwhile, there was no overlap between the levels of TNF-α in the infected patients and that of TNF-α in the uninfected groups of cirrhotic patients, as can be seen from Fig. 4. However, the non-stability at room temperature similar to that of other cytokines limits the prognostic significance of TNF-α determination in bacterial infection. The patients with compensated cirrhosis showed high circulating levels of endocan and cytokines but the median of this group was not significantly greater than for the control group (endocan; 1.98 ng/mL (0.3 – 2.78 ng/mL) versus 0.95 ng/mL (0.0 – 1.5 ng/mL). Although levels of endocan, CRP, PCT and cytokines were higher in decompensated group without infection when compared with compensated group, the difference was not statistically significant (Table 3). At admission 30 (59%) cirrhotic patients with infection had spontaneous bacterial peritonitis (SBP) (Table 1). Endocan concentrations in cirrhotic patients with SBP were not different to endocan concentrations measured in patients with other types of bacterial infections (urinary tract infection, pneumonia, skin infection) (data not shown).

Figure 1
Fig. 1. Serum concentrations of endocan in cirrhotic patients according to the presence of bacterial infection. Horizontal bars represent medians of the concentrations in the different figures.
 
Fig. 2. Serum concentrations of C-reactive protein (CRP) in cirrhotic patients according to the presence of bacterial infection. Horizontal bars represent medians of the concentrations in the different figures.
Figure 2
Fig. 3. Serum concentrations of procalcitonin (PCT) in cirrhotic patients according to the presence of bacterial infection. Horizontal bars represent medians of the concentrations in the different figures.
Table 3. Comparison between cirrhotic patients classified according to the presence of bacterial infection.
Table 3
Continuous variables are expressed as median (interquartile range; IQR).
Significance between groups: *P < 0.01; **P < 0.001; ***P < 0.0001 versus healthy controls; +P < 0.001; ++P < 0.0001 versus decompensated cirrhosis without infection; #P < 0.001; ##P < 0.0001 versus compensated cirrhosis without infection.
CRP, C-reactive protein; IL, interleukin; PCT, procalcitonin; TNF-α, tumor necrosis factor-α.
Figure 4
Fig. 4. Serum concentrations of tumor necrosis factor-a (TNF-α) in cirrhotic patients according to the presence of bacterial infection. Horizontal bars represent medians of the concentrations in the different figures.
Figure 5
Fig. 5. Serum concentrations of interleukin-6 (IL-6) in cirrhotic patients according to the presence of bacterial infection. Horizontal bars represent medians of the concentrations in the different figures.

In patients with decompensated cirrhosis regardless whether infection was present or not, endocan levels were not correlated with WBC (P = 0.61), PCT (P = 0.28) or with IL-6 (P = 0.21). C-reactive protein (r = 0.36; P < 0.05) and TNF-α (r = 0.42, P < 0.01) levels and Child-Pugh score (r = 0.48, P < 0.01) correlated only in the infected group of patients with endocan levels, while in the uninfected groups of cirrhotic patients no significant correlation could be detected. This was also true for compensated patients without infections at admission were no correlation of the investigated biomarkers with endocan could be detected. These results indicate that the elevation in endocan levels was greater than that of PCT and IL-6, and endocan was parameter that was independent of the PCT and IL-6 levels in patients with decompensated cirrhosis.

The diagnostic performance of endocan for identification of bacterial infection in decompensated cirrhotic patients

Using biomarkers at admission, receiver operator characteristic (ROC) analysis showed that areas under the ROC curve (AUC) of endocan, PCT, CRP and TNF-α were 0.832, 0.734, 0.584, 0.599, respectively. The serum endocan level had a specificity of 85%, a sensitivity of 76.1%, positive predictive value of 95.9%, and negative predictive value of 43.7% at the cut-off level of 2.05 ng/mL (Table 4). Furthermore, the serum PCT level had a specificity of 81.2%, a sensitivity of 69.9%, positive predictive value of 91.5%, and negative predictive value of 42% at the cut-off level of 0.5 pg/mL. The endocan-assay was the most powerful assay for early diagnosis of infection in decompensated cirrhotic patients.

Fig. 6 shows ROC curves of endocan for identification of patients with infection according to disease severity indicated by Child-Pugh stage. The accuracy of endocan for diagnosis of infection in cirrhotic patients increased in advanced liver disease. The diagnostic accuracy of endocan for identifying patients with infection was the best for Child C stage cirrhosis: AUC, (95%CI): 0.917 (0.880 – 0.939). Endocan showed a slightly lower accuracy in patients with Child B: AUC, (95%CI): 0.900 (0.857 – 0.929), whereas the performance of endocan in patients with Child A: AUC, (95%CI): 0.794 (0.761 – 0.817), was clearly inferior to the values of endocan for identification of infected patients with Child B or C stage cirrhosis.

Table 4. The accuracy of endocan, PCT, CRP and TNF-α for the diagnosis of bacterial infection in decompensated cirrhotic patients.
Table 4
The optimal cut-off value was calculated from the ROC analysis for endocan, PCT, CRP, and TNF-α and subsequently the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of the markers was calculated.
CRP, C-reactive protein; IL, interleukin; PCT, procalcitonin; TNF-α, tumor necrosis factor-α.
Figure 6
Fig. 6. Receiver operating characteristics (ROC) curves of endocan for identification of cirrhotic patients with bacterial infection according to disease severity indicated by Child-Pugh stage.

Using the DeLong method (27), pairwise comparison of ROC curves was performed. There were significant differences between the AUC of endocan in patients with Child A and that of endocan in patients with Child C (P = 0.019) for identification of bacterial infection. There was no difference between the area under the ROC curves of endocan for identification of bacterial infection AUC = 0.794 (95%CI 0.761 – 0.817) versus 0.900 (95%CI 0.857 – 0.929; P = 0.592), in cirrhotic patients with Child A and Child B, respectively. There were also no significant differences between the AUC of endocan in patients with Child B and that of endocan in patients with Child C (P = 0.094) (Fig. 6) for identification of infection, but the AUC of endocan in patients with Child C was higher than that of endocan in patients with Child B (0.917 versus 0.900).

DISCUSSION

Endocan is naturally expressed by endothelial cells, is highly regulated in presence of proinflammatory cytokines and proangiogenic molecules and may be considered an accurate marker of endothelial activation (7). Since endothelial injury is pivotal in the development of liver failure (28) and in sepsis (18, 19), an endothelial marker such as endocan might not only reflect the severity of liver disease but also represent a promising diagnostic marker of bacterial infection in cirrhotic patients.

The present study provides several lines of evidence to suggest that endocan acts as mediator of inflammatory state associated with bacterial infection in liver cirrhosis. First, serum endocan levels in decompensated cirrhotic patients with infections were significantly higher than in uninfected patients. Second, and more importantly, a significant positive correlation between endocan levels and both Child-Pugh score and inflammatory markers (TNF-α, CRP) was observed in cirrhotic patients with infection.

In cirrhosis, systemic inflammation, in form of activated circulating immune cells and increased serum levels of both proinflammatory cytokines (e.g. TNF-α, IL-6) (29, 30) and cell activation markers, is the result of persistent episodic activation of circulating immune cells from damage-associated molecular patterns, released from necrotic liver cells and, as infection occurs, from pathogen associated molecular patterns, released from the leaky gut (1). In our study, we found a correlation between serum endocan level and disease severity and a high level of circulating endocan was associated with TNF-α and its secondary mediator (i.e CRP) in infected cirrhotic patients. In a recent report, endocan appeared to reflect the degree of endothelial cell injury (31). Furthermore, endocan expression in primary cultured HUVECs is regulated by TNF-α, a cytokine that is known to stimulate endothelial cell activation and injury (32), although the precise mechanism of endocan expression in infection has not been elucidated. TNF-α is a known attractant for leukocytes, enhances expression of adhesion molecules on endothelial cells, and, therefore, may play an important role in hepatic inflammatory responses and cirrhosis progress. In addition, the proinflammatory activities of endocan on endothelial cells is mediated by in part by local release of TNF-α (25), which may induces a systemic release of CRP by stimulation of IL-6. Even if there was no correlation between endocan and IL-6, we demonstrated weak association with inflammatory CRP in cirrhotic patients. Although endocan has not been found to be specific for any systemic inflammatory diseases, it is known to mediate recruitment of circulating lymphocytes and monocytes to inflammatory sites (7). As a consequence, these effects of endocan on inflammation status may cause the deterioration of hepatic function in patients with advanced cirrhosis and bacterial infection. Collectively, elevated endocan may reflect degree of endothelial cell injury induced by a systemic inflammatory response, a pathologic process that could modify the course of advanced liver failure (33).

Several studies suggested a possible role of endocan in vascular contribution to organ-specific inflammation and in endothelium-dependent pathological disorders (26, 34, 35). Some other studies also suggest that an elevated serum endocan level can be used to predict future or worsening hepatic decompensation and consequent mortality (36, 37, 38). As hepatocellular carcinoma often develops in cirrhotic liver, death from the underlying disease constitutes a competitive risk of death from the tumor. In this context, the combined use of the Child-Pugh score and the serum endocan level enables better prognostic stratification of patients (36). Meanwhile, in our patients with cirrhosis, the MELD score was significantly related to the occurrence of bacterial infection, indicating that the hepatic decompensation was associated with an increased risk of infection. Accordingly, the early diagnosis of bacterial infection in decompensated patients with cirrhosis remain a major challenge, where time plays a crucial role. Some unique characteristics of these patients make the diagnosis of bacterial infections challenging. For example, the presence of leukocyturia does not always correlate with urinary tract infection and the diagnosis of spontaneous bacteremia can only be established once the results of blood cultures are received; since dyspnea, as the predominant presenting symptom in a pulmonary complication of advanced cirrhosis, is frequent in the subclinical forms of hepatopulmonary syndrome (39), difficulties exists to diagnose pneumonia. Thus, an early diagnosis of bacterial infection would certainly help to initiate adequate and timely antibiotics and would possibly improve outcome of decompensated patients with cirrhosis. Antibiotics can successfully modify the sequence linking alterations in gut microbiota and intestinal permeability with bacterial translocation and pro-inflammatory state, mainly through their effect on intestinal microbiota. However, infections caused by multiresistant bacteria are a growing threat in patients with advanced liver cirrhosis. As previous antibiotic administration is strongly related to the development of such infections, alternatives to antibiotics are urgently needed in the prevention of bacterial translocation and its consequences. The evidence is mounting that probiotic treatment through modulation of intestinal microbiota have the potential to affect the course of advanced liver disease (40). It has been observed that VSL#3 probiotic therapy was well tolerated and led to improvement of several clinical parameters in patients with decompensated cirrhosis (41). Interestingly, none of the cirrhotic patients developed bacterial infection, which would require administration of antibiotics. Therefore, the surrogate marker for bacterial infection which also acts as a guide to the effectiveness of therapy is needed. Thus, in the constant search for biomarkers of bacterial infection, endocan has arisen as an attractive candidate (16-19).

Although, increased levels of endocan in advanced liver disease had been previously reported (36, 37, 42), its potential value as a diagnostic tool has not been studied. To our knowledge, ours is the first study in which endocan levels revealed significant associations with the severity of liver disease within the infected group, as demonstrated by significant correlations with the Child-Pugh score. In addition, the accuracy of endocan for diagnosis of infection in cirrhotic patients increased in advanced liver disease. Diagnostic accuracy of endocan for identifying infected patients was the best for Child C stage cirrhosis (AUC, 0.917), reaching a sensitivity of 81.8%.

Recently Mosevoll et al. (43) demonstrated that the plasma levels of endocan show a considerable overlap when comparing healthy subjects, patients with suspected thrombosis and the patients without verified thrombosis. Therefore, they suggested that the plasma endocan alone do not seem to be useful in the diagnostic evaluation of these patients, but it may become useful in combination with other markers. In our study, a relatively small overlap of endocan concentrations existed between the 8 decompensated patients without infection and the 18 patients with bacterial infections, which had no a negative impact on the specificity of the test. Additionally, the best cut-off level for endocan in predicting infection in patients with decompensated cirrhosis was superior in its diagnostic accuracy compared with the optimum PCT, CRP and TNF-α cut-off levels. These results indicate that serum endocan alone may be useful marker for diagnosing bacterial infection in cirrhotic patients with a high level of specificity (85%) and sensitivity (76.1%). This proteoglycan differs from other biomarkers of bacterial infection. Serum endocan levels are detectable as early as 2 hours after an inflammatory response begins (23), which is earlier than the increase in the PCT or CRP (44, 45). Endocan shows kinetic properties that enable it to serve as an early as well as a follow-up (72 hours and beyond) marker of inflammation and infection (9). Endocan is still detectable when screened on a daily basis, whereas CRP or PCT will have already disappeared from the blood. Our results suggest that serum endocan is an independent parameter capable of distinguishing infected from uninfected decompensated patients with cirrhosis and that it is useful biomarker for early diagnosis of infection in these patients with a high level of specificity and sensitivity.

To understand our findings better, some limitations should be considered. First, although the number of patients enrolled might seem small, it adequately represents the sample size estimated to provide the specific power. Second, this study was not designed and powered to assess the ability of endocan to predict the incidence of bacterial infections in patients without overt infections. Lastly, we excluded some diseases that may influence endocan levels; however, some diseases may be unrecognized in our study group.

Life expectancy of patients with liver cirrhosis is closely linked to the degree of liver dysfunction and the occurrence of bacterial infection. Our study identified serum endocan as a powerful diagnostic marker to assess the severity of liver disease and cirrhotic patients with bacterial infection. It may be useful to implement endocan in future diagnostic algorithms for gauging the prognosis of patients with advanced liver disease, and larger prospective studies should investigate the practical clinical value of serum endocam measurements.

Acknowledgments: The project described was supported by a grant from the Wroclaw Medical University No. ST-D020.16.004.

Conflict of interests: None declared.

REFERENCES

  1. Albillos A, Lario M, Alvarez-Mon M. Cirrhosis-associated immune dysfunction: distinctive features and clinical relevance. J Hepatol 2014; 61: 1385-1396.
  2. Whang KT, Vath SD, Becker KL, et al. Procalcitonin and proinflammatory cytokine interactions in sepsis. Shock 2000; 14: 73-78.
  3. Le Moine O, Deviere J, Devaster JM, et al. Interleukin-6: an early marker of bacterial infection in decompensated cirrhosis. J Hepatol 1994; 20: 819-824.
  4. Navasa M, Follo A, Filella X, et al. Tumor necrosis factor and interleukin-6 in spontaneousbacterial peritonitis in cirrhosis: relationship with the development of renal impairment and mortality. Hepatology 1998; 27: 1227-1232.
  5. Propst T, Propst A, Herold M, et al. Spontaneous bacterial peritonitis is associated with high levels of interleukin-6 and its secondary mediators in ascitic fluid. Eur J Clin Invest 1993; 23: 832-836.
  6. Xing K, Murthy S, Liles WC, Singh JM. Clinical utility of biomarkers of endothelial activation in sepsis-a systematic review. Crit Care 2012; 16: R7. doi: 10.1186/cc11145
  7. Bechard D, Scherpereel A, Hammad H, et al. Human endothelial-cell specific molecule-1 binds directly to the integrin CD11a/CD18 (LFA-1) and blocks binding to intercellular adhesion molecule-1. J Immunol 2001; 167: 3099-3106.
  8. Kali A, Shetty KS. Endocan: a novel circulating proteoglycan. Indian J Pharmacol 2014; 46: 579-583.
  9. Grigoriu BD, Depontieu F, Scherpereel A, et al. Endocan expression and relationship with survival in human non-small cell lung cancer. Clin Cancer Res 2006; 12: 4575-4582.
  10. Maurage CA, Adam E, Mineo JF, et al. Endocan expression and localization in human glioblastomas. J Neuropathol Exp Neurol 2009; 68: 633-641.
  11. Leroy X, Aubert S, Zini L, et al. Vascular endocan (ESM-1) is markedly overexpressed in clear cell renal cell carcinoma. Histopathology 2010; 56: 180-187.
  12. Chen LY, Liu X, Wang SL, Qin CY. Over-expression of the Endocan gene in endothelial cells from hepatocellular carcinoma is associated with angiogenesis and tumour invasion. J Int Med Res 2010; 38: 498-510.
  13. Kim JH, Park MY, Kim CN, et al. Expression of endothelial cell-specific molecule-1 regulated by hypoxia inducible factor-1a in human colon carcinoma: impact of ESM-1 on prognosis and its correlation with clinicopathological features. Oncol Rep 2012; 28: 1701-1708.
  14. Roudnicky F, Poyet C, Wild P, et al. Endocan is upregulated on tumor vessels in invasive bladder cancer where it mediates VEGF-A-induced angiogenesis. Cancer Res 2013; 73: 1097-1106.
  15. El Behery MM, Seksaka MA, Ibrahiem MA, Saleh HS, El Alfy Y. Clinicopathological correlation of endocan expression and survival in epithelial ovarian cancer. Arch Gynecol Obstet 2013; 288: 1371-1376.
  16. Seo K, Kitazawa T, Yoshino Y, Koga I, Ota Y. Characteristics of serum endocan levels in infection. PLoS One 2015; 10: e0123358. doi: 10.1371/journal.pone.0123358
  17. Scherpereel A, Depontieu F, Grigoriu B, et al. Endocan, a new endothelial marker in human sepsis. Crit Care Med 2006; 34: 532-537.
  18. Mihajlovic DM, Lendak DF, Brkic SV, et al. Endocan is useful biomarker of survival and severity in sepsis. Microvasc Res 2014; 93: 92-97.
  19. Pauly D, Hamed S, Behnes M, et al. Endothelial cell-specific molecule-1/endocan: Diagnostic and prognostic value in patients suffering from severe sepsis and septic shock. J Crit Care 2016; 31: 68-75.
  20. Hatfield KJ, Lassalle P, Leiva RA, et al. Serum levels of endothelium-derived endocan are increased in patients with untreated acute myeloid leukemia. Hematology 2011; 16: 351-356.
  21. Akkok CA, Hervig T, Stamnesfet S, et al. Effects of peripheral blood stem cell apheresis on systemic cytokine levels in patients with multiple myeloma. Cytotherapy 2011; 13: 1259-1268.
  22. Lindas R, Tvedt TH, Hatfield KJ, Reikvam H, Bruserud O. Preconditioning serum levels of endothelial cell-derived molecules and the risk of posttransplant complications in patients treated with allogeneic stem cell transplantation. J Transplant 2014; 2014: 404096. doi: 10.1155/2014/404096
  23. Lassalle P, Molet S, Janin A, et al. ESM-1 is a novel human endothelial cell-specific molecule expressed in lung and regulated by cytokines. J Biol Chem 1996; 271: 20458-20464.
  24. Sarrazin S, Adam E, Lyon M, et al. Endocan or endothelial cell specific molecule-1 (ESM-1): a potential novel endothelial cell marker and a new target for cancer therapy. Biochim Biophys Acta 2006; 1765: 25-37.
  25. Lee W, Ku SK, Kim SW, Bae JS. Endocan elicits severe vascular inflammatory responses in vitro and in vivo. J Cell Physiol 2014; 229: 620-630.
  26. Albers I, Hartmann H, Bircher J, Creutzfeldt W. Superiority of the Child-Pugh classification to quantitative liver function tests for assessing prognosis of liver cirrhosis. Scand J Gastroenterol 1989; 24: 269-276.
  27. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988; 44: 837-845.
  28. Ceciliani F, Giordano A, Spagnolo V. The systemic reaction during inflammation: the acute-phase proteins. Protein Pept Lett 2002; 9: 211-223.
  29. Albillos A, Hera Ad Ade L, Reyes E, et al. Alvarez-Mon M. Tumour necrosis factor-αlpha expression by activated monocytes and altered T-cell homeostasis in ascitic alcoholic cirrhosis: amelioration with norfloxacin. J Hepatol 2004; 40: 624-631.
  30. Munoz L, Albillos A, Nieto M, et al. Alvarez-Mon M. Mesenteric Th1 polarization and monocyte TNF-αlpha production: first steps to systemic inflammation in rats with cirrhosis. Hepatology 2005; 42: 411-419.
  31. Su YH, Shu KH, Hu CP, et al. Serum Endocan correlated with stage of chronic kidney disease and deterioration in renal transplant recipients. Transplant Proc 2014; 46: 323-327.
  32. Li S, Wang L, Wang C, et al. Detection on dynamic changes of endothelial cell specific molecule-1 in acute rejection after renal transplantation. Urology 2012; 80: 738. doi: 10.1016/j.urology.2012.03.019
  33. Iwakiri Y. Endothelial dysfunction in the regulation of cirrhosis and portal hypertension. Liver Int 2012; 32: 199-213.
  34. Balta I, Balta S, Koryurek OM, et al. Serum endocan levels as a marker of disease activity in patients with Behcet disease. J Am Acad Dermatol 2014; 70: 291-296.
  35. Ziol M, Sutton A, Calderaro J, et al. ESM-1 expression in stromal cells is predictive of recurrence after radiofrequency ablation in early hepatocellular carcinoma. J Hepatol 2013; 59: 1264-1270.
  36. Nault JC, Guyot E, Laguillier C, et al. Serum proteoglycans as prognostic biomarkers of hepatocellular carcinoma in patients with alcoholic cirrhosis. Cancer Epidemiol Biomarkers Prev 2013; 22: 1343-1352.
  37. Toshikuni N, Ozaki K, George J, Tsutsumi M. Serum endocan as a survival predictor for patients with liver cirrhosis. Can J Gastroenterol Hepatol 2015; 29: 427-430.
  38. Ozaki K, Toshikuni N, George J, et al. Serum endocan as a novel prognostic biomarker in patients with hepatocellular carcinoma. J Cancer 2014; 5: 221-230.
  39. Boryczka G, Hartleb M, Rudzki K, Janik MA. Influence of an upright body position on the size of intrapulmonary blood shunts in patients with advanced liver cirrhosis. J Physiol Pharmacol 2015; 66: 855-861.
  40. Gupta N, Kumar A, Sharma P, et al. Effects of the adjunctive probiotic VSL#3 on portal haemodynamics in patients with cirrhosis and large varices: a randomized trial. Liver Int 2013; 33: 1148-1157.
  41. Marlicz W, Wunsch E, Mydlowska M, et al. The effect of short term treatment with probiotic VSL#3 on various clinical and biochemical parameters in patients with liver cirrhosis. J Physiol Pharmacol 2016; 67: 867-877.
  42. Tok D, Ekiz F, Basar O, Coban S, Ozturk G. Serum endocan levels in patients with chronic liver disease. Int J Clin Exp Med 2014; 7: 1802-1807.
  43. Mosevoll KA, Lindas R, Wendelbo O, Bruserud O, Reikvam H. Systemic levels of the endothelium-derived soluble adhesion molecules endocan and E-selectin in patients with suspected deep vein thrombosis. Springerplus 201; 3: 571. doi: 10.1186/2193-1801-3-571
  44. Dandona P, Nix D, Wilson MF, et al. Procalcitonin increase after endotoxin injection in normal subjects. J Clin Endocrinol Metab 1994; 79: 1605-1608.
  45. Pepys MB, Hirschfield GM. C-reactive protein: a critical update. J Clin Invest 2003; 111: 1805-1810.
R e c e i v e d : February, 7, 2017
A c c e p t e d : April 10, 2017
Author’s address: Assoc. Prof. Jolanta Zuwala-Jagiello, Department of Pharmaceutical Biochemistry, Wroclaw Medical University, 211 Borowska Street, 50-556 Wroclaw, Poland. E-mail: jolanta.zuwala-jagiello@umed.wroc.pl