Objectives: Liver function is affected by ischemia-reperfusion. Ischemia-reperfusion injury to the liver often follows hepatobiliary surgery. Here, we investigated biomarkers of liver ischemia-reperfusion injury using an animal model.
Materials and Methods: For this study, 24 male Sprague Dawley rats (146-188 g) were divided into 4 groups: group A was the control group, group B was the partial hepatic ischemia-reperfusion group, group C was the total hepatic ischemia-reperfusion group, and group D was the intermittent total hepatic ischemia-reperfusion group. Laboratory liver function levels were measured before ischemia, after ischemia, and after reperfusion. We used liver and renal biopsies for histopathological examination at the end of the study.
Results: After clamping and reperfusion, alanine aminotransferase and cystatin C levels in groups B, C, and D were significantly higher than levels in group A. In group B, after clamping, neutrophil gelatinase-associated lipocalin levels were higher than in groups A and D, with significantly higher level than in group D after reperfusion. Neutrophil gelatinase-associated lipocalin levels decreased significantly in groups B, C, and D after reperfusion. There was significantly greater hepatic damage in groups B, C, and D compared with group A but no significant differences in renal injury scores among the groups. There was a significant positive correlation between hepatic damage and renal injury. With regard to histopathological examination versus laboratory results, a statistically significant positive correlation was shown between grade of hepatic damage and serum alanine aminotransferase and cystatin C levels. Similarly, there was a positive correlation between renal damage score and alanine aminotransferase level.
Conclusions: In our animal model, alanine aminotransferase and cystatin C levels tended to increase with ischemia-reperfusion injury levels but neutrophil gelatinase-associated lipocalin decreased during reperfusion. In liver ischemia, we suggest that neutrophil gelatinase-associated lipocalin may be an important biomarker for distinguishing the reperfusion phase.
Key words : Cystatin C, Hepatobiliary surgery, Liver, Neutrophil gelatinase-associated lipocalin
Ischemia-reperfusion injury (IRI) is a common problem that is encountered in clinical practice during hepatectomies, liver transplantation, and hypovolemic shock. Ischemia-reperfusion injury is an inevitable complication of liver transplantation following restoration of perfusion after ischemic insult1 and can compromise liver function, leading to liver failure after hepatectomy and liver transplant. Thus, finding ways to reduce IRI is necessary to maintaining liver function.1,2
Neutrophil gelatinase-associated lipocalin (NGAL) is a member of the small extracellular protein family named lipocalin. Lipocalin family members were previously referred to as transport proteins, such as the retinol-binding and the fatty acid-binding protein, but were subsequently discovered to have multiple functions, including roles in olfaction, pheromone transfer, and prostaglandin synthesis.3 Neutrophil gelatinase-associated lipocalin is a monomeric protein synthesized from cells under stress and can be used as a biomarker for acute renal failure.4 Infection, inflammation (secreted from the secondary granules from activated neutrophils), ischemia, and neoplastic transformation increase the expression of NGAL.5 Many different tissues produce NGAL, including fat cells, the liver, lung, kidney, thymus, intestine, breast tissue, macrophages, and neutrophils. The production of NGAL in the liver, fat tissue, and macrophages is stimulated by proinflammatory activity.6
Cystatin is a 122-amino acid-long cysteine proteinase inhibitor that is produced at a constant rate in all cells with a nucleus. Because of its low molecular weight and basic structure, cystatin is filtrated easily from the glomerulus to be totally absorbed and catabolized. Cystatin, which is freely filtrated and not affected by body mass, is distinguished as a sensitive parameter for evaluation of the glomerular filtration rate. In addition, cystatin C serum levels are not altered by age or sex.7,8
Ischemic injury is a localized process resulting from cellular metabolic deficiencies, such as lack of oxygen support, glycogen consumption, and ATP depletion, resulting in cell death of the parenchyma. However, reperfusion injury after ischemia has both metabolic deficiencies and a strong inflammatory process, which ends in cytotoxicity.9 Advances in molecular biology have helped to advance our understanding of the underlying mechanisms of the liver, but more remains to be discovered. Hepatic IRI includes both warm and cold IRI. These 2 types of IRI share similar pathophysiological processes that start with anaerobic metabolism and involve mitochondria, oxidative stress, intracellular calcium overload, liver cells, neutrophils, cytokines, and chemokines.10 As the blood supply to the liver decreases, the metabolic pattern shifts from an aerobic to a nonaerobic state, thereby depleting the adenosine triphosphate molecules inside the cell and causing metabolic acidosis. During the beginning of the process, this decrease in pH plays a protective role against cell death. However, once reperfusion starts and pH is normalized, the latent proteases and phospholipases become activated, causing cell damage.11 The deprivation of ATP slows the ATP-dependent sodium-potassium pump, and this results in an increase of intracellular sodium ions followed by swelling of Kupfer cells, which are the key cells in IRI initiation.12 These cells, which are activated during early stages of reperfusion, release a large amount of both proinflammatory and anti-inflammatory mediators and reactive oxygen species.12,13
In this study, using a liver IRI animal model, we investigated the histopathological changes that take part in the liver and kidneys during IRI. We analyzed alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin, NGAL, and cystatin C levels in rat sera. Our aim was to investigate biomarkers related to IRI in the liver.
Materials and Methods
This study received approval from our animal ethics committee. All experimental procedures complied with the Guide for the Care and Use of Laboratory Animals. Rats were used to create the liver IRI model. To define a significant difference in the probability with 95% strength, 24 animals were required consisting of 40-day-old rats with midsize effect (effect size = 1.0). The 24 male Sprague Dawley rats (weighing 146-188 g) were divided into 4 groups of 6 rats per group (groups A to D).
Surgical procedures and experimental design
Briefly, after 24 hours of fasting, animals were anesthetized with pentobarbital sodium (60 mg/kg via intraperitoneal route) and were placed below a heating lamp to maintain constant temperature 37 °C). An identical midline abdominal incision was performed in each operation. After completion of procedures particular to each group, rats were humanely killed.
Group A, the control-sham group, received a midline laparotomy, with liver liberated and hepatic pedicle exposed and no further surgical intervention. Two hours later, a blood sample and liver and kidney biopsies were obtained.
Group B, the partial hepatic ischemia group, received a midline laparotomy, with liver liberated from its ligaments and subsequently all structures of the portal triad of the left and median hepatic lobes occluded for 60 minutes with a microvascular clamp (Aesculap). In group B rats, mesenteric congestion was prevented by allowing intestinal blood flow through the right and caudate lobes.14 Blood flow was stopped with a clamp, which was opened after 60 minutes to release the blood flow. At the time of the release of flow, at 60 and 120 minutes, a total of 3 blood samples were obtained. Finally, tissue samples were obtained from the kidney and liver. For this group, the aim was to achieve ischemia up to 70%.
Group C, the total hepatic ischemia group, also received a midline laparotomy. After the hepatic hilum was identified, complete warm hepatic ischemia was induced by the Pringle maneuver15 using microvascular bulldog clamps. Ischemia was noticed by color changes in the liver and intestinal tissue. Hepatic ischemia was maintained for 20 minutes, and then clamps were removed to allow reperfusion with a duration of 2 hours. Hence, at the time of release of flow and at 120 minutes thereafter, 2 blood samples were obtained. Finally, tissue samples from the liver and kidneys were taken at the time when the second blood sample was obtained (at 120 min). In group C, portal circulation was fully blocked. Thus, intestinal ischemia was visible based on changes in intestinal color.
Group D, the intermittent or reperfusion hepatic ischemia group, also received a midline laparotomy. Intermittent ischemia was maintained for 20 minutes by clamping (Pringle maneuver). The pedicle was liberated for reperfusion of 20 minutes, and then a second Pringle maneuver with a duration of 20 minutes was conducted. At the end of that second Pringle maneuver and after 2 hours of release of flow, 2 blood samples were obtained. Finally, tissue samples were taken at the time when the second blood sample was obtained (at 120 min).
Bilirubin, ALT, NGAL, and cystatin C levels were analyzed from obtained blood samples. Total bilirubin and ALT were analyzed using an AU 680 Biochemistry Analyzer (Beckman Coulter). On the same day, NGAL and cystatin C were analyzed manually by enzyme-linked immunosorbent assay using an analyzer from Bioassay Technology Laboratory. Results were expressed in ng/mL for NGAL and cystatin C, mg/dL for total bilirubin, and U/L for ALT. Kidney and liver pathology specimens were examined to scale level of ischemia.
Tissue sampling and histopathological examination
Kidney specimens were fixed in a 10% formaldehyde solution. The sections were examined for complete-ness, and 1 representative section of each kidney was selected for tissue processing. Samples were cut into slices, stained with hematoxylin-eosin dye, and examined under a light microscope by a pathologist who was blinded to the study design. All sections were examined for evidence of tissue injury. Results were evaluated with the use of the Endothelial, Glomerular, Tubular, and Interstitial (EGTI) scoring system.16 Tubules, glomeruli, tubulointerstitial area, and endothelial cells were evaluated, and the degree of injury was scored from 0 to 3.
Liver specimens were fixed in 10% formalin. Blocks were sliced at 4-mm thickness in paraffin, dyed with hematoxylin-eosin, and examined under a light microscope. Hepatic histological damage and hepatocellular necrosis were evaluated according to scales described by Shen and colleagues17 and Chen and colleagues,18 respectively. Grade 0 indicated no or minimal hepatic injury; grade 1 indicated light hepatic damage, which includes vacuolization of the cytoplasm and focal nuclear pyknosis; grade 2 indicated midlevel liver damage, which includes extensive nuclear pyknosis, the loss of intercellular borders, and light or medium level neutrophil infiltration; and grade 3 indicated severe hepatic damage, which includes distortion in the hepatic chords, neutrophil clusters, and apoptotic bodies.
We used SPSS for Windows version 15.0 for all statistical analyses. The descriptive statistics are presented as mean, median, and standard deviation. When parametric prerequisites were not met, the Kruskal Wallis test was preferred in the independent groups. For subgroup analysis, the Mann-Whitney U test with Bonferroni correction was used. In dependent groups, the Wilcoxon test was used. The relationships of ordinal variable and numerical variables were analyzed using Spearman correlation analysis. P < .05 indicated statistical significance.
We observed no significant difference in renal injury EGTI scores among the groups. Lowest and highest EGTI scores in all groups were 0 and 2, respectively. With regard to hepatic damage grade, a statistically significant difference was found among the groups (P = .002), with grades 0 and 1 shown in group A but increased grades 2 and 3 shown in group D (Table 1). Hepatic damage grade was significantly higher in groups B, C, and D than in group A (P = .005, P = .008, and P = .002, respectively) but not statistically different among groups B, C, and D (Table 2). There was a significant positive correlation between hepatic damage and renal injury (P = .023; Table 3).
Among all groups, when we looked at the relationship between histopathological examination and laboratory results for the whole study, there was a statistically significant positive correlation between grade of hepatic damage and serum ALT and cystatin C levels (P = .001, P = .013). Similarly, there was a positive correlation between renal injury EGTI score and ALT (P = .036). Differences between groups with regard to ALT, NGAL, and cystatin C levels after clamping and reperfusion were significant (P = .018, P < .001, P = .005, P = .001, P = .003, P< .001, and P = .003; Table 4). After the clamping procedure, total bilirubin level was significantly lower in group C than in group A (P = .005). Although there was no significant difference among groups after reperfusion, mean level of total bilirubin in group D was significantly higher than in group A (P = .005; Table 5). For total bilirubin, in group D, we observed a significant increase in postreperfusion levels compared with postclamping levels (P < .001; Table 4).
After clamping and reperfusion in groups B, C, and D, ALT levels were significantly higher compared with level shown in the control group (group A; Table 5). After reperfusion in groups B, C, and D, ALT levels were significantly higher than levels shown after clamping (P = .002, P <.001, and P = .005, respectively; Table 4).
After clamping, the NGAL level in group B was higher than levels in group A and group D. After reperfusion, mean NGAL level in group B was significantly higher than level observed for group D (Table 5). After reperfusion, there was no significant difference in mean NGAL levels versus reperfusion and postclamping levels (P > .05; Table 4).
The mean postclamping and postreperfusion cystatin C levels in groups B and C were significantly higher than levels shown in control group A (P = .004; Table 5). In group B, cystatin C levels were higher in the postreperfusion period than in the postclamping period (P = .006; Table 4).
Alanine acts as a catalyzer in the reversible transamination of L-alanine and 2-oxoglutarate to pyruvate and L-glutamate. Alanine aminotransferase, which is found abundantly in the cytosol of hepatocytes, during hepatocellular injury or death, is released from damaged liver cells. This leads to an increase in measured ALT activity in the serum.19
Compared with the control group (group A), ALT levels were significantly higher in groups B, C, and D (which all underwent arterial blockage to induce hepatic ischemia) both after clamping and after reperfusion. There were statistically significant differences among the groups with regard to values shown with ischemia and reperfusion. This finding suggests that ALT levels correlated with the portion of liver damaged and the severity of IRI. On the other hand, just as in hepatorenal syndrome,hepatic damage also causes renal damage. This is why there was a positive correlation in our study between ALT elevation and renal injury. Alanine aminotransferase levels have been found to correlate with the degree of hepatic injury in patients with nonalcoholic fatty liver disease.20 In parallel to this, acetaminophen hepatotoxicity is one of the rare liver injuries that can raise serum aminotransferases to greater than 10 000 IU/L. Patients who fall above, defined as AST > 1000 IU, are at risk of severe hepatotoxicity.21
During both the ischemia and reperfusion phases, as shown in Table 4, both ALT and total bilirubin levels had increased in our animal model. Thus, these provided us with no data on the reperfusion phase.
High levels of bilirubin may be expected because of deterioration in the elimination in acute ischemia. Although we encountered higher levels of bilirubin in group D, these levels were lower in group B, which had partial ischemia. This finding suggests a continuation of bilirubin elimination following partial ischemia. There were no differences between bilirubin levels obtained after ischemia and reperfusion, which may have been as a result of insufficient time between 2 samplings.
We observed no significant correlation between degree of hepatic or renal damage and bilirubin level. This could indicate that hepatocyte damage rather than cholestatic toxicity developed in our IRI experimental model.
Bilirubin has been found to be elevated in hepatic failure, and this finding has been used in various scoring systems, including the Model for End-Stage Liver Disease (MELD) and Child-Pugh scoring systems, which are used to classify cirrhotic patients according to risk group.22 In addition, in a retrospective analysis of 65 patients who were diagnosed with acute-on-chronic liver failure, high levels of bilirubin were found to help predict short-term mortality and used as a biochemical marker to improve discrimination of high-risk patients with acute failure of liver versus patients with chronic liver failure.23
In our study, we encountered statistical differences among the groups for both NGAL and cystatin C levels. After ischemia, NGAL levels were lower than levels after reperfusion. However, compared with our control group, this finding was statistically significant only for group B. In the ischemia-induced groups, we detected a significant difference only when we compared group B with group D (P < .05; Table 5).
In our literature search, we did not encounter any study that addressed the relationship between NGAL levels and hepatic injury. Because of its small size, NGAL is completely filtered from the glomerulus and absorbed by proximal tubules with megalin-dependent endocytosis. Thus, in acute kidney damage, NGAL increases rapidly in the urine because of failure to absorb it from the proximal tubules.24
With liver IRI, NGAL levels may be elevated not just because of a failure in reabsorption.24-26 This may be because of the increased secretion of NGAL locally in the liver cells and from neutrophils and macrophages, which have migrated to the inflamed liver and from inflamed vessel walls, as suggested by observations of pathological specimens.27
Both NGAL and cystatin C are biomarkers that reflect acute kidney damage in the early phase and are more specific for kidney injury than serum creatinine or urea levels.24-26 Serum creatinine levels do not increase until a moderate to severe reduction in the glomerular filtration rate occurs. However, NGAL and cystatin C levels begin to rise in the early phase of kidney failure.27 Metabolical changes take place resulting from hepatocyte destruction after the blood supply to the liver is impaired. Kidney functions are also severed due to liver damage. Failure in kidneys correlates with severity of liver ischemia. High levels of NGAL and cystatin C are not related to acute renal injury, but to liver ischemia, and this finding was revealed by pathological evaluation of kidneys in our study.
Our finding is in line with findings from Ariza and colleagues, who studied NGAL in patients with acute exacerbations of chronic liver failure; the study found that both urine and plasma NGAL levels were moderately and significantly directly correlated with liver function tests, MELD score, and markers of systemic inflammation, including blood leukocytes and C-reactive protein.28
In another study, Yoshikawa and colleagues showed that blood NGAL levels were significantly correlated with liver levels, indicating that blood NGAL was derived from the liver. In patients with chronic liver disease, blood NGAL levels were also shown to be associated with liver function and renal function. The investigators also suggested that blood NGAL levels can be used as a prognostic factor in patients with chronic liver diseases.29
The relationship between liver diseases and NGAL was further demonstrated by the overexpression of NGAL in hepatocellular carcinoma tissues and the close association between the marker and proliferation and invasion of hepatocellular carcinomacells.30
In our study, we may have not have allowed sufficient time to induce ischemia, which could have accounted for failure to find elevated NGAL levels in group D (recurrent short period ischemia-induced group). In addition, there was a significant difference between NGAL levels after clamping versus levels after ischemia and reperfusion. This finding suggests that NGAL is a better biomarker for ischemia than for reperfusion injury in warm liver ischemia. The continuing elevation of ALT during reperfusion may indicate that damage and hepatocyte destruction had continued; however, NGAL levels tended to be reduced by reperfusion, indicating that the reperfusion phase had started in acute liver ischemia. This shows that NGAL is a potentially useful indicator for initiation of the reperfusion phase of liver ischemia. However, NGAL is also a biomarker of renal failure, although we demonstrated that renal damage had not yet occurred at the tested stages. Thus, these values are more likely to be related to liver damage.
Cystatin C is a reliable marker of glomerular filtration because it is synthesized in all cells and at a constant rate and is not affected by muscle mass. In addition, it may increase as a result of the inflammatory process at the inflammation site.7,8,27 When we consider that the kidney specimen was spared, we believe that this is likely to have been elevated because of inflammation in the liver and tissue necrosis.
Compared with observations in the control group, cystatin C levels were significantly increased in groups B and C, both after clamping and after reperfusion. There was also an increase in group D, but this did not reach statistical significance. We also detected a significant difference in levels after clamping when we compared group B versus group D. Cystatin C may be a suitable biomarker for IRI. In Group D, in which arterial clamping was done for 20 minutes instead of 2 hours, cystatin C was not found to be elevated; this may be because there was insufficient time for the elevation of the molecule.
In a study that investigated the relation of cystatin C versus liver or kidney findings in children with hepatic fibrosis, El-Sayed and colleagues did not observe a relationship between cystatin C and true function of the liver (albumin and prothrombin) and different stages of hepatic fibrosis.31 However, elevated cystatin C levels were found to be significantly higher in patients with cholestatic disease than in their healthy controls and may be able to discriminate between intrahepatic and extrahepatic cholestasis.32 Chu and colleagues showed that the average cystatin C concentration of patients with hepatic diseases was significantly higher, with a linear relationship with severity of disease.33 Thus, along with our study findings, cystatin C may be a useful parameter for hepatic IRI.
In our study, ALT levels increased at both the occlusion and reperfusion phases, becoming particularly high in the reperfusion phase. This occurrence may be because of the continuation of hepatocyte damage. In contrast, NGAL and cystatin C levels tended to increase in the occlusion phase (higher in groups B and C than in group D). In the reperfusion phase, NGAL tended to be lower, whereas cystatin C levels tended to be relatively high. Thus, we believe that, in warm liver ischemia, NGAL may be a good biomarker for distinction between postischemia and reperfusion damage in the liver.
DOI : 10.6002/ect.2021.0023
From the 1Department of General Surgery and the 2Department of Pediatric Surgery, University of Health Sciences, Haseki Training and Research Hospital, Istanbul; the 3Department of General Surgery, Istanbul Provincial Health Directorate, Ministry of Health, Istanbul; the 4Department of Medical Biochemistry, University of Health Sciences, Istanbul Training and Research Hospital, Istanbul; and the 5Department of Pathology, Bezmi Alem Vak?f University, Medical Faculty, Istanbul, Turkey
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Author contributions: DY, FS, MC, and MSD provided conception and design; MSD and OMA acquired and analyzed the data; MC, ZMIS, and MOG interpreted the data; FS, MC, MSD, and OMA prepared the manuscript; and DY, MSD, and HO provided critical revisions.
Corresponding author: Mahmut Said Degerli, Department of General Surgery, University of Health Sciences, Haseki Training and Research Hospital, Istanbul, Turkey
Phone: +90 536 957 86 88
Table 1. Hepatic and Renal Damage in Animal Groups
Table 2. Comparison of Hepatic Damage Grade Between Groups
Table 3. Relationship Between Histopathological Findings and Laboratory Parameters
Table 4. Total Liver Function Levels in Blood Samples From Groups
Table 5. Comparison of Laboratory Parameters Between Groups