- Email: [email protected]
Bioartificial liver attenuates intestinal mucosa injury and gut barrier dysfunction after major hepatectomy: Study in a porcine model
Liver Bioartificial liver attenuates intestinal mucosa injury and gut barrier dysfunction after major hepatectomy: Study in a porcine model Constantinos Nastos, PhD,a Konstantinos Kalimeris, PhD,b Nikolaos Papoutsidakis, PhD,b George Defterevos, PhD,a Agathi Pafiti, PhD,c Eleni Kalogeropoulou,d Loukia Zerva, PhD,d Tzortzis Nomikos, PhD,e Apostolos Papalois, PhD,f Georgia Kostopanagiotou, PhD,b Vasillios Smyrniotis, PhD,g and Nikolaos Arkadopoulos, PhD,g Athens, Greece
Background. The aim of this study was to evaluate whether bioartificial liver support can attenuate gut mucosa injury in a porcine model of posthepatectomy liver dysfunction. Methods. Posthepatectomy liver failure was induced in pigs combining major (70%) liver resection and ischemia/reperfusion injury. An ischemic period of 150 minutes was followed by reperfusion for 24 hours. Animals were divided randomly into 2 groups: a control group (n = 6) that received standard critical care and a bioartificial liver support group (Hepat, n = 6) that were subjected to extracorporeal liver support for 6 hours during reperfusion. Intestinal mucosal injury, bacterial translocation, and endotoxin translocation were evaluated in all animals. Intestinal mucosa was also evaluated with markers of oxidative injury and immunohistochemical staining for caspase 3. Results. When compared with median values, the control group, animals in the Hepat group had a lesser intestinal mucosal injury score (4.0 [range:2.0–5.0] vs 1.0 [range:1.0–2.0]; P <.01), decreased bacterial translocation in the portal and the systemic circulation at 24 hours of reperfusion (P <.05), and decreased portal and systemic endotoxin levels at 24 hours (P < .05). At 24 hours after reperfusion, mucosal protein carbonyls and malondialdehyde levels were decreased in Hepat animals (0.57 nmol/mg [range:0.32– 0.70] vs 0.33 nmol/mg [range:0.03–0.53] and 3.85 nmol/mg [range:3.01–6.43] vs 3.27 nmol/mg [range:1.46–3.55], respectively; P < .05). There was no difference in tissue caspase staining. Conclusion. Bioartificial liver support seems to attenuate intestinal mucosal injury and gut barrier dysfunction after major hepatectomy. (Surgery 2016;159:1501-10.) From the Second Department of Surgery,a School of Medicine, National and Kapodistrian University of Athens, Aretaieion University Hospital; Second Department of Anesthesiology,b School of Medicine, National and Kapodistrian University of Athens, Attikon University Hospital; Department of Pathology,c School of Medicine, National and Kapodistrian University of Athens, Aretaieion University Hospital; Laboratory of Biopathology,d Attikon University Hospital; Department of the Science Nutrition – Dietetics,e Harokopio University; Experimental Research Department of ELPEN-Pharmaf; and Fourth Department of Surgery,g School of Medicine, National and Kapodistrian University of Athens, Attikon University Hospital, Athens, Greece
DURING THE LAST FEW DECADES, improvements in surgical techniques and intensive care support have facilitated more extensive liver resections as
a treatment of primary and metastatic tumors. Posthepatectomy septic complications occur in #20% of patients, causing increased morbidity
Funding: This project was co-funded by the European Social Fund and National Resources – (EPEAEK II) PYTHAGORAS. There is no conflict of interest from any of the authors.
Kapodistrian University of Athens, Aretaieion University Hospital, 76 Vassilisis Sofia’s Avenue, Athens 11528, Greece. E-mail: [email protected].
Accepted for publication December 9, 2015.
0039-6060/$ - see front matter
Reprint requests: Constantinos Nastos, MD, PhD Second Department of Surgery, School of Medicine, National and
Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2015.12.018
SURGERY 1501
1502 Nastos et al
and mortality.1-3 Microorganisms derived from the gut lumen possibly owing to the disruption of the gut barrier seem to be involved.4 Although techniques of vascular control are invaluable in preventing excessive blood loss, many of these techniques induce a temporary ischemia and result in the production of reactive oxygen and nitrogen species, which are responsible for induction of oxidative and nitrosative stress.5,6 Release of cytokines and inflammatory mediators can promote remote injury.7 Moreover, in the setting of liver failure, several factors that help to maintain the integrity of the gut mucosal barrier have been shown to be compromised.4,8-11 Although many studies have examined the effect of bioartificial liver devices on liver function, liver regeneration, and hepatic encephalopathy,12 there is no study on the effect of any bioartificial liver support on gut barrier function and gut-derived sepsis. The aim of the present study was to evaluate the effect of extracorporeal bioartificial liver support on gut barrier function in an experimental model of major hepatectomy associated with ischemia/reperfusion injury of the liver remnant. METHODS This protocol was approved by the Animal Research Committee of the University of Athens and the Committee of Bioethics of Aretaieion Hospital. Care, and handling of the animals was in accordance with European guidelines for ethical animal research. Twelve female Landrace pigs weighing 30–35 kg were used. Anesthetic protocol. After a 24-hour deprivation of food with free access to water, animals were premedicated with intramuscular administration of ketamine 10 mg/kg and midazolam 0.2 mg/kg. General anesthesia was induced with thiopental sodium (5 mg/kg intravenously). The trachea of the animals was intubated and positive pressure ventilation initiated (Model Sulla 808 V, Drager, Lubeck, Germany). Ventilatory settings were as follows: inspired oxygen fraction (in air), 0.4–0.6; tidal volume 8–10 mL/kg; and respiratory rate adjusted to maintain arterial pH within 7.35–7.45. End-tidal CO2 was measured with a sidestream infrared CO2 monitor (CD-102 Normocap, Datex Inc., Helsinki, Finland) to document adequate ventilatory patterns. Operative procedure. All operative procedures were performed under sterile conditions. Immediately after endotracheal intubation, a gastric tonometer catheter was inserted orogastrically into the stomach. Partial pressure of CO2 in the gastric mucosa was measured using the continuous
Surgery June 2016
intragastric capnometry with the Tonocap device (Datex Ohmeda). The right femoral artery was catheterized percutaneously using a 20-G catheter for sampling of blood and measurement of arterial pressure. The right internal jugular vein was also catheterized after surgical dissection using a polyurethane, 1-lumen sheath 6.5-Fr (Arrow International, Reading, PA). After laparotomy, a urinary catheter was inserted in the bladder through a cystotomy. Afterwards, a side-to-side portacaval anastomosis was performed using continuous 5-0 Prolene suture to prevent splachnic congestion during the subsequent Pringle maneuver. During the creation of the anastomosis, care was taken not to interrupt blood flow through the vessels, and the portal vein was side clamped using a Satinsky clamp. After the creation of the anastomosis, the left hepatic artery was ligated and the hepatoduodenal ligament was clamped (Pringle maneuver). This maneuver initiated the ischemic phase, during which a 70% hepatectomy was performed by resection of the median and left liver lobes, as described previously.13,14 Hepatectomy was performed within 30 minutes, and blood loss was <100 mL in all animals. The liver was kept ischemic during the completion of the hepatectomy for a total of 150 minutes, after which the portacaval anastomosis was occluded using a vascular clamp, and portal blood flow was redirected back to the liver remnant by unclamping the hepatoduodenal ligament as described previously12; this step initiated the reperfusion phase. A 20-G catheter was then inserted in the portal vein through a side branch for monitoring of portal pressure and sampling of portal blood and was exteriorized through the abdominal incision before closing the abdomen. The abdomen was closed, and the liver was reperfused for a total of 24 hours during which the animals were kept under mechanical ventilation and monitored. No volatile anesthetics or vasoactive drugs were given. Anesthesia was maintained with continuous intravenous infusion of fentanyl 10–20 mg/kg/h, ketamine 5–8 mg/kg/h, and vecuronium bromide 0.5 mg/kg/h. Animals were also given a continuous intravenous infusion of 5 mL/kg/h of a standard solution consisting of a 6:3:1 mixture of lactated Ringer’s, hydroxyethylstarch (Hesteril 6%, Fresenius, Louviers, France), and 5% dextrose in water solution to maintain normoglycemia, normal electrolytic values, and central venous pressure between 4 and 9 mm Hg. Blood temperature was maintained at 388C throughout the procedure by means of a heating pad and warming of the infused solutions.
Surgery Volume 159, Number 6
Mean arterial pressure and portal pressure were recorded, and blood samples from the systemic and portal circulation were taken at the beginning of the reperfusion period and at 6, 12, and 24 hours of reperfusion. At the end of the experiment, all animals were killed with intravenous infusion of thiopental 5 mg/kg and 2 g KCl, and the last 10 cm of the ileum was sampled for histologic studies and measurement of malondialdehyde (MDA) and protein carbonyls content. Distal ileum has been shown to be more susceptible to oxidative injury compared with the large bowel in studies evaluating bacterial translocation in the setting of acute liver injury.15 In addition, other studies have also shown that the ileum is affected more than the proximal small intestine in experimental protocols of gut barrier dysfunction in hemorrhagic shock.16 Study design and hepatocyte treatment. After 6 hours of reperfusion, the 12 animals were allocated randomly to a control group, which did not receive any treatment (n = 6) or to the Hepat group, which underwent a 6-hour session of bioartificial liver support (n = 6). In all animals, a double-lumen catheter was inserted into the right femoral vein after surgical dissection. The animals of the Hepat group were connected to a plasma separation system (COBE Spectra, Gambro BCT, Lakewood, CO) where blood was withdrawn continuously at 50 mL/min and separated into plasma and cellular elements. Plasma was then diverted at 30 mL/min to a reservoir, from where it was recirculated with a pump at 800 mL/min through an oxygenator (Capiox 308, Teruma Inc, Tokyo, Japan); the intrafiber space of the bioreactors which was loaded with 10 billion hepatocytes, which were isolated from 6 donor pigs weighing 20–25 kg by twin stage, in situ collagenase digestion as described previously.17 Hepatocyte viability was >90%. After recirculating through the bioreactors, plasma was collected in the reservoir and returned to the plasma separation device where it was reconstituted with red cells and reinfused to the animal at 50 mL/min by way of the double-lumen venous catheter. Temperature in the extracorporeal circuit was maintained at 388C using a water heater connected to the membrane oxygenator. The extracorporeal circuit was a modification of the system designed by Demetriou et al,18,19 with the exception of the increased number of hepatocytes and the increased recirculation speed (800 mL/min) through the hollow fiber columns. Duration of the extracorporeal circulation was 6 hours. All animals were then disconnected from the plasma separation system and monitored for 12 more hours
Nastos et al 1503
(until postoperative hour 24). During the same period, control animals received supportive care. We chose to start the liver support session 6 hours after reperfusion, because this period is when the consequences of acute liver injury begin to manifest in the model that we employed.14 The duration of the session was set to 6 hours because it has been shown that the viability of the hepatocytes that are incorporated into the device decreases dramatically after this period of time.20 Bacteriologic analysis. Quantitative bacteriologic analysis of portal and systemic blood was performed using the Isolator pediatric tube (Oxoid Limited, Hants, England) culture system. We used this system because it yields an improved rate of isolation and culture of bacteria as well as faster identification of sepsis with positive blood cultures in comparison with other systems of bacterial culture analysis.21,22 In detail, after blood sampling and under sterile conditions, 1 mL of blood was transferred to the isolator tubes containing polypropylene glycol 8 mg/L, sodium polyanetholsulphonate 9.6 g/L, and purified saponin 40 g/L kept in room temperature for 2 hours. The blood samples were inoculated in blood agar plates (Bioprerare, Athens, Greece) and were incubated aerobically at 378C for 24 hours. Colonies were identified with conventional bacteriologic methods. Enteric Gram-negative bacteria were identified by the API 20 System and Gram positive bacteria by the API STAPH and API 20 Strep System (BioMerieux, Marcy-l’Etoile, France). The results were expressed as base-10 logarithm of colony-forming units (CFU) cultured per milliliter of blood sampled. Endotoxin measurement. Endotoxin levels were determined in systemic and portal blood with the lumilus amebocyte lysate test using the kinetic turbidimetric method (Pyrogent-5000 bulk kit, Lonza Walkerville Inc, Walkersville, MD) as described previously.23,24 The bacterial strain of standard endotoxin was Escherichia coli O55:B5. The range of the assay was 0.01–1 EU/mL. All samples were measured in duplicate and were subjected to spiked concentration measurements. Sample treatment for inhibition consisted of 1/ 100 dilution and heating at 758C for 15 minutes. MDA and protein carbonyl content determination. Tissue MDA and protein carbonyls are sensitive markers of oxidative injury.25 Tissue was sampled at the end of the experiment and stored at 808C until analysis. For the determination of tissue MDA, ileal mucosa was suspended in an ice-cold buffer containing 100 mmol/L NaCl, 0.5 mmol/L KCl, 3.1 mmol/L CaCl2, 1 mmol/L
1504 Nastos et al
MgSO4, 0.55 mmol/L KH2PO4, and 50 mmol/L Tris-HCl pH 7.4. The final concentration of the tissue in the homogenization buffer was 10% w/v. The tissues were homogenized by sonication, and the resulting suspension was centrifuged at 5003g for 10 minutes. The pellets were discarded, and the supernatants were centrifuged at 20,0003g for 20 minutes. The supernatant was discarded, and the pellet (membrane fraction) was suspended in the aforementioned buffer to a final concentration of 10% w/v. The total protein of the membrane fraction was determined by the Bradford method,26 and MDA content was determined according to the method of Jentzsch et al27 using 100 mg of membrane protein. The levels of protein carbonyls were measured using the colorimetric assay kit from Cayman Chemical (Ann Arbor, MI). Results are expressed as nanomoles per milligram of tissue homogenate protein. Intestinal pathology. Ileal biopsy specimens were stored immediately in 4% formaldehyde, embedded in paraffin, cut in 3–5 mm sections sagitally to the serosa, stained with hematoxylin and eosin, and processed for microscopy analysis. Microscopic evaluation was performed by 2 independent expert pathologists who were blinded to the treatment groups. The intestinal mucosal injury (IMI) score was based on the pathology scoring of Chiu et al.28 This scoring system has been shown to be effective in evaluating injury to the gut mucosa.28 The recorded score was the mean score of the 2 pathologists. Statistical analysis. Mann–Whitney tests were used to compare median values and determine statistical significance. All calculations were carried out using SPSS 17.0 for Windows (SPSS, Inc, Chicago, IL). CFU were tested after logarithmic transformation and are expressed in logarithmic form. Other continuous as well as the IMI score values are expressed as median (range). RESULTS All of the animals survived the 24-hour reperfusion period. There were no major hemorrhagic events in any of the experiments. Hemodynamic parameters. Both groups remained hemodynamically stable throughout the extracorporeal circulation period (Fig 1). There were no differences in mean arterial pressure between the 2 groups at all time points studied. Mean arterial pressure decreased gradually during the 24-hour reperfusion period. Portal pressure did not have differences between groups throughout the experiment. Partial pressure of CO2 in the gastric mucosa was also not different between the 2
Surgery June 2016
groups in any timepoint, suggesting no differences in splachnic perfusion. Data are depicted in Fig 1. Intestinal mucosa injury. In animals of the control group, there was massive detachment of the epithelial layer from the basal membrane until the lateral part of the villi and even complete destruction and digestion of the villous architecture (Fig 1, D). Injury was attenuated in animals of the Hepat group (Fig 2, A and B). When compared with the control group, the Hepat group had a lesser IMI score (4.0 [2.0–5.0] vs 1.0 [1.0–2.0]; P = .007; Fig 1, D). Permeability to bacteria and endotoxin. Quantitative blood cultures from the portal circulation expressed as median values of log10 CFU/mL revealed positive cultures (Fig 3), suggesting decreased bacterial translocation in animals in the Hepat group that was statistically significant at 24 hours after reperfusion (3.32 log10 CFU/mL [0.00–4.36] vs 0.60 log10 CFU/mL [0.00–2.51]; P = .05). The same effect was noted in the systemic circulation as well at the end of the 24-hour reperfusion period (3.48 log10 CFU/mL [0.78–4.12] vs 0.00 log10 CFU/mL [0.00–1.48]; P = .008; Fig 3, A). Endotoxin levels in the portal circulation were decreased in animals of the Hepat group compared with controls at 24 hours after reperfusion (3.04 EU/mL [0.98–6.95] vs 0.14 EU/mL [0.05–0.83]; P = .004). Compared with control animals, endotoxin levels in the systemic circulation were also less in animals of the Hepat group at 24 hours after reperfusion (3.08 EU/mL [0.83–5.00] vs 0.74 EU/mL [0.45–2.60]; P = .025; Fig 3, B). Tissue markers of oxidative injury. The levels of MDA in the intestinal mucosa were also decreased in the Hepat group (3.85 nmol/mg [3.01–6.43] vs 3.27 nmol/mg [1.46–3.55]; P = .04), as were the levels of protein carbonyl in the Hepat group (0.57 nmol/mg [0.32–0.70] vs 0.33 nmol/mg [0.03–0.53]; P = .01; Fig 4, A). Intestinal mucosa apoptosis. Intestinal mucosa cells were positive for caspase-3 staining in the tips of the enteric villi 24 hours after liver reperfusion. There was a trend toward decreased staining positivity in the Hepat group (3.5 % stained cells [1.0–8.0] vs 1.25 % stained cells [0.5–2.0]; P = .10; Fig 4, B). DISCUSSION Bioartificial liver support systems improve postoperative liver dysfunction after major hepatectomy.12,29 Nevertheless, there has been no study evaluating the role of these systems in gut barrier dysfunction and gut-derived sepsis, which leads frequently to septic complications, multiorgan failure, and death after major liver resections. Hybrid
Surgery Volume 159, Number 6
Nastos et al 1505
Fig 1. Hemodynamic and histologic data. A, Mean arterial pressure (mm Hg). B, Portal circulation pressure (mm Hg). C, Partial pressure of CO2 in the gastric mucosa (mm Hg). D, Mucosal injury score. Data are expressed as median values ± 95% CI. Hepat, Bioartificial liver support group.
Fig 2. Intestinal mucosa from the terminal ileum 24 hours after liver reperfusion from animal of the (A) control group (injury score 3; arrow) and (B) bioartificial liver support group (injury score 0). Stain: hematoxylin and eosin; original magnification, 3100.
liver support devices could be used potentially in patients after major hepatectomy with small liver remnants in an attempt to prevent septic
complications thought to derive from translocation of gut flora. The use of these devices in such clinical scenarios might allow more aggressive liver
1506 Nastos et al
Surgery June 2016
Fig 3. Bacterial translocation and endotoxin levels. (A) Bacterial translocation in the portal and systemic circulation. Data are expressed as base-10 logarithm of colony-forming units (CFU) cultured per milliliter of blood sampled. *P < .05; **P < .01. (B) Endotoxin concentration in portal and systemic circulation expressed as endotoxin units per milliliter of blood. Data are expressed as median values ± 95% CI. *P < .05. Hepat, Bioartificial liver support group.
resections and potentially assist in the management of patients who are now considered inoperable. The aim of this study was to evaluate the role of extracorporeal bioartificial liver in the attenuation of bacterial translocation and gutderived sepsis after major hepatectomy in a large animal model. After major hepatectomy, septic complication can occur in #20% of patients.1-3 The microorganisms involved are thought to derive from the gut microflora.30,31 Normal mucosal barrier function prevents bacteria and their toxins from translocating into the circulation. A number of normal defensive mechanisms that contribute to the intestinal mucosal integrity and gut barrier are compromised post hepatectomy, as described previously.32,33,34
In addition, portal hypertension has been documented after major hepatectomy and can lead to congestion of the intestine with failure of the gut mucosal barrier.35 Also, when liver ischemia/reperfusion takes place, reactive oxygen species production may cause remote injury to the intestinal mucosa via the systemic circulation.32 Abnormal hemodynamic parameters of the systemic circulation can also compromise gut barrier function. As we have shown, mean arterial pressure can decrease after major hepatectomy and with postoperative liver dysfunction, thus compromising intestinal mucosal integrity. Extracorporeal bioartificial liver support seems to attenuate this decrease in arterial pressure, by helping to maintain adequate intestinal perfusion.12 In our study, gastric tonometry did not suggest any significant
Surgery Volume 159, Number 6
Nastos et al 1507
Fig 4. Markers of intestinal oxidative injury. (A) Intestinal mucosa tissue protein carbonyl and malondialdehyde concentration expressed in nanomoles per milligram 24 hours after reperfusion. Data are expressed as median values ± 95% CI (*P < .05). (B) Percentage of positive caspase-3 immunohistochemical staining of enterocytes 24 hours after liver reperfusion. Data are expressed as median percentage of cells ± 95% CI. Hepat, Bioartificial liver support group.
differences in splachnic perfusion (at least to the stomach) between groups, suggesting that bioartificial liver improved gut barrier function through other pathways. Acute portal hypertension causes splachnic venous congestion and injury to the pancreas and intestine.36,37 In our study, portal pressure was monitored during ischemia and reperfusion and did not show differences between groups, thus excluding portal hypertension as a confounding factor for the differences in gut mucosa injury and bacterial and endotoxin translocation. During reperfusion, an increasing trend in portal pressure was noted within each group. This early trend, which began after reperfusion, was probably owing to the reperfusion of the small liver remnant. After major liver resection, there is increased vascular resistance in the portal venous system owing to the decreased total intrahepatic vasculature.38 In addition, microcirculatory failure owing to accumulation of platelets and macrophages can contribute to this phenomenon.39 In the present study, we demonstrated attenuation of bacterial translocation in both the portal and systemic circulations in animals treated with bioartificial liver support after the treatment session. The bioreactor filter has a small pore size and by design is not permeable to bacteria. As a result, the decrease in bacteria cultures in the portal ad systemic circulation in the Hepat group cannot be attributed to clearance of bacteria from the system, but rather in the protection of the integrity of the mucosal barrier by the system. In contrast, the decrease in endotoxin translocation in animals subjected to extracorporeal bioartificial liver
support shown in our study could be explained by endotoxin clearance from cells within the bioreactor. The size of the pores of the bioreactor was 0.2 mm, allowing passage of molecules of #3,000 kDa. Endotoxins have a size that ranges from 10 to 1,000 kDa.40 Endotoxin clearance from the liver has been shown to correlate with functional hepatic mass,41 and as a result could be increased during bioartificial liver support. These findings also correlate with our findings for IMI score. Control animals developed moderate to severe mucosal damage, whereas animals in the Hepat group only developed mild mucosal injury, and in 1 animal only rare foci of grade 1 injury were noted. IMI score after major hepatectomy has been reported by Wang et al and Xu et al in rats as well,33,42 correlating with endotoxemia and the degree of liver resection. Oxidative injury results from liver ischemia and reperfusion. Oxidative and inflammatory markers “spilled” in the systemic circulation can lead to injury of other distant organs, including the intestinal mucosa.43-49 In our study, the concentration of MDA and protein carbonyls where measured to associate tissue injury with oxidative stress produced by the 150 minutes of vascular occlusion of the liver. Tissue oxidative markers were decreased in the Hepat group. A possible explanation for this observation is that the hepatocytes in the bioreactor contributed to the intrinsic liver antioxidative mechanisms which were compromised in the animals of the control group. This effect has been studied in the past as well in bioartificial liver devices, leading to amelioration
1508 Nastos et al
of the oxidative stress induced by liver failure, including hepatic encephalopathy and pulmonary injury.29,50 Apoptosis plays a major role in mucosal injury. Oxidative stress can result in injury of cellular membranes and DNA in epithelial cells of intestinal mucosa, activating the apoptotic path and disrupting the integrity of the gut mucosa barrier.51,52 Ikeda et al53 have shown that the typical manifestation of ischemia/reperfusion injury in the intestinal mucosa is the detachment of enterocytes from the basal membrane and that apoptosis is the major mechanism involved in this process concurrently with necrosis. Morphologically, the detached enterocytes create a subepithelial space and intracellular “bubbles,” which are responsible for the creation of the space of Gruenhagen, which results in the final detachment of the epithelial layer form the basal membrane.54 Recent studies have associated apoptosis with the interaction of the cellular molecules and the extracellular matrix with cellular death being noted when these interactions are lost.55 In the present study, the aforementioned pathologic space was observed in the control group but was absent in the Hepat group. Despite these differences, however, there was no difference in the immunohistochemical expression of caspase-3. This ostensible discrepancy may be explained by the fact that ischemia/reperfusion and oxidative stress induces a type II route of apoptosis in which caspase-3 is only partially involved.56 Another possible mechanism that may explain the findings of our study is the possible clearance of cytokines and other inflammatory mediators by the hepatocyte bioreactor. Because molecular weights of these substances are relatively small, these mediators could easily pass through the pores of the bioreactor. Although clearance of circulating cytokines is not proven by our study, such cytocine clearance has been shown in studies with bioartificial and nonbioartificial extracorporeal liver support systems in patients with acute liver failure.57,58 In the study using a molecular adsorbent recycling system, Guo et al58 showed a significant decrease in tumor necrosis factor-a, interleukins 6 and 8, and interferon-g, together with marked decreases in other, non–water-soluble albumin-bound toxins and water-soluble toxins that were associated with an improvement in hepatic encephalopathy. These systems, however, have failed to show any substantial benefit in septic parameters and complications of acute liver failure.59 In contrast, Liu et al studied a bioartificial liver support system with porcine hepatocytes and
Surgery June 2016
showed that proinflammatory cytokines were removed by the bioreactor, which seemed to be effective in blocking the transition of the systemic inflammatory response syndrome to multiorgan dysfunction syndrome in patients with acute and subacute liver failure.57 A limitation of our study is that the monitoring period after the induction of reperfusion was restricted to 24 hours; this is a relatively short period, and no conclusions can be made for the survival or the late evolution of bacterial translocation and gut barrier function. In addition, we only evaluated the hematogenous route of translocation and not the lymphatic route which has been described to contribute to the development of gut-derived sepsis60; we were unable to assess the mesenteric lymph nodes which would compromise the integrity of the gut barrier by itself. Finally, anaerobic bacteria were not cultured in our samples, but their role in bacterial translocation remains controversial.61 In conclusion, our study shows that dysfunction of the gut mucosal barrier is ameliorated by the use of bioartificial liver in the setting of acute liver injury after major hepatectomy. This finding could imply that the use of such devices postoperatively could permit even more aggressive liver surgery with more extensive liver resections which could be tested in future clinical trials. REFERENCES 1. Shigeta H, Nagino M, Kamiya J, Uesaka K, Sano T, Yamamoto H, et al. Bacteremia after hepatectomy: an analysis of a single-center, 10-year experience with 407 patients. Langenbecks Arch Surg 2002;387:117-24. 2. Rolando N, Harvey F, Brahm J, Philpott-Howard J, Alexander G, Gimson A, et al. Prospective study of bacterial infection in acute liver failure: an analysis of fifty patients. Hepatology 1990;11:49-53. 3. Wyke RJ, Canalese JC, Gimson AE, Williams R. Bacteraemia in patients with fulminant hepatic failure. Liver 1982;2:45-52. 4. Wang XD, Soltesz V, Andersson R, Bengmark S. Bacterial translocation in acute liver failure induced by 90 per cent hepatectomy in the rat. Br J Surg 1993;80:66-71. 5. Smyrniotis V, Farantos C, Kostopanagiotou G, Arkadopoulos N. Vascular control during hepatectomy: review of methods and results. World J Surg 2005;29:1384-96. 6. Garcea G, Gescher A, Steward W, Dennison A, Berry D. Oxidative stress in humans following the Pringle manoeuvre. Hepatobiliary Pancreat Dis Int 2006;5:210-4. 7. Yang JC, Ji XQ, Li CL, Lin JH, Liu XG. Impact of earlystage hepatic ischemia-reperfusion injury on other organs of rats. Di Yi Jun Yi Da Xue Xue Bao 2004;24: 1019-22. 8. Li LJ, Wu ZW, Xiao DS, Sheng JF. Changes of gut flora and endotoxin in rats with D-galactosamine-induced acute liver failure. World J Gastroenterol 2004;10:2087-90. 9. Wang X, Guo W, Wang Q, Soltesz V, Andersson R. Effects of a water-soluble ethylhydroxyethyl cellulose on
Surgery Volume 159, Number 6
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21. 22.
23.
24.
25.
gut physiology, bacteriology, and bacterial translocation in acute liver failure. J Invest Surg 1995;8:65-84. Wang X, Soltesz V, Axelson J, Andersson R. Cholecystokinin increases small intestinal motility and reduces enteric bacterial overgrowth and translocation in rats with surgically induced acute liver failure. Digestion 1996;57:67-72. Wang XD, Soltesz V, Andersson R. Cisapride prevents enteric bacterial overgrowth and translocation by improvement of intestinal motility in rats with acute liver failure. Eur Surg Res 1996;28:402-12. Arkadopoulos N, Kostopanagiotou G, Nastos C, Papalois A, Papoutsidakis N, Kalimeris K, et al. Reversal of experimental posthepatectomy liver failure in pigs: a new application of hepatocyte bioreactors. Artif Organs 2011;35:29-36. Ladurner R, Hochleitner B, Schneeberger S, Barnas U, Krismer A, Kleinsasser A, et al. Extended liver resection and hepatic ischemia in pigs: a new, potentially reversible model to induce acute liver failure and study artificial liver support systems. Eur Surg Res 2005;37:365-9. Arkadopoulos N, Defterevos G, Nastos C, Papalois A, Kalimeris K, Papoutsidakis N, et al. Development of a porcine model of post-hepatectomy liver failure. J Surg Res 2011;170:e233-42. Kasravi FB, Wang L, Wang XD, Molin G, Bengmark S, Jeppsson B. Bacterial translocation in acute liver injury induced by D-galactosamine. Hepatology 1996;23:97-103. Chang JX, Chen S, Ma LP, Jiang LY, Chen JW, Chang RM, et al. Functional and morphological changes of the gut barrier during the restitution process after hemorrhagic shock. World J Gastroenterol 2005;11:5485-91. Demetriou AA, Brown RS Jr, Busuttil RW, Fair J, McGuire BM, Rosenthal P, et al. Prospective, randomized, multicenter, controlled trial of a bioartificial liver in treating acute liver failure. Ann Surg 2004;239:660. Rozga J, Holzman MD, Ro MS, Griffin DW, Neuzil DF, Giorgio T, et al. Development of a hybrid bioartificial liver. Ann Surg 1993;217:502-9; discussion 9–11. Khalili TM, Navarro A, Ting P, Kamohara Y, Arkadopoulos N, Solomon BA, et al. Bioartificial liver treatment prolongs survival and lowers intracranial pressure in pigs with fulminant hepatic failure. Artif Organs 2001;25: 566-70. Demetriou AA, Brown RS Jr, Busuttil RW, Fair J, McGuire BM, Rosenthal P, et al. Prospective, randomized, multicenter, controlled trial of a bioartificial liver in treating acute liver failure. Ann Surg 2004;239:660-7; discussion 7–70. Washington J. The use of lysis-based procedures for blood cultures. Eur J Clin Microbiol Infect Dis 1991;10:618-9. Welch D, Scribner R, Hensel D. Evaluation of a lysis direct plating method for pediatric blood cultures. J Clin Microbiol 1985;21:955. Oishi H, Fusamoto M, Hatayama Y, Tsuchiya M, Takaoka A, Sakata Y. An automated analysis system of Limulus amebocyte lysate (LAL)-endotoxin reaction kinetics using turbidimetric kinetic assay. Chem Pharm Bull (Tokyo) 1988;36: 3012-9. Yokota M, Kambayashi J, Tanaka T, Tsujinaka T, Sakon M, Mori T. A simple turbidimetric time assay of the endotoxin in plasma. J Biochem Biophys Methods 1989;18:97-104. Alexandris I, Assimakopoulos S, Vagianos C, Patsoukis N, Georgiou C, Nikolopoulou V, et al. Oxidative state in intestine and liver after partial hepatectomy in rats. Effect of bombesin and neurotensin. Clin Biochem 2004;37:350-6.
Nastos et al 1509
26. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-54. 27. Jentzsch AM, Bachmann H, Furst P, Biesalski HK. Improved analysis of malondialdehyde in human body fluids. Free Radic Biol Med 1996;20:251-6. 28. Chiu CJ, McArdle AH, Brown R, Scott HJ, Gurd FN. Intestinal mucosal lesion in low-flow states. I. A morphological, hemodynamic, and metabolic reappraisal. Arch Surg 1970; 101:478-83. 29. Arkadopoulos N, Kalimeris K, Papalois A, Nastos C, Papoutsidakis N, Defterevos G, et al. Treatment with bioartificial liver improves lung injury in a swine model of partial hepatectomy and ischemia/reperfusion. Int J Artif Organs 2010;33:105-13. 30. Wang X, Andersson R, Soltesz V, Bengmark S. Bacterial translocation after major hepatectomy in patients and rats. Arch Surg 1992;127:1101-6. 31. Ren W, Wang X, Zhang A, Li C, Chen G, Ge X, et al. Selective bowel decontamination improves the survival of 90% hepatectomy in rats. J Surg Res 2015;195:454-64. 32. Nastos C, Kalimeris K, Papoutsidakis N, Defterevos G, Pafiti A, Kalogeropoulou H, et al. Antioxidant treatment attenuates intestinal mucosal damage and gut barrier dysfunction after major hepatectomy. Study in a porcine model. J Gastrointest Surg 2011;15:809-17. 33. Wang XD, Parsson H, Andersson R, Soltesz V, Johansson K, Bengmark S. Bacterial translocation, intestinal ultrastructure and cell membrane permeability early after major liver resection in the rat. Br J Surg 1994;81:579-84. 34. Hamazaki K, Sato S, Yunoki M, Noda T, Moreira LF, Mimura H, et al. Kupffer cell function in chronic liver injury and after partial hepatectomy. Res Experiment Med 1994;194:237-46. 35. Chen WZ, Hu KP, Xu RY, Pan WD. Beneficial effect of splenic artery ligation on bacterial translocation after major liver resection in rats. Dig Liver Dis 2013;45: 233-7. 36. Garcia-Tsao G, Albillos A, Barden GE, West AB. Bacterial translocation in acute and chronic portal hypertension. Hepatology 1993;17:1081-5. 37. Miyagawa S, Makuuchi M, Kawasaki S, Kakazu T. Changes in serum amylase level following hepatic resection in chronic liver disease. Arch Surg 1994;129:634-8. 38. Wang H, Ohkohchi N, Enomoto Y, Usuda M, Miyagi S, Masuoka H, et al. Effect of portocaval shunt on residual extreme small liver after extended hepatectomy in porcine. World J Surg 2006;30:2014-22; discussion 23–4. 39. Farmer DG, Amersi F, Kupiec-Weglinski J, Busuttil RW. Current status of ischemia and reperfusion injury in the liver. Transplant Rev 2000;14:106-26. 40. Carpentier B, Gautier A, Legallais C. Artificial and bioartificial liver devices: present and future. Gut 2009;58: 1690-702. 41. Caruana JA, Camara DS, Schneeberger GJ, Nolan JP. The clearance capacity of the canine liver for a portal vein endotoxin infusion. J Surg Res 1984;37:197-201. 42. Xu CP, Liu J, Liu JC, Han DW, Zhang Y, Zhao YC. Dynamic changes and mechanism of intestinal endotoxemia in partially hepatectomized rats. World J Gastroenterol 2007; 13:3592-7. 43. Filos KS, Kirkilesis I, Spiliopoulou I, Scopa CD, Nikolopoulou V, Kouraklis G, et al. Bacterial translocation, endotoxaemia and apoptosis following Pringle manoeuvre in rats. Injury 2004;35:35-43.
1510 Nastos et al
44. Bedirli A, Gokahmetoglu S, Kucuk K, Soyuer I, Guler I, Sakrak O. Bacterial translocation after partial hepatic resection under ischemia and reperfusion in rats: incidence and time course. Turk J Med Sci 2003;33:135-40. 45. Nastos C, Kalimeris K, Papoutsidakis N, Tasoulis MK, Lykoudis PM, Theodoraki K, et al. Global consequences of liver ischemia/reperfusion injury. Oxid Med Cell Longev 2014;2014:906965. 46. Varsos P, Nastos C, Papoutsidakis N, Kalimeris K, Defterevos G, Nomikos T, et al. Desferrioxamine attenuates pancreatic injury after major hepatectomy under vascular control of the liver: experimental study in pigs. HPB Surg 2012;2012:714672. 47. Papoutsidakis N, Arkadopoulos N, Smyrniotis V, Tzanatos H, Kalimeris K, Nastos K, et al. Basic research Early myocardial injury is an integral component of experimental acute liver failure–a study in two porcine models. Arch Med Sci 2011;7:217-23. 48. Arkadopoulos N, Nastos C, Defterevos G, Kalimeris K, Papoutsidakis N, Andreadou I, et al. Pancreatic injury after major hepatectomy: a study in a porcine model. Surg Today 2012;42:368-75. 49. Kalimeris K, Nastos C, Papoutsidakis N, Xanthopoulou MN, Defterevos G, Tympa A, et al. Iron chelation prevents lung injury after major hepatectomy. Hepatol Res 2010;40:841-50. 50. Shi Q. On bioartificial liver assist system: theoretical exploration and strategies for further development. Artif Cell Blood Subst Biotechnol 2000;28:535-46. 51. Zhang C, Sheng ZY, Hu S, Gao JC, Yu S, Liu Y. The influence of apoptosis of mucosal epithelial cells on intestinal barrier integrity after scald in rats. Burns 2002;28:731-7. 52. Noda T, Iwakiri R, Fujimoto K, Matsuo S, Aw TY. Programmed cell death induced by ischemia-reperfusion in rat intestinal mucosa. Am J Physiol 1998;274:G270-6.
Surgery June 2016
53. Ikeda H, Suzuki Y, Suzuki M, Koike M, Tamura J, Tong J, et al. Apoptosis is a major mode of cell death caused by ischaemia and ischaemia/reperfusion injury to the rat intestinal epithelium. Gut 1998;42:530-7. 54. Shah KA, Shurey S, Green CJ. Apoptosis after intestinal ischemia-reperfusion injury: a morphological study. Transplantation 1997;64:1393-7. 55. Frisch SM, Francis H. Disruption of epithelial cell-matrix interactions induces apoptosis. J Cell Biol 1994;124:619-26. 56. Fujise T, Iwakiri R, Wu B, Amemori S, Kakimoto T, Yokoyama F, et al. Apoptotic pathway in the rat small intestinal mucosa is different between fasting and ischemia-reperfusion. Am J Physiol Gastrointest Liver Physiol 2006; 291:G110-6. 57. Liu Q, Duan Z, Huang C. The clinical research of hybrid bioartificial liver for blocking the SIRS to MODS stage in patients with fulminant liver failure. J Emerg Med 2002;1:003. 58. Guo LM, Liu JY, Xu DZ, Li BS, Han H, Wang LH, et al. Application of molecular adsorbents recirculating system to remove NO and cytokines in severe liver failure patients with multiple organ dysfunction syndrome. Liver Int 2003; 23:16-20. 59. Rittler P, Ketscher C, Inthorn D, Jauch KW, Hartl WH. Use of the molecular adsorbent recycling system in the treatment of postoperative hepatic failure and septic multiple organ dysfunction–preliminary results. Liver Int 2004;24: 136-41. 60. Lemaire LC, van Wagensveld BA, van Gulik TM, Dankert J, van Lanschot JJ, Gouma DJ. Bacterial translocation to the thoracic duct in a setting of ischemia, partial resection and reperfusion of the porcine liver. Dig Surg 1999;16: 222-8. 61. Berg RD. Bacterial translocation from the gastrointestinal tract. Trends Microbiol 1995;3:149-54.
Background. The aim of this study was to evaluate whether bioartificial liver support can attenuate gut mucosa injury in a porcine model of posthepatectomy liver dysfunction. Methods. Posthepatectomy liver failure was induced in pigs combining major (70%) liver resection and ischemia/reperfusion injury. An ischemic period of 150 minutes was followed by reperfusion for 24 hours. Animals were divided randomly into 2 groups: a control group (n = 6) that received standard critical care and a bioartificial liver support group (Hepat, n = 6) that were subjected to extracorporeal liver support for 6 hours during reperfusion. Intestinal mucosal injury, bacterial translocation, and endotoxin translocation were evaluated in all animals. Intestinal mucosa was also evaluated with markers of oxidative injury and immunohistochemical staining for caspase 3. Results. When compared with median values, the control group, animals in the Hepat group had a lesser intestinal mucosal injury score (4.0 [range:2.0–5.0] vs 1.0 [range:1.0–2.0]; P <.01), decreased bacterial translocation in the portal and the systemic circulation at 24 hours of reperfusion (P <.05), and decreased portal and systemic endotoxin levels at 24 hours (P < .05). At 24 hours after reperfusion, mucosal protein carbonyls and malondialdehyde levels were decreased in Hepat animals (0.57 nmol/mg [range:0.32– 0.70] vs 0.33 nmol/mg [range:0.03–0.53] and 3.85 nmol/mg [range:3.01–6.43] vs 3.27 nmol/mg [range:1.46–3.55], respectively; P < .05). There was no difference in tissue caspase staining. Conclusion. Bioartificial liver support seems to attenuate intestinal mucosal injury and gut barrier dysfunction after major hepatectomy. (Surgery 2016;159:1501-10.) From the Second Department of Surgery,a School of Medicine, National and Kapodistrian University of Athens, Aretaieion University Hospital; Second Department of Anesthesiology,b School of Medicine, National and Kapodistrian University of Athens, Attikon University Hospital; Department of Pathology,c School of Medicine, National and Kapodistrian University of Athens, Aretaieion University Hospital; Laboratory of Biopathology,d Attikon University Hospital; Department of the Science Nutrition – Dietetics,e Harokopio University; Experimental Research Department of ELPEN-Pharmaf; and Fourth Department of Surgery,g School of Medicine, National and Kapodistrian University of Athens, Attikon University Hospital, Athens, Greece
DURING THE LAST FEW DECADES, improvements in surgical techniques and intensive care support have facilitated more extensive liver resections as
a treatment of primary and metastatic tumors. Posthepatectomy septic complications occur in #20% of patients, causing increased morbidity
Funding: This project was co-funded by the European Social Fund and National Resources – (EPEAEK II) PYTHAGORAS. There is no conflict of interest from any of the authors.
Kapodistrian University of Athens, Aretaieion University Hospital, 76 Vassilisis Sofia’s Avenue, Athens 11528, Greece. E-mail: [email protected].
Accepted for publication December 9, 2015.
0039-6060/$ - see front matter
Reprint requests: Constantinos Nastos, MD, PhD Second Department of Surgery, School of Medicine, National and
Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2015.12.018
SURGERY 1501
1502 Nastos et al
and mortality.1-3 Microorganisms derived from the gut lumen possibly owing to the disruption of the gut barrier seem to be involved.4 Although techniques of vascular control are invaluable in preventing excessive blood loss, many of these techniques induce a temporary ischemia and result in the production of reactive oxygen and nitrogen species, which are responsible for induction of oxidative and nitrosative stress.5,6 Release of cytokines and inflammatory mediators can promote remote injury.7 Moreover, in the setting of liver failure, several factors that help to maintain the integrity of the gut mucosal barrier have been shown to be compromised.4,8-11 Although many studies have examined the effect of bioartificial liver devices on liver function, liver regeneration, and hepatic encephalopathy,12 there is no study on the effect of any bioartificial liver support on gut barrier function and gut-derived sepsis. The aim of the present study was to evaluate the effect of extracorporeal bioartificial liver support on gut barrier function in an experimental model of major hepatectomy associated with ischemia/reperfusion injury of the liver remnant. METHODS This protocol was approved by the Animal Research Committee of the University of Athens and the Committee of Bioethics of Aretaieion Hospital. Care, and handling of the animals was in accordance with European guidelines for ethical animal research. Twelve female Landrace pigs weighing 30–35 kg were used. Anesthetic protocol. After a 24-hour deprivation of food with free access to water, animals were premedicated with intramuscular administration of ketamine 10 mg/kg and midazolam 0.2 mg/kg. General anesthesia was induced with thiopental sodium (5 mg/kg intravenously). The trachea of the animals was intubated and positive pressure ventilation initiated (Model Sulla 808 V, Drager, Lubeck, Germany). Ventilatory settings were as follows: inspired oxygen fraction (in air), 0.4–0.6; tidal volume 8–10 mL/kg; and respiratory rate adjusted to maintain arterial pH within 7.35–7.45. End-tidal CO2 was measured with a sidestream infrared CO2 monitor (CD-102 Normocap, Datex Inc., Helsinki, Finland) to document adequate ventilatory patterns. Operative procedure. All operative procedures were performed under sterile conditions. Immediately after endotracheal intubation, a gastric tonometer catheter was inserted orogastrically into the stomach. Partial pressure of CO2 in the gastric mucosa was measured using the continuous
Surgery June 2016
intragastric capnometry with the Tonocap device (Datex Ohmeda). The right femoral artery was catheterized percutaneously using a 20-G catheter for sampling of blood and measurement of arterial pressure. The right internal jugular vein was also catheterized after surgical dissection using a polyurethane, 1-lumen sheath 6.5-Fr (Arrow International, Reading, PA). After laparotomy, a urinary catheter was inserted in the bladder through a cystotomy. Afterwards, a side-to-side portacaval anastomosis was performed using continuous 5-0 Prolene suture to prevent splachnic congestion during the subsequent Pringle maneuver. During the creation of the anastomosis, care was taken not to interrupt blood flow through the vessels, and the portal vein was side clamped using a Satinsky clamp. After the creation of the anastomosis, the left hepatic artery was ligated and the hepatoduodenal ligament was clamped (Pringle maneuver). This maneuver initiated the ischemic phase, during which a 70% hepatectomy was performed by resection of the median and left liver lobes, as described previously.13,14 Hepatectomy was performed within 30 minutes, and blood loss was <100 mL in all animals. The liver was kept ischemic during the completion of the hepatectomy for a total of 150 minutes, after which the portacaval anastomosis was occluded using a vascular clamp, and portal blood flow was redirected back to the liver remnant by unclamping the hepatoduodenal ligament as described previously12; this step initiated the reperfusion phase. A 20-G catheter was then inserted in the portal vein through a side branch for monitoring of portal pressure and sampling of portal blood and was exteriorized through the abdominal incision before closing the abdomen. The abdomen was closed, and the liver was reperfused for a total of 24 hours during which the animals were kept under mechanical ventilation and monitored. No volatile anesthetics or vasoactive drugs were given. Anesthesia was maintained with continuous intravenous infusion of fentanyl 10–20 mg/kg/h, ketamine 5–8 mg/kg/h, and vecuronium bromide 0.5 mg/kg/h. Animals were also given a continuous intravenous infusion of 5 mL/kg/h of a standard solution consisting of a 6:3:1 mixture of lactated Ringer’s, hydroxyethylstarch (Hesteril 6%, Fresenius, Louviers, France), and 5% dextrose in water solution to maintain normoglycemia, normal electrolytic values, and central venous pressure between 4 and 9 mm Hg. Blood temperature was maintained at 388C throughout the procedure by means of a heating pad and warming of the infused solutions.
Surgery Volume 159, Number 6
Mean arterial pressure and portal pressure were recorded, and blood samples from the systemic and portal circulation were taken at the beginning of the reperfusion period and at 6, 12, and 24 hours of reperfusion. At the end of the experiment, all animals were killed with intravenous infusion of thiopental 5 mg/kg and 2 g KCl, and the last 10 cm of the ileum was sampled for histologic studies and measurement of malondialdehyde (MDA) and protein carbonyls content. Distal ileum has been shown to be more susceptible to oxidative injury compared with the large bowel in studies evaluating bacterial translocation in the setting of acute liver injury.15 In addition, other studies have also shown that the ileum is affected more than the proximal small intestine in experimental protocols of gut barrier dysfunction in hemorrhagic shock.16 Study design and hepatocyte treatment. After 6 hours of reperfusion, the 12 animals were allocated randomly to a control group, which did not receive any treatment (n = 6) or to the Hepat group, which underwent a 6-hour session of bioartificial liver support (n = 6). In all animals, a double-lumen catheter was inserted into the right femoral vein after surgical dissection. The animals of the Hepat group were connected to a plasma separation system (COBE Spectra, Gambro BCT, Lakewood, CO) where blood was withdrawn continuously at 50 mL/min and separated into plasma and cellular elements. Plasma was then diverted at 30 mL/min to a reservoir, from where it was recirculated with a pump at 800 mL/min through an oxygenator (Capiox 308, Teruma Inc, Tokyo, Japan); the intrafiber space of the bioreactors which was loaded with 10 billion hepatocytes, which were isolated from 6 donor pigs weighing 20–25 kg by twin stage, in situ collagenase digestion as described previously.17 Hepatocyte viability was >90%. After recirculating through the bioreactors, plasma was collected in the reservoir and returned to the plasma separation device where it was reconstituted with red cells and reinfused to the animal at 50 mL/min by way of the double-lumen venous catheter. Temperature in the extracorporeal circuit was maintained at 388C using a water heater connected to the membrane oxygenator. The extracorporeal circuit was a modification of the system designed by Demetriou et al,18,19 with the exception of the increased number of hepatocytes and the increased recirculation speed (800 mL/min) through the hollow fiber columns. Duration of the extracorporeal circulation was 6 hours. All animals were then disconnected from the plasma separation system and monitored for 12 more hours
Nastos et al 1503
(until postoperative hour 24). During the same period, control animals received supportive care. We chose to start the liver support session 6 hours after reperfusion, because this period is when the consequences of acute liver injury begin to manifest in the model that we employed.14 The duration of the session was set to 6 hours because it has been shown that the viability of the hepatocytes that are incorporated into the device decreases dramatically after this period of time.20 Bacteriologic analysis. Quantitative bacteriologic analysis of portal and systemic blood was performed using the Isolator pediatric tube (Oxoid Limited, Hants, England) culture system. We used this system because it yields an improved rate of isolation and culture of bacteria as well as faster identification of sepsis with positive blood cultures in comparison with other systems of bacterial culture analysis.21,22 In detail, after blood sampling and under sterile conditions, 1 mL of blood was transferred to the isolator tubes containing polypropylene glycol 8 mg/L, sodium polyanetholsulphonate 9.6 g/L, and purified saponin 40 g/L kept in room temperature for 2 hours. The blood samples were inoculated in blood agar plates (Bioprerare, Athens, Greece) and were incubated aerobically at 378C for 24 hours. Colonies were identified with conventional bacteriologic methods. Enteric Gram-negative bacteria were identified by the API 20 System and Gram positive bacteria by the API STAPH and API 20 Strep System (BioMerieux, Marcy-l’Etoile, France). The results were expressed as base-10 logarithm of colony-forming units (CFU) cultured per milliliter of blood sampled. Endotoxin measurement. Endotoxin levels were determined in systemic and portal blood with the lumilus amebocyte lysate test using the kinetic turbidimetric method (Pyrogent-5000 bulk kit, Lonza Walkerville Inc, Walkersville, MD) as described previously.23,24 The bacterial strain of standard endotoxin was Escherichia coli O55:B5. The range of the assay was 0.01–1 EU/mL. All samples were measured in duplicate and were subjected to spiked concentration measurements. Sample treatment for inhibition consisted of 1/ 100 dilution and heating at 758C for 15 minutes. MDA and protein carbonyl content determination. Tissue MDA and protein carbonyls are sensitive markers of oxidative injury.25 Tissue was sampled at the end of the experiment and stored at 808C until analysis. For the determination of tissue MDA, ileal mucosa was suspended in an ice-cold buffer containing 100 mmol/L NaCl, 0.5 mmol/L KCl, 3.1 mmol/L CaCl2, 1 mmol/L
1504 Nastos et al
MgSO4, 0.55 mmol/L KH2PO4, and 50 mmol/L Tris-HCl pH 7.4. The final concentration of the tissue in the homogenization buffer was 10% w/v. The tissues were homogenized by sonication, and the resulting suspension was centrifuged at 5003g for 10 minutes. The pellets were discarded, and the supernatants were centrifuged at 20,0003g for 20 minutes. The supernatant was discarded, and the pellet (membrane fraction) was suspended in the aforementioned buffer to a final concentration of 10% w/v. The total protein of the membrane fraction was determined by the Bradford method,26 and MDA content was determined according to the method of Jentzsch et al27 using 100 mg of membrane protein. The levels of protein carbonyls were measured using the colorimetric assay kit from Cayman Chemical (Ann Arbor, MI). Results are expressed as nanomoles per milligram of tissue homogenate protein. Intestinal pathology. Ileal biopsy specimens were stored immediately in 4% formaldehyde, embedded in paraffin, cut in 3–5 mm sections sagitally to the serosa, stained with hematoxylin and eosin, and processed for microscopy analysis. Microscopic evaluation was performed by 2 independent expert pathologists who were blinded to the treatment groups. The intestinal mucosal injury (IMI) score was based on the pathology scoring of Chiu et al.28 This scoring system has been shown to be effective in evaluating injury to the gut mucosa.28 The recorded score was the mean score of the 2 pathologists. Statistical analysis. Mann–Whitney tests were used to compare median values and determine statistical significance. All calculations were carried out using SPSS 17.0 for Windows (SPSS, Inc, Chicago, IL). CFU were tested after logarithmic transformation and are expressed in logarithmic form. Other continuous as well as the IMI score values are expressed as median (range). RESULTS All of the animals survived the 24-hour reperfusion period. There were no major hemorrhagic events in any of the experiments. Hemodynamic parameters. Both groups remained hemodynamically stable throughout the extracorporeal circulation period (Fig 1). There were no differences in mean arterial pressure between the 2 groups at all time points studied. Mean arterial pressure decreased gradually during the 24-hour reperfusion period. Portal pressure did not have differences between groups throughout the experiment. Partial pressure of CO2 in the gastric mucosa was also not different between the 2
Surgery June 2016
groups in any timepoint, suggesting no differences in splachnic perfusion. Data are depicted in Fig 1. Intestinal mucosa injury. In animals of the control group, there was massive detachment of the epithelial layer from the basal membrane until the lateral part of the villi and even complete destruction and digestion of the villous architecture (Fig 1, D). Injury was attenuated in animals of the Hepat group (Fig 2, A and B). When compared with the control group, the Hepat group had a lesser IMI score (4.0 [2.0–5.0] vs 1.0 [1.0–2.0]; P = .007; Fig 1, D). Permeability to bacteria and endotoxin. Quantitative blood cultures from the portal circulation expressed as median values of log10 CFU/mL revealed positive cultures (Fig 3), suggesting decreased bacterial translocation in animals in the Hepat group that was statistically significant at 24 hours after reperfusion (3.32 log10 CFU/mL [0.00–4.36] vs 0.60 log10 CFU/mL [0.00–2.51]; P = .05). The same effect was noted in the systemic circulation as well at the end of the 24-hour reperfusion period (3.48 log10 CFU/mL [0.78–4.12] vs 0.00 log10 CFU/mL [0.00–1.48]; P = .008; Fig 3, A). Endotoxin levels in the portal circulation were decreased in animals of the Hepat group compared with controls at 24 hours after reperfusion (3.04 EU/mL [0.98–6.95] vs 0.14 EU/mL [0.05–0.83]; P = .004). Compared with control animals, endotoxin levels in the systemic circulation were also less in animals of the Hepat group at 24 hours after reperfusion (3.08 EU/mL [0.83–5.00] vs 0.74 EU/mL [0.45–2.60]; P = .025; Fig 3, B). Tissue markers of oxidative injury. The levels of MDA in the intestinal mucosa were also decreased in the Hepat group (3.85 nmol/mg [3.01–6.43] vs 3.27 nmol/mg [1.46–3.55]; P = .04), as were the levels of protein carbonyl in the Hepat group (0.57 nmol/mg [0.32–0.70] vs 0.33 nmol/mg [0.03–0.53]; P = .01; Fig 4, A). Intestinal mucosa apoptosis. Intestinal mucosa cells were positive for caspase-3 staining in the tips of the enteric villi 24 hours after liver reperfusion. There was a trend toward decreased staining positivity in the Hepat group (3.5 % stained cells [1.0–8.0] vs 1.25 % stained cells [0.5–2.0]; P = .10; Fig 4, B). DISCUSSION Bioartificial liver support systems improve postoperative liver dysfunction after major hepatectomy.12,29 Nevertheless, there has been no study evaluating the role of these systems in gut barrier dysfunction and gut-derived sepsis, which leads frequently to septic complications, multiorgan failure, and death after major liver resections. Hybrid
Surgery Volume 159, Number 6
Nastos et al 1505
Fig 1. Hemodynamic and histologic data. A, Mean arterial pressure (mm Hg). B, Portal circulation pressure (mm Hg). C, Partial pressure of CO2 in the gastric mucosa (mm Hg). D, Mucosal injury score. Data are expressed as median values ± 95% CI. Hepat, Bioartificial liver support group.
Fig 2. Intestinal mucosa from the terminal ileum 24 hours after liver reperfusion from animal of the (A) control group (injury score 3; arrow) and (B) bioartificial liver support group (injury score 0). Stain: hematoxylin and eosin; original magnification, 3100.
liver support devices could be used potentially in patients after major hepatectomy with small liver remnants in an attempt to prevent septic
complications thought to derive from translocation of gut flora. The use of these devices in such clinical scenarios might allow more aggressive liver
1506 Nastos et al
Surgery June 2016
Fig 3. Bacterial translocation and endotoxin levels. (A) Bacterial translocation in the portal and systemic circulation. Data are expressed as base-10 logarithm of colony-forming units (CFU) cultured per milliliter of blood sampled. *P < .05; **P < .01. (B) Endotoxin concentration in portal and systemic circulation expressed as endotoxin units per milliliter of blood. Data are expressed as median values ± 95% CI. *P < .05. Hepat, Bioartificial liver support group.
resections and potentially assist in the management of patients who are now considered inoperable. The aim of this study was to evaluate the role of extracorporeal bioartificial liver in the attenuation of bacterial translocation and gutderived sepsis after major hepatectomy in a large animal model. After major hepatectomy, septic complication can occur in #20% of patients.1-3 The microorganisms involved are thought to derive from the gut microflora.30,31 Normal mucosal barrier function prevents bacteria and their toxins from translocating into the circulation. A number of normal defensive mechanisms that contribute to the intestinal mucosal integrity and gut barrier are compromised post hepatectomy, as described previously.32,33,34
In addition, portal hypertension has been documented after major hepatectomy and can lead to congestion of the intestine with failure of the gut mucosal barrier.35 Also, when liver ischemia/reperfusion takes place, reactive oxygen species production may cause remote injury to the intestinal mucosa via the systemic circulation.32 Abnormal hemodynamic parameters of the systemic circulation can also compromise gut barrier function. As we have shown, mean arterial pressure can decrease after major hepatectomy and with postoperative liver dysfunction, thus compromising intestinal mucosal integrity. Extracorporeal bioartificial liver support seems to attenuate this decrease in arterial pressure, by helping to maintain adequate intestinal perfusion.12 In our study, gastric tonometry did not suggest any significant
Surgery Volume 159, Number 6
Nastos et al 1507
Fig 4. Markers of intestinal oxidative injury. (A) Intestinal mucosa tissue protein carbonyl and malondialdehyde concentration expressed in nanomoles per milligram 24 hours after reperfusion. Data are expressed as median values ± 95% CI (*P < .05). (B) Percentage of positive caspase-3 immunohistochemical staining of enterocytes 24 hours after liver reperfusion. Data are expressed as median percentage of cells ± 95% CI. Hepat, Bioartificial liver support group.
differences in splachnic perfusion (at least to the stomach) between groups, suggesting that bioartificial liver improved gut barrier function through other pathways. Acute portal hypertension causes splachnic venous congestion and injury to the pancreas and intestine.36,37 In our study, portal pressure was monitored during ischemia and reperfusion and did not show differences between groups, thus excluding portal hypertension as a confounding factor for the differences in gut mucosa injury and bacterial and endotoxin translocation. During reperfusion, an increasing trend in portal pressure was noted within each group. This early trend, which began after reperfusion, was probably owing to the reperfusion of the small liver remnant. After major liver resection, there is increased vascular resistance in the portal venous system owing to the decreased total intrahepatic vasculature.38 In addition, microcirculatory failure owing to accumulation of platelets and macrophages can contribute to this phenomenon.39 In the present study, we demonstrated attenuation of bacterial translocation in both the portal and systemic circulations in animals treated with bioartificial liver support after the treatment session. The bioreactor filter has a small pore size and by design is not permeable to bacteria. As a result, the decrease in bacteria cultures in the portal ad systemic circulation in the Hepat group cannot be attributed to clearance of bacteria from the system, but rather in the protection of the integrity of the mucosal barrier by the system. In contrast, the decrease in endotoxin translocation in animals subjected to extracorporeal bioartificial liver
support shown in our study could be explained by endotoxin clearance from cells within the bioreactor. The size of the pores of the bioreactor was 0.2 mm, allowing passage of molecules of #3,000 kDa. Endotoxins have a size that ranges from 10 to 1,000 kDa.40 Endotoxin clearance from the liver has been shown to correlate with functional hepatic mass,41 and as a result could be increased during bioartificial liver support. These findings also correlate with our findings for IMI score. Control animals developed moderate to severe mucosal damage, whereas animals in the Hepat group only developed mild mucosal injury, and in 1 animal only rare foci of grade 1 injury were noted. IMI score after major hepatectomy has been reported by Wang et al and Xu et al in rats as well,33,42 correlating with endotoxemia and the degree of liver resection. Oxidative injury results from liver ischemia and reperfusion. Oxidative and inflammatory markers “spilled” in the systemic circulation can lead to injury of other distant organs, including the intestinal mucosa.43-49 In our study, the concentration of MDA and protein carbonyls where measured to associate tissue injury with oxidative stress produced by the 150 minutes of vascular occlusion of the liver. Tissue oxidative markers were decreased in the Hepat group. A possible explanation for this observation is that the hepatocytes in the bioreactor contributed to the intrinsic liver antioxidative mechanisms which were compromised in the animals of the control group. This effect has been studied in the past as well in bioartificial liver devices, leading to amelioration
1508 Nastos et al
of the oxidative stress induced by liver failure, including hepatic encephalopathy and pulmonary injury.29,50 Apoptosis plays a major role in mucosal injury. Oxidative stress can result in injury of cellular membranes and DNA in epithelial cells of intestinal mucosa, activating the apoptotic path and disrupting the integrity of the gut mucosa barrier.51,52 Ikeda et al53 have shown that the typical manifestation of ischemia/reperfusion injury in the intestinal mucosa is the detachment of enterocytes from the basal membrane and that apoptosis is the major mechanism involved in this process concurrently with necrosis. Morphologically, the detached enterocytes create a subepithelial space and intracellular “bubbles,” which are responsible for the creation of the space of Gruenhagen, which results in the final detachment of the epithelial layer form the basal membrane.54 Recent studies have associated apoptosis with the interaction of the cellular molecules and the extracellular matrix with cellular death being noted when these interactions are lost.55 In the present study, the aforementioned pathologic space was observed in the control group but was absent in the Hepat group. Despite these differences, however, there was no difference in the immunohistochemical expression of caspase-3. This ostensible discrepancy may be explained by the fact that ischemia/reperfusion and oxidative stress induces a type II route of apoptosis in which caspase-3 is only partially involved.56 Another possible mechanism that may explain the findings of our study is the possible clearance of cytokines and other inflammatory mediators by the hepatocyte bioreactor. Because molecular weights of these substances are relatively small, these mediators could easily pass through the pores of the bioreactor. Although clearance of circulating cytokines is not proven by our study, such cytocine clearance has been shown in studies with bioartificial and nonbioartificial extracorporeal liver support systems in patients with acute liver failure.57,58 In the study using a molecular adsorbent recycling system, Guo et al58 showed a significant decrease in tumor necrosis factor-a, interleukins 6 and 8, and interferon-g, together with marked decreases in other, non–water-soluble albumin-bound toxins and water-soluble toxins that were associated with an improvement in hepatic encephalopathy. These systems, however, have failed to show any substantial benefit in septic parameters and complications of acute liver failure.59 In contrast, Liu et al studied a bioartificial liver support system with porcine hepatocytes and
Surgery June 2016
showed that proinflammatory cytokines were removed by the bioreactor, which seemed to be effective in blocking the transition of the systemic inflammatory response syndrome to multiorgan dysfunction syndrome in patients with acute and subacute liver failure.57 A limitation of our study is that the monitoring period after the induction of reperfusion was restricted to 24 hours; this is a relatively short period, and no conclusions can be made for the survival or the late evolution of bacterial translocation and gut barrier function. In addition, we only evaluated the hematogenous route of translocation and not the lymphatic route which has been described to contribute to the development of gut-derived sepsis60; we were unable to assess the mesenteric lymph nodes which would compromise the integrity of the gut barrier by itself. Finally, anaerobic bacteria were not cultured in our samples, but their role in bacterial translocation remains controversial.61 In conclusion, our study shows that dysfunction of the gut mucosal barrier is ameliorated by the use of bioartificial liver in the setting of acute liver injury after major hepatectomy. This finding could imply that the use of such devices postoperatively could permit even more aggressive liver surgery with more extensive liver resections which could be tested in future clinical trials. REFERENCES 1. Shigeta H, Nagino M, Kamiya J, Uesaka K, Sano T, Yamamoto H, et al. Bacteremia after hepatectomy: an analysis of a single-center, 10-year experience with 407 patients. Langenbecks Arch Surg 2002;387:117-24. 2. Rolando N, Harvey F, Brahm J, Philpott-Howard J, Alexander G, Gimson A, et al. Prospective study of bacterial infection in acute liver failure: an analysis of fifty patients. Hepatology 1990;11:49-53. 3. Wyke RJ, Canalese JC, Gimson AE, Williams R. Bacteraemia in patients with fulminant hepatic failure. Liver 1982;2:45-52. 4. Wang XD, Soltesz V, Andersson R, Bengmark S. Bacterial translocation in acute liver failure induced by 90 per cent hepatectomy in the rat. Br J Surg 1993;80:66-71. 5. Smyrniotis V, Farantos C, Kostopanagiotou G, Arkadopoulos N. Vascular control during hepatectomy: review of methods and results. World J Surg 2005;29:1384-96. 6. Garcea G, Gescher A, Steward W, Dennison A, Berry D. Oxidative stress in humans following the Pringle manoeuvre. Hepatobiliary Pancreat Dis Int 2006;5:210-4. 7. Yang JC, Ji XQ, Li CL, Lin JH, Liu XG. Impact of earlystage hepatic ischemia-reperfusion injury on other organs of rats. Di Yi Jun Yi Da Xue Xue Bao 2004;24: 1019-22. 8. Li LJ, Wu ZW, Xiao DS, Sheng JF. Changes of gut flora and endotoxin in rats with D-galactosamine-induced acute liver failure. World J Gastroenterol 2004;10:2087-90. 9. Wang X, Guo W, Wang Q, Soltesz V, Andersson R. Effects of a water-soluble ethylhydroxyethyl cellulose on
Surgery Volume 159, Number 6
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21. 22.
23.
24.
25.
gut physiology, bacteriology, and bacterial translocation in acute liver failure. J Invest Surg 1995;8:65-84. Wang X, Soltesz V, Axelson J, Andersson R. Cholecystokinin increases small intestinal motility and reduces enteric bacterial overgrowth and translocation in rats with surgically induced acute liver failure. Digestion 1996;57:67-72. Wang XD, Soltesz V, Andersson R. Cisapride prevents enteric bacterial overgrowth and translocation by improvement of intestinal motility in rats with acute liver failure. Eur Surg Res 1996;28:402-12. Arkadopoulos N, Kostopanagiotou G, Nastos C, Papalois A, Papoutsidakis N, Kalimeris K, et al. Reversal of experimental posthepatectomy liver failure in pigs: a new application of hepatocyte bioreactors. Artif Organs 2011;35:29-36. Ladurner R, Hochleitner B, Schneeberger S, Barnas U, Krismer A, Kleinsasser A, et al. Extended liver resection and hepatic ischemia in pigs: a new, potentially reversible model to induce acute liver failure and study artificial liver support systems. Eur Surg Res 2005;37:365-9. Arkadopoulos N, Defterevos G, Nastos C, Papalois A, Kalimeris K, Papoutsidakis N, et al. Development of a porcine model of post-hepatectomy liver failure. J Surg Res 2011;170:e233-42. Kasravi FB, Wang L, Wang XD, Molin G, Bengmark S, Jeppsson B. Bacterial translocation in acute liver injury induced by D-galactosamine. Hepatology 1996;23:97-103. Chang JX, Chen S, Ma LP, Jiang LY, Chen JW, Chang RM, et al. Functional and morphological changes of the gut barrier during the restitution process after hemorrhagic shock. World J Gastroenterol 2005;11:5485-91. Demetriou AA, Brown RS Jr, Busuttil RW, Fair J, McGuire BM, Rosenthal P, et al. Prospective, randomized, multicenter, controlled trial of a bioartificial liver in treating acute liver failure. Ann Surg 2004;239:660. Rozga J, Holzman MD, Ro MS, Griffin DW, Neuzil DF, Giorgio T, et al. Development of a hybrid bioartificial liver. Ann Surg 1993;217:502-9; discussion 9–11. Khalili TM, Navarro A, Ting P, Kamohara Y, Arkadopoulos N, Solomon BA, et al. Bioartificial liver treatment prolongs survival and lowers intracranial pressure in pigs with fulminant hepatic failure. Artif Organs 2001;25: 566-70. Demetriou AA, Brown RS Jr, Busuttil RW, Fair J, McGuire BM, Rosenthal P, et al. Prospective, randomized, multicenter, controlled trial of a bioartificial liver in treating acute liver failure. Ann Surg 2004;239:660-7; discussion 7–70. Washington J. The use of lysis-based procedures for blood cultures. Eur J Clin Microbiol Infect Dis 1991;10:618-9. Welch D, Scribner R, Hensel D. Evaluation of a lysis direct plating method for pediatric blood cultures. J Clin Microbiol 1985;21:955. Oishi H, Fusamoto M, Hatayama Y, Tsuchiya M, Takaoka A, Sakata Y. An automated analysis system of Limulus amebocyte lysate (LAL)-endotoxin reaction kinetics using turbidimetric kinetic assay. Chem Pharm Bull (Tokyo) 1988;36: 3012-9. Yokota M, Kambayashi J, Tanaka T, Tsujinaka T, Sakon M, Mori T. A simple turbidimetric time assay of the endotoxin in plasma. J Biochem Biophys Methods 1989;18:97-104. Alexandris I, Assimakopoulos S, Vagianos C, Patsoukis N, Georgiou C, Nikolopoulou V, et al. Oxidative state in intestine and liver after partial hepatectomy in rats. Effect of bombesin and neurotensin. Clin Biochem 2004;37:350-6.
Nastos et al 1509
26. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-54. 27. Jentzsch AM, Bachmann H, Furst P, Biesalski HK. Improved analysis of malondialdehyde in human body fluids. Free Radic Biol Med 1996;20:251-6. 28. Chiu CJ, McArdle AH, Brown R, Scott HJ, Gurd FN. Intestinal mucosal lesion in low-flow states. I. A morphological, hemodynamic, and metabolic reappraisal. Arch Surg 1970; 101:478-83. 29. Arkadopoulos N, Kalimeris K, Papalois A, Nastos C, Papoutsidakis N, Defterevos G, et al. Treatment with bioartificial liver improves lung injury in a swine model of partial hepatectomy and ischemia/reperfusion. Int J Artif Organs 2010;33:105-13. 30. Wang X, Andersson R, Soltesz V, Bengmark S. Bacterial translocation after major hepatectomy in patients and rats. Arch Surg 1992;127:1101-6. 31. Ren W, Wang X, Zhang A, Li C, Chen G, Ge X, et al. Selective bowel decontamination improves the survival of 90% hepatectomy in rats. J Surg Res 2015;195:454-64. 32. Nastos C, Kalimeris K, Papoutsidakis N, Defterevos G, Pafiti A, Kalogeropoulou H, et al. Antioxidant treatment attenuates intestinal mucosal damage and gut barrier dysfunction after major hepatectomy. Study in a porcine model. J Gastrointest Surg 2011;15:809-17. 33. Wang XD, Parsson H, Andersson R, Soltesz V, Johansson K, Bengmark S. Bacterial translocation, intestinal ultrastructure and cell membrane permeability early after major liver resection in the rat. Br J Surg 1994;81:579-84. 34. Hamazaki K, Sato S, Yunoki M, Noda T, Moreira LF, Mimura H, et al. Kupffer cell function in chronic liver injury and after partial hepatectomy. Res Experiment Med 1994;194:237-46. 35. Chen WZ, Hu KP, Xu RY, Pan WD. Beneficial effect of splenic artery ligation on bacterial translocation after major liver resection in rats. Dig Liver Dis 2013;45: 233-7. 36. Garcia-Tsao G, Albillos A, Barden GE, West AB. Bacterial translocation in acute and chronic portal hypertension. Hepatology 1993;17:1081-5. 37. Miyagawa S, Makuuchi M, Kawasaki S, Kakazu T. Changes in serum amylase level following hepatic resection in chronic liver disease. Arch Surg 1994;129:634-8. 38. Wang H, Ohkohchi N, Enomoto Y, Usuda M, Miyagi S, Masuoka H, et al. Effect of portocaval shunt on residual extreme small liver after extended hepatectomy in porcine. World J Surg 2006;30:2014-22; discussion 23–4. 39. Farmer DG, Amersi F, Kupiec-Weglinski J, Busuttil RW. Current status of ischemia and reperfusion injury in the liver. Transplant Rev 2000;14:106-26. 40. Carpentier B, Gautier A, Legallais C. Artificial and bioartificial liver devices: present and future. Gut 2009;58: 1690-702. 41. Caruana JA, Camara DS, Schneeberger GJ, Nolan JP. The clearance capacity of the canine liver for a portal vein endotoxin infusion. J Surg Res 1984;37:197-201. 42. Xu CP, Liu J, Liu JC, Han DW, Zhang Y, Zhao YC. Dynamic changes and mechanism of intestinal endotoxemia in partially hepatectomized rats. World J Gastroenterol 2007; 13:3592-7. 43. Filos KS, Kirkilesis I, Spiliopoulou I, Scopa CD, Nikolopoulou V, Kouraklis G, et al. Bacterial translocation, endotoxaemia and apoptosis following Pringle manoeuvre in rats. Injury 2004;35:35-43.
1510 Nastos et al
44. Bedirli A, Gokahmetoglu S, Kucuk K, Soyuer I, Guler I, Sakrak O. Bacterial translocation after partial hepatic resection under ischemia and reperfusion in rats: incidence and time course. Turk J Med Sci 2003;33:135-40. 45. Nastos C, Kalimeris K, Papoutsidakis N, Tasoulis MK, Lykoudis PM, Theodoraki K, et al. Global consequences of liver ischemia/reperfusion injury. Oxid Med Cell Longev 2014;2014:906965. 46. Varsos P, Nastos C, Papoutsidakis N, Kalimeris K, Defterevos G, Nomikos T, et al. Desferrioxamine attenuates pancreatic injury after major hepatectomy under vascular control of the liver: experimental study in pigs. HPB Surg 2012;2012:714672. 47. Papoutsidakis N, Arkadopoulos N, Smyrniotis V, Tzanatos H, Kalimeris K, Nastos K, et al. Basic research Early myocardial injury is an integral component of experimental acute liver failure–a study in two porcine models. Arch Med Sci 2011;7:217-23. 48. Arkadopoulos N, Nastos C, Defterevos G, Kalimeris K, Papoutsidakis N, Andreadou I, et al. Pancreatic injury after major hepatectomy: a study in a porcine model. Surg Today 2012;42:368-75. 49. Kalimeris K, Nastos C, Papoutsidakis N, Xanthopoulou MN, Defterevos G, Tympa A, et al. Iron chelation prevents lung injury after major hepatectomy. Hepatol Res 2010;40:841-50. 50. Shi Q. On bioartificial liver assist system: theoretical exploration and strategies for further development. Artif Cell Blood Subst Biotechnol 2000;28:535-46. 51. Zhang C, Sheng ZY, Hu S, Gao JC, Yu S, Liu Y. The influence of apoptosis of mucosal epithelial cells on intestinal barrier integrity after scald in rats. Burns 2002;28:731-7. 52. Noda T, Iwakiri R, Fujimoto K, Matsuo S, Aw TY. Programmed cell death induced by ischemia-reperfusion in rat intestinal mucosa. Am J Physiol 1998;274:G270-6.
Surgery June 2016
53. Ikeda H, Suzuki Y, Suzuki M, Koike M, Tamura J, Tong J, et al. Apoptosis is a major mode of cell death caused by ischaemia and ischaemia/reperfusion injury to the rat intestinal epithelium. Gut 1998;42:530-7. 54. Shah KA, Shurey S, Green CJ. Apoptosis after intestinal ischemia-reperfusion injury: a morphological study. Transplantation 1997;64:1393-7. 55. Frisch SM, Francis H. Disruption of epithelial cell-matrix interactions induces apoptosis. J Cell Biol 1994;124:619-26. 56. Fujise T, Iwakiri R, Wu B, Amemori S, Kakimoto T, Yokoyama F, et al. Apoptotic pathway in the rat small intestinal mucosa is different between fasting and ischemia-reperfusion. Am J Physiol Gastrointest Liver Physiol 2006; 291:G110-6. 57. Liu Q, Duan Z, Huang C. The clinical research of hybrid bioartificial liver for blocking the SIRS to MODS stage in patients with fulminant liver failure. J Emerg Med 2002;1:003. 58. Guo LM, Liu JY, Xu DZ, Li BS, Han H, Wang LH, et al. Application of molecular adsorbents recirculating system to remove NO and cytokines in severe liver failure patients with multiple organ dysfunction syndrome. Liver Int 2003; 23:16-20. 59. Rittler P, Ketscher C, Inthorn D, Jauch KW, Hartl WH. Use of the molecular adsorbent recycling system in the treatment of postoperative hepatic failure and septic multiple organ dysfunction–preliminary results. Liver Int 2004;24: 136-41. 60. Lemaire LC, van Wagensveld BA, van Gulik TM, Dankert J, van Lanschot JJ, Gouma DJ. Bacterial translocation to the thoracic duct in a setting of ischemia, partial resection and reperfusion of the porcine liver. Dig Surg 1999;16: 222-8. 61. Berg RD. Bacterial translocation from the gastrointestinal tract. Trends Microbiol 1995;3:149-54.