- Email: [email protected]
Beneficial effect of splenic artery ligation on bacterial translocation after major liver resection in rats
Digestive and Liver Disease 45 (2013) 233–237
Contents lists available at SciVerse ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Beneficial effect of splenic artery ligation on bacterial translocation after major liver resection in rats Wen-Zhe Chen, Kun-Peng Hu, Rui-Yun Xu, Wei-Dong Pan ∗ Department of Hepatobiliary Surgery, the Third Affiliated Hospital of Sun Yat-sen University, Guangzhou 510630, PR China
a r t i c l e
i n f o
Article history: Received 16 April 2012 Accepted 23 September 2012 Available online 15 November 2012 Keywords: Major hepatectomy Splenic artery ligation Bacterial translocation Gut barrier Endotoxin
a b s t r a c t Objective: In major liver resection, bacterial translocation appears to be an important mechanism in the pathogenesis of spontaneous infection. This study was designed to investigate the effects of splenic artery ligation on bacterial translocation after major liver resection. Materials and methods: Rats were divided into three groups: the sham operation group (SO group), the two-thirds partial hepatectomy group (PHx group) and the two-thirds partial hepatectomy plus splenic artery ligation group (PHx + Sp group). Bacterial translocation, endotoxemia, d-lactic acid and intestinal histology were analyzed among three groups. Results: The rate of bacterial translocation was higher in the PHx rats than in the SO rats (65.0% vs. 6.67%; P = 0.001), so that in the PHx + Sp rats (25.0%; P = 0.011). Endotoxemia was not evident in the SO rats (0 pg/ml) and blood endotoxin levels decreased in the PHx + Sp rats (1.47 pg/ml) compared with the PHx rats (4.05 pg/ml, P < 0.001). d-lactic acid was also higher in both the PHx and PHx + Sp rats compared with the SO rats (39.09 mg/ml, 23.36 mg/ml, and 1.68 mg/ml; P < 0.01). Conclusion: Splenic artery ligation enhanced intestinal barrier function and diminished blood endotoxin levels and bacterial translocation in rats with major liver resection. © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Major liver resection has been widely used in various types of benign and malignant liver disease. Interestingly, infectious complications represent the most frequent cause of morbidity and mortality following major liver resection [1–6]. Clinical and experimental evidence have indicated that gut barrier dysfunction and translocation of bacteria from the intestinal lumen to the bloodstream are directly involved in the pathogenesis of these infections [7]. A compromised intestinal barrier function promotes the escape of enteric bacteria and endotoxins into the portal circulation, and a reduction in the functional reticuloendothelial volume permits their systemic spread [8]. As a result, septic events complicate the outcome of hepatectomized patients, which leads to high morbidity rates [9]. In major liver resection, Kupffer cells are significantly reduced. The intestinal mucosa is very important because it is the first line of defense in the prevention of bacterial translocation (BT). Therefore, maintaining the integrity of the intestinal barrier is an important goal in the management of liver resection. Portal pressure has been shown to be elevated after major liver resection [10–12], and acute
∗ Corresponding author. Tel.: +86 020 85253178; fax: +86 020 85253336. E-mail address: [email protected] (W.-D. Pan).
portal hypertension can lead to congestion and edema of the bowel wall, which could increase BT from the intestinal lumen to regional lymph nodes and/or the systemic circulation [13]. Splenic artery ligation has shown beneficial effects in rat models of major liver resection [14], but the impact on BT has not been investigated. Splenic artery ligation abscises the blood flow from the splenic vein, which might reduce portal pressure, improve intestinal mucosa and reduce BT. In this study, we investigated whether splenic artery ligation could reduce portal pressure, improve intestinal histology, decrease the permeability of the gut barrier and reduce BT and endotoxemia in rats after major liver resection. Moreover, we investigated whether the immune function changed after splenic artery ligation because the spleen is a part of the immune system.
2. Materials and methods 2.1. Animals Eighty-five female Sprague-Dawley rats with an initial weight of 180–200 g were included in this study. The rats were housed in a 12:12-h light–dark cycle under specific pathogen-free conditions and fed standard rat diet with water ad libitum (except for an overnight fast before surgery). All experiments were conducted in accordance with the National Institutes of Health Guidelines for the
1590-8658/$36.00 © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.dld.2012.09.009
234
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
Care and Use of Laboratory Animals, and all protocols were approved by the SunYat-sen University Animal Care and Use Committee.
spectrophotometric assay was used to determine the level of dlactic acid in portal vein blood [18].
2.2. Experimental protocols
2.7. Plasma endotoxin levels
The present study comprised 2 protocols. The goal of protocol 1 was to determine the relationship between splenic artery ligation and portal pressure after major liver resection. Thirty rats in protocol 1 were randomly divided into 3 groups to receive sham operation (SO group, 10 rats), two-thirds partial hepatectomy (PHx group, 10 rats) or two-thirds partial hepatectomy plus splenic artery ligation (PHx + Sp group, 10 rats). Fifty-five rats in protocol 2 were randomly divided into 3 groups to receive sham operation (SO group, 15 rats), two-thirds partial hepatectomy (PHx group, 20 rats) or two-thirds partial hepatectomy plus splenic artery ligation (PHx + Sp group, 20 rats). The goal of protocol 2 was to investigate the effects of splenic artery ligation on the pathology of the intestinal mucosa, intestinal permeability, endotoxin levels in the peripheral blood, immune function and BT after major liver resection.
Endotoxin levels were quantified by photometry using an MB80 microbiology kinetic rapid reader (Gold Mountain River Tech Development Co., Ltd., Beijing, China) and gram-negative endotoxin determination reagents according to the manufacturer’s instructions. Endotoxin concentrations are expressed as pg/ml [19]. 2.8. Light microscopy After fixation in neutral buffered formalin, the resected distal ileum segments underwent histological processing. The slides were stained with hematoxylin and eosin and analyzed under optical microscopy (Leica DMI4000B) by a single pathologist who was blinded to the sample identities. The tissue samples were classified according to the degree of tissue lesion based on the system of Chiu et al. [20] The following scores were used to define the degree of villous atrophy: 1, absent; 2, mild; 3, moderate; and 4, severe.
2.3. Experimental design 2.9. Electron microscopy Before surgery, all rats were fasted for 8 h, but water was allowed ad libitum. All experiments were performed under strict sterile conditions. The operations were performed under light ether anesthesia between 8:00 and 11:00 A.M. Before surgery, all rats received 5 ml of sterile normal saline subcutaneously to prevent dehydration. Afterward, a midline abdominal incision was made. In the SO group, the liver was only separated from its ligaments. Resection of two-thirds of the liver was performed by removal of the middle and left lateral liver lobe by placing 3/0 ligatures at the respective vascular pedicle as previously described [15]. In the PHx + Sp group, rats underwent liver resection and splenic artery ligation. After another 8-h fast, the animals were anesthetized with ether, and the abdomen was opened again under strict aseptic conditions within 48 h after surgery. In protocol 1, a small polyethylene catheter was inserted into the portal vein to determine portal pressure. In protocol 2, mesenteric lymph nodes (MLNs), ileum and blood samples were taken. 2.4. Portal pressure determination At laparotomy, portal pressure was measured by direct intraportal insertion of a small polyethylene catheter, which was connected to a pressure transducer, and this was registered using a multichannel recorder (BL-420F biological function experiment system, Chengdu Technology and Market Co., Ltd. Chengdu, China). The results are expressed as cm H2 O. 2.5. Assessment of bacterial translocation Viable bacteria in the MLNs represent BT from the lumen of the intestine [16]. Three MLNs (particularly those draining lymph from the ileum and cecum) of each rat were removed aseptically and dissected free of fat. The nodes were washed free of blood with sterile saline solution, weighed and immediately homogenized in sterile phosphate-buffered saline (PBS) for the preparation of bacterial cultures. A 0.1-ml aliquot of each homogenate was plated onto blood agar and incubated at 37 ◦ C, and the positive rate was determined after 48 h of incubation. 2.6. Intestinal permeability The d-lactic acid level in the portal plasma was determined to evaluate the changes in the permeability of the intestine [17]. A
Tissues for transmission electron microscopy (TEM) were immediately fixed in 0.1 mol/l sodium cacodylate buffer (pH 7.4) containing 2.5% glutaraldehyde for 2 h at room temperature, rinsed overnight (4 ◦ C) with 0.05 M Tris buffer (pH 7.6), washed and subsequently processed for TEM. Ileum tissues were cut into longitudinal sections, and randomly chosen enterocytes were examined in defined areas of the apical region (above the nucleus) by a single observer who was blinded to the sample identities. In sections from the ileum, the middle third of the upper half of the villi was studied. 2.10. Measurement of serum immunoglobulin Enzyme-linked immunosorbent assays (ELISAs) was performed for quantitative detection of total IgM, IgG and IgA using commercially available rat ELISA kits (Genway, San Diego, CA, USA). Peripheral venous blood specimens (1 ml) were collected from each rat, and the sera were separated and stored at −20 ◦ C until IgM, IgG and IgA titers were measured. The sera samples were diluted 1:5000, 1:80,000 and 1:2000 for the IgM, IgG and IgA measurements, respectively. A 100-l aliquot of each diluted test serum was pipetted into predesignated wells, and the microtiter plate was incubated at room temperature for 30 min. After washing the plates 4 times with appropriately diluted wash solution, 100 l of appropriately diluted enzyme-antibody conjugate (horseradish peroxidase-labeled anti-rat IgG, anti-rat IgM, or anti-rat IgA) was added to each well. After incubation for 30 min at room temperature, the plates were washed 4 times, and 100 l of the substrate solution (3,3 ,5,5 -tetramethylbenzidine and hydrogen peroxide in citric acid buffer) was added to each well. The reaction was stopped after 10 min with 100 l of stop solution (0.3 M sulfuric acid). The absorbance of each well was determined at 450 nm using a Thermo Labsystems Multiskan MK3 plate reader (Thermo Fisher Scientific Shanghai Co., Ltd, Shanghai, China). 2.11. Flow cytometry Peripheral blood samples were drawn from each rat, and the cells were evaluated by flow cytometry (BD FACSCalibur cytometer) using commercially available antibody (eBioscience, San Diego, CA, USA) for tagging. Peripheral blood cells were collected to determine the following markers in all rats: total T cells (CD3+),
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
Fig. 1. Portal pressure in the three experimental groups of rats. SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + SP = 2/3 partial hepatectomy + splenic artery ligation. The rats in the PHx group had significantly greater portal pressure compared with the SO and PHx + Sp rats. n = 10 in all groups. Data are represented by the mean ± SEM.*P < 0.01, **P < 0.01.
helper T cells (CD4+), and cytotoxic T cells (CD8+). For tagging, fluorescein isothiocyanate (FITC) anti-rat CD3 was used for total T cells, Allophycocyanin (APC) anti-rat CD4 for T helper cells, and phycoerythrin (PE) anti-rat CD8 for cytotoxic T cells. After extracellular tagging was completed, samples were incubated at room temperature for 15 min in the dark. Leukocytes were isolated from peripheral blood samples via centrifugation using a standard cell lysing protocol (flow cytometry lysing solution, Multisciences Biotech Co., Ltd, Hangzhou, China) in accordance with the manufacturer’s instructions. The cells were then rinsed once by centrifugation with PBS and resuspended in PBS for flow cytometry analysis.
2.12. Statistics Results are expressed as the mean ± SEM, median, or mean rank. Comparisons of quantitative variables among groups were made using one-way analysis of variance (ANOVA) or the corresponding nonparametric (Kruskal–Wallis) test as required. Post hoc comparisons were performed with the LSD-t or Mann–Whitney nonparametric tests. The chi-squared test or Fisher’s exact test were used for comparisons of percentages. Results with a P < 0.05 were considered to be significant.
3. Results 3.1. Portal pressure The portal pressure was 7.64 ± 0.44 cm H2 O in the SO group, and the pressure was significantly greater (13.34 ± 0.64 cm H2 O; P < 0.01) in the PHx group. In the PHx + Sp group, the portal pressure was significantly lower (10.30 ± 0.69 cm H2 O; P < 0.01) than the PHx group (Fig. 1).
235
Fig. 2. Percentage of bacterial translocation in the three experimental groups of rats. SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + Sp = 2/3 partial hepatectomy + splenic artery ligation. Bacterial translocation is presented as the percentage of rats in which bacterial translocation to the lymph nodes was detected. Bacterial translocation was significantly higher in the PHx rats (n = 20) compared with the SO (n = 15) and PHx + Sp rats (n = 20). *P = 0.001, **P = 0.011.
3.2. Bacterial translocation, plasma endotoxin and d-lactic acid levels Bacterial translocation to the MLNs was only found in 1 out of 15 (6.67%) rats in the SO group, whereas BT to the MLNs was observed in 13 out of 20 (65.0%) rats in the PHx group. The difference between the SO rats and the PHx rats was statistically significant (P = 0.001). In the PHx + Sp animals, BT only occurred in 25.0% (5/20) of the rats. Interestingly, the difference between the PHx and PHx + Sp rats was also statistically significant (P = 0.011) (Fig. 2), which indicated that splenic artery ligation markedly reduced the rate of BT. The median levels of plasma endotoxin and d-lactic acid in the SO group were 0 pg/ml and 1.68 mg/ml, respectively. In the PHx group, the median values were higher (4.05 pg/ml, P < 0.01 and 39.09 mg/ml, P < 0.01, respectively). In the PHx + Sp rats, however, the median plasma endotoxin and plasma d-lactic acid levels were lower than the PHx group (1.47 vs. 4.05 pg/ml and 23.36 vs. 39.09 mg/ml, respectively, P < 0.01 for both), and this was associated with a lower prevalence of BT (Table 1). 3.3. Ileal histology Ileal structure was altered by the presence of villous atrophy. In the nonparametric Kruskal–Wallis and Mann–Whitney tests, the mean rank of the degree of villous atrophy was significantly greater in the PHx group compared with the SO group (39.92 vs. 15.40; P < 0.01). In addition, the mean rank of the degree of villous atrophy was lower in the PHx + Sp group compared with the PHx group (25.53 vs. 39.92; P < 0.01). According to Chiu’s score, the mean rank of the degree of tissue lesion in the PHx group was significantly higher compared with the SO group (40.20 vs. 15.53; P < 0.01) and the PHx + Sp group (40.20 vs. 25.15; P < 0.01) (Table 1 and Fig. 3). Transmission electron microscopy confirmed the light microscopy findings. The ultrastructure of the distal small intestine appeared normal in the rats subjected to sham operations. A clear and intact continuous mucous layer covering the
Table 1 Effect of splenic artery ligation on plasma endotoxin, intestinal permeability and intestinal histology of major liver resection rats.
SO rats PHx rats PHx + Sp rats
Number of animals
Median of plasma endotoxin (pg/ml)
Median of plasma d-lactic acid (mg/ml)
Mean rank of the degree of villous atrophy
Mean rank of Chiu’s score
15 20 20
0 4.05* 1.47†
1.68 39.09* 23.36†
15.40 39.92* 25.53†
15.53 40.20* 25.15†
Abbreviations: SO, sham operation; PHx, two-thirds partial hepatectomy; PHx + Sp, two-thirds partial hepatectomy plus splenic artery ligation. * P < 0.01 vs. SO rats. † P < 0.01 vs. PHx rats.
236
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
Fig. 3. Ileal histology in tissues from the three experimental groups (hematoxylin and eosin staining). SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + Sp = 2/3 partial hepatectomy + splenic artery ligation. Arrow (1) shows villous fall off, atrophy, and Arrow (2) shows Grunhagen’s subepithelial space, which were more prominent in the PHx group. The images are 50× magnification.
Fig. 4. The ultrastructure of the ileum in the three experimental groups. SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + Sp = 2/3 partial hepatectomy + splenic artery ligation. Arrow (1) shows that the continuous mucous layer covering the epithelial cells was unclear, the enterocyte microvilli were shortened and damaged, while Arrow (2) shows the cellular mitochondria were swollen, and Arrow (3) displays the cell gaps were widened in the PHx rats, while these phenomena were obviously improved in the PHx + SP rats. In the SO rats, the ultrastructure of the ileum appeared normal. Also a space between 21,000 and “X”.
Table 2 Effect of splenic artery ligation on immune function of major liver resection rats.
SO rats PHx rats PHx + Sp rats
Median of Concentration of Serum IgA (g/L)
Median of Concentration of Serum IgM (g/L)
Median of Concentration of Serum IgG (g/L)
Mean of percentage of CD3 cells
Median of percentage of CD4 cells
Median of percentage of CD8 cells
Mean of the ratio CD4/CD8
0.12 0.48* 0.40†
0.21 0.17 0.20†
20.51 15.67 32.0†
24.32 ± 2.12 17.31 ± 1.71* 14.02 ± 1.99†
63.37 67.43 65.46†
38.52 35.22 37.0†
1.66 ± 0.18 1.88 ± 0.14 1.85 ± 0.18†
Abbreviations: SO, sham operation; PHx, two-thirds partial hepatectom; PHx + Sp, two-thirds partial hepatectomy plus splenic artery ligation; CD3, total T cells; CD4, helper T cells; CD8, cytotoxic T cells. * P < 0 .0167 vs. SO rats. † P > 0.05 vs. PHx rats.
epithelial cells was observed in the SO rats, whereas it was unclear or had disappeared in the rats subjected to major liver resection. Interestingly, the enterocyte microvilli were shortened or had focal damage in the PHx rats. Cellular mitochondria were either swollen with shortened, broken, or nonexistent mitochondrial ridges or showed degeneration. In addition, the cell gaps were widened. In the PHx + SP rats, these 4 index above were significantly improved (Fig. 4).
The rats in the PHx group had significantly reduced values of CD3 compared with the rats in the SO group (17.31 ± 1.71 vs. 24.30 ± 2.11; P < 0.0167); however, there was no significant difference between the PHx group and the PHx + Sp group (P = 0.214). The values for CD4, CD8 and the CD4/CD8 ratio did not show any significant differences between any two of the random group (Table 2).
3.4. Serum immunoglobulins
4. Discussion
The median value for IgA was greater in the PHx rats compared with the SO rats (0.48 vs. 0.12 g/L; P < 0.01), and the median IgA value in the PHx + Sp rats was comparable with the PHx rats (0.40 vs. 0.48 g/L, P = 0.55). The Kruskal–Wallis H test of IgM and IgG did not show any significant differences between any two of the random group (Table 2).
Although the incidence of hospital death after liver resection has generally been decreasing, postoperative infections remain common in patients who have had liver resection, especially in those undergoing major liver resection [9]. Severe postoperative infections often result in prolonged hospitalizations, rising costs of medical care, and even postoperative death [6]. Many studies
3.5. Cell population analysis using flow cytometry
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
have shown that BT is the main reason for infection after major liver resection. Therefore, lowering the rate of BT for major liver resection could play an important role in the prevention of postoperative infection. Bacterial translocation is the passage of bacteria from the gastrointestinal tract through the intestinal mucosal barrier to sterile tissues, such as the MLNs [16]. This process is stimulated by altered permeability of the intestinal epithelium, increased populations of luminal bacteria, and decreased host immune defenses. In major liver resection animal models, several mechanisms may be involved in BT. Acute portal hypertension may play a role by promoting the disruption of the gut barrier. This hypothesis was supported by the rate of BT in rats with acute prehepatic portal hypertension [13]. BT did not develop unless a severe impairment in intestinal permeability was present. The relationship between BT and intestinal permeability is supported by studies in burn patients where the impairment in permeability during the postburn period has been shown to be associated with infection by enteric organisms [21]. Consequently, lowering portal hypertension after operation may provide adequate protection of the intestinal mucosa and significantly reduce bacterial translocation. In the present study, we investigated the effects of splenic artery ligation on the portal pressure, the permeability of the intestine, and the rate of BT after major liver resection in rats, and we obtained data regarding changes in immune function. The PHx-Sp rats exhibited a significant decrease in portal pressure compared with the PHx rats. In addition, this study showed that intestinal histopathological changes were improved in the PHx-Sp group rats (e.g., partial villous atrophy was partially prevented). In parallel with these findings, we observed that bacterial translocation, blood endotoxin levels and d-lactic acid were significantly reduced in the PHx-Sp rats. These favorable effects on the intestine can be partially explained by the reduction of the portal pressure and the protection of the integrity of the mucosal barrier. The difference of the ultrastructure of the ileum between the PHx group and the PHx + Sp one was obvious. In the PHx group, the continuous mucous layer covering the epithelial cells was unclear, the cellular mitochondria were swollen, the cell gaps were widened. However, in the PHx + Sp group the continuous mucous layer covering the epithelial cells was clearer, the cellular mitochondria were normal, the cell gaps were narrowed. These results deduced that splenic artery ligation was able to improve the ultrastructure of the ileum in rats with major liver resection. Determining whether the spleen should be removed concomitantly with liver resection is controversial because the spleen is an immune organ. Traditional views lean towards keeping the spleen, but many recent studies have shown the beneficial effects of splenic artery ligation on liver resection. Splenic artery ligation has been shown to ameliorate liver injuries, promote preferable liver regeneration, and reduce the likelihood of bleeding complications and bilirubin overload, which increases the safety of hepatectomies [14,22]. The effect of splenic artery ligation on BT had not been previously reported in rats with major liver resection. Our overall findings support the beneficial effects of splenic artery ligation on major liver resection. We found that splenic artery ligation lowered postoperative portal pressure, improved histopathological changes, reduced BT and did not compromise immune functions compared with rats subjected to major liver resection without splenic artery ligation.
237
Our data indicated that splenic artery ligation did not cause changes in the immune function of the rats with liver resections, but splenic artery ligation did decrease endotoxemia and prevent BT. These effects may have been related to lower acute portal hypertension and less damage of intestinal morphology and function and permeability compared with the liver resection rats without splenic artery ligation. Conflict of interest No sources of support in the form of grants, equipment, drugs. References [1] Coppa GF, Eng K, Ranson JH, et al. Hepatic resection for metastatic colon and rectal cancer. An evaluation of preoperative and postoperative factors. Annals of Surgery 1985;202:203–8. [2] Ekberg H, Tranberg KG, Andersson R, et al. Major liver resection: perioperative course and management. Surgery 1986;100:1–8. [3] Sesto ME, Vogt DP, Hermann RE. Hepatic resection in 128 patients: a 24-year experience. Surgery 1987;102:846–51. [4] Fortner JG, Lincer RM. Hepatic resection in the elderly. Annals of Surgery 1990;211:141–5. [5] Yeh D-C, Wu C-C, Ho W-M, et al. Bacterial translocation after cirrhotic liver resection: a clinical investigation of 181 patients. Journal of Surgical Research 2003;111:209–14. [6] Capussotti L, Viganò L, Giuliante F, et al. Liver dysfunction and sepsis determine operative mortality after liver resection. British Journal of Surgery 2009;96:88–94. [7] Wang X, Andersson R, Soltesz V, et al. Bacterial translocation after major hepatectomy in patients and rats. Archives of Surgery 1992;127: 1101–6. [8] Wang XD, Pärsson H, Andersson R, et al. Bacterial translocation, intestinal ultrastructure and cell membrane permeability early after major liver resection in the rat. British Journal of Surgery 1994;81:579–84. [9] Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1803 consecutive cases over the past decade. Annals of Surgery 2002;236:397–406. [10] Kanematsu T, Takenaka K, Furuta T, et al. Acute portal hypertension associated with liver resection. Analysis of early postoperative death. Archives of Surgery 1985;120:1303–5. [11] Bruix J, Castells A, Bosch J, et al. Surgical resection of hepatocellular carcinoma in cirrhotic patients: prognostic value of preoperative portal pressure. Gastroenterology 1996;111:1018–22. [12] Iida T, Yagi S, Taniguchi K, et al. Improvement of morphological changes after 70% hepatectomy with portocaval shunt: preclinical study in porcine model. Journal of Surgical Research 2007;143:238–46. [13] Garcia-Tsao G, Albillos A, Barden GE, et al. Bacterial translocation in acute and chronic portal hypertension. Hepatology 1993;17:1081–5. [14] Arakawa Y, Shimada M, Uchiyama H, et al. Beneficial effects of splenic artery ligation on massive hepatectomy model in rats. Hepatological Research 2009;39:391–7. [15] Minato M, Houssin D, Morin J, et al. Surgically induced acute hepatic failure in the rat. European Surgical Research 1982;14:185–91. [16] Berg RD. Bacterial translocation from the gastrointestinal tract. Trends in Microbiology 1995;3:149–54. [17] Li Y, Chen Y, Zhang J, et al. Protective effect of glutamine-enriched early enteral nutrition on intestinal mucosal barrier injury after liver transplantation in rats. American Journal of Surgery 2010;199:35–42. [18] Szalay L, Umar F, Khadem A, et al. Increased plasma d-lactate is associated with the severity of hemorrhagic/traumatic shock in rats. Shock 2003;20: 245–50. [19] Dunér KI. A new kinetic single-stage limulus amoebocyte lysate method for the detection of endotoxin in water and plasma. Journal of Biochemical and Biophysical Methods 1993;26:131–42. [20] Chiu CJ, McArdle AH, Brown R, et al. Intestinal mucosal lesion in low-flow states, I. A morphological, hemodynamic, and metabolic reappraisal. Archives of Surgery 1970;101:478–83. [21] Ziegler TR, Smith RJ, O’Dwyer ST, et al. Increased intestinal permeability associated with infection in burn patients. Archives of Surgery 1988;123: 1313–9. [22] Oh JW, Ahn SM, Kim KS, et al. The role of splenic artery ligation in patients with hepatocellular carcinoma and secondary hypersplenism. Yonsei Medical Journal 2003;44:1053–8.
Contents lists available at SciVerse ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Liver, Pancreas and Biliary Tract
Beneficial effect of splenic artery ligation on bacterial translocation after major liver resection in rats Wen-Zhe Chen, Kun-Peng Hu, Rui-Yun Xu, Wei-Dong Pan ∗ Department of Hepatobiliary Surgery, the Third Affiliated Hospital of Sun Yat-sen University, Guangzhou 510630, PR China
a r t i c l e
i n f o
Article history: Received 16 April 2012 Accepted 23 September 2012 Available online 15 November 2012 Keywords: Major hepatectomy Splenic artery ligation Bacterial translocation Gut barrier Endotoxin
a b s t r a c t Objective: In major liver resection, bacterial translocation appears to be an important mechanism in the pathogenesis of spontaneous infection. This study was designed to investigate the effects of splenic artery ligation on bacterial translocation after major liver resection. Materials and methods: Rats were divided into three groups: the sham operation group (SO group), the two-thirds partial hepatectomy group (PHx group) and the two-thirds partial hepatectomy plus splenic artery ligation group (PHx + Sp group). Bacterial translocation, endotoxemia, d-lactic acid and intestinal histology were analyzed among three groups. Results: The rate of bacterial translocation was higher in the PHx rats than in the SO rats (65.0% vs. 6.67%; P = 0.001), so that in the PHx + Sp rats (25.0%; P = 0.011). Endotoxemia was not evident in the SO rats (0 pg/ml) and blood endotoxin levels decreased in the PHx + Sp rats (1.47 pg/ml) compared with the PHx rats (4.05 pg/ml, P < 0.001). d-lactic acid was also higher in both the PHx and PHx + Sp rats compared with the SO rats (39.09 mg/ml, 23.36 mg/ml, and 1.68 mg/ml; P < 0.01). Conclusion: Splenic artery ligation enhanced intestinal barrier function and diminished blood endotoxin levels and bacterial translocation in rats with major liver resection. © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Major liver resection has been widely used in various types of benign and malignant liver disease. Interestingly, infectious complications represent the most frequent cause of morbidity and mortality following major liver resection [1–6]. Clinical and experimental evidence have indicated that gut barrier dysfunction and translocation of bacteria from the intestinal lumen to the bloodstream are directly involved in the pathogenesis of these infections [7]. A compromised intestinal barrier function promotes the escape of enteric bacteria and endotoxins into the portal circulation, and a reduction in the functional reticuloendothelial volume permits their systemic spread [8]. As a result, septic events complicate the outcome of hepatectomized patients, which leads to high morbidity rates [9]. In major liver resection, Kupffer cells are significantly reduced. The intestinal mucosa is very important because it is the first line of defense in the prevention of bacterial translocation (BT). Therefore, maintaining the integrity of the intestinal barrier is an important goal in the management of liver resection. Portal pressure has been shown to be elevated after major liver resection [10–12], and acute
∗ Corresponding author. Tel.: +86 020 85253178; fax: +86 020 85253336. E-mail address: [email protected] (W.-D. Pan).
portal hypertension can lead to congestion and edema of the bowel wall, which could increase BT from the intestinal lumen to regional lymph nodes and/or the systemic circulation [13]. Splenic artery ligation has shown beneficial effects in rat models of major liver resection [14], but the impact on BT has not been investigated. Splenic artery ligation abscises the blood flow from the splenic vein, which might reduce portal pressure, improve intestinal mucosa and reduce BT. In this study, we investigated whether splenic artery ligation could reduce portal pressure, improve intestinal histology, decrease the permeability of the gut barrier and reduce BT and endotoxemia in rats after major liver resection. Moreover, we investigated whether the immune function changed after splenic artery ligation because the spleen is a part of the immune system.
2. Materials and methods 2.1. Animals Eighty-five female Sprague-Dawley rats with an initial weight of 180–200 g were included in this study. The rats were housed in a 12:12-h light–dark cycle under specific pathogen-free conditions and fed standard rat diet with water ad libitum (except for an overnight fast before surgery). All experiments were conducted in accordance with the National Institutes of Health Guidelines for the
1590-8658/$36.00 © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.dld.2012.09.009
234
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
Care and Use of Laboratory Animals, and all protocols were approved by the SunYat-sen University Animal Care and Use Committee.
spectrophotometric assay was used to determine the level of dlactic acid in portal vein blood [18].
2.2. Experimental protocols
2.7. Plasma endotoxin levels
The present study comprised 2 protocols. The goal of protocol 1 was to determine the relationship between splenic artery ligation and portal pressure after major liver resection. Thirty rats in protocol 1 were randomly divided into 3 groups to receive sham operation (SO group, 10 rats), two-thirds partial hepatectomy (PHx group, 10 rats) or two-thirds partial hepatectomy plus splenic artery ligation (PHx + Sp group, 10 rats). Fifty-five rats in protocol 2 were randomly divided into 3 groups to receive sham operation (SO group, 15 rats), two-thirds partial hepatectomy (PHx group, 20 rats) or two-thirds partial hepatectomy plus splenic artery ligation (PHx + Sp group, 20 rats). The goal of protocol 2 was to investigate the effects of splenic artery ligation on the pathology of the intestinal mucosa, intestinal permeability, endotoxin levels in the peripheral blood, immune function and BT after major liver resection.
Endotoxin levels were quantified by photometry using an MB80 microbiology kinetic rapid reader (Gold Mountain River Tech Development Co., Ltd., Beijing, China) and gram-negative endotoxin determination reagents according to the manufacturer’s instructions. Endotoxin concentrations are expressed as pg/ml [19]. 2.8. Light microscopy After fixation in neutral buffered formalin, the resected distal ileum segments underwent histological processing. The slides were stained with hematoxylin and eosin and analyzed under optical microscopy (Leica DMI4000B) by a single pathologist who was blinded to the sample identities. The tissue samples were classified according to the degree of tissue lesion based on the system of Chiu et al. [20] The following scores were used to define the degree of villous atrophy: 1, absent; 2, mild; 3, moderate; and 4, severe.
2.3. Experimental design 2.9. Electron microscopy Before surgery, all rats were fasted for 8 h, but water was allowed ad libitum. All experiments were performed under strict sterile conditions. The operations were performed under light ether anesthesia between 8:00 and 11:00 A.M. Before surgery, all rats received 5 ml of sterile normal saline subcutaneously to prevent dehydration. Afterward, a midline abdominal incision was made. In the SO group, the liver was only separated from its ligaments. Resection of two-thirds of the liver was performed by removal of the middle and left lateral liver lobe by placing 3/0 ligatures at the respective vascular pedicle as previously described [15]. In the PHx + Sp group, rats underwent liver resection and splenic artery ligation. After another 8-h fast, the animals were anesthetized with ether, and the abdomen was opened again under strict aseptic conditions within 48 h after surgery. In protocol 1, a small polyethylene catheter was inserted into the portal vein to determine portal pressure. In protocol 2, mesenteric lymph nodes (MLNs), ileum and blood samples were taken. 2.4. Portal pressure determination At laparotomy, portal pressure was measured by direct intraportal insertion of a small polyethylene catheter, which was connected to a pressure transducer, and this was registered using a multichannel recorder (BL-420F biological function experiment system, Chengdu Technology and Market Co., Ltd. Chengdu, China). The results are expressed as cm H2 O. 2.5. Assessment of bacterial translocation Viable bacteria in the MLNs represent BT from the lumen of the intestine [16]. Three MLNs (particularly those draining lymph from the ileum and cecum) of each rat were removed aseptically and dissected free of fat. The nodes were washed free of blood with sterile saline solution, weighed and immediately homogenized in sterile phosphate-buffered saline (PBS) for the preparation of bacterial cultures. A 0.1-ml aliquot of each homogenate was plated onto blood agar and incubated at 37 ◦ C, and the positive rate was determined after 48 h of incubation. 2.6. Intestinal permeability The d-lactic acid level in the portal plasma was determined to evaluate the changes in the permeability of the intestine [17]. A
Tissues for transmission electron microscopy (TEM) were immediately fixed in 0.1 mol/l sodium cacodylate buffer (pH 7.4) containing 2.5% glutaraldehyde for 2 h at room temperature, rinsed overnight (4 ◦ C) with 0.05 M Tris buffer (pH 7.6), washed and subsequently processed for TEM. Ileum tissues were cut into longitudinal sections, and randomly chosen enterocytes were examined in defined areas of the apical region (above the nucleus) by a single observer who was blinded to the sample identities. In sections from the ileum, the middle third of the upper half of the villi was studied. 2.10. Measurement of serum immunoglobulin Enzyme-linked immunosorbent assays (ELISAs) was performed for quantitative detection of total IgM, IgG and IgA using commercially available rat ELISA kits (Genway, San Diego, CA, USA). Peripheral venous blood specimens (1 ml) were collected from each rat, and the sera were separated and stored at −20 ◦ C until IgM, IgG and IgA titers were measured. The sera samples were diluted 1:5000, 1:80,000 and 1:2000 for the IgM, IgG and IgA measurements, respectively. A 100-l aliquot of each diluted test serum was pipetted into predesignated wells, and the microtiter plate was incubated at room temperature for 30 min. After washing the plates 4 times with appropriately diluted wash solution, 100 l of appropriately diluted enzyme-antibody conjugate (horseradish peroxidase-labeled anti-rat IgG, anti-rat IgM, or anti-rat IgA) was added to each well. After incubation for 30 min at room temperature, the plates were washed 4 times, and 100 l of the substrate solution (3,3 ,5,5 -tetramethylbenzidine and hydrogen peroxide in citric acid buffer) was added to each well. The reaction was stopped after 10 min with 100 l of stop solution (0.3 M sulfuric acid). The absorbance of each well was determined at 450 nm using a Thermo Labsystems Multiskan MK3 plate reader (Thermo Fisher Scientific Shanghai Co., Ltd, Shanghai, China). 2.11. Flow cytometry Peripheral blood samples were drawn from each rat, and the cells were evaluated by flow cytometry (BD FACSCalibur cytometer) using commercially available antibody (eBioscience, San Diego, CA, USA) for tagging. Peripheral blood cells were collected to determine the following markers in all rats: total T cells (CD3+),
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
Fig. 1. Portal pressure in the three experimental groups of rats. SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + SP = 2/3 partial hepatectomy + splenic artery ligation. The rats in the PHx group had significantly greater portal pressure compared with the SO and PHx + Sp rats. n = 10 in all groups. Data are represented by the mean ± SEM.*P < 0.01, **P < 0.01.
helper T cells (CD4+), and cytotoxic T cells (CD8+). For tagging, fluorescein isothiocyanate (FITC) anti-rat CD3 was used for total T cells, Allophycocyanin (APC) anti-rat CD4 for T helper cells, and phycoerythrin (PE) anti-rat CD8 for cytotoxic T cells. After extracellular tagging was completed, samples were incubated at room temperature for 15 min in the dark. Leukocytes were isolated from peripheral blood samples via centrifugation using a standard cell lysing protocol (flow cytometry lysing solution, Multisciences Biotech Co., Ltd, Hangzhou, China) in accordance with the manufacturer’s instructions. The cells were then rinsed once by centrifugation with PBS and resuspended in PBS for flow cytometry analysis.
2.12. Statistics Results are expressed as the mean ± SEM, median, or mean rank. Comparisons of quantitative variables among groups were made using one-way analysis of variance (ANOVA) or the corresponding nonparametric (Kruskal–Wallis) test as required. Post hoc comparisons were performed with the LSD-t or Mann–Whitney nonparametric tests. The chi-squared test or Fisher’s exact test were used for comparisons of percentages. Results with a P < 0.05 were considered to be significant.
3. Results 3.1. Portal pressure The portal pressure was 7.64 ± 0.44 cm H2 O in the SO group, and the pressure was significantly greater (13.34 ± 0.64 cm H2 O; P < 0.01) in the PHx group. In the PHx + Sp group, the portal pressure was significantly lower (10.30 ± 0.69 cm H2 O; P < 0.01) than the PHx group (Fig. 1).
235
Fig. 2. Percentage of bacterial translocation in the three experimental groups of rats. SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + Sp = 2/3 partial hepatectomy + splenic artery ligation. Bacterial translocation is presented as the percentage of rats in which bacterial translocation to the lymph nodes was detected. Bacterial translocation was significantly higher in the PHx rats (n = 20) compared with the SO (n = 15) and PHx + Sp rats (n = 20). *P = 0.001, **P = 0.011.
3.2. Bacterial translocation, plasma endotoxin and d-lactic acid levels Bacterial translocation to the MLNs was only found in 1 out of 15 (6.67%) rats in the SO group, whereas BT to the MLNs was observed in 13 out of 20 (65.0%) rats in the PHx group. The difference between the SO rats and the PHx rats was statistically significant (P = 0.001). In the PHx + Sp animals, BT only occurred in 25.0% (5/20) of the rats. Interestingly, the difference between the PHx and PHx + Sp rats was also statistically significant (P = 0.011) (Fig. 2), which indicated that splenic artery ligation markedly reduced the rate of BT. The median levels of plasma endotoxin and d-lactic acid in the SO group were 0 pg/ml and 1.68 mg/ml, respectively. In the PHx group, the median values were higher (4.05 pg/ml, P < 0.01 and 39.09 mg/ml, P < 0.01, respectively). In the PHx + Sp rats, however, the median plasma endotoxin and plasma d-lactic acid levels were lower than the PHx group (1.47 vs. 4.05 pg/ml and 23.36 vs. 39.09 mg/ml, respectively, P < 0.01 for both), and this was associated with a lower prevalence of BT (Table 1). 3.3. Ileal histology Ileal structure was altered by the presence of villous atrophy. In the nonparametric Kruskal–Wallis and Mann–Whitney tests, the mean rank of the degree of villous atrophy was significantly greater in the PHx group compared with the SO group (39.92 vs. 15.40; P < 0.01). In addition, the mean rank of the degree of villous atrophy was lower in the PHx + Sp group compared with the PHx group (25.53 vs. 39.92; P < 0.01). According to Chiu’s score, the mean rank of the degree of tissue lesion in the PHx group was significantly higher compared with the SO group (40.20 vs. 15.53; P < 0.01) and the PHx + Sp group (40.20 vs. 25.15; P < 0.01) (Table 1 and Fig. 3). Transmission electron microscopy confirmed the light microscopy findings. The ultrastructure of the distal small intestine appeared normal in the rats subjected to sham operations. A clear and intact continuous mucous layer covering the
Table 1 Effect of splenic artery ligation on plasma endotoxin, intestinal permeability and intestinal histology of major liver resection rats.
SO rats PHx rats PHx + Sp rats
Number of animals
Median of plasma endotoxin (pg/ml)
Median of plasma d-lactic acid (mg/ml)
Mean rank of the degree of villous atrophy
Mean rank of Chiu’s score
15 20 20
0 4.05* 1.47†
1.68 39.09* 23.36†
15.40 39.92* 25.53†
15.53 40.20* 25.15†
Abbreviations: SO, sham operation; PHx, two-thirds partial hepatectomy; PHx + Sp, two-thirds partial hepatectomy plus splenic artery ligation. * P < 0.01 vs. SO rats. † P < 0.01 vs. PHx rats.
236
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
Fig. 3. Ileal histology in tissues from the three experimental groups (hematoxylin and eosin staining). SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + Sp = 2/3 partial hepatectomy + splenic artery ligation. Arrow (1) shows villous fall off, atrophy, and Arrow (2) shows Grunhagen’s subepithelial space, which were more prominent in the PHx group. The images are 50× magnification.
Fig. 4. The ultrastructure of the ileum in the three experimental groups. SO = sham operation, PHx = 2/3 partial hepatectomy, PHx + Sp = 2/3 partial hepatectomy + splenic artery ligation. Arrow (1) shows that the continuous mucous layer covering the epithelial cells was unclear, the enterocyte microvilli were shortened and damaged, while Arrow (2) shows the cellular mitochondria were swollen, and Arrow (3) displays the cell gaps were widened in the PHx rats, while these phenomena were obviously improved in the PHx + SP rats. In the SO rats, the ultrastructure of the ileum appeared normal. Also a space between 21,000 and “X”.
Table 2 Effect of splenic artery ligation on immune function of major liver resection rats.
SO rats PHx rats PHx + Sp rats
Median of Concentration of Serum IgA (g/L)
Median of Concentration of Serum IgM (g/L)
Median of Concentration of Serum IgG (g/L)
Mean of percentage of CD3 cells
Median of percentage of CD4 cells
Median of percentage of CD8 cells
Mean of the ratio CD4/CD8
0.12 0.48* 0.40†
0.21 0.17 0.20†
20.51 15.67 32.0†
24.32 ± 2.12 17.31 ± 1.71* 14.02 ± 1.99†
63.37 67.43 65.46†
38.52 35.22 37.0†
1.66 ± 0.18 1.88 ± 0.14 1.85 ± 0.18†
Abbreviations: SO, sham operation; PHx, two-thirds partial hepatectom; PHx + Sp, two-thirds partial hepatectomy plus splenic artery ligation; CD3, total T cells; CD4, helper T cells; CD8, cytotoxic T cells. * P < 0 .0167 vs. SO rats. † P > 0.05 vs. PHx rats.
epithelial cells was observed in the SO rats, whereas it was unclear or had disappeared in the rats subjected to major liver resection. Interestingly, the enterocyte microvilli were shortened or had focal damage in the PHx rats. Cellular mitochondria were either swollen with shortened, broken, or nonexistent mitochondrial ridges or showed degeneration. In addition, the cell gaps were widened. In the PHx + SP rats, these 4 index above were significantly improved (Fig. 4).
The rats in the PHx group had significantly reduced values of CD3 compared with the rats in the SO group (17.31 ± 1.71 vs. 24.30 ± 2.11; P < 0.0167); however, there was no significant difference between the PHx group and the PHx + Sp group (P = 0.214). The values for CD4, CD8 and the CD4/CD8 ratio did not show any significant differences between any two of the random group (Table 2).
3.4. Serum immunoglobulins
4. Discussion
The median value for IgA was greater in the PHx rats compared with the SO rats (0.48 vs. 0.12 g/L; P < 0.01), and the median IgA value in the PHx + Sp rats was comparable with the PHx rats (0.40 vs. 0.48 g/L, P = 0.55). The Kruskal–Wallis H test of IgM and IgG did not show any significant differences between any two of the random group (Table 2).
Although the incidence of hospital death after liver resection has generally been decreasing, postoperative infections remain common in patients who have had liver resection, especially in those undergoing major liver resection [9]. Severe postoperative infections often result in prolonged hospitalizations, rising costs of medical care, and even postoperative death [6]. Many studies
3.5. Cell population analysis using flow cytometry
W.-Z. Chen et al. / Digestive and Liver Disease 45 (2013) 233–237
have shown that BT is the main reason for infection after major liver resection. Therefore, lowering the rate of BT for major liver resection could play an important role in the prevention of postoperative infection. Bacterial translocation is the passage of bacteria from the gastrointestinal tract through the intestinal mucosal barrier to sterile tissues, such as the MLNs [16]. This process is stimulated by altered permeability of the intestinal epithelium, increased populations of luminal bacteria, and decreased host immune defenses. In major liver resection animal models, several mechanisms may be involved in BT. Acute portal hypertension may play a role by promoting the disruption of the gut barrier. This hypothesis was supported by the rate of BT in rats with acute prehepatic portal hypertension [13]. BT did not develop unless a severe impairment in intestinal permeability was present. The relationship between BT and intestinal permeability is supported by studies in burn patients where the impairment in permeability during the postburn period has been shown to be associated with infection by enteric organisms [21]. Consequently, lowering portal hypertension after operation may provide adequate protection of the intestinal mucosa and significantly reduce bacterial translocation. In the present study, we investigated the effects of splenic artery ligation on the portal pressure, the permeability of the intestine, and the rate of BT after major liver resection in rats, and we obtained data regarding changes in immune function. The PHx-Sp rats exhibited a significant decrease in portal pressure compared with the PHx rats. In addition, this study showed that intestinal histopathological changes were improved in the PHx-Sp group rats (e.g., partial villous atrophy was partially prevented). In parallel with these findings, we observed that bacterial translocation, blood endotoxin levels and d-lactic acid were significantly reduced in the PHx-Sp rats. These favorable effects on the intestine can be partially explained by the reduction of the portal pressure and the protection of the integrity of the mucosal barrier. The difference of the ultrastructure of the ileum between the PHx group and the PHx + Sp one was obvious. In the PHx group, the continuous mucous layer covering the epithelial cells was unclear, the cellular mitochondria were swollen, the cell gaps were widened. However, in the PHx + Sp group the continuous mucous layer covering the epithelial cells was clearer, the cellular mitochondria were normal, the cell gaps were narrowed. These results deduced that splenic artery ligation was able to improve the ultrastructure of the ileum in rats with major liver resection. Determining whether the spleen should be removed concomitantly with liver resection is controversial because the spleen is an immune organ. Traditional views lean towards keeping the spleen, but many recent studies have shown the beneficial effects of splenic artery ligation on liver resection. Splenic artery ligation has been shown to ameliorate liver injuries, promote preferable liver regeneration, and reduce the likelihood of bleeding complications and bilirubin overload, which increases the safety of hepatectomies [14,22]. The effect of splenic artery ligation on BT had not been previously reported in rats with major liver resection. Our overall findings support the beneficial effects of splenic artery ligation on major liver resection. We found that splenic artery ligation lowered postoperative portal pressure, improved histopathological changes, reduced BT and did not compromise immune functions compared with rats subjected to major liver resection without splenic artery ligation.
237
Our data indicated that splenic artery ligation did not cause changes in the immune function of the rats with liver resections, but splenic artery ligation did decrease endotoxemia and prevent BT. These effects may have been related to lower acute portal hypertension and less damage of intestinal morphology and function and permeability compared with the liver resection rats without splenic artery ligation. Conflict of interest No sources of support in the form of grants, equipment, drugs. References [1] Coppa GF, Eng K, Ranson JH, et al. Hepatic resection for metastatic colon and rectal cancer. An evaluation of preoperative and postoperative factors. Annals of Surgery 1985;202:203–8. [2] Ekberg H, Tranberg KG, Andersson R, et al. Major liver resection: perioperative course and management. Surgery 1986;100:1–8. [3] Sesto ME, Vogt DP, Hermann RE. Hepatic resection in 128 patients: a 24-year experience. Surgery 1987;102:846–51. [4] Fortner JG, Lincer RM. Hepatic resection in the elderly. Annals of Surgery 1990;211:141–5. [5] Yeh D-C, Wu C-C, Ho W-M, et al. Bacterial translocation after cirrhotic liver resection: a clinical investigation of 181 patients. Journal of Surgical Research 2003;111:209–14. [6] Capussotti L, Viganò L, Giuliante F, et al. Liver dysfunction and sepsis determine operative mortality after liver resection. British Journal of Surgery 2009;96:88–94. [7] Wang X, Andersson R, Soltesz V, et al. Bacterial translocation after major hepatectomy in patients and rats. Archives of Surgery 1992;127: 1101–6. [8] Wang XD, Pärsson H, Andersson R, et al. Bacterial translocation, intestinal ultrastructure and cell membrane permeability early after major liver resection in the rat. British Journal of Surgery 1994;81:579–84. [9] Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1803 consecutive cases over the past decade. Annals of Surgery 2002;236:397–406. [10] Kanematsu T, Takenaka K, Furuta T, et al. Acute portal hypertension associated with liver resection. Analysis of early postoperative death. Archives of Surgery 1985;120:1303–5. [11] Bruix J, Castells A, Bosch J, et al. Surgical resection of hepatocellular carcinoma in cirrhotic patients: prognostic value of preoperative portal pressure. Gastroenterology 1996;111:1018–22. [12] Iida T, Yagi S, Taniguchi K, et al. Improvement of morphological changes after 70% hepatectomy with portocaval shunt: preclinical study in porcine model. Journal of Surgical Research 2007;143:238–46. [13] Garcia-Tsao G, Albillos A, Barden GE, et al. Bacterial translocation in acute and chronic portal hypertension. Hepatology 1993;17:1081–5. [14] Arakawa Y, Shimada M, Uchiyama H, et al. Beneficial effects of splenic artery ligation on massive hepatectomy model in rats. Hepatological Research 2009;39:391–7. [15] Minato M, Houssin D, Morin J, et al. Surgically induced acute hepatic failure in the rat. European Surgical Research 1982;14:185–91. [16] Berg RD. Bacterial translocation from the gastrointestinal tract. Trends in Microbiology 1995;3:149–54. [17] Li Y, Chen Y, Zhang J, et al. Protective effect of glutamine-enriched early enteral nutrition on intestinal mucosal barrier injury after liver transplantation in rats. American Journal of Surgery 2010;199:35–42. [18] Szalay L, Umar F, Khadem A, et al. Increased plasma d-lactate is associated with the severity of hemorrhagic/traumatic shock in rats. Shock 2003;20: 245–50. [19] Dunér KI. A new kinetic single-stage limulus amoebocyte lysate method for the detection of endotoxin in water and plasma. Journal of Biochemical and Biophysical Methods 1993;26:131–42. [20] Chiu CJ, McArdle AH, Brown R, et al. Intestinal mucosal lesion in low-flow states, I. A morphological, hemodynamic, and metabolic reappraisal. Archives of Surgery 1970;101:478–83. [21] Ziegler TR, Smith RJ, O’Dwyer ST, et al. Increased intestinal permeability associated with infection in burn patients. Archives of Surgery 1988;123: 1313–9. [22] Oh JW, Ahn SM, Kim KS, et al. The role of splenic artery ligation in patients with hepatocellular carcinoma and secondary hypersplenism. Yonsei Medical Journal 2003;44:1053–8.