Portal Modulation Effects of Terlipressin on Liver Regeneration and Survival in a Porcine Model Subjected to 90% Hepatectomy

Excessive postoperative portal pressure is associated with post-hepatectomy liver failure and small-for-size syndrome after partial liver transplantation. This study aimed to identify the portal modulation effects of terlipressin on liver regeneration and survival in a porcine model subjected to 90% hepatectomy.


Results
The 7-day survival rate was signi cantly higher in the terlipressin group than in the control group. Portal pressure in the terlipressin group was lower than that in the control group at 30 min and 1 h after hepatectomy. Total bilirubin level was lower in the terlipressin group than in the control group at 1 h and 6 h after hepatectomy. Proliferating cell nuclear antigen expression was higher in the control group than in the terlipressin group at 6 h after hepatectomy, while the proportion of Ki-67-positive cells was higher in the terlipressin group than in the control group at 7 days after hepatectomy. Endothelin-1 levels re ecting liver injury were lower in the terlipressin group than in the control group at 1 h and 6 h after hepatectomy.

Conclusion
Terlipressin could optimize liver regeneration and improve survival through rapid and effective portal modulation after extensive hepatectomy.

Background
Liver resection and transplantation are the primary strategies of curative treatment for various hepatic tumors and end-stage liver disease [1]. These surgical interventions are the best treatment options considering the long-term outcomes but could cause severe and fatal complications. Post-hepatectomy liver failure (PHLF) occurs when the small remnant liver fails to meet metabolic demands after extensive hepatectomy [2]. Small-for-size syndrome (SFSS) occurs when hepatic dysfunction develops without a speci c cause after partial liver transplantation using a small-for-size graft [3]. In the early postoperative period, increased portal pressure functions as an initial signal for liver regeneration [4]. However, excessively high portal pressure may cause the small remnant liver or graft to experience shear stress, resulting in sinusoidal endothelial injury with microcirculatory impairment and irreversible liver failure due to the disturbance of effective liver regeneration [5,6]. Therefore, excessively high portal pressure is regarded as a common pathophysiology in the development of PHLF and SFSS [7,8].
Several invasive procedures, including splenic artery ligation, splenectomy, and hemi-portocaval shunt, have been used to modulate increased portal pressure after partial liver transplantation [9,10]. However, such procedures increase the risk of severe complications and cannot be adjusted in situations where the portal pressure needs to be modi ed [11]. Alternatively, some studies have focused on pharmacologic portal modulation using splanchnic vasoactive agents [12]. A previous study in which a porcine model was subjected to 70% hepatectomy reported that terlipressin and octreotide effectively reduced excessive portal pressure and attenuated liver injury after massive hepatectomy [13]. However, to validate the effectiveness of pharmacologic portal modulation for the treatment or prevention of aggressive PHLF or SFSS in clinical settings, its effects have to be proven in a more extensive hepatectomy animal model. Therefore, in the present study, we aimed to elucidate the mechanism underlying the portal modulation effects of terlipressin on liver regeneration and survival using a porcine model subjected to 90% hepatectomy.

Study design
This study was approved by the Korea University Institutional Animal Care and Use Committee (KOREA-2016-0129-C1) and strictly followed the guide for the "Animal Research: Reporting in Vivo Experiments" [14]. The study included 65 to 85 days-old female domestic pigs [median weight 34.9 (range 28.0−39.4) kg]. Pigs were housed in a room with regulated temperature and humidity. All pigs were exposed to a 12 h light/dark cycle. Pigs fasted for 8 h prior to surgery, and all procedures were conducted under sterile conditions.
A total of 20 pigs were divided into the control (n = 10) and terlipressin (n = 10) groups. As previously described [13], the initial dose of terlipressin was determined based on that used for variceal bleeding or hepatorenal syndrome treatments in clinical settings. However, the dose was reduced due to side effects such as peripheral cyanosis or sustained tachycardia in initial cases. Therefore, 0.5 mg terlipressin (Glypressin®, Ferring, Switzerland) was injected intramuscularly three times a day just before hepatectomy to 7 days after surgery for surviving pigs. Blood samples and liver tissues were obtained, and portal pressure was measured according to the scheduled time frame (Fig. 1). All pigs were followed up until 7 days after surgery, at which time the surviving pigs were sacri ced.
Surgical procedure for 90% hepatectomy in the porcine model General anesthesia and 70% hepatectomy were performed as previously described [13]. Additional to 70% hepatectomy, half of the right lateral lobe was resected to achieve 90% hepatectomy. Glisson's pedicles to each lobe except the right lateral lobe were ligated and divided. According to the ischemic color change on the liver surface, hepatic parenchyma was transected using the clamp crushing technique. After performing 70% hepatectomy, the lateral half of the right lateral lobe was transected, paying attention not to damage the hepatic in ow and out ow. Glisson's pedicle to the right lateral lobe was temporarily clamped to reduce bleeding during parenchymal transection.
Portal pressure measurement and biochemical analysis Portal pressure was measured via direct puncture using a 24-gauge needle connected to an invasive pressure monitoring device (Vigileo Monitor, Edwards Lifesciences, Irvine, CA, USA) before administration of terlipressin, 30 min, and 1 and 6 h after hepatectomy, and on the seventh postoperative day before sacri ce. Blood samples were collected from the femoral vein before administering terlipressin, 1 and 6 h after hepatectomy, and on the seventh postoperative day. Serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), and total bilirubin levels were measured using a biochemical analyzer (Chemical analyzer AU5800, Beckman Coulter Inc., Brea, CA, USA), and prothrombin time (PT) using automated coagulation analyzer (Blood Coagulation Diagnosis analyzer, Diagnostica Stago Inc., Parsippany, NJ, USA).
The reaction was stopped by adding 50 µl of stop solution. Optical density at 450 nm was determined using a microplate reader.

Western blot
Suppressor of cytokine signaling 3 (SOCS3), total signal transducer and activator of transcription 3 (total-STAT3), phospho-STAT3, and proliferating cell nuclear antigen (PCNA) protein expression were examined using western blot. After mixing the protease inhibitor cocktail (Sigma Aldrich) in RIPA buffer, 1 mL of the mixture was added per tissue sample, and the tissue was smashed with a biomasher. Samples were centrifuged at 15,000 rpm for 15 min at 4 °C, and the supernatants were sonicated four times with 30 s bursts. Samples were re-centrifuged at 15,000 rpm for 15 min at 4 °C, and the supernatant was used for western blot. According to the manufacturer's protocol, 20 µg protein from each tissue sample was quanti ed using BCA protein assay. and used for the experiment. As shown in Table 1, the primary antibody was diluted in 1X TBST with 5% skim milk and incubated overnight at 4 °C. Furthermore, secondary antibodies (mouse anti-rabbit IgG-HRP, sc-2357, Santa Cruz Biotechnology and peroxidase anti-mouse IgG (H + L), PI-2000, Vector Laboratories) were all diluted to 1:1000 in the blocking solution (1X TBST with 5% skim milk solution) and incubated at room temperature for 1 h. Intensity for each protein was quanti ed using Image J (National Institutes of Health, Bethesda, USA). Data were normalized to the reference protein glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression.

Statistical analysis
All continuous values were presented as mean with standard deviation or median with interquartile range. Cumulative survival curves were analyzed using the Kaplan-Meier method and compared using the longrank test. Continuous variables were compared between two groups using paired Student's t-test or Mann-Whitney U test as appropriate depending on the normality of data distribution. P < 0.05 was considered to denote statistical signi cance, and P between 0.05 to 0.1 was deemed to be borderline signi cant. All statistical analyses were performed using SPSS software (version 24.0, IBM Corp., Armonk, NY, USA).

Survival
The terlipressin group showed higher 7-day survival rates than the control group (60% vs. 20%, P = 0.035, Portal pressure Portal pressure measured before hepatectomy was comparable between the terlipressin and control groups (6.6 ± 1.1 vs. 6.3 ± 0.8 mmHg, P = 0.631) (Fig. 3). In the control group, portal pressure increased 30 min after hepatectomy than that before hepatectomy with a borderline signi cance (6.3 ± 1.1 vs. 7.7 ± 2.1, mmHg, P = 0.083). Portal pressure in the terlipressin group tended to be lower than in the control group at all time points after hepatectomy. The terlipressin group showed lower portal pressure than the control group at 30 min after hepatectomy with a borderline signi cance (6.2 ± 0.9 vs. 7.7 ± 2.1 mmHg, P = 0.052), and a notable signi cance 1 h after hepatectomy (5.8 ± 1.1 vs. 7.7 ± 2.2 mmHg, P = 0.027).

Biochemical analysis
Serum AST, ALT, total bilirubin, and PT levels were evaluated to assess the extent of liver injury and functional status after extensive hepatectomy (Fig. 4) Protein expression by ELISA and western blot IL-6 and HGF are known to promote liver regeneration after hepatectomy, and ET-1 is a marker of liver injury with vascular endothelial damage [15]. IL-6 and HGF serum levels evaluated by ELISA showed no signi cant difference between the terlipressin and control groups. However, ET-1 levels were signi cantly lower in the terlipressin group than in the control group at 1 and 6 h after hepatectomy (P = 0.002 and P < 0.001, respectively) (Fig. 5).
Total-STAT3 and phospho-STAT3 play an important role in the process of liver regeneration, whereas SOCS3, one of the target genes of STAT3, functions as a negative regulator of liver regeneration [16][17][18]. There were no signi cant differences in total-STAT3, phospho-STAT3, and SOCS3 between the groups in western blot. However, PCNA expression was higher in the control group than in the terlipressin group at 6 h after hepatectomy (P = 0.043) (Fig. 6). PCNA re ects cell proliferation activity in the regenerating liver [19].

Sinusoidal hemorrhage and Ki-67
Sinusoidal hemorrhage area increased in the control group than in the terlipressin group 6 h after hepatectomy; however, the difference was not signi cant (Fig. 7a). Representative images are shown in Fig. 7B. The Ki-67-positive cell number in liver tissues was higher in the terlipressin group than in the control group on the seventh postoperative day with a borderline signi cance (P = 0.071) (Fig. 8).

Discussion
Recent advances in surgical techniques and perioperative management have enabled more extensive liver resection and transplantation with acceptable morbidity and mortality [20][21][22]. However, PHLF remains a major cause of mortality that lacks an effective treatment, while SFSS continues to be a fatal complication after partial liver transplantation [23]. Several studies have reported that excessive portal pressure to the small remnant liver or graft plays a critical role in the development of both PHLF and SFSS [4,24]. Therefore, various invasive procedures have been tried to prevent or treat these conditions. However, their effects on portal modulation are not only irreversible and unpredictable but also responsible for other severe complication [25]. Consequently, several studies have explored pharmacologic portal modulation as an alternative strategy. In addition to being non-invasive, drug dose and duration can be adjusted in pharmacologic portal modulation depending on different clinical situations. In the current study, the portal modulation effects of terlipressin, a vasopressin analog, were identi ed. Terlipressin acts selectively on the V1a receptor in the portal venous system, resulting in splanchnic vasoconstriction with limited impact on systemic circulation [26]. Terlipressin is already safely used to treat hepatorenal syndrome and acute variceal bleeding [27,28]. In a previous study conducted using rats subjected to 90% hepatectomy, terlipressin lowered portal pressure and promoted liver regeneration, resulting in the highest 1-week survival rate among various splanchnic vasoactive agents [29]. Therefore, the present study aimed to explore the clinical applicability of pharmacologic portal modulation by determining its effects using a large animal model while investigating its mechanism. The 70% hepatectomy model has been used to identify pharmacologic portal modulation effects by resecting the liver as much as the subject could survive [13,30]. In contrast, the 90% hepatectomy model has been used to investigate the effects of the pharmacologic intervention on liver regeneration and survival under extreme conditions wherein most subjects are expected to die [31]. In previous studies, portal pressure signi cantly increased after 90% hepatectomy in the porcine model, and all animals died within 51 h after hepatectomy [31,32]. In the current study, portal pressure in the control group was higher at 30 min after hepatectomy than before hepatectomy, and the pressure was maintained during the study period. On the other hand, the terlipressin group showed lower portal pressure than the control group at 30 min and 1 h after hepatectomy. The portal modulation effect of terlipressin was rapid and highly effective, considering that only one injection immediately before hepatectomy induced the changes in portal pressure for 1 h after hepatectomy in the terlipressin group.
One of the most important effects of terlipressin was the optimized modulation of the timing and degree of the liver regeneration process. It was validated based on the trends in the expression of PCNA and Ki-67, which re ected cellular proliferation activity in the regenerating liver. PCNA expression signi cantly increased in the control group than in the terlipressin group at 6 h after hepatectomy. However, the proportion of Ki-67-positive cells was higher on postoperative day 7 in the terlipressin group than in the control group. In the early stages after hepatectomy, quiescent hepatocytes enter the cell cycle (G0 to S phase), and cell division occurs to initiate liver regeneration [33]. Consequently, explosive cell division occurs as portal pressure and ow abruptly increase immediately after hepatectomy [34]. In contrast, the liver regeneration process in the terlipressin group occurred slowly; however, it was prolonged due to the portal modulation effect of terlipressin. Since dividing cells rarely function until normal microarchitectures are reformed, we assumed that well-controlled liver regeneration, especially in the early postoperative period, could be more favorable for the functional recovery of the liver [12,35]. Furthermore, after liver resection, not only parenchymal cells, such as hepatocytes, but also non-parenchymal cells (e.g., Kupffer cells and hepatic stellate cells) proliferate for liver regeneration at different time points [36]. Hepatocytes begin cellular proliferation within 24 h, followed by biliary ductal cells, Kupffer cells, and hepatic stellate cells in the subsequent 2 days. Lastly, sinusoidal endothelial cells usually start active regeneration 4 days after hepatectomy [16]. Therefore, the portal modulation effect of terlipressin in slowing down the initial process of hepatocyte proliferation may balance and optimize cellular proliferation after extensive hepatectomy.
Cytokines activated during liver regeneration were also analyzed to determine the molecular mechanism underlying the effects of terlipressin on portal modulation. IL-6 is an inducer gene involved in liver regeneration that functions by binding to its receptors in the hepatocytes and promotes STAT 3 expression [37]. Activated STAT3 induces the expression of SOCS3, and activated SOCS3 arrests IL-6induced STAT3 through negative feedback [17]. In the present study, IL-6 and SOCS3 levels showed no signi cant difference between the two groups. However, these genes were mostly involved in the early stages of liver regeneration, and other mechanisms, such as those associated with growth factors and metabolic pathways, could also affect liver regeneration collectively. Furthermore, the limitation of our study that experimental values were not measured in short intervals of time makes it di cult to identify serial changes over time.
The portal modulation effect of terlipressin has led to reduced liver injury and improved survival. The expression of ET-1, a potent vasoconstrictive peptide, is activated by sinusoidal endothelial injury [15,38].
In the current study, ET-1 levels were lower in the terlipressin group than in the control group. Although not a statistically signi cant difference, taking into account the results of histological examinations showing consistently decreased degree of sinusoidal hemorrhage in the terlipressin group, these ndings suggested that terlipressin could attenuate liver injury. Furthermore, total bilirubin levels in the terlipressin group were signi cantly lower than those in the control group. This could be due to the less endothelial injury in the terlipressin group as well as because terlipressin modulated the proportion of hepatocytes entering the cell cycle in the early postoperative period. Similar to previous studies, the study showed a high mortality rate (80%) in the control group within two days of performing 90% hepatectomy [39]. This result suggested the importance of early intervention after extensive hepatectomy. However, six pigs survived to 7 days in the terlipressin group, presenting a 7-day survival rate of 60%, which was three times higher than that of the control group.
A limitation of this study included the long time interval of measurement between 6 h and 7 days after hepatectomy. If hemodynamic changes and liver generation marker levels had been examined using a shorter interval, the mechanism of terlipressin on portal modulation could have been understood more clearly. Another limitation was that the effects of terlipressin at various doses and durations were not evaluated. Nevertheless, we con rmed the bene cial effect of terlipressin on portal modulation after extensive hepatectomy using a large animal model. Thus, this study could serve as the foundation of clinical trials for determining the effects of terlipressin on preventing or treating PHLF and SFSS.

Conclusion
Terlipressin rapidly modulated excessive portal pressure in the early postoperative period after extensive hepatectomy in a large animal model. Consequently, the modulated portal pressure could optimize the liver regeneration process, resulting in reduced liver injury and improved survival.  Figure 1 Schematic diagram of the experimental design. A total of 20 pigs were divided into two groups: control (n = 10), terlipressin (n = 10). Portal pressure measurement, blood sampling, and liver biopsy were performed according to the scheduled time frame shown.

Figure 2
Cumulative survival curves in each group after 90% hepatectomy. The 7-day survival was 20% in the control group and 60% in the terlipressin group (P = 0.035).

Figure 3
Portal pressure change in each group after 90% hepatectomy. The portal pressure of the terlipressin group was lower than that of the control group at 30 min after hepatectomy with a borderline signi cance (6.2 vs. 7.7 mmHg, P = 0.052) and a notable signi cance 1 h after hepatectomy (5.8 vs. 7.7 mmHg, P = 0.035). Dots indicate means, and whiskers indicate the standard error of means. *P < 0.05 and **P < 0.1 vs. control group at the same time point.

Figure 4
Postoperative evolutions of liver function test. a serum aspartate aminotransferase (AST); b alanine aminotransferase (ALT); c total bilirubin; d prothrombin time. Total bilirubin levels were demonstrated as fold changes over the preoperative values due to individual differences in preoperative value. Data are expressed as the median, with the 25%-75% percentiles in boxes and the 5%-95% percentiles as whiskers. *P < 0.05 and **P < 0.1 vs. control group at the same time point.

Figure 5
Postoperative change in protein levels evaluated using enzyme-linked immunosorbent assay. a interleukin (IL)-6; b hepatocyte growth factor (HGF); c endothelin-1 (ET-1). Data are expressed as the median, with the 25%-75% percentiles in boxes and the 5%-95% percentiles as whiskers. *P < 0.05 vs. control group at the same time point.  Postoperative change of sinusoidal hemorrhage. a proportion of sinusoidal hemorrhage in each group; b representative images. Data are expressed as the median, with the 25%-75% percentiles in boxes and the 5%-95% percentiles as whiskers.