Experimental model of portal hypertension and esophagogastric varices in minipigs: pressure and endoscopic pilot study

ABSTRACT Introduction: Portal hypertension still represents an important health problem worldwide. In the search for knowledge regarding this syndrome, experimental studies with animal models have proven to be useful to point the direction to be taken in future randomized clinical trials. Purpose: To validate the experimental model of portal hypertension and esophagogastric varices in a medium-sized animal. Methods: This study included five minipigs br1. Midline laparotomy with dissection of the portal vein and production of a calibrated stenosis of this vein was performed. Measurement of pressure in the portal venous and digestive endoscopic were performed before and five weeks after the production of a stenosis. Results: All animals were 8 months old, average weight of 17 ± 2.5 kg. The mean pressure of the portal vein immediately before the partial ligation of the portal vein was 8.9 + 1.6 mm Hg, with 26.6 + 5.4 mm Hg in the second measurement five weeks later (p < 0.05). No gastroesophageal varices or hypertensive portal gastropathy were seen at endoscopy procedures in our sample at any time in the study. Conclusion: Portal vein ligation in minipigs has been validated in the production of portal hypertension, but not in the formation of esophageal varices.

Esophageal and gastric varices were the origin of bleeding in 18.6 and 3.6% of these cases, respectively. Together, they share the first place of source UGB, together with peptic ulcer disease 4 . Between 2005 and 2007, 480 patients with UGB caused by rupture of esophageal varices were treated at the same center, with average of 20 cases per month 5 . It is noteworthy that, even with the recommended therapy, mortality caused by esophageal varices rupture in 50 patients seen at HCFMUSP reached 32% 5,[6][7][8][9] . This data underlines the importance of continuous refinement of the best therapeutic for the treatment of variceal bleeding.
In the search for knowledge regarding the portal hypertension syndrome, experimental studies with animal models have proven to be essential. The results of these studies are useful to point the direction to be taken in future randomized clinical trials, which are not only difficult to carry out, but longer and more costly. Rodents are the most used animals for experimental models of portal hypertension, probably because their lower cost, simplicity, space requirements and reproducibility 10 . There are several methods capable to reproduce one of the three types of portal hypertension (i.e., prehepatic and posthepatic), each with its own characteristics, results and applications 11 . Biliary tract ligation, hepatic artery embolization with 80% alcohol, infusion of non-radioactive colored tracer microspheres, intoxication by carbon tetrachloride (CCL4) and thioacetamide (TAA) are the most used intrahepatic portal hypertension models, while the gradual occlusion of the inferior vena cava (gOIVC) is the standard model for posthepatic portal hypertension [12][13][14][15][16][17][18][19] .
However, when the focus is on the study of esophageal varices, models of prehepatic portal hypertension are the ones of choice.
The most used method for this model is the partial portal vein ligation (PPVL). It is a simple, highly reproducible surgery, in which a puncture is performed along the portal vein (19-21G needle), with subsequent partial ligation of this (3-0 silk), leading to the formation of a calibrated stenosis. This is the method of choice for the study of esophageal varices because it produces more pronounced portosystemic collateral circulation in a more prevalent and rapid way (around seven days), compared with other models 10,11,20 .
When we focus on the study of esophageal varices, only medium-sized animals allow upper digestive endoscopy. A special breed of pig, known as minipig, has been used in numerous experimental studies, due to its characteristics 21,22 . This animal model has splanchnic hemodynamics and liver metabolism that are like humans' , in addition to tolerating partial occlusion of the portal vein for up to six months, a period sufficient to analyze the acute and chronic effects of portal hypertension 23 .
In addition, they have adequate size, reaching an average weight of 18 kg at 8 months, while other pig breeds reach over 100 kg at the same age. The PPVL method in minipigs was first described by Jensen et al. in 1986. In that study, there was significant increase in portal vein pressure in the five ligated animals, with varices formation after five weeks. The esophagus varices formed exhibited an estimated diameter of 5-7 mm, remaining unchanged at week 24 of the study [24][25][26] .

Animal facilities and veterinary care
Five minipigs br1 were included. The animals were provided by Minipigs (Pesquisa e Desenvolvimento LTDA) and transported to the vivarium of the Faculdade de Medicina, Universidade de São Paulo, where they were kept in conventional standard individual pens, at room temperature. The stalls were cleaned and disinfected twice a day, according to the department protocol.
The animals were weighed on arrival and fasted of solid food for 24 hours, receiving water as needed. The next day, they were transported to the LIM-37 using a specific cart and accompanied by a vivarium technician. Before the procedure, the animals were anesthetized.

Anesthetic procedure
General anesthesia was administered by an anesthesiologist. Initially, the animals were sedated (Table 1), using venoclysis in the ear. After anesthetic induction, the pigs were intubated and the endotracheal tube coupled to an inhalation anesthesia device (Takaoka ® brand, model Origami Ergo System), with a universal vaporizer, using isoflurane anesthetic liquid for maintenance of general anesthesia, along with the intravenous anesthetics listed in Table 1. The animals were monitored throughout the procedure, with continuous measurement of oxygen saturation by hemoglobin and heart rate with a Dixtal ® monitor (model DX 2010). During anesthesia, venous access was performed in the external jugular for postoperative analgesia.

Surgical procedure and pressure measurements of the portal venous system
The abdominal cavity was accessed by midline laparotomy, with subsequent dissection of the portal vein. After dissection, a catheter was introduced into the splenic vein and advanced to the portal vein, allowing the measurement of pressure in the portal venous system. Three successive pressure measurements were taken. We then placed a 3-mm diameter catheter parallel to the portal vein and ligated the two structures with a 3-mm teflon band (ASD-Bard ® ). After removal of the catheter, we produced a calibrated stenosis of the portal vein. The stenosis was proximal to the tip of the catheter previously introduced for pressure measurement. The other end of the catheter was adapted to a port-a-cath ® system and implanted subcutaneously in the anterior abdominal wall. The duration of the procedure ranged from 40 to 50 minutes.
Five weeks after the operation, the animals were again anesthetized, following the protocol described before, in order to perform three successive splenic vein pressure measurements upstream of the stenosis through the puncture of the porta-cath ® with a 19G needle.

Endoscopic procedures
The upper digestive endoscopy exams, for evaluation and photographic recording of esophagogastric varices and portal hypertensive gastropathy, were performed at two different times. The first exam, with the animal already anesthetized, was performed immediately before the surgical approach. The second followed a period of five weeks, right after the second measurement of the pressure of the portal venous system.

Post-operative care
After returning from the first anesthesia, the animals were transported to the swine maintenance sector of the Bioterium Center, using a specific cart accompanied by a vivarium technician. In the first four days after surgery, the animals received tramadol, 4 mg/kg, every 12 hours, through an intravenous catheter in the external jugular vein.

Statistical analysis
Qualitative variables were described in absolute numbers (n) and frequency (%). Quantitative variables were described as means (± standard deviation). For analysis of quantitative variables dependents, paired t-test was used. Significance level was defined as p < 0.05. Statistical analysis was performed using the IBM Statistical Package for the Social Sciences (SPSS) for Windows software, Version 24.0 (Armonk, NY, United States of America).

Results
A total of five minipigs br1 was included in our experiment. All animals were 8 months old, with an average weight of 17 ± 2.5 kg. The five animals (100%) were successfully submitted to surgical intervention for partial ligation of the portal vein. The entire cohort survived and returned for a new procedure after five weeks, with subsequent euthanasia.
The mean pressure of the portal vein immediately before the partial ligation of the portal vein was 8.9 ± 1.6 mm Hg, with 26.6 ± 5.4 mm Hg in the second measurement five weeks later (p < 0.05). No gastroesophageal varices or hypertensive portal gastropathy were seen at endoscopy procedures in our sample at any time in the study. The data are presented in Table 2.

Discussion
This is the first study that aimed to validate the use of the partial portal vein ligation technique to induce portal hypertension and esophageal varices formation in minipigs. This method was described by Jensen et al. 23 in the 1980s. In our study, we observed that this technique produced significant portal hypertension (26.6 + 5.4 mm Hg p < 0.05) in our sample (five Brazilian br1 minipigs), but without formation of esophageal varices.
The validation and development of this animal model, especially in pigs, are currently important beyond the study of portal hypertension. Partial portal vein ligation, even if selective, has been widely used in cancer surgery, specifically in extensive hepatectomies for the treatment of liver metastases. Its importance lies in the fact that it induces rapid liver hypertrophy contralateral to the ligated segment, thus reducing mortality by acute liver failure 27,28 . Therefore, better knowledge of the morbidity of this technique, particularly as it relates to the formation of esophageal varices, is specially significant.
About 15 years after the initial description of the method, the same Danish group evaluated the role of recombinant human epidermal growth in preventing complications secondary to esophageal varices sclerosis 23 . This study included 18 minipigs, all of which underwent partial ligation of the portal vein. After five weeks, this technique replicated the previous results, with the development of portal hypertension and formation of esophageal varices 29 . Similar results were also found in other animal species, ranging in size from small, such as rodents (with esophageal varices identified by portography), to medium, such as dogs [30][31][32][33][34] .
However, the formation of esophageal varices was not identified in all species. In a study by Inglés et al. 35  The main hypothesis for similar levels of portal hypertension causing or not esophageal varices in different animal models is related to the anatomical differences of the portal system among species. However, this does not apply to our work, since human and porcine portal system anatomies are similar, even when analyzing the morphological and microstructural characteristics of the portal vein [35][36][37][38] . We believe that other factors, such as the differences in the physiology and hemodynamics of the portal system among animal species, also influence the formation of esophageal varices. An example would be the predilection for different routes of portosystemic shunt. The most prevalent and prominent in humans is the coronary vein shunt, which is directly related to the formation of esophageal varices, as it drains the submucosal and paraoesophageal veins. However, the same may not hold for some animal models, which would account for an absence or lower prevalence of esophageal varices formation 39 .
Another factor that may play a relevant role in the formation of esophageal varices is the thickness of the esophageal mucosa and submucosa. Therefore, species that have such thicker esophageal layers would tend to produce less engorgement of varicose collateral veins. Ultimately, the minipigs used in Jensen et al. 25 research were the Gottingen minipigs. There are at least 14 pure breeds of minipigs, and Gottingen is one of them. We speculate that the differences between minipig br1 and the Gottingen breed may have also played a role in our findings, even though there are no studies comparing the portal vascular anatomy between these two breeds.
An issue that also deserves reflection when we analyze the absence of esophageal varices formation in our study, compared to previous studies by Jensen et al. 25 , is the subjectivity of the diagnosis of esophageal varices, especially those of small caliber. In two of their studies, there was a description of the formation of small-caliber varices, with a luminal projection of 3 mm 25,29 . In this diameter, the diagnosis of esophageal varices via upper digestive endoscopy has great interand intra-observer variability. Therefore, it is fair to conjecture that the formation of esophageal varices in this model may not be as prevalent as we initially believed.
This study has a limitation related to the small number of animals included in the experiment and few results. Other limitation is that the data obtained during the study are not supported by macroscopic or endoscopic photos. On the other hand, it has the advantage of being pioneer in the initiative of developing studies with medium-sized animals in our environment.

Data availability statement
All dataset were generated or analyzed in the current study.

Funding
Not applicable.