Successful ligation of the left and right gastric vein in a dog with congenital portosystemic shunts type Aii

Abstract

The type Aii shunt is a congenital extrahepatic portosystemic shunt (ePSS) involving the left and right gastric vein and the caudal vena cava (CVC). This report describes the case of a 6-month-old Italian greyhound diagnosed with a type Aii large-diameter ePSS. Staged surgeries were employed to completely ligate the 2 gastric veins and to avoid the risk of traumatizing the shunt vessel, CVC, and celiac artery. Clinical signs improved postoperatively, and after 3 years, ultrasonography demonstrated no evidence of reoccurrence. This procedure provides an alternative surgical option for correction of ePSS type Aii.

Key clinical message:

This case report demonstrates congenital PSS involving the left and right gastric vein and the caudal vena could be treated with both ligation of left and right gastric vein. This technique could decrease the risk of traumatizing the shunt vessel, CVC, and celiac artery.

Several types of common congenital extrahepatic portosystemic shunts (ePSS) have been described in detail, based on a combination of computed tomographic angiography (CTA), intra-operative mesenteric portovenography (IOMP), and gross anatomical findings (1–4). Extrahepatic portosystemic shunts involving the right gastric vein have been further subclassified into types Ai, Aii, Aiii, Aiv, and B (5). Extrahepatic portosystemic shunt type Aii is characterized by a double shunting conformation with 2 anastomosing shunting loops. A large right gastric vein extends left from the gastroduodenal vein, where it is joined by a large left gastric vein arising from the splenic vein. These vessels fuse to form a common stalk, which is short but large in diameter, and this enters the caudal vena cava (CVC) (6). In previous reports, this type of PSS was called a right gastric-caval shunt when composed of 2 shunt loops (7), or when composed of a single caudal loop (8). Extrahepatic portosystemic shunts Type Aii constitute 12% of congenital ePSS in the dog (8), and 50% of congenital ePSS involving the right gastric vein (5). The morphology of ePSS that arise from the right gastric vein are reported to be very wide and short, with a diameter comparable to that of the CVC, and they enter the CVC near but cranial to the celiac artery (7). A previous report suggests that ePSS type Aii should be attenuated via the epiploic foramen at the site of communication between the anomalous vessel and the CVC (9). However, in this location, the CVC, celiac artery, and shunt vessel are closely associated. Therefore, fine surgical technique and careful dissection are required before the shunt vessel can be reduced (9). Specific incidences of intraoperative hemorrhage in ePSS surgery are rarely reported; however, intraoperative hemorrhage should be considered as a possible complication (10), as hemorrhage could be major if encountered (11). It is important for surgeons to be aware of an alternative reduction or ligation site if dissection of the shunt vessels is challenging due to the anatomical location. There have been no reports that describe the attenuation or ligation site, considering the unique anatomy of ePSS type Aii. In this case report, we describe an alternative attenuation or ligation site in a case of ePSS type Aii, to avoid the risk of traumatizing the shunt vessel, CVC, and celiac artery.

Case description

A 6-month-old, intact, 4.0 kg, male Italian greyhound dog was referred to the Animal Medical Center at the Tokyo University of Agriculture and Technology, with a history of anorexia and failure to thrive. Physical examination was normal, and the complete blood (cell) count (CBC) results were within normal ranges. Serum abnormalities were detected, including low urea nitrogen concentration [7.3 mg/dL; reference range (RR): 9.2 to 29.2 mg/dL]; low creatinine (0.3 mg/dL; RR: 0.4 to 1.4 mg/dL); high alanine aminotransferase (118 U/L; RR: 17 to 78 U/L); hyperammonemia (209 μg/dL; RR: 16 to 75 μg/dL); low total protein concentration (4.8 g/dL; RR: 5.0 to 7.2 g/dL); and high total serum bile acids concentration (preprandial: 144 μmol/L; upper reference limit: 7.9 μmol/L; postprandial: 350 μmol/L; upper reference limit: 24.5 μmol/L). Computed tomographic angiography was used with the patient anesthetized to diagnose and evaluate the ePSS. One short but large diameter ePSS was identified arising from the left and right gastric veins. The left gastric vein showed communication with the dilated right gastric vein before its confluence with the pre-hepatic CVC. The shunt before entering the CVC was extremely short with a large diameter (length: 4.8 mm: diameter: 7 mm). The celiac artery was adjacent to the combined vessels. Surgery was performed 17 d after the initial consultation. Following pre-anesthetic administration of atropine sulfate (atropine sulfate injection 0.5 mg; Nipro, Osaka, Japan), 0.05 mg/kg, SQ, general anesthesia was induced and maintained using 5% isoflurane (Isoflurane for Animal Use; MSD Animal Health, Tokyo, Japan). Following midline laparotomy, the mesenteric vein was cannulated with an IV catheter; the baseline portal venous systolic pressure was 11 mmHg. To evaluate the portal blood flow (Figure 1 A), IOMP was used with an iodine contrast agent (Iopamidol 300 injection; Fuji Pharma, Tokyo, Japan). The epiploic foramen was located by isolating and retracting the duodenum in a left, ventrolateral direction; the anatomy was evaluated. The shunt vessel with large diameter could be seen entering the CVC from the left at the level of the epiploic foramen, with the celiac artery adjacent to the shunt. The proximity of these vessels made it challenging to dissect the shunt vessel safely. With a perceived increased risk of iatrogenic vessel trauma and subsequent hemorrhage, alternative surgical approaches were considered. Because the shunt vessel was arising from the right and left gastric veins, a direct approach to each vessel was elected. The dilated right gastric vein was easily located, near the left lobe of the pancreas, by retracting the duodenum in a left, ventrolateral direction (Figure 2). The dilated left gastric vein was positioned on the dorsal aspect of the stomach (Figure 3). A right-angle forceps was used to dissect around these 2 veins, and 0 silk braided sutures (Nesco suture; Alfresa Pharma, Tokyo, Japan) were placed around each vein. The right gastric vein was completely occluded temporarily. The portal systolic pressure increased to 23 mmHg. The ligated right gastric vein suture was released, and the left gastric vein was then completely occluded temporarily. The portal systolic pressure was 10 mmHg, and IOMP showed complete occlusion of blood flow in the left gastric vein and increased blood flow to the hepatic portal arborization (Figure 1 B). After 5 min, there was no visual evidence of portal hypertension and thus a complete ligation of the left gastric vein was performed using 0 silk braided suture. The patient recovered uneventfully. A second surgery was scheduled about 90 d following the first surgery, but it was postponed until 147 d following the first surgery due to financial constraints. During this period, the dog was managed with moderate protein-restricted food, and administration of lactulose. However, mild hyporexia remained. At this time, low urea nitrogen concentration (6.5 mg/dL; RR: 9.2 to 29.2 mg/dL), hyperammonemia (262 μg/dL; RR: 16 to 75 μg/dL) were detected. Based on these findings, sufficient improvement in liver function had not occurred with the first surgery and medical treatment. A second surgery was undertaken. Prior to the second surgical procedure, anesthesia was induced and maintained in the same manner as the first surgical procedure. Following midline laparotomy, the mesenteric vein was cannulated with an IV catheter; the baseline portal venous systolic pressure was 3 mmHg. Intra-operative mesenteric portovenography showed slight blood flow to the hepatic arborization and shunting blood flow from the right gastric vein into the CVC (Figure 1 C). The right gastric vein was located on the lesser curvature of the stomach and was temporarily completely occluded. Portal systolic pressures measured 6 mmHg. Intra-operative mesenteric portovenography showed completely occluded right gastric blood flow and increased blood flow to the hepatic portal arborization (Figure 1 D). After 5 min, there was no visual evidence of portal hypertension and thus the right gastric vein was completely ligated using 0 silk braided suture. Recovery from anesthesia was uneventful. The dog was discharged 3 d after the second surgery. The same medical treatment as after the first surgery was continued. On Day 17 following the second surgery, the patient had good appetite, and serum biochemistry values had improved to within normal ranges, including urea nitrogen concentration (21.5 mg/dL; RR: 9.2 to 29.2 mg/dL) and ammonia concentration (22 μg/dL; RR: 16 to 75 μg/dL). Medical management was discontinued after this follow-up. The most recent postoperative follow-up in our hospital was performed 3 y after the second surgery. There was no evidence of shunting vessels from both gastric veins to CVC based on abdominal ultrasound. Abdominal radiography showed an increased size of liver compared to before surgery. Hematologic evaluation remained improved, including urea nitrogen concentration (13 mg/dL; RR: 9.2 to 29.2 mg/dL); ammonia concentration (21 μg/dL; RR: 16 to 75 μg/dL); total protein concentration (6.5 g/dL; RR: 5.0 to 7.2 g/dL); and total serum bile acids concentration (preprandial: 10.2 μmol/L; upper reference limit: 7.9 μmol/L; postprandial: 24.5 μmol/L; upper reference limit 24.5 μmol/L).

Figure 1.

A — Dilated left (dashed arrow) and right (solid arrow) gastric veins fused with the CVC (*). B — Following complete, temporary ligation of the left gastric vein. The arrow shows the dilated right gastric vein which continues to the CVC (*). The dashed arrow shows the hepatic portal arborization. C — The shunted blood flow from the right gastric vein (dashed arrow) into the CVC (*). The arrow shows slight blood flow in the hepatic arborization. D — IOMP showing complete ligation of the right gastric vein with increased blood flow to the hepatic portal arborization.

Figure 2.

The dilated right gastric vein (arrow) near the left lobe of the pancreas (*) seen by retracting the duodenum (arrowhead). The dashed arrow indicates the stomach. Cranial is at the top of the image.

Figure 3.

The dilated left gastric vein (LGv) located on the dorsal aspect of the stomach (arrow). The spleen (arrowhead) is retracted in a left, ventrolateral direction towards the dog’s right side. Cranial is to the top of the image.

Discussion

In this case report, the right and left gastric veins were separately ligated or attenuated as an alternative to ligating the combined veins near the CVC. This report highlights that this type of surgery is an option when surgeons are concerned that there is the risk of iatrogenic trauma to the CVC, short and large PSS, or celiac artery when treating ePSS type Aii. A previous report suggests that the best attenuation site for ePSS type Aii is prehepatic CVC at the level of the epiploic foramen (9). It is the opinion of the authors that the right and left gastric veins are easier to approach in our described procedure than an approach of prehepatic CVC at the level of the epiploic foramen. Previous reports have suggested that all cases of ePSS that arise from the right gastric vein are very wide and seen to follow the lesser curvature of the stomach (5,7). Based on this information, we easily located the right gastric vein. Moreover, the dilated left gastric vein was easy to approach from the dorsal side of the stomach. Both approaches were relatively simple, requiring minimal dissection and organ manipulation due to their increased length and their isolated anatomical location, compared with the insertion of the common stalk into the CVC. Commonly, IOMP is performed following the localization and temporary complete occlusion of the shunt vessel. These IOMP studies provide different and clinically useful information compared with IOMP studies obtained before surgical manipulation of the shunt (12). In this case report, IOMP and portal pressure measurements were used to decide which veins to ligate and to assess the blood flow to the hepatic portal arborization following ligation. At the first surgery, complete ligation of the right gastric vein was avoided due to the excessive increase of portal pressure. Portal hypertension is a fatal complication of surgical management of ePSS with a 2 to 14% incidence of portal hypertension reported with acute suture ligation (13). The first surgery in this case achieved attenuation of overall shunt blood flow with a complete ligation of the left gastric vein. The degree of attenuation was based on portal pressure measurements and visual inspection for evidence of portal hypertension, such as pallor or cyanosis of the intestines, to reduce the risk of postoperative portal hypertension (14). As shunting persisted via right gastric vein following complete ligation of left gastric vein alone, additional procedures such as partial ligation or attenuation of the right gastric vein with an ameroid constrictor or cellophane band could have been considered to avoid requirement of a second surgical procedure. At the second surgery, following complete ligation of both gastric veins, IOMP showed no residual shunting into the systemic circulation despite the ligation of these vessels being remote from the insertion site into the CVC. Intra-operative mesenteric portovenography should reveal all tributaries related with ePSS vessels. If there is a perceived risk to approach the site of insertion into the systemic circulation directly, alternative attenuation sites may be considered based on the IOMP findings. In this case report, the alternative sites were the right and left gastric veins. Moreover, at the second surgery, complete ePSS ligation was safely managed with portal pressure measurements and careful observation of intestinal and abdominal organs. Attenuation with gradual occluding devices rather than complete ligation could have been an option in the presence of findings indicating portal hypertension.

In conclusion, this case report has identified another possible technique for safe and straightforward ligation/attenuation of an ePSS type Aii, based on IOMP and portal pressure measurements, considering the unique anatomy of this shunt configuration.