Laparoscopic Portosystemic Shunt Attenuation in Two Dogs*** Multimeda Content ***
Laparoscopic portosystemic shunt attenuation was successfully performed in two dogs. Aberrant vessels were noted on visual examination of the abdominal vasculature. Cellophane bands were placed around the vessel laparoscopically for attenuation. The dogs had minimal postoperative morbidity, and there was biochemical evidence of adequate shunt ligation at follow-up examination.
Introduction
During the last 10 years, the diagnostic and therapeutic uses of laparoscopy in surgery have increased dramatically in dogs and cats.1,2 Advantages of laparoscopy include the magnified viewing of intraabdominal structures, decreased surgical morbidity, reduced postoperative pain, and small integumentary scars.1,2 Disadvantages include the cost of equipment, specialized training required, and reduced tactile sensation on the part of the surgeon.1,2
Portosystemic vascular shunts are congenital defects in the normal venous return of gastrointestinal blood.3 This abnormal blood flow allows toxins and hepatotrophic agents to bypass the liver and enter the systemic circulation, resulting in a milieu of clinical signs such as hepatic encephalopathy and reduced growth.4 Attenuation of the aberrant vessel can be accomplished by suture ligation, ameroid constrictor application, intravascular coil embolization, or cellophane banding.5–8
This report presents two cases of attenuation of extrahepatic portosystemic shunt vessels using laparoscopic visualization and laparoscopically placed cellophane bands.a
Case Reports
Case No. 1
A 7-month-old, 19-kg, female Labrador retriever was evaluated at Gulf Coast Veterinary Specialists for an acute onset of ptyalism, restlessness, and mental agitation of 3 days’ duration. On physical examination, the dog was subjectively considered to be smaller in stature than normal. A complete blood count (CBC) revealed no significant abnormalities. Serum biochemical profile results showed mild elevations in alanine transferase (ALT) (83 IU/L, reference range 5 to 60 IU/L) and aspartate aminotransferase (AST) (68 IU/L, reference range 5 to 55 IU/L), albumin at the low end of normal (2.5 g/dL, reference range 2.5 to 3.6 g/dL), and a decreased globulin level (2.3 g/dL, reference range 2.8 to 4.5 g/dL). Fasting serum bile acids were normal (1.6 μmol/L, reference range <7.0 μmol/L), but postprandial bile acids were markedly elevated at 280 μmol/L (reference range <25 μmol/L). Abdominal ultrasonography revealed a subjectively small liver with no evidence of abnormal vasculature. Nuclear scintigraphy using technetium-99m-pertechnetate administered rectally revealed increased uptake in the heart as compared to the liver, with a calculated shunt fraction of 99% (reference range <15%). Results were consistent with a portosystemic shunt. The dog was given neomycinb (10 mg/kg per os [PO] q 12 hours) and lactulosec (3 mL PO q 12 hours), along with a low-protein diet for 1 week prior to surgery.
Preoperatively, the dog was medicated with oxymorphoned (0.05 mg/kg intramuscularly [IM]). Anesthesia was induced with propofole (6 mg/kg intravenously [IV]) and maintained with isofluranef in oxygen with a mechanical ventilator. Cefazoling (22 mg/kg IV) was administered q 2 hours during surgery.
The dog was placed in dorsal recumbency with no tilt to the operating table. A Veress needleh was introduced through a stab incision 1 cm caudal to the umbilicus, and the abdomen was insufflated with carbon dioxide to maintain a pressure of 12 mm Hg. Sites for trocar cannulae placement were selected to provide adequate visualization of the cranial abdomen and to allow instrumental manipulation of the liver. Each trocar/cannula unit was inserted through the skin and subcutaneous tissue via a stab incision made with a no. 11 scalpel blade. Bladeless trocarsi were then gently pushed through the abdominal musculature or linea alba, with care taken to avoid penetrating the underlying organs in the abdominal cavity. A 5-mm cannula was placed on the midline 1 cm caudal to the umbilicus. A laparoscopej was introduced through this portal, which provided visual monitoring of the introduction of two more 5-mm cannulae. These cannulae were inserted through the left and right paramedian abdominal wall, equidistant between the costochondral arches and the umbilicus. A fourth 5-mm cannula was placed in the right caudal abdomen equidistant from the umbilicus and pubic bone. The two right abdominal cannulae were used as instrument portals, while the left abdominal cannula was used for insertion of a 5-mm laparoscopic Babcock forceps.k These forceps were used to pull the duodenum toward midline and facilitate shunt dissection.
The abdomen was visually explored with the laparoscope, and a single extrahepatic portocaval shunt vessel was seen entering the caudal vena cava at the epiploic foramen. The vessel was dissected from the surrounding tissue using laparoscopic right-angled forceps and Maryland forcepsl [Video 1]. A single cellophane band,a 4 mm in width, was introduced into the abdomen and positioned around the shunt vessel. The vessel was temporarily occluded by grasping both ends of the cellophane band and applying tension. The pancreas and jejunum were evaluated visually for any congestion, change in color, or change in motility. No gross signs of portal hypertension were observed during the 5-minute observation period. The cellophane band, with an internal diameter of approximately 3 mm, was then secured with a hemostatic clipm to completely occlude the shunt vessel [Video 2].
A postoperative contrast portogram was performed to confirm the shunt location and degree of occlusion. A loop of jejunum was grasped with large, laparoscopic Babcock forceps and pulled up to the umbilical portal. The portal opening was enlarged with Metzenbaum scissors to 3 cm in length, the cannula was removed, and the jejunum was exteriorized. A mesenteric vein was catheterized with a 20-gauge over-the-needle cathetern and secured with 3-0 polydioxanone.o The remaining cannulae were removed, allowing complete deflation of the abdomen. These portals were closed using 2-0 polydioxanoneo suture in the abdominal wall, 3-0 polydioxanoneo suture in the subcutaneous tissue, and staples. The dog was moved to radiology, where contrast portography revealed complete functional shunt occlusion and radiographically normal hepatic portal perfusion. The jejunal venous catheter was removed, the vein was ligated with 3-0 polyglactin,p the jejunum was replaced into the abdomen, and the final portal was sutured as previously described.
The dog recovered uneventfully from anesthesia following the 120-minute procedure. Six hours after laparoscopy, the dog was eating and drinking, with no vomiting. Twenty-four hours after surgery, the dog’s abdomen was not painful on palpation. The dog was discharged 24 hours after surgery on lactulose and a low-protein dietq until normalization of serum bile acids. At a recheck examination 6 weeks postoperatively, the dog was clinically normal. Fasting serum bile acids were 18.1 μmol/L (reference range <13 μmol/L), and postprandial bile acids were 18.0 μmol/L at that time. A follow-up phone interview with the referring veterinarian 8 months after ligation revealed the dog had no clinical abnormalities.
Case No. 2
A 4-month-old, 1.6-kg, male Yorkshire terrier was evaluated at Gulf Coast Veterinary Specialists for acute onset of vomiting, weakness, and episodes of disorientation. Physical examination and CBC revealed no significant abnormalities. A serum biochemical profile showed mild elevations in ALT (125 IU/L) and alkaline phosphatase (389 IU/L) and decreases in albumin (1.7 g/dL) and globulin levels (2.3 g/dL). Serum ammonia was elevated at 403 μmol/L (reference range 0 to 99 μmol/L). Urinalysis revealed numerous ammonium biurate crystals and a normal (1.037) specific gravity. Nuclear scintigraphy demonstrated increased uptake in the heart as compared to the liver after rectal administration of technetium-99m-pertechnetate with a calculated shunt fraction of 95%. Results were consistent with a portosystemic shunt. The dog was given neomycinb (10 mg/kg PO q 12 hours) and lactulosec (0.5 mL PO q 12 hours), along with a low-protein dietq for 3 days prior to surgery.
Preoperatively, the dog was medicated with glycopyrrolater (0.2 mg/kg IM) and diazepams (0.2 mg/kg IV), followed by oxymorphoned (0.05 mg/kg IV). Anesthesia was induced with propofole (6 mg/kg IV) and maintained with isofluranef in oxygen with a mechanical ventilator. Cefazoling (22 mg/kg IV) was administered q 2 hours during surgery.
The dog was positioned on the operating table, and cannulae were placed in the abdomen as described for case no. 1. The abdomen was visually explored using a 5-mm 30° laparoscope,j and a single extrahepatic portocaval shunt vessel was noted at the epiploic foramen [Figure 1]. The vessel was dissected as described for case no. 1, and a single 5-mm cellophane banda was introduced into the abdomen and placed around the shunt vessel. The vessel was temporarily occluded by placing traction on the cellophane band, and the pancreas and jejunum were evaluated visually for congestion, change in color, or change in motility. No gross signs of portal hypertension were observed during a 5-minute observation period. The cellophane band, with an internal diameter of 3 mm, was secured with a hemostatic clipm to occlude the shunt by an estimated 75% [Figure 2].
The cannulae were removed, allowing abdominal deflation. Each portal was closed with 2-0 polydioxanoneo suture in the abdominal wall, 3-0 polydioxanoneo suture in the subcutaneous tissue, and staples in the skin. The dog recovered from anesthesia without complication following the 90-minute procedure. Eight hours after the laparoscopic procedure, the dog was eating and drinking. Twenty-four hours after surgery, the dog’s abdomen was not painful on palpation. Treatment with amoxicillin/clavulanic acidt (19 mg/kg PO q 12 hours for 5 days) was instituted. The dog was discharged from the hospital 24 hours after surgery on lactulosec and a low-protein dietq until normalization of serum bile acids. At a recheck examination 6 weeks postoperatively, the dog was clinically normal. Fasting serum bile acids were 0.3 μmol/L, and postprandial bile acids were 4.3 μmol/L at this time.
Discussion
In these two cases of laparoscopic portosystemic shunt ligation, the aberrant vessels were recognized on visual examination. Extrahepatic portosystemic shunting vessels are most commonly located between the portal vein and the caudal vena cava and between the left gastric vein and the caudal vena cava.4 Other locations include vessels that empty into the caudal vena cava from the splenic vein, gastroduodenal vein, or mesenteric veins.2 Portoazygos shunts and intrahepatic shunts are encountered less commonly.3,4 In these two cases, the origin of the shunt vessel was not determined with certainty, but the shunt vessels were found to enter the caudal vena cava cranial to the right phrenicoabdominal vein at the level of the epiploic foramen. This location allowed simple visual inspection and dissection, facilitated by left lateral retraction of the duodenum. Portoazygos shunts, in the authors’ experience, are more typically found in the region of the esophageal hiatus, and they are often more difficult to visualize and dissect laparoscopically.
Ideally, shunts should be identified and the anatomical location determined by ultrasonography prior to laparoscopic surgery. Although this may be possible in many cases, the shunt vessel was not identified with certainty pre-operatively in either of the cases reported here.
The most common location of entry into the caudal vena cava in cases of single extrahepatic portosystemic shunts is at the level of the epiploic foramen immediately cranial to the phrenicoabdominal vein.3 This area can be easily visualized upon laparoscopy following left lateral retraction of the duodenum. If a shunt vessel is not identified in this location, exploration of the region of the esophageal hiatus, facilitated by caudal retraction of the stomach, may identify portoazygos shunts. The omental bursa can also be opened, and branches of the portal vein can be viewed after cranial retraction of the stomach and caudal retraction of the bowel when searching for other types of portocaval shunts.
Potential complications not seen in these cases, but which might be encountered when performing laparoscopic portosystemic shunt attenuation, include inability to locate the vessel, misidentification of the shunting vessel, inadvertent penetration of organs during trocar placement, and subcutaneous emphysema.1 If experienced in locating portosystemic shunting vessels visually during traditional open abdominal approaches, a clinician can easily locate the aberrant vessel laparoscopically.
Cellophane banding was chosen for shunt attenuation in the dogs reported here. The physical nature of the cellophane band allowed easy introduction through the laparoscopic portal. Laparoscopic hemostatic clips were utilized to fasten the cellophane band in place. In previous studies, cellophane banding has provided favorable results via gradual occlusion of portosystemic shunt vessels.4,8 In a study by Hunt et al., 87% of banded extrahepatic shunt cases had normal liver function at the postoperative evaluations.4 Reasons for the remaining dogs having residual abnormalities included incomplete closure of the shunt vessel and development of multiple acquired shunts.4 A recommended internal diameter of the band ranges from 2 to 3 mm, depending on the size of the dog and the vessel.4,8 Postoperative serum bile acids in the two cases reported here were compatible with successful shunt attenuation and return of functional hepatic portal blood flow.
Intraoperatively, the response to shunt attenuation was determined by evidence of portal hypertension (e.g., intestinal and pancreatic color changes, intestinal hypermotility). Portal pressure monitoring during shunt occlusion was not performed in these cases because of the difficulty in introducing an invasive portal pressure monitoring device laparoscopically. Using intraoperative clinical parameters to assess portal hypertension has been previously reported.4,9 The estimated occlusion of shunt vessels by 75% and 100% in these two dogs was not associated with signs of portal hypertension, and both dogs recovered without incident. It may be possible to indirectly assess portal hypertension laparoscopically by measuring splenic parenchymal pressure using a percutaneously placed catheter.10 However, to the authors’ knowledge, the association between portal hypertension and splenic parenchymal pressure in dogs has not been documented.
The midline laparoscopic portal had to be enlarged to approximately 3 cm in the dog that underwent postoperative portography. This portal was arguably smaller than an open laparotomy incision, which would have been required in both dogs for traditional surgical abdominal exploration and portosystemic shunt ligation. No complications were encountered with abdominal insufflation or deflation in these two cases. The postoperative recovery time was short for both dogs, which allowed for an early release from the hospital.
These two cases demonstrated that laparoscopic portosystemic shunt attenuation can be successfully performed in the clinical setting. Laparoscopic abdominal exploration may be initially employed to locate the shunt vessel. Conversion to an open procedure is required if a shunt vessel is not definitively identified or if its location precludes straightforward laparoscopic dissection. A cellophane banding procedure was conducive to a laparoscopic approach, whereas ameroid constrictors exceed the diameter of most laparoscopic portals. The operative time in these two cases was somewhat longer than with open attenuation of single extrahepatic portosystemic shunts. Longer operative time was not associated with morbidity in these two dogs, and a decrease in this time is expected as the surgeons gain more experience.
Conclusion
Results of these cases suggest that laparoscopic portosystemic shunt attenuation is possible in dogs with extrahepatic portocaval vessels. Cellophane banding allowed for slow, gradual occlusion of vessels and alleviated the concern regarding immediate hypertensive complications when applied with partial ligation. Laparoscopic visualization of pancreatic and intestinal color changes was sufficient to assess for acute portal hypertension. Future evaluation of laparoscopy for identifying and banding portoazygos and intrahepatic vessels, and percutaneous transplenic catheterization for monitoring portal pressures are warranted.
Cellophane from commercially available tobacco packages sterilized with ethylene oxide
Biosol; Pharmacia & Upjohn, Kalamazoo, MI 49005
Lactulose solution, USP; Morton Grove Pharmaceuticals, Inc., Morton Grove, IL 60053
Numorphan; Endo Pharmaceutical, Inc., Chadds Ford, PA 19317
Rapinovet; Schering-Plough Animal Health Corp., Union, NJ 07083
Isoflo; Abbott Laboratories, Abbott Park, IL 60064
Cefazolin sodium; Apothecon Products, Princeton, NJ 08543
26120JL Verress needle; Karl Storz Veterinary Endoscopy, Goleta, CA 93117
Endopath 355; Ethicon Endosurgery, Inc., Somerville, NJ 08876
5-mm 30° laparoscope; Stryker Endoscopy, San Jose, CA 95138
Endopath 5BB; Ethicon Endosurgery, Inc., Somerville, NJ 08876
Endopath 5DCD; Ethicon Endosurgery, Inc., Somerville, NJ 08876
Ligaclip LS200; Ethicon Endosurgery, Inc., Somerville, NJ 08876
Venocath; Abbott Laboratories, Abbott Park, IL 60064
PDS; Ethicon, Somerville, NJ 08876
Vicryl; Ethicon, Somerville, NJ 08876
L/D; Hill’s Pet Nutrition, Inc., Topeka, KS 66601
Robinul; A. H. Robbins Co., Richmond, VA 23220
Diazepam; Hoffman-La Roche, Inc., Nutley, NJ 07110
Clavamox; SmithKline Beecham Pharmaceuticals, Philadelphia, PA 19101
Citation: Journal of the American Animal Hospital Association 42, 2; 10.5326/0420160
Citation: Journal of the American Animal Hospital Association 42, 2; 10.5326/0420160
Laparoscopic view of a 4-month-old, male Yorkshire terrier (case no. 2). The portosystemic shunt vessel is under the Maryland forceps, to the right of the caudal vena cava (arrow).
