• Open Access

Diagnosis and Surgical Cellophane Banding of an Intrahepatic Congenital Portosystemic Shunt in a Foal


  • No financial support was contributed.
  • The study was presented as a scientific poster at the annual ECEIM Meeting in Hannover, Germany, 2011.

Corresponding author: Colin C. Schwarzwald, PD Dr. med. vet., PhD, Dipl. ACVIM & ECEIM, Equine Department, Vetsuisse Faculty, University of Zurich, Winterthurerstrasse 260, 8057 Zurich, Switzerland; e-mail: cschwarzwald@vetclinics.uzh.ch.


congenital portosystemic shunt


constant rate infusion


computed tomography angiography


hepatic encephalopathy


portal hypertension


portosystemic shunt

A 9-week-old mixed breed Arab filly (75 kg) was presented to the Vetsuisse Faculty of the University of Zurich with a history of intermittent episodes of circling around the mare, incoordination, absence of menace reflex, apparent blindness, inability to nurse, lethargy, unresponsiveness, ptyalism, and bruxism starting soon after birth. The intervals between clinical episodes had become shorter and the episodes gradually increased in duration. Previous diagnostic tests, including complete blood count (CBC), plasma biochemistry profile, cerebrospinal fluid analysis, gastroscopy, and ultrasonographic examination of the liver, were indicative of hepatic encephalopathy (HE) and gastric ulcerations. A congenital portosystemic shunt (CPSS) or ascending cholangiohepatitis associated with duodenal ulcers was suspected, but not confirmed. Subsequent treatment with antibiotics, nonsteroidal anti-inflammatory drugs, and proton pump inhibitors PO had not been successful.

On the day of admission, the foal had been on pasture longer than usual and developed signs of incoordination, compulsive circling, hypersalivation, and apparent blindness. Clinical examination disclosed mild apathy, circling, ataxia, hypermetric gait of the forelimbs, and intermittent high head carriage. Facial nerve responses, including menace responses, were normal. A CBC showed mild leukocytosis (11,000 cells/μL; normal, 4,700–8,200). Plasma biochemistry analyses indicated hyperammonemia (208 μmol/L; normal, <50), increased lactate concentration (3.7 mmol/L; normal, <2), hyperbilirubinemia (4.8 mg/dL; normal, 0.5–2.3), hypoproteinemia (5.4 g/dL; normal, 6–7), and hyperglycemia (211 mg/dL; normal, 81–106). Urea, creatinine, and albumin concentrations were at the lower end of the reference range. Liver enzyme activities were within normal limits. Hepatic encephalopathy was diagnosed based on clinical and laboratory findings. Congenital PSS or (less likely) persisting ascending cholangiohepatitis, toxic hepatopathy (eg, iron intoxication, ingestion of toxic plants), or hyperornithinemia, hyperammonemia, and homocitrullinuria syndrome (an inherited disorder previously described in humans and suspected in 2 Morgan weanlings[1]) were considered as possible causes.

Medical treatment was initiated with lactulose1 (0.25 mL/kg q6h PO), metronidazole2 (25 mg/kg q12h PO), omeprazole3 (1 mg/kg q24h PO), penicillin G4 (30,000 IU/kg q6h IV), gentamicin sulfate5 (7 mg/kg q24h IV), flunixin meglumine6 (1.1 mg/kg q12h IV), and IV fluid treatment with lactated Ringer's solution7 and balanced maintenance solution8 at a rate of 60 mL/kg/d. Over the first 4 days, sarmazenil9 was administered at a dosage of 0.04 mg/kg q4h IV in an attempt to attenuate clinical signs of HE.[2]

Two days later, the foal was circling less, but was more lethargic and intermittently somnolent, had negative menace responses, and showed bruxism and salivation. Ammonia concentration was 166 μmol/L and bile acid concentration was 57 μmol/L (normal, <15) (Fig. 1). Transcutaneous ultrasonographic examination disclosed normal liver, spleen, duodenum, and kidneys and a lack of free abdominal fluid. No shunt vessel could be identified. Two transcutaneous liver biopsy specimens were obtained under ultrasonographic guidance using a core biopsy needle.10 Histopathologic examination of 2 liver biopsy specimens indicated severe lobular atrophy characterized by decreased distance between central veins and portal areas. There was marked arteriolar proliferation in the portal tract and mild proliferation of interlobular arteries. Multifocal, mild to moderate ductular hyperplasia was observed, and many of the periportal sinusoids and portal lymphatics were dilated. Portal veins were mostly normal in size and were only absent in single portal tracts. Reticulin staining indicated a mild increase of collagen in portal areas. These findings were interpreted as indicative of portal vein hypoperfusion. Bacterial cultures of the liver biopsy specimens were negative.

Figure 1.

Course of bile acid and ammonia concentrations before and after surgery. A rapid decrease in plasma concentrations is evident after surgery. The upper limits of the reference range for bile acid and ammonia concentrations, respectively, are indicated by horizontal dotted lines.

Computed tomography angiography (CTA) then was performed under general anesthesia, with the foal in sternal recumbency. The CTA study consisted of precontrast and postcontrast scans. The postcontrast scan was obtained after IV injection of an iodinated high osmolar contrast medium11 at a dosage of 700 mg I/kg. The postcontrast scan was synchronized with the peak portal enhancement using a test bolus technique. The injection duration was set to be the same as the scan duration, resulting in an injection rate of 5 mL/s. The CT study was performed on a multirow unit12 with a beam collimator of 24 data channels, using a detector row width of 1.2 mm (24 × 1.2 mm), a tube potential of 120 kV (peak voltage), and a tube current of 300 mA. Images were reconstructed with a section thickness of 1.5 mm, using a soft tissue kernel. In the postcontrast study, the portal vein tributaries could be indentified and followed to the liver hilus, ruling out presence of an extrahepatic shunt. An abnormal vessel could be identified, originating from the intrahepatic portion of the portal vein, entering the most ventral aspect of the caudal vena cava immediately caudal to the diaphragm (Fig 2). Unfortunately, breathing related motion artifacts prevented visualization of the margins of this abnormal vessel clearly enough to unquestionably diagnose an intrahepatic shunt. Nevertheless, a large abnormal vessel connecting the portal vein to the caudal vena cava was considered likely.

Figure 2.

Computed tomography angiography of thorax and abdomen. In the sagittal reconstruction, a dorsally orientated, contrast-filled vessel running toward the caudal vena cava was identified, suggestive for an intrahepatic congenital portosystemic shunt (arrow). The diagnostic quality of the image is impaired by breathing motion artifacts. CVC, caudal vena cava.

Diagnosis of a CPSS was further corroborated by ultrasound-guided percutaneous transsplenic injection of agitated saline and simultaneous echocardiography of the right heart in a right parasternal long-axis view, performed immediately after CTA, with the foal still under general anesthesia.[3, 4] Briefly, 10 mL of agitated normal saline was injected percutaneously into the splenic parenchyma, close to 2 large splenic veins with a 18g/40mm needle13 under ultrasonographic guidance. Immediately after injection, a simultaneous echocardiogram showed marked echo contrast in the right atrium and ventricle (Fig 3A–C and supporting information online only Video S1). Based on data available for dogs, agitated saline microbubbles should not cross the sinusoidal barrier and are absorbed by the liver parenchyma if no shunt is present.[3] Therefore, immediate appearance of microbubbles in the right heart was considered indicative of a direct communication between the portal vein and caudal vena cava in this foal.

Figure 3.

Echocardiographic right-parasternal long axis view of the heart before trans-splenic saline injection (A) and immediately after injection (B and C), showing marked echo contrast in the right heart. This finding was considered highly suggestive for a congenital portosystemic shunt. (See also supporting information online only Video S1, showing a video loop of the echocardiographic right-parasternal long axis view during trans-splenic saline injection.)

After diagnosis of CPSS, medical treatment was continued for 11 days. Although ammonia and bile acid concentrations remained increased (Fig 1), clinical signs improved until day 15, when the foal started showing neurologic signs again. Therefore, surgical correction of the CPSS was recommended. Eighteen days after admission, a ventral midline celiotomy was performed to determine the exact location of the intrahepatic shunt and to occlude the shunt using the cellophane banding procedure. The foal was sedated with medetomidine14 (1 μg/kg IV) and induced with propofol15 (3 mg/kg IV). General anesthesia was maintained with isoflurane,16 a constant rate infusion (CRI) of fentanyl17 (1–3 μg/kg/h IV) and 2 additional boluses of fentanyl (0.25 μg/kg IV). After aseptic preparation, a ventral midline incision from the umiblicus to the xiphoid was performed. The falciform ligament and teres hepatis then were transected and the liver was reflected to reach its parietal surface. The caudal vena cava was easily identified as it exited the liver. Subsequently, identification of the shunt was attempted by intraoperative ultrasonography.18 The liver first was scanned from the parietal surface, but the anomalous vessel could not be identified. Therefore, an approach to the portosystemic shunt (PSS) through the visceral surface of the liver was chosen, which allowed ultrasonographic visualization of the shunt vessel (Fig 4). The abnormal vessel could be identified originating from the intrahepatic portion of the portal vein and entering the most ventral aspect of the caudal vena cava, immediately caudal to the diaphragm. Because of the position of the shunt vessel deep within the liver parenchyma, it was difficult to locate it exactly within a particular lobe. However, the visualized part of the shunt was running parallel to the vena cava and was most likely located within the caudate lobe. The stomach was caudally retracted, a 3 cm incision was performed on the liver capsule using a number 10 scalpel blade over the shunt vessel, and the liver parenchyma was bluntly dissected to reach the vessel. When blood vessels were detected during the approach, they were examined ultrasonographically. The CPSS was identified about 2 cm inside the liver parenchyma, running parallel and in contact with the caudal vena cava. A tunnel was created between both vessels with ultrasonographic guidance and under visual control. A band of cellophane, 10 cm long and 1.2 cm wide, was folded longitudinally to form a 3-layered strip. The strip was passed around the shunt and the cellophane ring was secured, slightly decreasing the lumen of the shunt with 2 positioned titanium clips.19 Over a time period of 25 minutes after cellophane placement, heart rate increased slightly from 51 to 59/min and systolic/diastolic (mean) systemic blood pressures dropped slightly from 111/72 (87) mmHg to 95/57 (70) mmHg. These findings possibly were related to some degree of transient portal hypertension (PH),[5] but portal blood pressures were not measured. The cellophane band was loosened as much possible to avoid compression of the shunting vessel. Although repeated ultrasonographic examination still suggested slight narrowing of the shunt vessel (Fig 4B), heart rate and systemic blood pressures normalized, indicating that PH, if it had occurred, had most likely resolved or at least was no longer clinically relevant. The ventral midline incision then was closed by suturing the linea alba with a continuous suture pattern using an absorbable multifilament suture material.20 The subcutaneous layer was closed in the same fashion using absorbable monofilament suture material.21 The skin was stapled with a single use skin stapler22 and the suture line was covered using a stent. Fentanyl CRI was stopped 30 minutes before the end of anesthesia and further analgesia was provided with morphine23 (0.12 mg/kg IM). The foal recovered uneventfully.

Figure 4.

Ultrasonographic examination on the visceral surface of the liver. The congenital portosystemic shunt was identified about 2 cm inside the liver parenchyma, running parallel and in contact with the caudal vena cava before (A) and after cellophane placement (B). HV, hepatic vein; CVC, caudal vena cava.

After surgery, maintenance solution8 was administered at a dosage of 60 mL/kg/d IV for 5 days. Because plasma protein concentration (3.4 g/dL) and colloid osmotic pressure (9.4 mmHg; normal, >16) were low the day after surgery, the foal also was given 1500 mL of fresh frozen plasma.24 Medical treatment with lactulose1, omeprazole3, penicillin G4, and gentamicin5 were continued for 10 days after surgery. Flunixin meglumine6 was given for 5 days, metronidazole2 was given for 3 weeks, and trimethoprim sulfonamide25 (30 mg/kg q12h PO) was given for an additional 2 weeks after discontinuing IV antibiotics.

Within 2 days after surgery, neurologic signs improved and blood ammonia (50.0 μmol/L) and bile acid concentrations (24 μmol/L) had decreased markedly (Fig 1). Bruxism, ptyalism, and slight apathy persisted for 1 week. Three days after surgery, the foal developed a grade 1 rectal prolapse, which was managed with a purse-string suture that was left in place for 1 week and was intermittently loosened to allow passage of the feces. The foal did not show any signs of colic, abdominal distention, or diarrhea. Neither ascites nor thickened intestinal walls were seen on repeated ultrasonographic examinations. The foal was released from the hospital 16 days after surgery.

On re-evaluation 6 weeks after surgery, the foal appeared healthy. Blood ammonia concentration (54 μmol/L) still was slightly increased whereas bile acid concentration (5 μmol/L) was normal. The foal had gained 21 kg since surgery and had not shown any clinical signs after discharge. The liver was unremarkable on abdominal ultrasonography; the cellophane band and the shunt could not be visualized. At telephone follow-up with the owner 5 and 7 months after discharge, the foal was reported to be bright and alert and very lively. It had not shown any clinical signs and was in good body condition and growing well.

This report describes a novel, safe, and straightforward ultrasound-based diagnostic technique for the detection of CPSS in foals. Furthermore, it is the first report of successful surgical cellophane banding of an intrahepatic shunt in a foal.

A portosystemic shunt is a connection between the portal vessels and the systemic circulation, causing blood flow to bypass the liver. Intrahepatic CPSS represent the persistence of the ductus venosus or communications between the portal vein and hepatic vein or caudal vena cava. Congenital portosystemic shunts have been described in several species, including horses. The heritability of CPSS is unknown in horses, but in dogs and cats certain breeds are overrepresented.[6, 7] Clinical signs in horses usually are noticed before 1 year of age, varying from slight apathy to severe neurologic signs of HE, coma, and death.[8] Few cases of CPSS in foals have been reported in the literature[8-12] and their prevalence appears to be lower than in dogs and cats.

The clinical presentation of the foal was consistent with findings previously reported in foals and other species with HE caused by CPSS, including lethargy, disorientation, compulsive circling, ptyalism, central blindness, and bruxism.[8-12] Clinical signs in affected foals may be present from birth. However, they often get worse once the foals start eating more roughage. Because grass or hay contains proteins with lesser biologic value than milk, this may favor bacterial protein fermentation and production of ammonia in the large intestine. This also could explain why the foal showed marked clinical signs when it was on pasture for longer periods of time. Typical laboratory findings with CPSS, some of which were observed in the present case, include anemia, microcytosis, increased serum ammonia and bile acid concentrations, as well as low blood urea nitrogen, plasma protein, and albumin concentrations.[8-12] Although hyperammonemia is a hallmark of CPSS, it correlates poorly with clinical signs or severity of disease, because ammonia may be only one of many factors contributing to the development of HE.

Presence of a CPSS can be suspected based upon history, physical examination, and laboratory findings, but definitive diagnosis requires advanced imaging methods. Abdominal ultrasonography has been considered the method of choice to confirm CPSS in dogs and cats, but is very operator dependent.[13, 14] Ultrasonography performed by an experienced operator may allow identification of the exact position and morphology of the shunt vessel and assessment of abnormalities of portal venous flow, portal branches, liver parenchyma, and kidney size.[13] In the present case, transcutaneous ultrasonographic examination did not allow identification of a shunt vessel or any other abnormalities.

Liver biopsy specimens can be helpful in the diagnosis of a CPSS.[10, 11] However, the response of the liver to portal vein hypoperfusion is similar, regardless of the underlying cause. Therefore, additional clinical information is often required for final diagnosis. Furthermore, the degree of changes depends on residual portal blood flow, and exact localization of the shunt and may vary between liver lobes. Therefore, the biopsy sampling site may influence histologic findings. Nevertheless, the histologic changes seen in this foal were clearly indicative of abnormal portal blood flow, consistent with CPSS.

In the past, ultrasonography along with intraoperative mesenteric portography had been considered the standard for diagnosis of CPSS in veterinary medicine.[13, 15] In foals, CPSS has been confirmed by nuclear hepatic scintigraphy,[10, 12] operative mesenteric portography,[9-11] or postmortem examination.[8, 11] In human medicine, portovenography has been replaced by CTA in patients with spontaneous portosystemic shunts, which is equally accurate and allows 3-dimensional reconstruction and exact delineation of the shunt vessel.[16] CTA also has been shown to be a noninvasive and accurate method to diagnose and locate PSS in dogs.[17] Other commonly used techniques to detect CPSS in small animals are per rectal, transabdominal, or transsplenic portal scintigraphy[15, 18, 19] as well as contrast-enhanced magnetic resonance angiography.[15] In 1 previously reported foal with CPSS, the diagnosis could be confirmed by transsplenic portal scintigraphy.[12] In the present case, CTA was chosen as advanced imaging technique, because no shunt vessel could be identified by transcutaneous ultrasonography. However, because of difficulties in producing apnea long enough to complete the abdominal scan, breathing motion artifacts limited the diagnostic value of the CTA images. Therefore, ultrasound-guided percutaneous transsplenic injection of agitated saline with simultaneous echocardiography was performed in addition to CTA to confirm the presence of a CPSS. This novel diagnostic technique recently has been reported for the ultrasonographic diagnosis of PSS in dogs.[3, 4] It is minimally invasive, inexpensive, safe, and easy to perform, provided that standard ultrasound equipment is available and the operators are sufficiently experienced to perform the percutaneous transsplenic injection and simultaneous ultrasonography. Although this procedure was performed immediately after CTA, with the foal still under general anesthesia, it does not necessarily require general anesthesia and can be performed on a sedated foal in standing or recumbent position. In this case, appearance of microbubbles was only monitored in the right heart and before shunt occlusion, and no attempts were made to determine shunt location or to assess shunt closure after surgery. Conversely, Gomez-Ochoa et al were able to identify microbubbles in the portal vein, the shunt vessel, hepatic veins, the caudal vena cava, and the right heart.[4] This allowed them to differentiate between extrahepatic portocaval shunt, intrahepatic portocaval shunt, and portoazygous shunt, respectively, using 3 different acoustic windows and transducer positions during saline injection in dogs with PSS. Furthermore, they were able to document absence of microbubbles in the caudal vena cava and the right atrium in 7 dogs reevaluated after surgical correction. Accuracy, sensitivity, and specificity are currently unknown, both in dogs and foals. Also, we do not know whether microbubbles are completely retained by the sinusoidal barrier in healthy foals, as they are in healthy dogs.[3] However, appearance of marked echo contrast in the right heart immediately after transsplenic injection was certainly highly suggestive of CPSS in this foal. Nevertheless, additional studies should be conducted on healthy foals and foals with CPSS to establish this technique and to determine its diagnostic performance in this species.

Sole medical treatment of PSS with dietary changes, lactulose, and antibiotics has been associated with a high mortality rate in small animals,[14] although the results of a recent study in dogs with CPSS suggest that medical management can be associated with a reasonable long term outcome.[20] Medical treatment of CPSS in 2 foals described in the literature resulted in euthanasia because of recurrence of neurologic signs.[8, 10] In addition to standard treatment of hyperammonemia, sarmazenil, a partial inverse benzodiazepine receptor agonist, has been described to result in a substantial improvement of chronic HE in a canine model.[2] In the present case, sarmazenil treatment subjectively alleviated neurologic signs, such as compulsive circling. However, its efficacy remains unknown in horses.

Definitive treatment of PSS requires anatomic correction and occlusion of the shunt vessel.[14] Acute, complete surgical ligation of the shunt vessel can result in life-threatening PH, which is considered the most severe complication of surgical correction of CPSS.[5, 21] Conversely, partial attenuation of the shunt might prevent PH, but often results in persistence of clinical signs.[21] Therefore, gradual shunt occlusion techniques, including surgical cellophane banding[22, 23] and the use of ameroid constrictors,[24] have become the treatment of choice for PSS correction in small animals, decreasing the risk of potentially fatal PH. More recently, minimally invasive techniques, including coil embolization[25] and the use of percutaneous occluder devices,[5] also have been described as successful permanent treatments of PSS in dogs. Coil embolization leads to progressive closure of the shunt over 1–2 months, but bears the risk of migration of coils to heart and lungs.[25] If sized correctly, percutaneous occluder devices have less potential to migrate, but may lead to complete acute shunt occlusion with risk of PH, although the majority of dogs in a recent study could be acutely corrected without developing severe PH.[5]

So far, there have only been 2 successful corrections of extrahepatic CPSS in foals, 1 with transvenous coil embolization[12] and 1 with surgical ligation by paracostal laparotomy.[9] In another foal, surgical ligation of an intrahepatic CPSS failed after 16 days and an attempt to religate the shunt lead to uncontrollable hemorrhage.[9] Other reasons for low success in the treatment of CPSS in foals were inability to identify or to access the shunt intraoperatively.[11]

Cellophane banding frequently is used in small animals to occlude CPSS[23] and currently is considered the method of choice at our institution. During surgery, the cellophane band is fitted loosely around the shunt vessel, without occluding blood flow. Over the following weeks or months (depending on shunt size, initial degree of vessel attenuation, cellophane band width, and the patient's natural inflammatory response), the cellophane band gradually attenuates the shunt whereas avoiding life-threatening PH.[22] In the present case, cellophane banding was chosen to avoid acute complete occlusion of the shunt vessel that could have led to fatal complications. Identification of the shunt vessel and placement of the cellophane band were complicated by the fact that the shunt was located intrahepatically. Successful banding only was possible by close collaboration between the imaging radiologist and the surgeon exposing the shunt vessel. Portal or central venous pressures were not monitored during surgery, because the shunt was not occluded immediately and the risk for acute PH during surgery was considered small. Nevertheless, the slight, transient increase in heart rate and drop in systemic blood pressures immediately after cellophane placement were suggestive of development of mild acute PH during surgery. Furthermore, the marked decrease in ammonia and bile acid concentrations within 2 days after surgery suggested relatively rapid shunt occlusion. However, clinical recovery was uneventful and during the days after surgery, the foal did not show any signs of PH such as pain, abdominal distension, diarrhea, ascites, gastrointestinal edema, or cardiovascular compromise.

In conclusion, this is the 1st report of the diagnosis of an intrahepatic CPSS in a foal using CT angiography and ultrasound-guided percutaneous transsplenic injection of agitated saline. This case demonstrates that surgical cellophane banding is feasible in foals with intrahepatic CPSS and may lead to a complete resolution of clinical signs soon after surgery. Additional cases are necessary to gain more experience with the use of these novel diagnostic and therapeutic techniques in foals with CPSS.


  1. 1

    Rudolac, Streuli Pharma AG, Uznach, Switzerland

  2. 2

    Metronidazol, Haenseler AG, Herisau, Switzerland

  3. 3

    Gastrozol Daily, Axon Animal Health, Belrose NSW, Australia

  4. 4

    Penicillin Natrium Streuli, Streuli Pharma AG

  5. 5

    Vetagent, Veterinaria AG, Pfaeffikon, Switzerland

  6. 6

    Fluniximin, Dr. E. Graeub AG, Bern, Switzerland

  7. 7

    Ringer-Lactat-Loesung, Fresenius Kabi (Schweiz) AG, Stans, Switzerland

  8. 8

    Equi-Biserol, Laboratorium Dr. G. Bichsel AG, Interlaken Switzerland

  9. 9

    Sarmasol, Dr. E. Graeub AG

  10. 10

    Bard Biopty-Cut, 18 gauge, Bard Limited, Crawley, UK

  11. 11

    Telebrix, 350 mg I/mL, Guerbert, Zurich, Switzerland

  12. 12

    Somatom Sensation Open, Siemens AG, Medical Solutions, Erlangen, Germany

  13. 13

    Terumo Neolus, Terumo Europe N.V., Leuven, Belgium

  14. 14

    Dorbene, Dr. E. Graeub AG

  15. 15

    Propofol 1% Fresenius, Fresenius Kabi (Schweiz) AG, Stans, Switzerland

  16. 16

    Isofluran Baxter, Baxter AG, Volketswil, Switzerland

  17. 17

    Sintenyl i.v., Sintetica SA Pharmaceuticals, Mendrisio, Switzerland

  18. 18

    Logiq e with 8C-RS 4–10 MHz microconvex transducer, GE Medical Systems, Glattbrugg, Switzerland

  19. 19

    Weck Horizon Ligation System, Teleflex Medical AG, Belp, Switzerland

  20. 20

    Polysorb 2, Ethicon, Johnson & Johnson Medical GmbH, Norderstedt, Germany

  21. 21

    Biosyn 2/0, Syneture, Covidien, Norwalk, CT

  22. 22

    Appose UCL, Covidien, Norwalk, CT

  23. 23

    Morphin HCI Sintetica 0.1%, Sintetica SA Pharmaceuticals, Mendrisio, Switzerland

  24. 24

    FFP, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland

  25. 25

    Rota TS Oraldoser, Vétoquinol, Ittingen, Switzerland