Budd-Chiari-like Syndrome in a Dog Due to Liver Lobe Entrapment Within the Falciform Ligament

Introduction

Budd-Chiari syndrome includes a group of disorders in which hepatic venous outflow obstruction is at the level of the hepatic venules, large hepatic veins, caudal vena cava, or right atrium. The term “Budd-Chiari syndrome” was originally used in human medicine to describe obliterating endophlebitis of the terminal hepatic veins, resulting in veno-occlusive disease.¹ Causes in humans are typically associated with states of hypercoaguability. ¹ Inherited deficiencies of protein C, protein S, and antithrombin III, as well as acquired myeloproliferative disorders, paroxysmal nocturnal hemoglobinuria, antiphospholipid syndrome, pregnancy, and oral contraceptive use have all been reported as causing Budd-Chiari syndrome in humans.¹ Less common causes include tumor invasion of the inferior vena cava, systemic aspergillosis, caudal vena caval webs, trauma, inflammatory bowel disease, dacarbazine therapy, and idiopathic causes.¹

The term “Budd-Chiari-like syndrome” (BCLS) has been proposed for use in veterinary medicine to refer specifically to mechanical obstruction causing hepatic venous outflow obstruction between the hepatic sinusoids and the heart, including the caudal right atrium that functions as an extension of the caudal vena cava.² This definition excludes disorders such as right heart failure, pericardial effusion, and restrictive pericarditis. ² Disorders causing BCLS result in increased lymph production and leakage through the fenestrated sinusoids into the space of Disse.² This lymph is high in protein, and when it escapes the space of Disse through the hepatic capsule, abdominal effusion results. Dogs with such a postsinusoidal obstruction will have hepatomegaly with a high protein and low cellular ascites consistent with a modified transudate.² Causes reported in dogs include dirofilariasis, cor triatrium dexter, intracardiac tumors, extra- and intraluminal neoplasia of the caudal vena cava, kinking of the caudal vena cava, caudal vena caval fibrosis secondary to a needle foreign body, congenital diaphragmatic hernia, and intrahepatic postsinusoidal obstruction.^2–14 Reports of BCLS in cats are few; one case involves membranous obstruction of the caudal vena cava, and another case involves histopathological changes consistent with veno-occlusive disease as described in humans.^15,16

The World Small Animal Veterinary Association (WSAVA) liver standardization group states that the term BCLS is misleading, because veno-occlusive disease as described in humans has not been described in veterinary medicine.¹⁷ Budd-Chiari syndrome in human medicine and BCLS in veterinary medicine have both been applied to a variety of different etiologies and not one particular disease or pathogenesis. For this reason, the author has chosen to use this term, despite recommendations to the contrary from the WSAVA. To the author’s knowledge, a suitable alternative to classifying these diseases has not been developed.

This report describes the diagnosis and management of a dog with BCLS due to liver lobe entrapment in the falciform ligament.

Case Report

A 13-year-old, spayed female golden retriever was referred for evaluation of acute collapse and ascites. According to the owner, on the morning of day 1, the dog had been normal until immediately following a walk, when the dog vomited twice and collapsed. The dog was presented to the referring veterinarian in lateral recumbency with a heart rate of 200 beats per minute (bpm) and abdominal distension.

Abdominocentesis was performed, and an in-house evaluation revealed a serosanguineous fluid with total solids of 4 g/dL. An electrocardiogram (ECG) revealed sinus tachycardia. Systolic blood pressure was 89 mm Hg. A complete blood count (CBC) and serum biochemical profile submitted to a human laboratory revealed azotemia (blood urea nitrogen [BUN] 33 mg/dL [reference range 6 to 26 mg/dL]; creatinine 2.1 mg/dL [reference range 0.4 to 1.5 mg/dL]), hypoalbuminemia (albumin 2.5 g/dL; reference range 3.5 to 5.0 g/dL), and a mature neutrophilic leukocytosis (white blood cells 26.9 × 10³/μL; reference range 4.30 to 10.0 × 10³/μL). A urinalysis revealed isosthenuria (specific gravity 1.015) and 3+ proteinuria. A heartworm antigen test was negative. Thoracic radiographs were taken but were not available for review.

Therapy for suspected right-sided heart failure was begun, which included administration of digoxin^a (0.007 μg/kg per os [PO] q 12 hours) and furosemide^b (3 mg/kg PO q 12 hours). Therapy for possible bacterial endocarditis was also added (cephalexin^c 15 mg/kg PO q 12 hours; enrofloxacin^d 3 mg/kg PO q 12 hours). The dog’s condition continued to deteriorate, and referral for further evaluation occurred on day 2 of illness.

On presentation to the referral hospital, the dog was recumbent, had weak pulses, prolonged capillary refill time (3 seconds), increased respiratory rate (60 breaths per minute), and a heart rate of 180 bpm. Systolic blood pressure was 70 mm Hg, and an ECG revealed ventricular tachycardia. Differential diagnoses at presentation included hypoxemia with ischemia, hypovolemia, pain, systemic inflammation resulting in disseminated intravascular coagulation, cardiac contusion, electrolyte abnormalities, and drug reaction toxicity.

The dog was initially treated with lidocaine because of the tentative diagnosis of cardiac disease, concerns that the ventricular arrhythmia was the cause of the weakness, and the presence of hypotension and poor perfusion. An intravenous catheter was placed, and a bolus of lidocaine^e (2 mg/kg intravenously [IV]) was administered with no improvement in the ventricular tachycardia. After two additional boluses of lidocaine (2 mg/kg IV), the ventricular arrhythmia slowed to 165 bpm. A constant-rate infusion (CRI) of lidocaine was started (50 μg/kg per minute IV).

Within an hour of presentation, thoracic radiographs were taken that revealed microcardia consistent with hypovolemia. Clinical signs and parameters on presentation supported hypovolemic shock. Abdominal radiographs revealed mild spondylosis deformans at the first and second lumbar vertebrae and no evidence of abdominal effusion. A bolus of hetastarch 6%^f (4 mL/kg IV once) was administered, followed by a bolus of balanced crystalloid solution^g (22 mL/kg IV once).

A CBC and serum biochemical analysis revealed a mildly increased hematocrit (54.3%; reference range 35% to 54%) consistent with dehydration, and mild leukocytosis (15.7 × 10³/μL; reference range 8.0 to 14.5 × 10³/μL). Thrombocytopenia was also present (68 × 10³/μL; reference range 220 to 600 × 10³/μL). Differentials for the thrombocytopenia included increased consumption, destruction, sequestration, or decreased production. Prothrombin time, partial thromboplastin time, and plasma fibrinogen concentration were assessed and found to be within normal limits. Serum biochemical analysis revealed elevations in aspartate transaminase (129 U/L; reference range 0 to 50 U/L), alanine aminotransferase (451 U/L; reference range 0 to 60 U/L), and alkaline phosphatase (126 U/L; reference range 0 to 100 U/L). These abnormalities were consistent with hepatocellular damage, possibly secondary to hypovolemia and ischemic injury.

The previously noted azotemia had worsened (BUN 57 mg/dL [reference range 8 to 22 mg/dL]; creatinine 2.8 mg/dL [reference range 0.5 to 1.7 mg/dL]), and the potassium and chloride were found to be low (potassium 3.4 mmol/L [reference range 3.8 to 5.5 mmol/L]; chloride 101 mmol/L [reference range 107 to 115 mmol/L]). Furosemide administration was suspected to be the cause of hypokalemia and hypochloremia; however, additional loss of chloride secondary to vomiting may also have been a contributing factor. Replacement IV fluids were given at 120 mL per hour, with 40 mEq KCl/L to replenish body potassium levels. Serum albumin concentration was normal (3 g/dL; reference range 2.5 to 4.2 g/dL). Resolution of the previously noted hypoalbuminemia may have been due to false elevation of the albumin secondary to dehydration. The initial albumin level also may have been inaccurate, because that value was obtained at a human laboratory lacking canine controls.

Following stabilization with fluid therapy, on the morning of day 2, an echocardiogram was performed that revealed no abnormalities of the heart or intrathoracic vena cava. The dog was maintained in the intensive care unit with continuous ECG monitoring and was treated with a lidocaine CRI (increased to 75 μg/kg per minute IV), hetastarch (20 mL/kg per day IV), and crystalloid solution (85 mL/kg per day IV). As the dog’s hypovolemia resolved, progressive abdominal distension and a palpable fluid wave were noted. The ventricular arrhythmia resolved following 6 hours of fluid therapy, and the lidocaine CRI was slowly discontinued over 24 hours with no recurrence of the arrhythmia. The arrhythmia was suspected to be secondary to a combination of digoxin administration, hypovolemia, and decreased myocardial perfusion; therefore, additional antiarrhythmic drugs were not administered.¹⁸

An abdominal ultrasound was performed the afternoon of day 2 and revealed a moderate amount of free-fluid hypoechogenicity of the pancreas, hyperechoic renal cortices, and an area of hyperechoic omentum adjacent to the left lateral lobe of the liver. No evidence of caudal vena caval thrombosis was detected. Fluid obtained by abdominocentesis during the ultrasound was evaluated to be characteristic of a modified transudate (nucleated cells 800/μL; protein 4.3 g/dL).

The dog continued to improve clinically on medical management; however, the abdominal effusion and azotemia persisted. Based on the absence of heart disease and the high-protein ascites, a post-sinusoidal venous obstruction causing a BCLS was suspected. An abdominal exploratory was recommended to further evaluate noncardiac causes of the abdominal effusion.

On day 7 (5 days after admission to the referral hospital), surgery was performed. Initial approach into the abdomen immediately revealed that the distal portion of the left lateral liver lobe was strangulated by a piece of the falciform ligament. Entrapment and compression of the hepatic veins supplying the left lateral liver lobe could result in the highprotein abdominal effusion and BCLS. Other abnormalities included an 8-mm diameter nodule in the caudate liver lobe and an irregular splenic surface. The kidneys appeared grossly normal. Submitted for biopsy were samples from the strangulated left lateral liver lobe, nodule from the caudate liver lobe, left limb of the pancreas, spleen, and cranial pole of the left kidney. A liver sample also was submitted for culture.

Histopathology of the strangulated liver lobe revealed periportal and bridging fibrosis, marked vacuolar degeneration, and macrophage aggregates with brown pigmentation consistent with bile. The hepatic capsule was markedly edematous, and dilatation of lymphatics was marked. The nodule in the caudate liver lobe contained minimal portal fibrosis and diffuse vacuolar change. Evidence of fibrosis in both the strangulated section and the nodule, and the diffuse vacuolar change, could suggest more chronic and diffuse hepatic disease. The liver culture was negative. Biopsy results of the pancreas were consistent with mild interstitial pancreatitis, and those of the spleen were consistent with a splenic hematoma. Histopathological changes in the left kidney were significant and included multifocal renal tubular epithelial necrosis and sloughing of cells into tubular lumens. Multifocal areas within the renal medulla were characterized by loss of tubules and interstitial fibrosis. These changes support acute on chronic renal injury secondary to the hypovolemia, and they may explain the progressive azotemia despite aggressive fluid therapy.

The dog recovered well following surgery, and the abdominal effusion resolved. On day 9, the dog was discharged. Reassessment 2 weeks postsurgery revealed no abdominal effusion on ultrasound examination, but azotemia was persistent (BUN 65 mg/dL, reference range 8 to 22 mg/dL; creatinine 3.2 mg/dL, reference range 0.5 to 1.7 mg/dL).

Discussion

The cause of the liver lobe entrapment in this case is unknown. The dog had no known history of trauma prior to the collapse. One possible cause could be related to the anatomy of the left lateral liver lobe. Liver lobe torsions are considered rare in all reported species. However, of the documented cases, nearly 50% of torsions occur in the left lateral liver lobe.^19,20 Reasons this lobe is more commonly affected include its increased mobility, large size, and separation from other lobes.¹⁹ Most affected dogs are medium to large breeds. One report indicated that three of 13 affected dogs were golden retrievers; however, because of the small numbers, a true breed predisposition could not be assessed.²⁰ Similar anatomical factors may have contributed to the left lateral liver lobe becoming entrapped in the falciform ligament. No evidence of any abnormality was noted that could have led to increased mobility of the liver lobe or falciform ligament. Such abnormalities can include congenital or traumatic loss of triangular or coronary ligaments of the liver lobe or loss of the ligament attachments of the falciform.

While BCLS is considered a rare clinical entity, it should be considered as a differential diagnosis in a dog with abdominal effusion characterized as a modified transudate, after heart and pericardial diseases have been ruled out. Right heart failure is a more common cause of such effusion and may explain why the dog of this report was started on therapy for heart disease prior to referral. In human medicine, patients with Budd-Chiari syndrome are commonly suspected of having cholecystitis because of the presence of abdominal pain and thickening of the gallbladder on ultrasound. ¹ These patients often undergo cholecystectomy prior to the appropriate diagnosis of Budd-Chiari syndrome.¹

Initial diagnostics in human patients include Doppler ultrasonography, contrast-enhanced computed tomography (CT), and, less often, magnetic resonance imaging.¹ A “spiderweb” pattern following hepatic venography confirms the diagnosis.¹ Liver biopsy can be supportive of, but is not considered necessary to confirm the diagnosis.¹ In small animals, diagnostic tools have varied depending upon the location of the post-sinusoidal obstruction. Echocardiography, cardiac catheterization and angiography or venography, ultrasonography, nuclear scintigraphy, hepatic biopsy, measurement of wedged hepatic vein pressure, and necropsy have been used in making a diagnosis.^3–16 Contrast-enhanced CT has not been reported in the evaluation of BCLS in animals, but it might have been useful in the current case. Doppler blood-flow assessment of the left lateral liver lobe during the initial ultrasound examination might have revealed decreased blood flow, suggesting torsion or entrapment. The left liver lobe appeared normal, with the exception of the area thought to be adjacent hyperechoic omentum. This was correctly identified as the falciform ligament at surgery.

Treatment and prognosis for animals with BCLS vary upon whether the underlying cause can be identified and corrected. Surgical removal of the entrapped liver lobe in this case resolved the BCLS. The dog’s long-term prognosis was considered guarded, however, because of significant renal disease. The initial blood work and isosthenuria supported preexisting renal disease that was likely worsened by the dog’s hypovolemia.

Conclusion

To the author’s knowledge, this is the first case reported of BCLS due to liver lobe entrapment in the falciform ligament. This case also emphasizes the importance of ruling out more common causes of modified transudate abdominal effusion, such as right heart failure or pericardial disease, prior to commencing therapy.