Congenital portosystemic shunts in Maltese and Australian Cattle Dogs

Congenital portosystemic shunts in Maltese and Australian Cattle Dogs PLC TISDALL*, GB HUNT*', CR BELLENGERt and R MALIKt Veterinary Cardiovascular Unit, University of Sydney, New South Wales 2006 SUMMARY Congenital portosystemic shunts were definitively diagnosed in 62 dogs over a period of 15 years. Maltese and Australian Cattle Dogs were significantly overrepresented, accounting for 14 and 13 cases, respectively. Maltese invariably had a single extrahepatic shunt derived from the left gastric or gastrosplenic vein, whereas Cattle Dogs usually had large intrahepatic shunts involving the right liver lobes. The clinical syndromes resulting from anomalous portosystemic communications were indistinguishable in the 2 breeds. Fasting blood ammonia concentration was elevated in 20 of 22 dogs tested, providing a minimally invasive and effective means of diagnosis. Complete or partial shunt attenuation was performed successfully in all 9 Maltese and in 2 of 6 Cattle Dogs in which it was attempted. Aust Vet j 71:174 - 178

Introduction Congenital portosystemic shunts (CPS) are anomalous venous communications that allow portal blood rich in absorbed nutrients, toxins and hormones to avoid first-pass hepatic metabolism, so that these constituents then enter the systemic circulation in high concentrations. Shunts are typically single vessels and may be either intrahepaticor extrahepatic (Payneet a1 1990).Generally, intrahepatic shunts occur in large breeds of dogs and extrahepatic shunts in small and toy breeds (Rothuizen et af 1982; Breznock and Whiting 1985). Resulting syndromes of hepatic dysfunction, encephalopathy and urate urolithiasis are similar in both cases, and have been well described (Ewing et af 1974; Rothuizen et a1 1982; for a review see Martin 1993).Disease associated with CPS is generally progressive. The liver is deprived of trophic factors in portal venous blood and atrophiesprogressively (Starzl etaf 1973,1976).Surgery to partially or completely attenuate the shunting vessel and thereby redirect portal blood flow to the liver is therefore the treatment of choice (Johnson et af 1987). Shunt attenuation usually ameliorates neurological signs immediately, while liver function generally improves but may remain abnormal (Lawrence et a1 1992). A previous analysis of cases from the Sydney metropolitan area (Maddison 1988) showed Maltese and Australian Cattle Dogs (ACD) were over-represented among diagnosed cases of CPS. More recent accessions suggest that this is a continuing trend. This report compares and contrasts the clinical and anatomical features of CPS in these 2 breeds.

Materials and Methods Diagnosis of Congenital Portosystemic Shunts Sixty-nine dogs with CPS were examined at Sydney University Veterinary Teaching Hospital between October 1978 and August 1993. Diagnosis was based on historical, clinical and biochemical criteriaonly in 7 dogs (including2 ACD), while in 62 dogs there also was confirmation by visualisation of the shunt at surgery, necropsy or with angiography.Only animals in which the shunt and its location

* t

Department of Veterinary Anatomy Department of Veterinary Clinical Sciences 2 Author for correspondence 0 OmnipaqueEO, 350 mg.mL-' iodide, Winthrop Laboratories, Ermington, NSW 2 115

174

were confirmed visually or radiographically were included in the subsequent analysis (14 Maltese, 13 ACD). Eleven of these dogs were included in a previous report (Maddison 1988).Their files were reviewed, paying particular attention to clinical and laboratory findings, radiographic images, case management and outcome. Where possible, precise anatomical features of shunting vessels were noted. Shunts were classified as intrahepatic or extrahepatic according to their location at surgery, necropsy or angiography. Fasting ammonia concentration was determined in 23 dogs and in 13 of these ammoniatolerance was measured 20 to 40 min after rectal or oral administrationof 100rng.kg-' of ammoniumchloridesolution (50 mg.mL-'; Rothuizen and Van den Ingh 1982). Liver and kidney size was assessed radiographically in 15 dogs. Kidneys were considered enlarged if their length was more than 3.5 times that of the second lumbar vertebra (Lee and Leowijuk 1982). Liver size was assessed subjectively according to the position of the pyloric gas shadow and the gastric axis in relation to the ribs and spine. Portal venograms were obtained in 11 dogs, either by catheterisation of splenic or jejunal veins at laparotomy (8 cases), or by catheterisation of the cranial mesenteric artery (3 dogs) under image intensificationvia the femoral artery using a right-angled tip catheter (6 French). Contrast agents used varied between cases, but typically 0.5 to 1 mL.kg-' of iohexol' was injected rapidly by hand. Images were recorded on a rapid sequence camera or a single radiographic exposure as the injection was completed. Shunts were considered intrahepatic if they were cranial to the body of the thirteenth thoracic vertebra and extrahepatic if caudal to this level (Birchard et a1 1989).

Surgical Approach f o r Shunt Attenuation Surgical shunt attenuation was attempted in 9 Maltese and 6 ACD. A cranial midline laparotomy was performed in all cases, and in 5 ACD the incision was extended cranially via a partial median sternotomy to adequately expose the liver. In all cases the portal vein and caudal vena cava (CVC) were inspected to identify abnormal extrahepatic tributaries. Intrahepatic shunts were identified by catheterisation, palpation or portal venotomy to expose the shunt orifice. Silk ligatures (3/0 to 0) were placed and tightened gradually. The degree of attenuation was limited by rises in portal pressure (no more than 10 cm H2O; Birchard 1984) measured with a water manometer via a catheter placed in a splenic or jejunal vein. Changes in systolic arterial blood pressure (all dogs) and central venous pressure (10 dogs) were monitored (Swalec and Smeak 1990; Auriralian Veterinary Journal Vol. 71, No. 6 , June 1994

Komtebedde et a1 1991),and the intestines observed for cyanosis or hypermotility (Mathews and Gofton 1987)indicating development of splanchnic congestion. Complete attenuation was possible in one Maltese and one ACD, shunts being partially attenuated in the remaining dogs.

Post-surgical Follow-up of Cases Information was obtained for 7 of the 9 Maltese and the 2 ACD that survived corrective surgery. Reports were obtained by telephone or personal interview between 3 months and 6 years after surgery.

Statistical Analysis Records of all dogs (n = 14828)admitted since February 1988were examined to estimate breed and sex distributions during the study period. Most cases of CPS were admitted during this period (15 of 27;56%), from which computerised data were available. Expected and actual numbers of cases were compared, using Fisher's exact test. Age at presentation and results of biochemical analyses (alkaline phosphatase [ALP], alanine aminotransferase [ALT], urea, glucose, albumin, protein and ammonia) for Maltese and ACD with CPS were compared, using the Mann-Whitney U test. P values less than 0.01 were considered significant. All values in the text are given as the mean k SD.

Results Maltese and ACD were significantly over-represented among cases of CPS. Maltese comprised 23% of cases (14of 62dogs) compared with an expected frequency of 2% (P < 0.00001).Similarly, the observed frequency for ACD (21%; 13 of 62cases) was greater than expected (1%; P c 0.ooOOl).

Age at Diagnosis Age at diagnosis was similar in Maltese (range 3 months to 8.5 years, median 9 months) and ACD (range 5 to 36 months, median 12 months). In both breeds the majority (18of 27 dogs, 67%) were diagnosed at or before 12 months of age, most being detected between 6 and 12 months (13 of 17 dogs). In 5 Maltese and 4 ACD the condition was first diagnosed when they were older than one year. Sex There was a preponderance of males among affected dogs (10 of 14Maltese and 10of 13 ACD), although this failed to reach statistical significance (P = 0.26 and 0.15,respectively). The sex distribution in all dogs presenting at the hospital during this period was 54.7% male and 45.3% female.

Biochemical variable" ~

_

_

_

_

_

_

151 f 101

13

ALP (UL')

209 f 149 3.3f 2 4.5f 2 160 f 96 492 f 178 51 f 6 30f6

Urea (mmol.L-') (rnmo1.C')

Fasting ammonia (pmo1.L') AT (prno1.C') TPP (g.L-') Albumin (9.C')

* t

Dogs of both breeds presented similarly, although signs in individual cases varied considerably; vague neurological and gastrointestinal signs (depression, inappetence, vomiting and diarrhoea) were most common. One Maltese and 2 ACD presented with dysuria, stranguria or haematuria associated with urate urolithiasis, but were considered normal in all other respects by their owners. There is a well-described association between CPS and urate calculi (Maretta eta1 1981;Brain 1988).One Maltese, an 8.5-year-old breeding bitch that had successfully reared several litters, was asymptomatic and was detected by biochemical screening. The most consistently recorded historical finding was episodic depression. Behavioural abnormalities included pacing, tail chasing, circling and head pressing. Some dogs exhibited inappropriate aggression, vagueness and mood changes. Less common neurological signs included torticollis, nystagmus, opisthotonus, muscle tremor and blindness. Behavioural and neurological signs were typically interspersed with periods of normality. Generalised motor seizures were observed in 3 Maltese, 2 of which were hypoglycaemic (0.7mmol.L-' and 2.6mmol.L-'). The remaining dog had a blood glucose concentration of 5.0 mmol.L-'. An unsubstantiated report of seizure-like activity was given by one ACD owner. Polyuria or polydipsia was reported in 9 dogs, but excessive water intake ( > 100 m~.kg-'.ciay") was confirmedin only one. Prolonged recovery from barbiturate anaesthesia was reported in one Maltese, and behavioural alteration in the post-anaesthetic period in another. In 3 ACD the onset of clinical signs was related to feeding. Ascites referable to hypoproteinaemia (25 gL") was encountered before therapy in one ACD, while hypoalbuminaemia (range 6 to 19 g.L") and ascites developed in 3 other dogs after dietary protein restriction.

Biochemical Abnormalities Biochemical changes were similar to those previously described in dogs with CPS (Griffths et a1 1981;Table l), and were not significantly different in the two breeds studied. No single biochemical test was able to detect all dogs with CPS. Frequent findings included mild to moderate increases in ALP and ALT activities, and decreases in urea, cholesterol and albumin concentrations. The most consistent finding was hyperammonaemia. Fasting ammonia concentrations were increased in 20 of 23 dogs tested, while an ammonia tolerance test demonstrated decreased ammonia clearance in the remaining cases (1 ACD, 2 Maltese). Fasting bile acid concentrations were determined in only 5 dogs (3 ACD, 2 Maltese) and were invariably increased (range, 34 to 437 pmol.L-'). Hypoglycaemia ( < 3.3 mmo1.L-') was detected at least once in 4 of the 10 Maltese in which the blood glucose was measured. In contrast, all 7 ACD tested were normogly caemic.

TABLE 1 Plasma biochemicalmeasurements in Maltese and Australian Cattle Dogs (mean f SD) Maltese n Cattle Dogs n P value Reference ranaef _ ~

ALT (UC')

GIUCOS~

Presenting Complaint

12

0.82

12 10 7

0.15 0.40 0.65 0.82 0.13

13 12 10 13 8 11

153 f 94 135 f 84 3f2 5fl 171 It 134 364 f 133 45f 14

10 5 10

10

20f8

9

0.50 0.02