The surgical attenuation of an intrahepatic portosystemic shunt in 45 dogs is described. Twenty-nine (64 per cent) had left divisional shunts consistent with a patent ductus venosus (PDV), 15 (33 per cent) had central divisional shunts and one had a right divisional shunt. In the dogs with a PDV, the shunt vessel could be most easily manipulated at a posthepatic site, whereas in those with central and right divisional shunts the manipulation could be more easily made intrahepatically but sometimes involved demanding intravascular surgical techniques. Eight dogs (18 per cent) died during the surgery or shortly afterwards. Of the 37 dogs surviving longer term, 28 (76 per cent) became clinically normal and required no medication or diet control. In a further three animals the shunt was ligated completely only during a second surgical procedure. The remaining six dogs were euthanased because of clinical signs of encephalopathy which were either surgically or medically uncontrollable.

Six dogs had signs of pelvic limb weakness, pain and collapse as a result of occlusion of the distal aorta and/or the iliac arteries by a thrombus. Antemortem diagnosis was made on the basis of clinical signs, angiography and ultrasonography. Five dogs had concurrent disease that probably predisposed to thrombosis, including hyperadrenocorticism (three dogs), neoplasia and cardiac disease. Two dogs died shortly after the episode of thrombosis. Dogs that survived the acute episode received aspirin in an attempt to prevent thrombosis occurring again and all regained pelvic limb function. For dogs that survived longer than one month after the acute episode, repeat thrombosis was uncommon; hence the prognosis was related to the underlying disease. Aortic and iliac thrombosis in dogs is an uncommon condition that usually arises secondarily to a predisposing disease process; it carries a more favourable prognosis than feline aortic thromboembolism.

Survival times in this study exceed those in other studies published previously. In part, this is explained by an improved remission duration following surgery compared with previous reports. More striking though is the longevity of patients following institution of medical therapy. These data give strong objective support to the role of medical therapy in the management of canine insulinoma, including following relapse after surgically induced remission.