Anatomical classification of canine congenital extrahepatic portosystemic shunts based on CT angiography: A SVSTS and VIRIES multi-institutional study in 1082 dogs.

Abstract

Canine congenital extrahepatic portosystemic shunt (EHPSS) morphologies have not been fully elucidated. The goal of this retrospective, multi-institutional study was to use CT angiography to create an anatomical-based nomenclature system for canine congenital EHPSS. These shunt morphologies were then evaluated to identify any significant association with patient age, sex, breed, weight, or subjective portal perfusion score. Data collected respectively from the SVSTS and VIRIES list-serves included patient DOB, sex, breed, weight, CT date, and reported diagnosis. A single author (C.W.) viewed all CT scans and classified shunts based on the shunt portal vessel(s) of origin, the shunt systemic vessel(s) of insertion, and any substantial portal vessels contributing to the shunt. Additionally, hepatic portal perfusion was subjectively scored between one (poor/none) and five (good/normal) based on the caliber of the intrahepatic portal veins. A total of 1182 CT scans were submitted from 13 different institutions. Due to exclusion criteria, 100 (8.5%) were removed, leaving 1082 CT scans to be included. Forty-five different EHPSS anatomies were identified with five classifications accounting for 85% of all shunts (left gastric-phrenic [27%], left gastric-azygos [19%], left gastric-caval [15%], aberrant left gastric-caval with right gastric vein [12%], and aberrant left gastric-caval with right gastric vein and short gastric vein [11%]). Shunt origin involved the left gastric vein in 95% of the described classifications. Significant differences were identified among the five most common shunt types with respect to age at the time of the CT scan (P<.001), sex (P=.009), breed (P<.001), weight (P<.001), and subjective portal perfusion score (P<.001). An anatomical classification system for canine EHPSS may enable improved understanding, treatment comparisons, and outcome prediction for these patients.