LAUSR

OBJECTIVE Distinguishing retrohepatic end-to-side portocaval shunts (ES-PCS) and side-to-side portocaval shunts (SS-PCS) can be difficult, but it is essential for determining the treatment strategy. Our experience with retrohepatic PCS is analyzed. METHODS Since 2007, 9 children (5/9 ES-PCS and 4/9 SS-PCS) were surgically treated. Radiology studies included Doppler-ultrasound, CT/MRI and angiography/occlusion test (8/9). RESULTS CT in 5/5 ES-PCS revealed the portal vein (PV) entering the left side of the vena cava with a uniform shape. 4/4 SS-PCS showed aneurysmal PV containing the origin of the main intrahepatic portal branches (IHPB) entering the cava anterior aspect or slightly to the right with a variable length (from long to short/wide). ES-PCS anatomy showed caudate lobe absence with the fistula entering the left cava aspect free of parenchyma, but anterior through the caudate lobe in SS-PCS. With the angiography/occlusion test, the IHPB was undeveloped in ES-PCS (portal pressure > 38 mmHg) and hypoplasic in SS-PCS (portal pressure < 25 mmHg). ES-PCS treatment included: 1/5 hepatectomy and 4/5 definitive banding (one by laparoscopy); while in SS-PCS: 1/4 liver transplantation, 2/4 1-step closure (one by laparoscopy), and 1/4 definitive banding. CONCLUSION Fistula shape, cava relationship, IHPB and portal pressures differ between the two shunt types. A question arises regarding the need for secondary complete closure after banding. LEVEL OF EVIDENCE Level IV.