Experience with the Rex shunt (mesenterico-left portal bypass) in children with extrahepatic portal hypertension

Abstract

Extrahepatic portal vein thrombosis (EPVT) in children can lead to severe bleeding from gastrointestinal varices, ascites, thrombocytopenia from hypersplenism, and other coagulation disorders. The authors have used the superior mesenteric vein to intrahepatic left portal vein (Rex) shunt in 5 children with symptomatic EPVT and report their results with this novel technique. Children with symptomatic portal hypertension were screened for the underlying cause. All children with essentially normal livers and obstruction of the extrahepatic portal vein were considered for the Rex shunt. Evaluation included liver function tests, liver biopsy, and radiological evaluation of the intrahepatic vascular anatomy. Five patients between the ages of 2.8 and 10.5 years underwent evaluation for portal hypertension secondary to extrahepatic portal vein obstruction. Three patients had idiopathic extra hepatic portal vein thrombosis with cavernous transformation, 1 had thrombosis after a living-related liver transplant, and 1 had compression and obstruction of the main portal vein from enlarged lymph nodes after treatment of systemic histoplasmosis. All patients were symptomatic. Three patients had intermittent bleeding from esophageal and gastric varices, and all 5 had relative degrees of hypersplenism with enlarged spleens and thrombocytopenia (11,000 to 77,000). Three patients had significant leukopenia. Results of imaging studies suggested that 3 patients had inadequate intrahepatic portal veins for shunting, but all patients at exploration underwent successful shunting. There were no serious intraoperative complications. Postoperative complications included ascites in 2 patients that resolved within 1 month. There were no early shunt thromboses. The median postoperative length of stay was 7 days. Clinical follow-up ranged from 7 to 21 months. Gastrointestinal bleeding did not recur in any patient, and ascites resolved in all. Spleen size decreased significantly (P < .01) from 9.4 +/- 4.0 cm to 5.0 +/- 3.7 cm below the left costal margin. Mean platelet count and white blood cell count rose after shunting from 79 +/- 42 to 176 +/- 73 (P < .02) and 5.4 +/- 2.3 to 7.5 +/- 3.9 (P = .06), respectively. All shunts were studied at 1 and 7 days, and 3 and 6 months after the procedure. Shunt patency was documented in all cases. Subsequently, shunt blockage occurred in 2 patients. The Rex shunt has proven to be an effective method of resolving portal hypertension caused by EPVT including thrombosis after living donor transplantation. This shunt is preferable to other surgical procedures because it eliminates portal hypertension and its sequelae by restoring normal portal flow to the liver.