Intravascular US-guided direct intrahepatic portacaval shunt with a PTFE-covered stent-graft: Feasibility study in swine and initial clinical results

PURPOSE: To determine the feasibility of the creation of a direct intrahepa tic inferior vena cava (IVC)-to-portal-vein shunt with puncture guided by a transfemorally placed intravascular ultrasound (IVUS) probe and use of a p olytetrafluoroethylene (PTFE)-covered stent-graft. MATERIALS AND METHODS: In five swine, transjugular access was used to perfo rm a direct puncture from the IVC to the portal vein with use of a modified Rosch-Uchida Portal Access set directed with real-time IVUS (9 MHz) introd uced from a transfemoral venous approach. The direct intrahepatic portocava l shunt (DIPS) was then created with single or overlapping PTFE-covered Pal maz stents placed through a 10-F sheath and dilated to a diameter of 8 mm. Follow-up was performed with transhepatic portography at 2, 4, and 8 weeks. Animals were killed when shunts occluded or at the termination of the stud y at 8 weeks. Gross and microscopic histologic study was performed on sacri ficed animals. A similar technique was used to create DIPS in five patients with intractable ascites, with follow-up by US and venography. RESULTS: All experimental DIPS created in swine were created without compli cations. Portal vein punctures were achieved in four of five swine on the f irst or second pass of the needle. Follow-up transhepatic portography at 2 weeks demonstrated occlusion of two shunts, both explained by technical rea sons at sacrifice. At 4 and 8 weeks, the remaining three shunts were patent on portography. Histology showed a thin neointimal lining with no signific ant tissue ingrowth or hyperplasia. Clinically, in five patients, successfu l puncture of the portal vein from the IVC was achieved in one to three pas ses. Creation of DIPS led to a reduction of mean portosystemic gradient fro m 18-29 mm Hg (mean, 24 mm Hg) to 9-10 mm Hg (mean, 9 mm Hg). One patient d ied of liver failure 2 days after creation of DIPS. The other four patients were doing well 2-15 months (mean, 8 months) after the procedure, with pat ency confirmed by US and venography. CONCLUSION: Creation of DIPS is technically feasible, and the direct IVC-to -portal-vein puncture can be done accurately with real-time IVUS guidance. Further studies and longer follow-up are necessary to determine if the shor t length of the PTFE-covered stent-graft and avoidance of the hepatic vein will increase the long-term patency compared to standard transjugular intra hepatic portosystemic shunt creation.