The purpose of this study was to evaluate the feasibility, safety, and efficacy of portal vein recanalization (PVR) and propose a new classification for better selecting candidates with portal vein occlusion (PVO) in whom PVR could be feasible.

The charts of 15 non-cirrhotic patients in whom stent placement using a trans-hepatic approach was attempted for the treatment of PVO with cavernous transformation were reviewed. There were 12 men and 5 women with a mean age of 47±12 years (range: 22–60 years). Intrahepatic involvement was classified into 3 groups according to the intrahepatic extent of PVO: type 1 included occlusions limited to the origin of the main portal vein and/or the right or left portal branches, type 2 included type 1 plus extension to the origin of segmental branches, type 3 included type 2 plus extension to distal branches.

There were 6 patients with PVO type 1, 7 patients with PVO type 2, and 2 patients with PVO type 3. Indications for PVR were gastrointestinal bleeding (n=6), portal biliopathy (n=2), reduce portal pressure before surgery (n=4), or other (n=3). PVR was successful in 13 patients (87%) with no severe side effects. Failure of PVR or early stent thrombosis occurred in 100% of type 3 vs. 8% of type 1 and 2 patients (P=0.03). During a mean follow-up of 42±28 months (range: 6–112 months), patients with a permeable stent had resolution of portal hypertension-related manifestations. In 13 patients in whom PVR was feasible, stent permeability was 77% at 2 years (87% vs. 60% in patients who received anticoagulation or not, respectively; P=0.3).

PVR is feasible in most patients with non-cirrhotic, non-tumoral portal vein occlusion when there is no extension of the occlusion to distal branches.