Abstract No. 73: What variables predict the need for revisions in patients undergoing TIPS with PTFE-covered stents?

Abstract

To determine what clinical, laboratory, or procedural variables predict the need for TIPS revisions in individuals with PTFE-covered TIPS. Between 2003–2009, a total of 141 patients underwent TIPS venographic evaluation following TIPS placement with PTFE-covered stents. Of these, a total of 43 patients required TIPS revisions. Twenty-three (23) different variables were modeled using univariate analysis, and then evaluated using a Cox proportional hazards model. Of the variables tested, only the presence of a previous intervention predicted the need for further interventions (p=0.03). This variable also had an interaction with time (e.g. the hazard ratio increased over time), so that there was a greater likelihood of a need for a second revision as the time from the previous intervention increased. Other variables, including the size of the stent, etiology of portal hypertension, post-TIPS portosystemic gradient, and intrahepatic stent location failed to predict the need for future interventions. Individuals requiring one TIPS reinterventions should be aggressively screened for TIPS failures because of the increased likelihood that they will require another intervention. No correlation with other clinical, laboratory, or procedural variables was noted with the need for TIPS revisions.