Original Article
Risk factors for stent graft thrombosis after transjugular intrahepatic portosystemic shunt creation
Abstract
Background: To identify risk factors of stent graft thrombosis after transjugular intrahepatic portosystemic shunt (TIPS) creation.
Methods: Patients who underwent TIPS creation between June 2003 and January 2016 and with followup assessing stent graft patency were included (n=174). Baseline comorbidities, liver function, procedural details and follow-up liver function tests were analyzed in association with hazards of thrombosis on followup. Competing risk cox regression models were used considering liver transplant after TIPS creation as the competing risk variable.
Results: One-, 2- and 5-year primary patency rates were 94.1%, 91.7% and 78.2%, respectively. Patient age [sub-hazard ratio (sHR): 1.13; P=0.001], body mass index (BMI) <30 (sHR: 33.08; P=0.008) and a higher post-TIPS portosystemic pressure gradient (sHR: 1.14; P=0.023) were significantly associated with TIPS thrombosis in multivariate analysis. A higher rate of TIPS thrombosis was observed in those for whom the procedure was clinically unsuccessful (P=0.014). A significant increase in incidence of thrombosis was noted with increasing tertiles of post-TIPS portosystemic gradients (P value for trend=0.017).
Conclusions: Older age, lower BMI and higher post-TIPS portosystemic gradients were associated with higher hazards of shunt thrombosis after TIPS creation using stent grafts. Higher rates of shunt thrombosis were seen in patients for whom TIPS creation was clinically unsuccessful. The association between TIPS thrombosis and higher post-TIPS portosystemic gradients may indicate impaired flow through the shunt, a finding which may be technical or anatomic in nature and should be assessed before procedure completion.
Methods: Patients who underwent TIPS creation between June 2003 and January 2016 and with followup assessing stent graft patency were included (n=174). Baseline comorbidities, liver function, procedural details and follow-up liver function tests were analyzed in association with hazards of thrombosis on followup. Competing risk cox regression models were used considering liver transplant after TIPS creation as the competing risk variable.
Results: One-, 2- and 5-year primary patency rates were 94.1%, 91.7% and 78.2%, respectively. Patient age [sub-hazard ratio (sHR): 1.13; P=0.001], body mass index (BMI) <30 (sHR: 33.08; P=0.008) and a higher post-TIPS portosystemic pressure gradient (sHR: 1.14; P=0.023) were significantly associated with TIPS thrombosis in multivariate analysis. A higher rate of TIPS thrombosis was observed in those for whom the procedure was clinically unsuccessful (P=0.014). A significant increase in incidence of thrombosis was noted with increasing tertiles of post-TIPS portosystemic gradients (P value for trend=0.017).
Conclusions: Older age, lower BMI and higher post-TIPS portosystemic gradients were associated with higher hazards of shunt thrombosis after TIPS creation using stent grafts. Higher rates of shunt thrombosis were seen in patients for whom TIPS creation was clinically unsuccessful. The association between TIPS thrombosis and higher post-TIPS portosystemic gradients may indicate impaired flow through the shunt, a finding which may be technical or anatomic in nature and should be assessed before procedure completion.
