Depending on the anatomy

of the targeted SPSS and preference of the interventional radiologist, the approach and occlusive type of intervention was determined. In case of recanalized paraumbilical veins, the approach was mainly percutaneous (occasionally transhepatically) and under local anesthesia, whereas for SRS, mesenterico-renal, or -caval shunts, the access was primarily transhepatically or less frequently by way of the femoral vein and under general anesthesia. In any case, the SPSS was first confirmed and evaluated by way of conventional angiography. Embolization was subsequently performed using either find more coils, amplatzer plugs or matrix, or a combination of these latter. Occlusion was angiographically confirmed at the end of the procedure. In patients fulfilling inclusion and lacking exclusion criteria, medical history, demographic and biochemical characteristics, drug history, specifics of the SPSS, details of the embolization procedure including potential associated complications, immediate and long-term outcome, and survival were reconstructed and completed according to medical records and clinical databases and/or by contacting the general physician check details in charge

of the patient. The data were retrieved per center by way of a standardized case-report form and centrally processed. Efficacy was evaluated by direct (primary) and indirect (secondary) outcome parameters. learn more The primary outcome measure was to evaluate the number of patients free of HE within 100 days pre- and postintervention (short-term efficacy) and during overall time of follow-up pre- and postintervention (long-term efficacy). Secondary parameters involved assessment

of the worst grade of encephalopathy, number and days of hospitalizations because of HE, changes in medical therapy, and the degree of disability on a short- and long-term basis, as defined above. The degree of disability or dependence in daily activities was assessed through the modified Rankin Scale (mRS).20 Safety was assessed by evaluating immediate postprocedural complications (bleeding, thromboembolic events, infection, anaphylaxis, hypotension, etc.) and in the long-term by monitoring portal hypertensive complications: de novo occurrence or aggravation of preexisting gastroesophageal varices or portal hypertensive gastropathy (with or without bleeding), ascites (with or without need of paracentesis), or spontaneous bacterial peritonitis. Changes in liver function were assessed by alterations in the MELD score. Statistical analysis was performed using MedCalc Statistical Software. Data are given as mean ± standard error of the mean (SEM) or as range between brackets.