Portal Hypertensive Gastropathy and MELD‐Na Score Predict Recurrent Gastrointestinal Bleeding After TIPSS: An ALTA Group Study.

Background: Transjugular intrahepatic portosystemic shunt (TIPSS) is highly effective for treatment of variceal bleeding; however, factors contributing to rebleeding complications remain unclear. Aims: In this study, we aim to determine risk factors for recurrent portal hypertensive gastrointestinal bleeding following TIPSS. Methods: Utilising the Advancing Liver Therapeutic Approaches multicentre database, we retrospectively identified adult patients who underwent TIPSS for secondary prophylaxis of variceal bleeding and had a gastrointestinal rebleeding event within 1 year. We developed multivariable logistic regression models to identify clinical/procedural characteristics associated with rebleeding. Results: We identified 476 patients, predominately middle‐aged (mean age 57), male (62%) and White (65%), with mean MELD‐Na 16. 16% (n = 77) had a rebleeding event; these patients were more likely to be male (p = 0.016), with higher serum creatinine (p = 0.005), MELD‐Na (p = 0.0002), portal hypertensive gastropathy on pre‐TIPSS upper endoscopy (p = 0.000) and with higher incidence of TIPSS revision (p = 0.000). There were no significant differences in type of TIPSS endoprosthesis, concurrent embolotherapy, and post‐TIPSS pressure gradients between those who experienced rebleeding and those who did not. After adjusting for TIPSS revision, multivariable analysis revealed MELD‐Na and presence of portal hypertensive gastropathy on pre‐TIPSS endoscopy were independently associated with rebleeding. Conclusions: In this retrospective analysis of a multicentre, nationally representative database, we found that apart from TIPSS‐related factors, high MELD‐Na and portal hypertensive gastropathy on pre‐TIPSS endoscopy were independent predictors of rebleeding within 1 year following TIPSS. These variables may be used to identify high‐risk patients who may require additional monitoring following TIPSS. [ABSTRACT FROM AUTHOR]