1 - Factors Impacting Post-TIPS Liver Failure (PTLF): A Retrospective Cohort Study

Monday, April 13, 2026

10:00 AM - 10:09 AM ET

Presenting Author(s)

JC

Jirapa Chansangrat, MD

Assistant Professor
Medical Institute, Suranaree University of Technology, Thailand

Author/Co-author(s)

Amrit Khooblall, MBA

Student
Division of Interventional Radiology, The Cleveland Clinic Imaging Institute, United States
LK

Layth M. Khawaja, BS

Medical Student
Northeast Ohio Medical University, United States
MN

Monika Neale, DO

Resident
Division of Interventional Radiology, Imaging Institute, Cleveland Clinic, United States
LW

Logan Worley, MD

Resident
Division of Interventional Radiology, Imaging Institute, Cleveland Clinic, United States
YL

Yu An Lin, DO

Radiology resident
Cleveland Clinic, United States
Baljendra Kapoor, MD

Staff Physician
Michigan Medicine, United States
Sameer Gadani, MD, FSIR

Associate Professor and Staff
Cleveland Clinic Foundation , United States

Purpose: To identify clinical and laboratory predictors of PTLF

Materials and Methods: A retrospective analysis was performed in 950 patients who underwent TIPS. Baseline demographics, pre-TIPS clinical factors, and procedural details were collected. Serial measurements of laboratory data (AST, ALT, bilirubin, albumin, INR, MELD) were obtained at baseline, daily for the first 7 days, and at 1-, 3-, 6-, and 12-months post TIPS. Key dynamic variables included post-TIPS peak and % change of enzyme values, time to peak, and time to recovery. PTLF was defined according to the criteria proposed by Gaba et al {1}. Dynamic patterns were compared between PTLF and non-PTLF patients using Mann-Whitney U and X² tests. Predictors were assessed using univariable and multivariable logistic regression. Cox proportional hazards regression was applied to evaluate predictors of AST and ALT recovery.

Results: PTLF occurred in 169 patients (17.8%). Patients with PTLF were older and more likely to have pre-TIPS spontaneous portosystemic shunt (SPSS) (12.4% vs 6.4%, p=.007) and celiac stenosis (3.0% vs 0.3%, p< .001). PTLF patients showed higher post-TIPS peak AST (144 vs 107 U/L, p=.001), peak ALT (178 vs 128 U/L, p=.001), greater % change in AST and ALT (325% vs 206%, p< .001, and 303% vs 193%, p< .001, respectively). In multivariable regression, independent predictors of PTLF included older age (OR 1.04, p=.027), previous hepatic encephalopathy (HE) (OR 0.12, p< .001), post-TIPS peak MELD and %change of INR (OR 1.03, p=.042). In Cox regression, PTLF was not independently associated with AST or ALT recovery.

Conclusion: Both baseline factors (age, HE) and post-TIPS peak MELD and % change of INR independently predicted PTLF. In contrast, Cox regression analyses showed that PTLF was not associated with AST or ALT recovery. These findings highlight potential factors causing PTLF, and support the use of combined clinical and dynamic laboratory markers for risk stratification and early transplant evaluation after TIPS.