Background and aim: pre-emptive transjugular intrahepatic portosystemic shunt (p-TIPS) and salvage TIPS are the two main indications of TIPS insertion in the setting of portal hypertension (PHT) related gastrointestinal bleeding in patients with cirrhosis. We aimed to evaluate the factors associated with outcome after p-TIPS and salvage TIPS insertion in our centre including sarcopenia parameters.
Patients and methods: We retrospectively included all patients with cirrhosis who underwent p- TIPS and salvage TIPS in the Lausanne University Hospital between 2010 and 2019. We quantified the lumbar-3-skeletal muscle index (L3SMI) using a deep learning-based method. Sarcopenia was defined as L3SMI ≤ 50 cm2/m2 in men and ≤ 39 cm2/m2 in women.
Results: A total of 15 patients who underwent p-TIPS and 23 who underwent salvage TIPS fulfilled the inclusion criteria. Patients were predominantly male (n=32, 84.2%), with a median age of 60.2 years (51.5-66.0), and median MELD of 14.5 (11.0-16.8). Lactate level (p=0.009), circulation failure (p=0.02), vasopressors use (p=0.02), and mechanical ventilation requirement (p=0.02) were significantly lower for patients who underwent p-TIPS compared to patients with salvage TIPS with a greater 1-year survival 92.8% vs. 37.3% (p=0.002). In p-TIPS patients, none of the available scores were significantly performant in identifying patients at risk to develop severe HE or to have clinical ascites at 1 year. In salvage TIPS patients, MELD score (p=0.04) presence of ACLF (p=0.04) and its grade (p=0.05) was the most discriminant variable. The 42-day survival as following in patients without ACLF, ACLF grade 1, 2 and 3: 85.7% vs. 66.7% vs. 62.7% vs. 0%, p=0.05.
Conclusion: In our experience, patient who underwent p-TIPS had excellent outcome and high survival rate at 1-year and no discriminatory predictive factors for poor outcome were identified. In patients who underwent salvage TIPS the 6-weeks mortality remains high and MELD score and ACLF grade are the best predictive for 6-weeks survival identified. Sarcopenia parameters were not associated with outcome with however a limited sample size for these analyses.