Targeted albumin infusions in hospitalized patients with cirrhosis receiving terlipressin: A post-hoc analysis of ATTIRE.

Abstract

Introduction and objectives: The ATTIRE trial showed increased severe adverse events with targeted albumin therapy. Safety concerns exist regarding albumin-terlipressin use for hepatorenal syndrome-acute kidney injury (HRS-AKI), but terlipressin is also used for variceal bleeding and hypotension. We evaluated the safety of terlipressin and albumin for any indication using ATTIRE data.

Materials and methods: In ATTIRE, hospitalized decompensated cirrhosis patients were randomized to daily 20 % albumin (serum albumin ≥30 g/L) or standard care for up to 14 days post-randomization. Of 777 patients, 42 received terlipressin at randomization in the targeted albumin arm and 41 in standard care. We studied death and fluid-related complications from serious adverse event reporting during the 15-day trial period.

Results: Indications for terlipressin were variceal bleeding (73 %), HRS-AKI (23 %) and sepsis-induced hypotension (3 %). Median albumin dosing was higher with targeted albumin than standard care for variceal bleeding (200 g vs. 0 g) and sepsis-induced hypotension (180 g vs. 0 g), but similar for HRS-AKI (220 g vs. 230 g). A composite of death and fluid-related complications was more common with targeted albumin compared to standard care (log-rank: p = 0.011), where the increased risk persisted when adjusting for baseline MELD. This composite outcome occurred more often in variceal bleeding patients treated with targeted albumin (n = 7) compared to standard care (n = 2), although the difference was not statistically significant (p = 0.064).

Conclusions: In hospitalized cirrhosis patients, targeted albumin infusions with terlipressin may increase the risk of death and fluid-related complications, particularly in variceal bleeding. Caution is warranted when using albumin in this subgroup.

Clinical trial number: EudraCT number: 2014-002,300-24 and International Standard RCT Number: 14,174,793 Research Ethics Committee Number: 15/LO/0104.