Document resource
Background Variceal bleeding (VB) is a serious complication of liver cirrhosis. In addition to endoscopy, various medications are used to reduce the severity or incidence of VB. However, each of these medications have been associated with adverse events in these patients.Methods In the ATTIRE (Albumin to Prevent Infection in Chronic Liver Failure) trial, 777 patients were randomised to daily 20% albumin infusions or standard of care during hospitalisation with decompensated cirrhosis across 35 United Kingdom hospitals from 2016–2019. We performed a descriptive analysis of baseline characteristics, medications and outcomes for the 115 patients with suspected VB at baseline to evaluate the effects of proton pump inhibitors (PPIs), non-selective beta-blockers (NSBBs) and terlipressin (TP).Results PPIs were almost universally prescribed in patients admitted with VB, in 100/115 patients. PPIs were associated with lower blood pressure (BP) parameters at baseline (mean systolic BP 112mmHg vs 123mmHg, p=0.018; mean arterial pressure (MAP) 81mmHg vs 91mmHg, p=0.0024), possibly reflecting a more unwell patient cohort. No significant difference was seen in the incidence of adverse outcomes, including new infection, hepatic encephalopathy or renal dysfunction. These findings held true in a subcohort of 41 patients prescribed intravenous (IV) PPI.TP was prescribed in just under half of the patients (54/115). Use was not associated with any significant difference in blood pressure parameters, and baseline heart rate per minute (82 vs 92, p=0.0008), bilirubin (75μmol/L vs 115μmol/L, p=0.017) and overall MELD scores (16.12 vs 18.29, p=0.042) were actually lower in the TP group. There were no significant differences between groups in incidence of new renal dysfunction and infection, nor in-hospital, 28-day and 90-day mortality.28/115 patients were prescribed NSBBs. Blood pressure parameters at baseline and subsequent renal dysfunction were no different between groups. Patients on NSBBs had significantly lower 28-day (0/28 vs 18/87, p=0.009) and 90-day (2/28 vs 22/87, p=0.04) mortalities.Conclusion In this real-world cohort of patients hospitalised with suspected variceal bleeding, PPIs, TP and NSBBs were not associated with adverse events in hospital or increased mortality following discharge. Interpretation is limited by not knowing whether PPIs were existing prescriptions or initiated at presentation and that almost all patients received PPIs, but short-term use appears safe, supported by similar findings in a subcohort prescribed IV PPIs. NSBBs in the context of acute VB likely reflects existing prescriptions and previous interactions with a hepatology team and use was associated with improved mortality.