Abstract
Sodium-glucose cotransporter 2 inhibitors (SGLT2i) may have favourable neurohumoral and metabolic effects in patients with chronic liver disease. However, studies examining SGLT2i in this population have been limited to patients with non-alcoholic fatty liver disease and have focused on surrogate biomarkers. Our aim was to evaluate whether SGLT2i can reduce the incidence of ascites and death over a period of 36 months in patients with cirrhosis and diabetes mellitus. Using electronic health data from Veterans Affairs hospitals in the United States, we conducted a propensity score matched intention-to-treat analysis among veterans on metformin who subsequently received either SGLT2i or dipeptidyl peptidase-4 inhibitors. Among 423 matched pairs (in total, 846 patients), we found no significant difference in the risk for ascites (hazard ratio 0.68 for SGLT2i, 95% confidence interval 0.37-1.25; p = .22) but did find that SGLT2i users had a reduced risk for death (adjusted hazard ratio 0.33, 95% confidence interval 0.11-0.99; p < .05). In comparison with dipeptidyl peptidase-4 inhibitors, SGLT2i may improve survival for patients with cirrhosis who require additional pharmacotherapy for diabetes mellitus beyond metformin, but confirmatory studies are necessary.