Long-term metformin use may improve clinical outcomes in diabetic patients with non-alcoholic steatohepatitis and bridging fibrosis or compensated cirrhosis.

BACKGROUND: Metformin may protect against hepatocellular carcinoma and mortality among patients with type 2 diabetes. AIM: To investigate whether long-term use of metformin improves survival and reduces liver-related outcomes among patients with type 2 diabetes and non-alcoholic steatohepatitis.
METHODS: A total of 191 diabetic patients with biopsy-proven non-alcoholic steatohepatitis and bridging fibrosis or compensated cirrhosis were retrospectively identified at Indiana University Medical Center between October 2004 and January 2016. Of them, 110 were users and 81 never-users of metformin. Primary outcomes were transplant-free survival, development of hepatocellular carcinoma or a first event of hepatic decompensation.
RESULTS: Cirrhosis was present in 85% of metformin users and 88% of nonusers. Metformin dose was greater than or equal to 1 g/d in 104 out of 110 users and its median duration of use was 6 (95% CI: 4.4-7.9) years. The mean follow-up was 6.92 and 6.80 years for metformin users and non-users, respectively. During follow-up, 28 patients developed hepatocellular carcinoma (metformin users: 7, nonusers: 21), and 52 died (metformin users: 7, nonusers: 24) or were transplanted (metformin users: 13, non-users: 13). Metformin use was associated with lower risk of overall mortality or transplant (HR: 0.42; 95% CI: 0.24-0.74, P = 0.003) and hepatocellular carcinoma (sHR: 0.25; 95% CI: 0.11-0.58, P = 0.001), and remained independently associated with both outcomes after propensity-score and covariate-adjusted analyses. No instances of hepatotoxicity or lactic acidosis were observed.
CONCLUSION: Our study demonstrated an association between long-term metformin use and reduced the risk of all-cause mortality/transplant and hepatocellular carcinoma in diabetics with non-alcoholic steatohepatitis and advanced fibrosis.