Randomised clinical study: acute effects of metformin versus placebo on portal pressure in patients with cirrhosis and portal hypertension.

Background Portal hypertension is the main determinant of clinical decompensation in patients with liver cirrhosis. In preclinical data metformin lowers portal pressure, but there are no clinical data for this beneficial effect. Aims To investigate the acute effects of metformin on hepatic venous pressure gradient (HVPG) and liver perfusion. Methods In a randomised, double-blinded study design, we investigated 32 patients with cirrhosis before and 90 minutes after ingestion of 1000-mg metformin (n = 16) or placebo (n = 16). Liver vein catherisation was performed to evaluate HVPG and indocyanine green (ICG) infusion for investigation of hepatic blood flow. Results The mean relative change in HVPG was -16% (95% CI: -28% to -4%) in the metformin group compared with 4% (95% CI: -6% to 14%) in the placebo group (time × group interaction, P = 0.008). In patients with baseline HVPG ≥12 mm Hg clinically significant improvements in HVPG (HVPG 20% reduction in HVPG) were observed in 46% (6/13) of metformin-treated and in 8% (1/13) of placebo-treated patients (P = 0.07). There were no changes or differences in systemic blood pressure, heart rate, hepatic plasma and blood flow, hepatic ICG clearance, hepatic O2 uptake or inflammation markers between groups. Conclusions A single oral metformin dose acutely reduces HVPG in patients with portal hypertension without affecting systemic or liver hemodynamics or inflammatory biomarkers. This offers a promising perspective of a safe and inexpensive treatment option that should be investigated in larger-scale clinical studies with long-term outcomes in patients with cirrhosis and portal hypertension.