A justification for less restrictive guidelines on the use of metformin in stable chronic renal failure

Aim The aim was to justify less restrictive use of metformin in stable chronic renal failure, because a literature review reveals metformin is associated with a significantly lower incidence of cardiovascular events and mortality compared with other hypoglycaemic agents, and metformin-associated lactic acidosis is rare and causation uncertain. Studies on intentional metformin overdose and metformin bioavailability, renal clearance and plasma metformin in renal impairment provide evidence in support of a less restrictive use of metformin. Methods In metformin overdose (n = 22), lactic acidosis was not inevitable with a plasma metformin > 40 mg/l (therapeutic level c. 1 mg/l): Severe lactic acidosis (pH ≤ 7.21, plasma lactate ≥ 11 mmol/l, n = 8) did not occur unless plasma metformin was > 40 mg/l. Plasma lactate was a more consistent predictor of pH than plasma metformin, with plasma lactate ≤ 4.7 being associated with a pH ≥ 7.34. A likely ‘safe’ plasma lactate is < 3.5 mmol/l and plasma metformin < 10 mg/l. Results Plasma metformin can be predicted from estimated glomerular filtration rate and metformin dose. Reported plasma metformin in renal failure was always less than predicted plasma metformin. Predicted plasma metformin (mg/l), with an estimated glomerular filtration rate of 30 ml/min and metformin 2000 mg/day was 6.8; an estimated glomerular filtration rate of 20 ml/min and metformin 1500 mg/day was 5.1; an estimated glomerular filtration rate of 10 ml/min and metformin 500 mg/day was 4.4. Conclusion Metformin accumulates in renal failure and, although accumulation does not always lead to lactic acidosis, dose modification to achieve a predicted plasma metformin < 10 mg/l is suggested. As plasma metformin is not routinely available, plasma lactate should be useful in monitoring the use of metformin in renal failure.