Lactic acidosis in metformin therapy: searching for a link with metformin in reports of 'metformin-associated lactic acidosis'.

OBJECTIVE: The link between metformin and lactic acidosis in metformin therapy may be causal, associated or coincidental. Our objective was to investigate this link by studying and analysing published reports of so-called 'metformin-associated lactic acidosis'. RESEARCH DESIGN AND METHODS: systematically searched in the BIOSIS, DERWENT, EMBASE, MEDLINE, and PASCAL databases of the English language and non-English language literature for all reports of so-called 'metformin-associated lactic acidosis' published from May 1995 through January 2000. We did not include reports related to metformin overdose or contrast media-induced renal failure. Metformin accumulation and concurrent pathologies were critically reviewed as precipitating factors for metformin-associated lactic acidosis. Metformin accumulation was assessed in terms of the recorded measurement of metformin concentration in plasma or, if not available, by the presence of primary renal failure, i.e. renal failure that was not secondary to a shock syndrome.
RESULTS: We found 21 reports describing a total of 26 patients. Criteria of lactic acidosis (lactate > 5 mmol/l, pH <or= 7.35) were not met in four patients. In the remaining 22 patients, plasma metformin concentration was determined in only four, of whom one had a normal value. In the 18 patients with lactic acidosis where plasma metformin concentration data was not available, the presence of primary renal failure was absent or unlikely in six patients, uncertain in two, and likely or proven in 14. With regard to these 14 patients, the precipitating factor was metformin in 12 patients (in the context of renal failure either chronic or acute) and intercurrent pathologies in two others. Overall, lactic acidosis was either absent (n = 4), precipitated by concurrent pathology (n = 8), precipitated by metformin without apparent associated pathology (n = 12) or of uncertain origin (n = 2). Death occurred 10 times but only once in the 12 patients with metformin-induced lactic acidosis and this was not related to metformin.
CONCLUSIONS: While the term 'metformin-associated lactic acidosis' is commonly used to depict all situations of lactic acidosis in metformin therapy, true metformin-associated lactic acidosis, i.e. one which refers to metformin and concurrent pathologies as co-precipitating factors, was never observed in the studied reports. As there was no mortality due to metformin alone, it is important that physicians are familiar with the range of other risk factors that contribute to lactic acidosis in patients treated with metformin.