6 Pearls About Metformin and Lactic Acidosis
February 13, 2014, 4:00 pm
Metformin accumulation: Lactic acidosis and high plasmatic metformin levels in a retrospective case series of 66 patients on chronic therapy. Vecchio S et al. Clin Toxicol 2014 Feb;52:129-135.
Metformin is frequently used alone or in combination to treat type 2 diabetes. It lowers blood glucose by decreasing hepatic gluconeogenesis, predominantly by inhibiting mitochondrial respiratory chain complex I. The drug is eliminated mainly by the kidneys, and acute or chronic renal insufficiency may allow accumulation of the drug with increasing levels.
A small percentage of patients on metformin develop severe lactic acidosis. There has been an ongoing controversy as whether this acidosis is metformin-associated or
metformin-induced. This paper, from the Pavia Poison Control Centre in Northern Italy, helps shed light on this question.
The authors retrospectively reviewed patients admitted to their toxicology unit over a 5-year period. Eligible patients were on chronic metformin therapy at the time of admission, had lactic acidosis (pH < 7.35, arterial lactate > 5 mmol/L), and elevated metformin levels (plasma metformin > 4 mcg/ml). Cases of acute overdose were excluded. The study objective was to correlate the metformin levels with measured pH, lactate levels, renal function, and mortality rate.
Sixty-six eligible patients were identified. All patients presented with acute renal failure and severe lactic acidosis (mean pH 6.91, mean lactate 14.36 mmol/L). About half the patients had a pre-existing contraindication to metformin therapy, predominantly renal failure and/or heart disease. Approximately 75% presented after several days of a mild gastrointestinal prodrome with nausea, vomiting, and diarrhea; this may either have represented the initial manifestations of metformin poisoning, or caused dehydration that precipitated toxicity. Common presenting signs and symptoms included neurological impairment or coma, dyspnea, and shock. The early mortality rate was 26%, and did not correlate with either lactate level or metformin concentration. This may be because the main determinants of prognosis were the patients condition and underlying medical problems, or because — since metformin is a mitochondrial poison — mortality correlates with levels in the cells, not the blood. Virtually all the patients (94%) were treated with hemodialysis; the four who weren’t either died before it could be instituted (1 patient) or spontaneously recovered (3).
This paper is well worth reading, as is the accompanying commentary by David Juurlink (@DavidJuurlink) and D. M. Roberts. The following are some pearls and take-home lessons about metformin-associated metabolic acidosis (MALA) from these two articles:
- Evaluate patients started (or continued) on metformin for contraindications to the drug.
- Screen patients on metformin who present with a gastroenteritis-type syndrome or other conditions that predispose to dehydration for metabolic acidosis (my feeling is that an initial blood gas would be unnecessary since checking the electrolytes and anion gap should suffice).
- Consider early hemodialysis in patients presenting with MALA — this would both help remove the drug and correct severe acidosis.
- Add metformin toxicity to the differential diagnosis in appropriate patients suspected of having sepsis, mesenteric ischemia, or respiratory failure.
- Know that metformin-induced metabolic acidosis does indeed exist, and that these patients are typically extremely ill.
- Realize that with proper care these patients can survive, even if they’ve presented with amazingly low pH readings.