Can we love Metformin? Need we to fear it? Or perhaps we must give it Machiavelli’s ultimate accolade, and love and fear the drug with equal weight?

Why to love Metformin? It is endorsed in the US, UK and Europe as the initial drug treatment for adults with type 2 diabetes. It treats diabetes, does not cause weight gain, does not increase risk of hypoglycaemia, and might reduce heart attacks and strokes. One tablet costs about £0.20 ($0.29/€0.26) and the drug has been widely used for upwards of 60 years.

So, must the love be tempered by fear? Metformin inhibits the mitochondrial respiratory chain, driving anaerobic metabolism and increasing lactic acid production. The drug is excreted almost entirely unchanged in urine so reduced kidney function may lead to accumulation of both metformin and lactate and therefore, a metformin-associated lactic acidosis (MALA). Where ingestion of ‘overdose’ doses of metformin is seen, acutely or intentionally, this may be termed metformin-induced lactic acidosis (MILA).

Notoriously, a similar drug, Phenformin, was withdrawn in the late 1970s after catastrophic lactic acidosis occurred in patients. The fear of metformin has lingered ever since, despite the overall incidence of lactic acidosis in metformin users being somewhere between 3 and 10 cases per 100,000 patient years and generally indistinguishable from the base rate in diabetics. Admittedly, reliable data specifically in patients with CKD is hard to come by and UK and US recommendations are to review the dose and not start the drug at an eGFR <45ml/min and stop the drug at <30ml/min.

If you’ve seen a case of MALA, it might have extinguished any love you had for metformin; mortality is 30-50%, serum lactate is often over 20mmol/l and pH can fall below 7. Those are frightening numbers. Luckily, the EXTRIP workgroup has produced management recommendations in 2015 based on a careful review of 175 articles.

The group managed to distil the following conclusions. Firstly, if you have a patient who is severely poisoned by metformin, then you should consider an extracorporeal therapy (ECTR). Why? Generally, because it brings all the benefits of ECTR: electrolyte management, reversal of hypothermia and support of renal function. Specifically, because it allows metformin removal, provides rapid and safe correction of acidosis over and above supportive care and bicarbonate, and reduces lactate.

But severe metformin poisoning is a vague concept, ingestions are often hard to quantify in the emergency setting, and metformin levels are hard to come by and cannot presently be said to be useful in management, so when should we start ECTR?

If any of lactate >20mmol/l, pH <7.0, or failure of standard supportive care (including bicarb), are present, the group recommend ECTR. Additionally, if lactate is 15-20mmol/l and/or the pH 7.0-7.1, then they suggest ECTR be considered. Finally, they recommend that the threshold for ECTR should be lowered in patients with impaired renal function, shock, a low GCS or liver failure.

The group suggest that intermittent haemodialysis with bicarbonate buffer is preferred over continuous modalities, and that dialysis achieves both better correction of acidaemia and superior removal of metformin and lactic acid. They also make the point that if no ultrafiltration is required, this group of critically unwell patients may tolerate intermittent dialysis.

In terms of when to stop ECTR, the group suggest that once lactate is below 3mmol/l and pH above 7.35, ECTR can be stopped safely.

Perhaps unsurprisingly for a critical illness, all of the group’s recommendations are assessed as based on D grade, or “very low level” evidence. However, the recommendations are clear and concise, and when added to FDA/NICE advice about the use of metformin in CKD, and the rarity of MALA, physicians everywhere should choose Machiavelli’s third option and both fear and love metformin.

Click here for EXTRIP Reference

About the author

I am an NIHR Clinical Lecturer at the Centre for Nephrology at University College London. I have clinical interests throughout nephrology and research interests in genome technologies and novel educational techniques in medicine.