metformin and lactic acidosis- How strong of a relation?

How much should we be worried about.I felt that lactic acidosis has been hyped a little by everyone in clinical practice. In fact I am yet to see a case of lactic acidosis that can be attributed to Metformin use.
The relation between the two is a legacy of Phenformin, which was taken out of market in 1978 due to high incidence of lactic acidosis.It is still prescribed in Brazil ,Greece, China & Portugal.
Metformin also possesses this risk, but to a lesser extent, mainly because it is not metabolised in the liver(as phenformin).


Lets recap lactic acid metaboism....
Lactic acid is the a dead end product beacuse pyruvate, its immediate precursor is also its only route of metabolism. Lactate is buffered by equal amount of HCO3, and the liver's job is to oxidise the lactate and replace the HCO3.This is a formidable task, as 1400 mmol of lactate are produced everyday, but the plasma concentration is maintained around 1 mmol. The kidneys are responsible for excretion of 10 -20% of lactate. If we look at the picture above, increase in NADH (from alcohol) or pyruvate(from glycolysis) would shift the reaction towards lactate. Metformin may inhibit oxidative metabolism and increase NADH, and thus worsen lactic acidosis. But this happens usually if excretion or buffering is deficient.


Lets look at some data. Of the first one million patients who took Metformin in US, 47 developed lactic acidosis. Of them 43 had renal failure or CHF. Only 4 didn't have any other reason apart from Metformin, and these were not fatal (NEJM 98). Based on this observation, CHF was included in the CI list for Metformin.
Another study done my Bristol Meyers (COSMIC study) comparing 1 year treatment of 7200 pts on Metformin Vs 1500 pts on other antdiabetics showed no cases of lactic acidosis. A meta analysis in 2003, of 194 studies didn't reveal any case of lactic acidosis in 36,893 pt years in Metformin group Vs 30,109 pt years in non-metformin group. Many of the studies in this analysis were of short duration, and included very few number of patients though.
The most recent Cochrane review 2008 on this subject concluded "There is no evidence from prospective comparative trials or from observational cohort studies that metformin is associated with an increased risk of lactic acidosis" . 


The main accusations on Metformin and lactic acidosis have been only from case reports, and they still continue to be reported. 
Usually pts on Metformin who develop lactic acidosis are sick, and have other underlying problems.I will go back to my first statement,that this relation has been overhyped, and that Metformin does not seem to cause lactic acidosis IF USED IN THE RIGHT POPULATION.( avoid if CKD 4 & 5, caution in CKD 3, avoid if Creat >1.3 in females, and >1.4 in males, CHF, and any underlying reason that can cause tissue hypoxia--> lactic acidosis)