Abstract

• Many guidelines recommend metformin as first-line therapy for patients with type 2 diabetes. This recommendation is primarily based on the results of the Ukpds trial published in 1998. However, the methodology of this trial has been criticised.
   
• In 2014, does the harm-benefit balance of metformin still justify its first-line use in type 2 diabetes? To answer this question, we conducted a review of the literature using the standard Prescrire methodology.
   
• In the Ukpds trial, involving about 1700 overweight diabetic patients, metformin monotherapy for about 10 years was more effective in reducing mortality than glycaemic control based mainly on dietary measures, and also more effective than treatment with a sulphonylurea such as chlorpropamide or glibenclamide, or with insulin. However, these results are undermined by several methodological flaws.
   
• In the Adopt trial, in which about 4400 patients were followed for 4 years, metformin, glibenclamide and rosiglitazone did not have significantly different effects on the risk of death or cardiovascular events.
   
• A meta-analysis of ten randomised trials versus placebo or other hypoglycaemic drugs did not show that metformin monotherapy had a statistically significant effect on mortality.
   
• In the Cosmic trial, including more than 5000 patients, metformin monotherapy for one year was not more effective in reducing mortality than another oral hypoglycaemic drug.
   
• In the Spread-Dimcad trial in 304 diabetic patients with coronary artery disease, metformin monotherapy appeared to be more effective in preventing cardiovascular complications than glipizide after 5 years of follow-up.
   
• The adverse effects of metformin mainly consist of dose-dependent gastrointestinal disorders and rare cases of life-threatening lactic acidosis. Kidney failure reduces metformin elimination.
   
• Metformin rarely causes hypoglycaemia and has no effect on body weight. It does not increase cancer-related mortality. It sometimes causes vitamin B12 deficiency leading to macrocytic anaemia or peripheral neuropathy.
  
  
• Metformin mainly carries a risk of interactions with drugs that impair renal function, such as non-steroidal anti-inflammatory drugs and iodinated contrast media. Renal failure can lead to metformin accumulation and an increased risk of lactic acidosis.
  
  
• In mid-2014, the only study to show a reduction in mortality and complications of diabetes remains the Ukpds trial. Taken together, the available data suggest that metformin monotherapy tends to reduce mortality and cardiovascular morbidity and mortality. Its adverse effects have been extensively studied and are acceptable provided renal function is monitored, especially in situations in which patients are at risk of kidney failure. The harm-benefit balance of metformin monotherapy remains favourable in most patients with type 2 diabetes when dietary measures alone are not sufficient.  

©Prescrire 1 November 2014

"Type 2 diabetes and metformin. First choice for monotherapy: weak evidence of efficacy but well-known and acceptable adverse effects " Prescrire Int 2014; 23 (154): 296-272. (Pdf, subscribers only)