Is Metformin Safe for Patients With Chronic Kidney Disease?

Lynda Szczech, MD, MSE


January 30, 2017

This feature requires the newest version of Flash. You can download it here.

Hello. I am Lynda Szczech, a practicing nephrologist in Durham, North Carolina. I choose the topics for these video viewpoints based on my own clinical questions. If I scan through a journal and see an article on a subject I want to learn more about, that is how I choose a topic.

This video viewpoint is on a very practical subject: metformin. Metformin is an oral anti-hyperglycemic with a long history. In many clinical practice guidelines, metformin is the first-line oral therapy for patients with type 2 diabetes mellitus, and it is the first-line therapy for a reason. Metformin achieves good control of blood sugar without causing significant weight gain and without significant risk for hypoglycemia. It also provides a potential for mortality benefits.

We do not see a lot of metformin in our nephrology clinics and practices; metformin has carried a black box warning for a very long time because a related medication, phenformin, has been associated with lactic acidosis. The black box warning specifically advises against using metformin in patients with an underlying medical condition that may predispose to lactic acidosis. Renal insufficiency is on that list of conditions because of the fear that decreased kidney function could result in a failure to excrete the lactic acid once it is produced. Also on the list are congestive heart failure and chronic liver disease.

A Brief History

A considerable amount of medical literature has examined the potential effect of metformin on outcomes and the potential risk for metformin to cause lactic acidosis compared with other oral anti-hyperglycemics, including case reports of lactic acidosis associated with metformin. During the past 10 years, two reviews have been published that enhance our understanding of the relative risk for lactic acidosis with metformin.

The first was a large Cochrane Database Systematic Review[1] published in 2006, comparing the risk for lactic acidosis in almost 48,000 patient-years (metformin) versus just over 38,000 patient-years (non-metformin). The other analysis[2] was published in 2008 and looked at about 50,000 patients. Both analyses failed to find an increased risk for lactic acidosis in patients receiving metformin above and beyond the baseline risk, when compared with patients receiving other sulfonylureas.

Are there limitations to these analyses? Yes. First and foremost, neither specifically looked at the risk related to glomerular filtration rate (GFR) or creatinine clearance. These were background, all-comer analyses. They included people with normal kidney function and people with compromised kidney function, and thus cannot prove that the risk is not elevated in people with chronic kidney disease (CKD). But these very large population-based studies did not find an elevated risk for lactic acidosis associated with metformin. They deliver a risk-benefit analysis, weighing the benefit of good glycemic control, potential impact on hard cardiovascular outcomes, the risk for hypoglycemia, and the risk for lactic acidosis.

Does Metformin Provide a Mortality Benefit?

Several observational studies suggest that metformin does provide a mortality benefit. I want to specifically draw your attention to a Swedish National Diabetes Register study. Ekström and colleagues[3] looked at more than 50,000 patients with type 2 diabetes, using patients' levels of kidney function to understand the effect of metformin within the strata of estimated GFR (eGFR). This analysis demonstrated a significant association between metformin and lower mortality in patients with an eGFR of > 60 mL/min/1.73 m2 and those with an eGFR between 45 and 60 mL/min/1.73 m2. The hazard ratio for both subgroups was about 0.87 and both were statistically significant. In patients with an eGFR between 30 and 45 mL/min/1.73 m2 the hazard ratio was 1.02 with a confidence interval beautifully straddling 1, suggesting that there was a neutral association with mortality. This represented a smaller sample size but clearly showed a difference between those subgroups.

Is there a benefit to metformin? These observational studies suggest that there could be, at least in patients whose eGFRs are > 45 mL/min/1.73 m2. The benefit did not seem to differ among various categories of patients with multiple medical variables. Overall, these studies found a potential for benefit and not necessarily solid evidence for a risk of harm.

A Re-examination by the FDA

Why are we talking about metformin now, and why is a new systematic review[4] relevant? On re-examination of the data associated with metformin, the US Food and Drug Administration (FDA) published a safety announcement[5] in April 2016, providing additional information on how to dose metformin in patients with kidney disease. We may be seeing more metformin use in patients who are referred to our nephrology clinics, given their recommendations.

The systematic review by Crowley and colleagues[4] summarizes the outcomes literature of metformin in patients with CKD, congestive heart failure, or chronic liver disease. I will review the findings in patients with moderate to severe kidney disease. I encourage you to look at the article to review the findings in patients with heart failure and liver disease.

The investigators identified all English-language studies that have examined outcomes of adults with type 2 diabetes and CKD (defined as an eGFR < 60 mL/min/1.73 m2), compared metformin vs other anti-hyperglycemic regimens, and reported mortality and MACE (major adverse cardiovascular event) outcomes in the analysis. They abstracted the data, performed meta-analytic techniques, and summarized it for our consumption.

They found six studies that included all of this information in people with CKD. Although six is not a large number of studies, it nonetheless comprised a substantial number of people who had type 2 diabetes and CKD. Five of the six studies included information on all-cause mortality. The meta-analysis of these five studies revealed a 22% lower mortality risk for patients receiving metformin compared with those who received a different anti-hyperglycemic agent, with a P value < .001—highly significant.

Similar to the Swedish cohort study, two of the studies provided mortality information on CKD categories specifically, and these studies suggested potentially less benefit in patients with eGFRs 30-45 mL/min/1.73 m2, indicating that this is potentially a subgroup where the risk-benefit ratio may be slightly different from that in people whose eGFRs are > 45 mL/min/1.73 m2.

Only two of the studies provided information on MACE outcomes. One study failed to demonstrate a difference in these events. The other study demonstrated a lower risk for readmission for congestive heart failure associated with metformin, with an odds ratio of .091 and a confidence interval of 84-99. If there is a benefit here, it is relatively small. The benefit that this study found to be associated with metformin was in all-cause mortality, which arguably is a good outcome.

Finally, fewer patients with CKD who received metformin had an elevated risk for severe hypoglycemia, compared with those patients on other anti-hyperglycemic regimens. I believe that this is of great importance, given that many of the studies we have seen in recent years show that severe hypoglycemic events are related to a higher risk for all-cause mortality.

Overall, this timely and well-performed systematic review comes at a time when we may be observing more metformin use in patients seen in our nephrology clinics. I encourage you to look at the FDA safety announcement.[5] The announcement provides recommendations on when and how to start, when to temporarily stop, and when to discontinue metformin altogether based on kidney dysfunction. The FDA suggests that starting metformin in patients who have stage 1 or 2 kidney disease should be safe, given the available safety information.

For patients who have stage 3A disease, with an eGFR of 45-60 mL/min/1.73 m2, the risk-benefit ratio appears to be the same as for patients with stage 2 or 1 disease. For patients with stage 3B disease (eGFR 30-45 mL/min/1.73 m2) who are already on metformin, this FDA safety announcement suggests that it is appropriate to reassess the risk-benefit ratio at that point. It further suggests that metformin would not be started de novo in a patient who presents with an eGFR of 30-45 mL/min/1.73 m2. On the basis of the renal clearance parameters, the FDA continues to state that metformin is contraindicated in patients with an eGFR < 30 mL/min/1.73 m2.

I hope that I have provided a bit of history of metformin and kidney disease, information about the risks, and a lot of new information about the potential benefits. As we see more of our patients using metformin, we will be able to assist them to manage their blood sugars with a good understanding of their kidney function, its trends, and when to start and stop this medication on the basis of this risk-benefit ratio.

With that, I wish you a very good day and I hope that you found this informative.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.