Diabetes Discoveries & Practice Blog

Metformin and Chronic Kidney Disease (CKD)

An iPad and stethoscope

Systematic review evaluates benefits and harms of metformin use in people with diabetes and CKD.

Blake Cameron, MD, discusses his team’s research on clinical outcomes of metformin use in people with chronic kidney disease (CKD) and explains what health care professionals need to know about changes to the U.S. Food and Drug Administration’s (FDA’s) guidance on metformin. These changes make some people who have both diabetes and CKD eligible to take metformin, and research suggests that the medication may have benefits for this population.

Q: Why did the U.S. Food and Drug Administration (FDA) add a boxed warning about metformin use in people with CKD when it approved the medication in 1994?

A: The story on metformin is really a story of guilt by association. Metformin belongs to a class of drugs called biguanides. From the 1950s to the 1970s, the leading drug in the biguanide class was phenformin. Globally, metformin was rarely used because phenformin was much more effective in lowering blood sugar. However, phenformin had a big problem: it was killing people. In 1977, phenformin was pulled from the U.S. market because it caused a fatal condition called lactic acidosis.

Biguanide research continued, and by the early 1990s, studies showed that metformin was safer than phenformin. However, because metformin is cleared from the body through the kidneys, there was concern that it would build up to toxic levels in people with CKD, and like phenformin, cause lactic acidosis. When the FDA approved metformin in 1994, they attached a boxed warning that cautioned against use of the drug in people who had even the slightest bit of abnormal kidney function, that is, men who had serum creatinine at or above 1.5 mg/dL and women who had serum creatinine at or above 1.4 mg/dL. However, for many years, clinicians have been concerned that this restriction might be excessively cautious.

Q: How has the FDA changed the boxed warning about metformin use in people with CKD?

A: In 2016, the FDA relaxed restrictions on metformin use in people with CKD in two ways. First, they included people who had worse kidney function. Second, they moved away from using the serum creatinine blood test as the marker of kidney disease, and they instead used the estimated glomerular filtration rate (eGFR). For a variety of reasons, eGFR is a much more accurate method of assessing kidney function.

The FDA’s new boxed warning indicates that metformin can be initiated in patients who have an eGFR greater than 45 mL/min, and treatment can be continued in existing patients as long as the eGFR remains above 30 mL/min.

Estimates suggest the changed guidance allows 1 million more people to take metformin. So, it's a substantial shift in the population that's eligible for the drug. Due to nuances in the conversion between creatinine and eGFR, people newly eligible to take metformin will tend to be younger and/or African American.

Q: Could you describe the goals of your team’s study of the clinical outcomes of metformin use in people with CKD and other historical contraindications?

A: Our main goal was to evaluate the benefits and harms of metformin use among populations with type 2 diabetes and moderate to severe CKD, congestive heart failure, or chronic liver disease with hepatic impairment. We wanted to confirm whether the published literature supports the change in FDA guidance.

Our study, a systematic review led by Dr. Matthew Crowley, Dr. Clarissa Diamantidis, and Dr. John Williams, was part of a report that was commissioned by the U.S. Department of Veterans’ Affairs (VA) as part of the VA Evidence-based Synthesis Program. While the VA was the primary audience, the study has much broader implications.

Also, even before the FDA changed its guidance, we knew that about 20 or 30 percent of patients who were taking metformin fell inside the original boxed warning.

Q: What did you and your coauthors find?

A: We found that the evidence supports the FDA’s change in the boxed warning. Among people with diabetes and moderate CKD, there is a 22 percent reduction in death due to any cause in those taking metformin compared with those who weren't taking metformin.

So, not only does metformin appear to be safe for people with diabetes and moderate CKD, but it appears to improve health and survival compared to alternative treatments.

Q: What future research is needed in this area?

A: Even though we found this consistent benefit of metformin, the quality of the studies that we reviewed was fairly poor. What we really need are randomized controlled trials for metformin and CKD to establish—with a much greater degree of certainty—the risks and benefits of the drug.

Further research would have a couple benefits.

  1. We need to figure out exactly what degree of CKD is permissible for metformin use. Our study systematically reviewed all the recently published literature on this question, and these studies by and large covered people with eGFRs between 30 and 60 mL/min. The question is still unanswered for people with eGFRs below 30 mL/min.
  2. Across the entire spectrum of kidney disease, we need to understand tradeoffs in metformin dosing. Even though metformin could potentially be used in people with more advanced CKD, we don't know if the dose should be reduced for those patients.

Q: What should health care professionals know about prescribing metformin for patients with diabetes and CKD?

A: All prescribers should take a look at the current FDA guidance on metformin, if they haven't done so. Again, this is a substantial change in terms of who and how metformin should be prescribed to these patients. As a practicing clinician, I still see patients who could potentially benefit from metformin but are steered away from it because of the previous boxed warning. It's my personal belief that far more people experience harm from not taking metformin than do from taking it.

Has the revised FDA guidance changed the way you think about metformin for your patients with diabetes and CKD?

Comments