Q&A with Matthew Crowley: Systematic Review of Clinical Outcomes for Metformin

Friday, January 20, 2017

When managing diabetes, there are many possible approaches to treat a given patient. Factoring a patient’s other health issues into his/her diabetes management routine is essential in choosing the best options.

Metformin is the first-line medication used to treat type 2 diabetes because it is affordable, doesn't cause patients to gain weight, and doesn't cause excessively low-blood sugar. However, when metformin was first approved for use in type 2 diabetes in 1994, the Food and Drug Administration gave the drug a “boxed warning” label for patients with chronic kidney disease, congestive heart failure, and chronic liver disease. The warning informed medical practitioners that metformin use can cause lactic acidosis in these patients, which can be fatal.

Over the years, the FDA slowly changed its guidelines for metformin. In 2006, heart failure was removed as an absolute contraindication and was changed to a precaution.

Matthew J. Crowley, MD, assistant professor of medicine in the Division of Endocrinology, Metabolism, and Nutrition, and the Durham Veteran Affairs Evidence-based Synthesis Program (ESP) team spent roughly a year conducting a systematic literature review on metformin’s benefits in patients with chronic kidney disease, congestive heart failure, and chronic liver disease – health issues that have traditionally kept people from using metformin.

The paper, "Clinical Outcomes of Metformin Use in Populations With Chronic Kidney Disease, Congestive Heart Failure, or Chronic Liver Disease: A Systematic Review," was published in Annals of Internal Medicine in January 2017.

Crowley on the review

Q: What were your goals for this study?
A: Our team's goals were to survey the existing literature in a systematic and rigorous fashion to identify all relevant evidence about the benefits and harms of metformin in patients with chronic kidney disease, congestive heart failure, and chronic liver disease. Because others have recently looked at lactic acidosis, this paper focused on mortality and cardiovascular outcomes. There were a series of other outcomes of interest, as well, but unfortunately, we didn't find much evidence addressing them.

Q: Did you find any noticeable differences between the three diseases?
A: Even though metformin has traditionally been withheld from these patients due to concerns about lactic acidosis, the available evidence shows that it's actually associated with lower mortality in all three groups. What we found was exciting and it supports what the FDA has been doing with by expanding patients' access to this medication.

Q: As for the results, what expectations did you and your team have during the systematic review?
A: Because these patients have not used metformin routinely, there hasn't been a great understanding about how metformin affects long-term outcomes for populations with traditional contraindications and precautions. One prior review in congestive heart failure gave us the sense that metformin may be useful, but for the most part, we weren't sure what to expect.

One surprise was that the data available on this topic are fairly limited – all the evidence we found was observational, so subject to various biases. That added the additional challenge of trying to interpret the data appropriately.

Q: Why is it relevant for medical practitioners to understand this knowledge about metformin?
A: Primary care doctors are one group that we hope will find this review useful, because most type 2 diabetes is managed in primary care settings. Helping patients find effective, affordable options for diabetes can be challenging. Metformin is a generic medication, which makes it very affordable, and it doesn't cause weight gain, hypoglycemia, or many other severe side effects. It's exciting that we have this added data to support the idea that metformin may benefit some patients who historically have not been able to use it.

Q: Why was it interesting to you to take on such a big project regarding metformin?
A: As an endocrinologist, I care for a lot of patients with diabetes. Chronic kidney disease is a common complication of diabetes, and patients with chronic kidney disease unfortunately haven't been able to use metformin in the past. I've had many patients whose kidney function worsened over time, and we've had to discontinue metformin – this led to deterioration in their diabetes control. Knowing that metformin not only seems to be safe from the standpoint of lactic acidosis in mild-to-moderate chronic kidney disease, but also appears to be associated with lower mortality, will really help us in the clinical setting. It already has – I've been able to restart metformin for some patients who had previously stopped it.

Q: What were the most challenging aspects of using the systematic method?
A: When performing a systematic review, there is a lot you need to do before you even start looking at literature. This includes gathering input from clinical experts in the field, scoping your questions, and working with a search librarian to refine your search strategy.

The biggest challenge was going from a massive number of articles and identifying which studies were the most accurate and relevant to our questions. It's definitely a laborious process. You're starting with thousands and thousands of possible articles that you may need to include. Fortunately, this is something that the Durham VA Evidence-Based Synthesis Group (ESG) has a lot of experience with. We work with a search librarian to create a comprehensive search strategy to identify all potential articles that are relevant to our topic and questions. Then, the team has to screen every article for relevance. We ultimately identified 17 observational studies that were relevant to our review.

Q: Why was it more relevant to create a systematic literature review for this research topic?

A: It's important to understand the current state of the literature for important topics like this – not only to guide patient care, but to identify gaps in the literature so that we have an idea of what kinds of research questions need to be answered going forward.

During the course of this review it became clear that there were no randomized trials in this domain. So the question of whether we need randomized trials to examine clinical outcomes of metformin in patients with historical contraindications and precautions is important for discussion. Even without randomized trials, additional observational studies will be needed to make sure that no new concerns emerge as metformin use increases in these groups. 

Q: What does your team plan to do next to expand this review’s research?
A: Our group will continue to do systematic reviews that will inform medical practice in other areas. As far as metformin, we've discussed how we might contribute to this literature through new studies using Veterans Affairs data and other databases, which could answer some of the questions we identified during our review. 

This story was written by Tia Mitchell, communications intern for the Department of Medicine.