Statins: The Story Behind the Label Changes
In late February 2012, the US Food and Drug Administration (FDA) issued new labeling changes for the entire statin drug class. First, all statins must now carry a warning noting that there have been reports of increased blood sugar and glycosylated hemoglobin (A1c) levels with statin use. Additionally, labels must include new information on the potential for usually mild and reversible cognitive side effects. Finally, the labels of certain statins as well as protease inhibitors for HIV and hepatitis C virus (HCV) will now warn about interactions between the 2 sets of drugs that could increase the risk for myopathy and kidney failure. The FDA says that it is also eliminating the recommendation that patients on statins undergo routine periodic monitoring of liver enzymes. Medscape interviewed Amy G. Egan, MD, MPH, Deputy Director for Safety in the Division of Metabolism and Endocrinology Products at the FDA, regarding these changes.
Do Statins Cause Diabetes?
Medscape: Recent studies of popular statins showed a significant increase in the risk for diabetes associated with statin therapy. The Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin (JUPITER) trial showed an increase in the incidence of diabetes mellitus in patients taking rosuvastatin compared with placebo. Also, in the Pravastatin or Atorvastatin Evaluation and Infection Therapy -- Thrombolysis in Myocardial Infarction 22 (PROVE IT-TIMI 22) substudy, worsening glycemic control occurred more frequently in the intensive lipid-lowering therapy (atorvastatin) arm than in the moderate lipid-lowering therapy (pravastatin) arm. Were these trials the primary bases for the changes made by the FDA? Did the recent results from the Women's Health Initiative (WHI), which showed a 48% increased risk for diabetes in women taking statins, also play a role in these changes? Can you describe the strengths and limitations of these studies?
Dr. Egan: This labeling change, with regard to an increase in A1c and fasting plasma glucose, was added to the rosuvastatin (Crestor®; AstraZeneca Pharmaceuticals; Wilmington, Delaware) label in February 2010 subsequent to our review of data from the JUPITER trial that showed a 27% increase in investigator-reported diabetes in rosuvastatin-exposed patients vs placebo-exposed patients. We were also aware of the results of the substudy of the PROVE IT-TIMI 22 trial, which reported worsening glycemic control in patients receiving intensive lipid-lowering therapy with atorvastatin relative to moderate lipid-lowering therapy with pravastatin. At the time of our review of the JUPITER trial, there were already meta-analyses appearing in the medical literature that showed an effect of statins on fasting plasma glucose and incident diabetes.[5-12] During the course of our labeling negotiations with the statin sponsors, we became aware of the publication of the results of the WHI study. The WHI study was provocative though there were methodological flaws that limited the interpretability of the findings. While we certainly do acknowledge the results of the study, they did not provide the most robust data leading to the labeling change.
Medscape: Were there differences between statins in the effect on A1c?
Dr. Egan: While we don't have as much data on the most recently approved statin, pitavastatin, we know that pharmacokinetically it is quite similar to rosuvastatin. While there are no large cardiovascular outcome trials for pitavastatin, we did not consider the absence of data to be evidence of an absence of the signal, so pitavastatin was required to make the same labeling change with regard to an effect on A1c and fasting plasma glucose.
We did not require this labeling change for pravastatin. Our reasoning was based on the results of the West of Scotland Coronary Prevention Study (WOSCOPS) which reported a 30% decrease in the incidence of diabetes in pravastatin-exposed patients vs unexposed patients. We found it difficult to consider the data from the meta-analyses as more compelling than data from this clinical trial. You will see from the Website and from the posted labels that all of the statins have the language regarding increases in A1c and fasting plasma glucose with the exception of pravastatin.
Medscape: The labeling change did not indicate that avoidance of these agents in prediabetic persons or newly diagnosed diabetics was warranted.
Dr. Egan: We do know from our review of the JUPITER data that many of the patients who went on to develop investigator-diagnosed diabetes mellitus had impaired glucose tolerance at baseline so they do seem to be more susceptible. However, again, we did not make any specific recommendations as to particular groups because we also see impaired tolerance in patients who do not have impaired fasting glucose at baseline; we see it in nondiabetic patients, prediabetic patients, and diabetic patients. However, keep in mind that we consider these changes, and the resultant risk, to be very small. We do not know whether discontinuation of the statin results in a reduction of fasting plasma glucose back to pretherapy levels because this was not a discontinuation criterion in the trials -- ie, if a patient's fasting plasma glucose went up or that individual was diagnosed with diabetes, statins were not discontinued in that person. That is the same message that we want to convey: We are not recommending that patients be discontinued from their statin therapy based on a small increase in blood sugar levels. Rather, elevations in blood sugar levels should be treated with dietary and lifestyle management and/or adjustment or initiation of antidiabetic therapies. We do not consider this a reason to not continue or not initiate statins, particularly in the diabetic population where patients are at increased risk for major adverse cardiovascular events and statin therapy has been shown to reduce that risk.
Medscape: Do the label changes apply to any specific populations, such as women?
Dr. Egan: No. Despite the higher hazard ratios observed in the WHI study, we do not have strong evidence suggesting that there is a gender effect for the development of this adverse effect.