"The data presented by Juurlink do not support the suggestion that you should favor one drug over the other," de Vries said. "There is no reason to believe yet that there is a difference."
People who have a preexisting heart condition should not be started on either of these drugs, de Vries stressed. Patients who develop heart failure should have the drug discontinued, because the condition is usually reversible, she added.
Another expert, Dr. Carl J. Lavie, medical director for cardiac rehabilitation and prevention director of the Stress Testing Laboratory at the Ochsner Heart and Vascular Institute in New Orleans, thinks the study gives more reasons not to use Avandia.
"I believe that most clinicians have stopped using Avandia -- some will use Actos instead or go to another class completely," Lavie said.
Lavie noted that an ongoing study assessing Avandia and Actos in a head-to-head comparison should provide more answers about these drugs.
"Until then, this current study provides further ammunition against using Avandia or at least for trying Actos first, which is along the same lines as the current American Diabetes Association and European guidelines anyway," Lavie said. "Most will, and should, avoid either agent in a patient with heart failure or high heart failure risk."
"Both drugs have a place in treating diabetes, but not if you have a preexisting heart condition," de Vries said.
For more information on diabetes, visit the U.S. National Institute of Diabetes and Digestive and Kidney Diseases.SOURCES: David Juurlink, M.D., Ph.D., Division Head, Clinical Pharmacology and Toxicology, Sunnybrook Health Sciences Centre, Toronto; Steven E. Nissen, M.D., chairman, Department of Cardiovascular Medicine, Cleveland Clinic, Ohio; Corinne de Vries, Ph.D., professor, pharmacoepidemiology, Department of Pharmacy and Pharmacology, University of Bath, England; Carl J. Lavie, M.D., medical director, cardiac rehabilitation, and prevention director, Stress Testing Laboratory, Ochsner Heart and Vascular Institute, New Orleans; Aug. 19, 2009, BMJ, online