In response to the study by Dr. Nissen and his colleagues in the New England Journal of Medicine
which we linked to last week) GlaxoSmithKline issued a letter in the Lancet defending the safety of Avandia
based what they believe is the best available data. The Lancet has also issued its own editorial on the issue
The Wall Street Journal reports on the growing dispute:
Ronald Krall, [GSK] chief medical officer, wrote in the letter that Glaxo did a "meta analysis" similar to the one conducted by Cleveland Clinic cardiologist Steven Nissen, whose article in the New England Journal of Medicine last week linked the drug to a potentially increased risk of heart attacks. Glaxo's own metaanalysis also found indications of increased risk, Dr. Krall wrote, but he said the number of adverse events was low.
Dr. Krall also discussed results from two large Glaxofunded studies of the drug. Neither trial, called Dream and Adopt, was designed primarily to assess the drug's heart risks. But Glaxo's analysis of the Adopt trial showed major adverse cardiovascular events, such as heart attacks and strokes, were "rare," and heart risks were similar to those of two other diabetes drugs. The Dream trial also showed "no significant difference" in cardiovascular events between the drug and placebo, the letter said.
Meanwhile, an independent safety board recently reviewed an interim analysis of an ongoing study, Record, designed specifically to assess the drug's impact on the heart. The board determined the trial should continue, the Glaxo letter said.
In an interview, Christopher Viehbacher, Glaxo's president of U.S. pharmaceuticals, said the trial's interim results were "giving us the confidence to say that we stand behind this product." He said there is a chance the interim results will become public before the Record trial is complete, though the tradeoff would be that the publicity could weaken the statistical power of the final results.
Mr. Viehbacher said it wouldn't surprise the company if the Food and Drug Administration were to call for "some labeling changes" for Avandia. One possibility would be elevating a heartfailure warning on the label to a more severe "black box" warning, he said.
Dr. Nissen criticized Glaxo's letter to the Lancet, saying the company was slicing the data differently from the Adopt and Dream results originally published. Dr. Nissen also said the company was referring to such small subsets of data in the Adopt and Dream trials, that no firm conclusion could be drawn.
"Somebody went back and looked for something that would support their contention," Dr. Nissen said. "This is not a scientifically proper way to analyze data."
It is ironic, however, for Dr. Nissen to make this charge given the admitted limitations of his own study on Avandia's safety. Experts agree that metaanalyses of aggregated clinical trial databy GSK or Dr. Nissenare problematic, and that the best way to adjudicate safety issues is with a prospective, randomized trial, which GSK is currently conducting.