Polemic intensifies as Merck & Co admits Vioxx study error
Last Updated: 2006-05-31 14:42:22 -0400 (Reuters Health)
By Nick Hudson
LONDON (Agence de Presse Medicale for Reuters Health) - Merck & Co Inc has admitted it made a mistake in the description of how it analysed key data in a study that led to the withdrawal of Vioxx in 2004 following cases of heart attacks and strokes.
The New York Times said the admission could undermine one of the company's core defences in the 11,500-strong Vioxx-related lawsuits.
Merck & Co Inc said in a statement issued late on Tuesday that it will hire an independent statistician to review the data.
The New York Times said outside scientists found that Merck's admission, when considered along with other clinical trials of the drug and studies tracking real-world Vioxx use, supported critics' long-standing claims that Vioxx could cause heart problems quickly.
18-MONTH RULE
"There never was any evidence for the 18-month story," said Dr. Alastair J. J. Wood, a drug safety expert at Vanderbilt University.
In defending the lawsuits, Merck has consistently taken the position that Vioxx can cause heart problems only if it is used continuously for more than 18 months.
But in the five Vioxx court cases in which juries have reached decisions so far, Merck has lost three times - twice in cases in which the heart attack victim had taken the drug for fewer than 18 months.
Merck has based the 18-month theory largely on data from the APPROVe study, in which the company tracked 2,600 patients in a test to see whether Vioxx could prevent colon polyps, the New York Times reported.
In the APPROVe study, twice as many patients taking Vioxx suffered heart attacks or strokes as those taking placebo. But most of the difference appeared to occur after patients had taken Vioxx for 18 months in a row, Merck has said, according to the newspaper.
When it reported the APPROVe results in The New England Journal of Medicine early last year, Merck said it examined if Vioxx's risk changed over time. They found that the drug had a significantly higher risk than placebo only after the 18-month benchmark -- but no increased risk before that time.
Merck said in Tuesday's statement that it made a mistake in reporting this last year.
In reality, there was a 7% chance that Vioxx has the same risk of causing heart attacks before and after the 18-month benchmark is reached.
That 7% figure may seem like a relatively small chance of error, but scientists say it is high enough to mean that Merck did not proved its theory, the paper reported.
Critics recently raised new questions about the APPROVe trial's results, after the company submitted new data to the Food and Drug Administration as part of a one-year follow-up study of patients in the trial.
Some outside scientists said that the new data indicated that Vioxx's dangers became evident after only about four months of use, the paper added. Merck disputed that interpretation and said that its newly discovered mistake was unrelated to that new data.
INSISTS ERROR DOES NOT AFFECT CONCLUSION
Dr. Peter Kim, Merck's chief scientist, said that Merck found its mistake only last week and immediately told the FDA and the non-Merck researchers who had taken part in the APPROVe study, which was published in the March 17, 2005, issue of The New England Journal of Medicine.
"We came forth quickly to let the study authors know about the error (and) let the regulatory agencies know about the error," Kim said.
"The correction being made here is a correction of a description of a method," Kim told the Chicago Tribune.
"The data itself has not changed. We believe it does not affect the results or conclusion of the APPROVe study."
CRITICS NOT SO SURE
But Dr. Steven Nissen, the interim chairman of cardiovascular medicine at the Cleveland Clinic, and a frequent critic of Merck, said the mistake the company disclosed on Tuesday was yet another example of Merck's mishandling data to make Vioxx seem safer.
"They're acknowledging that they misrepresented the APPROVe data, when they reported that there was a statistically significant difference between the first and the second 18 months," Nissen said.
"There is no biologically plausible reason to expect an 18-month delay. I never thought it made any sense."
Dr. John Baron, a professor at the Dartmouth medical school and one of the authors of the original journal article about APPROVe, said Merck's mistake was relatively minor. In general, trying to calculate whether risks change over time is very difficult, he added.
"The assessment of this constancy is something of an art. It's not a 'cut and dry, it's there or it's not' business," he said.
"The evidence that the Vioxx effect begins immediately or quickly, or however you want to describe it, is far from conclusive. The evidence that it doesn't begin until 18 months is also not something that has been proven."
