Thursday, February 14, 2008

"A Catastrophic Failure" - More on the Ketek Trial Debacle

We previously posted (also here, here and here) about the ill-fated clinical trial, study 3014, of the antibiotic telithromycin (Ketek) made by Sanofi-Aventis, run by Pharmaceutical Product Development Inc. (PPD). Problems with the trial included fabrication of data at one clinical site, and allegations of manipulation of data at another. The physician in charge of the first site was convicted of mail fraud, and the physician in charge of the second had his license suspended. Although the results of this trial were never published, it still crept into the clinical literature: it was cited in a review article in the New England Journal of Medicine.

The case is now the subject of a US congressional investigation. As reported by the Associated Press (via the International Herald Tribune),

Democrats in the U.S. House of Representatives are threatening to hold a member of President George W. Bush's Cabinet in contempt for not turning over documents in a probe of a Sanofi-Aventis antibiotic.

Dingell, who chairs the House Energy and Commerce Committee, said he would support holding Health and Human Services Secretary Michael Leavitt in contempt for refusing to turn over FDA briefing documents subpoenaed by the committee.

The documents were used to prepare FDA Commissioner Andrew von Eschenbach for his appearance before lawmakers last year. Von Eschenbach testified the FDA did not use the flawed safety study to approve Ketek. Dingell and other Democrats say that statement may be untrue.

Robert West, an FDA agent who first investigated Ketek, said he tried to get permission in 2002 to look into whether Aventis was aware of fraudulent data when it submitted the study. West said his request was blocked by senior FDA officials, although he said he did not know which ones.

"A catastrophic failure" was how another FDA investigator, Douglas Loveland, described the company's handling of the study.

"The decision-making process Aventis used to investigate these problems was illogical and ineffective and it could have led them to come to the wrong conclusion," Loveland told House lawmakers.

The investigator stopped short of saying Aventis, Sanofi-Aventis' corporate predecessor, knew the results were false when they submitted them but said it "should have known." Loveland said the results contained all the hallmarks of sham data, including forged signatures and crossed out information.

Sanofi President for Research and Development Paul Chew testified that the company submitted the study in "good faith."

The company was not able to spot the fake patient results....

It seems clear that this case involved a very sloppy clinical trial sponsored and run by a pharmaceutical company. Such sloppy science breaks promises made to human subjects of clinical research that their participation would go to advance science and improve health care. Sloppy trials that produce dubious data will do neither. Furthermore, sloppy science that produces dubious results provides patients and doctors pseudoevidence rather than the best possible evidence relevant to decisions about health care interventions, and hence may also lead to bad decisions that harm patients.

It is bad enough that clinical trials sponsored by pharmaceutical, biotechnology, and device companies are often sloppy, and designed to show their products in the most favorable light, rather than provide valid and unbiased data testing hypotheses most important for patients' and physicians' decisions. This case now raises the more disturbing possibility that high government officials entrusted with protecting the public's health and safety may have hid the sloppiness and unreliability of the Ketek trial. Time and further investigation will tell if this possibility is more than speculation. But the latest events in the Ketek saga clearly show the need for more open and transparent clinical research on drugs and devices, and more open and transparent regulation of such companies by the government.

No comments: