A leading US institute has asked for a review of the trial results published in 1995 that led to a change in the American Heart Association's guidelines on use of the thrombolytic agent alteplase. It follows a report in the BMJ earlier this year which showed that the majority of the association's experts on stroke had ties to the manufacturers of the drug.
The National Institute of Neurological Diseases and Stroke (NINDS) has invited an independent committee to re-analyse the data from a five year, multisite trial, the results of which were published in the New England Journal of Medicine on 14 December 1995.
The results were cited by the American Heart Association as the main reason for changing their advice in 2000 on the use of alteplase (tPA, also known as rt-PA) for stroke, moving it up from “optional (class IIb)” to “definitely recommended (class I),” despite continuing controversy over the safety and efficacy of the treatment.
W Michael O'Fallon, professor of biostatistics and chairman of the independent committee, said the review was triggered by an outpouring of letters to the editor of the BMJ after the article in the journal .
The committee has been asked to determine whether “eligible stroke patients may not benefit from rt-PA given according to the protocol used in the [NINDS] trials, and whether any subgroup imbalances invalidate the trial as claimed by some of the critics,” according to the October 2002 issue of NINDS Notes (see www.ninds.nih.gov).
The institute has denied concerns that financial conflicts of interest might explain positive claims during its earlier analysis.
The six member review committee hopes to submit a report to the institute in the next few months. The institute states: “None of the committee members has a connection with the previous published study or with the manufacturer of tPA.”
But some observers are already voicing concerns about the way in which the committee was selected.
Professor O'Fallon was appointed by Dr John Marler, the project leader of the original NINDS trial. He asked Dr Marler for help in finding an emergency physician to sit on the committee and says that Dr Marler gave him a “short list of names.” Dr Marler also “vetted” Professor O'Fallon's selection of two biostatisticians.
“I guess they would also support having the judicial system set up so defendants could pick their own jury,” commented Dr Robert McNamara, past president of the American Academy of Emergency Medicine.
Professor O'Fallon defended the procedure, responding to Dr McNamara's comments with: “I found a very impartial committee that is competent to do the task.”
Dr Marler did not deny that he had given Professor O'Fallon a list of names but said: “I had nothing to do with the selection of the committee members.” Responding to Dr McNamara's comments, he said he had thought about whether the committee might not be seen as independent as he had appointed Professor O'Fallon, but that he “couldn't think of any way to get round that.”
Meanwhile the institute has fiercely denied that it has refused to release the raw data from the trial, although the BMJ requested the data to be released under the Freedom of Information Act last year and still has not received it. The request was sent on 28 June 2001 to the Food and Drug Administration, and the FDA acknowledged receipt on 10 July 2001. The institute says it never received the request from the FDA.
Meanwhile, Professor O'Fallon has been asked to “pledge confidentiality” regarding the data and to assure that he and other committee members “would retain no copies of the data.”
When asked why committee members could not release the data, Professor O'Fallon said, “We have no right to retain [the data]. Whether the raw data will be released is an issue for NINDS—we were not asked to make a recommendation on it.”
However, the institute has told the BMJ that this was because they want researchers requesting the data to be assured that they are getting good data. It also wanted to ensure the institute could control, and vouch for, the raw data.
Dr Michael Meyer, a principal investigator and member of the steering committee of the original NINDS trial, has, for the first time, spoken out to the BMJ about his concerns over the trial.
He says concerns he raised about fatal haemorrhages and outcomes in elderly patients with acute severe strokes “were ignored.” He believes that the raw data for this trial “should be put on a CD and made available.”
He says he was ignored when he tried to raise questions about the safety of alteplase for stroke—particularly in elderly patients with acute, severe strokes.
In response to Dr Meyer's allegations, Dr Marler told the BMJ: “If there was any ‘ignoring’ it was done by the New England Journal of Medicine. They turned [a letter from Dr Meyer] down. The other committee members felt [Dr Meyer's analysis] was just very repetitious of what they said.
“The concept he was trying to convey had already been published in several articles including in the Physician's Desk Reference, where it says to use t-PA with extreme caution in very elderly patients and in patients with a high stroke score,” said Dr Marler.