In Debate Over Antidepressants, FDA Weighed Risk of False Alarm
Doubting Data on Suicide And Kids, Officials Stopped Presentation by Staffer Avoiding a 'Shouting Match'
(May 25, 2004) - Last December, drug-safety analyst Andrew Mosholder delivered a blunt message to his colleagues at the Food and Drug Administration: Young people taking antidepressants were more likely to become suicidal.
That conclusion, based on Dr. Mosholder's analysis of clinical trials comparing youths who took the drugs with those who took a placebo, set off a still-raging debate over whether the benefits of antidepressants outweigh the risks for children and teens. It's a question of keen interest for the pharmaceutical industry and for American families, since the drugs are prescribed to millions of young people.
Dr. Mosholder's bosses criticized his findings as premature and based on unreliable data. They barred him from reporting his conclusion to an FDA advisory committee that met in February. Above all, they didn't want to drive people away from the medications "when we weren't sure that was the right thing to do, and it may be the wrong thing to do," says Robert Temple, associate director for medical policy in the FDA's drug-evaluation center.
The clash illustrates the FDA's challenge in balancing the need for prompt warnings about potential side effects from drugs against the risk of discouraging the use of helpful medicine. Although critics have accused the FDA of a cover-up to protect a lucrative market for drug makers, the events suggest a more complex situation, in which the agency took the initiative in uncovering much of the data but hesitated to make a declaration that was sure to alarm many patients and doctors.
Antidepressants and suicide are particularly tough to analyze because it's hard to tease out whether suicide is a side effect or a consequence of depression itself. Even defining a "suicidal event" -- crucial to figuring out whether an antidepressant induces suicidal behavior -- is subjective.
In a rare glimpse of the FDA's struggle, a series of internal e-mails reviewed by The Wall Street Journal shows how one of Dr. Mosholder's superiors was worried he would condemn the drugs too strongly before the advisory committee. Unlike most FDA deliberations, the committee meetings are public. The superior, Anne Trontell, deputy director of the drug-safety office, sent Dr. Mosholder a list of questions for which she wanted him to prepare and advised him "to acknowledge the limitations of your analyses." She also offered to "provide guidance on how to finesse the likelihood of a shouting match at the table."
The FDA's handling of Dr. Mosholder's conclusions is drawing fire from Capitol Hill, and committees in both the House and Senate are investigating. "You don't just ask someone to clam up," says Iowa Republican Charles Grassley, chairman of the Senate Finance Committee. "If there's any doubt, they ought to put out the caution to the public at large, not try to muzzle it."
Mr. Grassley says his investigators are looking at other possible instances where FDA drug safety officials weren't allowed to make their conclusions public. The investigators are interviewing Dr. Mosholder, who has hired a lawyer.
Dr. Mosholder declined requests for an interview. The FDA also didn't make any officials from its drug-safety office, including Dr. Trontell, available.
Antidepressant prescriptions for young people are up sharply even though few studies have been published for most of the drugs demonstrating that they work in that population. According to an analysis by Washington State University researchers, the rate of antidepressant prescriptions for children and adolescents more than tripled in the U.S. from the early 1990s to 2001. In 2002, an estimated 10.8 million prescriptions for the most widely used antidepressants were dispensed for patients under 18 years old, according to an FDA analysis. Some psychiatrists credit the drugs for the sharp decline in the suicide rate among young people since 1994.
At the FDA, the issue of antidepressants and young people surfaced in the fall of 2002. That's when Dr. Mosholder noticed that in some studies, patients taking the antidepressant Paxil experienced more incidents of "emotional lability" -- a category that includes suicide attempts and suicidal thoughts -- than did those taking placebo pills. The FDA division that reviews antidepressants asked Paxil's maker, GlaxoSmithKline PLC, for more information. When it arrived, Dr. Mosholder was assigned to analyze it.
It was an assignment for which he was well prepared. A 46-year-old psychiatrist with experience treating children and adolescents, Dr. Mosholder joined the FDA in 1992 and spent about a decade reviewing antidepressants and other drugs that companies were seeking approval to market. One former official who supervised Dr. Mosholder calls him a careful and diligent reviewer. Later Dr. Mosholder moved to the drug-safety office, which often deals with safety problems after drugs go on the market.
In June 2003, British authorities, who had been conducting their own review, announced that Paxil shouldn't be used to treat depression in children under 18. The FDA issued a similar recommendation shortly afterward.
Then the stakes grew higher. In late July, the FDA asked manufacturers for information about pediatric studies of eight other depression drugs: Prozac, Zoloft, Luvox, Celexa, Wellbutrin, Effexor, Serzone and Remeron. Several of those fall into the category of selective serotonin reuptake inhibitors, or SSRIs. Although only Prozac is approved by the FDA specifically for pediatric use in depression, doctors can legally prescribe any of the drugs to patients of all ages.
Many of the companies had done studies of their drugs on young people so they could get an extra six months to market their drugs without competition from generic copycats. U.S. law provides for the extra exclusive period to encourage companies to conduct pediatric studies. The FDA asked the drug makers to search their pediatric study databases for key words such as "attempt; cut; gas; hang" or partial words such as "suic" or "overdos."
In October, the FDA told doctors it couldn't "rule out an increased risk" from antidepressants and announced it would hold an advisory committee meeting of outside experts the following Feb. 2 to discuss the issue. Usually the agency follows its committees' advice, in part because the committees are themselves swayed by the data that FDA staff experts present.
But as the drug-company data trickled in, senior FDA officials became concerned that the studies were too unreliable to produce a clear conclusion. The skeptics came to include two influential officials: Dr. Temple, a 32-year agency veteran who is considered perhaps the agency's leading authority on clinical trials, and Thomas Laughren, who has led the agency's psychiatric drug review team for 17 years and oversaw the approval of many of the antidepressants at issue.
The agency was reviewing incidents classified as "possibly suicide related," since no child in any study actually committed suicide. But the skeptics thought some of those incidents -- for example, a child "who hit her head with her hand" and others who engaged in "superficial cutting behavior" -- might not be suicidal behavior at all. They also wondered, conversely, whether certain injuries classified by the drug companies as "accidental" should have been included. One example: A child who "stabbed himself in the neck with a pencil while taking a test."
Another confounding factor: Sometimes one study of a drug would show a big suicide risk but another study of the same drug would show no risk.
To try to get a clearer picture, the FDA asked drug makers for more details on individual patients. And it commissioned Columbia University researchers to lead a review to sort out which incidents were truly suicidal.
On Dec. 10, British regulators dropped a bombshell: They urged doctors not to prescribe any of the antidepressants in the SSRI class except Prozac to depressed patients younger than 18.
About a week later, as Dr. Laughren remembers it, FDA scientists met to discuss the agenda for the February advisory committee meeting. Using a one-page outline, Dr. Mosholder declared his conclusion: Young people who took antidepressants were far more likely to show suicidal tendencies than those who took placebo pills -- 3.2% vs. 1.7%, according to data from more than 4,000 trial subjects that Dr. Mosholder would later compile in a report. Dr. Mosholder told his colleagues he wanted to present his conclusion to the advisory committee and urge the advisers to make a recommendation for dealing with the risk.
Dr. Laughren says he was startled at his colleague's determination. What about the problems with the classification of the incidents? he asked Dr. Mosholder. Or the fact that different studies showed different results, even for the same drug? Dr. Mosholder didn't budge, Dr. Laughren recalls: He was "quite firm in his views on it."
Dr. Mosholder felt the data were so clear that his conclusion was unlikely to be altered by the Columbia review, as he wrote in subsequent e-mails and a report. Even if some incidents were inaccurately classified, he believed the mistakes ought to affect both the drug group and the placebo group equally, leaving the bottom line unchanged.
Dr. Temple's view was, "Garbage in, garbage out," he says in an interview. If the data were unreliable, so was the conclusion. He thought some of the supposed "suicide-related" events might merely show that kids taking antidepressants had more energy. "Some of the things that have been represented as suicide-related were self-evidently not that," he says. "There were enough of those so we weren't sure how it would add up."
On Dec. 23, high-ranking FDA officials, including Dr. Laughren and Dr. Temple, had a meeting that Dr. Mosholder didn't attend. They decided that he wouldn't be permitted to speak about his findings to the advisory committee. Instead, Dr. Laughren would describe the data that Dr. Mosholder had analyzed, without offering any conclusion.
Drs. Temple and Laughren had a good sense of how high-profile and emotional the meeting would likely be. Both had attended a similar committee meeting in 1991, when the FDA was weighing warnings about antidepressants and suicidal tendencies in adult patients. That one had been sparked partly by a study by a Harvard psychiatrist and others, which argued that six patients had become suicidal because of Eli Lilly & Co.'s Prozac. The ensuing publicity helped provoke a flood of similar reports to the FDA. But the panel, after hearing both sides, voted against a warning label for Prozac.
This time, the FDA would carefully control its message. Dr. Mosholder was told that he couldn't present his data, although he was attending the committee meeting to make another presentation. In case the advisers sought his views at the meeting, Dr. Trontell, one of his bosses, sent him a list of questions for which she wanted him to prepare. She asked to see his planned answers.
Dr. Mosholder offered a lengthy response in an e-mail dated Jan. 28, 2004. It said that despite the data's shortcomings, his findings were "difficult to dismiss." In a revised version dated the next day, he added some caveats, saying "obviously, one can't predict" whether the new examinations of the data would change his results.
When the advisory committee met Feb. 2, Russell Katz, the head of the FDA division that reviews psychiatric drugs, opened by saying the agency had not yet been able to "adequately and comprehensively address the critical question before us." He noted briefly that "some within the agency" thought the data were sufficient for a "definitive analysis ... that this class of drugs is associated with a risk of suicidal behavior in pediatric patients."
Dr. Laughren discussed the clinical-trial data that Dr. Mosholder had originally analyzed, but he stressed the unreliability of the data instead of the possible risk from the drugs. Dr. Mosholder was never asked about his analysis of the clinical trials.
Punctuating the meeting was emotional testimony from relatives of people who had committed suicide while taking antidepressants. One woman showed a series of slides of her daughter, including one at her Stanford University graduation. Two weeks after beginning to take Paxil, the young woman went to the kitchen late at night and stabbed herself with a large chef's knife: "The only noise was a slight yelp and a thump when she fell on the floor," her mother said.
At the end, the committee voted that the FDA should continue its analysis and issue a public warning in the interim. On March 22, the FDA called on makers of 10 antidepressant drugs to add to the labels a warning that patients should be watched for "worsening depression or the emergence of suicidality."
Dr. Mosholder issued a confidential report to Dr. Katz of the FDA review division Feb. 18, which was later leaked to the press, calling on the FDA to follow the lead of British regulators and discourage the use of antidepressants except Prozac for depressed children and adolescents. He noted that many young people would be taking the drugs during the months that the FDA was re-examining the data.
FDA officials now are awaiting the results of the Columbia research, due in June. And later this summer, the FDA's advisory committee on antidepressants will meet again to take a look at the updated information.
Source: Wall Street Journal
Comprehensive information on children and teens with depression can be found in the HealthyPlace.com Depression Center.
HealthyPlace.com Depression Center Links
home ~ site map ~ causes ~ types ~ people ~ living with
treatments ~ self-help ~ support ~ suicide ~ related issues
We subscribe to the HONcode principles of the Health On the Net Foundation.