Antidepressant Makers Withhold Data on Children

By Shankar Vedantam
Washington Post Staff Writer
Thursday, January 29, 2004; Page A01

Makers of popular antidepressants such as Paxil, Zoloft and Effexor have refused to disclose the details of most clinical trials involving depressed children, denying doctors and parents crucial evidence as they weigh fresh fears that such medicines may cause some children to become suicidal.

The companies say the studies are trade secrets. Researchers familiar with the unpublished data said the majority of secret trials show that children taking the medicines did not get any better than children taking dummy pills.

Although the drug industry's practice of suppressing data unfavorable to its products is legal, doctors and advocates say such secrecy distorts the scientific record.

"Conflicts of interest and the company control of the data have thrown out the scientific method," said Vera Hassner Sharav, a critic of the drugs and a patients' rights advocate. "If hundreds of trials don't work out, they don't publish them, they don't talk about them."

"We need a journal of negative findings," agreed Darrel Regier, director of the American Psychiatric Association's division of research, who believes the drugs save children's lives. "The probability of those negative findings being published is far less than the chances of positive studies -- even journals are not interested in negative studies."

Concerns over the safety of antidepressants among children have been heightened after a December warning by British regulators that the drugs may trigger suicidal thoughts and increase the rate of self-injury. An expert advisory panel of the Food and Drug Administration is scheduled to meet Monday to examine the issue, but the agency's full U.S. analysis of the data is not likely to be completed until summer.

One industry executive, Philip Perera, a medical director at GlaxoSmithKline, said that his preference was to publish all trials but that negative studies could lead doctors to prematurely reject a medicine.

"If you start publishing negative data, will it be concluded by practitioners and others that the drug is ineffective?" he asked, saying that genuinely effective medicines sometimes do no better than placebos, or dummy pills, in trials -- at least half of all children seem to get better on placebos.

The U.S. psychiatric establishment largely supports the use of antidepressant medicines in children, with many arguing that abandoning the drugs would lead to more suicides in children with depression. But its critics, including consumer advocates and some psychiatrists , question whether mainstream psychiatry is biased by widespread financial ties to the pharmaceutical industry.

The answer lies hidden in a maze of secret data, conflicting scientific interpretations and a corporate-funded clinical trial system that is not primarily designed to answer questions of public health.

"If the companies wanted to publish negative studies they could, but companies don't like to publish negative studies," said Russell Katz, director of the neuropharmacology division at the FDA , which has access to all the data. "It's amusing so many people are making pronouncements about the data -- scientists and physicians -- . . . without seeing the data."

Advocates say openness about studies is important because, apart from Prozac, no antidepressant has been approved by the FDA for treating children with depression. Doctors writing prescriptions do not have approved labeling to guide them: They must rely on their own judgment and the available scientific knowledge -- even as information is being withheld.

The medicines under scrutiny belong to a class of drugs called selective serotonin reuptake inhibitors, or SSRIs. Led by Prozac, the first to be approved, the medicines caused a revolution in psychiatry.

Recent analyses suggest that as many as 1 percent of children in the United States are treated for depression in any year, said Mark Olfson, a professor of clinical psychiatry at Columbia University. Of those, 57 percent are on antidepressants.

The lack of information is the one issue about which advocates and critics of the medicines agree. Lawrence Diller, a Walnut Creek, Calif., pediatrician and author of "Should I Medicate My Child?," said that "as a front-line doctor dependent on research, it seems so contaminated by potential conflicts of interest. . . . The smoking gun is revelations from the British that negative studies were not published."

Keeping data secret, critics said, has led to conflicting information, contradictory advice and heightened fears.

For example, GlaxoSmithKline, which makes Paxil, has conducted three trials on depressed children. Company officials said all turned out negative -- the children on the drug did not do better than those on placebos -- but only one was published. Based on its data, the company warned British doctors that Paxil, sold there as Seroxat, "should not be prescribed as new therapy" to depressed children younger than 18. Its letter last June cited the risk of increased hostility, agitation, and suicidal thoughts and attempts.

No such warning was issued in the United States, though Paxil is identical to Seroxat. Here, the company's official line on giving Paxil to children is "No recommendations can be made regarding the use of Paxil or Paxil CR in these patients."

"There are differences" between the two recommendations, GlaxoSmithKline's Perera acknowledged in an interview. "They reflect the message that we receive from the respective regulatory bodies." British regulators have essentially prohibited the use of Paxil for children. The FDA is conducting a review of eight drugs, including Paxil.

Perera said the company would await the FDA advisory panel's verdict before considering whether to make all its data public.

Cathryn Clary, vice president for psychiatry and neurology at Pfizer, which makes Zoloft, said it had sponsored two trials in children. One had a negative result, but the company pooled it with a positive study and only published the combined result, which was positive.

"We certainly understand the wish of academics and researchers and physicians to understand all of this data," she said. But small sample sizes in trials "run the risk of magnifying or diminishing a signal. Releasing an individual study can be as misleading as it is helpful."

Graham Emslie, a professor of psychiatry at the University of Texas Southwestern Medical Center, who has helped conduct several trials for drug companies, counted nine recent trials of Prozac, Paxil, Zoloft and Celexa in depressed children. Results of two Prozac trials, one Paxil and the pooled Zoloft data have been published -- meaning that data from five trials, including the stand-alone Zoloft trial that was negative, have not.

Emslie also counted six other studies on the related antidepressants Effexor, Serzone and Remeron -- none of which has been published, he said. Data from several of the unpublished studies have been presented at scientific meetings, and one has been submitted for publication, he said.

Studies reported at conferences are not subject to rigorous advance "peer review" by independent researchers, as are studies published by well-regarded journals. Emslie said he would like to see all the data published but he said the research had been paid for -- and belongs to -- the companies. "They have a legitimate right to do what they want with the data," he said.

But David Healy, a Welsh psychiatrist and author of "The Antidepressant Era," rejected the notion that the safety information could be treated like any other private property. Healy prescribes the medicines but has campaigned for more cautious use and more accurate labeling.

"On a pressing issue like this," he said, "there is no reason these data could not be put into the public domain in their entirety."

The FDA said it is evaluating 20 studies in all, but agency officials have declined to identify them.

In the end, some scientists believe, the only way to ensure that science is conducted in the public interest is for it to be funded with public dollars. The National Institutes of Health is therefore ramping up funding for clinical trials.

"We have been dependent on the pharmaceutical industry to provide the answers," said Thomas R. Insel, director of the National Institute of Mental Health. "The questions they want answered are different than the public health questions."