Washington Post (April 23, 2004)

Antidepressants Called Unsafe For Children.

4 Medications Singled Out In Analysis of Many Studies.

By Shankar Vedantam

Four popular antidepressants being used to treat thousands of depressed American children are unsafe, ineffective or both, according to the first comprehensive scientific review to include all available studies, including negative data that have long been withheld from public scrutiny by the pharmaceutical industry.

It is especially dangerous to prescribe Paxil, Zoloft, Effexor and Celexa for children who are suicidal, said British researchers who conducted the analysis published yesterday in the journal the Lancet, because the data show a clear increase in the risk of suicidal behavior among children taking the drugs -- and no benefit.

The study calls into question the repeated assurances of the American psychiatric establishment, which has regularly encouraged use of the medications in depressed children. It also contrasts sharply with the position of the U.S. Food and Drug Administration, which has had access to the same data but has never identified such serious problems.

The analysis involved no new data, but it is the first scientific "meta-analysis" of all available trials to be published in a peer-reviewed journal. As such, it puts the scientific debate over the medications on a new footing and deepens the chasm between the predominantly positive American view of the drugs and a growing stream of negative reviews by Australian, Canadian and British psychiatrists. The Lancet analysis backs up the warning by British regulators last year not to prescribe the medications to depressed children.

In a sharply worded editorial, the Lancet's editors said the trust of patients had been abused by doctors and the pharmaceutical industry, and that safety had been compromised in the search for profits. The state of the research, the editorial concluded, is riddled with "confusion, manipulation and institutional failure."

"If I wanted to introduce a new drug for children who are suicidal and said this has very little proof of efficacy and it has an increased risk of suicide, people would say I was mad," said Tim Kendall, director of the National Collaborating Centre for Mental Health in London, and one of the authors of the new study.

Yet, Kendall said, that is precisely the situation with the four drugs. Kendall's analysis, which was funded by the British government, contrasted the largely positive results of studies published by drug manufacturers with negative data that the companies call proprietary and have not published. British regulators recently made the unpublished data available for study.

"In each of the published articles, the authors concluded the drug was either effective or safe or both," Kendall said. "When you look at the combined evidence, it is ineffective, unsafe or both."

The review found a positive risk-benefit ratio for only one medicine -- Prozac, which the researchers said did not carry an elevated risk of suicide and had been proved to benefit depressed children. The article said that while the other antidepressants should not be given to children with depression, it is possible that children with anxiety or other disorders might benefit from them. The number of American children being treated for depression in any year has surged in recent years; a majority are being treated with antidepressants.

The FDA said it was continuing to evaluate whether the antidepressant trials showed an increased risk of suicidal behavior. Although an internal FDA analysis found an increased risk of suicidal behavior among children taking the drugs, senior officials at the agency have said they do not have confidence in the conclusion. The matter has been referred to a team at Columbia University for further evaluation.

"Some of the things called suicidal thinking or behavior are not," said Robert Temple, FDA's associate director of medical policy, in a recent interview. "There is one child banging her head against the wall -- that is not suicide. There are quite a few children who did superficial cutting. That doesn't show intent to kill oneself."

FDA officials have consistently said the large number of failed antidepressant trials do not prove that the drugs are worthless, a position in conflict with the Lancet study.

Mariann Caprino, a spokeswoman for Pfizer Inc., which makes Zoloft, categorically rejected the charge that the company had compromised the safety of children in the pursuit of profits.

"The use in children of our drug in particular, is a very, very, very small percentage of the overall total prescriptions of this product," she said in an interview yesterday. "To suggest that we are motivated by profiting off of children is ludicrous."

While Caprino said all the company's data had been disclosed, the Lancet article said a published study that combined two Zoloft trials omitted information that, when factored in, showed an "unfavorable risk-benefit balance."

Wayne D. Blackmon, a Washington psychiatrist who has long said clinicians cannot rely on the integrity of the data they are being given, said Congress should force the FDA to take unpublished negative trials into account and force the companies to make all data -- positive and negative -- available for public scrutiny.

In the meantime, he said, clinicians should go back to the Hippocratic oath -- "First, do no harm" -- and "recognize that you are flying by the seat of your pants."