Depression drug data leave many questions
October 04, 2004
These days, when the parent of a depressed child asks Ellen Sholevar which treatment is best, she doesn't know what to say. And she's a child and adolescent psychiatrist.
"At this point, we don't have science to guide us," said Sholevar, who works at Temple University Hospital. "It puts all of us in a tough situation."
Sixteen years after the approval of Prozac - the first of a new generation of antidepressants - doctors who treat kids still have many questions: Which drug should they use? How long should a child take it? Do the medicines really cause suicides and, if so, who is most at risk? The list goes on and on.
No matter how the Food and Drug Administration responds to a recommendation to issue stronger warnings about the drugs, many of those questions likely will remain unanswered for years.
The crux of the problem, doctors say, is that pharmaceutical companies fund studies aimed at FDA approval for specific uses for their drugs, not at giving doctors the facts they need.
"The big question is, how do we get the information that decision-makers - teens, parents, adults, doctors, policy-makers - need to make decisions about patients. And the database is woefully inadequate for that purpose," said John March, a Duke University child psychiatrist. He led a government-funded study on Prozac and talk therapy in children hailed by many as the kind of work that's needed.
Child psychiatrists hold out little hope that private industry will do what's needed, in part because only about 7 percent of prescriptions for antidepressants go to children and because companies have little to gain, and much to lose, from head-to-head comparisons.
The National Institute of Mental Health (NIMH) did few clinical trials before 1999, but is now funding a handful of studies aimed at measuring real-life functioning in kids with complex emotional problems.
March's Prozac study, a rare, long-term look at kids on the drugs, has already published results showing that in the short run, Prozac combined with talk therapy worked better than either Prozac, talk therapy, or placebo alone. In the future, data from the study will tell doctors more about long-term results.
Another study compares how well three antidepressants work in adolescents, and a third study makes a stab at figuring out how to treat adolescents who have already tried suicide, a group usually excluded from research.
With NIMH support, March is assembling the first large network of doctors to participate in trials of drugs or therapies for mental illness, an approach similar to that used for comparing cancer medicines.
At last month's FDA hearings, discussions centered on an analysis by the FDA and Columbia University of pooled data from antidepressant studies in children. The analysis found that 2 percent to 3 percent of kids on medication were more likely to consider suicide, though there were no suicides in the studies.
Many studies of antidepressants in children, experts said, were done to secure patent extensions, a congressional incentive encouraging drug testing in kids. Companies could get extensions whether the drugs worked or not, and that may have affected study design and execution, doctors said.
The results of the studies aren't clear, and give doctors no clues about which children are most at risk. Answering that question would take very large studies. March said that figuring out, say, whether young girls on antidepressants are at especially high risk could require 10,000 patients.
Psychiatrists say they would like to see studies like the one March did with Prozac in other drugs. Though other drugs work in adults, so far, only Prozac has proven better than a dummy pill in depressed kids.
Because suicide in children can be an impulsive act, doctors want to know more about how often the drugs lead to agitation or trigger manic behavior in kids with previously undiagnosed bipolar disorder - and what to do about it. They want to know if young kids respond differently than adolescents and how to treat kids who also have other problems.
Gail Edelsohn, a child psychiatrist at Thomas Jefferson University Hospital, said drug company trials cover treatment for, at most, 12 weeks. "Then the data stops. In real life we're looking at a longer treatment course and a longer course of having symptoms."
Irwin Lucki, a pharmacologist who studies antidepressants at the University of Pennsylvania, said there are many basic science questions yet to be answered. Is it possible, for example, that a certain amount of brain development is necessary before the newer antidepressants work? And, he said, very little work has been done on how the drugs affect developing organisms.
The biggest need, though, is for better drugs, said Insel. At best, current antidepressants help 60 percent to 70 percent of depression sufferers. That leaves a lot of people who don't get better. "I don't think anybody would settle for a 60 percent response rate in diabetes or heart disease."