Schizophrenia drugs work poorly, study suggests

The biggest study ever of drugs for schizophrenia should help psychiatrists get away from trial and error in treating their patients with the debilitating mental illness, its authors said yesterday. But the $44 million federally funded research project also reached a much darker conclusion: None of the drugs works very well.

Nearly three-quarters of the patients in the study stopped taking their assigned schizophrenia drugs before the 18-month study was finished, mainly because the medications didn't improve their condition enough or they couldn't stand side effects ranging from weight gain to nervous tics. Nearly two-thirds of those who took the most effective drug, Zyprexa, nonetheless switched to a different medication during the study, in part because it caused patients to put on an average of 9 pounds -- the most of the five drugs in the study.

The study, known by its acronym, CATIE, also raised doubts about whether new drugs that now dominate treatment of schizophrenia are significantly better than older, less expensive medicines. To the researchers' surprise, a little prescribed older drug called Trilafon worked as well as all the newer drugs except Zyprexa, at a fraction of the cost and with no more side effects.

''We have a lot more to do here," Dr. Thomas R. Insel, director of the National Institute of Mental Health, which funded CATIE, said in a telephone news conference. ''These results should cry out for the need for us to develop a whole new generation of drugs to treat schizophrenia."

CATIE, which stands for Clinical Antipsychotic Trials of Intervention Effectiveness, is likely to have an immediate effect on the treatment of schizophrenia, a disease characterized by delusions and disordered thinking that afflicts 3.2 million Americans. Even with treatment, 85 percent of schizophrenics are unemployed, most never marry, and 10 percent of them take their own lives, Insel said. Until now, psychiatrists have relied on anecdotes and intuition to find a drug that both controls the disease and that the patient will continue taking.

Psychiatrists not connected to the study praised CATIE for giving them detailed head-to-head comparisons of the growing number of treatments, allowing them to tailor drugs to individual patients. For example, if a patient's family has a history of diabetes, a psychiatrist might prescribe Geodon, which causes the smallest increase in blood sugar.

''While we're searching for new ways to treat people, we'd better have the best available evidence on the drugs that are available," said Dr. Bruce Cohen, president of McLean Hospital in Belmont. ''This takes us another step along the way."

Fifty years ago, many people with schizophrenia were sent to state mental hospitals, but the development of antipsychotic drugs such as Haldol controlled the delusions and sometimes violent behavior enough to allow them into the community. Unfortunately, the drugs often carried a heavy price, causing tremors and muscle rigidness like Parkinson's disease, or constant involuntary movements such as chewing or writhing motions, a condition called tardive dyskenisia. As a result, psychiatrists struggled to keep patients on their medications.

Since 1993, six new drugs, called atypical antipsychotics, have swept aside the older drugs with the promise of fewer neurological side effects and potentially better control of the disease. The atypical antipsychotics, which include Zyprexa and Seroquel, are now the fourth-largest-selling class of prescription medications, bringing in a projected $10 billion in sales this year. But the available studies of these drugs generally looked at short-term effectiveness and did not compare one new treatment against another.

So the National Institute of Mental Health launched the ambitious CATIE study, pitting the new drugs against an older drug in the treatment of 1,493 chronic schizophrenia patients at 57 sites around the country. Each patient received one of five antipsychotic drugs -- all identical in packaging and appearance -- and researchers tracked their progress over 18 months.

After 18 months, 36 percent of the Zyprexa patients were still taking their drug compared with 26 percent of those on Risperdal, 25 percent of those on the older drug, Trilafon, 21 percent of patients taking Geodon, and 18 percent of the patients on Seroquel. By that measure, lead CATIE investigator Jeffrey Lieberman of Columbia University Medical Center in New York City, said Zyprexa was the most effective drug.

However, CATIE also revealed important differences between the drugs that could help doctors make treatment decisions. For instance, patients taking Risperdal and Trilafon saw the smallest increases in their cholesterol levels, which might make them the best choice for patients with a history of heart disease. On the other hand, Zyprexa patients had the fewest hospitalizations, suggesting the drug was most successful at controlling the most severe schizophrenia cases.

''This wasn't a horse race," said Dr. Robert Rosenheck, one of the CATIE investigators from Yale University Medical School. ''No one drug was hands-down superior to all the others. . . . You need to look at the situation of your patient and figure out which drug works best."

Scott Allen can be reached at allen@globe.com.