Generic Fares Well
In Big Psychiatry Study

In a surprising finding, a government study comparing schizophrenia treatments found that an older generic medicine was as effective as all but one of the newer and more-expensive brandname drugs widely used to treat the mental illness.

Weight gain was a side effect with brand name Zyprexa.

The $67 million federally funded study also exposed just how poorly current antipsychotic drugs really work: Nearly three-quarters of people treated stopped taking the medicine they had been given within 18 months, due to side effects or poor control of symptoms. The results, from the experience of 1,500 patients, are to be published in this week's New England Journal of Medicine.

The findings may have significant implications for how doctors treat the 3.2 million people in this country suffering from schizophrenia. The newer, costlier antipsychotics make up 90% of the market today. They are also used for bipolar disorder, and for severe cases of depression, kids with extreme behavioral problems, and dementia.

But psychiatrists have been prescribing these drugs with incomplete information about the benefits and risks of each drug. The studies that companies do to get drugs approved aren't designed to compare available treatments or shed light on the differences between similar drugs. When companies do compare their drugs to others, the studies have often been subject to criticism that they were designed in favor of a particular drug. So psychiatrists have had no head-to-head, impartial comparison to help them weigh treatments.

Now this trial, part of a six-year push by the National Institutes of Health to examine a range of psychiatric drugs, "provides a comprehensive set of data that were obtained independently of the pharmaceutical industry and in a scientifically rigorous way," says Jeffrey Lieberman, head of psychiatry at Columbia University and principal investigator on the trial.

Schizophrenia patients in the study were randomly assigned one of five drugs. The older antipsychotic perphenazine was found to be just as effective as three newer, so-called atypical antipsychotics: Johnson & Johnson's Risperdal, AstraZeneca PLC's Seroquel and Pfizer Inc.'s Geodon, although each had slightly different side-effect profiles.

Eli Lilly & Co.'s Zyprexa was more effective than the other drugs. But the trial confirmed a concern that has emerged in recent years: The new antipsychotics can cause extreme weight gain and lead to heart disease and diabetes, and Zyprexa causes more of these side effects than the other medicines. Zyprexa's manufacturer Eli Lilly had long argued that all the newer antipsychotics caused these problems and the FDA eventually mandated a warning for the whole class.

It remains to be seen whether the findings will lead psychiatrists to change their prescribing habits. One thing to watch is whether public programs like Medicaid or private insurers use the findings to justify trying older generic medicines before the new ones. The researchers plan to analyze the cost-effectiveness of the various treatments, using data on the rates of hospitalizations, doctors' visits and drug costs.

Although the older drug, perphenazine, worked just as well as several of the newer drugs, some doctors may resist prescribing the older drugs because of long-held fears about side effects: In some people they cause involuntary movements, jerkiness and tremors. In the 1990s, drug makers came out with the new generation of drugs, known as "atypical" antipsychotics, which supposedly caused fewer neurological side effects.

Backed by huge marketing efforts, use of these newer antipsychotics has exploded, reaching $10.1 billion in U.S. sales last year, according to IMS Health. But some psychiatrists and researchers have been critical of how drug companies developed and promoted the medications.

Studies done by drug companies are too small and short-lived to pick up long-term safety problems, and they often exclude the sickest patients and people who have other diseases in addition to the illness being treated. Moreover, the companies' studies have generally compared the new drugs to high doses of Haldol, a potent older antipsychotic available as a generic that is known to cause relatively high rates of movement side effects.

The new trial aimed to eliminate some of the guesswork in treatment. In the first stage of the study, if patients did well on the drug they were assigned, they stayed on it for the 18-month-long trial. But if the patient felt the drug wasn't working or experienced bad side effects, they were switched to another antipsychotic.

The primary measure of the drugs' effectiveness was how long patients stayed on them. The researchers chose this somewhat unusual trial design to reflect patients' and doctors' overall judgments on whether the benefits were worth any undesirable effects. Most psychiatric clinical trials measure a drug's effectiveness based on whether it relieves symptoms as measured by questionnaires and rating scales.

Patients on Zyprexa stayed on the drug for longest, for a median of 9.2 months. Patients on perphenazine, the older drug, stayed on for 5.6 months, Risperdal for 4.8 months, Seroquel for 4.6 months, and Geodon for 3.5 months. Nearly a quarter of people who stopped taking Seroquel, Risperdal, perphenazine and Geodon stopped because the drug wasn't working.

The study found no significant difference among the drugs in the incidence of neurological side effects like shaking. (However, the patients who discontinued perphenazine because of side effects were more likely to do so because of movement side effects.) The finding is notable because it undercuts the prevailing view shaped by drug-company marketing that the newer drugs cause fewer movement side effects. The researchers acknowledged that the finding may not represent the whole picture since the most serious of side effects can take years to emerge.

Patients on Zyprexa gained an average of two pounds per month. One-third of patients on Zyprexa gained more then 7% of their initial body weight compared with 16% of patients taking Seroquel, 14% taking Risperdal, 12% taking perphenazine, and 7% taking Geodon. Patients who discontinued Zyprexa because of side effects were more likely to do so because of weight gain.

The drug makers defended their drugs. Pfizer said Geodon, which has long had small market share because of nagging cardiac safety concerns, performed well and without weight gain or metabolic side effects. Lilly said the results proved that Zyprexa was superior to the other drugs, while Johnson & Johnson said the study didn't adequately reflect Risperdal's strengths because the doses given were too low. AstraZeneca said that Seroquel balanced efficacy and tolerability.

"This is one of the first studies that clearly says we ought to rethink our use of the first generation antipsychotics," said Robert Conley, a professor of psychiatry and pharmacy science at the University of Maryland, Baltimore, who isn't affiliated with the study. "There are safe ways to use them."

Write to Leila Abboud at leila.abboud@wsj.com¹