Donald F. Klein of Columbia University remembers handing out the first pills for schizophrenia in 1955, when the only treatment was electroshock therapy. One patient, an institutionalized war veteran who had not spoken in decades, swallowed the medication and a few days later suddenly asked, “When am I getting out of this place?”
According to a new study, this early class of drugs may work just as well at reducing hallucinations and delusions as modern pills that cost up to 10 times more. In the early 1990s the new medications, such as Zyprexa and Seroquel, took over, promising greater effectiveness and less debilitating side effects, including tremors. The federal government sponsored the $43-million study as a check against the results of previous tests that had been sponsored by drug companies and as a way to evaluate Medicaid’s payout for high-priced schizophrenia medication, one of the program’s biggest expenses.
Overseen by Columbia’s director of psychiatry, Jeffrey A. Lieberman, the study randomly assigned one of five drugs—four from the new class, one from the old—to 1,460 chronically ill patients and followed them for 18 months. Lieberman’s team found that all five compounds were about equally effective. And at moderate doses, at least, adverse reactions from the old medication, Trilafon, were no worse than for the others.
Perhaps the biggest surprise, however, was that none of the pharmaceuticals satisfied patients. Almost 75 percent of the participants quit treatment, were taken out by their doctors or asked to be switched to another pill, because symptoms were not improving or side effects were intolerable. “We clearly need better drugs,” Lieberman says.
Although the old medications can cause tremors, stiffness and jerky muscle movements, the new medicines bring problems, too: weight gain and increased blood glucose, cholesterol and triglyceride levels, which put patients at risk for diabetes or heart disease. The rates of these side effects vary, so until better pills are actually devised, treatment may continue to be less than optimum.