With schizophrenia, you pick your poison. But until last week, it often felt like a game of chance, played with a loved one's precious marbles. Now, thanks to a landmark study by the National Institutes of Health (NIH) comparing four of the new drugs and one older one, we finally know how these potions stack up. All in all, the results are pretty heartbreaking. Three-quarters of the nearly 1,500 patients in the 18-month trial stopped taking the drug they were assigned, often because it wasn't working or had intolerable side effects. Zyprexa did best (merely 64% dropped out), but, yup, it carries a high risk of weight gain and diabetes. Three other second-generation drugs turned out to be no more effective than the old one, Trilanfon (perphenazine)—a cousin of the sledgehammers but with a somewhat better picture on side effects. As a matter of fact, by one measure, you're better off with the old drug: it costs about $45 a month, while the new ones cost $400 to $500 more.
But wait, haven't we heard this story before? Does the phrase "water pill" ring a bell? Three years ago another big NIH study showed that a cheap, old-fashioned diuretic (a.k.a. water pill) worked better for most folks with high blood pressure than did costly, cutting-edge medications. (These included a calcium-channel blocker and an ACE inhibitor). Then there's the sad lesson of Vioxx and its ilk. That category of painkillers captured a $5 billion-a-year market on the celebrated promise that they were safer than older, cheaper analgesics like Tylenol or Advil. In this case, as the nation learned when Vioxx and Bextra were withdrawn and Celebrex got slapped with a black-box warning, we were paying a premium to trade a sizable risk of tummy trouble for a smaller but still troubling risk of heart attacks and strokes.
Results as disappointing as these raise big questions. The first one I asked myself with respect to my stepsister is: Why didn't we know this sooner? Why do we keep putting new drugs on the market, promote them like crazy, get everybody excited, sell billions of dollars worth, and then find out that they're not the giant step forward we'd hoped for? Isn't the Food and Drug Administration obliged to find out this stuff in the approval process?
The surprising answer is no. Sure, the FDA is supposed to make sure that new drugs are safe and effective and that promotional claims about them don't get out of hand, but the agency defines efficacy as outperforming a sugar pill. "You do not have to show one drug is better than another," explains Dr. Robert Temple, who heads the FDA's office of medical policy. (The exception: new drugs are tested against older therapies in devastating ailments like cancer or AIDS, when it would be unethical to give a control group placebos.) While Temple plainly sees the value of studies comparing new drugs to old, it might take an act of Congress, he says, to assign that task to the FDA: comparative studies, which cost big money, "are hard to execute under the current law."
So whose job is it to compare pills? "I don't think there's a clear answer right now," Temple concedes. "And with the new Medicare drug benefit about to soak up large amounts of money, we need to answer this."
The vast majority of U.S. drug research is conducted by pharmaceutical companies, but few experts trust them to do head-to-head matchups. It's just too easy for a company to tilt the odds in its own favor by choosing the weakest rival drug or by playing around with dosages or the patient-selection process.
So it is the NIH that rides to the rescue. The National Institute of Mental Health (NIMH) paid $42.6 million to study existing drugs for schizophrenia. It's money well spent, says institute director Dr. Thomas Insel, but "that's money we can't put into developing the next generation of these compounds or understanding the biology of this disease."
And it means waiting until long after drugs have been approved and desperate hopes raised to learn what might have been gleaned much sooner. The FDA has come under increasing criticism for mishandling the drug-approval process, culminating last week in the resignation of its chief, Lester Crawford. Maybe it's time for Congress to redefine that process. "The FDA should not approve a drug unless it is shown to be better in some way than the existing drugs," suggests Dr. Marcia Angell, former editor of the New England Journal of Medicine and author of The Truth About the Drug Companies. A better drug, she notes, could be defined broadly: fewer side effects, easier to take, longer lasting, more effective in a subset of patients. Such a policy would not only speed vital information to doctors, it would also spur drug companies to focus on creating truly novel medications rather than minor variations on existing themes, what many researchers call "me too" drugs. "The pace of innovation has been slow," says Dr. Jeffrey Lieberman, chairman of the psychiatry department at Columbia University Medical Center and lead author of the study.
If there's one lesson from last week's drug news, it's this: what we need—for those with schizophrenia and so many other ailments—is something new and improved that truly lives up to the hype.