Medication Under a Microscope

Evaluating 'Surrogates'

Part of the problem, Krumholz, Nissen and others say, is how the government sometimes evaluates drugs, relying on "surrogate endpoints" instead of medications' ability to treat or prevent illnesses. Cholesterol-lowering drugs, for example, can be approved based on their power to cut cholesterol and not on whether they protect against heart attacks or strokes. Diabetes drugs can be approved based on whether they reduce a protein known as hemoglobin A1C -- a measure of blood sugar -- and not on their effect on complications caused by high blood sugar.

"There are a lot of things we do in this country where we treat these surrogate measures with very little evidence that we are actually treating the patient," said Nortin M. Hadler, a professor of medicine at the University of North Carolina.

If one drug prevents heart attacks by lowering cholesterol, that does not guarantee that another will do the same if it works differently, critics argue. And drugs may have unforeseen hazards that outweigh any benefits, or they may not produce secondary effects, such as lowering inflammation, that add to their usefulness.

"It's not necessarily how low you go, but how you get there," Nissen said.

Pharmaceutical industry representatives argue that the use of surrogate endpoints has been shown to be reliable and helps speed valuable drugs to patients. It would take much longer and cost much more to conduct studies that measure the effect on diseases.

"If you took that approach, in all likelihood you'll be denying patients the opportunity to be successfully treated because you'll increase the time it would take to make a drug more widely available," said Alan Goldhammer, deputy vice president of regulatory affairs for the Pharmaceutical Research and Manufacturers of America, an industry trade group.

One solution, some experts say, may be to approve drugs based on surrogate measures but require drug companies to quickly follow up with studies to prove that the medications are helping patients.

"It should be done only with the proviso for following up with a trial to prove that the drug improved the things we really care about -- survival, heart attack, stroke -- major symptoms that make people feel better or live longer," Nissen said.

Food and Drug Administration officials said that is something the agency might consider but added that evaluating surrogate endpoints for drug approvals has been very useful.

"Some people don't respond adequately to available therapy," said Robert Temple of the FDA's center for drug evaluation and research. "You have to make a judgment about whether you want to have a way of treating that now or you want to wait five years, 10 years -- who knows? That's a judgment that has to be made."

Pushing Prescriptions?

Some critics go even further than questioning the use of surrogate measures to approve drugs. They argue that the evidence supporting the widespread use of many cholesterol-lowering and blood-sugar-lowering drugs is weak, except for select patients. The popularity of these drugs has been driven by industry-funded studies that overstate their value for preventing disease, they argue, exposing far too many people to the potential risks of the medications themselves.

"What's going on here is our research enterprise is almost completely controlled by the pharmaceutical industry," said John Abramson, a clinical instructor at Harvard Medical School and author of the book "Overdosed America." "It's their job to create a need for their products. Their job is not to maximize public health."