Vaccination News Home Page

http://www.nytimes.com/2002/09/17/health/17DOSE.html?tntemail0

Caution: That Dose May Be Too High

By ABIGAIL ZUGER

wenty percent of prescription drugs are marketed with instructions for use that must later be corrected by the manufacturer, researchers say — almost always to lower the recommended dose or to warn that the drug may be hazardous to certain patients.

Drugs released in the last five years are even more likely to have their instructions changed than older ones.

Advertisement

These findings by Georgetown University researchers, along with similar data from Europe, support longstanding concerns among some experts that drugs are being studied at excessively high doses to emphasize their effects, and then marketed at the same high doses to maximize profits.

Some experts say these patterns may contribute to the high frequency of serious and life-threatening drug side effects.

The Georgetown study examined the recommended doses of 354 prescription drugs released from 1980 to 1999. Of those drugs, the instructions on the label were corrected for 73, or 21 percent, after the drug came to market.

Eighty percent of the corrections consisted of a reduction in the original dose or a new restriction for certain groups of patients, like those with liver or kidney disease or those taking certain other medications.

Drugs approved by the Food and Drug Administration through its so-called fast-track system adopted a decade ago for lifesaving compounds seemed no more likely to have corrections than other drugs.

Drugs with label corrections came from all drug classes and include the potency drug Viagra, which was restricted from certain cardiac patients after the drug was first approved; the H.I.V. drug AZT, whose dose was cut in half after its release in 1987; and the antidepressant Prozac, which should be started in the elderly in smaller doses than usual.

The Georgetown study did not include drugs often used at doses that are different from the ones recommended by manufacturers (like pain relievers), or drugs that have been reintroduced in low-dose variants (like birth control pills), and for this reason its figures represent a low-end estimate of the number of drugs with dosage corrections after they are on the market.

The study will be published in the medical journal Pharmacoepidemiology and Drug Safety, and was released on the its Web site in August.

"We've seen a lot of situations where drugs are approved by the F.D.A. and subsequent important information about their optimal dose is not determined until afterward," said James Cross, the study's lead author, who did the research at Georgetown's Center for Drug Development Science and now works for the food and drug agency.

"I don't think that the results are alarming," Mr. Cross said, "but they shouldn't be taken with indifference either. Patients and prescribers need to be aware that a drug label is a dynamic piece of information; it doesn't just get approved and then sit there on a shelf never to be heard from again."

Dr. Wayne Ray, a professor of preventive medicine at Vanderbilt and its director of pharmacoepidemiology, said: "These results don't surprise me at all. It's sort of an unintended consequence of the way we approve drugs. They've put their finger on an important policy issue."

When a drug is being studied for possible marketing, the recommended dose is set very early in the process, based on small studies, so that manufacturers can efficiently carry out large studies of the drug's safety and efficacy.

Because manufacturers have a strong incentive for the drug to show an effect in these studies, they often choose to study a relatively high dose, which may prove to be too high, Dr. Ray said.

The process has other inherent limitations that help explain why so many drug doses change, Mr. Cross said. "Manufacturers can't test drugs in millions of people, so inevitably things are detected after approval," he said.

For instance, if a new drug interacts badly with a drug for a different purpose already on the market, the recommended dose of the older drug may have to change for people taking both, a situation that could not have been foreseen when the old drug was first marketed.

"Still," Mr. Cross said, "tools are available to the drug companies about better analyzing the data from premarketing trials." Such tools include laboratory analyses of drug molecules and sophisticated analysis of clinical data from the first people to take the drug, so that some dosing corrections might be made earlier.

Drug makers and regulatory agencies anticipate learning new details of a drug's effects and interactions after it goes to market. Even the largest clinical trials cannot always identify patterns that emerge when drugs are used by millions of people.

Drug companies maintain that the high rate of label corrections indicates the progress of science. "The central issue is that we're constantly learning about a drug throughout its life cycle," said Dr. Alan Goldhammer, an associate vice president at Pharmaceutical Research and Manufacturers of America, a trade group. "We can't know everything about a drug before it goes to market. That would mean long regulatory delays and the American public would not want that."

Manufacturers always continue to assess a drug's safety after marketing, Dr. Goldhammer said.

The Georgetown team found that when its data were adjusted for the length of time a drug was on the market, those released from 1995 to 1999 had a chance of label correction that was roughly threefold greater than the rate for those released from 1980 to 1984. The corrections occurred after an average of 16 years on the market for the older drugs, but after only 3 for newer ones.

Under the F.D.A.'s fast-track policy, some watchdog groups have said drugs are being rushed to market without adequate testing. But Mr. Cross said that most accelerated drug approvals were in cancer and H.I.V. drugs and that these were no more likely to have labeling corrections than others.

Rather, he said, the recent spate of label corrections may reflect more vigilance from manufacturers and the F.D.A. to find dosage problems early and correct them.

Further evidence that the recent acceleration in dose corrections is not directly related to federal regulatory policy comes from a new European study with very similar findings for drugs marketed there.

Published with the American study, the European research used marketplace data compiled by the World Health Organization to show that from 1982 to 2000, 115 drugs used in Europe had their standard doses corrected. Most were reductions, and they became more frequent in the middle to late 1990's. Drugs that were first marketed in the United States were no more likely to have dose corrections than were drugs first sold in Europe.

Advertisement

"We can't say if it's the mistakes that are happening at a rising rate or if it's the mechanisms for finding them that are getting better," said Dr. John Urquhart, a professor of pharmacoepidemiology at Maastricht University in the Netherlands and an author of the European study.

Economic factors also come into the equation. Although some drugs have similar prices set for a range of doses, most "are priced on a per-milligram basis," Dr. Urquhart said. This explains manufacturers' desire to sell as many milligrams as possible, he said, while giving groups that pay for prescription drugs an incentive to finance studies supporting lower doses.

For some experts, these results only confirm long-held views that drug-dosing and marketing policies are fundamentally flawed.

"It's long been known that for individual subjects the dosage listed on a drug label is not necessarily the right one," said Dr. Carl C. Peck, the director of Georgetown's drug development center and an author of the its study. "We're aware of the pressures drug developers feel to select a dose and invest in that dose. But one in five is a high error rate. Both doctors and patients should be concerned."

Dr. Peck said that, except in medical emergencies, for many patients a policy of "start low and go slow" is the best way to begin a prescription medicine, with frequent visits to a doctor who is committed to readjusting doses based on the patient's response. In the future, he said, genetic-based analysis of a patient's metabolic idiosyncrasies may help streamline this process.

Dr. Jay S. Cohen, an associate professor of family and preventive medicine at the University of California at San Diego, said the Georgetown data illuminated only part of the problem. "The findings don't include a lot of drugs whose doses should be reduced but aren't," he said. In his book "Overdose" published last year, he argued that for many people the recommended drug doses were much too high.

Dosage recommendations are based on small studies done on young, healthy people, usually men, Dr. Cohen said. But people have such variability in size, shape, age and patterns of metabolizing drugs, not to mention concurrent illnesses and necessary medicines, that the notion of a single optimal treatment makes no sense, he said. About 75 percent of the serious side effects occur at recommended doses, he added.

"If a dose is reduced 7 or 10 years after the drug was first marketed," Dr. Cohen said, "you wonder how many side effects, what great costs to patients and to the medical system could have been avoided."

Vaccination News Home Page

ALL INFORMATION, DATA, AND MATERIAL CONTAINED, PRESENTED, OR PROVIDED HERE IS FOR GENERAL INFORMATION PURPOSES ONLY AND IS NOT TO BE CONSTRUED AS REFLECTING THE KNOWLEDGE OR OPINIONS OF THE PUBLISHER, AND IS NOT TO BE CONSTRUED OR INTENDED AS PROVIDING MEDICAL OR LEGAL ADVICE.  THE DECISION WHETHER OR NOT TO VACCINATE IS AN IMPORTANT AND COMPLEX ISSUE AND SHOULD BE MADE BY YOU, AND YOU ALONE, IN CONSULTATION WITH YOUR HEALTH CARE PROVIDER.