A License for Safer Drugs

By Marsha N. Cohen
Saturday, September 2, 2006

We suffer from a migraine, have a chronic illness, pick up an infection -- our doctor prescribes a drug, we fill the prescription, we swallow the medicine. We hardly think about a drug's risks while waiting anxiously for it to make us feel better. Because this country has what is generally deemed the world's best system for ensuring the safety of prescription drugs, the risks are usually well worth the benefits.

And yet our confidence has been shaken in recent years by revelations about risks emerging in such drugs as Vioxx, Ketek, Baycol and all the antidepressants. Each was licensed by the Food and Drug Administration, and at the time each was approved, its effectiveness in treating conditions for which it was marketed outweighed its safety risks.

But all that the FDA knew at the time of approval came from clinical testing on just a few hundred to a few thousand people. Dangerous, even deadly, side effects that might arise only once in a thousand uses are extremely unlikely to be revealed in pre-market testing. If the danger is otherwise common in the population (for example, of heart attacks), then the causal connection between the harm and use of the drug might remain obscured for a long time. And yet a drug's license has no expiration date.

Turn that fact over in your mind. The state grants you a driver's license, but you must renew it periodically. Teenagers are given only probationary licenses in many states. But drugs are in essence approved forever. True, the FDA might remove a drug from the market because of safety concerns. But without any automatic review system in place, too little attention is paid -- by both manufacturers and the FDA -- to developing information to determine whether the predicted benefit-to-risk ratio of a drug has remained the same after exposure to it has become widespread.

Drug development is extremely expensive; the investment is largely returned in the early years of marketing, when patent protection rules out price competition. Why would a rational drug company voluntarily look for reasons that its drug should be removed from the market?

Congress is looking into this situation, and a number of interesting ideas for legislative change have been discussed. But they apparently do not include this simple notion: limited-term licensing and a license renewal process.

While the threat of lawsuits ought to provide some incentive for the drug companies to do more post-marketing scrutiny, it seems to have had a limited impact. And the FDA lacks the financial and human resources for the kind of marketplace surveillance that could yield timely information about the safety of newly marketed drugs.

Looming license expiration dates would give both the drug companies and the FDA a much greater incentive for the post-marketing studies critical to confirming drug safety. A drug approved after the traditional three phases of testing (in which the last step usually includes several thousand patients) could receive a five-year license. Before that time elapsed, the company would be required to conduct studies of the safety and effectiveness of its product in use and to demonstrate to the FDA that the expected benefit-to-risk ratio had held firm. At that point it could receive a permanent license, subject, of course, to even later developments.

Many new drugs are released to the market after accelerated approval -- and after testing on far fewer people than is standard. A shortened test protocol enables new drugs for serious or life-threatening conditions to get to patients sooner. But the inevitable result is that the post-marketing period will yield critical information, including some that might suggest that approval of the drug should be rescinded. These drugs ought to receive a shorter period of initial licensure than others, perhaps just three years, during which the manufacturer would be required to complete at least the equivalent of the traditional Phase 3 testing. After that the drug could be licensed for five years, and the company would be required to submit data from continued active surveillance to get a permanent license.

This simple change in our drug approval laws would yield a bonanza of important information on the safety and effectiveness of drugs in use, in comparison to the anemic feedback on newly marketed drugs now available.

Early users of every new drug, however thoroughly tested, are in essence additional test subjects. But their experiences rarely translate into helpful information. For one thing, if the patient suffers some common adverse effect (such as a heart attack in an elderly patient), or a drug-caused effect manifests itself only after the drug is discontinued or after long use, the connection between the drug and the adverse event might not be recognized. And even if the connection is suspected or made, there is no requirement that patients or physicians notify the FDA or drug companies.

Under current practices, drug companies, while they need to be responsive to reports of adverse events, are not required to go looking for them. Without an organized and purposeful attempt to obtain feedback (positive and negative) about the impact of a new drug, information will necessarily be incomplete and inaccurate.

Limited-time drug licensure would furnish incentives to ensure that all new drugs are properly scrutinized. Companies would need to formulate post-marketing surveillance plans as well as promotional plans when a drug is approved. The FDA would be subject to firm deadlines in reviewing post-marketing information, because prescribers and patients would be relying upon a seamless system of licensure renewal in the absence of negative information.

This system would not be cost-free, but neither is the current one. The timely discovery of the cardiac implications of the use of Vioxx, for example, would have saved countless dollars and prevented many serious health effects, including deaths. It would have dramatically reduced the litigation that is so reviled by business but needed for the redress of patients who suffer harm.

Congress ought to enact a license renewal requirement that will be safe and effective.

The writer teaches food and drug law at the University of California's Hastings College of the Law in San Francisco.

© 2006 The Washington Post Company