Miller has argued that the FDA regulators make decisions “defensively–in order to avoid approvals of harmful products at any cost–and so they tend to delay or reject new products of all sorts, from fat substitutes to cancer drugs and painkillers.” He asserts that what is needed is a fundamental reform of the FDA’s culture of risk-aversion. He says that “the FDA’s long-standing hostility and regulatory excesses toward drug companies have caused the number of submissions for marketing approval to decline steadily since 1995, so there is now relatively little in the pipeline awaiting approval. Not even the most highly motivated bureaucrats can approve applications that haven’t been submitted.”
Miller seems agitated that people have not adopted his one and only major FDA reform idea: turning the FDA into a body that certifies the ability of nongovernmental organizations to approve drugs–analogous (in his view) to the DMV’s allowing private mechanics to conduct auto inspections. This is Miller’s way of trying to place pressure on overworked, underpaid bureaucrats who, he thinks, dream up reasons not to approve medicines. He maintains that delegating a significant portion of the oversight responsibility to nongovernmental certifiers will generate competition within a fantasy drug-certifier industry and create pressure for greater efficiency in the oversight process.
Apparently Miller believes that somehow–even though they would be legally liable for every step of the drug-development process—these new certifiers would be less cautious than the agency itself. In fact, however, when you compare the FDA to the groups that would have the most expertise and trust–Institutional Review Boards of medical-research centers–the FDA is a speedster. Miller seems to think that such private entities would compete on how fast they approved medicines. But if a drug deemed unsafe by the media had gone through one of these Miller drug mills, you can bet the approving facility would have ended up facing an investigative firestorm—and maybe even have been forced to shut down. Does he really think the overreaction to Vioxx would have been any different if a private shop had approved the drug?
Miller is not really railing at the FDA, but rather at the risk-averse culture of our society as a whole. And he has no other idea about how to combat it than either his mini-FDAs or having a commissioner with the toughness to fire civil servants for being . . . risk-averse. In fact, when the FDA has tried to stick a cork in those raising unwarranted safety concerns (David Graham is an example), Congress and the media make them heroes, no matter how wrong they are. How do you fire people in a situation like that?
To be sure, the FDA has to make more than merely administrative adjustments in the drug-evaluation process. Since 2003 the FDA has tried to reduce drug-development times by avoiding multiple review cycles. Reviewers must reach out to companies and hold regular meetings throughout the review process itself and are evaluated on doing so. But such managerial shifts are not sufficient.
The major problem with drug development today–and the principal reason the number of new drug applications being approved is at an all-time low–is that researchers simply do not know until too late in the drug-development process which drugs will work and for whom. Senior FDA leadership has been attempting to extend to all drugs the same approach the FDA has used to fast-track HIV medicine, through what is called the Critical Path Initiative: developing molecular measures of drug safety and effectiveness, applying different statistical approaches that require fewer clinical trials, and encouraging more real-time analysis of post-market drug information that can be used to refine prescribing instead of pulling medicines from the market. Indeed, drug companies and the FDA are in agreement that the science of drug evaluation has to be reformed to keep up with the science of drug discovery.
Transitioning from the old-style approach, which relies on subjective probability estimates, to a mechanistic model of evaluation, which confirms what is happening at the cellular level to change the course of disease, is not a glamorous process. The FDA is partnering with companies, academics, and other government agencies to implement nearly 100 approaches to make drug development more personalized and predictive. This isn’t as exciting as firing FDA employees at dawn, or executing them at dusk, for being risk-averse. But it does have the advantage of making all of drug development more accurate and less expensive, which means that many more medicines can be developed and tested more quickly.
For example, early this year, the FDA announced the launch of a new biomarker-based way for evaluating whether drugs are toxic in humans; it’s called the exploratory IND (short for Investigational New Drug, or a drug that has yet to be tested in people). Why is this important? Despite extensive preclinical screening using test-tube experiments and animal models, many compounds continue to fail in phase I clinical studies. But now drug developers no longer have to depend solely on such preclinical data for phase I candidate selection. The FDA and the National Cancer Institute developed a set of biomarkers based on the relationship between genetic variations and drug response at a molecular level; now the agency is seeking to combine biomarkers with new ways of using powerful methods of analyzing clinical data to speed up drug development and make medicines safer. Reliance on a single controlled trial–a technique widely used by the FDA in other settings–is one approach for integrating these new tools.
These efforts are but the beginning of a fundamental shift in the way medical innovations are evaluated. Some economists have estimated that reducing the number of clinical trials and time required by 10 percent could reduce the cost of development by hundreds of millions of dollars and time required by years. Here’s what one observer said about the impact of these technologies: “Why is drug development so expensive? Certainly not because of a dearth of technology: Biotechnology, combinatorial chemistry, sequencing of the human genome, improved purification techniques, and other innovations have made drug research more efficient.”
The FDA leadership over the past several years–Mark McClellan, Andy von Eschenbach, Janet Woodcock, Scott Gottlieb–have being making these new techniques part and parcel of the drug-evaluation process. Yet the same observer who made that last comment–Henry Miller–has nothing good to say about them. Indeed, in his piece for NRO he sought out the most critical and personal attacks on them in the writings of those who support more, not less, FDA regulation. In truth, Miller’s baseless criticisms are insignificant when compared to the accomplishments and dedication of such people as Gottlieb, Woodcock, and von Eschenbach.
Robert Goldberg and Peter Pitts are vice president and president of the Center for Medicine in the Public Interest, which was founded early this year.