As Congress tries to figure out how to speed up the drug approval process, a pair of economists are asking a different question: Are the criteria we use to choose which drugs are safe and effective too inflexible?
At the core of this question are two kinds of mistakes that federal regulators can make when deciding whether or not to approve a new drug. The first is pretty straightforward: If they set the evidence bar too low, drugs that don't work and carry dangerous side effects could end up on the market. The second error is less obvious: If they set the bar too high, effective medicines may end up in the dustbin instead of in the hands of patients.
The new study suggests that for many devastating diseases, the current standards are too risk-averse, erring on the side of keeping drugs off the market even when patients, facing a dire prognosis, would accept higher risks that the drug doesn't work or comes with toxic side effects.
"For certain kinds of contexts, that doesn't make any sense," said Andrew Lo, a professor of finance at the Sloan School of Management at the Massachusetts Institute of Technology who did the study with his student, Vahid Montazerhodjat. Lo said that pancreatic cancer is the poster-child of the problem that motivated the team to propose an alternative method for calculating the level of acceptable risk.
"It's a death sentence for people that have it, and when I talk to pancreatic oncologists, they tell me they feel more like undertakers than doctors. They nave nothing to offer," Lo said. "So do you really want to be as stringent in those cases where patients are going to die anyway? You'd take a bigger chance of making a mistake."
Lo praised the FDA overall. It's not a surprise that the agency would strive hard to avoid the first kind of mistake; the repercussions of allowing bad drugs on the market are lawsuits, unfavorable press and deaths. And the idea that effective drugs are languishing due to risk thresholds is mostly hypothetical; no one really knows how many drugs haven't been brought forward. A recent analysis commissioned by Forbes found that rejecting drugs has become a rarity at the agency, and the drug approval rate has gone up over the past seven years.
Still, for diseases that have no therapies and the prognosis is death, Lo said, the typical threshold of risk may be overkill. Traditionally, that allows for a 2.5 percent chance that a drug is approved that doesn't work. He proposed using a statistical method that would take into account the prevalence and severity of a disease and determine a new, individualized risk threshold for each disease. A higher threshold of acceptable risk for an illness that is untreatable today could allow for smaller clinical trials to be designed that take less time.
The team found that for relatively common diseases that do have treatments, such as prostate cancer, the current statistical standard may actually be a little too lenient -- we may be accepting riskier drugs than we should. But for devastating diseases such as pancreatic cancer, the tolerance for risk that a drug doesn't work should, according to their calculations, be about 10 times higher.
Jeff Ventura, a spokesman for the FDA said that the agency had no comment on the study.
Hamilton Moses III, the former chief physician at Johns Hopkins Hospital, said that this kind of an approach fits with clinical experience. Neurologists treat patients with diseases like Parkinson's, for which there are already a range of treatments to manage the symptoms and improve quality of life, and ALS, which often strikes young people and for which physicians have virtually no treatment options. Patients in these groups both need more effective therapies, but their acceptance of risk is likely to be different -- the problem has been figuring out just how to factor in that acceptance.
"From the perspective of policy, this is a real problem with the FDA, since so many patient advocacy groups would like to see the risk threshold lowered for 'their' diseases," Moses said by e-mail. The paper could offer a quantitative way to determine what the risk threshold is for different disease.
But Moses also said that there are possible problems. When companies call for changing the risk threshold, they point to surveillance after the drug is approved as a safeguard against adverse events. That is often insufficient, he said.
Watchdogs have been critical of efforts to loosen and ease the regulatory process, but economists say that the effect of the current threshold for drug approval is not just in the number of drugs approved, but in what drugs get developed in the first place.
"A conservative FDA reduces the flow of new drugs to be tested. In a sense, failing to approve a good drug has two costs, the opportunity cost of lives that could have been saved and the cost of reducing the incentive to invest in R&D," economist Alex Tabarrok wrote on the blog, Marginal Revolution.
Lo similarly argued that a uniform risk threshold has an effect on whether companies even try.
"I've spoken to pharma companies and biotech companies that have said, look, it’s not worth it for them to take the risk; it costs a lot of money, takes a lot of time and at the end, there's a small probability of getting an effect," Lo said. "If the threshold were lower -- if you increased the odds of success by a factor of 10 -- that’s got to affect, from a rational economic perspective, your desire, your motivation, your funding of these kinds of projects."