Some days were presumably more fun than others. Take the controversy surrounding a drug for Duchenne muscular dystrophy. Studies did not show a clear benefit to the drug and an FDA advisory committee voted against approving it, but patient advocacy groups pushed for the drug. After much infighting, the drug was granted accelerated approval this fall by the agency's top drug-review official, and Califf allowed the decision to stand.
Meanwhile, Califf strongly defended the FDA's requirement that a drug must be shown to be effective for approval. Jim O'Neill, an associate of PayPal co-founder Peter Thiel who has been mentioned as a possible successor to Califf in the Trump administration, has suggested that the agency require only safety testing for approval.
Califf also talked about his mother, who has multiple myeloma.
(This interview has been edited for brevity and clarity.)
What do you think are the biggest challenges facing the agency?
We're seeing increasingly complex information, so we need really good people to deal with it, and there were limitations in terms of hiring and information technology and infrastructure that are really critical to a science-based public health agency with a regulatory mandate. So that’s something I have really focused on, and it’s going to be an ongoing challenge because the world is moving faster, not slower.
The second big need is . . . evidence generation. I started off as a doc, running intensive-care units. You make decisions about individual people, multiple decisions a day, your decisions have consequences. It’s exciting. There’s a lot at stake. The middle part of my career has been as a researcher and research administrator where at the end of every article you say, “More research is needed.” And that's wonderful, too, because the goal is open exploration.
But then at the FDA, you have a lot of the same science issues but you have to make decisions. And it’s very noticeable when we make decisions with good evidence. It's still emotionally charged because we regulate such a large part of the economy, and there are winners and losers. But when we have good evidence, it’s easy to defend the decisions and the arguments are typically good arguments to have about how you interpret good evidence.
But we are often in situations where we have to make decisions because there are time limits and we don't have the kind of evidence we really need to make a good decision, and then things are not so pretty. Here we are in a time when everything is digitized, we have scads of information, but as a society we haven't been so good at turning that information into knowledge, particularly in the medical arena.
When you look back at the controversy over the Duchenne muscular dystrophy drug, do you have any second thoughts about the way that turned out?
We didn’t have the evidence we really wanted and a decision had to be made. But the steps that should be taken in the future to reduce the number of times we get into this situation, I think I clearly laid out. By law, the FDA is given a lot of discretion in cases of serious and life-threatening diseases with no available treatment. We are instructed by law to consider all sources of evidence beyond the traditional and to use unvalidated biomarkers [such as molecules that may be affected by a drug and measured] if we think that it’s reasonably likely they will predict a clinical benefit. So as I pointed out in the document, the definition of “reasonably likely” is not defined and “all sources of evidence” includes a lot of possible things.
We have been working closely with NIH because with 5,000 rare genetic diseases and this being such an exciting time of not just quantifying the genome, which Francis Collins started, but now understanding the functional aspects, we are going to have cures for a lot of rare genetic diseases.
I hope people will look at the Sarepta case [Serapta Therapeutics produces the Duchenne muscular dystrophy drug], and say there are some things that regardless of whether we think FDA made the right decision, let’s look at what needs to be done for the next 4,999 rare genetic diseases because patients and their families have reason to want things to get done.
What’s your reaction to the argument that drugs should not have to be proven effective before getting approved by the FDA?
There’s a history to proving effectiveness of drugs that goes back to pre-1962, that I think many people are aware of, and the requirements are written into law. I think what many people lose sight of is that what’s been called a standard is actually quite flexible. As an intensive-care-unit doc, and now I’m of the age where I’m a patient, too (I’m on drugs for lipids and hypertension), I think people will want to take drugs that have a benefit because all drugs have a risk. I don’t know of any that are completely free of risk.
Almost 90 percent of drugs that get into Phase One [the earliest stage of clinical trials] don’t make it to market because of toxicity or they actually don’t work or they can’t be manufactured on a scale that’s needed to be safely produced and distributed in a global market. And so that means if you are not demonstrating that a drug has a benefit, most of what they take won’t work.
And I think we have pretty clear evidence from the public that they would like to have a system that's giving them some assurance that the treatments they are given work.
But again, to get back to the flexible standard, that means in the case of rare genetic disease with no available treatment, “working” means changing a biomarker, it doesn't mean improving a clinical outcome, and then the proof of that is developed post-market.
I think 21st Century Cures worked out quite well for us. The personnel issues are clear and we are very happy with what Congress agreed to do there. [The bill allows FDA to give pay raises to scientists and other expert staff.] And the sections on real-world evidence and modernizing clinical trials are all quite consistent with the direction we hoped we would be going in and I think will be helpful.
In the debate over high drug prices, the FDA sometimes gets criticized for not approving generic drugs quickly enough. Some in Congress want to set a 150-day deadline for some generic-drug approvals. What's your reaction?
We need to approve generics as quickly as we can, keeping in mind that the laws for generics are pretty simple in terms of what a company needs to do for a generic drug. But a company also needs to be able to manufacture the drug and show that it’s high quality throughout the supply chain. In some of the dialogue about this, people may have lost sight of the fact that if you look at the backlog in generics it’s almost all compounds that have had multiple applications because there were deficiencies that needed to be corrected. I don't know that 150 days is necessarily the right number. But there’s no objection at FDA to going as fast as we can as long as we keep a quality standard.
What else is on your mind?
My mom has multiple myeloma. If you were a cancer patient, you would want the industry to do well, but you want there to be a rational approach that optimizes getting the right treatments out there at the right time. She’s a beneficiary of accelerated approval of a drug for multiple myeloma. Just recently, I got a call saying it’s time for hospice for your mom. She has had 89 good years. I'm saying all this with her agreement and permission.
I called [FDA oncology chief Richard] Pazdur and said, What’s new? And he said, We have a few things that look good and they are tolerated well and even for someone age 89 might work. And it's worked in her. And I’m insistent that we need to then bring in the evidence after accelerated approval. For this particular drug, I won't go into the specifics of the drug, but I was pleased a couple weeks ago to see they did randomized trials in follow-up and they were positive and got additional indications.
She wanted to be alive to vote. I won’t say for whom.