This article first appeared in Re/Code on August 28, 2014, 5:30 AM PDT
Silicon Valley loves to hate the Food and Drug Administration.
In my eight months on the health beat at Re/code, the agency comes up repeatedly in interviews, cast as the plodding, paternalistic bureaucrats stifling technological and medical progress.
While health is suddenly a hot topic in the region, a handful of venture capitalists have said they simply won’t invest in companies that require FDA approval — opting for the relative safety (or more obvious exits) of health apps and infrastructure plays over drugs and diagnostics. Others are looking for end runs around the system altogether.
In large part it’s a cultural clash. The ask-forgiveness-not-permission ethos of an Uber or Airbnb, which counts on consumer sentiment to swing policy, doesn’t play with the FDA. You’re not allowed to move fast and break things when it comes to people’s health. And the agency doesn’t take kindly to companies that violate its rules while ignoring repeated requests for information.
But some in Silicon Valley, particularly those who have no choice but to work with the FDA, are striving to locate what is ultimately some obvious middle ground: Yes, the FDA is slow, underfunded and understaffed. But beyond the invisible hand fringes, few are eager for a return to the days when the free market decided what was medicine and what was snake oil (or worse).
I caught Dr. Beth Seidenberg, a partner in the life sciences group at Kleiner Perkins Caufield & Byers and a physician who trained at Johns Hopkins University, sounding utterly constructive on the issue during a conference earlier this summer.
Seidenberg is the former chairperson and co-founder of the National Venture Capital Association’s Medical Innovation and Competitiveness Coalition, created in 2010 specifically to push for FDA reform. Before joining Kleiner, she was the chief medical officer at Amgen and a senior executive in research and development at Bristol-Myers Squibb.
I recently visited her office along Sand Hill Road in Menlo Park, Calif., to discuss the topic of FDA reform further. The interview that follows has been edited for space and clarity.
Re/code: What are MedIC’s key areas of concern with how the FDA regulates startups today?
Beth Seidenberg: MedIC was formed when FDA was in somewhat of a crisis. We were seeing a tremendous slow down in drug approvals. And the device division, [the Center for Devices and Radiological Health], was having great variability in not only product approval for market authorization but also product approval for initiating clinical trials. It was sort of a black box.
We put MedIC together with NVCA to bring entrepreneurs and venture capitalists together to basically start an effort around providing insight to FDA about the consequences of the lack of predictability and the slow progress to get products approved.
And now, three years later, it’s been remarkable. We’re very, very proud of both NVCA-MedIC and FDA for the accomplishments that have been made.
The most recent PDUFA [the Prescription Drug User Fee Act, which is renewed every five years] advanced to where there’s more transparency, more risk is being taken and the quality of the reviews has gone up significantly.
Can you be a little more specific? How has the process actually changed?
One way is there was a big effort to focus on innovation and provide a faster regulatory path for innovative products. There was a clause put in for a “breakthrough therapy” designation.
It allows for the sponsor and the FDA to work hand-in-glove in advancing a product to market as quickly as possible, balancing efficacy, unmet need and safety.
In 2013, there were 37 designations granted and three approvals: Gazyva, a Genentch drug for [chronic lymphocytic leukemia]; Imbruvica, which is approved for [mantle cell] lymphoma; and the Gilead hepatitis C drug Sovaldi.
(Another five have been approved so far in 2014.)
With all that said, is it perfect now, or are there areas where the FDA still falls short?
No. I think areas where we need to focus and need more help is in neurodegenerative disorders like Alzheimer’s and Parkinson’s, the silent killers of our elderly population that are now, and are going to continue to be, a huge burden to society.
Part of it is that we don’t have the tools. But the other part is that the FDA has an opportunity to work with sponsors to develop those tools and provide a pathway to help get those drugs to market earlier.
In the case of neurodegenerative illnesses, we don’t have good endpoints, we don’t have good measures of the progression of the diseases. So we do trials where patients are very far advanced but we need to move the trials to early onset of the diseases, where you can have a more significant impact.
To me that’s the big unmet area on the drug side. There’s also a lot of work needed on the device side. The industry would generally say the CDRH is still understaffed and underperforming relative to drugs and that’s having an impact on the industry from an investment perspective.
Meaning that money is not flowing in because of the amount of uncertainty?
It’s uncertainty at both ends. They have something called the IDE, investigational device exemption. But there’s unpredictability on how long it will take to get your IDE approved and what requirements will be requested for a specific device. There’s a lack of a consistent pathway.
And then on the approval side, I can tell you for one of our companies, it was noninvasive, it was considered a device, but it was more of a diagnostic. It took a year and a half to get the FDA to approve something that was basically no risk. It was more of a question around a statistical analysis than efficacy or safety.
Those kind of uncertainties create barriers for people to invest and for entrepreneurs to innovate.
It’s clear to everyone why we want the process to go as smoothly, efficiently and quickly as possible. But there are examples, like Vioxx and Clyert, drugs that had some horrible side effects, that did get through the process. So how do we accelerate and streamline without sacrificing safety and efficacy?
When you have a new therapy, they’ve now put in place what are called REMS, risk evaluation and mitigation strategies. They are, I’d say, required more frequently now for new products as they’re approved.
It allows the sponsor to do either specific data collection in an ongoing fashion as a new drug is released and communicate that to the FDA on an ongoing basis — or, in some cases, it requires a sponsor to run another trial. You can never take the risk away, but it allows good oversight of risks so that issues are highlighted earlier.
The FDA has been working on guidelines around health apps and digital health products. And we’ve seen companies emerge describing themselves as digital therapeutics like Mango, which Kleiner invested in. They claim they can help people on the edge of developing serious diseases stay healthy by steering them toward healthier lifestyles and better drug compliance. At what point does information and guidance tip into medical advice that requires scientific scrutiny and regulatory oversight?
As I look at it from a physician perspective, I think anything that is a diagnostic test or provides a therapeutic recommendation should be regulated, because you don’t want to get that wrong.
Let’s use a simple example that’s out there every day: Measuring blood glucose and advising a patient on how much insulin they should take. In that case, you’re basically saying, based on this test, here’s a therapeutic intervention.
You want to make sure you get that right and that should be regulated. But that’s distinct from a diabetic who needs to take their Metformin every day being reminded, “Take your Metformin today.”
If there is an app that is an enabler of better health outcomes, like Mango — where we know the cost of not complying with medication is in the billions of dollars every year — in that case we’re providing a tool that facilitates or enables a consumer to actually have hopefully better outcomes.
In that case I don’t think you need regulatory oversight, because you’re not putting somebody at risk. You’re actually probably lowering their risks.
The most timely and local example of the FDA and a startup butting heads is 23andMe, where the FDA required it to stop selling the health version of its personal genetics product. Because you look at these issues closely, I’m curious whether you had an opinion on whether that was an appropriate or inappropriate decision?
That’s between the FDA and the sponsor. I never want to opine [on] a decision where I’m not privy to all the back and forth information.
But in that case, where they’re providing an analysis of someone’s genome and advising on risk relative to the general population — right? That’s essentially what they’re doing. If they’re providing a diagnostic that says that you have the BRCA1 gene and you are at risk for breast cancer, therefore you should go receive mammograms every six months — in that case, that is a very directed path relative to a diagnostic, and FDA should make sure that the information being provided is accurate.
I don’t want a lot of false positives or false negatives — and [do want to know] that the recommended path of action or intervention is something that has been vetted by the medical community and accepted.