Reader Question #1: Why doesn't every Duchenne drug get an accelerated approval?
An explainer on fast-tracking drugs
Most of us in the Duchenne community know at least a little bit about the FDA. We know that the FDA is responsible for approving the drugs on which we rely, like gene therapy and new steroids. We know—and complain often—about the glacial pace of those approvals. That aside, the FDA is a rather opaque institution and its processes labyrinthine.
The other day, I had reason to give the agency some thought. A reader and fellow DMD mom asked me about accelerated approval, the program that speeds up decisions for rare and serious disease drugs.
Why don't all Duchenne drug companies seek accelerated approval? And how quick is an accelerated approval compared with a traditional approval?
This week, I'd like to answer this reader's questions, because I think they are important for all of us to understand. Whether a drug gets an accelerated approval can mean gaining access to potentially life-changing drugs substantially earlier—and in our situation, the passage of time ticks loudly.1
So at the risk of getting wonky, let's dive in.
How does the FDA approve drugs?
The FDA’s job is to evaluate drugs before they can be sold to patients. To do this, the agencies’ scientists and statisticians look at reams of data collected by pharma companies that track whether a drug is having any positive effects on a patient’s health as well as any side effects. This data comes from clinical trials.
For a long time, the FDA required drug companies to conduct two full clinical trials. In order to be approved, drugs had to show that there was clear “clinical benefit.” (Remember this term; it’s important. “Clinical benefit” is defined as a favorable change in how a patient feels, functions, or survives as a result of a treatment. In Duchenne clinical benefit is often measured by changes in function, like the NSAA, the PUL, and the four-stair climb.) If they did, they could apply for a traditional approval for their drug.
Then in 1992, amid the desperation of the HIV epidemic, the FDA introduced a pathway called “accelerated approval” for very serious diseases where there was a clear need and no good treatments. This was intended for drugs that could show a change in a biomarker—a thing you measure in bloodwork or imaging, e.g., dystrophin production in the case of DMD, or viral load in the case of HIV or tumor size in cancer—that would be likely to result in clinical benefit even if the drug had not yet shown those changes in trials.
Over time, the accelerated approval process changed. For instance, there was no longer a requirement that drugs go through two trials; one could suffice so long as the company had already begun enrolling a second trial that would confirm that “clinical benefit.”
Recently, the program has attracted negative attention, too. There have been some valid criticisms that the FDA has fast-tracked drugs that fail clinical trials because pharma pays for much of the agency’s budget; that it’s encouraging people to spend huge sums on unproven drugs; that the FDA is not holding companies accountable to complete trials that would show whether drugs actually work.
So why don't all DMD drugs get an accelerated approval?
Generally speaking, DMD treatments fall into one of two camps: The first is treatment that restores dystrophin, either by exon-skipping or by gene therapy. The second is drugs that limit inflammation.
Let's take a look at the DMD drugs that have received an accelerated approval: four exon skippers, and Elevidys, Sarepta's gene therapy.
These are the drugs that went through the longer, traditional approval process: Duvyzat (aka Givinostat), and the steroids Agamree and Emflaza.
There's a pretty obvious division here. The drugs with that could use dystrophin as a biomarker received accelerated approval—and the drugs that targeted inflammation didn't.
Remember how clinical benefit is defined by how a patient feels, functions or survives? Well, an increase in dystrophin does not necessarily improve a patient's function. (The reason for this is complicated. For more on the faultiness of dystrophin as a biomarker, jump to this footnote.2) But since dystrophin is the missing protein-in-question, it’s reasonable to expect that increases in dystrophin would improve muscle strength and function. (Again, this stuff is complicated, and it hasn't always worked out that way.3)
The other basket of drugs—those that try to stabilize muscle or limit inflammation, like steroids—doesn’t affect dystrophin. The trouble is that those drugs also don't meaningfully change any molecules that are reliable proxies for muscle health or progression.4 Without biomarkers, these companies have to show that boys are getting stronger—or at least getting weaker more slowly than you'd expect. And that can take years.
Back to my reader's other question: Is an accelerated approval always faster?
There's no magic number for how quick either of those two processes could be, but “in the context of DMD, accelerated approval is certainly faster than traditional full approval,” said Tim Franson, an infectious disease doctor and pharmaceutical consultant who has worked on drug approvals for decades.
Two recent developments illustrate this pretty clearly.
Avidity is an exon-skipping company whose stunningly good dystrophin readings I wrote about a few months ago. They said in a call with investors that they would be initiating talks with the FDA to chart a path toward accelerated approval, presumably using increases in dystrophin as a biomarker. Bear in mind, the company has run one trial, which began less than two years ago, and their dystrophin readings come from muscle biopsies taken four months after dosing. That's right, they are using data from a change seen after four months.
Let's compare that to Givinostat, an anti-inflammatory drug that was approved earlier this year. Without an FDA-approved biomarker, the company had to prove a change in boys’ strength, which was measured by the time it took them to climb four stairs. To do this, the company ran two trials, the second of which was an 18-month placebo-controlled trial. Their results clearly showed that boys who took the drug maintained their speed—i.e., strength—at climbing stairs, when compared with boys on placebo. But getting those results took 30 months across two trials.
What'll it take for every Duchenne drug to go for an accelerated approval?
The short answer? Better biomarkers.
Scientists and companies are aware of the need for biomarkers, ones that represent changes in muscle inflammation and cell health. A better biomarker could help us quickly see if drugs are working.
The long answer: DMD has been called the Mount Everest of diseases for a number of reasons, one of which is that it progresses quite slowly. Especially if you compare it to Spinal Muscular Atrophy type 1, many cancers, or ALS. In those diseases, survival is a meaningful, and binary, metric. If people taking a drug are still alive after two years when history suggests they wouldn't be, it's obvious the drug is doing something.
In Duchenne, you can't run a trial based on survival. People live decades (a great thing!). You can, however, run it according to loss of function, as measured by strength and ability-based tests, like the stair climb or walking tests. But those readings don't change all that quickly. Companies can't afford to run trials for 2-3 years, and no one would sign up to participate if they could be on placebo for that long either.
Biomarkers that better represent muscle health could resolve some of this. And they would—hopefully—enable effective anti-inflammatory drugs reach patients much, much more quickly.
Have a question about anything DMD-related? Send it in, and I'll try to report back :)
Thank you to Dr. Franson for chatting with me for this story!
Here's what else I'm reading:
First Day of a 'New Life’ for a Boy with Sickle Cell, by Gina Kolata. The rollout of a gene therapy is not so simple, and can be painfully slow—as the 20,000 Americans living with the most severe form of sickle cell are learning, more than nine months after the FDA approved two treatments. This profile of a 12 year old shows the bureaucratic tedium and immense toll of his wait and preparation for infusion day. (New York Times. Link to a gift article, so no subscription required!).
The Canary, by Michael Lewis. This time Michael Lewis takes on an unlikely coal miner-turned-mine safety innovator with the Department of Labor. Even if reading about mine safety isn't your idea of a good time, bear with me. It’s good. Really good. It’s part of a refreshing Washington Post series celebrating the successes of government, rather than its failures, as the media is wont to do. (In fairness, failures are usually more newsworthy and/or make for a better story.) I thought it was fitting this week, since I had the oft-critiqued FDA on the mind. (Washington Post)
Stay Curious and Keep Exploring, by Emily Calandrelli. I've been trying to find ways to engage my kids on the weekends. This book is fun, design-y and has clear instructions for experiments you can do with stuff you already have lying around the house.
Gaining access to drugs earlier is a double edged sword. It is of course important. But in the rare disease space, it can also mean taking a drug about which there is hardly any data. In DMD, the first three exon skipping drugs were approved on the basis of just one trial each, with a dozen or so participants. Follow-up efficacy still has not been submitted to the FDA.
Two reasons why it’s a faulty biomarker. The first is that we actually have no idea how much dystrophin enough to confer benefit. 2 percent? 5 percent? 10 percent? 20 percent? Jury's out.
The second reason is that none of the currently approved products result in full-length dystrophin. Current DMD gene therapies produce a “microdystrophin,” which is a protein about a third of the size of full-length. With exon-skipping, the dystrophin protein that DMD patients produce is probably longer than a microdystophin, but depending on a boy's mutation, the skipped dystrophin will be a little bit shorter than full length. And because it is not full-length dystrophin, we don't actually know how well it will function, how stable it is, etc.
If you take a peek at figure 13 on page 34 in this FDA review of Elevidys, you can see that there's no obvious correlation between levels of microdystrophin and change in NSAA.
A commonly used biomarker in DMD is an enzyme called creatine kinase, or CK, which is released into the bloodstream when muscles are damaged. It's used for diagnosis, including newborn screening, because almost all people with DMD have elevated levels of CK in their blood. However, CK is wildly variable depending on how much someone moves. It also declines as a person ages and they have less muscle.
another very informative article! keep it coming!!!