For the companies developing treatments for Duchenne muscular dystrophy (DMD), it’s not just drugs they have had to invent – it’s whole new ways to prove they work.
In 2007 the US Food and Drug Administration (FDA) told PTC Therapeutics there wasn’t yet a measure good enough for its drug to get approval for use. With little previous research to go on, the New Jersey, USA, company chose a test that tracks the change in how far young boys can walk in six minutes.
Although the trial that followed came close, it – perhaps inevitably – failed to reach the standard of evidence usually required. “Regulators like you to pre-specify the outcome, say, ‘Here’s the bet: we put our chips on Five’,” says Stuart Peltz, CEO of PTC. “However, in a brand new disease where trials have never been done before, how do you know what number to put [your bet] on?”
For doctors to be able to prescribe a drug, it usually has to be approved by a regulator like the FDA or the European Medicines Agency (EMA). The regulators scrutinise progress, from tests in cells and animals to phase I human trials testing factors such as safety and how long the drug stays in the body. Phase II trials normally look at whether the drug actually has a biological effect and how much is needed to see that effect. Then large-scale, long-term phase III trials study how well the drug actually works, although only around a fifth of phase II trials reach this stage.
In some phase II and most phase III trials, some of the participants are given the existing conventional treatment and an inactive placebo instead of the drug being tested. Companies then compare the main trial and placebo-control groups, looking for a ‘statistically significant’ effect – one that could only happen by chance once in every 20 identical trials.
“In general we ask for two phase III trials because every 20th trial is a false positive, showing an effect where there is none,” says Spiros Vamvakas, head of the scientific advice office at the EMA. But for rare diseases like DMD, where recruiting enough patients for a phase III trial is difficult, the EMA will accept smaller trials (preferably ones comparing the drug to a placebo).
Giving someone an inactive placebo instead of a real drug seems cruel when dealing with such a horrible disease, but Vamvakas emphasises the efforts made to look after the participants. All trials have a series of ‘stopping rules’ supposed to avoid patients being harmed – for example, switching everyone to the real drug as soon as an effect has been demonstrated.
In rare diseases, the EMA is also less strict about how companies test for statistical significance. Rather than all analyses being pre-specified, it allows the companies to reconsider their initial bet. For example, PTC was able to split up its trial to focus on the patients where the effect was most obvious. But Vamvakas stresses that not all drugs can do this: “If companies just recycle old product and give it to Duchenne patients because they’re desperate, then I’ll be quite strict in terms of believing an effect.” In PTC’s case, scientists are still debating the exact way the drug works, but the potential good is enough that it’s worth taking a risk. “These new DMD products are truly innovative developments that have really good science behind them,” says Vamvakas, “so I’m more willing to accept the weaker methodology.”
This approach is helping get PTC’s drug to boys who can still walk on the basis of phase II trials alone. But that’s little comfort to the patients who are too ill to walk and hence can’t do the one thing that’s currently used to measure progress; for them, there isn’t even a test that could show whether a drug is effective. Netherlands-based Prosensa had a taste of the challenge involved in testing older, weaker patients when pharma giant GSK ran a phase III, 186-participant trial using its drug. “One out of every four boys in the phase III trial had a baseline six-minute walk distance of less than 300 metres and was older than seven,” explained Prosensa CEO Hans Schikan. “That’s where you see huge variability in the outcome of those tests.”
Trying to deal with that variability is what imposes frustrating limitations on who gets the drugs. To get the best trial data, the developing companies have to be very careful about who they recruit. And the drugs can only be approved for those they’ve been shown to work for in trials.
“From the patient point of view, it’s difficult to accept,” says Vamvakas. “They think: ‘This is for Duchenne, whatever my stage is,’ but that’s not the reality. The evidence has to be taken from that particular stage of the disease – that is when we know it works. If we go too far, we end with confounding data and we cannot be so certain.”
Researchers are hunting for a test that can show a treatment has an effect in older boys, on top of all the other ways they’re pushing to get drugs to patients. There are many candidates: tests of lung and heart function or arm and finger strength, or imaging muscle deterioration using MRI. But because there’s so much variability in the effects drugs have in older boys, it’s not a given that there will be a test that can help them get access to a drug.
In that case, the focus will be on treating future generations of DMD sufferers, says Kay Davies from the University of Oxford, UK. “You do what you can for the current generation. If you could do something that allowed them to keep their arm function going, it would make a huge difference to their quality of life, although it may not extend it. But the aim has to be ultimately that you get in early, perform screening, and identify those that can maximally benefit from these drugs.”
The way the DMD community is tackling such challenges serves as a test case for other rare diseases, Davies adds. “How you set up a small clinical trial, how you develop new biomarkers, and the new technologies that are being put in place… All of these things are going to be helpful in other fields. It’s a great paradigm.”
Kay Davies sits on Mosaic’s Editorial Advisory Group and is Deputy Chair of the Wellcome Trust, which publishes Mosaic.