I am writing to you about the forthcoming meeting to consider the marketing application for eteplirsen, a morpholino-oligo that induces exon-skipping to generate functional dystrophin protein and therefore, it is hoped, improved outcomes for patients with Duchenne Muscular Dystrophy (DMD).
I should, before continuing, declare my conflict of interest. DMD is a terrible disease, and understandably the prospect of a first treatment able to slow or arrest the progress of the disease arouses strong emotions among patients and their families. In voicing their support for the approval of eteplirsen, the conflict of interest of these patient advocates is clear. Similarly, the vested interest of the company that owns the rights to eteplirsen, Sarepta Therapeutics (NASDAQ:SRPT), is straight-forward and obvious. Less clear, but nonetheless relevant, are the strong links between doctors who have been treating DMD patients with eteplirsen and the company behind the drug. In writing to voice their support, these doctors did not fully disclose this conflict, so in setting out the case against approval I do not want to make the same error. My conflict of interest is that I am a taxpayer, and it will fall to me, and people like me, to pay for the use of this drug should it be approved.
My concern is that eteplirsen may be granted an accelerated approval because of the undue pressure being brought to bear by those parties with an understandable interest in seeing it approved for use: patients (and their families) and the company (and its consultants), despite not meeting the usual requirements for such an approval.
Let me be absolutely clear: I do not oppose approval of eteplirsen because I believe it is either unsafe or that it does not work. Indeed, the available data is encouraging on both counts. Crucially, however, the studies that have been performed to date do not reach the generally accepted bar for demonstrating efficacy. It is possible, but far from certain, that when further data has been collected this agent will prove to be the long-sought-after breakthrough for DMD patients.
The key data that supporters of this agent point to is a year-long study in 12 boys with DMD. This study has the gold-standard double-blind placebo-controlled design, although the placebo patients were randomized to drug after 24 weeks complicating the assessment. While small studies are unavoidable in rare genetic diseases, approving a drug on the basis of just eight treated patients would set a new precedent.
The study had two key end-points: the first was levels of dystrophin protein in muscle biopsies, and the second (and more important) was the change in muscle function, measured by the 6 Minute Walk Test (6MWT). The study clearly showed that treatment with eteplirsen resulted in exon-skipping, although the amount of exon-skipped dystrophin was still much less than the amount of mutant dystrophin in the muscle. It also showed an increase in the percentage of muscle fibres that stained with an antibody against dystrophin (although the use of fibre percentage, rather than a quantitative assessment of the amount of functional dystrophin probably exaggerates the apparent magnitude of the effect). The changes in fibre percentage were statistically significant beyond 12 weeks of treatment, and it while one could argue about the relevance of this ‘fibre percentage’ as a chosen end-point, it would be difficult to argue there has not been a clear demonstration of increased functional dystrophin protein in the muscle of the boys treated with eteplirsen.
However, the same cannot be said of the key 6MWT data. Unfortunately, two of the boys who received the lower of the two doses of eteplirsen lost the ability to walk very quickly after being recruited into the trial. While the treatment increased dystrophin-positive fibres in these boys, just as it did in the others, this did not prevent loss of ambulation. The authors of the paperdescribing the study argue, quite reasonably, that there was little prospect of any drug being able to arrest the rapid decline in these boys, where the disease was more advanced, and that they should therefore be excluded from the analysis. Having excluded them, they find a significant improvement in the 6MWT among the remaining six boys compared to the placebo.
These kind of statistical gymnastics give the best possible estimate of what is happening when boys with DMD are treated with eteplirsen, given the paucity of available information. They provide a suggestion that the drug does indeed work on this key functional end-point. But we have to remember it is a best-guess based on just 6 patients selected, post-hoc, from 8 that were treated. However good the justification for making that selection, it weakens the conclusion (which was already weakened by the small numbers involved) still further.
The risk of over-interpreting the findings from the 6MWT in this study is illustrated by the trajectory of the placebo/delayed treatment group. Almost all of the loss in 6MWT distance occurred in the 8 weeks between week 24 and week 32, in marked contrast to the complete lack of decline in the first 12 weeks of the study (which was just 3 months earlier). What that illustrates is the noise in the measurement – there is not a steady decline, but rather a step-wise loss of function (at least when assessed using the 6MWT). Exactly when a given patient shows a marked decline varies (even though the overall pattern of decline over several years is broadly similar for all the patients with the disease). This raises the concern that the eteplirsen-treated boys just happened to avoid a measured decline in the study period. To argue that cannot be true would be a classic example of “having your cake and eating it”: to select out data from two boys who showed an unusually large decline, while allowing to remain data from boys who showed an usually small decline simply because “that’s what we expect the drug to achieve”.
None of this is to say that eteplirsen doesn’t work. The data is consistent with a complete abrogation of functional decline for a period of a year – a miraculous outcome. But it is also consistent with a statistical fluke. Quite simply, there is not enough information available to distinguish the two with any kind of reliability.
What then should the FDA do?
If it rejects the pending application, it risks denying boys with DMD an effective treatment. If it approves, then it risks saddling healthcare providers – and ultimately taxpayers – with big bills for something that doesn’t work. On the face of it, it is an unappealing choice.
But the reality is that regulators are always denying patients “potentially effective” drugs. That is the inevitable consequence of having regulators at all. While the evidence is accumulated that an agent is safe and effective, patients are denied access – so that, in return, we can have confidence that the medicines we take are safe and effective. Without a regulator, we are back in an age of snake-oil salesmen.
And the regulator is not just there to protect patients. It is there to protect the precious, and always limited, healthcare budgets – which is where, as a taxpayer, I have a vested interest in the outcome of the decision, just as patients and their families do. Indeed, avoiding paying for a drug that may not work allows that same money to be spent on other drugs that do work that might otherwise be denied to patients on economic grounds. The emotional arguments, therefore, simply cannot be applied – someone will miss out on beneficial treatment, and approving eteplirsen on the current dataset only increases the risk that someone will be denied an effective drug in order to pay for a placebo.
What about an accelerated approval or a compassionate-use exemption?
Both are dangerous constructs. With an accelerated approval in hand, will it be possible to recruit young DMD patients to a randomized trial needed to prove the benefit of eteplirsen once and for all? Listening to the lobbying patient groups, it’s hard to believe any of them will be volunteering to receive placebo. And others will be discouraged from developing better therapies by the lack of clarity as to how good eteplirsen really is. In the event the current data owe more to statistical fluke than genuine efficacy, the long-term interests of DMD patients may be substantially harmed by an accelerated approval. It is not simply a matter of “no cost compassion” – the costs, not only in financial terms, are real and affect taxpayers and DMD patients alike.
We can, and should, have an open debate about what we want the drug regulation regime to achieve for us as potential patients and as actual taxpayers. But we should not vary the standards applied to different drugs on the grounds of need – doing so is bad for patients and taxpayers alike. Applying the current standards dispassionately would see eteplirsen fall below the bar for approval on the available data.
Read the original article here
RxCelerate Ltd is an outsourced drug development platform based near Cambridge, UK. We specialize in delivering an entire road map of drug development services from discovery and medicinal chemistry through to formal preclinical development and clinical up to Phase IIa. In the last five years, we have witnessed dramatic changes in the drug development …