A Rebuttal Of The Sarepta Bull Rebuttal

| About: Sarepta Therapeutics, (SRPT)

The bulls certainly came out in force after my Sarepta Therapeutics (NASDAQ:SRPT) short article last week, offering up a new round of scientifically ungrounded arguments to keep the stock in orbit.

As I stated before, the clinical trial data for eteplirsen warrants further study, but is not nearly compelling enough for an accelerated approval. As I'll show later on, this is the general consensus among scientists that are going to make the call at the FDA, despite what financial journalists may say.

So in a rebuttal to the recent bull articles, I'd like to expound upon a couple of points that got a good bit of feedback in my earlier article. I hope this helps clarify why the scientific community isn't jumping on the Sarepta bandwagon - at least not until after a robust Phase III trial is performed showing the drug provides a clinically meaningful benefit to DMD patients.

You said before that the 6MWT is a highly variable endpoint. What does that mean?

The 6MWT is a measure of performance. Biologists tend to think about performance in two ways: Average and Maximal performance. Why? Because there are a variety of factors that influence how one performs on a given task at a given time.

Think about it this way. Athletes don't always go out and perform their best. Sometimes Usain Bolt blows away the competition and other times he loses the occasional race - despite probably being the best sprinter on the planet. The truth is that individuals, even athletes, can perform sub-optimally at any time, and the causes of sub-optimal performance are numerous.

So when biologists measure an individual's performance, we need to account for intra-individual variability, or sub-optimal performance. To do so, we tend to measure an individual's performance multiple times on the same task with breaks in-between trials. From a statistical purview, this gives us the ability to analyze both an individual's average among trials, and their best (maximal) effort.

The problem with this measure in boys with DMD is that they often fall during the walking trials. So safety is a big issue, making repeated measures a tad problematic.

Even so, the current protocol for the 6MWT is to test the boys twice, one-week apart. Although not as rigorous as one would like, this protocol is fair enough given the potential safety issues at hand. And according to the methods reported in Sarepta's peer-reviewed work, this is the protocol they followed, more or less.

Just so there is no doubt over the issue, here is exactly what was said about the 6MWT in Sarepta's studies:

The 6MWT was administered using the protocol established for DMD.

The "established" protocol is the McDonald et al. (2010) paper published in Muscle Nerve.

Moreover, Sarepta's article indicated that they used "maximal" performance in the analysis, not average.

With that in mind, you need to pay attention to the standard deviation for each group. What you'll see is that standard deviation is rather high for the 6MWT, which is probably due to differences in leg length among patients. Remember, leg length is roughly equivalent to stride length in both humans and animals. This shows the variability in the 6MWT as an endpoint, and its underlying cause is likely morphological (i.e., height, leg length, etc.).

Keeping this in mind, if you look at the standard deviations for height, it also shows a fairly large standard deviation within trial groups as well. In other words, bigger boys can walk faster because they have longer legs and stride lengths.

What does this mean?

It means that you must correct for size in the analysis. To deal with this size-effect, you need to remove it statistically, which can be done by either using leg length as a covariate, or calculating residuals (variation in the 6MWT independent of size). To be clear, such an analysis was not reported for Sarepta's dataset. They sort of dealt with this issue by nesting subject within treatment, but that doesn't explicitly handle the size issue. Individual effects could be motivational issues, not biomechanical determinants of maximal performance.

Why didn't they explicitly account for size even though they are fully aware of it? Again, their sample size is just too small. Such a small sample size does not permit a proper calculation of residuals, especially using least-squares techniques such as ANCOVA and a highly variable measure of performance.

Bottom line: What matters is size-adjusted performance on the 6MWT. This would directly account for differences in stride length among patients, and lead to lower standard deviations within and among groups. Simply put, we can look directly at the effects of eteplirsen on walking performance independent of a boy's size.

Why is the 6MWT the appropriate primary endpoint?

While some folks, including Sarepta, have argued that dystrophin production should be the primary endpoint, it is not widely accepted in the field. A recent peer-reviewed article on DMD has this to say about the 6MWT:

The 6MWT was been shown to be an integrated global measure of ambulatory function in DMD that is influenced by decreased lower extremity strength, biomechanical inefficiencies during gait, diminished endurance, and compromised cardio-respiratory status.

In other words, it's pretty telling about a lot of things, whereas dystrophin production can't even predict ambulatory ability. That's why the 6MWT is the gold-standard in DMD studies, not dystrophin production.

You're way off base with your bearish outlook. The FDA won't care about sample size or a different primary endpoint.

Oh yeah? Make sure to read this recent review from none other than Nature Drug Reviews. The article contains particularly interesting tidbits such as:

The key challenge for Sarepta now is the small size of its development programme. The company only enrolled 12 patients because it had a limited quantity of eteplirsen when it launched the trial.

Moreover, the clinical value of dystrophin production - as opposed to the clinically grounded 6MWT end point - remains a hot topic of debate, with observers keenly awaiting input from the FDA.

Simply put, professional scientists are doubtful about Sarepta's chances for early approval. And not having enough of a drug to increase your sample size is no excuse at all.

You said that oligonucleotides have a poor track record? What does that mean?

For over three decades now, oligonucleotide based drugs were supposed to revolutionize medicine. Yes, three decades. To my knowledge, there are only three FDA approved AO drugs still on the market, with one gaining approval just last January and its being closely monitored for safety issues. Compared to the large number of AOs tested thus far and the limited indications for the handful of FDA approved AOs, that isn't exactly a revolution.

But I did a lit search and it was full of articles trumpeting the power of AOs?

I would recommend checking the author affiliations of these articles. Companies like Isis Pharmaceuticals (ISIS), Prosensa (NASDAQ:RNA), and Sarepta have a vested interest in keeping investor interest high. Sadly, however, these drugs have a failure rate close to 99% in clinical trials. Prosensa's drisapersen was only the latest to fail.

And if you dig a bit deeper, you will see that every few years a review article pops up trumpeting the next generation of AO drugs. Predictably, the review is followed up by a new round of clinical failures. Indeed, some of the recent reviews were particularly optimistic about drisapersen, and we all saw how that unfolded.

Ok, so what would make you a Sarepta Bull?

Honestly, nothing. The stock has already built in most of the potential value for an FDA approval, and a commercial launch will require significant resources. If eteplirsen gets the green light, Sarepta plans on ramping up its clinical efforts for its remaining DMD drug candidates, meaning their cash burn rate will increase dramatically. So even in the unlikely event of an accelerated approval, Sarepta is going to have to find big chunks of money to commercialize eteplirsen and advance their clinical programme. That means a sizeable secondary would be coming down the pike. In sum, the potential remaining upside in Sarepta is far outweighed by its risks.


Overall, you can see that it's not just some crackpot Seeking Alpha authors that are doubtful. Nature is a pre-eminent scientific journal, and the article referenced above shows that the concerns over the trial design, endpoint, etc. are pervasive in the industry. If financial journalists and investors want to ignore the sentiment within the scientific community, they do so at their own peril.

Perhaps the FDA will gloss over these issues and happily approve eteplirsen next year. Stranger things have happened. But if you watched the Amarin (NASDAQ:AMRN) Advisory Committee Meeting recently, where they thrashed the trial design and even questioned the type of placebo used, I wouldn't bet the farm on it. Moreover, I have a hard time believing the FDA is going to put themselves in the position of having to pull eteplirsen if its confirmatory Phase III trial fails. That would be a PR disaster.

By contrast, I suspect the FDA is going to wait until there is more data in order to make a decision. When that happens, it's going to have a dramatic impact on a company whose shares are already trading close to 15X their book value, and most of that valuation is based on the potential of eteplirsen.

Worse still is that if eteplirsen fails, it essentially scuttles the remainder of Sarepta's exon-skipping DMD pipeline. What is Sarepta worth then? I highly doubt anything near $1.3 billion.

But go ahead, defend a biotech bubble stock, and its significant risks.

Disclosure: I am short SRPT. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.