Sarepta Setback: FDA Doubts About DMD Biomarker Crushes Early Approval Dreams

FDA’s rejection of dystrophin as a biomarker in Duchenne muscular dystrophy and doubts about the six-minute walk test endpoint put Sarepta back at square one.

Sarepta Therapeutics Inc. is facing the daunting task of designing a new, placebo-controlled trial of its Duchenne muscular dystrophy therapy in a very limited population without a clearly valid endpoint following FDA’s rejection of its plan for gaining early approval of eteplirsen based on a surrogate marker.

The company announced Nov. 11 that FDA had informed it that the failure of other drug programs had cast doubt on the value of dystrophin as a biomarker. DMD, a muscle-wasting disease that affects young boys and gradually diminishes their ability to walk, is caused by mutations in certain exons that makes them unable to produce the protein dystrophin. Eteplirsen, as well as GlaxoSmithKline PLC/Prosensa Holding BV’s drisapersen, is an RNA antisense drug that uses exon-skipping technology to restore expression of the protein. Sarepta had intended to use dystrophin levels as a surrogate marker to submit for accelerated approval (Also see "Sarepta’s Latest Bid In Race For DMD Indication: Submit NDA With Phase IIB Data" - Pink Sheet, 24 Jul, 2013.).

The agency cited the Phase III failure of GSK/Prosensa’s drisapersen, which has “a similar mechanism of action [and] was reported to be negative, despite increased expression of dystrophin,” the agency said in pre-meeting comments to the company (Also see "DMD Development Landscape Shifts After Prosensa Failure; Sarepta Continues Strong" - Pink Sheet, 7 Oct, 2013.) (See sidebar).

“The disconnect between increased expression of dystrophin and clinical efficacy for drisapersen, combined with previous negative reports for PTC124 [PTC Therapeutics’ ataluren failed in a Phase IIB trial in 2010], another drug thought to act by increasing dystrophin, raises considerable doubt about the biomarker, and consequently, its ability to reasonably likely predict clinical benefit,” the agency said.

“To be clear, we will be pursuing a more traditional path of approval while we concurrently try to persuade the FDA to reconsider the potential of an early filing strategy for eteplirsen,” CEO Chris Garabedian said during a Nov. 12 call with investors.

The company met with FDA on Nov. 8 to finalize the trial design for the confirmatory study to support the accelerated approval submission.

Garabedian noted that executives were pleased with the level of involvement from FDA, which implies the agency considers the program a priority. The meeting was extended from 60 to 90 minutes and there were 25 participants, including CDER Director Janet Woodcock, Office of New Drugs Director John Jenkins and CDER Deputy Director of Clinical Science Bob Temple, as well as the office and division directors.

However, FDA used the meeting to communicate “unsolicited comments related to a change in their thinking on the regulatory status of eteplirsen” since their meeting in July, based on the drisapersen trial and the earlier ataluren trial, as well as recently published natural history data that raised questions about the expected decline in DMD patients.

“We do not fully understand nor do we agree with many of the FDA's comments and new perspectives as we believe our dystrophin data has been well characterized and compares favorably to and is differentiated to other dystrophin producing technologies, including drisapersen and ataluren,” Garabedian said.

Tripped Up By Walk Test

The six-minute walking distance test has been a standard, accepted endpoint in other settings, like pulmonary arterial hypertension, and while there is no regulatory precedent for DMD, it is being used in all of the trials across sponsors. Sarepta believes it is “the most validated, reliable and predictive endpoint” for the disease. “We don’t believe there are better validated endpoints in the DMD population,” Garabedian said. Pulmonary function is being assessed, but “anything beyond that has not been very well characterized.”

But FDA is reconsidering the use of the six-minute walk test to support the NDA because of the recent natural history data, which indicated that patients who can walk over 350 meters at baseline generally will be stable, not have the decline that the company predicts. Sarepta had submitted historical data indicating a decline of 76 meters to 83 meters, consistent with the drisapersen and placebo arms in the GSK/Prosensa study.

The new data led the agency to call for a re-examination of study enrollment criteria and endpoints for the confirmatory trial.

“The agency has reiterated their demand for a placebo-controlled study, because in their view ‘efficacy endpoints in DMD are effort dependent and susceptible to bias and the natural history is highly variable and has recently improved with steroid use and advances in ancillary care,’” Garabedian noted.

Switching to a placebo control would require doubling the planned sample size for the confirmatory trial “and could pose problems obtaining IRB approval to obtain two surgical biopsies with general anesthesia in placebo patients in order to protect the blinding,” he added.

As that “would significantly delay” recruitment of an ambulatory age-appropriate exon 51 amenable population in the U.S., FDA suggested exploring additional subpopulations and endpoints “and were open to discussing novel approaches to designing a placebo-controlled study, although the type and extent of these endpoints and subpopulations remains open.”

Sarepta has scheduled a follow-up meeting with FDA to discuss the confirmatory study design later this month. Between the need for a placebo control, questions about dystrophin as a biomarker and the request to discuss alternate endpoints and subpopulations, the company said, dosing in the confirmatory trial will not begin “until at least the second quarter of 2014.”

An Ongoing Dialogue

At the moment it is unclear what alternate endpoints could be used, Garabedian said. “They indicate that they would still accept six-minute walk in an appropriately powered study but ... they understand that the feasibility of enrollment could be hampered because of their requirement of a placebo control. So this is where we don't believe we can easily or at all potentially find enough patients to do the placebo-control that they want in a timely manner or may not be able to fully recruit that many patients to power” the analysis.

One option would be to enroll patients of different statuses – “the younger patients, older patients, ambulatory patients that might require us to look at different endpoints in a six-minute walk test,” Garabedian said. “Maybe there is a way to combine these endpoints. Again, we think that's a challenge because these are generally unvalidated endpoints and [wonder] how to mathematically combine them and then power a study, but we can have that discussion with the FDA.”

Garabedian stressed that the company is not giving up on re-convincing FDA of its original plan. During the Nov. 8 meeting, the Sarepta team focused on describing how differentiated eteplirsen is from drisapersen, in the hopes of dissuading FDA from extrapolating from the failed drisapersen trial. “For example, we highlighted how we have data that shows our chemistry has multi-fold, better exon-skipping, that our sequence versus their sequence, all things being controlled, has multi-fold activity. We talked about the higher dose we used in our studies, five- to eight-fold higher dose in our clinical studies and we shared contrasting data on RT-PCR dystrophin as best as we could do apples-to-apples side-by-side,” he said.

He noted that FDA has requested additional data from the failed drisapersen trial and has not completed its analysis of that trial.

“We are continuing our scale-up manufacturing activities and we had a productive CMC meeting with the FDA, where they demonstrated the flexibility we were hoping for on stability data, comparability and process validation,” Garabedian noted. "While the FDA decision to reconsider the potential for an NDA filing in 2014 will require us to re-think the need for scaling up for commercial demand, we have not altered our plans at this time to prepare for large-scale production in the hope that the agency may reconsider being opened to an early NDA filing on our existing dataset.”

Ramifications For The Rest Of The Pipeline

The outcome of the eteplirsen discussions will have far-reaching impact for Sarepta, which has a stable of exon skipping candidates – and the entire program has been “largely predicated” on dystrophin being accepted as a biomarker.

If FDA does decide dystrophin doesn’t work as a biomarker, “it would render us unable to get other drugs approved for rare exon because we would not be able to power studies on six-minute walk or these other endpoints,” Garabedian acknowledged.

“This is why we need to work with the FDA closely to come to an agreement and assessment of how to capture dystrophin in a way that they believe is validated, so that we can have streamlined approvals of the other exon [candidates],” he said.

Still, during the presentation of the quarterly results Chief Medical Officer Edward Kaye noted that Sarepta is on track to submit two or more INDs by the end of 2014, including its exon 53 candidate.