Memorandum

Novo Nordisk announces decision to resubmit Tresiba (insulin degludec) and Ryzodeg (insulin degludec/insulin aspart) in the US – March 31, 2015

Executive Highlights

  • Novo Nordisk announced on Thursday that it has decided to resubmit Tresiba (insulin degludec) and Ryzodeg (insulin degludec/insulin aspart) to the FDA within the next month based on interim results from the DEVOTE CVOT.
  • A Class 2 resubmission means that the FDA should provide a decision towards the end of the year, barring any unforeseen surprises.
  • The company arrived at this decision with input from the FDA, giving us greater confidence that we might have otherwise that Tresiba can make it through the regulatory process this time around.

Novo Nordisk announced on Thursday that it has decided to resubmit its ultra-long-acting basal insulin Tresiba (insulin degludec) and Ryzodeg (insulin degludec/insulin aspart) to the FDA within the next month based on interim results from the DEVOTE CVOT. The trial accumulated enough events for the prespecified interim analysis in January, ahead of the company’s expected timeline that had itself been advanced multiple points during the past year. Following a novel protocol designed with FDA involvement (see below), it was a small firewalled team within the company made the decision to resubmit with FDA input. To preserve the integrity of the DEVOTE trial, which is expected to complete in 2H16, Novo Nordisk management remains blinded to the interim data. In addition to the DEVOTE data, the resubmission includes a safety update from clinical trials and postmarketing surveillance of Tresiba. The press release stressed that these interim results come with a higher level of uncertainty than the final results due to the lower number of events; however, the fact that the company decided to proceed with resubmission suggests that the interim results are not particularly worrisome. The FDA is expected to accept the NDA or issue an Incomplete Response Letter within a month of resubmission – given their active involvement in the resubmission decision, we would be shocked if the Agency did not accept the application. From that point, a Class 2 Resubmission is expected to take six months; an Advisory Committee meeting is possible, but given the FDA’s desire to keep the interim data blinded we are not sure how this would go.

As a reminder, Novo Nordisk outlined its planned approach to the DEVOTE interim analysis during its 3Q14 update following the August 11th FDA hearing on interim CVOT data disclosure and direct discussions with the Agency. At the time, management stressed that resubmission based on interim results was not a given and would depend heavily on FDA input – this gives us greater confidence than we might have otherwise that Tresiba can make it through the rest of the review process, as ideally any worries on the Agency’s part would have been conveyed to the company already. If that is the case, it may help convince other companies that using interim trial data to support regulatory filings is a viable practice, an argument that Novo Nordisk has made repeatedly since the August hearing. When this process is over and done with, we will be interested to hear feedback from Novo Nordisk management (both those inside and outside the firewalled group) and the agency on how smoothly things went.

  • Excitingly, this resubmission also opens the door for an FDA submission of the potentially transformative GLP-1 agonist/basal insulin combination Xultophy (insulin degludec/liraglutide). However, Novo Nordisk has not yet indicated whether that would occur immediately or after a final Tresiba regulatory decision. If the company takes the firewalled group’s decision to resubmit the Tresiba NDA as a vote of confidence in the data (as a reminder, it cannot actually see the interim results), it may feel more confident in submitting Xultophy sooner rather than later. The pressure is already on, as Sanofi has stated that it plans to file its own GLP-1 agonist/basal insulin combination LixiLan (insulin glargine/lixisenatide) in the US and EU in 4Q15. As a reminder, Xultophy is already approved in Europe and is seeing its first launches.
  • If Tresiba is approved in the US towards the end of this year, that would place it a few quarters behind the approval of Sanofi’s Toujeo (insulin glargine U300). That approval occurred a month ago, and Sanofi plans to launch Toujeo in the US by the beginning of 2Q15. The EMA is expected to provide a final decision on Toujeo in 2Q15, while Tresiba has been approved in Europe since 2013. The other next-gen basal insulin that was previously closest to regulatory filing that had been Lilly’s novel basal insulin peglispro (BIL), but that candidate recently suffered a major setback that pushes FDA and EMA filing back until 2017 at the earliest. There is a group of once-weekly basal insulins slowly approaching, but those are either in early clinical development or preclinical development.
  • The broader issue of interim data disclosure from CVOTs remains very much unresolved. While Novo Nordisk clearly believes that submissions based on interim data can be done without compromising trial integrity or raising FDA ire, other companies have had very different responses. Most recently, Orexigen received a reprimand from FDA Director of the Office of New Drugs Dr. John Jenkins for disclosing data from the 25% interim analysis of the Light Study (the initial CVOT) for Contrave (naltrexone/bupropion) that appeared to show cardiovascular benefit, even though that trial was not part of a post-marketing requirement. J&J had to initiate a second CVOT (CANVAS-R) for Invokana (canagliflozin) due to concerns about the disclosure of interim data from the first trial (CANVAS) to support approval. Sanofi withdrew its NDA for Lyxumia (lixisenatide) in 2013 to avoid the risks of interim data disclosure from the CVOT ELIXA, for which we recently saw topline data indicating cardiovascular neutrality (i.e.: safety).
    • Overall, we feel there is a significant need for clearer FDA guidance on how interim data should be used in the regulatory process. It is conceivable that the sort of approach Novo Nordisk is taking could become a more standardized way to facilitate (relatively) efficient drug approvals within the limits of the 2008 CV Guidance. However, we do not believe this represents anything close to an ideal solution, as it involves a great deal of uncertainty on the part of manufacturers. Essentially, up until the point of this announcement, Novo Nordisk’s senior management did not know whether Tresiba and Ryzodeg would be resubmitted now or in late 2016 – this is a lot to ask. It is also difficult to know how enforceable the “firewalling” process is, although we have few doubts that the major companies in diabetes would follow all the rules.  
    • We continue to hope that the FDA will revisit its broader CVOT requirements in the near future. We are interested to see if any suggestions to that effect emerge during the upcoming April 14th EMDAC meeting on the results of the SAVOR and EXAMINE CVOTs.

-- by Emily Regier, Manu Venkat, and Kelly Close