Obesity drug maker
That’s amazing news, and it sent Orexigen shares up 46% today.
Except that it might not be such good news at all. These data were not supposed to be released publicly because the study is still ongoing, as per an agreement with the Food and Drug Administration. And Orexigen decided to release the data without consulting Steven Nissen, the chair of cardiology at the Cleveland Clinic and the lead investigator of the study whose imprimatur had previously meant a great deal to the small biotechnology company. (Nissen is best known for calling out what he believes are unsafe drugs, including Vioxx, from Merck, and Avandia, from
Nissen says, in a statement:
The interim results of the Light Trial were released to a small unblinded Orexigen team by the Data Monitoring Committee under strict rules that allowed access for the purposes of regulatory filing. No public release of these data was anticipated and was not approved by the DMC or Executive Committee. It would not be appropriate for me to comment on preliminary interim data derived from an ongoing outcomes trial that is still underway.
The study that provided the positive results was a Nissen brainchild. He and University of Washington statistician Thomas Fleming, who was heading a safety committee for the Orexigen study, had proposed to the FDA a new strategy for testing the safety of diabetes and obesity drugs. A large study testing their use in preventing heart attacks and strokes would be started. If the risk was less than 80% elevated at the 25% mark, the drug could be approved, but details of the results would be kept secret. The reason is that disclosing the results for an approved drug could cause people to drop out of the study, meaning that good data on the drug’s safety will never be collected.
Last August, the FDA held an all-day advisory committee meeting on how to deal with these disclosures of interim data. Here’s what Fleming who chairs the safety committee (known as a Data Monitoring Committee) said at that meeting, according to a transcript:
In August of 2012, Dr. Steve Nissen and I sent a letter to FDA and in that letter, we indicated 0that in current ongoing cardiovascular safety trials where consideration of marketing approval may be based on interim data, access to unblinded data should be limited to a small core group in order to preserve the integrity of the trial, of the final analysis of the ongoing trial.
Within sponsors, a data access management plan should be created to ensure that only members of an unblinded team have access to unblinded interim data with the composition of this team determined with the input from the FDA review division, the DMC, and the trial's academic executive committee. Data supporting the sponsors' submission should be tightly restricted to the members of the unblinded team until after completion of the fully study. Unblinded team members should not participate in the subsequent conduct, and management [should remain unaware of the results] of the study until the final database is locked and the trial is unblinded.
Doesn't Orexigen's data release sound like the exact opposite of what Fleming is describing? (Fleming did not respond to requests for comment.)
The FDA advisory panel was asked “do stakeholders agree that disclosure of detailed analyses (such as point estimates of hazard ratios and the associated confidence intervals) could undermine the integrity of an ongoing trial and jeopardize its continuation, potentially eliminating or substantially delaying the agency’s ability to obtain needed long-term safety information? “ Their answer was, pretty much, yes. (You can see slides from the meeting here, and a webcast is available here.)
This morning, in an 8-k filed with the Securities and Exchange Commission, Orexigen disclosed that it had filed for a patent on using Contrave to prevent cardiovascular events, and included the following information: There were 35 major cardiovascular events in the 4,455-patient Contrave group, compared to 59 in the 4,450-patient placebo group, with a hazard ration of 0.59 and a confidence interval of between 0.39 and 0.9. (That’s a 40% reduction.) Orexigen also included the results, including hazard ratios and confidence intervals, for heart attacks, strokes, and CV deaths. In other words, this is exactly what the FDA did not want Orexigen to do.
Orexigen's disclosure could result in patients dropping out of the study. If you have a 50% chance of getting placebo, why not be sure you get the medicine that seems to give a benefit? The results might not be nearly as rosy when the trial finishes -- Fleming's DMC could have recommended stopping the study, and apparently didn't. If the results do hold up, the FDA isn’t going to ignore the data or punitively take Contrave off the market. But at the very least Orexigen has endangered what was an accelerated path to market for diabetes and obesity drugs for other companies. If the FDA is really worried about such disclosures, it may not allow other companies to run studies in this way. This move likely damages Orexigen’s relationship with the FDA and could hurt its relationship with its marketing partner, Takeda.
Orexigen did not respond to repeated requests for comment. It’s likely that the company will say that it felt it had to file for a patent, and that it had to disclose that fact. It seems equally likely, no matter what Orexigen said, that it wanted to boost its stock price. This was a mistake on the part of Chief Executive Mike Narachi, and potentially a big one that could hurt biotechnology companies for years to come.
Update: Adam Feuerstein points out that the data are in the patent. This means that when the patent was published, the data would be disclosed as well, which explains how this happened. It doesn't stop it from being a mistake -- and even if it was unintentional on Orexigen's part, it endangers the whole approval mechanism they used to get Contrave approved.
