Two weeks ago, the NIH ran a workshop to examine the safety of several widely used diabetes drugs called GLP-1 inhibitors and whether a definitive link can be established to acute pancreatitis and pancreatic cancer, which were the subject of recent studies that generated considerable controversy. The outcome was inconclusive, but the FDA may want further studies. Meanwhile, the American Diabetes Association called for drugmakers that sell these meds to release patient-level data that can be used for an independent review (back stories here, here and here). To what extent the drugmakers - Merck, Bristol-Myers Squibb, AstraZeneca, Eli Lilly, Novo Nordisk and Boehringer Ingelheim – will comply remains to be seen. With the exception of GlaxoSmithKline, the pharmaceutical industry has resisted releasing such data over concerns that proprietary information will be compromised. We spoke with Alan Moses, corporate vp and global chief medical officer at Novo Nordisk, about the safety of its Victoza medicine and the ADA request...
Pharmalot: I’m curious to know what you think the ADA is asking for.
Moses: This is being defined as we go. The first request, I’m interpreting, is really an expedition into what might be available, what sponsors will be to share with an independent body that the ADA, through some as-yet undefined mechanism, might identify as an arbiter of the data is available. So they’re basically asking companies... here’s a spectrum of everything we may want – and it’s a fairly inclusive list that includes specific inquiries based on specific fields of a database relating to specific issues. And I think it’s very much just being formulated. And I believe that there intent is, once they have a sense of what’s possible, to then define an RFA that will be used to inform the field. I think the purpose here, and it’s an appropriate purpose, is to have an independent review of available data to cast aside some of the emotions that have been injected into this discussion and come up with an unbiased assessment of risk with DPP-4 inhibitors and GLP-1 receptor agonists. That’s my interpretation.
Pharmalot: Let’s carve out what’s being asked or meant. Anonymized data, for instance. That’s a contentious point of debate. And Glaxo is, so far, the only company to agree to provide patient level) data (see this). I’m curious to know how far you’re willing to go to provide data. If the ADA were to go in the direction of making this more publicly available – beyond just the organization chosen to conduct the review – is that something you see as a possibility?
Moses: There are two levels of discussion here. One is specific to these agents and pancreatic safety and the other is data transparency or disclosure. I would say as a company that Novo Nordisk is supportive of the concepts behind data disclosure, sharing the data, but recognizing there are limits in terms of how much or what kind of data can be shared, mostly for the protection of the individuals involved, because even anonymized data can become non-anonymous with the right kind of probing, depending on what kinds of trials were done and the number of individuals involved. And we’re very conscious of that. In Europe, the climate is even more restrictive, in terms of legislation in certain countries in regard to what can be made available, because of privacy protection issues.
And you have to be careful not to compromise intellectual property or property of commercial value that may, at first blush, may not be obvious when you look at the data. So I think as long as there are reasonable protections in regard to those things, we certainly are open to more sharing rather than less sharing, and more than has been shared to date. We’ve been supportive actually, in some respects, leading discussions with EFPIA concerning the EMA discussions, which are being prompted by the European parliament. So it’s a really complex social-political issue in Europe right now. And it’s been promulgated by the fact that there’s a perception some companies have not shared data that should have been shared in regard to safety issues. As always, when there’s a transgression, there tends to be an over-exuberant response to try to correct the situation. And I think our goal is to just make sure there are not unintended consequences that can, in themselves, harm the public in ways they don’t necessarily expect.
Pharmalot: Of course, the ADA doesn’t really have any authority, so it ends up being a voluntary effort. So the cynic might say, ‘what good does it accomplish?’ And this (review) will take a certain amount of time that will elapse.. Are we going to be left with a bunch of unanswered questions?
Moses: It’s all about the quality of the data… I think one of the interesting revelations at the NIH meeting was the presentation by the two presenters from the FDA. One, a non-clinical guy, if you will, got up there and said, ‘look, there are some 18,000 animals and multiple species, including 2,500 non-human primates, that have been studies and for which the data have been submitted as part of regulatory files. There are 20 compounds in this area. That’s a lot of data. And we have found nothing in the data to suggest a clear relationships between these classes of compounds and evidence of pancreatitis or pancreatic cancer. I was a little bit surprised and reassured that they made such a bold statement in terms of what they studies and were unable to establish a relationship.
For me, the fact that they would stand up at a public forum and make that straightforward statement was quite reassuring. Does that absolutely rule out any risk? No, it does not. You can never rule out the negative. But what I think it does do is say that if this is an issue, it’s not an issue that rises to the level of being detectible in a large number of animals or humans with enough clarity to say that there is a relationship.
We know that, essentially, all drugs have the potential to cause some adverse effects in some individuals… The question is how the benefit-risk affected by adverse events, if they occur, or the potential if adverse events, if you can’t prove them. What I was trying to relay to the audience there was (Victoza) in my mind, there is no question the benefits outweigh the risks, particularly in regard to the pancreas, where we don’t have any evidence. Certainly, we have no evidence for any risk for pancreatic cancer. And we can’t be 100 percent about the risk of pancreatitis – yea or nay – but the risk, if it occurred, would still be acceptable given the benefit of this drug…
Pharmalot: What would you like to see happen from here then?
Moses: I think the ADA is in the position of being a patent champion… and to do what they can to reassure physicians and patients about the safety of these drugs, or identify a problem if one really exists. And I think that’s an appropriate stand for them to take. It’s unclear whether this approach will actually accomplish that , beyond what the FDA has already done and, which is actually being accomplished in Europe with a safeguard, which has been in place since 2011 and is looking at the safety of anti-hyperglycemic agents for Type 2 diabetes. So there’s a lot going on and we’re just waiting for answers.
Pharmalot: To me, what was interesting about the ADA statement was that, compared with March when one of the studies came out (concerning pancreatitis and pancreatic cancer), they pooh-poohed it. Within the course of three months, the ADA changed course.
Moses: I don’t know that they changed course – I can’t speak for them - but what I think they recognized was this discussion was becoming destructive and primarily for physicians and patients. And they said we can play a role in this and help to solve it. Whether the plan is sufficiently timely or will provide the answers is not known now and that’s why they’re exploring what can be done. I think they’re taking a very responsible position here, but it’s still unclear whether it will resolve the situation any more than what’s already being done by a number of different companies and certainly by the FDA.
Pharmalot: But the FDA didn’t reach a conclusive finding.
Moses: So you could say ‘shouldn’t the FDA pool all this data that they already have and make this judgment?' That’s up to the FDA. What was clearly apparent from the presentation is that they’re not overly concerned about it because of the nature of the data they have already.
Pharmalot: So is the ADA stance just window dressing?
Moses: That’s your phrase, not mine…. Whether in the end this will resolve it or not, and what ‘it’ is hasn’t been defined. Part of the problem is that without knowing what’s going to be done, it’s hard to say it’s the right or the wrong solution to the issue. I’m not being evasive. There is no answer in my mind.
Pharmalot: Do you have any preconditions to participation?
Moses: No, but we’re not even that far. I would say the way ADA request has been interpreted in the press has been a little bit strong. It’s been played that companies should come forward with patient-level data. And they are suggesting it’s the right thing to do, but it’s come across in the language a little stronger than what they’re asking for now. I think they’re asking what’s possible – what are companies willing to share? And that’s not defined yet.
Pharmalot: What is Novo doing on its own toward answering this?
Moses: Since the beginning of the development program, when it was recognized that GLP-1 directly affected beta cells, we’ve been looking carefully at the pancreas, because there was concern this compound might cause cells to develop benign or even malignant tumors over time. And we’ve seen no evidence of that. In the pre-clinical program, we did the usual toxicology. We looked at animals for up to two years in carcinogenicity studies – in mice and rats – and non-human primates at doses over 60-fold the exposure a human would have. Did we see anything? There was no differentiation between animals that didn’t receive (Victoza) and animals that did.
After some of the warnings about pancreatitis were included in the label, we committed to a number of studies in the post-marketing requirements – FDA and EMA – that are in the process of being executed. Two of them are five-year studies and we haven’t gone five years yet. And the other was a non-clinical study in rats that were made diabetic so we could look at the effects in these animals, specifically on the pancreas. That study has been completed and published. We took very careful sections of the pancreas. It was divided into four sections and… measured the various cell sizes and compartments. We saw no changes that differentiated control animals from animals treated with active drug. There was just no signal…
At the same time, we have a preliminary results of a pharmacovigilance study, where we prospectively defined where we’re going to look at and we matched each new initiator on the (drug) with a patient treated with another compound for diabetes. They were matched for age, duration of diabetes, gender, background medications. And you had to have a good match to be included… A five-year study and we’ll finish in late 2015. We did a cut through May of this year, and if you look at the hazard ratio, or relative-risk ratio, in no case was that ratio greater than one…. So we’re very reassured. Now, you could say there were too few cases. But there were not a huge number of cases of pancreatitis and certainly less of pancreatic cancer. We studied 25,000 initiators. .. The confidence intervals included 1. There was no statistical differences – you wouldn’t expect them – and no signal either. That was very reassuring and that study will continue. I would suspect by the end we’ll have 50,000-plus patients initiated.
The other study is our CV outcome trial… part of a post-marketing commitment to the FDA… One of the principle secondary outcomes from that are pancreatitis and pancreatic cancer. Unlike the clinical development program, cases of pancreatitis are adjudicated – have to submit all the data to confirm the diagnosis. That was not in the clinical development program because there was no indication at the time that could be a relationship. This will report in ’16… So we’re doing that. And also looking at pancreatic enzyme levels in all ongoing trials.
The challenge is this so dangerous that we can’t wait to get this data. Our response, my response, would be we can’t replicate a signal in all the studies we’ve done and we believe – the randomized, controlled, blinded studies – are the best way of determining whether there is anything there or not… If the ADA or other independent bodies want to pool all of the data from the studies… if you can’t answer the question from that data, the likelihood of ever having a clean answer is not high.