Onglyza and the Type 2 Diabetes Approval Path: The Map is Still Sketchy

Biopharma companies hoping to develop drugs to compete in the type 2 diabetes market can breathe a bit easier with the approval of Bristol-Myers Squibb/AstraZeneca’s Onglyza (saxagliptin): the final go ahead from the Food & Drug Administration July 31 unequivocally shows that it is indeed possible to meet FDA’s new cardiovascular safety guidelines for the class.

The outcome is not surprising, after an advisory committee overwhelmingly supported approval of the drug earlier this year—but given the delays and uncertainty, sponsors may be forgiven if they weren’t going to believe it until they saw it.

And now that FDA has posted the approval letter for the drug, sponsors can get to work on figuring out exactly what it will take in terms of post-marketing requirements in the type 2 market of the future.

The letter makes one thing clear: a mandatory Risk Evaluation & Mitigation Strategy will not be an inevitable price for entering the class. Onglyza was approved with a host of post-marketing study commitments, but no REMS.

As for those post-marketing studies themselves, the key commitment is exactly what sponsors would have expected: a prospective clinical trial to assess cardiovascular outcomes. That section is bound to be read carefully by sponsors eager for more clues about how onerous post-marketing trials will be in the class. But they probably won’t learn much that they don’t already know.

Recall that during the advisory committee, Bristol declared its intention to run a study to demonstrate a superior cardiovascular safety profile for saxagliptin. The approval letter (in keeping with the safety focus of the new post-marketing authorities) frames the trial differently, with the goal of showing a relative risk of 1.3 or less versus the control population.

The letter also does not include any details about numbers of patients, duration of therapy or other key trial design elements. However, FDA has made clear its preference for an event-driven trial design, which (assuming Bristol agrees) would make some of the those elements irrelevant.

The letter does detail a number of other specific risks to test for (most triggered by signals in the NDA for Onglyza, but some clearly of a classwide concern.)

And it does set some deadlines. Bristol must finish designing the trial in November, and then complete the study in less than five years.

But beyond that, there’s not much to learn.

Still, the Onglyza letter is certain to be a template for future approvals in the type 2 diabetes class, so here is how FDA describes the pivotal post-marketing trial:

"A randomized, double-blind, controlled trial evaluating the effect of saxagliptin on the incidence of major adverse cardiovascular events in patients with type 2 diabetes mellitus.

The primary objective of this trial is to establish that the upper bound of the 2-sided 95% confidence interval for the estimated risk ratio comparing the incidence of major adverse cardiovascular events observed with saxagliptin to that observed in the control group is less than 1.3.

Secondary objectives must include an assessment of the long-term effects of saxagliptin on lymphocyte counts, infections, hypersensitivity reactions, liver, bone fracture, pancreatitis, skin reactions, and renal safety. For hypersensitivity reactions, especially angioedema, reports should include detailed information on concomitant use of an angiotensin-converting enzyme inhibitor or an angiotensin-receptor blocker. For cases of pancreatitis, serum amylase and/or lipase concentrations with accompanying normal ranges and any imaging study reports should be included in the narratives.

Because renal impairment is an important complication of diabetes, you must ensure that there is a minimum of 1 year of exposure for at least 200 saxagliptin-treated patients with moderate renal impairment and at least 100 saxagliptin-treated patients with severe renal impairment.

The timetable you submitted on July 15, 2009, states that you will conduct this trial according to the following timetable:

Final Protocol Submission: by November 30, 2009


Study Completion: by July 31, 2015


Final Report Submission: by January 31, 2016.”

While You Were Watching Your Sugar

Or maybe instead, while you were skipping sessions at ADA to get sugar-coated beignets at Cafe Du Monde. Either or. Mmmmmmmmmmm.

In any case we've got your weekend wrap-up below, a few morsels of general industry news sprinkled with the sweet taste of clinical diabetes announcements. Or you know, the other way around.

While you were should have been sweeping the Dodgers ...

• Aileron, a biotech developing a new class of molecules it calls 'stapled peptides', raised a $40 million Series D. What's particularly noteworthy is the participation by SR One (GSK's corp. VC arm), which led the round, as well as Lilly Ventures, Novartis Venture Fund and Roche Venture Fund. For the kind of analysis that predicts deals like this, see this May START-UP feature by Ellen Licking.
  
• ADA: From late on Friday, a new GSK-sponsored study, RECORD, says Avandia doesn't up heart-attack risk. Not everyone is sold. Hmmm, we wonder who might have a comment. According to a WSJ report: “The Record trial is seriously flawed,” Nissen said, adding that there weren’t enough patients who remained in the study to say conclusively that the drug doesn’t pose a high risk for heart attacks and other cardiovascular problems.
  
• ADA: GLP-1 candidate from GSK, Syncria, better than Byetta in Phase II trial. Good enough for a major Phase III program, anyway. Reuters has the wrap up here.
  
• ADA: Meanwhile, Novo's liraglutide was beating up on Sanofi's Amaryl. Details on that two-year study here.
• ADA: Hi, we're from Boehringer Ingelheim. Did we mention we're in diabetes now?
• The NYT Magazine on the challenge of passing health care reform.
• EHA: ADA wasn't the only game in town this weekend, depending of course on what town we're talking about. At the European Hematology Assn meeting in Berlin Calistoga Pharma touted new results on its PI3k inhibitor.
  

image by flickr user jasonrowland.org used under a creative commons license

While You Were Watching Football

You didn't miss anything this weekend. Really, nothing happened, we looked. Oh, but the Eagles destroyed the St. Louis Rams 38-3 and the Phils took the weekend set against the Mets at Shea two games to one. But to dig up anything interesting in the biopharma universe, we had to wait til this morning. So here you go ... while you were driving to work:

• BMS and AstraZeneca released positive data from a study of their DPP-4 inhibitor saxagliptin (Onglyza) at the European Association for the Study of Diabetes meeting in Rome this morning. The companies said that the drug, which was submitted to FDA over the summer, "produced significant reductions across all key measures of glucose control studied."
• Also at EADS, Novo Nordisk said that its liraglutide human GLP-1 analog plus metformin and rosiglitazone led to improved blood glucose lowering, weight loss, blood pressure reduction and improvement in beta cell functioning. For more news out of this diabetes meeting, click here.
  
• Velcade was approved in the EU for combination with melphalan and prednisone for the treatment of patients with previously untreated multiple myeloma (MM) who are not eligible for high-dose chemotherapy with bone marrow transplant.
• Finally, a mini Press Release of (Last) Week: for the PR rep who is sending us pitches to cover the release of "Don't Mess With the Zohan" on DVD, this is for you.

photo of Lincoln Financial field, scene of yesterday's carnage, by flickr user MattP33, used under a creative commons license.

Whoa Nelly: FDA Puts the Brakes on Liraglutide

When FDA knows six months ahead of a user fee deadline that it won’t be able to make an on-time approval decision, it’s not exactly a positive sign of a drug’s future prospects.

But that is exactly what Novo-Nordisk is facing with its once-daily human GLP-1 analogue liraglutide—the company’s much anticipated type 2 diabetes therapy.

FDA isn’t scheduled to make a decision on the approvability of liraglutide until March 23 at the earliest. But despite being six months away from the user fee deadline, the agency has already warned the company not to expect an on-time approval.

The delay, Novo says, will allow time for FDA to absorb feedback from an advisory committee review of liraglutide, which has been scheduled for March 2. The agency told Novo that it doesn’t believe the three weeks between the meeting and the user fee deadline will be enough time to complete the review.

“Since the advisory committee meeting is held shortly before...[the user fee] date, the agency has indicated that it will most likely have to extend the date of completing their assessment by a couple of months,” Novo said in a press release.

While FDA has certainly been able to deliver approvability decisions within three weeks of an advisory committee meeting in the past, there are two reasons why it is different this time around for liraglutide.

First, the agency is simply missing more user fee deadlines—something that we’ve been tracking and writing about in The RPM Report. While FDA is allowed to miss approval deadlines under the Prescription Drug Fee User Act, drug reviewers were given a green light to start missing more than usual earlier this year.

Based on some pretty small numbers, we've found that the longer a drug sits at FDA, the less of a chance of a positive decision. So if the agency is indeed just a "couple of months" late on liraglutide, that seems to bode well for the drug's approvability prospects. That is, of course, barring any other unforeseen (or seen) circumstances.

Which brings us to our second point. FDA is obviously taking it a bit slow on liraglutide given the pancreatitis safety signal seen with another GLP-1: Amylin’s exenatide (Byetta). Novo’s challenge at what will surely be a closely watched advisory committee meeting will be to make a convincing case that its product does not carry the same signal as Byetta.

As the Diabetic Investor points out, we’ve already seen this play out in the diabetes market over the cardiovascular safety issues with GlaxoSmithKline’s rosiglitazone (Avandia). As Avandia was imploding, Lilly/Takeda were asked to defend the safety profile of their thiazolidinedione drug product, pioglitazone (Actos). And in that case, they succeeded.

Novo Scraps Inhaled Insulin

The dismal failure of Pfizer/Nektar’s Exubera loudly called into question whether inhaled mealtime insulin was commercially viable at all.

The answer—surprise, surprise--is that it’s not, at least according to Novo Nordisk. The Danish firm announced last night that it was scrapping its Phase III inhaled insulin program, which uses Aradigm’s AERx liquid aerosol system.

The slight irony here is that the (already painfully delayed) AERx program was killed because it failed to show “sufficient clinical or convenience benefits” over the various insulin analogs already available to patients—including, prominently, those using Novo’s own FlexPen, a discreet and simple-to-use injection device.

Novo’s decision doesn’t mean that other late-stage inhaled insulin wannabes, including Lilly/Alkermes and MannKind will follow suit. But as we argued in this IN VIVO feature, Pfizer’s snafu means the going will be tough. Novo was at best going to be third to market, and the brick-sized device (far larger than Lilly/Alkermes’) had long been recognized as a problem—that’s why Novo had begun a next-generation program in-house. And the product required refrigeration.

So the writing was on the wall. In fact it’s somewhat of a relief that Novo has finally put this long and expensive project to bed, taking a non-recurring cost of about $260 million (DKK 1.3 billion), which will hit 2007 operating profit. Mads Krogsgaard Thomsen, CSO and EVP of Novo Nordisk had already last year acknowledged that “this is not going to be a huge product.” Now it won’t be one at all.

Not that this spells the end of pulmonary delivery for Novo. The problem with all of the current batch of inhaled insulins, according to Thomsen, is that they’re short-acting, meal-time insulins that must be taken alongside basal insulin—the ones available with tiny, pain-free injection devices. Exubera's failure showed that patients (and payors) understandably, were reluctant to add onto that regime something even more complex. And Pfizer, for reasons we outline here, failed to reverse the treatment sequence by persuading physicians to prescribe insulin earlier on.

So given that meal-time insulin is typically a fifth or sixth step in diabetics’ chain of medication (which progresses from diet-and-exercise, through oral anti-diabetic drugs to GLP-1s and then basal insulin) why did drug companies focus their inhaled efforts on this and not basal insulin? “Becase we had no choice,” says Thomsen, technological limitations meant that prandial insulin was the only one which could be formulated for inhalation.

Those limitations are no longer, he continues. Novo now intends to focus on developing pulmonary forms of basal insulin and GLP-1. We outlined in a feature last summer the importance of glucagon-like-peptide (GLP-1) analogs (and Phase III GLP-1 analog liraglutide in particular) to Novo’s business, so it’s no surprise that GLP-1s feature in the firm’s fresh set of pulmonary delivery plans.

These are a way from the market, however—liraglutide itself can’t be formulated for inhaled delivery because its half-life is too short; nor can Lilly’s first-to-market Byetta. Still, “we’re not starting from scratch, either; we have an inhaled, bioavailable GLP-1 candidate in late-preclinical trials,” asserted Thomsen on a conference call following today’s news.

Novo’s shares were down nearly 4% this morning; chances are Aradigm might have a bad day when the US exchange opens. But Novo’s put on a brave face. “We’re going from being followers in a commercially unattractive area, inhaled meal-time insulin, to leaders in a highly commercially-attractive area—a new generation of inhaled long-acting basal insulins and GLP-1 analogs.”

Science Matters: A small personalized medicine bailout for Cox-2s?

There was little attention paid to last week's paper suggesting that PPAR delta agonists might be used to prevent the cardiovascular side effects of Cox-2 inhibitors (coxibs) such as Vioxx and Celebrex.

The study in the Journal of Experimental Medicine (JEM) showed that Cox-2 suppresses the expression of tissue factor (TF) -- the primary activator of blood clotting and a proximal cause of coxibs' CV problems -- via the activation of PPAR delta.

Of course, there are no approved PPAR delta drugs, although pharmas including GSK have tried developing them to treat cardiovascular disease. (One news outlet suggested GSK's drug could be an "unlikely savior" for Vioxx, but that's a far stretch.) And no one would think to couple a PPAR delta with a coxib for use as a combination analgesic--the risk/benefit ratio of that presumably is way off.

But there's another, intriguing aspect to this research result.

The problem with Vioxx is that it is associated with cardiovascular complications in a small number of patients. "We should look at these patients in terms of their TF levels and other clotting parameters," suggests Timothy Hla of the University of Connecticut Health Center and a principal author of the JEM paper. "Is the TF gene in these people somehow different? Is it regulated differently? Are they more sensitive or more resistant to the effects of the PPAR delta they produce? Instead of looking at the selectivity of Cox-2, let's look at patients' sensitivity."

Hla has a longstanding interest in Cox-2's role in normal blood vessel physiology and angiogenesis: he cloned the gene from human vascular cells in and named it Cox-2 in 1992, while at the American Red Cross Research Institute.

A first step would be to measure TF levels, which can be easily collected from plasma, in patients taking Celebrex and correlate them with treatment results. It's all well and good to talk about testing PPAR delta agonists for their therapeutic effects regulating the TF gene. (The most advanced may be GSK's GW 501516, which the Hla group used in its experiments. GSK in-licensed the compound from Ligand Pharmaceuticals, but its development has lagged at Phase II. Ligand's most recent 1o-K says the drug's development is 'on hold' pending the review of preclinical studies, and there is no mention of it on GSK's own clinical trials web site or in any recent publicity [clintrials.gov lists a 'completed' Phase II study], so for all we know it has been terminated.)

So that's a long way off. Most of the focus on the mechanism of Cox-2 has centered on its effect on platelets. A simple blood test might go a long way towards refining that effort.

More Insulin Problems

Pfizer isn't the only company having problems with an alternative delivery form of insulin.

Emisphere Technologies has been working on oral insulin for more than a decade--and finally its board got fed up.

In October 2006, the company reported disappointing Phase II results with its oral insulin--no difference from placebo. The stock tumbled by more than 50%. By January the long-time CEO, Michael Goldberg, MD, was out. The firing was led by Mark Rachesky, who joined Emisphere's board in 2005 when his fund, MHR Institutional Partners, loaned the company $15 million, later changing the straight debt to a convert--at $3.78 a share. That's unprofitably close to the company's current stock price (and less than half the Emisphere price when Rachesky did the covert deal).

Goldberg was blamed for poor execution of the key trial (as well as the fact that the company hasn't made much progress in its nearly nineteen years of life). The original trial design called for the oral insulin to be tested in very sick but stable patients; to accelerate the enrollment, the company apparently relaxed the criteria. In a subsegment look at patients who met the original criteria--an often unreliable analysis--oral insulin apparently did perform well. But now the company needs a new trial to prove the point. And this isn't the first trial that wasn't well executed: its oral heparin test was hurt by a poor liquid formulation whose taste turned off patients.

The company is by no means dead--it's managed to hire a new CEO, Michael Novinski, the former president of Organon USA. And it's brought on a board of big-name diabetologists to advise it on trial design and execution. It's got a new formulation with apparently three times the absorption of the older version--thus reducing cost-of-goods and increasing patient convenience.

But oral insulin won't win on convenience. And it won't win by eliminating the pain of injections. New smaller needles make taking insulin relatively painless. The big advantage will be eliminating embarassment, says one insider: "What diabetics hate is at a restaurant having to pull up their shirt and stick themselves in the stomach with a needle." Exubera doesn't solve that problem: puffing on a big device is no less inconspicuous than sticking one's belly with a needle.

A pill should be a lot more acceptable. And yet that's not enough. The market for insulin reformulations is being shaved by better needle technology and by new products, like the injectable Byetta and the oral Januvia. And as Pfizer is learning from Exubera, managed care doesn't want to pay for convenience. Emisphere's product, to be truly important, will have to show better outcomes. It may be able to get on the market by showing equivalency to injectable insulin--but to make Emisphere, and oral insulin, a success, it will need to keep diabetics healthier, too. That's a much more expensive task than most proponents of oral protein delivery ever figured they'd have to accomplish.

Take a Deep Breath, Pfizer, and Think Again

It's just what Pfizer didn't need: another expensive failure.

Less than six months after the much-vaunted Lipitor-replacement torcetrapib tanked, it's now official that the Exubera launch is a flop. The numbers say it, the newspapers say it, and even Pfizer’s own executives acknowledge that “we still need to figure out how to market Exubera.”

Not surprising, perhaps—Pfizer isn’t exactly a large molecule expert, and Exubera, as the first ever inhaled insulin, presents even more challenges for its sponsor than your average biologic.

First, doctors must be convinced of the drug’s advantages over existing treatments; not just longer-acting insulins such as Levemir, marketed by diabetes-leader Novo Nordisk, but also since late 2006, Merck & Co.’s star DPP-4 inhibitor sitagliptin (Januvia), a rare industry success story (to be followed by recently-approved Janumet, a combination of Januvia and metformin).

Then, time-constrained docs must learn how the inhaler device works (and adapt dosages, which are different when insulin is inhaled rather than injected). If they get past that hurdle--and aren’t swayed by detractors such as Dr. John Buse, president-elect of the American Diabetes Association, who says that Exubera may present a safety risk in Type I diabetics--then their patients must get used to carrying an umbrella-sized device around with them (and paying more for the privilege: Exubera costs $2-3 more per day than injectable insulin).

So Exubera is being chosen for just one in 500 insulin prescriptions in the US, despite Pfizer’s investment—900 part-time diabetes educators, over six months of doc-targeted marketing, and considerable sales rep resource, in part diverted from other more profitable drugs. Analysts have slashed Exubera sales forecasts, in some cases to barely more than $300 million by 2012.

But Pfizer’s not giving up. “Don’t write Exubera off,” warns a senior executive at Pfizer. “We’ve screwed up before on launch, and the drug has come back.”

The company will soon launch a fresh marketing campaign, a summer DTC advertising splurge (very unfashionable), and reckons that transferring Exubera promotion to its highly successful CV team might make the difference.

Chances are it won’t. Marketing inhaled insulin—effectively a niche drug, despite initial forecasts of peak-sales of up to $2 billion—isn’t like marketing a primary care pill. You can’t use the same tricks.

“Patients will be inhaling insulin over the next few years,” insist Pfizer executives. Sure, but how many, and whose product? By the time Pfizer figures out its promotional game, competition will be even worse, even if there is a market. The handful of inhaled insulin alternatives in Exubera's wake have smaller, more discreet devices, and will benefit from Pfizer's experience in figuring out how to make breathing in one's insulin more attractive than injecting it.

“You may wonder whether Exubera could be right for you,” says Pfizer’s product website to prospective customers. Despite its brave face, Pfizer must be wondering whether this drug is right for it, too.

Reporting on Exubera: an A-Buse

Many analysts have questioned the potential of Pfizer’s inhaled insulin, Exubera. Nonetheless, it was more than surprising to see the comments about the drug attributed to American Diabetes Association president-elect John Buse, MD, this week.

In a widely circulated AP story on March 21 about Exubera, Buse was quoted saying "I think Pfizer will wish they had never gotten into this. I doubt they'll regain their investment…There is no advantage to Exubera and there may be a safety risk. I see it as my job to talk people out of (using) it."

Buse has spoken to IN VIVO in the past, and we know him to be a highly credible and extremely careful commenter who focuses on medicine and avoids discussion of company issues. His statements on Exubera and Pfizer therefore seemed unusually pointed, all the more so given that he had said “There’s cause for tremendous optimism” following the product’s approval in 2006 for the treatment of type 1 and type 2 diabetes in adults.

When asked in an email last night about the accuracy of the AP story, Buse replied: “Not fair. The comments were specifically about treating type 1 diabetes. Not about treating type 2 diabetes.”

That makes sense. Indeed, last year he pointed out the possibility of a safety risk for type 1 diabetics, who must use Exubera in combination with longer-acting insulin. The inhalation device “is a little too cumbersome to be used in multiple daily therapies like we use in type 1 diabetes, but certainly that would be an option,” he said then, adding that the drug is “mostly going to be used in the type 2 population that's just getting started in the insulin business,” as an alternative to or in combination with rapid-acting insulin injections or oral drugs. Safety issues are also amplified in type 1 patients because “they are likely to be exposed for 50-plus years,” Buse added in his recent email to us.

As for the reported comments about Pfizer’s ability to recoup its investment, he would only say that “I may have said it” in the course of reflecting on the competition Exubera faces in both the type 1 and type 2 markets from Lilly-Amylin’s Byetta, “even better [incretin mimetics] on the horizon, better pens, much greater acceptance of long-acting insulin, increasing price pressure from managed care and the interest in long-term safety demonstration.” --Mark Ratner

Did the AP inhale?

Galvus Misses Its Window; Is this All Good for Merck?

The Galvus delay is terrible news for Novartis.

Competitor Januvia, from Merck, has already seen the single best launch in recent memory for an anti-diabetes product. With a 40% share of new written scrips just four months after introduction, it may be the best launch of any new drug into a crowded category. Given its labelling -- “similar to that reported with placebo” -- doctors are rushing to prescribe what one clinician called a "no-brainer drug": zero training required in administering it; once-a-day dosing, with or without food, with or without any other medication. Once they're comfortable with Januvia, why would doctors switch to anything else, unless they're dramatically differentated?

Originally estimated to be three months behind Januvia, Galvus was already at a big timing disadvantage. Things are now much worse. Galvus seemed to have little differentiation over Januvia before (and therefore little chance of gaining a market-share advantage--now its only differentiation is negative: the skin lesions in primates, linked in FDA's mind to the toxicities seen with the drug at very high doses. That won't encourage doctors to try new patients on Galvus, particularly if, as a number of experts believe, Galvus' label--granted the drug's ultimately approved--comes with restrictions on use.

The potential time bomb for Merck is that Galvus' problems will redound to its detriment--just as the problems around Merck's own Vioxx KO'd Pfizer's Celebrex. FDA's metabolic division has been under severe scrutiny and it's possible they could do the cautious thing and start looking at lot more closely at Januvia. And when they do, will doctors too start thinking a lot more before prescribing what was once a no brainer?

AZ-BMS Diabetes Deal: Two Paths for Big Pharma

If there is a schism among Big Pharma it is between those companies clinging on to the notion that pharma's future role is--as it is today--as a massive marketer of mass-market drugs, and those that see specialism as a means to avoid imploding under the weight of their own infrastructures.
The diabetes deal announced yesterday by AstraZeneca and Bristol-Myers illustrates the pursuit of each strategy: AstraZeneca, eager to play in what one pharma CEO described this week as "the disease of our epoch," has paid BMS $100 million upfront for worldwide (except Japan) co-development and co-commercialization rights to two late-clinical stage diabetes projects. For BMS the move is another big step back from primary care marketing and confirmation that the company's future lies along a specialist path.

AZ will fund the majority (75%) of development costs through 2009, the companies said, after which costs will be split 50-50. Should each of the two drugs--saxagliptin, a DPP-4 inhibitor currently in Phase III and dapagliflozin, a SGLT2 inhibitor in Phase IIb--reach global markets BMS will earn $650 million in pre-commercial milestones and could land an additional $300 million per drug in sales milestones. Post launch expenses and profits will be split evenly on a global basis and BMS will manufacture both products and book sales.

Acquisition of diabetes projects to shore up its primary care portfolio has been high on AZ's agenda since the PPAR agonist tesaglitazar (Galida) crashed out of clinical trials in May 2006; ironically the decision to yank Galida was based on thought-leader and regulatory reaction to BMS's own PPAR, muraglitazar (Pargluva) and intimations that Galida was in for similar treatment. Pargluva was killed after analysis published in JAMA by Cleveland Clinic CV chair Steve Nissen, MD, questioned the safety of PPARs and FDA said further long-term clinical studies would be needed to approve the product. (See "Anything but Academic: Lessons from the PPAR Failures," The RPM Report, June 2006.)

By the time saxagliptin hits the market the best AZ and BMS can hope for is only two entrenched competitors: Merck's Januvia and Novartis' Galvus will likely await. Dapagliflozen is a sodium glucose co-transporter-2 inhibitor, which blocks the re-absorption of glucose from urine in the kidney; a more novel, yet riskier prospect.