Sponsors must move beyond glycemic control if they want any new therapy for type 2 diabetes to be commercially viable, according to a Roche executive.

In an era of heightened focus on cardiovascular safety, new diabetes agents will have to prove more than just an ability to lower HbA1c, said Anders Svensson, Roche’s head of global clinical development for metabolism. New therapeutics will also have to demonstrate a beneficial effect on CV risk factors, such as lipids, blood pressure or body weight, or show a benefit in macrovascular outcomes to satisfy payers, he predicted.

Speaking at a DIA meeting in Washington, D.C., on diabetes drug development, Svensson discussed how new requirements in the U.S. and Europe for demonstrating CV safety are forcing companies such as Roche to whittle down their pipelines and shift resources into other therapeutic categories. Svensson's talk showed that the concern that FDA's increased requirements for diabetes drugs would drive companies from the field is a very real problem.

As industry reconsiders diabetes drug development in the new world of CV requirements, new endpoints are being considered.

The Attraction Is Gone

“Diabetes used to be an attractive area for industry,” Svensson said. “But for now … diabetes is not nearly as attractive as it was a couple of years ago, and we and others are transferring resources to other areas and maybe these other areas, including treatment of atherosclerotic complications, will actually benefit patients more than new ways of controlling HbA1c.”

Thus, while FDA asserts that HbA1c reduction remains an appropriate regulatory standard for approval – a view reiterated at the DIA meeting by FDA Metabolism and Endocrinology Products Division Director Mary Parks – industry is already looking beyond this endpoint.

The value proposition in diabetes has changed dramatically in the past two years due to the new regulatory requirements for demonstrating CV safety, which have increased the size, duration, cost and risk of both pre-approval and post-marketing clinical trials for new compounds.

FDA’s December 2008 guidance on diabetes drug development requires sponsors to exclude an 80% or higher excess CV risk before approval. For those products that meet this regulatory filing threshold but have not excluded an increased risk of 30% or more, a post-marketing CV outcomes trial will generally be required (Also see "FDA’s Diabetes Guidance On CV Risk Will Get First Test With Pending NDAs" - Pink Sheet, 22 December, 2008.).

The guidance’s requirements were immediately applied to all drugs in development and under review. Among the drugs pending at FDA at the time of the guidance’s release, data from the clinical programs for Bristol-Myers Squibb/AstraZeneca’s DPP-4 inhibitor Onglyza (saxagliptin) and Novo Nordisk’s GLP-1 analog Victoza (liraglutide) managed to satisfy the new requirements. The drugs were ultimately approved, albeit with requirements for long-term CV outcomes studies that are ongoing (see chart).

However, Takeda’s DPP-4 inhibitor alogliptin was ensnared by the new requirements, with FDA citing a need for additional CV safety data pre-approval (Also see "Takeda's Alogliptin Suffers Major Setback, Delaying Drug Launch Until 2012 Or Later" - Pink Sheet, 30 June, 2009.).

The European Medicines Agency also has issued new draft guidelines on CV risk assessments for diabetes drugs. While the draft contains no “goalposts” on the level of risk that must be excluded, it states that indications of increased risk of certain adverse events or unacceptable lack of precision may trigger a request for additional CV outcomes trials pre-approval (Also see "FDA Plans To Add Specifics To Statistical Part Of Antidiabetics CV Guidance" - Pink Sheet, 12 July, 2010.).

The EMA draft guideline further states that new glucose-lowering agents should show at least neutral or beneficial effects on associated CV risk factors, such as obesity, blood pressure and lipids.

Adjusting To A New Reality

When the FDA guidance went into effect, it was immediately recognized that sponsors needed to make adjustments to clinical trials, such as conducting longer and larger studies than were previously necessary for testing antidiabetics. But as more experience is gained, companies and regulators have realized that they must enroll patients at higher CV risk (Also see "CV Safety Gets Diabetes Drugs Approved, Not Sold, AstraZeneca Exec Notes" - Pink Sheet, 5 October, 2009.).

The requirements continue to raise a number of statistical and data integrity concerns that still must be grappled with ( Also see "Diabetes CV Studies Spur Questions On Statistical Methods, Data Integrity" - Pink Sheet, 18 October, 2010. ).

Discussing FDA guidance’s requirements, Svensson said that to satisfy the 1.8 cut point for the upper bound of the 95% confidence interval for the estimated risk, a sponsor must present data on 120-137 adjudicated CV events, depending upon the randomization scheme. For a drug that is believed to have a neutral CV profile, data on 600-700 adjudicated events will be needed to satisfy the post-approval upper bound of 1.3, assuming the endpoint used is a composite of major cardiovascular adverse events (MACE), consisting of nonfatal myocardial infarction, nonfatal stroke and CV death.

“Even if we go for high-risk populations with a [MACE] incidence of 2-3%, that means 25,000-30,000 patient years,” Svensson said. “We’re now in an area where we have traditionally been with outcomes studies with approved drugs like anti-hypertensives, statins and antiplatelets.”

Neutral Data Bring Market Punishment

Given the cost and time involved in a Phase III/IV program, such an investment must deliver evidence of positive benefit, Svensson asserted. “It’s not just enough to exclude harm, and neutral or equivocable data will be punished by the market. We’ve seen what happened with Zetia after the ENHANCE trial.”

Sales of Merck’s cholesterol drugs Zetia (ezetimibe) and Vytorin (ezetimibe/simvastatin) tumbled after the January 2008 release of results from the ENHANCE study, which showed Vytorin did no better than simvastatin alone in reducing atherosclerosis (Also see "Merck/Schering: The Next Wave Of Consolidation" - Pink Sheet, 16 March, 2009.).

As difficult and expensive as development is under the current paradigm, Svensson questioned what would happen if and when one or more antidiabetics actually demonstrates a CV benefit.

As difficult and expensive as development is under the current paradigm, Svensson questioned what would happen if and when one or more antidiabetics actually demonstrates a CV benefit.

“Even if we can get a regulatory approval based on HbA1c, it might no longer be a commercially valid surrogate in the market because if there are drugs that on top of that confer a CV benefit, that will be a new benchmark, and also because the question has been raised both by academia and regulatory authorities about the validity of HbA1c as the only surrogate marker for approval,” Svensson said.

As a result, Roche is already preparing to face a world in which diabetes drug development no longer relies solely on HbA1c. The company has slashed its early-stage pipeline and closed 14 projects in the preclinical phase because “they did not look like they could deliver anything beyond glycemic control and that will not, in the future, be good enough because we see that outcomes will likely be needed at or near launch.”

Roche’s recent experience with its late-stage pipeline further suggests the new development paradigm provides little breathing room for agents with non-CV-related safety problems.

In September, the company said it had halted dosing in all Phase III trials of the GLP-1 receptor agonist taspoglutide due to higher-than-expected discontinuation rates, particularly gastrointestinal tolerability and serious hypersensitivity reactions (Also see "Roche/Ipsen's Suspended Taspoglutide Could Become Another Diabetes Casualty" - Pink Sheet, 10 September, 2010.).

Looking For The “Safe Bets”

Svensson predicted industry will focus its R&D efforts on drugs that suggest a “safe bet.”

“We will need a reason to believe that they reduce CV risk prior to moving into a huge and expensive Phase III investment. This will also impact the size, length and cost of the Phase II program. We will most likely do larger and longer Phase II programs to try to evaluate this, kill more projects before moving into Phase III, and targets with just glucose reduction will not be investable and will drop out of the pipeline.”

Even if industry discovers new mechanisms for controlling glucose, payers ultimately will be disinterested unless a drug has other advantages, he predicted.

Center for Drug Evaluation and Research Deputy Director for Clinical Science Robert Temple said some concerns raised by Svensson have less to do with FDA’s guidance and more to do with the market and clinical community’s view of the kind of evidence that is needed.

The guidance’s 1.8 benchmark “struck me at the time it was promulgated as very moderate and very wise,” Temple said. “The 1.8 gets you into the marketplace. There’s still plenty to worry about. It’s only after you’re already selling it that you have to come up with something that’s rather more difficult.”

However, Svensson said the 1.3 post-approval mark, not the 1.8 pre-approval requirement, is the real hurdle. “You cannot bring something there and hope to recoup your investment if you’re taken off the market after four or five years when you fail demonstrating the 1.3 or actual benefit.”

Advantages To Being Later-In-Class?

Perhaps with an eye toward commercial viability of drugs now in the pipeline, questions were raised as to whether fourth or fifth agents in a particular class would be forced to undergo the same level of CV scrutiny as their predecessors if outcomes studies for the latter showed no signal of harm.

Hylton Joffe, Diabetes Team I lead medical officer in FDA’s Metabolism and Endocrinology Products Division, said the agency and industry are currently in a “learning stage.”

“We haven’t really looked carefully for CV safety with diabetes drugs to date,” FDA Medical Officer Joffe said. “If we don’t look, we’ll never know.”

“We haven’t really looked carefully for CV safety with diabetes drugs to date; event rates have been very low, basically yielding inconclusive data in traditional diabetes development programs,” Joffe said. “If we don’t look, we’ll never know. Maybe we’ll do this for 10 years and we’ll come back and every single drug will be perfectly fine and then we’ll have to scratch our head and say does this make sense to keep continuing along this road? Or maybe we’ll have a couple of surprises along the way and we’ll say it’s a good thing we looked.”

In a panel discussion that followed a debate over the CV safety of GlaxoSmithKline PLC’s Avandia, Cedars-Sinai Heart Institute cardiologist Sanjay Kaul, a member of FDA’s Cardiovascular and Renal Drugs Advisory Committee, questioned what would happen if a post-approval outcomes study failed to reach the 1.3 benchmark.

“This guidance document … enforces a hazard ratio of 1.3. I would be curious to see if [for] programs that don’t meet that criteria, whether FDA has the legislative authority to pull a drug from the market,” Kaul asked. None of the FDA officials on the panel addressed Kaul’s question.

By Sue Sutter

Glycemic Control Not Enough For New Diabetes Drugs, Roche Exec Says

More from Archive

More from Pink Sheet