Firms shouldn't lose sight of the need to build the commercial aspects of their type 2 diabetes drugs as they focus on reorienting clinical trials to include cardiovascular risk assessment, according to an AstraZeneca exec.

While new CV safety hurdles are making trials bigger, longer, more expensive, and - with the inclusion of sicker patients - more difficult to conduct and interpret, companies should still focus on the end goal of creating a marketable product, AstraZeneca Executive Director of Clinical Development Shamik Parikh stressed at a Sept. 23 conference hosted by the Drug Information Association on cardiovascular safety and type 2 diabetes drug development.

"At the end of the day, proving cardiovascular safety does not sell drugs." He added, "You have to go after your indications; you have to go after your place in the market. Just doing a big trial and coming up with [a point estimate of] 1.3 is not going to sell your drug."

"Saying 'my drug does not kill more patients than placebo' is not going to do the trick. You have to have your other indications."

Though having an acceptable safety profile is a valuable asset on the market (and having that robustly defined should help with payers and help give doctors the reassurance to prescribe the drug), "safety first" is not an appropriate strategy for drug development, even if it defines the regulatory review for FDA.

Since the release of FDA's guidance on CV risk assessment for type 2 diabetes drugs, developers have been focused on how to incorporate the new requirements into clinical trials - a complex undertaking. Parikh's point is a valuable one for companies to keep in mind, because that can't happen at the expense of the traditional aims of clinical trials: building an attractive cohort of clinical data to support your molecule.

Policy Is Still Young, And Having Growing Pains

FDA and industry are less than a year into working with the new standards, which specify that new diabetes drugs must demonstrate a hazard ratio of less than 1.8 compared to the control when measuring CV events (¹ Also see "FDA Diabetes Guidance Encourages Enrollment Of Sicker Patients" - Pink Sheet, 17 December, 2008.).

AstraZeneca has more direct experience than most firms. The first test of the new policy came with the advisory committee reviews of AstraZeneca/Bristol-Myers Squibb's Onglyza (saxagliptin) and Novo Nordisk's Victoza (liraglutide). The common critique from the expert panel was that neither program managed to detect enough events to do a true assessment, which led to the suggestion that sicker patients need to be studied (² Also see "Diabetes Drug Developers Urged To Study Higher Risk Patients By FDA Panel" - Pink Sheet, 27 April, 2009.).

FDA's Office of Medical Policy Director Robert Temple made clear to the conference attendees that the studies do need to be reoriented towards sicker patients to get a proper sense of the CV risk. "The whole thrust of the diabetes guidance ... is to see whether those people who already have a pretty good shot at stroke, heart attack, or death, have more risk on the drug," he said. "So these aren't clean patients - if you don't have enough events to have a confidence interval of 1.8 - if people are too clean - you lose. You won't show what you need to show."

How To Design Trials With Dual Purposes

Enrolling sicker patients is the clearest strategy for collecting more events in the clinical trials, but it's still unresolved how that will translate into simultaneously assessing the efficacy of the drug.

One way to fulfill the need to evaluate the CV safety without muddling the efficacy assessment, Parikh suggested, is to separate out those ends into different clinical trials.

That was one of three options Parikh outlined for how companies can creatively design their development program.

Firms could do two separate trials, one indication-seeking efficacy trial and one dedicated CV study. In fact, companies can stagger those trials, doing traditional monotherapy or metformin add-on studies and then, six to 12 months later, "to make sure you get some data and have more confidence in your compound, then you embark on your dedicated trial in high-risk patients," with the intent of finding the CV events in that trial. "This way you're getting your indications for use in low-CV risk patients and getting your high-risk patients," he said.

Another option would be to conduct a single large outcomes study that represents the entirety of the Phase III program - a real-world study with early and late-stage patients. But that would take "a lot more time," Parikh commented. It's also putting all of the eggs into one basket.

The third option would be to do a series of indication-seeking trials that incorporate high-risk patients, relying on all the trials to deliver the CV events to meet the standards of the guidance. That too has a higher degree of risk, though, in case the high-risk patients affect the collection of adequate data to support the indication.

Design Features To Help Along The Way

Regardless of the Phase III approach, companies need to do more in Phase IIb. For example, companies can ease the Phase III process by taking care of dose exploration before starting the late-stage trials. "Companies will pay a big price if they're not sure of their doses in Phase III," he said. "It was easier to do dose exploration in Phase III before - it's not going to be that easy under this guidance."

The guidance also pushes companies toward doing more robust Phase II trials by stating that Phase II and III designs should permit a pre-specified meta-analysis of major CV events.

Firms also need to consider different randomization schemes to get enough CV events in the study drug arm. Parikh said most of the diabetes development programs he now sees "are driven by safety exposure requirements, and those requirements are for how many patients should be on your drug ... at the end of the day you will end up with more patients on your drug than on the comparator."

During a panel discussion on clinical trial design considerations, the subject came up of what to use as the comparator, and how to account for confounding factors if the two arms required different levels or kinds of rescue therapy - a subject of greater relevance now that trials will enroll high-risk, rather than "clean," patients.

"At the advisory committee the issue of the comparator was raised, and the advice from the members did favor placebo on background therapy or standard of care," said FDA Director Division of Endocrinology and Metabolism Products Mary Parks. "Now, as we move forward and learn more about these trials, that may change. Particularly as we get to know more about the cardiovascular risks of other therapies," Parks said.

In The End, It's Still Harder Than Before

But all of the creative approaches to late-stage development still have the greater difficulty of diabetes drug development in the post-CV risk assessment reality - and there is still the danger that the policy will have a depressing effect on the field.

Parikh gave voice to the many concerns that have been circulating around the industry about the potential for the new CV risk requirements to discourage diabetes drug development.

Under the new policy, clinical trials for diabetes drugs are more difficult to conduct and have less certain outcomes. Already venture capitalists have re-evaluated type 2 diabetes investments because of the new regulatory hurdles and an economic climate that leaves them less nimble.

Parikh pointed out how those pressures are being felt as drug development companies look at their pipelines. "How many of us would move a drug with a point estimate of 1.26 into Phase III?" he asked. "The odds are not in your favor that you are going to automatically meet the 1.3 requirement."

"People will push for lower and lower point estimates and ... this is going to take out some of the drugs in development. Maybe it's a good thing, maybe it's not, but at the end of the day, it's this screening thing that is being applied so that automatically we can do a definitive trial."

Bigger, Longer, Costlier, More Uncertain

Parikh noted that Phase III trials will now be bigger, at least two to three years longer than before, and much more expensive - all with more uncertainty toward successful registration for a given timeline.

"There is going to be a lot more uncertainty because you are now relying on the cardiovascular events that are happening in the trials that are ongoing and we are relying on the relative risk that is shaping up. There is going to be a lot more uncertainty of the ability to file on a given timeline," Parikh said.

Also uncertain is whether the CV hurdle will keep rising as new drugs compete with old drugs that may have favorable CV risk point estimates - which could lead companies to drop development of more drugs for fear of not falling within the new risk thresholds.

Speakers at the conference noted that CV considerations already account for around 70 percent to 80 percent of internal compound attrition. Still, the approval of AstraZeneca/Bristol's Onglyza - the first new molecular entity to treat diabetes approved according to the new standards - proves that the hurdle is not insurmountable (³ Also see "AstraZeneca/Bristol's Diabetes Drug Onglyza Cleared" - Pink Sheet, 31 July, 2009.).

The lure of the diabetes market is still a strong pull that should be enough to keep investment in novel diabetes drugs flowing.

AstraZeneca and Bristol released the first Phase III data from another drug out of their diabetes collaboration on Oct. 2. Dapagliflozin is potentially the first in the new class of SGLT2 inhibitors, which facilitate elimination of glucose in the kidneys - a new approach (⁴ Pharmaceutical Approvals Monthly July 2009).

- Jamie Hammon ( ⁵ j.hammon@elsevier.com )

CV Safety Gets Diabetes Drugs Approved, Not Sold, AstraZeneca Exec Notes

More from Archive

More from Pink Sheet