No 'duh': Obvious answer to drug R&D is not so easy to implement

The answer to reducing the cost and time of developing drugs for unmet needs seems pretty obvious – a better understanding of the disease, the patient and the molecule being developed.

But putting that mechanistic answer into action is neither obvious nor easy, a point repeatedly made Tuesday at an FDA, Brookings Institute conference on improving productivity in drug R&D and reducing late-stage failures by advancing early learning.

One of the biggest barriers to moving forward is the tension that exists between using the empirical tradition of randomized, controlled trials – the recognized gold standard of drug development – and the mechanistic, scientific approach that seeks to understand the mechanism of action, the FDA's Janet Woodcock said in opening the conference. That tension must be brought to the surface and discussed.

Although it's the "triumph of clinical medicine," a randomized, controlled trial can only test a single hypothesis, Woodcock noted, adding, "That's not modeling. That's not mechanistic data." Instead, it results in the loss of a vast amount of information from each patient. That loss could doom a trial.

The primary reasons for costly clinical trial failures "are centered on efficacy and safety issues stemming from insufficient knowledge of key matters like the biological relevance of the molecular target or the dose-response relationship" between the study drug and that target, according to a Brookings backgrounder.

Woodcock acknowledged that the loss of data from preclinical and early stage development contributes to the 50 percent failure rate of phase III trials. "We can do better," she said. Part of the solution is recognizing that preclinical work and early trials must be more than a check-the-box process.

The challenge to get more out of early development and, thus, avoid late-stage failure has become greater with the creation of speedier tracks to approval. The new breakthrough designation, for instance, has resulted in a high stakes race so complex that no one has figured out how to win it, said Michael Maitland, an assistant professor of medicine at the University of Chicago Medicine.

Some sponsors run at breakneck speed from preclinical to phase I with an eye on turning phase II into the pivotal trial so they can get to the finish line and enter the market ahead of their competitors. Others are taking the time to figure out more about the drug they're developing, Maitland said.

INTEGRATING PRECLINICAL DATA

To get to the finish line, Woodcock said sponsors and regulators need to better integrate preclinical models into clinical development. "This is fantastically complex when you get into the details," she said, "but at least we're not blindly intervening" when the compound is moved into humans.

At the starting gate, the sponsor needs to identify the right target by looking closely at the legacy of successes and failures, said Paul Morgan, head of translational safety at Astrazeneca plc. Having the right target is the first step in a framework that also includes having the right tissue, the right safety, the right patients and the right commercial potential. Other members of the panel added a sixth "R" – the right industry culture that embraces collaboration internally and externally.

Strong, reproducible data are crucial to identifying the right target. Yet 50 percent of published studies can't be repeated with similar conclusions, said James Barrett, chairman of the pharmacology and physiology department at Drexel University. The prevalence of irreproducible preclinical results costs $28 billion in the U.S. alone, he added. (See BioWorld Today, June 12, 2015.)

To strengthen the quality and contributions of preclinical studies, the FDA's Karen Davis-Bruno described a paradigm shift that's occurring with the ultimate goal of reducing risk in the clinic. Instead of taking an observational/reactive approach to nonclinical studies, sponsors need to follow an integrative predictive/proactive approach, she said.

GETTING THE DOSE RIGHT

Another area that needs to be better exploited is dose selection and regimen. Woodcock encouraged sponsors to explore the widest range of doses possible, extracting the maximum amount of data from every patient, especially when targeting a rare disease.

Because of the cost implications and the rush to get to phase III, dose selection and the data generated in dosing trials haven't always received as much attention as they should, given that the wrong dose could sink a drug. And when sponsors try to turn phase II into their pivotal trial to speed approval, they have to use the same data to confirm their results and select their dosing. "It's difficult to do both," FDA statistician Lisa LaVange said.

In several instances, sponsors have been required to conduct postmarket dosing trials. Amita Joshi, senior director of clinical pharmacology at Genentech Inc., noted that a study of 41 oncology drugs showed that more than a fourth of those drugs had to go through dosing trials following approval.

Looking ahead, the FDA expects to see more efficient trial designs that integrate all the data, including the rationale for the dose selection, said Vikram Sinha, director of the agency's Division of Pharmacometrics. Identifying the right dose should be the goal of early development, he added, recognizing that each therapeutic area will have unique aspects that drive dose studies.

Dose selection, like other parts of drug development, is all about discovery and the patient, said Sandra Allerheiligen, a vice president at Merck Research Laboratories. "There's no checklist. Every drug is different," she added.

Allerheiligen stressed that drug development should be cyclical, not linear, with as much attention given to translating clinical results into the real world as is given to translating the basic science to clinical development.

Maitland agreed, saying that the challenge for the sponsor is getting the dose right for each patient. Drug development doesn't stop at approval; it must continue on in collaboration with physicians and patients, he said.

The conference was the first step in an ongoing discussion aimed at identifying unrealized opportunities to enhance existing and emerging approaches, along with the barriers, to the more systematic application of pharmacology clinical tools in drug development, according to Brookings.

Source