Krams: How to Popularize the Adaptive Approach

November 20, 2012 | Adaptive clinical trials might be globally embraced if the current preoccupation with fixed trial designs was viewed as an “engineering problem” requiring a lot of talk and teamwork between adaptive design experts, said Michael Krams, MD, vice president & head of the neurology franchise at Johnson & Johnson (J&J).  

“This type of innovation will happen once we get out of the Tower of Babel and start building Empire State Buildings.” The skyscraper in Manhattan went up in a surprisingly short 30 months. 

Krams was speaking last  month at Transforming Drug Development: Delivering Innovation and Efficiency Through Adaptive Clinical Trials* in Bethesda, Maryland. He is widely considered the ultimate champion of adaptively designed trials, having implemented them at Pfizer, Wyeth, and now J&J (see, “Michael Krams, Biting the Adaptive Trials Bullet,” May 2007). But more important than the trials themselves is the potential beneficial impact of adaptive designs on societal health, he said. 

Reformulating entrenched processes to favor adaptive designs requires top-down leadership and the enabling infrastructure at individual companies as well as industry-wide collaboration, said Krams.  A cooperatively developed and published “integrated blueprint for efficient support infrastructure” is among his key proposals for moving the industry from a “fast-to-milestone to a fast-to-patient” mindset. The education piece is also crucial and might include high-level workshops, university courses on the pharmacological basis for adaptive trials, and efforts directed at change opponents within individual companies, he adds. 

To take root, the adaptive approach needs to be implemented across therapeutic sectors, if not entire companies, Krams said. At J&J, every new molecular entity declaration prompts a roundtable meeting to map out a development strategy using scenario analysis to scout out opportunities across study programs. Support from not only drug supply management but also finance is critical, as study budgets might range from $1 million to $20 million versus traditional point estimates. 

Regulatory challenges are real, too, but not insurmountable. Quite possibly, the “right customers within the [regulatory agencies] aren’t being addressed,” Krams said. “I go to the people who own the problem, like Rusty Katz [director of the FDA’s neurology products division]. I explain to him for disease modification for Alzheimer’s…there is no way without combination therapies and preventive strategies that we will solve the problem.” Krams’ candor worked. He was invited back to discuss his adaptive approach and Katz is becoming one of its champions. 

“It’s all about change management,” said Krams. “If the objective is to get to the correct answer, then the decision-making framework needs to identify the best research question and the best way to answer that question in the most cost-efficient manner.”  Instead, large companies “incentivize for faster-to-next-milestone thinking.” The quality of phase III studies is consequently “a societal problem of incredible dimension,” he said. “Billions of dollars are wasted because we don’t talk enough.” Outcomes would likewise improve if more time was spent analyzing study results, planning subsequent ones, and key experts “endorsed” patients into phase III trials. 

Umbrella Study for Dementia  

“Can we think of building solutions for dementia as an engineering problem?” Krams challenges listeners. The problem with Alzheimer’s disease is that it starts a decade or two before it is clinically visible and yet trials involve patients already diagnosed with dementia. PET imaging could be looking for amyloid deposition in the brains of patients pre-diagnosis. Information could also be borrowed from Down’s syndrome research, where all clinical trial subjects produce sufficient amyloid protein precursor to guarantee the development of Alzheimer’s disease. 

A very large epidemiological and natural history study currently seeking funding will build a “biomarker-based GPS system to [triangulate] the position of an individual on the trajectory of developing dementia,” said Krams. “Eventually, we’ll take large longitudinal studies, like the Framingham [Heart] Study, and make a pharmacological interventional trial out of it.” 

Individual biomarkers are of no interest, Krams said. “It’s the relationship of many biomarkers to each other that gives away the fingerprint.” Development of a map for familial Alzheimer’s would be followed by a similar one for the sporadic variety. Eventually, it may be found that therapeutic intervention is of benefit before the first amyloid deposition occurs. “Implementation is not a trivial thing,” he adds. Biomarkers will have to “read out early and clearly demonstrate a relationship to something that happens later that is of relevance clinically.” 

The beauty of this proposed umbrella study is that a variety of treatments could be explored simultaneously and those that look promising would move into confirmatory trials, said Krams. It’s the sort of adaptive approach already happening in oncology and that the U.K.’s Department of Health is particularly keen on broadening to new therapeutic areas. When drug development becomes more adaptive and thus more “networky” [sic] in this way, the focus shifts from individual trials and study programs to what is scientifically sensible and of central interest to regulators: what is most beneficial to patients.  

* Transforming Drug Development: Delivering Innovation and Efficiency Through Adaptive Clinical Trials, October 22, 2012, Bethesda, Maryland. http://www.innovationsinclinicaltrials.com/