Making faster, better, more cost-effective clinical trials

February 1, 2008

Over the past few years, the number of new cancer drugs approved by FDA has been dwindling. Critics generally cite exorbitant costs, lack of accrual to trials, conflicts of interest, and government regulations as the primary reasons for the dearth of new molecular compounds.

Over the past few years, the number of new cancer drugs approved by FDA has been dwindling. Critics generally cite exorbitant costs, lack of accrual to trials, conflicts of interest, and government regulations as the primary reasons for the dearth of new molecular compounds.

According to a new study group from the University of North Carolina, Chapel Hill, School of Public Health, our clinical trial system itself diminishes our ability to make safe and effective drugs available to the public in a timely manner.

Led by Joseph G. Ibrahim, PhD, the UNC Center for Innovative Clinical Trials has an interdisciplinary focus and theme involving several departments at UNC as well as collaborators from other institutions. The Center received initial start-up funding through a gift from Dennis and Joan Gillings to the School of Public Health.

A primary focus of the Center is to advance statistical science in clinical trials and move that knowledge into clinical and statistical practice. "Our goal is to become globally recognized as the premier leader in research regarding clinical trials," Dr. Ibrahim told ONI.

The structure of the initiative is a multipronged approach of methodological research (see Table on page 27), applied interdisciplinary research, outreach to industry, education, and critical evaluation of clinical trial methods, he said.

"The Center is unique in that we are developing new statistical methods for clinical trials, " said Dr. Ibrahim, who is also Alumni Distinguished Professor of Biostatistics and director of the biostatistics core at the Lineberger Comprehensive Cancer Center.

Dr. Ibrahim said that phase III clinical trials require a multidimensional approach in which several endpoints are examined simultaneously. "We think this will make adverse events, toxicity, treatment interventions, and quality-of-life results more apparent earlier on in the conduct of the trials," he said.

In addition, he said, in order to cut the dismal rate of failed trials, we need to formulate clearer "stopping rules"—points at which a trial is halted because of safety concerns or efficacy concerns: Either the drug is so ineffective that it's unethical to continue the treatment, or it is so effective that it's unfair to withhold the treatment from other patients in the trial.

To effect substantial changes in measurement or analysis of trial data will not be easy. It will require building consensus among investigators, industry, regulatory agencies, and medical providers, both in the United States and globally, he said.

Dr. Ibrahim said that the main reason for creating the Center was to find ways to make the conduct of clinical trials better, faster, and safer, and to speed up the drug approval process. "That's the overarching goal, because right now trials are taking way too long to complete, whether they're successes or failures, and then once they're completed, the approval process at the FDA is too long," he said.

Frequentist or Bayesian?

"One of the biggest differences in perspectives of the clinical trial process involves statistical data analysis," Dr. Ibrahim said. There are two main paradigms for analyzing data: the frequentist and the Bayesian.

In the frequentist paradigm, you, in a certain sense, "start over" each time in the design and analysis of a new trial without using preconceived notions or incorporating previous data. The frequentist paradigm is the one that is most often used now, especially for designing studies and analyzing clinical trial data in FDA's review process, Dr. Ibrahim said.

On the other hand, Bayesian methods incorporate information you know from past experience into the trial design and analysis. Bayesian methods allow for direct probability (chance) statements, such as the chance that one treatment is more effective than another, and they are a more natural paradigm for making predictions about future outcomes with certain treatments or therapies, he said.

"The Bayesian approach is something FDA and NIH are currently grappling with," Dr. Ibrahim said.

Overcoming lack of software

A big issue for FDA concerning Bayesian methods is the lack of software to facilitate incorporation of preconceived notions, in statistical terms, into the design and analysis of trials. "Since frequentist methods have been accepted, there are well-established software packages to implement these methods. This isn't the case with Bayesian methods," he said. However, since many believe there is strong potential for using Bayesian methods to make trials more efficient, the software will ultimately follow.

In the interim, the Center will focus on its priorities of developing various partnerships and obtaining NIH funding and collaborative research projects with FDA. "We possess the potential to have a major global impact on research directions and practice in clinical trials," he said.