Donald Berry, Ph.D., professor and chair of the Department of Biostatistics and Applied Mathematics at M.D. Anderson Cancer Center stated in the January 2006 issue of Nature Reviews Drug Discovery, the statistical method used nearly exclusively to design and monitor clinical trials today (the frequentist method) is so narrowly focused and rigorous in its requirements that it limits innovation and learning.
He advocates adopting the Bayesian methodology, a statistical approach that is more in line with how science works. It is used routinely in physics, geology and other sciences. And he has put the approach to the test at M.D. Anderson, where more than 100 cancer-related phase I and II clinical trials were being planned or carried out using the Bayesian approach.
The main difference between the Bayesian approach and the frequentist approach to clinical trials has to do with how each method deals with uncertainty, an inescapable component of any clinical trial. Unlike frequentist methods, Bayesian methods assign anything unknown a probability using information from previous experiments. The Bayesian methods make use of the results of previous experiments, to do continuous updating as information accrues, whereas the frequentist approaches assume we have no prior results.
Doctors want to be able to use biomarkers to determine who is responding to what medication and look at multiple potential treatment combinations. They want to be able to treat a patient optimally depending on the patient's disease characteristics. Cancer is a diverse disease and what works to treat one person's disease may not work for another.
Clinical trials test the efficacy (not the accuracy) of a drug. The efficacy of a drug is to produce a desired effect, which is tumor response (shrinkage). Single arm clinical trials provide the tumor response evidence that is the basis for approving new cancer drugs. The Bayesian methology can bring some much-needed "accuracy" to the forefront of clinical trials.
Clearly, more effective cancer therapies are desperately needed, and after 30 years of investigation aimed at intensified multi-agent chemotherapy, we should look for other avenues of study. In an era of ever-increasing numbers of partially effective cancer therapeutics, there is an obvious need for more accurate methologies. We cannot afford any more 'trial-and-error' treatments.
http://www.mdanderson.org/department...fb00508b603a14
Threaded Mode