Statistical Approach to Clinical Trials may Accelerate Cancer Drug Development Process

A study published in JAMA Oncology by researchers from the Massachusetts Institute of Technology (MIT) and Mayo Clinic proposes a change to the framework for determining whether or not therapies are safe and effective.

"Randomized clinical trials, where patients are assigned randomly to two groups, one receiving a new treatment and the other receiving a placebo, are the gold standard for determining the safety and effectiveness of a treatment," says Andrew Lo, Ph.D., study author. "Only when the treatment group shows significant improvement over the placebo group, will regulators approve the therapy."

Dr. Lo says the current process is designed to protect the public by preventing ineffective and unsafe therapies, "false positives", from entering the marketplace.

At the core of this new framework, which was jointly developed by MIT researchers Lo, Shomesh Chaudhuri, M.S., and Vahid Montazerhodjat, Ph.D. and the late Mayo Clinic biostatistician Daniel J. Sargent, Ph.D, was how to quantify "significant improvement" in a clinical trial.

Dr. Lo says the traditional approach is to set the bar high enough so that the risk of a false positive is small, typically 2.5 percent. However, this value is considered arbitrary by Dr. Lo, and patients with fatal diseases like pancreatic cancer or glioblastoma (GBM) may be willing to take a greater risk of a false positive because their alternative is death.

The research team's method for computing the optimal risk of false positives on a case-by case basis accounts for the severity of the disease, the number of patients affected, and the value of an effective treatment to patients. For colon cancer, their method yields an optimal risk of 2.3 percent close to the commonly used value. But for GBM, the optimal risk is 47.5 percent, reflecting the fact that the median survival time for these patients is two to three years and there are currently no effective treatments for this disease.

This framework is flexible enough to include a variety of stakeholder perspectives, and the authors have provided open-source software to allow others to re-run their analysis under different sets of assumptions for risk preferences, disease burden and prevalence, and value delivered to patients.

By allowing higher rates of false positives, more drugs will get approved for the most critical diseases, but the impact of side effects may increase as well. To address this concern, the authors recommend creating a new category of "conditional approval" which lasts only for a few years. During this time, regulators, pharma companies, and doctors will monitor patients carefully and collect more data. Depending on how the patients respond, the conditional approval can either expire or be changed to full approval.