New advances in genomic testing have the potential to radically change the way cancer drugs are developed and prescribed, offering personalized therapy with treatment tailored to the patient’s tumor genome. With the advent of this “personalized medicine,” scientists are hopeful that drugs shown to work in a cancer caused by a specific genetic mutation that appears in one anatomical site – the breast, for example – may prove to be successful in treating a cancer caused by the same type of mutation that occurs in another anatomical site – the liver, for example, or the colon.
In a new Perspective piece in the New England Journal of Medicine, Harvard T.H. Chan School of Public Health Acting Dean David Hunter looks at the challenges of maintaining the rigorous current model of drug design — through randomized, controlled trials (RCTs) testing safety and efficacy — in this new paradigm.
The article appeared online August 19, 2015.
Hunter, who also is Vincent L. Gregory Professor in Cancer Prevention, notes that genomic sequence-informed drug design has been successful in several cases involving common cancer types and target tumor mutations. But each patient’s tumor has a spectrum of mutations, he writes, making it difficult to aggregate information from the hundreds or thousands of patients usually needed to get the kind of assessment of effectiveness and safety yielded by randomized controlled trials.
A concern is that with a growing number of centers and institutes for personalized cancer therapy springing up, the temptation will be to use existing cancer treatments “off-label” to treat tumors with apparently the same genetic basis, sidestepping the usual drug testing through clinical trials that has been the cornerstone of care to this point. The cost of some of the new drugs, greater than $10,000 per month of treatment, is another factor that should mandate high quality standards of effectiveness.
“Are we to replace guidelines for drugs based on rigorous standards of evidence derived from RCTs with educated guesswork based on anecdotal experience and case reports?” he writes. “As the saying goes, ‘the plural of anecdote is not data.’”
He, and co-author Ralph B. D’Agostino Sr. of Boston University, offer recommendations for addressing the need for data, starting with the use of “basket studies” that enroll patients with a tumor in any of various tissues that has a specific mutation thought to be responsive to a specific drug in order to prioritize drug/mutation/tumor combinations in which the larger more rigorous studies should be performed. Other recommendations include the development of large, transparent, databases of evidence on drug efficacy and long-term toxicity for new drug/mutation/tumorcombinations not tested in the clinical trial context.
Read the NEJM Perspective: Let’s not put all our eggs in one basket
Learn more
Greater focus needed on cancer prevention (Harvard Chan School news)
Most cancers not just ‘bad luck’ (Harvard Chan School news)
photo: iStock