Eric Topol, M.D., describes the mode of action of cancer therapeutics. Most of the successful therapeutics target mutations that perturb the normal regulatory control of cell function. He points out that the human genome has only 20,000 genes, but the number of regulators is over 4 million. Most regulators act by inhibiting growth.
In contrast, the majority of cancer therapeutics are described by primary location such as nonsmall cell lung cancer, ovarian cancer, etc. Organ location was set by the cohort selected in the clinical trials. Although they may be from the same tissue or organ, these cancers often show little commonality in biochemical function, i.e., cancer is usually heterogeneous.
Recognizing this leads to new questions that seem to be more relevant. For example, would it be more useful to characterize cancers by their mutation rather than by the location? Topol’s short answer is, “It’s the mutation, stupid!” He cites examples where recognizing a common mutation leads the physician to select a therapeutic that had originally been approved for a cancer in a different organ.
However, regulators and insurance companies, particularly in the U.S.A., appear to be in a dither about off-label use of therapeutics. Yes, there is potential for abuse. Also, key concerns about safety, efficacy, dose, and side effects will probably not be addressed. But with potentially lethal diseases, the option of waiting for results from a clinical trial on a known mutant is not attractive. Indeed, according to current practice, if the FDA were to demand a placebo-controlled clinical trial, half of the cohort would be sacrificed to provide an estimate of improved clinical effectiveness (extension of lifetime).
It seems that it would be better to start classifying cancers by their mutation(s) and correspondingly the mechanisms of action of the therapeutics. This would help match the disease to the therapeutic. Hopefully, one clinical trial would be sufficient to assure safety and efficacy. Over time, many real patients would provide data to support findings of efficacy.
Cancer patients are generally not looking for marginal improvements during their lifetime. They hope for remission and cure. Life extension is interesting, but much more difficult to evaluate, especially if the treatment entails pain and suffering. We all know torture when we see it.
In the future, patients presenting with diseases need to be stratified into cohorts that are useful in defining therapies. Initially, this would probably involve errors in cohort assignment. However, with refinement and study, the selection of “peers” will become strict and potentially diagnostically and therapeutically useful.
In the future, when there is a choice of therapies, perhaps one can use in vitro assays on live cells from the patient to estimate comparative efficacy.
Robert L. Stevenson, Ph.D., is a Consultant and Editor of Separation Science for American Laboratory/Labcompare; e-mail: firstname.lastname@example.org.