Paul Howard and Peter Huber argued in the fourth chapter of their new book, “Unlocking Precision Medicine,” that to realize what the National Institutes of Health calls an “approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person,” our society should rely on the tech sector—not the government—to spur precision-medicine innovations. The authors asserted that the “tech industry’s network of ‘permissionless innovation’ flourishes because its companies don’t have to navigate dense webs of federal regulations before launching new products...”

Howard and Huber claimed that the tech industry’s drive for relentless productivity is the very force that keeps its thumb firmly pressed on the pulse of organizations and individuals alike, which makes iteratively customizing and improving subsequent products and services possible.

Compare that to the costly, constrained, multi-year approval cycle of the U.S. Food and Drug Administration (FDA), which must give its okay before anyone outside of clinical trials can benefit from a given treatment.

“There is very little in the way of rapid or real-time feedback loops in the (healthcare) industry—even though the technology for it is available today,” wrote the authors.

Shockingly, Howard and Huber suggested that most drugs might not actually be failing the FDA’s tests, and that, rather, the drug-testing protocol that the FDA uses is probably faulty:

These trial designs were first used in the 1930s, and are built on the assumption that patients with similar clinical symptoms are suffering from the same disease. But that’s an assumption that has been disproven by modern science over the past few decades—especially after the deciphering of the human genome in 2000.

The authors praised new technologies that have revealed that “many seemingly common disorders—so defined by their apparent clinical symptoms—are in fact clusters of biochemically different diseases (breast cancer, for instance, is a combination of at least nine different diseases).” Further, Howard and Huber noted that “traditional symptom-based definitions of singular diseases are still used to frame most FDA-approved clinical trials, leading to the testing of one drug at a time in patients that are assumed to have the same condition.”

The authors suggested that the answer to this mishandling of data is precision medicine, and that its success “hinges on two deeply intertwined strategies.”

The first is targeting. The companies creating the drugs need to be empowered to “target disease-propelling molecular factors (‘biomarkers’) and pathways.”

The second is identification. Doctors must have the ability to use key tests to:

  1. Single out “patients in whom those factors or pathways are present” 
  2. Quickly match those patients with the right drug
  3. Follow up with those specific patients and adjust their treatments, if necessary

“Left to expert physicians with access to large patient databases that allow them to share reliable information quickly,” wrote Howard and Huber, “this feedback strategy can produce a rapid-learning health system that improves care based on what we learn from every patient’s response to treatment.”

Purchase the full book here.


Learn about the genomics-health IT gap for precision medicine: read the Chilmark report now!