P4 Medicine – How do we get from current practice to future progress?

In the current issue of Nature Reviews in Clinical Oncology, Stephen Friend and Leroy Hood – the man who first described predictive, personalized, preventive, and participatory medicine as “P4 medicine” emphasize that clinical trials of the future will need to be designed so they fully capitalize on P4 medicine. Specifically with the advent of high-throughput genomic medicine, Friend and Hood note that science is experiencing a paradigm shift: what we once thought were single diseases, are in reality multiple distinct molecularly defined disease states. The natural extension is that trials previously needing 10,000 patients to show a benefit will now need tens of thousands of patients to have adequate statistical power to detect the same level of benefit. Even in large academic medical centers, the authors note, it will be difficult or even impossible to rapidly accrue proportionally small patient subsets in which personalized interventions can be feasibly explored and tested. The authors suggest the use of patient driven networks to obtain the patient numbers needed for these trials.

We at the center typically think about patient-physician interaction and patient empowerment, activation, or participation when we talk about the “participatory” P of P4 medicine. Friend and Hood extend the definition of participatory to include efforts by patients to enroll themselves in networks whose goal it is to provide access to clinical trials that would otherwise be beyond the reach of single, brick and mortar institutions. Given the authors’ backgrounds, they are understandably preoccupied with the genetics and molecular aspects of such trials, but their concept still applies to models of personalized medicine that may not involve molecular data.

Even for purely non-genomic based disease models, we are beginning to appreciate the importance of individual variation in response to therapy. We already have the tools to perform personalized, risk-based prevention and treatment for many disease states using family history, biochemical markers, and patient preference. For prospective health care to reach its maximum potential in the future, we will need to develop new models that allow iterative improvements in personalization and intervention outside of current models of clinical trials. Much in the way that the physiology-based practice of medicine of the 21st century came to embrace evidence based medicine (EBM), personalized medicine will likely require some analogous, personalized equivalent to maximize its full potential . One possibility is patient driven networks as proposed by Friend and Hood. Another is novel statistical modeling that would allow us to personalize interpretation of conventional clinical trials.

All the tests in the world don’t mean much without a plan

I recently spoke to a representative from the Indian Health Service about “personalized medicine” in federal initiatives. Of particular note, I mentioned during the course of our conversation that the Patient Protection and Affordable Care Act includes two stipulations for personalized health planning. Under the new law, Medicare will pay for one personalized health planning visit annually. Additionally, personalized health planning was a key initiative for demonstration projects through community health centers. Interestingly, despite the IHS’ focus on personalized approaches to care, this individual knew nothing about personalized health planning within the law. So I decided to see what the federal government actually says about personalized health care. Here’s what I found.

The government has great hopes for personalized approaches to care. In fact, “personalized health care will improve the safety, quality, and effectiveness of healthcare for every patient in the US.” The government also has a very limited definition of personalized health care, which is “using ‘genomics’, or the identification of genes and how they relate to drug treatment” as the means to “enable medicine to be tailored to each person’s needs.” But how could this be? What about Sec. 4206 of PL 111-148, “Demonstration Project Concerning Individualized Wellness Plans.” Or Sec. 4103, “Medicare Coverage of Annual Wellness Visit Providing a Personalized Prevention Plan?”

The omission may in part be due to the fact that personalized health planning hasn’t caught on as a concept yet, despite its arguable potential to change the way health care is delivered – and perhaps more importantly, to improve health outcomes, as we’ve seen with coronary heart disease and type 2 diabetes here at Duke.  Instead, many people are wide eyed by the concept of fancy tests and SNPs and the prospect that if we decode a patient’s genetics we will have found the silver bullet and not only cure patients of disease, but prevent disease altogether.

Except we won’t.  At least not in the foreseeable future for the health problems that are currently costing us the most money.  Because science tells us that our genes are only one part of the story. Our environment is another – in many cases more important – contributor to our health. In fact, our environment can actually change the way in which our genes are expressed. We’ve see this phenomenon in nature. We’ve seen it in people.  And in both cases, we’ve found that there’s no way to predict exactly how the environment will change the phenotype.

But there is something we can do successfully for patients at risk for chronic disease. We can track those environmental factors known to affect gene expression – like nutrition and stress and environmental exposures to name a few. It’s called personalized health planning, a strategic approach to care that identifies all a patient’s risk factors for disease (genetic, environmental, behavioral, psychosocial) and very importantly tracks those factors that the patient  can do something about – either themselves or with the help of modern medicine.  Three quarters of national health care expenditures are for chronic diseases due to health behaviors, namely smoking and obesity. If we really want to improve the health of the nation while reducing health care costs, expensive genetic tests cannot be synonymous with, or the primary road to, personalized health care. We can run all the tests available in modern medicine. (Some might argue we already do.) But if we don’t help patients to plan for their good health in the context of their known risks from all those tests, we’ve failed them.