How do we deliver on the promise of personalized medicine?

In his State of the Union address, Barack Obama announced the intent to invest in personalized or “precision” medicine – the tailoring of treatments to an individual’s genetic code.  Canadian funding agencies have also enthusiastically supported this idea, and all University of Toronto-affiliated hospitals promote their institutions’ forays into personalized or precision medicine. Collectively, these pronouncements seductively portend a transformative style of medicine in which your doctor selects from a cornucopia of potential treatments to find the one that’s “right for you”. While this vision is worth working towards, to get there in our lifetime we’ll first need to know a lot more about the fundamental biology that underlies all disease, and we’ll also need to get a lot better at inventing medicines.

The push for personalized medicine is driven by treatments that have been shown to work in genetic sub-populations of patients. Herceptin was found to be effective in a genetically-defined subtype of breast cancer, and the wonder-drug Gleevec has largely turned a specific genetic type of leukaemia from a death sentence to a chronic disease. These life-saving advances connected genetics and drug discovery in a personal way, to some extent, and created the dream that all medicine will be practiced this way.

The hope is that doctors can stop managing clinical disorders that have all-encompassing names like “autism”, and based on each patient’s DNA sequence, precisely classify the disease – for example “autism-spectrum disorder 12.5.445.12”, and then simply match the genetic code with the appropriate treatment from a pharmaceutical cornucopia.

What people rarely talk about is the unfortunate truth that this cornucopia does not exist, it will not exist for decades (if ever) and it will cost tens of trillions of dollars to create. Even if we achieve a perfect genetic understanding of all diseases, there will never be enough medicines with which to practice personalized medicine.

The fundamental problem is that finding the genetic cause of a disease is easy compared to understanding what the faulty gene does normally, and how the mutation causes the disease. And understanding gene function is in turn far easier than discovering a new medicine to fix it.

A history lesson is instructive.  In fact, we already have a lot of the genetic pre-requisites to practice personalized medicine. Over the past decades, scientists and doctors have precisely classified thousands of diseases based on the genetic code. Yet for all but a handful of these diseases, there is no treatment – despite a lot of effort and funding.

Indeed over 50 years ago, sickle cell anemia was the first disease to be understood through its genetics – there is still no specific treatment.  In the late 1980’s, cystic fibrosis (CF) and Huntington’s disease (HD) were two of the diseases whose genetic faults were uncovered through the then new tools of molecular biology. Twenty-five years and billions of dollars later, there is one expensive (personalized) treatment that helps 1% of the CF patients and no treatments for HD. In fact, we still do not understand precisely how the HD mutation causes the disease.  And these are but 3 of 4,000 more diseases characterized precisely at the genetic level – with no medicine available.

We must face the facts.  History shows that knowing the mutation is just the beginning of the medical odyssey.  No amount of additional genetic classification or analysis of big data can make personalized medicine a reality in the near term, and no genomics revolution will change that fact.

To make “precision medicine” a reality, we first need to put genetics in perspective.  If we want to discover cures, we need to invest in science driven to understand fundamental mechanisms.  But that gets to an interesting question – we’ve been doing this for decades – so where are the cures?

I argue that the current model of research funding is part of the problem. Most funding systems around the world are risk-averse and promote “safe” science. As a result, for example, more than 60% of global (and Canadian) biomedical research investment in 2014 was focused on the relatively small number of genes that were exciting in the late 1990s.  Clearly, after 15 years and with no cures for most diseases in hand, the answers are not under those lights.  If we want cures, we need to be more daring with our science funding investments.

We also need to invest in better and more cost-effective approaches to invent medicines, because the current approach is too slow and expensive.  Each year, the public and private sectors invest tens of billions of dollars in drug discovery.  Unfortunately, most of this funding is spent by many organizations pursuing the same ideas in parallel and in secret, causing massive duplication of effort and resulting in no advancement of knowledge.  And since most of these ideas fail – this activity incurs great cost not only to each participant, but also to the system as a whole.  If we want to discover new medicines better and faster, we need to promote and support different organizational models of drug discovery, such as the open source models that have served other sectors brilliantly.

It is important to emphasize that there can be no personalized medicine without the medicines themselves.