Precision medicine, or personalized medicine, is an initiative of the US government that aims to adjust, according to individual characteristics, the prevention and treatment of diseases with genetic, phenotypic, clinical, environmental and lifestyle data. For this reason, the federal government, through the National Institutes of Health (NIH), is financing projects to sequence more genomes, create large bio-banks and generate big data studies from electronic medical records and all types of electronic devices of diagnosis and monitoring (for more information we advise you to visit the post of Cristina Roure on the subject).
The idea, in principle, seems interesting, because the current biomedical research is based on standard patients and offers unique measurement solutions, therefore many people, because of barely quantifiable individual circumstances, receive maladjusted treatments. Now, however, we’ve begun to accumulate data on the difficulties and limitations of what promised to be a radical revolution in medicine. For this reason it seemed appropriate to comment on the article "Seven Questions for Personalized Medicine" by Michael Joyner (Mayo Clinic) and Nigel Paneth (Michigan University). The seven reflections, proposed by these authors, are the following:
Having the genome contributes little or nothing to improving the forecast (and prevention) of risks
Notwithstanding the great expectations of the human genome project, the genetic variants known so far add little additional predictive power to the known environmental algorithms. On the other hand, it was thought that when people knew the risks inherent to their genome, they would be more predisposed to adhere to lifestyle change interventions (if it was deemed appropriate), but this, in practice, is not happening.
Pharmaco-genetics is not meeting expectations
There is evidence that personalized treatment of cancer (one of the jewels of the project) is not meeting expectations. To demonstrate this claim, the authors have chosen two examples: a) The genetic variation BRCA1/2 affects 5% of women with breast cancer, for this reason, the added value that this genetic finding brings to clinical outcomes, 25 years after its discovery, is still very limited to a small group of women, so it has had no influence on the fall, by one third, in the mortality from breast cancer that occurred in the same period; b) Along the same lines, in the 1980s it was discovered that some patients with cystic fibrosis had a mutation in a transmembrane conductivity regulatory gene (CFTR). Two drugs were developed to alleviate the effects of this mutation, with discrete results (an improvement in 5-10% of maximum forced expiratory volume in the small group of patients with the mutation), while survival and quality of life of people with cystic fibrosis has improved markedly thanks to increasingly more precise and adjusted clinical guidelines for each person.
The information in electronic medical records is of poor quality
The almost universal implementation of applications for medical records is not guaranteeing the quality of the data that doctors collect and, in addition, the lack of interoperability of computer systems between health institutions doesn’t help either. It seems, therefore, that health systems are far from being able to offer the phenotypic and clinical data that the precision medicine project would entail. To illustrate the problem, I suggest you undertake a test: try to search through the EHR of your institution for the abdominal perimeter and the level of physical activity of the people that they take care, just to put two examples of data (one phenotypic and the other lifestyle), which are considered essential for the project of precision medicine.
Big data is far from contributing value to biomedical research
Leaving aside the debate on the ethics of data donation, the suggestive big data methodology, in which, through data mining, the system, in a more or less intelligent way, discovers associations of variables without having to raise previous hypotheses, is having development difficulties, due to what has been raised previously: genomic data are difficult to translate in the clinic, while phenotypic, clinical and lifestyle data are of poor quality. For this reason, the article authors propose to strengthen, as long as things do not improve, the methodology of the clinical trial and the population studies of cohorts.
Conflicts of interest have escalated to institutions and companies
The large amount of public and private money that generated the project of precision medicine has caused research institutions and related companies to compete for investment funds and in doing so, feed a bubble of projects and patents with a more commercial than scientific spirit.
There is an increase in over diagnosis and overtreatment
Many healthy people, when they know the risks of their genome, are compelled to consume diagnostic and therapeutic resources, or even surgical (remember the case of Angelina Jolie). At the "Preventing Over diagnosis" (Quebec 17-19 August 2017), Gregory Feer (Darmouth Geisel School of Medicine and associate editor of JAMA) stated that, while the supply of genomic studies is growing, doctors, overwhelmed by this tendency, continue improvising recommendations, without knowing how they can develop reasonable criteria for action.
Low benefits for public health are detected
The determinants of health continue to be the factors that most strongly influence the longevity and quality of life of people, while precision medicine is a project with obvious limitations, which makes a lot of sense for certain groups and for certain novelty therapies, but doesn’t seem to be in the best position to combat community health problems, or to revolutionize the practice of medicine.
As Joyner and Paneth say, advocates of precision medicine should temper their narrative and therefore should communicate more reasonable expectations to an audience, today, astonished by the enthusiastic proclamations of so many commercial advertisements.