A simple definition of personalized medicine is healthcare designed for each individual patient. I think it is a worthwhile long-term healthcare objective, with emphasis on long-term, because clinical data on the benefits (to both patients and health systems) of such approaches, along with the required array of diagnostic tools and treatments, is currently not available. Personalized medicine has achieved a substantial buzz level despite limited, slow progress to date. This blog contains some of my personal thoughts on defining personalized medicine and its path forward.
The first step in current medical practice is generally the diagnosis of the presence or absence of various medical conditions. The standard tools include family medical history, blood and urine analysis, ECG and basic medical imaging. The first step in personalized medicine will also be diagnosis but with an added focus on the relative risk for disease development using a more complex set of tools. One idea is that an enhanced medical profile, probably including a genomic / proteomic profile, will allow the assessment of an individual’s relative risk of developing a series of serious medical conditions. However, many questions remain around how this will be implemented and what the benefits would be:
- Which serious medical conditions should be assessed in an enhanced medical profile?
- For which serious medical conditions could a relative risk be assigned today based on an enhanced medical profile?
- When will it be reasonably possible to assess relative risk for all serious medical conditions included in an enhanced medical profile? Are new diagnostic tests or technologies needed or just long term clinical data?
- What is the cost of the initial profiling?
- Will the profiling have to be repeated? If yes, at what cost?
The next step in medical practice is to treat the medical condition. There are serious medical conditions for which there are no cures and treatments have limited benefit. Will an enhanced medical profile provide any benefit for these patients? I believe that the answer is yes, since the patients, their families and the healthcare system can prepare for the care and treatment which these patients will need. In addition, there will be a profiled population available for research and clinical trials, involvement in which must still require the patient’s informed consent.
There are some serious medical conditions which can be prevented. The two major tools which are available now for disease prevention are vaccines and lifestyle adjustment.
- Vaccines have been able to eradicate some very serious diseases (i.e. smallpox), substantially reduce the incidence of most childhood infectious diseases, and more recently impact the incidence of some adult medical conditions (hepatitis B and cervical cancer). However, development of novel vaccines is difficult – nothing yet for HIV and hepatitis C – and expensive. It does not currently make economic or clinical sense to personalize childhood vaccination programs – the universal programs work at reasonable costs. Adult vaccination programs are personalized taking into consideration the potential exposure based on type of work and travel.
- Lifestyle adjustment – the worldwide healthcare system could potentially reallocate trillions of dollars spent on treating the results of only two lifestyle problems, obesity and smoking. These are conditions where the patient probably needs personalized human support during lifestyle adjustment more than they need personalized diagnostics and drugs. Healthcare systems are good at fast treatment with defined tools that produce immediate results – they are less effective with soft tools that take much longer to show any potential benefit.
There are many medical conditions for which most patients are adequately treated with the current basket of diagnostics and drugs. It would not make clinical or economic sense to do universal enhanced profiling for these medical conditions. It would probably make sense to do enhanced profiling for those patients where current therapies do not adequately control these medical conditions.
Personalized medicine can probably provide the greatest benefit for serious medical conditions such as cancer. New cancer therapies have specific targets, including over-expressed cell surface antigens and enzymes in the complex web of growth signaling pathways. The cancer drug basket was once filled primarily with non-specific cytotoxics, but new additions to the basket are usually targeted therapies. Instead of a simple visual assessment of biopsy samples looking for abnormal cells, personalized medicine might eventually include an assessment of all drug targets and even an assessment of the activity of various drug combinations against biopsy samples.
The buzz achieved by personalized medicine is a two-edged sword: It looks great during fund-raising campaigns and on grant applications, but if expectations are too high and progress continues to be both limited and slow, governments, philanthropists and investors will turn their attention elsewhere.
For any agency, fund or individual considering an investment in personalized medicine, there are many questions which need to be asked, including:
- What is the product or service being developed;
- Is there a path to regulatory approval (if needed);
- Assuming successful development, can it be used on approval or are other products or long-term data needed for it to be clinically useful;
- Will use of the new product require healthcare professionals to make major changes to the way they practice medicine; and
- Does the clinical benefit justify the proposed pricing?