There has been a significant interest in the development of drug therapies that are “tailored” to the individual. This is driven by the simple reality that each of us responds differently to pharmaceuticals. Some of us are more sensitive, needing less to effect a positive result, some are highly intolerant – leading to organ damage or even death. The pharmaceutical companies have a very difficult time in developing a “standard dose” that will provide treatment and not cause damage to the majority of patients. But even then, we find cases where medications deemed safe through trials are found at a later date to cause harm in a particular population cohort.
Personalized treatments can take on numerous forms.
One concept is to develop diagnostic feedback systems that measure the amount of active ingredient actually absorbed into the bloodstream. In this way, an effective dose can be maintained while ensuring that the patient never receives a damaging dosage.
Another focuses on treatment strategy. There are often a range of medications available for a specific condition for which people respond differently based upon body chemistry. By having a tool identifies the most effective medication based upon the particular body chemistry of the patient, one eliminates the typical guesswork that physicians are forced into. This opportunites is enabled through the identification of DNA and/or protein based “markers” that differentiate between different body chemistries, enabling the physicians to prescribe the most effective treatment without the guesswork.
Lastly, it is currently not possible to predict how medications affect particular patient cohorts, so the concept of developing medications with a specific body chemistry group in mind is not yet there. Instead, we currently simply try medications on a trial population and see how effective it is across the board. This is done in the “Phase 3″ trial stage for all medications being prepared for market. Many potential products never make it to market after Phase 3 trials – drugs are found to only benefit a small portion of the trial population, or are found to cause organ damage in other cohorts, or have significant side-effects in another group.
What is clear, however, is that some portion of the populations can and do see positive results, but since the trials are designed to find a treatment for the entire population, the smaller cohort may never be even discovered, which leads to a remarkable business opportunity – data mining of drug trial data from abandoned phase 3 trials to attempt to find cohorts that see benefits from the treatments, and then finding biomarkers that enable one to identify and select members from this cohort.
In order to accomplish this, one needs the following – knowledge of what trials have been conducted, their outcomes on an individual level, a general set of biomarkers that can be used as indicators, samples from each of the patients to test against, a simple way to test the samples for the markers, and statistical analysis tools to identify the markers with the greatest correlation with positive outcome. There are now straightforward ways of addressing each of these.
Trial Data and patient samples: Remarkably, both trial data and patient samples ( blood etc.. ) are kept long after phase 3 drug trials are completed – even if the trials turn out to be failures. Samples are kept in cold storage and patient records are kept by the pharmaceutical companies. To the extent that the pharma companies see an opportunity to recoup some of their losses from a failed trial, they will be very interested in providing access to this data
General biomarkers and testing processes: With the advent of microarray DNA analysis systems, it is possible to screen for hundreds, if not thousands of individual markers in a single test step. These types of screening systems are essentially off the shelf, with the only question being what set of markers one would want to test for.
Statistical analysis tools: Given that this procedure is nearly the same as what is being used by the bioinformatics community in their “shotgun” approaches to finding new medications, these tools are widely available, some even in the public domain.
Impacts on trials: what is unclear is how the FDA will react to the new data – will it be required to perform an additional Phase 3 trial based upon the combination of diagnostic/therapeutic, or will the evidence from the original trial data be considered sufficient to approve the product/process. In either case, the opportunity to resurrect potentially effective treatments with a relatively low initial investment is compelling.
Finally, as a key element of success, The FDA has confirmed that personalized medicine strategies that impact small cohorts qualify under the “orphan drug” program, meaning that there are significant reductions in barriers to market entry. This is extremely important, as it reduces time to market and additional costly clinical trial requirements.
For more information contact steven <dot> warwick <at> innovationscommercialization <dot> com