Gerry Higgins
“The FDA considers both unwanted harmful effects and lack of efficacy to be adverse events. When considering candidates for pharmacogenomic relabeling, the FDA evaluates the severity of the adverse reaction, the degree to which genetics predicts the reaction, and the total number of individuals likely to be affected. These factors are important in assessing the potential public health impact of a recommendation to consider genetics when using the drug.” (see http://www.fda.gov/Drugs/ScienceResearch/ResearchAreas/Pharmacogenetics/ucm083367.htm).
So, what are the implications for the hospital physician? To date, FDA labeling has required, recommended, or mentioned gene testing prior to prescription in about ten percent of the labels for drugs approved by the FDA. At the recent Personalized Medicine World Conference held at Stanford, an FDA official said “things are changing hugely” at the agency in this domain. Many hospital and private practice physicians are not familiar with the myriad of gene tests that are commercially available, how the consequences of an individual’s patient’s genetic variation impacts dosing, and may not be familiar with the terminology of personal genomic variation. Pharmacogenomic Clinical Decision Support systems, combining clinical and pharmacogenomic data, have the potential to support the clinician per FDA gene testing requirements. This presumes that such software applications have been developed, and are integrated with the Electronic Health Record or as stand-alone software – but that is certainly not the case as of early 2010. In addition, approved hardware platforms and certified laboratories are required for testing a patient’s genetic markers that are associated with variability in drug response, and this is another area in which hospitals need to develop or expand their own laboratories or outsource to companies like LabCorp and Quest Diagnostics – but time is often key in drug prescription. The only sure bet is that a surge of new FDA gene testing requirements will eventually impact clinical practice
Some definitions from the FDA: Pharmacogenomics (PGx) is defined as the study of variations of DNA and RNA characteristics as related to drug response. Pharmacogenetics (PGt) is a subset of pharmacogenomics (PGx) and is defined as the study of variations in DNA sequence as related to drug response.
Anthony Guerra says
Hi Gerry,
What are the liability implications here? One of the big concerns among physicians is that more information — delivered through the EMR — means more liability. If a patient’s genome sequence lives under a tab in the EMR, will every physician be required to understand its implications on the patient’s medication regimen?
Gerry Higgins says
Anthony-
That’s a great point. I might reverse the situation and say that if a physician disregards genomic, proteomic, metabolomic or other tests, they are liable for not using the best diagnostic information available to them. What I see as the really big challenge is how to educate the physician such that they can apply this information in the best possible way at the time of critical diagnosis, and there is a tremendous need for even basic education in the “new genomics” – for example, What is an allele? Or a SNP? Patients are already coming to doctors with their personal genomic information gathered from the DTC companies (e.g., 23andMe, Navigenics), and what is the busy physician going to do with these data?
Anthony Guerra says
Do physicians generally receive sufficient education in medical school to understand the genomic information you are talking about? Perhaps this problem goes back to the schools.
Getting that information into the EHR in a way that can be understood within a few seconds at the point of care is another issue.
Very complicated stuff.
marcdparadis says
Most physicians in this country had their basic biological and medical training over 20 years ago, before the human genome had been sequenced and when pharmacogenomics was a rogue academic theory. So even if today’s pre-med programs and medical schools are teaching the subject, there is a generational disconnect in fundamental science and medical training here.
Proteomics and metabolomics are even newer beasts. No doubt all of these -omics hold great clinical promise but I would argue that they almost deserve medical specialties in and of themselves.
Gerry Higgins says
marcdparadis-
Great point.
Even now, many medical schools in North America are not adequately covering the human genome, proteome, etc, whereas schools in Europe and Asia are putting more emphasis on training physicians in the basic science of these domains. I agree that most physicians today do not have a sufficient base of knowledge to understand and discriminate between the various results of the “white-hot” explosions in the biology of the post-genomic world that are immediately relevent to clinical care. It used to take 17 years for scientific breakthroughs in biomedical research to trickle down into clinical medicine. Now that timeline is on the order of 4-5 years. But, today’s physicians will soon will be required to use and understand these data with the advent of the Genome-Enabled EHR (see my note on – “Feds Enable Genome-based EHR.”). Where are we going to recruit the clinicians / scientists that actually understand both bioinformatics and clinical informatics?
jbormel says
One of the most prescribed drug classes is statins. I personally know at least a half dozen patients (mostly physicians) who cannot tolerate them. There are genomic indications and contraindcations that are known to drive this.
http://www.ncbi.nlm.nih.gov/pubmed/18946323?itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_RVDocSum&ordinalpos=1
No one to date as been offered genetic testing. I personally was switched twice to alternative statins. This suggests to me that the application of science in general is well below the power of interpretation now available. This is independent of EHRs although EHRs could be part of the treatment of poor access and application of available clinical knowledge.
Gerry Higgins says
jbormal-
The best recent review of the pharmacogenomics of statins is in the Nature publication called The Pharmacogenomics Journal. The review is free, and can be found at: http://www.nature.com/tpj/journal/v10/n1/full/tpj200954a.html
Genaissance Pharmaceuticals, Inc, was developing a PGx gene test for statin responsivity – you might check the Pharmacogenetics and Pharmacogenomics Knowledge Base web site for more information: http://www.pharmgkb.org/
Thanks – Gerry