personalized medicine: something old, something new EDITORIA from chronic myelogenous leukemia as a result of and a lot of genome-wide association studies. As the rare Philadelphia chromosome translocation, heady as that information is, it is a long way from we have a drug that addresses that phenotype, at being completely or even adequately descriptive least temporarily. If your form of breast cancer is for diagnostic, much less therapeutic purposes overexpressing a specific gene, we have a drug that One need look no farther than the uproar that may work better for you than for others without followed the founding of genomic information that genetic variation. And so forth. However, companies such as Navigenics and 23andMe for most of the diseases for which we are gaining although questions were loudly raised about insight, we are still struggling with therapeutic privacy and ownership of genomic information, options. But there is real hope, buried within the the real question is whether this information in confusing new genomics-based language emerg- its current woefully incomplete form is worth ing from the lab bench into the clinic, that we anything from a medical perspective. I suspect ay find more effective new treatments. This not, or at best its worth is minimal and is largely evolutionary change, perhaps a period of punc- misunderstood by patients and their doctors tuated equilibrium, is still in line with all that 17-9]. Perhaps most telling is the blind faith of has gone before some customers of these companies, who have, essence, replaced rational knowledge with an odd The ability to more precisely describe kind of genetic determinism metaphysics, mir- phenotypes has allowed us to change the roring the surge of spiritual and mystical prac specifics, but not the fundamental practice tices that arose in response to the incompleteness of the humors paradigm By no means am I trying to belittle the progress"We particularly need to be cautious about made. Like the early rise of toxicology and drug what our current state of genomic knowledge treatments in the 15th and 16th centuries, and tells us-and does not tell us-about human he ability to see beyond the limitations of the disease and its treatment human eye in the 17th, the toolbox of medicine btedly be changed Incorporation of new findings into regula in the 21st as well, thanks to our ability to ever medical practice requires that the language of the more precisely define phenotype with genomic omics be translated into clinically relevant and tools and insights. Personalized medicine, which meaningful terms, and be tied to some actionable has been around a long time, will be ever more outcome. Driven to date by molecular biologists exquisitely personalized based on better phe-( for the most part), there remains a significant gap notype definitions, though perhaps not as fast between genomic knowledge and clinical prac- as suggested in the President's words or by the tice. At its heart is the silence that meets the most various groups that have formed to promote this common question from practicing physicians newold field. I believe that the primary chal- what difference does this make in how I trear lenges we face are threefold, in decreasing order patient X? There are a few examples of answers urgenc for some diseases, but that fundamental medical Sorting out meaningful phenotypes from question is still unanswered in the vast major- ity of cases. This must be acknowledged, even as many academic centers are racing to focus on Incorporation of this knowledge and language the translational research needed to drive bench into medical ractice findings to the bedside, making genomic knowl Incorporation of this refined diagnostic ability edge relevant to practicing physicians and their into healthcare and regulatory systems built patients. And by relevant, I mean that it has to be around older medical paradigms more than a statement of molecular phenotyp in other words, it has to be accompanied with ar Previous(r)evolutions went through similar option for addressing the problem. steps,and there is no reason to believe the current One particularly interesting potential effect of one is any different. his new knowledge in driving clinical practice We particularly need to be cautious about what may be the shift of focus away from traditional our current state of genomic knowledge edical divisions based on organs and organ and does not tell us -about human pathology towards a more mechanistic disease and its treatment. Much of our current description based on cellular pathways [Iol edge is based largely on a few complete genomes is not too far a stretch to think about doctors www.futuremedicine.cfuture science group www.futuremedicine.com 3 Personalized medicine: something old, something new Editorial from chronic myelogenous leukemia as a result of the rare Philadelphia chromosome translocation, we have a drug that addresses that phenotype, at least temporarily. If your form of breast cancer is overexpressing a specific gene, we have a drug that may work better for you than for others without that genetic variation. And so forth. However, for most of the diseases for which we are gaining insight, we are still struggling with therapeutic options. But there is real hope, buried within the confusing new genomics-based language emerg￾ing from the lab bench into the clinic, that we may find more effective new treatments. This evolutionary change, perhaps a period of punc￾tuated equilibrium, is still in line with all that has gone before. “The ability to more precisely describe phenotypes has allowed us to change the specifics, but not the fundamental practice of medicine.” By no means am I trying to belittle the progress made. Like the early rise of toxicology and drug treatments in the 15th and 16th centuries, and the ability to see beyond the limitations of the human eye in the 17th, the toolbox of medicine will undoubtedly be changed in profound ways in the 21st as well, thanks to our ability to ever more precisely define phenotype with genomic tools and insights. Personalized medicine, which has been around a long time, will be ever more exquisitely personalized based on better phe￾notype definitions, though perhaps not as fast as suggested in the President’s words or by the various groups that have formed to promote this ‘new’ old field. I believe that the primary chal￾lenges we face are threefold, in decreasing order of urgency: n Sorting out meaningful phenotypes from irrelevant ones; n Incorporation of this knowledge and language into medical practice itself; n Incorporation of this refined diagnostic ability into healthcare and regulatory systems built around older medical paradigms. Previous (r)evolutions went through similar steps, and there is no reason to believe the current one is any different. We particularly need to be cautious about what our current state of genomic knowledge tells us – and does not tell us – about human disease and its treatment. Much of our current knowl￾edge is based largely on a few complete genomes and a lot of genome-wide association studies. As heady as that information is, it is a long way from being completely or even adequately descriptive for diagnostic, much less therapeutic purposes. One need look no farther than the uproar that followed the founding of genomic information companies such as Navigenics and 23andMe: although questions were loudly raised about privacy and ownership of genomic information, the real question is whether this information in its current woefully incomplete form is worth anything from a medical perspective. I suspect not, or at best its worth is minimal and is largely misunderstood by patients and their doctors [7–9]. Perhaps most telling is the blind faith of some customers of these companies, who have, in essence, replaced rational knowledge with an odd kind of genetic determinism metaphysics, mir￾roring the surge of spiritual and mystical prac￾tices that arose in response to the incompleteness of the humors paradigm. “We particularly need to be cautious about what our current state of genomic knowledge tells us – and does not tell us – about human disease and its treatment.” Incorporation of new findings into regular medical practice requires that the language of the omics be translated into clinically relevant and meaningful terms, and be tied to some actionable outcome. Driven to date by molecular biologists (for the most part), there remains a significant gap between genomic knowledge and clinical prac￾tice. At its heart is the silence that meets the most common question from practicing physicians: what difference does this make in how I treat patient X? There are a few examples of answers for some diseases, but that fundamental medical question is still unanswered in the vast major￾ity of cases. This must be acknowledged, even as many academic centers are racing to focus on the translational research needed to drive bench findings to the bedside, making genomic knowl￾edge relevant to practicing physicians and their patients. And by relevant, I mean that it has to be more than a statement of molecular phenotype; in other words, it has to be accompanied with an option for addressing the problem. One particularly interesting potential effect of this new knowledge in driving clinical practice may be the shift of focus away from traditional medical divisions based on organs and organ pathology towards a more mechanistic disease description based on cellular pathways [10]. It is not too far a stretch to think about doctors