The Right Drug for the Right Patient: Personalized Medicine Comes of Age
Time was when the phrase “personalized medicine” would have referred to something like the provision of medical care by a family doctor on a one-to-one basis, but for several years now indexers have had to take note of a special meaning—namely, the application of knowledge about an individual’s genetic make-up to a tailored approach to clinical management. “Molecular medicine” and terms such as genomics, proteomics and the rather less elegant pharmacometabolomics cover the same concept. Sometimes the focus is more on the genetics of the tumor in cancer cases, or on an infecting agent, such as the hepatitis C virus.
Although, as noted above, not every pertinent paper explicitly employs the phrase “personalized medicine,” tracking those that do provides a vivid illustration of the topic’s ascension in recent years. In the attached graph, ScienceWatch compiled the number of papers indexed each year by Thomson Reuters since 2001 that expressly use the phrase in their title, abstract, or keywords: from barely a dozen in 2001 to upwards of 750 in 2011—more than a 60-fold increase.
|(The most-cited paper of each year, of those featuring “personalized medicine” in the title, based on citations through May 2012)|
|2011||L. Chin, et al., “Cancer genomics: from discovery science to personalized medicine,” Nature Med., 17(3): 297-303.||14|
|2010||M.A. Hamburg, F.S. Collins, “The path to personalized medicine,” New Engl. J. Med., 363(4): 301-4, 2010.||94|
|2009||J.S. Ross, et al., “The HER-2 receptor and breast cancer: Ten years of targeted anti-HER-2 therapy and personalized medicine,” Oncologist, 14(4): 320-68, 2009.||109|
|2008||S.F. Zhou, et al., “Clinical pharmacogenetics and potential application in personalized medicine,” Current Drug Metabolism, 9(8): 738-84, 2008.||66|
|2007||W. Burke, B.M. Psaty, “Personalized medicine in the era of genomics,” JAMA, 298(14): 1682-4, 2007.||67|
|2006||J.K. Nicholson, “Global systems biology, personalized medicine and molecular epidemiology,” Molec. Systems. Bio., 2: Article no. 52, 2006.||72|
|2005||W. Sadee, Z.Y. Dai, “Pharmacogenetics/genomics and personalized medicine,” Human Molec. Genetics, 14(2): R207-14, 2005.||57|
|2004||A.D. Weston, L. Hood, “Systems biology, proteomics, and the future of health care: Toward predictive, preventative, and personalized medicine,” J. Proteome Res., 3(2): 179-86, 2004.||223|
|2003||M. Oscarson, “Pharmacogenetics of drug metabolizing enzymes: Importance for personalised medicine,” Clin. Chem. & Lab. Med., 41(4): 573-80, 2003.||39|
|2002||K.K. Jain, “Personalized medicine,” Curr. Opin. Molec. Therapeut., 4(6): 548-58, 2002.||48|
SOURCE: Thomson Reuters Web of Knowledge
Furthermore, for a modest sampling of literature on the topic, ScienceWatch assembled the attached table (which, like the graph, draws upon Thomson Reuters Web of Science). The table contains a selection of papers containing the term “personalized medicine” in their titles, listing each year’s most cited for 2002 through 2011.
FROM DISEASES TO DRUG EFFECTS
The idea that breast cancer is not a single disease is not new—for example, there are estrogen receptor positive/negative tumors and ones associated with the genes BRCA 1 and 2—but in April of this year much publicity attended the appearance in Nature of an Anglo-Canadian study revealing that breast cancer has, genetically speaking, 10 subtypes (C. Curtis, et al., Nature, April 18, 2012; DOI 10.1038/nature10983). Patients will not immediately benefit from findings of this sort, but in recent years Science Watch’s medicine section has provided examples hinting at practical benefits for this approach, such as antiplatelet agents and carriage of CYP2C19*2 and wild-type (non-mutated) KRAS and retuximab.
The principle, however, can be applied to unwanted drug effects as well. Indeed, the earliest contact I can recall with the potential of personalized medicine was 12 years ago in an essay by Dr. Allen D. Roses (Lancet, 355: 1358-61, 2000) in which one of the illustrations was the already well-established association between isoniazid-related peripheral neuropathy and alleles of the N-acetyltransferase 2 gene. Slow acetylators had higher than normal blood levels of the drug. Other examples are genetic screening for the avoidance of severe reactions to the anti-HIV drug abacavir and bleeding in patients on warfarin, where the right balance between wanted anticoagulation and unwanted hemorrhage is a delicate one. Few people have had their complete genome sequenced, but with the cost of sequencing falling all the time it might be feasible for us all to carry in our wallets and purses plastic cards containing a chip with personal genetic data, not complete perhaps but sufficient to assist physicians make better treatment decisions. But that’s a long way off.
At a research level, if not yet in routine clinical practice, personalized medicine has arrived, with all the trappings of a specialty.
At a research level, if not yet in routine clinical practice, personalized medicine has arrived, with all the trappings of a specialty. The European Commission even has a personalized medicine unit and the budget for facilitating research in this area over the five years 2007-11 was 900 million euros. In 2006 then-Senator Barack Obama introduced a Genomics and Personalized Medicine Act. This and other similar attempts failed but another try is expected soon. The objective is to smooth the path for personalized medicine—for example, by sorting out potential overlaps of agency interest and via specific research grants (A. Konski, Medicine Bulletin, Oct. 3, 2011; http://www.personalizedmedicinebulletin.com).
Since 2004, personalized medicine has had its own journal, and there are also plenty of dedicated international meetings and international collaborations. Personalized Medicine’s March, 2012 issue (available to all via www.futuremedicine.com) provides a good overview of this novel way of thinking in medicine and of some of the problems associated with it.
One difficulty could be genetic heterogeneity within a tumor, which might lead to “underestimation of the tumor genomics landscape portrayed from single tumor-biopsy samples and may present major challenges to personalized medicine (M. Gerlinger, et al., New Engl. J. Med., 366: 883-92, 2012.
Three years ago PricewaterhouseCooper was talking of a $232 billion personalized medicine market growing at 11% per annum. A more recent estimate puts annual growth at 9.5%, but either way this is big business. Not that profit is the only incentive for investment, for there are also important implications for the licensing of drugs , for healthcare providers, and for the design of clinical trials. Drug development is very expensive and time-consuming. If a new agent has a known molecular mechanism of action linked to specific genes or alleles of polymorphic genes, clinical trials could be done on patients selected on genetic grounds. Trials would thus be smaller, and successful studies would lead to the licensing of the drug for use in patients identified on genetic information and not for all with the disease in question—which, in theory at least, should reduce the drugs’ bill for healthcare providers.
Both the US Food and Drug Administration, which last year issued draft guidance for industry, and the European Medicines Agency via the work of its Pharmacogenomics Working Party, are starting to tackle the implications for licensing and for early-stage trials. If the evaluation of efficacy is potentially being simplified, the detection of adverse reactions would be more difficult with smaller clinical trials; remember, even large ones sometimes have occasionally failed to pick up side-effects so important that post-marketing withdrawal has been the unfortunate consequence.
A NEW WORLD
A glimpse of the new world of clinical trials can be had from a paper published online back in March and appearing in print on May 5th. One obstacle to translating findings in genetic research into clinical practice is the usually specialized and time-consuming nature of the tests, this being especially important in urgent settings such as percutaneous coronary intervention (where anti-platelet therapy is given after the procedure). In this Canadian study, led by Dr. Jason D. Roberts (J.D. Roberts et al., Lancet, 379: 1705-11, 2012), nurses with just a half-hour’s training did the mouth swabs and used the machine (Spartan RX CYP2C19), a result being obtained within the hour. This is true bedside technology.
Carriers of one or two 2 alleles of CYP2C19 in the rapid-testing group were given prasugrel; non-carriers and those randomized to standard management were given clopidogrel. The endpoint was platelet reactivity at seven days, and the findings supported the use of this genetic approach and avoidance of clopidogrel where indicated. If every genetic factor influencing the function of anti-platelet drugs were to be included, the testing would be more complex. Nonetheless, the trial size this time was only 187 and the focus was on just 23 in each group who were carriers. Even so, it may be a conventionally large trial, still recruiting but scheduled to finish in October, 2014, that will provide the evidence needed for a major shift in policy. Pharmacogenomics of Anti-platelet Intervention-2 (NCT 01452152) will have some 7200 patients. The idea is similar to that in the Canadian trial but this time the end-point is the frequency of clinical events. So, in respect of personalized antiplatelet treatment (as is discussed by Amber L. Beitelshees, Lancet, 379: 1680-2, 2012), clinicians are in a dilemma: what should they do today, when the results of trials seeking firmer evidence are still in progress?
Mr. David W. Sharp, M.A. (Cambridge), formerly deputy editor of The Lancet, is a freelance writer in Minchinhampton, U.K.
The data and citation records included in this report are from Thomson Reuters Web of ScienceTM. Web of ScienceTM is a registered trademark of Thomson Reuters. All rights reserved.