By Donna Dickenson

"We are in a new era of the life sciences, but in no area of research is the promise greater than in personalized medicine."
-Barack Obama, as a Senator introducing the bill that became the Genomics and Personalized Medicine Act 2007


The soaring promises made by personalized genetic medicine advocates are probably loftier than those in any other medical or scientific realm today. Francis Collins, former co-director of the Human Genome Project, wrote: "We are on the leading edge of a true revolution in medicine, one that promises to transform the traditional 'one size fits all' approach into a much more powerful strategy that considers each individual as unique and as having special characteristics that should guide an approach to staying healthy...You have to be ready to embrace this new world."1 Certainly vast sums are pouring into personalized medicine; plans to spend $416 million on a four-year plan were announced in December 2011 by the National Institutes of Health, and private sector interest is also intense.

But does the science bear out the claim that there's a genuine paradigm shift toward personalized genetic medicine? It has been said that ten years after the completion of the Human Genome Project, geneticists are almost back to square one in knowing where to look for the roots of common disease.2 As of March 2012, current genetic tests and molecular diagnostics have only been applied to about two per cent of the US population.3 A Harris poll of 2,760 patients and physicians in January and February 2012 indicated that doctors had only recommended personal genetic tests for four percent of their patients, hardly the stuff of a paradigm shift-at least not yet.

It has been asserted that a baby could have her genome fully sequenced at birth, revealing her susceptibility to particular diseases. She could then enjoy the benefits of made-to-order diagnostic tools and drugs throughout her lifetime. That really is the "Holy Grail" of personalized genetic medicine, but it makes huge and currently unfounded assumptions about how much we are actually able to predict. Most major diseases are caused by the interplay of many genes rather than one, and they arise from both environmental and genetic causes.

The most recent policy update from the American Society of Clinical Oncology accepts that genetic testing for personal cancer susceptibility is now a routine part of clinical care, especially for high-penetrance mutations like the alleles (variants) of the BRCA1 and BRCA2 genes implicated in some breast and ovarian cancers. However, the Society also notes that such cancers are comparatively uncommon. The Society believes that there is little clinical value in testing for the 100 or more relatively common single nucleotide polymorphisms (SNPs) linked to parts of the genome that are associated with cancer in a yet undetermined way, because the risk from each individual SNP variation is generally too small to serve as the basis for clinical decision-making. By contrast, a family history of breast and ovarian cancer could alert a clinician to order a direct and specific test for the BRCA1 and BRCA2 genes implicated in some such tumors. But BRCA 1 and 2 testing may be restricted by monopoly patent protection on those genes, leading to prices of up to $3,500 for the diagnostic tests. Although these patents were challenged in a recent court case,4 they still stand at present.

In pharmacogenetics or pharmacogenomics, clinical genetic typing is used to determine a patient's probable response to drugs such as cancer treatments and to tailor the pharmaceutical regime personally. It might be possible, for example, to identify patients who are genetically programmed to respond more quickly to chemotherapy and to give them lighter dosages, so as to avoid the worst side effects. Pharmacogenetics is not confined to oncology, but there the goal is also to adjust treatment to the sequenced genome of the cancer, which differs from the patient's normal cells. This double approach is crucial because cancer is so heterogeneous, even in patients with the same diagnosis. After sequencing the entire genomes of fifty patients' breast cancers, researchers found that only ten percent of the tumors had more than three mutations in common.5

Outside oncology, there has also been progress in pharmacogenetics. For example, the drug warfarin is an oral anticoagulant commonly used to prevent or manage venous thrombosis. It is sometimes difficult to determine the correct dosage for an individual patient, and thinning the blood excessively can be an unwanted side effect, carrying its own risks. But now warfarin dosage can be tailored to identify particular patients at increased risk of bleeding, by sequencing two genes that account for most of the variation in how people react to the drug. In public health, a major study-the five-year "Human Heredity and Health in Africa" (H3) study, jointly funded by the National Institutes of Health and the Wellcome Trust-aims to apply genome scanning and sequencing techniques to major communicable diseases such as HIV/AIDS, tuberculosis and malaria, as well as to non-communicable conditions such as cancer, stroke, heart disease and diabetes. The hope is that the project will finally bring some of the benefits of advanced genetics research to the world's poorest continent.

These and other developments give reason to be hopeful about pharmacogenetics, certainly more so than about direct-to-consumer retail genetic testing. However, a genome-wide analysis of biopsies done on four kidney cancer patients showed that a single tumor can have many different genetic mutations at various locations. Two-thirds of the genetic faults identified were not repeated in the same tumor, let alone in any other metastasized tumors in the body.6 That is quite discouraging, because if a pharmacogenetic drug targets one mutation in the tumor, it may not work on other mutations.

The former head of the American Society of Clinical Oncology, George Sledge, has gone so far as to declare that the only cancers that have been outwitted so far by pharmacogenetics are the "stupid" ones-the minority of cancers caused by mutations in only one or two genes. "One danger of stupid cancer is that it makes us feel smarter than we are," Sledge concedes ruefully.7 That overconfidence is obvious in many of the more exaggerated paeans to personalized medicine.

Trials in cancer pharmacogenetics additionally have to contend with an inherent paradox of personalization: The more unique or specific the proposed drug is to particular genetic sub-groups of patients, the harder it becomes to find enough patients for statistically significant results. This profound problem makes some commentators skeptical that individualized drug therapy will be possible for most conditions any time in the foreseeable future.

The continuous discoveries of new surprises about the genome call into question the claim that personalized medicine is almost here, or that individualized drug therapy will soon be a reality. In fact, it probably never will be, or at least not by DNA testing alone, because most genotype-phenotype associated studies are hampered by limited size and therefore decrease in statistical power.8

If the scientific evidence alone fails to bear out the bigger claims for personalized medicine, why is there such great interest? We need to look to social and economic factors as well as scientific ones. For a pharmaceutical industry facing the expiry of patent protection on many of its best-selling drugs, new markets have to be found. By breaking an existing medication down into different "size ranges," and by persuading customers that they cannot simply rely on a "one-size-fits-all product," pharmaceutical companies can create new niche markets.

It would be even more advantageous for the pharmaceutical industry if the individual patient could be persuaded to pay for genetic typing out of her own pocket, so that she would then know which of the niche pharmaceuticals is her "size." Now that the $1,000 whole-genome test is approaching reality, retail genetics may well extend its reach from subsets of SNPs to offering whole-genome mapping. Customers could thus have all their personalized genetic information ready for access when needed, so that prescribing on a pharmacogenetic model could become much more commonplace. In that event, diagnostic costs would be transferred from the public health system or insurers to the private individual, while some individuals might find themselves excluded from coverage on the basis of their genetic profile.

Patients' enthusiasm for pharmacogenetics would take quite a dent if they saw it as a rationale for denying them therapy, but in an era of cost-cutting, that could well happen. Cancers driven by a number of different genetic pathways may require different regimes of drug combinations for different patients. With drugs required by smaller-size patient groups, it may not be economical for drug companies to produce every drug required for the regimen of any particular patient. From the drug companies' point of view, big blockbuster drugs have traditionally been the money-spinners. Unless a stratified patient group is large (or wealthy) enough to constitute a niche market, it would not necessarily be in drug companies' interests to tailor medicines too narrowly. This is the largely ignored economic reality of personalized genetic medicine: The more personalized it becomes, the more its range of customers narrows-and therefore the less incentive there is for firms to produce the drugs.

Alternatively, pharmaceutical firms might pursue a strategy of high price increases for personalized cancer drugs. The pricing of a group of oral oncolytic (anti-tumor) drugs, including Gleevec, has gone up by over 76 per cent since 2006.9 The drug Xalkori, which was developed with a small group of patients whose lung cancers had a particular mutation, is being made available at a price of $9,600 per month.10 This high price is driven by the small size of the potential market; the total target population for the drug is expected to be fewer than 10,000 patients.11

Against the trend of genetic personalized medicine, some of the most promising research in cancer prevention actually comes not from the complexities and costs of individually tailored drugs, but from simple, cheap and comparatively safe "one size fits all" drugs, even for genetically caused conditions. In October 2011, a UK team found that a daily 600 mg dose of aspirin resulted in a 63 percent reduction in the number of colorectal cancers in patients with a hereditary disease called Lynch syndrome. This genetic condition increases the risk of colorectal and uterine cancer in about 2 to 7 percent of the population by affecting genes responsible for detecting and repairing DNA damage.12 Every one of the 861 people with this syndrome in the trial got the same dosage of the same simple drug against the same threat. It worked.


Donna Dickenson, MSc, PhD, is a fellow of the Ethox Centre in Oxford, Emeritus Professor of Medical Ethics and Humanities at the University of London, and honorary senior research fellow at the Centre for Ethics in Medicine at the University of Bristol.



1. Collins, Francis S. 2010. The Language of Life: DNA and the Revolution in Personalized Medicine. New York: Harper Collins, pp. xxiv-xxv.

2. Wade, Nicholas. 2010. A decade later, genetic map yields few cures. New York Times, June 12th.

3. United Health Center for Health Reform and Modernization. 2012. Personalized Medicine: Trends and Prospects for the New Science of Genetic Testing and Molecular Diagnostics. Working Paper 7, March, p. 3.

4. Association of Molecular Pathology et al. v United States Patent and Trade Office and Myriad Genetics Inc. 2010. 669 F Supp 2d 365.

5. Wadman, Meredith. 2011. Fifty genomes sequences reveal breast cancer's complexity. Nature News, April 2, doi:10.1038/news.2011.203.

6. Gerlinger, Marco, Rowan, Andrew J., Horswell, Stuart, et al. 2012. Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. New England Journal of Medicne 366(10) 883-892.

7. Quoted in Harper, Matthew. 2011. Cancer's new era of promise and chaos. Forbes, June 5,, p. 3

8. Nebert, Daniel W., Ge, Zhang, and Vessell, Elliott S. 2008. From human genetics and genomics to pharmacogenetics and pharmacogenomics: past lessons, future directions. Drug Metabolism Review 40(2): 187-224.

9. Chiang, Alex, and Milton, Ryan P. 2011. Personalized medicine in oncology: next generation. Nature Reviews Drug Discovery 10: 895-896, at 895.

10. Ibid.

11. Kwak, E.L., et al. 2010. Anaplastic lyphoma kinase inhibition in non-small-cell lung cancer. New England Journal of Medicine 363: 1695-1703.

12. Geddes, Linda. 2011. Daily aspirin cuts risk of colorectal cancer. New Scientist, October 28.

Search: GeneWatch
The Council for Responsible Genetics’ Genetic Privacy Manual: Understanding the Threats- Understanding Your Rights will be a comprehensive, electronic source of information for the consumer on these issues.
View Project
The Gene Myths series features incisive, succinct articles by leading scientists disputing the exaggerations and misrepresentations of the power of genes.
View Project