NEJM -- Ten Years On -- The Human Genome and Medicine

Ten Years On — The Human Genome and Medicine

Harold Varmus, M.D.

On a June day nearly 10 years ago, the leaders of the United States and the United Kingdom, accompanied by the leaders of the public and private teams deciphering the human genome, announced that a draft sequence had been completed. That occasion was rich with promises of new and more powerful ways to understand, diagnose, prevent, and treat disease. Two years later, the Journal published a series of articles that reviewed the status of medical genetics and the prospects for a new era of "genomic medicine." The tone was cautionary, if hopeful. An editorial (full disclosure: written by me) that accompanied the launch of the series reminded readers that "the full potential of a DNA-based transformation of medicine will be realized only gradually, over the course of decades."1

Now, after the first decade of a postgenome world, only a handful of major changes — some gene-specific treatments for a few cancers, some novel therapies for a few mendelian traits, and some strong genetic markers for assessing drug responsiveness, risk of disease, or risk of disease progression — have entered routine medical practice. And most of these can be traced to discoveries that preceded the unveiling of the human genome. As Francis Collins, formerly the leader of the publicly funded sequencing efforts, recently commented: "the consequences for clinical medicine . . . have thus far been modest . . . the Human Genome Project has not yet directly affected the health care of most individuals."2

In this issue, the Journal begins another series of articles on genomic medicine.3 Is it appropriate for the Journal to be taking stock so soon? It is, and for the following reasons.

First, readers will want to know the state of play — about the dramatic expansion of the knowledge of human genes and proteins and about the extent to which such knowledge is affecting the prevention and treatment of diseases — especially for the more common disorders in which inherited risk has long been implicated but rarely defined. As several articles in this series will describe, detailed maps of genetic markers of human variation, mostly single-nucleotide polymorphisms (SNPs), have facilitated many remarkable genomewide efforts to associate known SNPs with disease predisposition. But this approach has usually failed to reveal strongly influential haplotypes, and in general, other implicated susceptibility haplotypes collectively account for only a small fraction of the apparent heritable risk. Clearly, more than one decade of genomics will be required to understand the inborn risks of most common disorders, such as diabetes and hypertension.

Second, readers will enjoy learning from these articles how rapidly the engines of genomics and other "-omics" are turning. These engines are among the most important tools for discovery in biology today, and their horsepower has increased dramatically. Most notably, new methods have increased rates of DNA sequencing by at least five orders of magnitude and lowered costs dramatically. More efficient and more accurate platforms now also work with ever smaller samples to detect SNPs, measure gene copy numbers, find novel mutations (such as gene fusions), monitor gene expression (RNAs and proteins), characterize chemical modifications of DNA ("epigenomics"), and catalogue metabolic products ("metabolomics").

Still, genomics and related disciplines are more closely aligned with modern science than with modern medicine. They produce knowledge that is broad in its scope, but only a few selected items of that new information are now widely used as guides to risk, diagnosis, or therapy. Physicians are still a long way from submitting their patients' full genomes for sequencing, not because the price is high, but because the data are difficult to interpret.

Third, some articles will serve as reminders that when genomics and related sciences do yield findings of practical merit, adoption of new practices is often exasperatingly slow. Watching from my own corner of medical science, I have been troubled by the slow pace at which testing for a couple of relatively common mutations in our most lethal cancer, lung cancer, has entered clinical care, even when these mutations strongly predict responses to treatment with existing drugs and should more often guide practice. The forthcoming reviews that highlight successful applications of gene-based medicine might hasten adoption of the beneficial changes in medicine that will eventually, if gradually, come from gene-based sciences.

Through this series, readers will also become aware of ways that changes in the worlds of commerce, law, regulation, ethics, health insurance, and information technology are intersecting with the expanded role of genetics in medicine. For instance, just a couple of months ago, the legality of gene patents was thrown into doubt by a ruling from a federal district court4; if upheld, this decision would change — and, in my view, improve — the landscape for development of many tests involving genes that have been patent-protected. Passage of the Genetic Information Nondiscrimination Act in 2008 and health reform legislation this year has reduced concerns that genetic testing could have consequences for employment and insurance eligibility, enhancing acceptance by patients. In addition, the Obama administration's support of health information technology and comparative effectiveness research will doubtless influence the way genetic information is incorporated into practice, transmitted in medical records, and used to inform clinical investigations.

The direct-to-consumer marketing of genotypes, mainly identification of SNPs as putative markers of disease risk, has been among the most visible manifestations of genomics. Yet this practice is not regulated, lacks external standards for accuracy, has not demonstrated economic viability or clinical benefit, and has the potential to mislead customers.

The Genomic Medicine series will also reveal a change in the way the world refers to medicine informed by genes. The term "personalized medicine" has become nearly ubiquitous as a means of conveying how molecular tests can subdivide diagnostic categories and refine therapeutic choices. This phrase may also prove to be strategically successful — by preemptively warding off claims that an overreliance on genotypes in medical practice is deterministic and thus "impersonal," or that genetic approaches undermine more traditional approaches to "personalized" care that are based on knowledge of a patient's behavior, diet, social circumstances, and environment. Of course, both genetic and nongenetic information is important; the more we know about a patient — genes and physiology, character and context — the better we will be as physicians. By measuring the distance to a fuller integration of genomic knowledge into patient care, this new series of articles may encourage a more nuanced calibration of what it means to "personalize" medicine.

open here to see the full-text:
NEJM -- Ten Years On -- The Human Genome and Medicine