Since completion of the Human Genome Project in 2003, medical researchers have endeavored to translate genetic data and build an index of clinically relevant information. Now, they are turning their attention to the thorny problem of supplying clinicians with genetic and genomic information when they need it most. Here are some examples.
The Food and Drug Administration is modifying drug labels to include clinically relevant genetic information. For example, officials announced in summer 2007 that they would change the labeling for warfarin, a blood thinner for which carefully calibrated dosing is critical. Too much can cause hemorrhaging; too little can result in blood clots and stroke. New labels indicate that atypical metabolization of warfarin by patients with variations in the CYP2C9 and/or VKORC1 genes influences the drug’s sustained concentration in their blood.
The Veterans Affairs Department has undertaken a large-scale project to collect patients’ genetic information and link the data with information contained in its electronic medical record system.
VA and the Defense Department are working with the National Institutes of Health to integrate patients’ family histories into My HealtheVet, VA’s personal health record system.
In August, NIH released a new policy for the genome-wide association studies it will conduct and support, and the National Library of Medicine has created a database to distribute the information such studies generate.
In April, the American Health Information Community approved the creation of a framework for advancing clinical decision support tools. The project includes the formation of a multistakeholder group co-sponsored by the Agency for Healthcare Research and Quality, HHS’ Personalized Health Care Initiative and the Office of the National Coordinator for Health Information Technology.
— John Pulley
Securing genomic information
In May, President Bush signed the Genetic Information Nondiscrimination Act (GINA), which protects people from discrimination based on their genetic information.
The bill addresses concerns about the potential for the misuse of molecular medical data. The Senate passed it unanimously, and the House approved it by a vote of 414-1.
However, privacy advocates say the law doesn’t go far enough. Among its exclusions, GINA doesn’t offer protection to buyers of life insurance.
“There is great promise in genetic research and certain clinical decision-making tools based on that information,” said Deven McGraw, director of the Center for Democracy and Technology’s Health Privacy Project. “But we need to think about privacy and security protections, and we haven’t done that.”
As viewers of “CSI: Crime Scene Investigation” know, genetic information is a unique identifier, which makes securing such data inherently difficult. The stakes of security breaches are higher, too. Because blood relatives share numerous genes, unauthorized disclosure of a single patient’s data could compromise the privacy of family members, too.
Moreover, the tendency to think of translational medicine exclusively in terms of its technical challenges misses the magnitude and complexity of the undertaking. Using genetic and genomic data to improve clinical decision-making “is a really difficult problem — a technical, policy, societal and business problem,” said Dr. Vance Vanier, chief medical officer at Navigenics, which provides genotyping services to consumers.
— John Pulley
Editor's Note: This story has been updated Aug. 12 at 5:30 p.m. EST. Please go to "Corrections and Clarifications" for details.
Meet Joe’s genes. Residing within each of his body’s 100 trillion cells are 23 pairs of chromosomes containing as many as 25,000 genes made up of more than 3 billion base pairs of nucleotides, the structural units of DNA and RNA. If Joe could transfer that data to a CD, the file would occupy as much space as his copy of Led Zeppelin IV.
Joe’s genes are remarkably similar to those of Jacques and Jane, José and Josephine. From China to Chicago, Taiwan to Timbuktu, the genomes of individual people are 99.9 percent identical. Genetically speaking, we are all average Joes. Small genetic variations make all the difference. A few million nonconforming nucleotide base pairs determine or influence height and hair color, the ability to hit a hanging curveball and Joe’s predilection for classic rock.
Genetic variation also affects health. With the exception of trauma-induced afflictions, nearly all diseases have a genetic component. Acting alone, with other genes and in combination with environmental factors, genomes confer resistance to some illnesses and vulnerability to others. They are a critical factor in determining a patient’s prognosis and response to treatment.
A consensus is emerging on the front lines of molecular medical research that unlocking the secrets embedded deep within Joe’s genes will revolutionize the practice of medicine. The concept, known as translational medicine, seeks to link research more directly to patient care. It will require sophisticated information technology systems that don’t yet exist. And it will take years to develop networks that can securely store, analyze, retrieve, format, and transfer genetic and genomic data in a way that benefits physicians.
The challenges are steep but not insurmountable, said Dr. W. Gregory Feero, chief of the Genomic Healthcare Branch at the National Human Genome Research Institute. “It’s just a matter of time,” he said. “Ultimately, we will be able to sort these things out.”
In the federal government, key participants in the effort to advance the clinical application of molecular medicine include various agencies within the Health and Human Services, Veterans Affairs and Defense departments. HHS Secretary Mike Leavitt is an enthusiastic supporter of personalized, technology-enabled medicine.
However, efforts by public and private entities to develop next-generation systems have so far been largely uncoordinated. The old bugaboos of health care — information silos, reliance on paper records, resistance to new technologies — have slowed progress. Electronic prescribing — arguably the most advanced health IT application — still accounts for only a small percentage of prescriptions. And family medical histories are collected and stored in unstructured ways that often limit their clinical value.
In June, the American Health Information Community, a public/private group that advises Leavitt on health IT, urged the government to promote the use of genetic information — particularly pharmacogenomic tests — in clinical practice and devise standards for using that data in electronic health records. AHIC recommended development of “standard terminology, standard metrics, and structured information for outcomes analysis and research, [which] are needed to allow data exchange, interoperability and integration of [genetic and genomic data] into clinical decision-making.”
The promise Since the dawn of medicine, doctors have relied largely on experience and an organic database of self-contained medical knowledge to diagnose, test and reat patients. Unfortunately, the first symptom of some afflictions, such as cardiovascular disease, is often sudden death.
That prevailing clinical paradigm — a physician-centric witch’s brew of art and inexact science — is akin to one in which a mechanic trie s to fix the world’s most complex machine without the benefit of an owner’s manual. Indeed, physicians have been hamstrung by a lack of “access to the underlying biological processes, unique to each of us, that start with the ‘instructions’ in our genes,” Leavitt wrote in the foreword to last fall’s landmark HHS report “Personalized Health Care: Opportunities, Pathways, Resources.”
Those limitations began to fade in 2003, when the Human Genome Project published what is essentially a manual for the human body. However, its knowledge wasn’t immediately accessible. The genome is like a multivolume tome written in an unknown language.
Nevertheless, cracking the code opened the door to preventive genomic health care, “a field of medicine only dreamed about for years,” said Dr. Vance Vanier, chief medical officer at Navigenics, which is the latest company to provide direct-to-consumer genotyping.
The science Even before the human genome had been fully mapped, scientists knew that an aberrant gene called a single nucleotide polymorphism causes diseases such as cystic fibrosis and multiple sclerosis. If you carry a genetic marker for Huntington’s disease, for example, you have about a 100 percent chance of developing that neurological disorder when you’re in your 50s.
Genes combined with environmental factors, such as exposure to carcinogens or a steady diet of deep-fried Twinkies, cause other diseases. Breaking the human genetic code has allowed researchers to conduct genome-wide association (GWA) studies that investigate the links between gene clusters and disease onset, progression and responsiveness to treatment. Armed with genetic and clinical data, GWA investigators are using powerful comparative tools to discern previously unknown correlations.
Vast amounts of newly mined genetic and genomic data are enabling those investigations. That information lode has resulted from low-cost DNA sequencing, the price of which has plummeted along the same cost curve that informs the economics of the semiconductor industry. At the current pace, medical scientists predict that a full genetic scan that cost millions of dollars a few years ago will be available on a mass commercial scale for about $1,000.
“The more data you have to work with, the more correlations you can find,” said Samuel Aronson, director of IT at the Harvard Medical School-Partners HealthCare Center for Genetics and Genomics (HPCGG).
The challenge Consider the difficulty of solving a 10,000-piece jigsaw puzzle by committee, with each of 1,000 geographically dispersed participants possessing a few pieces. Assume that no one speaks the same language or knows what the finished puzzle should look like. The challenge of creating the IT infrastructure to realize the promise of translational and personalized medicine is even more daunting. It is “perhaps one of the most complex science-based endeavors in our history,” Leavitt said.
The goal is to assemble a dynamic and synergistic information-management system that will allow the easy exchange of data among patients, clinicians and laboratories. Of course, knowledge confers a certain power that isn’t always easily shared.
“It’s not just a problem of data fragmentation but one of data Balkanization,” said Dave Menninger, vice president of global marketing and product management at Inforsense, a data-analysis software company based in London. “Data is actively defended, often in hostile ways.”
That sentiment, Menninger said, originated with John Quackenbush, professor of biostatistics and computational biology at the Dana Farber Cancer Institute and Harvard Medical School.
Setting aside those issues for a moment, consider a scenario involving our friend Joe, who suffe rs migraine headaches. Eager for relief, he undergoes a full genome scan and gives permission for researchers to access his EHR. Armed with de-identified data from thousands of migraine sufferer esearchers uncover genetic correlations implicated in the onset of migraines. Further inquiries determine that sufferers whose genomes have certain genetic markers respond well to a new migraine drug. Joe’s doctor is alerted to the discovery and sends an electronic prescription to Joe’s pharmacy.
Aside from health care benefits, translational and personalized medicine also promises to rein in “health care spending that is not sustainable,” said Raju Kucherlapati, a professor of genetics at Harvard Medical School and scientific director of HPCGG. Seeking to provide a model for translational medicine, the center is building a system of linked databases that will generate rules to support clinical decision-making.
Impediments to implementing a systems-based approach to medicine include the lack of common data standards and language for exchanging medical and genetic data; insufficient infrastructure for sharing results of GWA studies; privacy concerns that could hamper data exchange; and the absence of tools for translating data and knowledge into useful clinical information.
In the absence of automated support, a constantly growing volume of new medical knowledge could easily overwhelm physicians.
“The big question is how you integrate databases of all the patient information you have and all the new genomic information that is coming at us at full force,” said Dr. Ronald Przygodzki, associate director of genomic medicine and acting director of biomedical laboratory research and development at VA. “How do you put it in some [form] that is readable and understandable to…physicians?”
“The tools we are working with are primitive compared to what we will ultimately need,” said Gregory Downing, program director of the Personalized Health Care Initiative at HHS. “We are talking about a different animal in terms of engineering and knowledge base.”
From the battlefield to the home front: Managing medical data
Government Health IT presents Col. Claude Hines Jr., program manager for the Defense Health Information Management System, in this recent InSight eSeminar. Col. Hines discusses the health information technology and tactical challenges faced by the military medical community in Iraq, Afghanistan and other areas of conflict. In doing so, he describes the current information technology solutions for transferring clinical data between battlefield care givers to health care personnel at military treatment facilities worldwide.
