Cardiac surgery patients get precisely the right dose of blood thinners to avoid clotting. Glaucoma patients use the least costly, most effective eye drops that lower pressure in their eyes. Diabetics get the therapy they need, based on their genetics, to avoid amputations. Personalized medicine projects already under way are working toward achieving these and many other quality improvement and cost-savings goals.

Pairing genetic information with clinical information in electronic health records is paving the way for a revolution in health care, pioneers in the field say. Ultimately, clinicians will be able to much more precisely pinpoint disease treatments that will yield the best results for a patient. And business intelligence and data mining applications are playing key roles in the early research efforts.

Here's how John Quackenbush, a Harvard University professor of computational biology and bioinformatics, describes what's at the end of the rainbow: "The goal is to try to use existing clinical information and new types of assays [tests] to better direct therapy and understand which patients are likely to respond to different therapeutic protocols. Then we can get patients involved in a protocol as early as possible with the greatest likelihood of success."

Quackenbush, who heads a personalized medicine project at Dana-Farber Cancer Institute in Boston, predicts the price of determining a genome, the entirety of an individual's hereditary information, will decline from the current $10,000 to about $1,000 within five years. "Genome sequencing will then become a regular part of clinical care," he says.

"In 10 years, we'll have a core of genetic variants that are a routine part of most people's electronic records," predicts Dan Roden, M.D., assistant vice chancellor for personalized medicine at Vanderbilt University Medical Center in Nashville. "Fast forward 30 years from now. We'll have a system where everybody's genome is part of their record." The genetic information, paired with electronic health records, will enable clinicians to, for example, target therapy for a diabetic to minimize the risk of an amputation, he says.

Roden also offers the example of potentially altering the current practice of prescribing Plavix or a similar blood clot prevention drug to all patients with stents to help avoid a coronary event, such as a heart attack. Researchers have determined that only 20% of stent patients who get a placebo instead of Plavix develop a coronary event within a year. "So by giving Plavix to all, we're wasting money on 80% of patients," he notes. "If we could use personalized medicine to identify the 20%, we could save a lot."

To achieve such results, however, will require researchers at numerous medical centers to work together, sharing genetic information and electronic health records, he contends. Vanderbilt and many other medical centers are doing early research on their own and taking small steps toward collaborative efforts for broader studies.

Today, most pioneering personalized medicine research involves gathering DNA samples and pairing them with electronic health records for the patients involved, but stripping all the information of personal identifiers. The data is then being used to do big-picture research on treatments for types of diseases, rather than creating patient-specific treatment plans.

But researchers are hopeful that today's aggregate research will lead to tomorrow's truly personalized medicine, where an individual's genetic information enables clinicians to pinpoint effective treatments, case by case.

"Personalized medicine is more than a bunch of genes in your record," Roden contends. "It includes more participation by the consumer in their own health care," he adds. And it requires that clinicians investigate the patient's social circumstances to determine whether they can actually afford the treatment recommended or need help addressing the cost.

Getting the Ball Rolling

To get rolling on personalized medicine research, Dana-Farber Cancer Institute leveraged a $1 million grant from Oracle Corp., Redwood Shores, Calif., to create a data mart to store de-identified information from multiple clinical databases. It's using Web-based business intelligence software from Inforsense, a unit of the British firm IDBS, to enable researchers to make their own queries without the aid of a programmer, Quackenbush explains.

The initial project involves collecting genetic information from patients with multiple myeloma, a type of cancer, pairing it with clinical data and attempting to pinpoint the therapies that yield the best results for specific types of patients. The cancer center is reaching out to other provider organizations across the globe, initially in France, to build a broad database to support the project.

Early this year, Dana-Farber plans to launch an ambitious effort to collect genetic information on virtually all of its cancer patients as a routine part of their assessments, Quackenbush says. "This will give physicians additional information that they can use to help direct therapy," he says. For example, if one gene carries a mutation, "we already know certain therapies simply won't work for you. We'll direct patients to the therapies that, based on our clinical trials, would be most useful."

Personalized medicine likely will make its biggest initial inroads in the area of cancer, Quackenbush predicts, because patients are motivated to support research. "Patients are very willing to sign consent forms enabling us to use tissues for research."

Quackenbush isn't sure whether, over the long haul, genetic information will be stored in a central database or within an individual patient's electronic record. "The ability of EHRs to cope with new types of information is lagging far behind our development of technologies that generate the data," he says. "So it's a research project for us."

The Role of EHRs

Vanderbilt University Medical Center is one of several organizations participating in a national project, called the eMERGE Network, that's attempting to determine the role EHRs can play in personalized medicine for patients with certain chronic conditions. For example, Vanderbilt is using EHRs, paired with genetic data, to research how genetic variations affect the electrical activity of the heart. Meanwhile, Marshfield (Wis.) Clinic is studying how genetics affects development of cataracts as well as the level of HDL cholesterol.

"The big challenge here is investigating how to use the EHR to the best advantage for genome science," says Vanderbilt's Roden. "We have the opportunity to work with others with similar resources and compare notes to get things done."

At the local level, Vanderbilt has built a BioVU database of more than 72,000 DNA samples that it expects could grow to as many as 200,000. Researchers extract DNA from discarded blood samples as authorized by volunteer patients. To ensure security and privacy, Vanderbilt developed a sophisticated application of cryptography to create a 128-character identifier for the DNA sample and the associated medical record that cannot be linked back to the individual's identity, Roden explains.

"My job now is to get the resources moving to allow Vanderbilt investigators to ask the questions they want to ask" by querying the data, Roden adds. The data mining software involved was developed at Vanderbilt. Research efforts leveraging the data were slated to begin early this year.

Vanderbilt earned a grant to pay for a robot that handles bar coded DNA samples stored in tubes. That way, when a researcher asks for 1,000 DNA samples of patients with a specific condition in a certain age range, the robot can automatically retrieve those samples.

Role of Robots

Like Vanderbilt, Marshfield Clinic has deployed a robot to manage retrieval of DNA samples from almost 20,000 volunteers, says Cathy McCarty, senior research director at the 750-physician practice's Center for Human Genetics. A pioneer in the field, the clinic has been involved in genetic research for decades.

Now that it has collected a diverse sampling of DNA, the clinic is focusing on collecting DNA from patients with specific conditions that are the target of treatment research. That way, it can compare their genetics with that of the broader population, McCarty explains.

The Wisconsin clinic has 20 research projects under way leveraging its rich genetic resources. For example, it's recruiting volunteers for a project attempting to develop a dosing calculator for Warfarin, a blood thinner. The calculator would take into account three genetic markers, age, gender and other factors.

Another project is investigating various eye drops used to lower the pressure in the eyes of glaucoma patients. "We've identified a genetic marker that predicts who responds best to the lower-cost generic drug vs. the more costly drug, and we're testing this," McCarty says.

As a participant in the Wisconsin Genomics Initiative, Marshfield soon will begin work in collaboration with several academic medical centers in the state to investigate genetic predictors of heart attacks.

Launching into Orbit

In the Milwaukee area, Aurora Health Care, an integrated delivery system, has launched ORBIT, an ambitious personalized medicine project that has collected 50,000 DNA samples since last March. Aurora expected to have 100,000 specimens by early this year, ultimately gathering about 500,000, says Randall Lambrecht, vice president of research.

Because Aurora is involved in about 600 clinical research trials, it's hopeful that the wealth of genetic information, tied to electronic records, will help fuel research on new drugs and therapies, Lambrecht says.

Until now, clinical trials have taken a "shotgun approach," using broad populations to test drugs or serve as control groups taking placebos. "In the future, we'll be able to screen out those whose genetics indicate they'd likely react adversely to the drug and make sure they're not included in the trial, or pinpoint the 5% who are good candidates for a drug."

At Aurora, the DNA samples are retrieved using robotics, much like at Vanderbilt and Dana-Farber. To test-drive the business intelligence technologies involved, Aurora has launched some preliminary, small-scale studies, including one on genetic biomarkers for cardiovascular disease and cancer.

"Longer-term, the idea is to go from our de-identified biorepository to one that's patient-specific," Lambrecht says. "Consenting patients will allow us to track the genetic reasons for their predisposition for a disease or investigate whether a new therapy is right for them."

So far, Aurora is funding its efforts on its own, without any grant support. "We saw the need for moving fast in the arena of trying to use genomic information in patient care."

A Futurist Gazes ahead

The single most significant change in health care in the next 20 years will be the movement toward personalized medicine fueled by genetic information, says futurist Jeffrey Bauer. "We'll see a shift from one-size-fits-all medicine to recognizing just how different each of us is," says Bauer, the Chicago-based partner in the futures practices at consulting firm ACS Healthcare Solutions. Once doctors have easy access to most patients' genetic information, they'll be able to make far better choices about what drugs, therapies and behavioral modification approaches to select, he adds.

"But we're a decade or more away from having the knowledge to use genetic predispositions to manage diseases," the futurist says.

Thanks to genetic research, however, clinicians already have more detailed information on the genetics of some diseases, like cancer, that are enabling them to pinpoint the specific type of disease a patient has to help guide treatment decisions.

Today, it costs about $10,000 to determine an individual's genome, the entirety of their hereditary information. But already, scientists can analyze specific gene locations associated with particular diseases for as little as $100, Bauer notes. Eventually, the price for genetic profiling will decline dramatically, he predicts. "But price isn't the real issue.

"The issue is how quickly we will get an affordable and accurate test," Bauer says, predicting that should happen within a decade. "The tests are not yet ready for prime-time. We need further research."

Once Americans feel assured that their genetic information will remain private, many will authorize having the data stored in their electronic health record, and some will grant permission for researchers to use it for clinical trials and other purposes, Bauer says. If it lives up to its promise, The Genetic Information Nondiscrimination Act, a federal privacy law which went into effect in December, could provide adequate reassurance to those concerned about privacy, he adds.

So where are we headed? Twenty years from now, primary care physicians will use patients' genetic information to help target and rank susceptibility to various diseases, Bauer predicts.

This article can also be found at

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access