Making Personalized Medicine PayJohn Carey
The dirty little secret about drugs is that they only work in about half of the people who take them. So says an educational nonprofit called the Personalized Medicine Coalition, and many drug executives concede as much. Of the $292 billion spent in the U.S. on prescription drugs in 2008, as much as $145 billion went to medications that didn't help individual patients, said Jerel Davis, project manager at McKinsey, at a recent conference. And billions more are being spent to treat adverse drug reactions and other complications. "When you look at the data, it's shocking," says Dr. Robert S. Epstein, chief medical officer at Medco Health Solutions (MHS), a $51 billion company that manages drug prescriptions for 60 million Americans.
Researchers know how to solve this problem. First, figure out the differences between those patients who respond to a drug and those who don't, then treat only to those who will benefit. But this personalized medicine approach "has been slower to develop than we thought 10 years ago," says Richard K. Schatzberg, CEO of Generation Health, a startup that offers targeted medicine services. Lack of enthusiasm in the drug industry is a big reason; companies would lose billions of dollars if only those who actually benefit were to use such blockbuster drugs as antidepressants, arthritis medicines, and cholesterol pills.
Now, however, the promise of personalized drug treatments appears more realistic, thanks to new players on the sceneand a new business model. The recent entrants are pharmacy benefit managers (PBMs) such as Medco and CVS Caremark (CVS). Medco is testing patients for genetic variations that explain why they respond differently to drugs like warfarin, a widely used blood thinner, and tamoxifen for breast cancer. CVS Caremark has taken a majority stake in Generation Health and expects to launch a similar testing program in May. The move by the PBMs "is transformative," says Edward Abrahams, executive director of the Personalized Medicine Coalition, whose members include scientists, health-care providers, payers, and patients' groups. "We are talking about better care for millions of people and keeping costs down for employers, whose insurance costs are exploding. It could be the tipping point."
PBMs plan to make money by selling personalized medicine services to employers, which are willing to pay them higher fees for improved health outcomes and lower prescription costs. Medco and other PBMs also hope to win market share from their slower-moving competitors. "It is a differentiator for us," says Dr. Jane Barlow, vice-president for Medco's personalized medicine business. Plus, they expect genetic testing will increase the percentage of patients using certain cheaper generic drugs, thus increasing profits. Medco has signed up 200 employers to its program, representing 7 million people. "This has been the fastest adoption of a new program in Medco history," says Epstein. One early—and eager—adopter: IBM (IBM), which expects better health outcomes and cost savings, says Dr. Martin Sepulveda, an IBM vice-president for health matters.
The idea took a long time to bear fruit at Medco, even though it was an obsession for Epstein, an epidemiologist by training. "Back in 2000, Rob Epstein explained to me this would change the face of medicine—and make all the pharmaceutical companies nervous," recalls Schatzberg. "It took longer than he thought." The company's first foray into personalized treatments in 2002 foundered. Epstein wanted to do genetic testing on asthma patients to predict better which ones might end up in the hospital, "but I couldn't see the return on investment," he says.
The business case improved as scientists identified more genes linked to drug responses. For many medicines, enzymes produced in the liver are crucial. Some enzymes change drugs so that they are excreted from the body. Others convert drugs that have no effect when first administered into a medically active form. Because of variations in genes, these enzymes may work quickly or slowly or not at all. One example is tamoxifen, used to prevent breast cancer recurrence. In 20% of people, the enzyme that usually activates this drug is partially or completely ineffective, and the drug provides little or no benefit.
A turning point for Medco came in 2005, when a Food & Drug Administration advisory committee recommended that genetic information be considered in making treatment decisions with warfarin. The blood thinner is widely prescribed to prevent clots, but it's notoriously difficult to get the dose right. "We know that we kill people with warfarin all the time," says Dr. Issam Zineh, associate director for genomics at the FDA's Center for Drug Evaluation Research. Too much warfarin raises chances of bleeding and strokes caused by bleeding; too little allows deadly clots to form. The cost of using the wrong doses is estimated to be in the billions of dollars per year. With a genetic test, doctors can determine if people will need more or less warfarin than the standard 5-milligram dose.
When Epstein looked at Medco's medical records of its million patients on the drug, he discovered something alarming: As many as 25% of them ended up in the hospital within six months of starting on warfarin. "Avoiding one hospitalization could underwrite the cost of the test for 100 patients," Epstein reasoned. Medco worked with the Mayo Clinic to measure the clinical benefits and cost savings from genetic tests for warfarin. The final data won't be released for several months, but Medco found that employers were eager to sign on for the testing service anyway. Now, when Medco sees a prescription coming in for warfarin, it recommends genetic testing to the doctor and patient. In Medco's experience, 67% of doctors and 82% of patients agree to testing.
"THE INCENTIVES ALIGN"
The next drug Medco personalized was tamoxifen. Identifying women who can't metabolize the drug into its active form and putting them on a different drug reduces the cancer's chances of recurrence—and the costs of future treatment. Coming soon is a test for another blood thinner, the blockbuster Plavix. Pinpointing those who benefit will enable Medco to keep more patients on the drug when it goes generic, instead of switching to a more expensive alternative that doesn't require a test. A bonus is that the results from any given genetic test are usually applicable to many drugs. The same variation that determines the response to Plavix, for instance, can help determine how Valium, heartburn drugs like Nexium, and the antidepressant Celexa should be used.
The PBMs' foray into individualized treatments "is where the business rubber meets the road," says Michael Stocum, managing director of consultant Personalized Medicine Partners. "The incentives align. Patients want to get the right drug, and payers are willing to pay if they get a benefit."
Targeting drugs to those who benefit will obviously cut revenues for some drugmakers. But the pharmaceutical industry itself has started to back away from trying to sell the same medicines to everyone, says former Pfizer drug researcher Dr. Bruce H. Littman, now president of consultant Translational Medicine Associates. "The blockbuster mentality is still in place, but drugmakers are coming around," he says.
If they don't, the FDA may not be pleased. In the future, the agency may balk at approving drugs that can't be directed to the right patients—and payers may decline to reimburse. Amgen, for one, strongly backs the use of a test for a gene called KRAS for its $8,400-per-month colon cancer drug, Vectibix. About 40% of people have a variation of KRAS that prevents the drug from working. Drugmakers "see a future business model where they just want all of the smaller market of appropriate patients," says Generation Health's Schatzberg.
Given these trends, the once overhyped idea of personalized medicine "is really starting to get legs," says Dr. Eric Topol, chief academic officer at Scripps Health. "The old way of giving therapeutics will be obsolete."
Business Exchange: Read, save, and add content on BW's new Web 2.0 topic networkA Revolution—or Not
The Language of Life, a new book from Dr. Francis Collins, director of National Institutes of Health, trumpets a revolution in personalized medicine. But Peter Aldhous, San Francisco bureau chief for New Scientist, finds the evidence thin. In a Jan. 7 entry on the New Scientist blog CultureLab, he calls Collins "a cheerleader for a revolution that hasn't arrived quite yet" and notes that many users of consumer gene testing services "have been underwhelmed by the insights gained."
For this and related articles, go to http://bx.businessweek.com/genetic-testing/reference/