MANAGED CARE November 2008. ©MediMedia USA
Expanded blood testing can reveal much about individuals’ health risks, but how should health plans decide on coverage?
Cardiologist George Rodgers, MD, still remembers the fiftyish woman he examined years ago after she reported chest discomfort. “As I was leaning in to listen to her heart, I encountered a peach-sized tumor in her breast,” he recalls. “When I asked about it, she told me she’d first noticed the growth when it was pea-sized. ‘Why didn’t you do anything about it?’ I said, and she replied: ‘Because I was afraid it was cancer.’”
That fearful woman, who died of breast cancer six months later, was a dramatic case of not putting health care information to effective use.
Today Rodgers is president and CMO of Biophysical Corp., a four-year-old Austin, Texas-based company that offers clients an advanced array of information about their health based on biomarkers — signs far subtler than that peach-sized growth that can nevertheless provide warnings of incipient illness.
“Broadly speaking, even blood pressure can be thought of as a biomarker, because it’s an objective, measurable index that can indicate potential disease,” says Rodgers. Biophysical Corp. concentrates on serum biomarkers in the blood. “Blood often shows signs of disease years before a person becomes aware of symptoms,” he explains. “We’ve been able to shed light for many people on what’s coming their way.”
Boosters of blood testing beyond the standard blood workup believe biomarkers may help bring the “personalized medicine” one hears so much about across the specialties these days. The markers have been used increasingly by clinicians to assess individual patients’ risk for diseases. What isn’t yet clear is whether — perhaps by helping to stratify risk and target subgroups for special interventions — they will soon become a practical tool in the hands of health plans.
Biomarkers vary by individuals
Scientists say biomarkers are personal: Each person’s body functions in a unique way, and biomarkers reflect this uniqueness. For some people, certain biomarker levels naturally run high. For others, they run low. For some, biomarker levels fluctuate widely, while for others they are fairly consistent. “When we look at a patient’s ‘personal reference range’ — his or her past recorded biomarker levels — we’re no longer just comparing results to the population, but also looking at what’s normal for that person,” says Rodgers. “That can be powerful personal health information. By knowing what’s coming, patients and their doctors can make specific treatment and lifestyle decisions aimed at slowing or preventing the onset of disease.”
The response from the managed care community? Yes, this knowledge is potentially powerful, but its practical application on a population level has yet to be demonstrated. “I guess the first question would be, ‘How would this information be used in practice, and is there evidence that its use is tied to better outcomes?’” says Eric B. Larson, MD, MPH, director of the Center for Health Studies at the Group Health Cooperative in Seattle. “Most markers,” he notes, “have not yet been shown to be of value in the management of patients.” For example, says Larson, the biomarker APO E4 appears to be found more often in people with Alzheimer’s disease, but no one argues that it can provide a definitive diagnosis or make a difference in decisions about treatment or family counseling.
Of course, some biomarkers could potentially stir patients to new vigilance in lifestyle choices, screenings, perhaps even prophylactic surgery. And Larson is interested enough in a kindred concept to be a co-investigator in the Multiplex Initiative, a study at the Henry Ford Hospital in Detroit that is exploring how more than 2,000 people who are offered genetic testing make use of the option, and its results. But the jury is out. Writing this year in the journal Nature, Larson and his co-authors note “the mixed success of current risk communications based on elevated cholesterol, blood pressure and other biomarkers in motivating behavior change” and worry that a headlong, market-driven rush to genetic testing might engender mistaken priorities and false hopes.
Still, says Larson, “there’s a lot of interest in moving the findings of genomics [genetic mapping] into what’s useful and what’s not, and that’s a role HMOs can play.”
Obesity, diabetes and heart disease
“Before I was tested I thought there was only one reason to lose weight — to look better,” a Biophysical client once told Rodgers. “Now you’ve given me 10 reasons!” Between 1976 and 2004, the prevalence of obesity among adults ages 20–74 increased from 15 percent to 32.9 percent, and there were similar hikes in diabetes and the pre-diabetes condition known as metabolic syndrome, which are harbingers of heart disease. Of course, the most effective treatment starts early in the disease process, before complications have set in. Last year, according to a new American Diabetes Association study, diabetes-related medical and economic costs in the United States hit $174 billion. That’s a 32-percent increase from five years earlier. There’s a lot of money at stake, as well as many lives. But not all obese patients develop diabetes, so, without symptoms, how does a physician determine whom to treat? The answer, says Rodgers, is biomarkers.
“Biomarkers can tell who has metabolic syndrome and who is on the way to diabetes,” he says. Metabolic syndrome is diagnosed when a person has at least three of five factors: high blood pressure, high fasting glucose, high fasting triglyceride, low HDL cholesterol, or large abdominal girth. As a person’s body mass index increases, so does the risk for metabolic syndrome.
Reviewing the results of more than 400 clients, Biophysical has noted that as BMI goes up, so does the percentage of people with risk factors for the metabolic syndrome. But the correlation isn’t complete — one reason that doctors today run blood tests instead of simply using a tape measure on the waist. Some 58 percent of obese people don’t have metabolic syndrome, says Rodgers, and 10 percent of their skinnier contemporaries do. Blood testing helps to pinpoint who is actually at risk.
| As BMI goes up, so do risk factors | ||||
|---|---|---|---|---|
| BMI | Glucose >100 | Triglycerides >150 | HDL <40 Men | HDL <50 Women |
| <25 | 11.5% | 6.2% | 10.9% | 11.3% |
| 25–30 | 28.7% | 20.2% | 14.7% | 25.0% |
| >30 | 41.89% | 33.7% | 29.0% | 31.5% |
Similarly, obese people experience lower levels of adiponectin, testosterone (for men), and HDL, and higher levels of ferritin, fasting blood sugar, cholesterol, LDL, triglycerides, and inflammation. But again, not all obese people experience these effects. These biomarkers, says Rodgers, can help identify who is in trouble.
And consider inflammation, “the match that starts the disease fire.” Three good indications of inflammation are the biomarkers C-reactive protein (CRP) and interleukins 6 and 8, which are strongly associated with risk for cardiovascular disease. Supporters say biomarkers can tell who has metabolic syndrome and who has or is at risk for diabetes, and of these people, who has the cardiovascular and inflammation profile that puts them most at risk. Once identified, each of these parts can be treated. And biomarkers can also register improvements over time, providing powerful psychological reinforcement for appropriate self-care.
| Indications for inflammation | ||||
|---|---|---|---|---|
| BMI | Glucose >100 | CRP>3 | IL-6>12 | IL-8>75 |
| <25 | 11.5% | 14.2% | 5.7% | 7.5% |
| 25–30 | 28.7% | 17.0% | 2.1% | 12.2% |
| >30 | 41.89% | 44.5% | 10.8% | 20.2% |
Could the aggressive use of serum biomarkers by health plans save lives and resources? Rodgers concedes that large studies to prove their efficacy have not yet been done, but he thinks the answer may be yes. To illustrate, he cites the year’s most famous early death from heart disease, that of “Meet the Press” host Tim Russert, who collapsed suddenly in June after plaque ruptured in a coronary artery, causing a clot that cut off circulation to part of his heart. His case reminds us that plaque buildup in arteries doesn’t tell the whole story of the risk of sudden heart attack; a rupture in that plaque is often the fatal precipitating event.
“There are biomarkers associated with an impending rupture, such as myloperoxidase (MPO), CD-40 ligan, and the matrix matella proteinases,” says Rodgers. “If a broader evaluation of Russert’s biomarkers had been done, it’s possible that medical intervention could have brought down the level of inflammation and prevented the rupture. And I’m not talking about necessarily putting in a stent or doing a bypass; I’m talking about medical therapy.” Rodgers points out that the costs of a thorough blood test would be trifling compared with the tragedy of Russert’s death, or the treatment costs for a similar heart attack patient who survived, but in a vegetative state.
Still, Carolyn Langer, MD, medical director of medical management and policy at Harvard Pilgrim Health Care, says the predictive value of biomarkers doesn’t yet have firm science behind it. “Many of these markers promise to assist in stratifying high-risk populations, which may in turn enable clinicians to intervene earlier in the disease process,” she says. “But even for stratification, I don’t feel they are ready for prime time yet.” If their predictive value were now supported by studies, testing for them would be ordered by leading cardiologists, she says. Indeed, her plan consults annually with such specialists to keep abreast of “the emerging standard of care.”
And Langer adds a practical concern: From a logistical standpoint, it would be very difficult for the health plans to acquire all of those test results, she says, because plan members don’t always get tested by contracted national labs.
“I don’t want to sound pessimistic,” Langer concludes. “To the extent that new biomarkers may someday help us to risk-stratify the population to assist doctors in intervening earlier, that would be a very useful tool for us.”
Early cancer detection
If biomarkers could be used as a screening device as mammograms are used now to screen breast cancer, there would be few doubters. Between 1990 and 2000, the breast cancer mortality rate decreased 2.3 percent annually because of mass screening that resulted in earlier detection of more cancers. Early detection is a key to saving lives and cost. Though only 25 breast cancers are diagnosed as a result every 10,000 mammograms, the test is broadly considered worthwhile. It isn’t clear exactly how and when biomarkers will claim a similar role.
Rodgers points out that the National Cancer Institute has established the Early Detection Research Network (EDRN) to focus on biomarkers and other methods to detect cancer. Biomarkers can help, he says, in two ways:
- Multiple markers can be viewed in combination. This process, called “stacking,” can be more revealing than checking one marker in isolation.
- The “velocity of change” in biomarkers can be tracked over time. Such investigation of “stacked” markers can be more revealing than the same group examined at just one or two comparison points.
Today’s researchers are refining these techniques to discover cancer at its earliest stages. And they are making inroads with ovarian cancer, the fourth most common cause of cancer deaths among women, which is generally symptom-free until its very late stages, yet has a high cure rate when diagnosed early. Biomarkers such as CA-125, CA 15-3, CA 19-9, interleukin-6, interleukin-8, monocyte chemotactic protein-1 (MCP-1), epithelial growth factor (EGF), and vascular endothelial growth factor (VEGF) are among those considered promising in screening for early ovarian cancer, says Rodgers.
But today, says Lee Newcomer, MD, executive vice president for oncology services at United HealthCare, cancer treatment gets a more potent assist from markers in malignant tissue itself than from those in the blood. For example, a positive test for human epidermal growth factor receptor-2 (HER-2) now reveals whether a breast-cancer patient will respond to the new drug Herceptin, with dramatic results.
“When you’re talking about screening lots of normal people,” warns Newcomer, “you need a test that’s very, very specific.”
He points out another practical obstacle. “The big problem for health plans today is the way these blood tests are coded and billed for, which isn’t keeping pace with technology,” he says. “We don’t know whether a given test is for ovarian cancer, diabetes, asthma, or heart failure, because the code only describes the process, and under that process there are hundreds of possible tests. There are a few serum-marker tests that probably should be covered today, but it’s very difficult to sort those out.”
Will serum biomarkers be part of a new “personalized” approach to organizing the care of populations by health plans? Science suggests that in some form they probably will, even if today’s plans are only beginning to figure out how. Someday, perhaps, ignoring a hundred tales the blood can tell will seem as unthinkable as neglecting a growth that keeps on growing.