Earlier Detection Means More Hope for Cure
Feb. 12, 2004
Imagine a future where a simple blood test during a physical exam might predict your risk for heart disease, cancer, multiple sclerosis, or even arthritis. This is the brave new world of biomarkers — proteins in the blood whose presence can signal disease — and some doctors say the future of biomarkers is now.
A "biomarker" test is a one that uses a protein in the blood or urine as a marker for what is going on inside the body. The best known biomarker test isn't even for disease: a home pregnancy test works on the same principles.
Doctors hope that with refinement biomarkers will soon be able to perform all sorts of tasks, from diagnosing an illness early to predicting which treatment will be most effective.
"[Biomarkers] will become increasingly important in determining therapy, will revolutionize the way we categorize disease and choose therapies," believes Dr. Bruce Chabner, clinical director of the Massachusetts General Hospital Cancer Center in Boston.
But critics of the biomarker boom say most tests are not as specific or accurate as a home pregnancy kit. "If you have a marker that goes up for many different diseases, it doesn't add much clinical value," argues Dr. Jerome Hoffman, professor of Emergency Medicine at University of California-Los Angeles.
The seemingly unlimited potential of biomarkers has caused an explosion of new studies on the subject and raised questions about how useful some of the biomarkers really are.
Two new studies this week in The New England Journal of Medicine focus on B-type natriuretic protein, or BNP as a marker of heart disease. The first study found measuring blood levels of BNP in a person suffering from labored breathing can help determine whether the patient has heart failure or not.
The second study examined BNP's effectiveness as a biomarker in healthy people. The researchers found people with elevated BNP levels were 77 percent more likely to suffer heart failure and 27 percent more likely to die over a five-year period. This suggests BNP may be able to predict as well as diagnose heart disease.
A third study in the most recent Journal of the American Medical Association found a link between a different biomarker, C-reactive protein or CRP, and age-related vision loss.
Finally, a fourth study in the journal Circulation found biomarkers for heart disease differ somewhat in men and women.
Researchers in that study found testing for two standard biomarkers of heart disease (troponins and CK-MB) missed a large percentage of high-risk women. However, adding two new biomarker tests — for high sensitivity CRP (hs-CRP) and BNP — identified 69 percent of the high-risk population in comparison to the 33 percent to 50 percent identified through standard tests alone.
Louise Cooper of Quincy, Mass., learned how important the new biomarker tests are when she went to the emergency room with increasing chest pain. The standard cardiac biomarker tests came back clean, but her physician, Dr. Christopher Cannon of Brigham and Women's Hospital, an author on the Circulation study, knew to look for other markers.
Cooper's high BNP levels signaled acute heart problems. She quickly had a cardiac stent inserted and is now doing fine. "I was back home the next morning," Cooper recalls, adding the biomarker tests are "amazing."
Tests Not Making the Grade?
For a biomarker to be effective, the test must meet two criteria. It must be sensitive, meaning it always detects the illness when the illness is there. It must also be specific, meaning the test registers "positive" for a single illness and no other conditions.
"The reality is that most biomarker tests are neither," maintains Hoffman.
For example, increased CRP levels have been linked to heart disease, diabetes, age-related vision loss, and dementia. What, then, does an elevated CRP level in your blood tell you about which disease you might have?
"Markers are only helpful when they guide therapy that improves patient outcomes," explains Dr. Mark Ebell, a Michigan family doctor and deputy editor of American Family Physician. "Otherwise they confuse and frighten more than they help."
Adds Dr. Daniel F. McCarter, associate professor of clinical family medicine at the University of Virginia, "The bottom line is that no test can tell you what disease a patient does or does not have … When you use even a very specific test in a population with low incidence of disease, if the test is positive, it is more likely to be a false positive (showing the patient has disease when in fact no disease is present), than a true positive."
High Stakes Testing
The problem of false positives becomes apparent, for instance, with the development of a new biomarker test, called OvaCheck, being developed for ovarian cancer.
Current methods of early detection for ovarian cancer are poor, but early detection is key to survival: detection of ovarian cancer early has a five-year survival rate of 90 percent, whereas late detection leads to a five-year survival rate of only about 35 percent.
OvaCheck, if it can improve early detection, has the potential to revolutionize ovarian cancer treatment. But if the test is not specific enough, women who test falsely positive could end up with their ovaries removed unnecessarily.
Some organizations are moving proactively to evaluate new biomarkers for clinical use. Dr. Sudhir Srivastava of National Cancer Institute heads a group dedicated to putting new biomarkers to the clinical test. The Early Detection Research Network (EDRN) was created five years ago in response to the onslaught of new biomarkers appearing in cancer research.
"There was no mechanism to take the research from basic science to clinical use," Srivastava explains.
The EDRN is a consortium of 52 institutions, each of which plays a different role in the biomaker research. Some labs work on identifying new biomarkers for cancer. Other labs make sure the results are reproducible. Many biomarkers fail at this stage.
If the biomarkers pass the first two stages, they move on to clinical trials to ensure that they actually add value to tests already in place. Of 10 biomarkers evaluated so far, only four have moved on to clinical trials.
Other fields are also moving to evaluate biomakers. There is an NIH-sponsored
workshop on biomarkers for multiple sclerosis scheduled for April. In rheumatology,
the NIH has funded a multicenter study that will recruit 5000 subjects
and follow them for four years to identify biomarkers that predict both
incident knee osteoarthritis and biomarkers that predict progression of
Copyright © 2004, ABC News