(c) 2006 
Dow Jones & Company, Inc. Reproduced with permission of copyright owner. Further reproduction or distribution is prohibited without permission.An experimental blood test has shown a glimmer of promise of one day addressing a major health-care challenge: detecting lung cancer at an early stage.
The test, developed by researchers at the University of Kentucky, is designed to identify several proteins in the blood associated with lung cancer, one of the deadliest forms of cancer. A recent small study suggested the test could detect lung cancer several years before it would be picked up by an imaging procedure known as a CT scan.
There are lots of caveats. The blood screen, which has been licensed to a small Maryland company, needs to be studied in a larger population and cleared by regulators before it becomes available. Even if it can detect lung cancer early, it hasn't been proven to save lives. But researchers, although cautious, say it has potential to help improve the dismally low survival rate for lung-cancer patients.
"We could actually have a way to detect lung cancer early," said Jonathan Cohen, chief executive of 20/20 GeneSystems Inc., Rockville, Md., which licensed the technology. "If that's true, it's a huge medical advance. . . . Having an early detection would in all likelihood substantially save lives."
There is no current standard for screening for lung cancer, as there is for breast and prostate cancers. Most people don't know they have lung cancer until symptoms emerge, such as persistent cough or recurring chest infections. Imaging and other techniques are then used to diagnose it. In many cases, the cancer already is at an advanced stage and may have spread to other organs.
Only 15% of people diagnosed with lung cancer live for another five years, one of the lowest survival rates among all forms of cancer, according to the American Cancer Society. More than 160,000 Americans are expected to die from the disease this year.
If lung cancer is caught early enough, the five-year survival rate improves to 50%, which suggests that a better tool for early detection could help. As it is now, only 16% of lung cancers are diagnosed at such early, localized stages. Treatment options for people with early stage lung cancer include surgery, chemotherapy and radiotherapy.
A newer imaging technique known as a spiral CT scan has shown promise in detecting lung cancer earlier than previous methods. But experts say there isn't yet definitive evidence that the use of spiral CT scans reduces death rates. Toward that end, the National Cancer Institute is sponsoring a long-term study of 50,000 current or former smokers to see whether widespread use of spiral CT scans can cut the death rate from lung cancer. The trial began in 2002 and is slated to last until 2009.
Another screening approach is to find so-called biomarkers, or substances in the body that suggest a cancer is present. This was the approach taken about five years ago by researchers at the University of Kentucky. Instead of focusing on a single marker -- as the blood screen for prostate cancer does -- the researchers decided to look for several markers associated with nonsmall-cell lung cancer. Nonsmall- cell lung cancer is a form of the disease that accounts for most lung- cancer deaths.
One of the Kentucky researchers, Li Zhong, said the presence of lung cancer induces the body's immune system to send out proteins known as antibodies that kill cancer cells. In cancers that progress, the cells aren't killed fast enough.
Mr. Zhong and his colleagues have identified five such antibodies, which were found in higher concentrations in blood samples from lung- cancer patients than in samples from people who don't have lung cancer. They developed a screen designed to spot the five antibodies.
In a recent study, the researchers tested the screen on 102 blood samples obtained from people in a separate lung-screening study at the Mayo Clinic in Rochester, Minn. Some of the blood samples had been taken from people one to five years before they were diagnosed with lung cancer. Some were from people not diagnosed with lung cancer, to serve as a control. Participants in the study were current or former smokers ages 50 to 75.
The blood screen was able to predict 32 of 40 cases of cancer in samples drawn one to five years before the cancer was diagnosed using CT scans. It correctly identified 49 of 56 control samples as being noncancerous. The screen also correctly identified all of six additional blood samples taken from patients at the time they were diagnosed with lung cancer using a CT scan.
Those results indicate the test was 82.6% accurate in the lung- cancer samples, and 87.5% accurate among noncancerous samples. Put another way, the screen produced false negatives in about 17.4% of the samples that were cancerous, and false positives in about 12.5% of healthy samples. The blood screen also was used in a separate set of blood samples, which suggested accuracy rates above 90%. The results of the study, which was funded by the National Institutes of Health and other organizations, were published last month in the Journal of Thoracic Oncology.
Although the blood screen produces some false positives and negatives, it appears to have a lower rate of false positives than that of CT scans, which also can pick up noncancerous lesions. CT-scan detection of benign nodules can lead to unnecessary treatment, including surgery, which carries risks of collapsed lung or death.
"The [accuracy] rates are higher than what we have in other tests, that's the advantage," said Michael Unger, director of pulmonary- cancer detection and prevention at Fox Chase Cancer Center in Philadelphia. "The study has tremendous merits as research," he said, adding that it was "very promising."
To improve accuracy, the researchers may try to add biomarkers to the screening test, Mr. Zhong said. And to further validate the test's effectiveness, it should be studied in a larger trial, with as many as 1,000 patients, said Mr. Cohen of 20/20 GeneSystems.
The blood test wouldn't replace imaging techniques, but rather complement them. People who tested positive with the blood screen would still need to get an imaging procedure such as a CT scan to identify the size and location of any tumors. But the screen could help cut down on unnecessary use of CT scans by identifying people with no signs of lung cancer.
"It could be useful in conjunction with a more comprehensive screening approach," said Norman Edelman, a physician and chief medical officer at the American Lung Association.
It's possible that a widespread screening program would recommend that people with certain risk factors -- such as being age 55 or older and a current or former smoker -- undergo routine screens such as the blood test.
Mr. Cohen of 20/20 GeneSystems estimates the cost of the blood test would be less than $200. CT scans can cost between $300 and $1,000 and usually aren't covered by insurance, according to the American Cancer Society.
Mr. Cohen said 20/20 GeneSystems is in the process of raising about $4 million from individual investors to continue the development of the test. It also is exploring a partnership with larger companies. It expects to conduct further studies over the next 12 to 18 months.
My Research
