Dec. 13, 2002 (San Antonio, Texas) -- In what's being called one of the biggest advances in cancer research in years, scientists have developed a blood test that can detect cancer with a greater than 90% accuracy. This artificial intelligence -- already tested for cancers of the breast, ovary, and lung -- could one day be used to detect many types of cancer.
The government researcher leading the development of the computer-assisted technology is optimistic that a blood test for ovarian cancer could be available as early as 2004. And tests for prostate, breast, and lung cancers could soon follow, predicted Emanuel Petricoin III, PhD, co-director of the Clinical Proteomics Program, a joint program of the FDA and the National Cancer Institute.
The blood test could prove one of the biggest developments in cancer research in years, he says. The benefits of the test would be twofold. Not only would it offer a way to detect some cancers earlier, when they're still curable, the test would also allow some patients to avoid unnecessary biopsies and all the anxiety and risks that come with them.
But before it's ready for prime time, doctors meeting at the 25th Annual San Antonio Breast Cancer Symposium cautioned the test needs to be validated in large numbers of men and women in clinical trials.
The test involves scanning tiny amounts of blood for hidden patterns of proteins that distinguish cancerous tissue from benign, much like the bar codes on food and household products that reveal their price at the supermarket checkout.
"All that's needed [for the quick fingerstick test] is a single drop of blood," Petricoin says. "The computer does the rest."
The feasibility of the approach was first proved in ovarian cancer, an often-deadly form of cancer because there is currently no way to detect it early, in its curable stages.
In tests on several hundred blood samples, some taken from women with ovarian cancer and others from healthy women, the test proved "an astonishing" 100% accurate in detecting cancer, even at the earliest stages, Petricoin said.
In contrast, the best screening method now available -- a blood test for levels of a protein known as CA-125 followed by ultrasound -- misses the vast majority of early tumors, he says. "By the time it's now diagnosed, ovarian cancer is too often deadly."
Based on these findings, the National Cancer Institute plans to begin a much larger clinical trial using the technology in women with ovarian cancer in early spring, Petricoin said. The object of that study will be to determine if the test can predict which women who are in remission will relapse.
But ovarian cancer is just one use of the technology, he stressed. "The beauty of this approach is that it's like building a platform for a house. Once we have the blueprint set up for ovarian cancer, it's easy to move into a clinical trial using the same platform, the same machine, for any type of cancer."
In fact, the test has already been studied in 317 women who underwent a breast biopsy after a suspicious mammogram. It was 90% accurate in detecting breast cancer in the samples. Plus, it accurately predicted that cancer was not present in 71% of the non-cancerous samples -- a finding that one day could help many women avoid unnecessary breast biopsies, Petricoin says.
The team is now working to improve the accuracy of the technology for spotting breast cancer before moving into larger clinical trials.
But the testing doesn't stop there. In a study of 50 lung cancer samples and 50 healthy samples, the test proved 95% accurate in detecting both cancerous and non-cancerous tissue, he says.
"A blood test for lung cancer would be a huge plus," Petricoin tells WebMD. "Right now, patients with suspicious [X-ray] findings have to undergo a chest biopsy, an invasive procedure that is painful and carries a risk of complications."
Other doctors at the meeting were cautiously optimistic.
William Gradishar, MD, associate professor of medicine at Northwestern University says, "A blood test for cancer would be very useful. But the track record with other blood tests has not been great. We need more data."
Petricoin offered assurance that the data would be forthcoming. "All the trials are proceeding. Our goal is to bring the test to the patients as quickly as possible while still thoroughly evaluating it," he said.