Imagine a blood test that could detect the earliest signs of ovarian cancer to help far more women survive. Or one that could prevent thousands of aging men from undergoing unnecessary biopsies for prostate cancer.
Those tests are moving toward reality, thanks to new technology that can spot early signals in drops of blood.
The National Cancer Institute has begun a major study to prove if the blood test detects early relapse in ovarian cancer patients. Relapse occurs dismally often, and if the test works as well as earlier research suggests, it could win Food and Drug Administration approval for that use within a few years.
It would take longer to prove to FDA’s standards whether the test also can spot ovarian cancer the first time it strikes.
Labs to begin offering test soon
Two national testing laboratories aren’t waiting. Later this year, Quest Diagnostics and LabCorp hope to begin offering the blood test, by prescription, for women at high risk of ovarian cancer because of genetic or family history.
Despite caution from the test’s own inventors that it’s not yet ready for wide use, federal law allows those labs to offer tests that aren’t FDA-approved provided they meet other government certification standards, which they’re now attempting to do.
How does the testing work? It’s called proteomics, the study of all proteins in living cells Proteins are molecules that do the body’s work by directing cells’ actions. Scientists have long used single aberrant proteins as a signal, or biomarker, for different diseases — such as PSA, or prostate specific antigen, used to screen men for prostate cancer.
But one protein gone bad seldom is definitive. Indeed, most men with elevated PSA levels don’t have cancer but a benign enlarged prostate. Too often, it takes a surgical biopsy to tell.
About the new method
The new method: Proteins usually work through networks of circuit boardlike interactions that leave behind microscopic patterns. In a unique collaboration, scientists at the cancer institute and FDA discovered how to measure those patterns with special technology that picks out protein fragments floating in blood, patterns that can show when normal cells have turned cancerous.
“There is a wealth of information in the blood that we didn’t know about before,” says NCI’s Dr. Lance Liotta, who co-directs the program. “We’re finding an ocean of biomarkers.”
First, Liotta and his partner, FDA microbiologist Emanuel Petricoin, pursued patterns that signal ovarian cancer, a disease in desperate need of better diagnosis.
Some 25,580 women will be diagnosed with ovarian cancer this year, and 16,090 will die, the American Cancer Society estimates. Caught in its earliest stages, five-year survival is around 95 percent. But there is no screening method, so three-quarters of patients are diagnosed in advanced stages, when they have only a 20 percent survival chance.
Reducing the risk of false positives
Liotta and Petricoin tested blood samples from 250 women, some with ovarian cancer and some healthy. So far, the test has spotted all the cancers but sometimes falsely indicated that healthy women had cancer, too.
Nailing down that error rate, with much more study, is crucial, says NCI ovarian cancer specialist Dr. Elise Kohn. Even if the test ultimately were 98 percent accurate, she says, for every cancer caught, 49 healthy women would have unnecessary biopsies because of false results.
So Kohn is enrolling women who just finished initial ovarian cancer treatment into a clinical trial to see if the test detects relapse early. That’s easier to prove.
“This is not a consumer test,” stresses Quest spokesman Gary Samuels. Doctors would have to order it.
Prostate cancer is next on the government’s agenda. Initial research suggests a proteomics blood test might prevent 55 percent of unnecessary biopsies while missing 3 percent of cancers when given to men whose PSA levels are in a diagnostic gray zone. Now scientists are working to prove that and to see if protein patterns also signal which prostate tumors are aggressive and which are unlikely to kill.
Next on the list: patterns that seem to signal lung cancer and the earliest beginnings of deadly pancreatic cancer.