A test that measures protein concentrations in the blood can signal the presence of ovarian cancer, a new study shows. The finding brings scientists a step closer to a diagnostic tool for catching this stealthy cancer early enough for effective treatment. The researchers caution, however, that they haven’t yet perfected the procedure.
There is currently no routine screening for ovarian cancer in the general population, even though more than 22,000 women are diagnosed with the disease each year in the United States. Doctors typically test for ovarian cancer when a woman experiences a combination of symptoms, such as abdominal pain, bloating, and abnormal vaginal bleeding, or if she has a family history of this cancer. Doctors use ultrasound or touch to examine the abdomen and can use a blood sample to look for a high concentration of the protein called CA125, which sometimes indicates ovarian cancer.
Unfortunately, these tests miss many early-stage ovarian cancers, a shortcoming that has tragic consequences. Most women found to have ovarian cancer don’t survive 5 years beyond the day of diagnosis, primarily because the cancer has spread unnoticed by the time it’s discovered. However, early-stage cancer that’s confined to the ovaries is highly treatable.
To devise an early-warning test, David C. Ward of the Nevada Cancer Institute in Las Vegas and his colleagues measured the relative concentrations of 169 proteins in the blood of ovarian cancer patients and healthy women. The amounts of 35 of the proteins varied significantly between the groups.
The researchers chose four of those proteins that are relatively simple to detect as the basis for a new ovarian cancer test. Two of the proteins, prolactin and osteopontin, showed up more abundantly in the cancer patients. The two others, leptin and insulin-like growth factor-II, were scarcer in the patients.
The researchers tried out the test on 246 blood samples, about half from cancer patients and half from healthy women. The four-protein analysis enabled the scientists, who didn’t know beforehand the source of a sample, to correctly identify the cancer status of the blood-sample donor 95 percent of the time. Notably, the test correctly spotted 26 of 27 early-stage ovarian cancers, the team reports in an upcoming Proceedings of the National Academy of Sciences.
By testing their approach on blood samples from unidentified women, the researchers “make a powerful argument for the potential of this type of strategy,” says John O. Schorge of the University of Texas–Southwestern Medical Center in Dallas.
Doctors at the Yale University School of Medicine are using the new analysis to test women who have a close relative with ovarian cancer or who harbor a mutation in one of the cancer-suppressing genes known as BRCA1 or BRCA2, Ward says. Such women are likely to be the first to benefit from a test for ovarian cancer, Schorge says.
Ward’s team is attempting to improve the test’s accuracy by adding other proteins to the array tested. “We absolutely have to increase the sensitivity of this test,” Ward says. Before such a test can make it to the clinic, it would also have to prove itself in evaluations involving blood samples from thousands of women.