Ovarian cancer is one of the deadliest cancers, mainly because it's difficult to diagnose in early, more curable stages and usually isn't detected until it's advanced. That may be changing: A new screening technique developed by researchers at University College London can detect twice as many women with the deadly cancer as conventional methods.

The new method interprets changes in the levels of a protein called CA125 in a woman's blood. CA125 is found in cancerous tumors, and is especially concentrated in ovarian cancer cells. The new method detected cancer in 86 percent of women with invasive epithelial ovarian cancer (iEOC) where the two conventional tests would have identified less than half of these women (41 percent or 48 percent).

The results came from the world's largest ovarian screening trial involving post-menopausal women aged 50 and older who were randomly assigned to different annual screening strategies (multimodal screening or transvaginal ultrasound) or no screening at all. The 46,237 women underwent yearly screenings for up to 14 years.

In addition, the blood of all women was tested once a year for CA125 levels, and then a computer algorithm was used to interpret their risk of ovarian cancer. Risk was based on factors including the woman's age, the original levels of CA125, and how the level changed over time. The pattern was compared with known cases of cancer and controls to estimate the woman's risk of having ovarian cancer.

CA125 levels have been used as biomarkers to screen for ovarian cancers, but there was an arbitrary cutoff level. Some women, however, have CA125 levels that are higher than the cutoff, and some women have naturally lower levels. What best signals the presence of ovarian cancer is a rise in CA125 levels, which the new screening method detects.

"These results are therefore very encouraging," said Professor Usha Menon. "They show that use of an early detection strategy based on an individual's CA125 profile significantly improved cancer detection compared to what we've seen in previous screening trials."

"Our findings indicate that this can be an accurate and sensitive screening tool, when used in the context of a woman's pattern of CA125 over time," said Professor Ian Jacobs, currently President of The University of New South Wales, Australia.

"What's normal for one woman may not be so for another," said Jacobs, who created the trial and co-developed the computer program to measure changes in CA125. "It is the change in levels of this protein that's important. My hope is that this approach will prove capable of detecting ovarian cancer early enough to save lives."

The study was published in the Journal of Clinical Oncology.