TORONTO — Cancer researchers have long been seeking a test that would allow them to accurately and easily diagnose ovarian cancer, one of the most lethal forms of the disease, in the early stages when it is highly treatable.

A study being published Tuesday by a group of Canadian and American scientists provides new hope that such a test may be in sight.

The researchers report that a blood test looking for four proteins or "biomarkers'' was highly effective in detecting ovarian cancer in a group of 242 women, roughly half of whom had the disease.

But while the results are extremely promising, they aren't yet good enough to suggest the test is ready for use as a screening tool on millions of seemingly healthy women, said senior author David Ward, a molecular biologist from Newfoundland who works at the Nevada Cancer Institute.

"It's not ready for prime time yet,'' Ward said Monday from Las Vegas.

"It has considerably higher sensitivity than anything we have at the moment, but there's still improvement to be made.''

Others agreed with that assessment, but hailed the results as a significant step forward.

"This is a very important study that requires confirmation in other groups' hands and on larger blinded sample sets to determine whether and where it has a role in early screening for ovarian cancer,'' said Dr. Gordon Mills, chair of molecular therapeutics at the M.D. Anderson Cancer Center in Houston, Texas.

"Being able to diagnose ovarian cancer early could have a major impact on the outcome.''

Ovarian cancer is often called a silent killer because the disease exhibits few or no symptoms in the early stages. Caught early, it has a cure rate of between 80 and 90 per cent. But fewer than 20 per cent of cases are caught before they progress to more serious disease. In later stages, the cure rate is 35 per cent or lower.

Adding to the difficulty of early diagnosis is the fact that at least 90 per cent of cases have no family history of the disease.

The Canadian Cancer Society estimates that 2,400 Canadian women will be diagnosed with ovarian cancer this year and 1,550 will die of the disease.

This research, reported in the Proceedings of the National Academy of Sciences, was done by scientists from Yale University in New Haven, Conn., George Washington University in Washington, D.C., and the Nevada Cancer Institute.

By measuring blood levels of proteins leptin, prolactin, osteopontin and IGF-II, the team was able to pinpoint with 95 per cent sensitivity and 95 per cent specificity the women in their group who had ovarian cancer.

Sensitivity measures a test's ability to detect all of the cases of whatever illness it is designed to find. Specificity indicates whether all the cases which tested positive actually have the illness -- in other words, whether it is turning up false positives.

High specificity is crucial in a screening test that would be used on large numbers of healthy people. That's because anyone who tests positive has to undergo additional testing to determine if they are a true case. That's expensive for the health care system and poses risks for individuals who may not have any disease at all.

In the case of ovarian cancer, confirmatory testing would likely require a laparoscopy, a surgical procedure in which a scope is inserted through an incision in the abdomen.

Even with a 95 per cent specificity, the test would turn up too many false positives to be acceptable, said co-author Patricia Bray-Ward, Ward's wife and a native of Kingston, Ont.

"You're going to end up doing laparoscopy on many women who don't have ovarian cancer,'' she said.

"And the problem with that is you're not only incurring anxiety, but there's clinical risk. Any time you put people to sleep and cut their bellies open, then there's unacceptable clinical risks to doing that for the number you would do.''

It is suggested that a test for early-stage ovarian cancer would have to have a specificity rating of 99.6 per cent before it could to be acceptable as a screening tool for the general public.

Ward said the group is currently studying the test's usefulness in women known to be at high risk of developing ovarian cancer. As well, it has identified other biomarkers which may help raise the specificity of the test to the necessary level.

Dr. Joan Murphy, a gynecological oncologist at Princess Margaret Hospital in Toronto, also found the research promising, but was reserving judgment.

Previous potential breakthroughs have failed to deliver, Murphy said, suggesting considerable additional study will be needed.

"Unfortunately, we've been down this road before and things have been reported that were promising,'' she said.

"In fact this does probably look a little better than some of the things we've been romanced by. But it takes lots and lots of women, years of observation.''