STANFORD, Calif. — Current diagnostic tests for ovarian cancer are woefully ineffective for early detection of the disease, say researchers at Stanford University School of Medicine. A new study finds that in order to make a significant dent in the mortality rate for the deadly cancer, the tests would have to be able to detect tumors of less than 1 cm in diameter, or about 200-times smaller in mass than those currently used to assess potential new tests. Still, if that hurdle can be overcome, there is good reason to believe that testing could make a big difference: The window of opportunity for treating these clinically undetectable cancers before they become life threatening is surprisingly long: about four years.
"We are miles away from detecting the most deadly ovarian tumors at this early stage," said Stanford biochemistry professor Patrick Brown, MD, PhD, "but now we have a chance of actually designing an effective test that will allow us to treat them before they become deadly." If a blood test is to be effective, said Brown, it will likely require identifying new markers that are never produced by normal cells — rather than testing for abnormally high levels of proteins detectable in normal blood, as current tests do. Other possible strategies might rely on new molecular imaging methods or fluid samples from the uterus or vagina — in which tumor markers are likely to be more concentrated.
The research will be published in the July 28 issue of the open-access journal PLoS Medicine. The article will be freely available to anyone after publication.
Ovarian cancer is particularly feared by women and their physicians because the disease is so difficult to detect in its early stages. Symptoms are vague, and often don't occur until the tumor is already several centimeters in diameter. At this point it may have already spread to surrounding organs and tissues. What's more, several published studies have indicated that the current screening tests deliver many false positive results and don't reduce mortality from the disease.
"Reliable early detection would save so many more lives than many new blockbuster anticancer drugs," said Brown, a Howard Hughes Medical Institute investigator and a member of the Stanford Cancer Center, who collaborated with the non-profit Canary Foundation to conduct the research. The foundation is dedicated to the early detection of many types of cancer. "If we can do this, which is no small challenge, the potential to go from a less than 20 percent chance of surviving five years to a relatively minor surgery that would have a very high cure rate is huge," he said.
Part of the difficulty in designing an effective test lies in the fact that there is more than one type of ovarian cancer. The most deadly, known as serous ovarian cancer, accounts for about 50 percent of all cases of ovarian cancer but it is responsible for at least 80 percent of deaths from the disease. In contrast to other types of ovarian tumors, which can grow to be quite large before spreading to other locations, serous ovarian tumors usually metastasize before they are diagnosed.
Cancers are classified, or staged, in part according to the degree of involvement of other organs; the less-deadly forms of ovarian cancer are usually diagnosed at an earlier stage of progression. Because researchers designing the diagnostic tests have assumed that these seemingly early cancers would eventually go on to become more invasive, they considered them to be good models for designing diagnostic tests aimed at detecting ovarian tumors in their infancy. However, these tumors are actually intrinsically different from the tumors that are diagnosed at a lethally advanced stage.
"It dawned on me at some point that we were being somewhat glib about what it was we were trying to detect," said Brown. "What we really needed to know is what the more-dangerous tumors looked like before we knew they were there."
Brown and his co-author, Chana Palmer, PhD, of the Canary Foundation, realized that it was possible to get just such a sneak peak at these tumors by looking at tissue from women carrying a genetic misstep called the BRCA-1 mutation. Because women with the mutation are very likely to develop breast or ovarian cancer, many elect to have their ovaries and Fallopian tubes removed as a preventive measure. Although these women appeared healthy at the time of their surgeries, close examination of the removed tissue indicated that some — about 8 percent, according to Brown and Palmer's analysis of previously published studies — had early, undiagnosed serous ovarian tumors.
The researchers combined the results of several previously published studies to estimate the prevalence, location, size and stage of the tumors. By comparing this information with the incidence of diagnosed serous ovarian tumors in a similar group of women, they calculated that the window of opportunity for early detection and possible successful treatment is about 4.3 years. During most of this time, the tumors were less than 1 cm in diameter; by the time the tumors reached 3 cm in diameter, more than half had advanced to stages III or IV (spread beyond the pelvis). As a comparison, the average diameter of an ovarian cancer tumor at the time of diagnosis is about 10 cm.
Brown estimated that most serous ovarian tumors in the study had progressed to an advanced stage almost a year before diagnosis. In order to halve the number of deaths from serous ovarian cancer, it will be necessary to have an annual screening test capable of detecting tumors about 0.5 cm in diameter — far beyond the capability of any currently available tests.
"This doesn't make me feel gloomy at all," said Brown, who is now studying whether it might be possible to detect ovarian cancer-specific markers in fluids sampled from the vagina or cervix. Such an approach may avoid the extreme dilution that interferes with detection of such a marker in blood samples. Based on the current research, Brown is also pursuing the Holy Grail of a truly cancer-specific molecular marker — a novel protein or DNA sequence that occurs only in cancer cells. Many current cancer-screening tests rely on changes in levels of particular markers that also occur, albeit at lower levels, on non-cancerous cells. It may also be possible to devise imaging techniques that could be useful screening tools, somewhat like mammograms for breast cancer.
Said Brown, "I was much more disheartened before we did this study, when we had no idea what we were looking for."