Welcome
Around Harvard
This article originally appeared in the March 2008 Harvard Health Letter and is provided courtesy of Harvard Health Publications.
Finding ovarian cancer early
Will symptom awareness lead to earlier detection and better outcomes?
If ovarian cancer is diagnosed early, it can be treated successfully with surgery, although it’s a major operation: the ovaries, fallopian tubes, uterus, and omentum, a fatty layer of tissue that covers the intestines, are removed.
More often, though, it is diagnosed at a later stage — and the survival statistics are grim.
Tumors are occasionally detected during a routine pelvic exam, but the ovaries are anatomically “deep,” so early-stage cancers are rarely found.
Ovarian cancer does produce a protein called CA-125 that can be measured in the blood. But in premenopausal women many noncancerous conditions, like endometriosis, can elevate levels. Studies in postmenopausal women have also indicated that there would be an unacceptably high number of false positives. Research has shown that CA-125 testing would miss cancers, too.
Another possibility for early detection is transvaginal ultrasound, which involves getting an image of the ovaries with an ultrasound device inserted into the vagina. But a large study published in 2000 found that it took 5,200 ultrasounds to find just one case of ovarian cancer, a result that made screening women with transvaginal ultrasound seem impractical.
So there’s a big void. The Gynecologic Cancer Foundation, the American Cancer Society, and other groups tried to fill it in 2007 with a consensus statement about ovarian cancer symptoms.
The silent killer speaks
For years, ovarian cancer was portrayed as a disease that tended to sneak up, symptomless, on patients and their doctors. But many women who have had the disease have challenged that notion, saying they had distinctive abdominal pain or digestive problems before they were diagnosed. Some had wrenching stories about their complaints being ignored, or being referred for tests and treatments for other conditions like irritable bowel syndrome, only to be diagnosed later. On the other hand, the kinds of complaints women had are common. Doctors were rightly concerned about ordering tests for ovarian cancer based on symptoms that could have so many other causes.
Some large, carefully done surveys of ovarian cancer patients started to sort the situation out. Researchers found that many of the symptoms that women had prior to an ovarian cancer diagnosis weren’t classically gynecological; instead, women said they felt bloated, for example, or started to feel the need to urinate frequently. Researchers also found that even women diagnosed with early disease remembered having symptoms before their diagnosis.
But studies that depend on people’s recollections after they’ve been diagnosed with a disease cause a problem that researchers call recall bias. It’s been shown that after a diagnosis, especially when it’s of a serious condition like cancer, people rather selectively — and unwittingly — pick out clues about why they got sick and symptoms that might have foreshadowed their condition.
Dr. Barbara Goff, a University of Washington researcher, and her colleagues came up with a way to get around this problem, at least in part. Instead of questioning women already diagnosed with ovarian cancer, they questioned women prior to surgery to remove an ovarian or pelvic mass that might have been cancer. That timing presumably lessened the recall bias and also allowed them to compare women who turned out to have a benign growth with women who were diagnosed with cancer. They also recruited women being seen in primary care clinics as a comparison group. Their results, published in 2004 in the Journal of the American Medical Association, formed the basis of the symptoms guidelines promulgated in 2007. The main finding was that 43% of the women with ovarian cancer had a combination of bloating, increased abdominal size, and urinary symptoms, compared with just 8% of the primary care clinic patients. Moreover, the symptoms were more frequent (20 to 30 times a month) and more severe than those experienced by the women with benign growths and those seen at the primary care clinic.
Too vague?
When the consensus statement came out, many doctors had serious qualms. The symptoms are too vague, they said, and will lead to unnecessary workups for ovarian cancer. Some aren’t convinced that early-stage ovarian cancer produces these or any other symptoms, so a symptom index will only serve to identify late-stage cancers that aren’t curable. The statement also doesn’t spell out what doctors should do if women have these symptoms.
But the symptoms really aren’t quite as vague as some say: women are supposed to see a doctor if pelvic pain, bloating, or one of the other symptoms is new and if they have had it nearly every day for more than two or three weeks. Dr. Goff has written that even if the index only identifies late-stage cancers a bit earlier, that may do some good, because treatment will be more effective. The key, say some doctors, is to use good clinical judgment and to accept that a cluster of symptoms is never going to be as precise as a lab or imaging test.
