Ovarian cancer is the second most common gynecologic malignancy in the US and—because more than 70% of women affected are diagnosed with advanced-stage disease—the deadliest. Contrary to what might be expected, screening has not lessened mortality even in high-risk populations. The idea of targeting women having specific symptoms to be screened is relatively recent because of the general belief that symptoms have limited specificity. The authors undertook a case-control study of 149 women with ovarian cancer. An exploratory group served to estimate odds ratios for self-reported symptoms of variable severity, frequency, and duration. The results of logistic regression analysis were used to develop a symptom index, which then was tested in a confirmatory group. The group with ovarian cancer included 55 women with early-stage disease, 88 with late-stage disease, and 6 whose disease stage was not known. Thirty-one of the women with early-stage disease and all but 2 of those with advanced-stage disease had invasive tumors.
In the exploratory group, symptoms that were significantly associated with ovarian cancer, when occurring more than 12 days a month and for less than a year, included pelvic/abdominal pain, urinary urgency/frequency, increased abdominal size/bloating, and difficulty eating/feeling full. On logistic regression analysis, pelvic/abdominal pain, increased abdominal size/bloating, and difficulty eating/feeling full were independently associated with the presence of ovarian cancer. The symptom index was considered to be positive if any of these 6 symptoms occurred more than 12 times a month and had been present for less than 12 months. In the confirmatory group, the symptom index was 56.7% sensitive for early-stage disease and 79.5% sensitive for advanced-stage disease. The index was 90% specific for women older than 50 years of age and 86.7% specific for younger women. When analyzing 1709 women presenting to a primary care clinic, 45 of them (2.6%) tested positive and would have been identified as having symptoms very suggestive of ovarian cancer. The proportions of women less than 50 years of age and those aged 50 and older who tested positive were 3.3% and 1.4%, respectively.
The presence of specific symptoms, taking their duration and frequency into account, can help point to women who harbor ovarian cancer. Calculating the symptom index is a relatively simple means of identifying women who should be promptly and thoroughly evaluated.
Department of Obstetrics and Gynecology, University of Washington School of Medicine, Seattle, Washington; Molecular Diagnostic Program, Fred Hutchinson Cancer Research Center, Seattle, Washington; University of Washington School of Public Health, Seattle; and Seattle Ultrasound Associates, Seattle, Washington