Characterization of prediagnostic symptoms among primary epithelial ovarian cancer cases and controls

The goal of this study was to characterize types, frequency, combinations, and relative onset of symptoms among ovarian cancer cases and controls.

Participants were from an ongoing, population-based, case-control study of primary epithelial ovarian cancer in a 48-country region of North Carolina. Incident cases (N = 267), aged 20-74, were identified by area hospitals on a rapid case ascertainment basis between April 1999 and March 2001. Population-based controls, with at least one intact ovary, frequency matched on age and race, were identified using random-digit dialing (N = 287) and Health Care Financing Administration (HCFA) phone lists (N = 30). Trained nurse-interviewers, using a standardized questionnaire, asked participants about specific symptoms experienced for at least 2 weeks in the year prior to diagnosis (cases)/interview (controls).

More than 90% of cases reported at least one symptom and symptoms were most often the reason for the doctor visit leading to diagnosis (74%), followed by routine examination (12%). Among invasive cases, symptoms with onsets longer before diagnosis (median 5-7 months) included gas/nausea/indigestion; urinary frequency/urgency; bowel irregularity; abnormal menstrual/vaginal bleeding or discharge; pain during intercourse; and ongoing fatigue. Symptoms with onsets closer to diagnosis (median of 2-4 months) included distended/hard abdomen; bloating/feeling of fullness; unexplained weight gain/loss; pelvic/abdominal discomfort; chest pain/respiratory difficulties; and "other" symptoms. Controls reported fewer symptoms than cases (median 1 vs 5 or 6). Control symptoms were of longer duration and much less likely to occur in combination.

Earlier diagnosis of ovarian cancer may be possible if women and physicians recognize the importance of combinations of seemingly unrelated symptoms, especially those identified as occurring longer before diagnosis.