ABSTRACT: Transvaginal ultrasonography (US) has been used as a screening method for early detection of suspected ovarian malignancies in asymptomatic women. Most ovarian abnormalities (cysts) detected by a single transvaginal US resolve without treatment. Previous studies have shown that unilocular ovarian cysts and uncomplicated septate cysts have a very low probability of malignancy when followed with serial US. No data are available, however, on the frequency and disposition (outcome and/or resolution) of cysts with a complex structure monitored with serial US.
This study examined the prevalence, incidence, persistence, and resolution of ovarian abnormalities among a large cohort of asymptomatic women using serial transvaginal US. The investigators reviewed imaging data from 39,337 women enrolled in the University of Kentucky Ovarian Cancer Screening Program who were monitored from 1987 to 2012 with 221,576 baseline and interval transvaginal US.
Of the 39,337 participants, 31,834 (80.9%) had a normal US for the first and all subsequent visits, whereas 6807 (17.3%) had an abnormal transvaginal US and were monitored over 21,588 follow-up US scans. The incidence of ovarian cyst formation was significantly higher in premenopausal (34.9%) than in postmenopausal women (17.0%, P < 0.001). Similarly, the incidence of cysts was significantly higher among premenopausal women (15.3% premenopausal, 8.2% postmenopausal; P < 0.001). Among the women with cysts, the initial US was abnormal in 46.7% (3180/6807); 63.2% (2009/3180) of these cysts resolved to normal on subsequent US. Ten percent (3480/35,314) of the cases classified as normal on the first US were abnormal on subsequent annual examinations. Ovarian cysts were categorized as follows: 11.5% unilocular, 9.8% with septations, 7.1% with solid areas, and 1.8% predominantly solid. Many cysts detected on the initial screen were followed to resolution. A total of 557 participants underwent surgery: 85 ovarian malignancies and 472 nonmalignancies. Over the course of the study, serial US monitoring reduced the false-positive results, raising the positive predictive value from 8.1% to 24.7%.
These data show that many ovarian abnormalities spontaneously resolve over time, even if the initial US appearance is complex, solid, or bilateral. Therefore, a single transvaginal US abnormality should not be used as the sole indication for surgery. Sequential observation of cysts over time can reduce false-positive results and increase the positive predictive value.
Division of Gynecologic Oncology, Department of Obstetrics and Gynecology and the Department of Biostatistics, the University of Kentucky, Chandler Medical Center-Markey Cancer Center, Lexington, KY