Ovarian cancer is one of the top five causes of cancer death for women in the United States, with more than 14,000 American women expected to die of this disease in 2013.1 Bimanual pelvic examination, especially in postmenopausal women, has limited accuracy in detecting subtle changes in ovarian size and morphology.2,3 Transvaginal ultrasonography has an accepted and significant role in the management of suspected ovarian malignancies.4–6 Transvaginal ultrasonography has also been explored in healthy women without symptoms for the detection of early-stage ovarian cancer, which is often curable by conventional therapy.7–13
The literature contains reports on the occurrence of ovarian abnormalities detected by transvaginal ultrasonography.14–18 However, absent from the literature is information on the frequency and disposition of ovarian abnormalities under serial ultrasonographic surveillance, specifically those with a complex structure. In this investigation, we review transvaginal ultrasonography data from a cohort of 39,337 women enrolled in the University of Kentucky Ovarian Cancer Screening Program, which has been ongoing since 1987.19–22 This study collates information on the prevalence, incidence, persistence, and resolution of ovarian abnormalities.
MATERIALS AND METHODS
Women enrolled in the University of Kentucky Ovarian Cancer Screening Program from 1987 to 2012 composed the study group (n=39,337). An abnormal transvaginal ultrasonogram received by women in the ongoing study group was used as the basis for the sample reported on here. Approval was received from the University of Kentucky institutional review board. Study eligibility, exclusions, instrumentation, protocol, criteria for designating an abnormality, data collection, and storage were as previously reported.3,9,12,23,24 In brief, criteria for eligibility were: 1) asymptomatic women aged 50 years or older and 2) asymptomatic women aged 25 years or older with a documented family history of ovarian cancer in at least one primary or secondary relative. Genetic testing was not included in this trial. All study participants completed a questionnaire that included medical history, surgical history, last menstrual period or menopausal status, hormonal use, and family history of cancer. Women with a known ovarian tumor or a personal history of ovarian cancer were excluded from the screening trial. Participants in the screening program received free annual screening. Criteria for abnormality were based on ovarian volume (>2 standard deviations above normal mean volumes)24 and the presence of cysts (with septations, solid areas, or papillary projections) as well as echogenic solid structures. After an abnormal ultrasonographic result, repeat ultrasonograms were scheduled at intervals ranging from 6 weeks to 6 months. When an abnormality resolved, women were placed on annual follow-up. The duration of surveillance for abnormalities that resolved averaged 276±8.6 (standard error of the mean) days, whereas the total duration of surveillance that included normal results averaged 7.3 years for a total of 207,002 screening years. Women underwent surgery if the complexity of the abnormality increased to cystic with solid areas or to mostly solid, if there was an increase in volume greater than 50 cm3 associated with constant or increasing complexity, or if they had newly reported regional pain after a second abnormal ultrasonogram.
Bilaterality was defined as the simultaneous occurrence of an ovarian abnormality on both sides of the body, whereas unilaterality was defined as the occurrence of an abnormality in one ovary only. In this context, bilateral and unilateral cysts and solid masses were examined by categorizing observations into one of three groups:
Group A: the unilateral group having ovarian abnormalities on one side or the other but never on both sides simultaneously
Group B: the intermittent group having ovarian abnormalities on one side or the other at different times (nonsynchronous findings); in this group, the measurements were confined to nonsynchronous, unilateral ovarian abnormalities
Group C: the bilateral group having ovarian abnormalities present simultaneously on both sides with analyses exclusively confined to bilateral ovarian abnormalities
Ultrasonographic observations were classified as the following: A) normal ovaries; B) simple, unilocular cysts; C) cysts with septation(s) (uniloculated and multiloculated); D) cysts with solid areas; and E) solid masses. Tumors were classified using the World Health Organization histologic system and were staged according to the International Federation of Gynecology and Obstetrics system.
Definitions used in this study were as follows: true-positive (surgically confirmed malignancy after ultrasonograms classified as abnormal), true-negative (two negative ultrasonograms over no less than 365 days during which no malignancy is reported), and false-positive (no malignancy at surgery after ultrasonograms were classified as abnormal). Ovarian tumors of low malignant potential were included as true-positive results. Fallopian tube malignancies were categorized with ovarian malignancies, whereas peritoneal cancers were excluded because they do not originate in the adnexa.
The number of women with a normal initial ultrasonogram that changed to a new abnormal ultrasonogram on any subsequent visit was used to define incidence. The number of women who had an abnormal ultrasonogram on any visit was used to define prevalence. Significance was determined at the .01 level or better. Proportions were compared using the χ2 test. Resolution of ovarian abnormalities was determined by survival analysis using the Kaplan-Meier method with differences tested using the log rank test and interactive factors tested using the Cox regression proportional hazards model and the Cox-Mantel log rank test. Life tables were used from the Kaplan-Meier analyses to obtain numeric values of resolution.
The demographic characteristics of the study population are shown in Table 1. The frequency and occurrence of abnormal ultrasonograms are listed in Table 2. The vast majority of ovarian abnormalities persisted for multiple ultrasonograms. Nearly half of all participants (46.7%) found to have an ovarian abnormality were discovered on the first ultrasonogram, whereas 63.2% of women with an abnormal first ultrasonogram had a normal finding on the subsequent ultrasonogram. The first transvaginal ultrasonogram was normal in 89.8% of women screened and only 9.9% of these women subsequently experienced an abnormal transvaginal ultrasonogram. Lastly, 80.9% of the women enrolled in the study group never had an abnormal ultrasonogram (Table 2).
The prevalence and incidence of abnormal transvaginal ultrasonograms are shown in Table 3. Of the 39,337 participants, 17.3% had an abnormal ultrasonogram on their first or later visits, defining an ultrasonographic prevalence of 21,588 abnormal ultrasonograms in this population that had received 221,576 ultrasonograms (Table 3). The prevalence of transvaginal ultrasonographic abnormalities was significantly higher in premenopausal (34.9%) than postmenopausal women (17.0%, P<.001); likewise, the incidence of cyst formation was significantly higher for premenopausal women (15.3% premenopausal; 8.2% postmenopausal, P<.001).
The ovarian abnormalities were categorized as follows: 11.5% unilocular cysts, 9.8% cysts with septations, 7.1% cysts with solid areas, and 1.8% predominantly solid (Table 4). The abnormalities were further classified according to risk for malignancy as low risk (unilocular and cysts with septations) and high risk (cysts with solid areas or predominantly solid masses) (Table 4). Complex abnormalities (cysts with solid areas) and solid masses were more likely to resolve in a year of surveillance (76.5–80.6%) than unilocular cysts and cysts with septations (32.8–43.9%, P<.001). Complex abnormalities and solid masses had a median time to resolution of 7.8–8.7 weeks, whereas unilocular cysts and cysts with septations had a median time to resolution of 53–55.6 weeks (Table 4).
The duration of incident complex ovarian abnormalities (cyst with solid compared with predominantly solid) is graphed in Figure 1. Ultrasonographic abnormalities that were unilateral, asynchronously bilateral, or simultaneously bilateral were observed to resolve. Although many resolved in less than 6–12 months, others resolved over considerably longer periods of time. The effect of the size of the abnormality on resolution is presented in Figure 2 for both ovarian abnormalities that are cystic with solid components or predominantly solid, showing that the largest abnormalities resolved more slowly but still resolved.
We divided the screening data into three time periods based on the evolution of the ovarian cancer screening program algorithm (Table 5). Period A (December 14, 1987, to December 31, 1999) represents an era when serial transvaginal ultrasonography was not used. Period B (January 1, 2000, to December 31, 2007) incorporated serial monitoring of simple, unilocular cysts with the understanding that they were at low risk for malignancy.14,15 Period C (January 1, 2008, to May 15, 2012) expanded serial observation to also include septate cysts after published data confirmed this group to be at low risk for malignancy.16
There have been 85 true-positive malignancies identified through the ovarian cancer screening program and 472 nonmalignancies for a calculated positive predictive value (PPV) of 15.3% (Table 5). Notably, before serial transvaginal ultrasonography was used, the PPV was only 8.1%; however, as serial ultrasonography was introduced, the PPV progressively improved to 24.7% (Table 5), representing an improvement of 304% compared with the first study period. Each woman who had a serial transvaginal ultrasonography true-positive averaged 2.9 abnormal ultrasonograms and 7.5 total ultrasonograms (normal+abnormal). Finally, 463.8 women needed to receive ultrasonography to detect one malignancy, whereas 92.5 women with an abnormal transvaginal ultrasonogram were needed to detect one malignancy.
The distribution of findings for women with persisting abnormal transvaginal ultrasonograms who underwent surgery is shown in Table 6. Fifty-three invasive epithelial cancers were detected: stage I, 24 (45%); stage II, 12 (23%); stage III, 17 (32%); and stage IV, 0 (0%). Using a two-tier classification system in which grades 2 and 3 are combined, 83.3% (30 of 36) of early-stage epithelial cancers detected by screening were aggressive type II ovarian malignancies.25–28 Of the benign tumors surgically removed, serous cystadenomas were the most frequent (46.6%), whereas endometriomas, mucinous cystadenomas, leiomyomas, and cystic teratomas each accounted for 5–7% of the benign findings (Table 6).
It has been reported previously that unilocular ovarian cysts and uncomplicated septate cysts have a very low probability of malignancy and can be monitored with serial ultrasonography.14–16 Many of these low-risk cystic tumors spontaneously resolve without surgical intervention, even in postmenopausal women. Here we report additional findings showing that complex ovarian abnormalities formally considered high risk may resolve when monitored. As a result of serial ultrasonographic observation, we have shown that the number of operations performed for benign ovarian tumors can be reduced, effectively decreasing false-positive results and increasing the PPV of recent screening efforts to 24.7%.
The Prostate, Lung, Colon, Ovary trial29 relied on a single ultrasonographic abnormality to trigger a recommendation for surgery. Not surprisingly, this strategy generated a large number of false-positive results for nonmalignant tumors, although the overall PPV was not calculated.29 In a previous Prostate, Lung, Colon, Ovary trial publication that used the same algorithm, 19.5 women underwent surgery for each malignancy detected, for a calculated PPV of 5.1%.30
Our data would suggest that monitoring an ovarian cyst over time is a novel and valuable clinical variable. Serial transvaginal ultrasonography permits a functional evaluation of a tumor, identifying nonmalignant tumors that decrease in size and complexity and malignant tumors that often increase in both volume and morphologic complexity. Thus, a sequential observation of tumors over time can help segregate benign from malignant tumors. Because the duration of ovarian abnormalities may vary considerably, in the absence of persistent growth or worsening tumor complexity, short-term surveillance is a reasonable alternative to surgery. The ultrasonographic interval can be lengthened once evidence of improvement is documented. Complex tumors that increase in size and complexity should be surgically removed.
This longitudinal approach has translated into a significant improvement in ultrasonographic screening (PPV=24.7%), where serial transvaginal ultrasonography now has a level of performance comparable to mammography,31–33 especially with regard to the PPV.34–36 Moreover, transvaginal ultrasonography with serial morphology evaluation seems to perform favorably to published outcomes of diagnostic ovarian algorithms (PPV=29.5% overall; 17.9% premenopausal, 39.1% postmenopausal).37 This is particularly compelling because the cancer prevalence in a diagnostic population, where there is a known ovarian tumor, far exceeds that of a screening population.
The exact time interval recommended for serial observation is difficult to define. Our data suggest that the ultrasonographic interval depends on both the initial tumor morphology and the change in volume and complexity over time. In general, for unilocular or septate cysts, a 6-month surveillance ultrasonogram is suggested. Although more complex tumors may resolve, they should be cautiously followed at 6-week to 3-month intervals. The result of each subsequent ultrasonogram determines the recommended course of action, which could include continued monitoring or surgery. Although many cysts resolved in fewer than 6–12 months, some took considerably longer and, when documented to be stable, were followed on an annual basis.
One of the strengths of this investigation is that it presents data from a large, 25-year prospective population-based study with carefully defined ultrasonographic metrics. All ultrasonography is monitored for quality control and each abnormal result is individually reviewed by the principal investigator. Some have questioned the inclusion of borderline ovarian tumors as a true-positive result. The ovarian cancer screening program has maintained this classification since the inaugural ultrasonogram, which seems justified given the potential for having micropapillary projections and the association with peritoneal implants that correlate with increased risk. Moreover, both serous and mucinous ovarian borderline tumors can be lethal.38,39 The study was performed at a single institution (although multiple, statewide screening sites are used) and thus is susceptible to the biases associated with such an investigation. The screening algorithm has been cautiously modified over time to reflect the evolving knowledge of ovarian ultrasonography.
In current practice, many clinicians are faced with a small ovarian abnormality having some complexity that continues to persist over a long time. When it is clear that there are no increases in size or complexity, surveillance at a longer 6- to 12-month interval would seem appropriate. The data presented here indicate that some of these ovarian abnormalities can have a lengthy persistence. However, we have published ultrasonographic images of a variety of complex ovarian structure as the last view before resolution40 so that complexity alone does not indicate against resolution. Serial ultrasonography has established that many ovarian abnormalities resolve, even if they are complex, solid, or bilateral. With serial observation, more than 60% of ovarian abnormalities discovered on the initial transvaginal ultrasonogram returned to normal on subsequent visits. The strategy of serial ultrasonographic monitoring reduced the false-positive results over time, raising the PPV from 8.1% to 24.7%. Thus, it is advantageous to avoid a single abnormality as a trigger for surgery and to take a measured serial approach to improve the performance transvaginal ultrasonographic screening for ovarian cancer.
1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin 2013;63:11–30.
2. Padilla LA, Radosevich DM, Milad MP. Accuracy of the pelvic examination in detecting adnexal masses. Obstet Gynecol 2000;96:593–8.
3. Ueland FR, DePriest P, DeSimone C, Pavlik EJ, Lele SM, Kryscio RJ, et al.. The accuracy of examination under anesthesia and transvaginal sonography in evaluating ovarian size. Gynecol Oncol 2005;99:400–3.
4. Management of adnexal masses. ACOG Practice Bulletin No. 83. American College of Obstetricians and Gynecologists. Obstet Gynecol 2007;110:201–14.
5. Graham L. ACOG releases guidelines on management of adnexal masses. Am Fam Physician 2008;77:1320–3.
6. NICE Guidelines. Ovarian cancer: the recognition and initial management of ovarian cancer. 2011. Available at: http://www.nice.org.uk/nicemedia/live/13464/54266/54266.pdf
. Retrieved February 22, 2013.
7. Jacobs IJ, Bridges J, Reynolds C, Stabile I, Kemsleya P, Grudzinskas JG, et al.. Multimodal approach to screening for ovarian cancer. Lancet 1988;1:268–71.
8. Bourne TH, Campbell S, Reynolds KM, Whitehead MI, Hampson J, Royston P, et al.. Screening for early familial ovarian cancer with transvaginal ultrasonography and colour blood flow imaging. Br Med J 1993;306:1025–9.
9. DePriest PD, Gallion HH, Pavilk EJ, Kryscio RK, van Nagell JR. Transvaginal sonography as a screening method for the detection of early ovarian cancer. Gynecol Oncol 1997;65:408–14.
10. Jacobs I, Skates SJ, MacDonald N, Menon U, Rosenthal A, Prys Davies A. Screening for ovarian cancer: a pilot randomised controlled trial of ovarian cancer screening. Lancet 1999;353:1207–10.
11. Sato S, Yokoyama Y, Sakamoto T, Futagami M, Saito Y. Usefulness of mass screening for ovarian carcinoma using transvaginal ultrasonography. Cancer 2000;89:582–8.
12. van Nagell JR, DePriest P, Reedy M, Gallion HH, Ueland FR, Pavlik EJ, et al.. The efficacy of transvaginal sonographic screening in asymptomatic women at risk for ovarian cancer. Gynecol Oncol 2000;77:350–6.
13. Buys SS, Partridge E, Greene MH, Prorok PC, Reding D, Riley TL, et al.. Ovarian cancer screening in the Prostate, Lung, Colorectal and Ovarian (PLCO) cancer screening trial: findings from the initial screen of a randomized trial. Am J Obstet Gynecol 2005;193:1630–9.
14. Modesitt SC, Pavlik EJ, Ueland FR, DePriest PD, Kryscio RJ, van Nagell JR. Risk of malignancy in unilocular ovarian cystic tumors less than 10 centimeters in diameter. Obstet Gynecol 2003;102:594–9.
15. Greenlee RT, Kessel B, Williams CR, Riley TL, Ragard LR, Hartge P. Prevalence, incidence, and natural history of simple ovarian cysts among women >55 years old in a large cancer screening trial. Am J Obstet Gynecol 2010;202:373.e1–9.
16. Saunders BA, Podzielinski I, Ware RA, Goodrich S, DeSimone CP, Ueland FR, et al.. Risk of malignancy in sonographically confirmed septated cystic ovarian tumors. Gynecol Oncol 2010;118:278–82.
17. Hartge P, Hayes R, Reding D, Sherman ME, Prorok P, Schiffman M, et al.. Complex ovarian cysts in postmenopausal women are not associated with ovarian cancer risk factors. Am J Obstet Gynecol 2000;183:1232–7.
18. Sharma A, Apostolidou S, Burnell M, Campbell S, Habib M, Gentry-Maharaj A, et al.. Risk of epithelial ovarian cancer in asymptomatic women with ultrasound-detected ovarian masses: a prospective cohort study within the UK collaborative trial of ovarian cancer screening (UKCTOCS). Ultrasound Obstet Gynecol 2012;40:338–44.
19. van Nagell JR, Higgins RV, Donaldson ES, Gallion HH, Powell DE, Pavlik EJ, et al.. Transvaginal sonography as a screening method for ovarian cancer: a report of the first 1000 cases screened. Cancer 1990;65:573–7.
20. van Nagell JR, DePriest PD, Puls LE, Donaldson ES, Gallion HH, Pavlik EJ, et al.. Ovarian cancer screening in asymptomatic postmenopausal women by transvaginal sonography. Cancer 1991;68:458–62.
21. van Nagell JR Jr, DePriest PD, Gallion HH, Pavlik EJ. Ovarian cancer screening. Cancer 1993;71(suppl):1523–8.
22. van Nagell JR Jr, Gallion HH, Pavlik EJ, DePriest PD. Ovarian cancer screening. Cancer 1995;76(suppl):2086–91.
23. Pavlik EJ, Saunders BA, Doran S, McHugh KW, Ueland FR, DeSimone CP, et al.. The search for meaning—symptoms and transvaginal sonography screening for ovarian cancer. Cancer 2009;115:3689–98.
24. Pavlik EJ, DePriest PD, Gallion HH, Ueland FR, Reedy MB, Kryscio RJ, et al.. Ovarian volume related to age. Gynecol Oncol 2000;77:410–2.
25. Malpica A, Deavers MT, Lu K, Bodurka DC, Atkinson EN, Gershenson DM, et al.. Grading ovarian serous carcinoma using a two-tier system. Am J Surg Pathol 2004;28:496–504.
26. Gershenson DM. The heterogeneity of epithelial ovarian cancer: getting it right. Cancer 2010;116:1400–2.
27. Bodurka DC, Deavers MT, Tian C, Sun CC, Malpica A, Coleman RL, et al.. Reclassification of serous ovarian carcinoma by a 2-tier system. A Gynecologic Oncology Group Study. Cancer 2012;118:3087–94.
28. Steffensen KD, Waldstrom M, Grove A, Lund B, Pallisgard N, Jakobsen A. Improved classification of epithelial ovarian cancer. Results of 3 Danish Cohorts. Int J Gynecol Cancer 2011;21:1592–600.
29. Buys SS, Partrideg E, Black A, Johnson CC, Lamerato L, Isaacs L, et al.. Effect of screening on ovarian cancer mortality—the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Randomized Controlled Trial. JAMA 2011;305:2295–303.
30. Partridge E, Kreimer AR, Greenlee RT, Williams C, Xu J-L, Church TR, et al.. Results from four rounds of ovarian cancer screening in a randomized trial. Obstet Gynecol 2009;113:775–82.
31. Humphrey LL, Hefland M, Chan BKS, Woolf SH. Breast cancer screening: a summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 2002;137:347–60.
32. Elmore JG, Armstrong K, Lehman CD, Fletcher SW. Screening for breast cancer. JAMA 2005;293:1245–56.
33. Lehman CD, Gatsonis C, Kuhl CK, Hendrick RE, Pisano ED, Hanna L, et al.. MRI Evaluation of the contralateral breast in women with recently diagnosed breast cancer. N Engl J Med 2007;356:1295–303.
34. Bennett RL, Blanks RG. Should a standard be defined for the Positive Predictive Value (PPV) of recall in the UK NHS Breast Screening Programme? Breast 2007;16:55–9.
35. Kopans DB. The positive predictive value of mammography. AJR Am J Roentgenol 1992;158:521–6.
36. Venkatesan A, Philip C, Kerlikowske K, Sickles EA, Smith-Bindman R. Positive predictive value of specific mammographic findings according to reader and patient variables. Radiology 2009;250:648–57.
37. Miller RW, Smith A, DeSimone CP, Seamon L, Goodrich S, Podzielinski I, et al.. Performance of the American College of Obstetricians and Gynecologists’ ovarian tumor referral guidelines with a multivariate index assay. Obstet Gynecol 2011;117:1298–306.
38. Uzan C, Kane A, Rey A, Gouy S, Pautier P, Lhomme C, et al.. How to follow up advanced-stage borderline tumors? Mode of diagnosis of recurrence in a large series stage II-III serous borderline tumors of the ovary. Ann Oncol 2011;22:631–5.
39. Kikkawa F, Nawa A, Kajiyama H, Shibata K, Ino K, Nomura S. Clinical characteristics and prognosis of mucinous tumors of the ovary. Gynecol Oncol 2006;103:171–5.
40. Pavlik EJ, van Nagell JR Jr. Early detection of ovarian tumors using ultrasound. Womens Health (Lond Engl) 2013;9:39–55.