Long-Term Survival of Women With Epithelial Ovarian Cancer Detected by Ultrasonographic Screening

Original Research

Long-Term Survival of Women WithEpithelial Ovarian Cancer Detected byUltrasonographic ScreeningJohn Rensselaer van Nagell Jr, MD, Rachel Ware Miller, MD, Christopher P. DeSimone, MD,Frederick R. Ueland, MD, Iwona Podzielinski, MD, Scott T. Goodrich, MD, Jeff W. Elder, MD,Bin Huang, PhD, Richard J. Kryscio, PhD, and Edward John Pavlik, PhD

OBJECTIVE: To estimate the effect of ultrasonographicscreening on stage at detection and long-term disease-specific survival of women with epithelial ovarian cancer.

METHODS: Eligibility included all asymptomatic womenaged 50 years and older and women aged 25 years andolder with a documented family history of ovarian can-cer. From 1987 to 2011, 37,293 women received annualultrasonographic screening. Women with abnormal screensunderwent tumor morphology indexing, serum biomarkeranalysis, and surgery.

RESULTS: Forty-seven invasive epithelial ovarian cancersand 15 epithelial ovarian tumors of low malignant poten-tial were detected. No women with low malignant po-tential tumors experienced recurrent disease. Stage dis-tribution for invasive epithelial cancers was: stage I, 22(47%); stage II, 11 (23%); stage III, 14 (30%), and stage IV,0 (0%). Follow-up varied from 2 months to 20.1 years(mean, 5.8 years). The 5-year survival rate for invasiveepithelial ovarian cancers detected by screening was:

stage I, 95%�4.8%; stage II, 77.1%�14.5%; and stage III,76.2%�12.1%. The 5-year survival rate for all womenwith invasive epithelial ovarian cancer detected byscreening as well as interval cancers was 74.8%�6.6%compared with 53.7%�2.3% for unscreened women withovarian cancer from the same institution treated by thesame surgical and chemotherapeutic protocols (P<.001).

CONCLUSION: Annual ultrasonographic screening ofasymptomatic women achieved increased detection of early-stage ovarian cancer cases and an increase in 5-year disease-specific survival rate for women with ovarian cancer.(Obstet Gynecol 2011;118:1212–21)DOI: 10.1097/AOG.0b013e318238d030

LEVEL OF EVIDENCE: II

Ovarian cancer remains the leading cause ofgynecologic cancer mortality in the United

States despite advances in radical surgery, postoper-ative care, and chemotherapy.1 In 2010, more than15,000 deaths from ovarian cancer were reported inthe United States alone.1 Although certain symptomshave been associated with ovarian cancer,2,3 thesesymptoms are nonspecific, and the majority ofwomen continue to present with advanced-stage dis-ease. Early-stage ovarian cancer, however, is highlycurable when treated by conventional therapy.4

Screening asymptomatic women at risk for ovariancancer has been proposed as a means to facilitateearlier diagnosis and improve survival.

In 1987, the University of Kentucky Ovarian Can-cer Screening Trial was initiated to assess the efficacy ofannual transvaginal ultrasonography as a screeningmethod for ovarian cancer.5 Since that time, free screen-ing has been provided to more than 37,000 women. Thefollowing investigation was performed to documentstage at detection and long-term survival of patients withovarian cancer detected by screening in this trial.

See related editorial on page 1209.

From the Division of Gynecologic Oncology, Department of Obstetrics andGynecology, and the Department of Statistics, the University of KentuckyChandler Medical Center-Markey Cancer Center, Lexington, Kentucky.

Supported by research grants from the Kentucky Department of Health andHuman Services and the Telford Foundation.

Portions of the Materials and Methods section have been published previously invan Nagell JR Jr, DePriest PD, Ueland FR, DeSimone CP, Cooper AL,McDonald JM, et al. Ovarian cancer screening with annual transvaginalsonography: findings of 25,000 women screened. Cancer 2007;109:1887–96.© 2007 American Cancer Society.

Corresponding author: John Rensselaer van Nagell Jr, MD, Division ofGynecologic Oncology, University of Kentucky-Markey Cancer Center, 800 RoseStreet, Lexington, KY 40536-0293; e-mail: [email protected].

Financial DisclosureThe authors did not report any potential conflicts of interest.

© 2011 by The American College of Obstetricians and Gynecologists. Publishedby Lippincott Williams & Wilkins.ISSN: 0029-7844/11

1212 VOL. 118, NO. 6, DECEMBER 2011 OBSTETRICS & GYNECOLOGY

MATERIALS AND METHODSPortions of the “Materials and Methods” have beenpublished previously6 and are repeated here. Womenwho enrolled in the University of Kentucky OvarianCancer Screening Trial from January 1987 to June2011 were evaluated. This trial was approved by theUniversity of Kentucky Human Investigation ReviewBoard for Human Studies. Eligibility criteria included1) all asymptomatic women aged 50 years or older;and 2) asymptomatic women aged 25 years or olderwith a documented family history of ovarian cancer inat least one primary or secondary relative. Genetictesting was not performed routinely as part of thistrial. Before study entry, all study participants com-pleted a questionnaire that included medical history,surgical history, menopausal status, hormonal use,and family history of cancer. Menopause was definedas the absence of menses for 12 months. Women witha known ovarian tumor or a personal history ofovarian cancer were excluded from this investigation.

After informed consent was obtained, each pa-tient underwent screening according to the algorithm

illustrated in Figure 1. Since 2000, transvaginal ultra-sonography has been performed using General Elec-tric Logiq 400 units with a 5-mHz vaginal probe asdescribed previously.6 Doppler flow and three-dimen-sional ultrasonography were performed in womenwith persisting ovarian abnormalities using a GeneralElectric Voluson 730 ProV unit with a 5- to 9-mHzvaginal probe. All ultrasonographic images were re-viewed by at least one of the authors. At each screen,both ovaries were measured in three dimensions.Ovarian volume was calculated using the prolateellipsoid formula (length�width�height�0.523). Allscreening information was entered into a Medlogdatabase on a local network. Criteria for abnormalityincluded an ovarian volume more than 20 cm3 forpremenopausal women and more than 10 cm3 forpostmenopausal women. These values were usedbecause they were more than 2 standard deviationsabove the mean volume of normal ovaries in pre-menopausal and postmenopausal women.7 In addi-tion, any cystic ovarian tumor with a solid or papillaryprojection into its lumen was considered abnormal.

Transvaginal ultrasoundscreening

Normal Abnormal

Repeat transvaginalultrasound screening

in 1 year


in 4–6 weeks

Normal Abnormal


in 1 year

Proceed with:Tumor indexingCA-125Color Doppler

ultrasonography

Cystic or septated tumorless than 10 cm in

diameter; CA-125 lessthan 35 U/mL

Complex or solid tumor


in 6 months

Diagnostic surgery

Fig. 1. Study algorithm from theUniversity of Kentucky OvarianCancer Screening Trial. Modifiedfrom van Nagell JR Jr, DePriestPD, Ueland FR, DeSimone CP,Cooper AL, McDonald JM, et al.Ovarian cancer screening withannual transvaginal sonography:findings of 25,000 women screened.Cancer 2007;109:1887–96. © 2007American Cancer Society.van Nagell. Ovarian CancerScreening and Long-Term Survival.Obstet Gynecol 2011.

VOL. 118, NO. 6, DECEMBER 2011 van Nagell et al Ovarian Cancer Screening and Long-Term Survival 1213

Women who had a normal screen were scheduled toreturn in 12 months for a repeat screen. Women whohad an abnormal screen underwent repeat ultrasono-gram 4–6 weeks later. Women who had abnormalsecond screens had a serum CA 125 determination,tumor morphology indexing, and Doppler flow ultra-sonography. Morphology indexing was performedaccording to the classification of Ueland and col-leagues.8 Each tumor was given a score from 1 to 10according to increasing morphologic complexity andvolume (Fig. 2). During the last 5 years of this trial,women with unilocular cystic ovarian tumors or sep-tated cystic tumors that measured less than 10 cm ingreatest dimension and a normal serum CA 125 havebeen followed without surgery by repeat ultrasonog-raphy at 6-month intervals. Women with a persistingsolid or complex ovarian tumor or women with acystic ovarian tumor and an elevated serum CA 125level were advised to undergo surgical removal of thetumor. During the past 15 years of this investigation,every effort was made to perform laparoscopic tumor

removal as the initial step in the operative evaluationof women with persisting ovarian tumors detected byscreening.

At the time of laparoscopy, ovarian tumors wereplaced in an endoscopic bag intra-abdominally andremoved through a midline subumbilical incision.Patients with ovarian cancer on frozen section orpatients with obvious metastatic disease at laparos-copy underwent immediate exploratory laparotomyand staging. Tumors were classified histologicallyaccording to the World Health Organization systemand were staged according to the International Fed-eration of Gynecology and Obstetrics system. Aftersurgery, patients were treated according to the celltype, histologic differentiation, and stage of eachtumor, usually with six cycles of combination plati-num-based chemotherapy. Patients with ovarian can-cer were followed at monthly intervals during treat-ment, every 3 months for 2 years after completion oftreatment, and every 6 months thereafter. Follow-updata on all patients were coordinated with the Amer-ican Cancer Society, the University of KentuckyTumor Registry, the Kentucky Cancer Registry, andthe Kentucky State Department of Vital Statistics.Cause of death was confirmed by analysis of relevantclinical records and state death certificates.

Patients with epithelial ovarian cancer entered inthe University of Kentucky Tumor Registry or thestatewide Kentucky Cancer Registry from 1995 to2011, but who did not receive screening, served ascontrol groups for this investigation. These womenwere from the same geographic area as women in thescreening group and were referred to the Universityof Kentucky-Markey Cancer Center or other institu-tions served by the Kentucky Cancer Registry afterclinical detection. The time period for evaluation ofthe control groups was chosen because after 1995, allpatients were surgically staged and received platinumand taxane chemotherapy. Women in the controlgroup were not matched case-by-case to women inthe screening group but represented disease-specificsurvival rates of local and regional patients withovarian cancer detected without screening. Controlpatients treated at the University of Kentucky-MarkeyCancer Center all underwent surgical staging andmaximum tumor cytoreduction and received at leastsix courses of combination carboplatin and taxanechemotherapy at this institution.

Proportions were compared by using �2 statistics.Statistical significance was determined at the .05 level.Survival rate was estimated using the Kaplan-Meiermethod with differences tested using the log rank testand interactive factors tested using the Cox regression

Fig. 2. Morphology index (numeric value 0–10). Reprintedfrom Gynecol Oncol, Vol. 91, Ueland FR, DePriest PD,Pavlik EJ, Kryscio RJ, van Nagell JR Jr, Preoperative differ-entiation of malignant from benign ovarian tumors: theefficacy of morphology indexing and Doppler flow sonog-raphy, 46–50, © 2003 with permission from Elsevier.van Nagell. Ovarian Cancer Screening and Long-Term Survival.Obstet Gynecol 2011.

1214 van Nagell et al Ovarian Cancer Screening and Long-Term Survival OBSTETRICS & GYNECOLOGY

proportional hazards model. Life table plots areshown in place of stepwise Kaplan-Meier plots be-cause screened patients had longer survival times thanunscreened members of the comparison groups.

RESULTSThirty-seven thousand two hundred ninety-threewomen enrolled in the University of Kentucky Ovar-ian Cancer Screening Trial from January 1987 to June2011. A positive family history of ovarian cancer wasdocumented in 8,309 (22%) of these women. Clinicalcharacteristics of the women screened are illustratedin Table 1. Women enrolled in this study underwenta total of 205,190 scans (mean, 5.5 scans per partici-pant). One or both ovaries were visualized in 169,197scans (87.6%). Ultrasonographically undetectableovaries were considered normal for purposes of thisinvestigation.

Five hundred twenty-three women (1.4%) with apersisting ovarian tumor on transvaginal ultrasonog-raphy underwent surgery (Table 2). The most com-mon histologic diagnoses were: serous cystadenoma,primary ovarian carcinoma, endometrioma, and mu-cinous cystadenoma. Fifteen epithelial tumors of lowmalignant potential, five nonepithelial ovarian malig-nances, and nine cancers metastatic to the ovary weredetected by screening.

The statistical definitions used in this investiga-tion are presented in Table 3. The 76 women with anabnormal screen and invasive epithelial ovarian can-cer (n�47), ovarian tumors of low malignant potential(n�15), nonepithelial ovarian cancer (n�5), or cancermetastatic to the ovary (n�9) were considered true-positives. The 447 women with positive screens whohad benign ovarian tumors were considered false-positives. The 36,758 patients with negative screenswho did not have ovarian cancer were classified astrue-negatives. The 12 women who developed ovar-ian cancer within 1 year of a negative scan wereconsidered false-negatives. Using these data, the sen-sitivity of transvaginal ultrasonography screening was86.4%, and the specificity was 98.8%. The positivepredictive value of an abnormal screen was 14.53%,and the negative predictive value of a normal screenwas 99.97%. After 2006, the positive predictive valueincreased to 20.17% when the screening algorithmwas changed such that surgery was not recommendedin women with unilocular cystic ovarian tumors orseptated cystic ovarian tumors less than 10 cm indiameter.

Forty-seven women with invasive epithelial ovar-ian cancer and 15 women with epithelial ovariantumors of low malignant potential detected by screen-

ing are the subjects of the present investigation. Thefive women with nonepithelial ovarian cancer and thenine women with cancer metastatic to the ovary arenot considered further in this study.

Demographic and histologic variables in womenwith low malignant potential epithelial ovarian tu-mors are as follows: the mean age of these women was61.7 years (median, 63 years; range, 48–78 years), themean weight was 167 pounds (median, 162 pounds;range, 100–240 pounds), and the mean gravidity was2.0 (median, 2; range, 0–6). Thirteen of these tumorscontained both cystic and solid components on ultra-sonography, and two were entirely solid. The meanmorphology index score was 5.1 (median, 5; range,3–7) compared with 3.5 (median, 3; range, 0–10) forpatients with benign tumors (P�.001). Forty percentof women with low malignant potential ovarian tu-mors had a palpable abnormality on clinical exami-nation, and only two had an elevated serum CA 125(more than 35 units/mL). Twenty percent of womenwith a low malignant potential tumor (3 of 15) re-ported a family history of ovarian cancer, and onlyone was diagnosed under the age of 50 years. Thestage distribution of women with low malignant po-tential ovarian tumors was as follows: stage IA, six;stage IB, two; stage IC, six; and stage IIIA, one.Twelve patients had serous ovarian tumors, one pa-tient had an endometrioid ovarian tumor, and twopatients had a mucinous ovarian tumor, all of border-line malignancy. An ovarian abnormality was de-tected on the first screen in 53.3% of these cases. Allwomen were treated primarily by surgery, and nonereceived postoperative chemotherapy. Patients havebeen followed 0.2–14.8 years (mean, 6.2 years; me-dian, 6 years) after diagnosis, and none has developedtumor recurrence or died of any cause.

Demographic and histologic variables in the 47women with invasive epithelial ovarian cancer de-tected by screening are as follows: the mean age ofthese women was 64.5 years (median, 66 years; range,36–86 years), the mean weight was 158 pounds(median, 151 pounds; range, 98–270 pounds), and themean gravidity was 2.3 (median, 2; range, 0–8). Anabnormality in ovarian morphology or volume wasdetected on the first screen in 16 patients (34%) andon subsequent screens in 31 patients (66%). All ovar-ian malignancies were solid or complex (both solidand cystic components) except one with cystic mor-phology and a thickened cyst wall. The mean mor-phology index score of ovarian malignancies was 6.1(median, 6; range, 1–10) compared with 3.5 (median,3; range, 0–10) for benign ovarian tumors and signif-icantly different (P�.001). Twenty-one of these


women (45%) had a palpable abnormality on clinicalexamination (note that a physical examination wasperformed only on patients in which the ultrasono-graphic examination was abnormal and was biased byknowledge of the ultrasound result.) Thirteen percentof women with invasive ovarian cancer detected byscreening (six of 47) reported a family history ofovarian cancer with two women diagnosed under theage of 50 years. At the time of surgery, 22 of the 47women (47%) with epithelial ovarian cancer detectedby screening had stage I disease and 11 (23%) hadstage II disease. In contrast, 701 of 2,560 (27%)unscreened women with clinically detected epithelialovarian cancer treated at hospitals throughout Ken-tucky and entered into the Kentucky Tumor Registryhad stage I or stage II disease at the time of diagnosis(P�.01). Fourteen women (30%) in the screeninggroup had stage III ovarian cancer at the time ofdetection as compared with 46% in the unscreenedgroup (P�.028). There was also a substage shift withinstage III ovarian cancer in the screening group. Halfof the women with stage III ovarian cancers detectedby screening had stage IIIA or stage IIIB tumors,whereas 84% of unscreened women with stage IIIovarian cancer presented with stage IIIC disease(P�.01). There were no cases of stage IV ovariancancer detected in women undergoing screening ei-ther at the initial prevalence screen or in subsequentincidence screens.

Serum CA 125 was elevated (more than 35units/mL) in 10 of 31 women (32%) with stage I orstage II ovarian cancer detected by screening and in11 of 14 women (79%) with stage III disease (note thattwo women did not have a preoperative CA 125determination). The cell types of the ovarian cancer

were as follows: serous carcinoma, 18; adenocarci-noma, 14; endometrioid carcinoma, 12; clear cellcarcinoma, two; and mucinous carcinoma, one. Fourtumors were well differentiated, nine were moder-ately differentiated, and 34 were poorly differentiated.According to the histologic criteria of Kurman andShih,9 10 women (21%) had type I ovarian cancer and37 women (79%) had type II ovarian cancer.

Patients with invasive ovarian cancer all under-went total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, maximal tumor cy-toreduction, and staging. Complete tumor debulking(no visible residual disease) was achieved in all pa-tients with stage I–IIIA disease. Eighty-six percent ofpatients with stages IIIB and IIIC disease were cytore-duced to less than 1 cm residual disease. Threepatients with stage IA grade 1 endometrioid carci-noma were treated by surgery alone. The remainingpatients all received a minimum of six cycles ofplatinum and taxane chemotherapy. For the past 10years of this investigation, patients requiring chemo-therapy usually have received six cycles of intrave-nous carboplatin and paclitaxel every 3 weeks as theirinitial postsurgery treatment regimen.

After therapy, patients were followed from 0.1 to20.1 years (mean, 5.8 years; median, 5 years) andnone have been lost to follow-up. To date, ninescreen-detected women have died of ovarian cancer,and six screen-detected women have died of othercauses with no evidence of disease. Thirty-one pa-tients are alive with no evidence of disease, and onepatient is alive with disease. The causes of death inpatients who died with no evidence of ovarian cancerare as follows: two patients died of a myocardialinfarction and one each of multiple myeloma, chronic

Table 1. Clinical Characteristics of Patients Screened

All All True-Negative True-Negative

n 37,293 36,758Age (y) 56.6 56 (25–95) 56.6 56 (25–95)Parity 2.4 2 (0–18) 2.3 2 (0–18)Weight (lb) 160.1 153 (80–450) 160.1 153 (84–450)Height (in.) 64.3 64 (49–78) 64.3 64 (49–78)

n % n %

Family history ofOvarian cancer 8,309 22.3 8,170 22.2Breast cancer 15,333 41.1 15,103 41.1Colon cancer 9,272 24.9 9,149 24.9

No history of hormone therapy 7,074 19.0 7,021 19.1Hormone therapy on last visit 4,756 12.8 4,703 12.8Nulliparous 5,547 14.9 5,470 14.9

Data are mean or median (range) unless otherwise specified.


obstructive airway disease, and colon cancer. One of22 patients (4.5%) with stage I disease, four of 11patients (36%) with stage II disease, and four of 14(29%) with stage III disease have died of ovariancancer. The 5-year survival rate for invasive epithelialovarian cancers detected by screening was: stage I,95%�4.8%; stage II, 77.1%�14.5%; and stage III,76.2%�12.1%. During this trial, 12 women havedeveloped interval ovarian cancer within 1 year of anormal screen (false-negative), and seven of themhave died of disease. These women generally hadpoorly differentiated cancers with minimal ovarianenlargement on transvaginal ultrasonography, whichwere diagnosed 6–12 months after a normal scan.The 2-year and 5-year survival rates for these interval

cases were 75%�12.5% and 57%�14.6%, respec-tively.

The 5-year disease-free survival rate for womenwith stage I and stage II epithelial ovarian cancerdetected by screening was 95%�4.8% and 77.1%�14.5%, respectively. The survival rates for womenwith stage I or stage II ovarian cancer did not varysignificantly in the screening or control groups (Fig.3). However, the survival rate for women with ad-vanced disease was significantly higher in the screen-ing group (Fig. 3). This is attributable to the fact thatthere were proportionally more cases of stage IIIAand IIIB ovarian cancers detected by screening andthat no patient developed stage IV disease whilebeing screened. The 5-year disease-free survival ratefor the entire group of women whose invasive epithe-lial ovarian cancers were detected by screening was

True-Positive True-Positive False-Positive False-Positive False-Negative False-Negative

76 447 1263.7 59 (36–86) 59 54 (29–97) 67.1 59 (54–80)2.1 2 (0–8) 2.3 2 (0–10) 3.4 3 (0–9)

159.4 152 (98–270) 162.8 157 (80–376) 151.1 146 (110–240)64.5 65 (60–70) 64.6 65 (55–72) 64.8 65 (62–68)

n % n % n %

12 15.8 128 28.6 4 33.340 52.6 196 43.8 6 50.013 17.1 110 24.6 1 8.35 6.6 47 10.5 2 16.79 11.8 44 9.8 1 8.3

16 21.1 60 13.4 2 16.7

Table 2. Histologic Findings in Patients WithPersistently Abnormal Screens (n�523)

Histologic Finding n

Epithelial ovarian cancer 47Nonepithelial ovarian cancer 5Epithelial ovarian tumor of LMP 15Metastatic cancer to ovary

Appendix 3Primary peritoneum 3Breast 1Endometrium 1Colon 1

Serous cystadenoma 208Endometrioma 31Mucinous cystadenoma 27Cystic teratoma 25Fibroma, thecoma, or Brenner tumor 38Leiomyoma 23Hydrosalpinx or paratubal cyst 31Hemorrhagic cyst or other 64

LMP, low malignant potential.

Table 3. Statistical Definitions in Ovarian CancerScreening

Term Screen FindingNo. of

Participants

True-positivePositive Histology confirms ovarian cancer 76

False-positivePositive Benign ovarian histology 447

True-negativeNegative No evidence of disease 12 months

after negative screen36,758

False-negativeNegative Ovarian cancer diagnosed within

12 months of negative screen12

Sensitivity: 86.4% (95% confidence interval [CI] 84.6–88.2);specificity: 98.8% (95% CI 98.7–98.9); positive predictivevalue positive: 14.53% (95% CI 12.99–16.07); positivepredictive value negative 99.97% (95% CI 99.95–99.99).


84.6%�5.6% as compared with 53.7%�2.3% in un-screened women with epithelial ovarian cancertreated at the University of Kentucky-Markey CancerCenter from 1995 to 2011 (P�.001). The women in

the control group all were surgically staged andtreated by the same surgical and chemotherapeuticprotocols as those women whose cancers were de-tected by screening. The 5-year survival rate forwomen in the screened group was also significantlyhigher (P�.001) than that of unscreened womentreated at multiple centers throughout the state andentered from 1995 to 2011 in the Kentucky CancerRegistry (84.6%�5.6% compared with 48.4%�1.2%).When the interval (false-negative) cases are included,the 5-year disease-free survival rate for women withovarian cancer in the screening group is 74.8%�6.6%,which remains significantly (P�.001) higher thanthose women treated at this institution who did nothave screening (53.7%�2.3%).

DISCUSSIONTheoretically, the use of screening to detect ovariancancer at an earlier stage should result in morecurable cases; however, there are limited data tosuggest that screening actually results in reducedovarian cancer mortality.10–15 Screening research in-volves analysis of large numbers of women and takesconsiderable planning, discipline, time, and financialsupport to perform. Ovarian cancer meets the criteriaof a disease that should benefit from screening.16 First,ovarian cancer is highly curable with conventionaltreatment methods when detected at an early stage.Numerous studies have confirmed that the 5-yearsurvival rate for women with stage I epithelial ovariancancer is as high as 95% in major cancer centers.4

Recent studies have indicated that symptoms such asearly satiety, abdominal bloating, and pelvic painoccur more commonly in women with ovarian cancerthan in control patients without the disease.2,3 How-ever, these symptoms are nonspecific and are oftenignored by many women, resulting in significantdelays before seeking medical evaluation. It is esti-mated that without screening, only 15% of ovariancancers are localized to the ovary at the time ofdiagnosis.4 This is understandable because ovaries arepalpable in less than 30% of postmenopausal womenat the time of examination under anesthesia.17 There-fore, it is extremely difficult to detect subtle changesin ovarian morphology or volume associated withearly-stage ovarian cancer by clinical examinationalone.

One of the difficulties in designing screeningstrategies for ovarian cancer is that it is not a commondisease. The incidence of ovarian cancer in the gen-eral population is 20–30 per 100,000 in womenyounger than 40 years of age and rises to 40 per100,000 at the age of 50 years.18–20 Therefore, screen-

Screened populationn=47

Unscreened populationUniversity of Kentucky Hospital Registry

n=741P<.001



n=488 P<.02



n=253

Not significant; P=.883

Epithelial ovarian cancer: Stages I and II

Epithelial ovarian cancer: Stages III and IV

Epithelial ovarian cancer: All stages

Disease-specific survival (years)



Pro

babi

lity

of s

urvi

val

Pro

babi

lity

of s

urvi

val

Pro

babi

lity

of s

urvi

val

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

0 1 2 3 4 5 6 7 8 9 10

0 1 2 3 4 5 6 7 8 9 10

0 1 2 3 4 5 6 7 8 9 10

Fig. 3. Ten-year survival rates of screened and unscreenedwomen with epithelial ovarian cancer by life table method;P value obtained by the log rank test.van Nagell. Ovarian Cancer Screening and Long-Term Survival.Obstet Gynecol 2011.


ing studies must focus specifically on women who areat high risk for the disease. Entry criteria for thepresent screening trial included only asymptomaticwomen older than age 50 years or women older thanage 25 years with a documented family history ofovarian cancer. The screening algorithm followed inthis investigation used transvaginal ultrasonographyas the initial screening test with serum CA 125obtained only in women with a persisting ultrasono-graphic abnormality. This algorithm detected 47 in-vasive epithelial ovarian cancers and 15 epithelialovarian tumors of low malignant potential in 37,293women screened or 1.6 epithelial ovarian cancercases detected per 1,000 women screened.

An important factor in screening is the intervalbetween the onset of ovarian cancer and the presenceof clinical symptoms (the preclinical phase) duringwhich screening intervention should lead to earlierdetection. Recently, Crum and coworkers21 and Kur-man and Shih9 have postulated that there are twotypes of epithelial ovarian cancer. Type I ovariancancers are typically low grade endometrioid, clearcell, or mucinous cell types and have profiles of geneexpression similar to those of low malignant potentialtumors. In contrast, type II ovarian cancers are high-grade serous carcinomas, endometrioid carcinomas,or undifferentiated carcinomas, which are biologicallyaggressive. Type II ovarian cancers typically displayp53 mutations and are thought to have a shorterpreclinical phase than type I tumors. According tothis model, annual screening would be expected to bemost effective in detecting type I tumors while miss-ing the more aggressive type II tumors, which areresponsible for the majority of ovarian cancer deaths.In the present investigation, 21 of the 33 women(64%) with stage I or II invasive epithelial ovariancancers detected by screening had poorly differenti-ated serous carcinomas, poorly differentiated endo-metrioid carcinomas, or undifferentiated adenocarci-nomas (type II cancers). Sixteen of these 21 women(76%) are alive and well with no evidence of diseaseafter conventional treatment with surgery and plati-num-based chemotherapy. One can postulate thatwithout screening, the majority of these womenwould have progressed rapidly to advanced-stageovarian cancer, and many would have died of disease.

A concern about ultrasonographic screening is alow positive predictive value in differentiating malig-nant from benign ovarian tumors. In the multicenterProstate, Lung, Colorectal, and Ovarian Cancer trial,for example, 19.5 surgeries were required to detectone ovarian cancer (positive predictive value 5.1%)when screening was performed using a combination

of transvaginal ultrasonography and serum CA 125.13

In the present investigation, 523 women with persist-ing ovarian tumors underwent surgery. Forty-seven ofthese women had invasive epithelial ovarian cancer,15 had epithelial ovarian tumors of borderline malig-nancy, five had nonepithelial ovarian cancer, andnine had cancer metastatic to the ovary. Therefore,6.9 surgical procedures were performed per ovariancancer detected (positive predictive value 14.53%).The difference in positive predictive value noted inthis trial and that reported in the Prostate, Lung,Colorectal, and Ovarian Cancer trial can be ex-plained partially by a higher cancer prevalence in thescreened population and by two modifications in theUniversity of Kentucky screening algorithm, whichwere adopted in the past 5 years. First, analysis ofmore than 3,200 unilocular cystic ovarian tumors lessthan 10 cm in diameter detected by screening re-vealed that the risk of malignancy in these tumors wasessentially nonexistent. Women with these tumorswere followed ultrasonographically for an average of6.3 years, and none developed ovarian cancer.22 As aresult, the screening algorithm was changed, andcystic ovarian tumors less than 10 cm in diameterwere no longer removed surgically. In a subsequentinvestigation,23 2,870 complex septated cystic ovariantumors identified by ultrasonographic screening wereevaluated for risk of malignancy. Surgery was per-formed in 128 women and no tumor with this mor-phologic pattern was malignant. The remainingwomen were followed ultrasonographically every4–6 months without surgery for an average of 6.4years. Thirty-eight percent of these tumors resolvedspontaneously, and no patient developed ovariancancer. Therefore, the screening algorithm was fur-ther amended in 2009 such that surgery was notrecommended in women with septated cystic ovariantumors without solid areas or papillary projections.Ovarian tumors with these two morphologic patternsmade up 55% of the total abnormalities detected byscreening. As a result of these changes, the positivepredictive value of screening in this trial increasedfrom 14.1% (1987–2006) to 20.2% (2006–2011). Fi-nally, a specific treatment algorithm was followed inthis trial for women with a persisting ovarian abnor-mality on transvaginal ultrasonography. This in-cluded careful morphologic evaluation of each tumor,Doppler analysis of tumor blood flow, and serumbiomarker analysis before a patient underwent sur-gery. In the Prostate, Lung, Colorectal, and OvarianCancer trial, there was no uniform treatment algo-rithm, and treatment was at the discretion of thepatient’s primary care physician.14 Clearly, further


research is needed concerning the relationship oftumor morphology, molecular genetic patterns, andbiomarker profiles to risk of malignancy so thatsurgery can be avoided in women with ovariantumors at minimal risk for neoplasia. This shouldresult in progressively increasing the positive predic-tive value for multimodality screening.

One of the objective measurements of an effec-tive screening test is its ability to lower stage atdetection and disease-specific mortality. In the Pros-tate, Lung, Colorectal, and Ovarian Cancer trial, bothtransvaginal ultrasonography and CA 125 were eval-uated as screening tests. Although transvaginal ultra-sonography was associated with a low positive predic-tive value, 72% of cases detected by ultrasonographyalone were stage I or stage II.13 Screening with serumCA 125 was associated with a higher positive predic-tive value, but 74% of biomarker-detected ovariancancers were stage IIIC or IV. Data from the presenttrial are consistent with these findings in that 33 of the47 cases (70%) of the women whose ovarian cancerswere detected by transvaginal ultrasonographyscreening had stage I or II disease as opposed to 27%of women with clinically detected ovarian cancerfrom the same population referred to institutionsthroughout the state (P�.01). Similarly, in the multi-institutional U.K. Collaborative Trial of Ovarian Can-cer Screening trial, 48% of ovarian cancers detected inthe prevalence screen by multimodal screening hadstage I–II disease compared with 26% in the un-screened population.24

The increase in long-term ovarian cancer survivalobserved in the screening group in the present trial isdirectly attributable to lower stage at detection. The5-year survival rate for women with stage I or IIepithelial ovarian cancer was 89%, and 70% of thetotal screening group had these two stages of diseaseat the time of detection. Whereas more than twothirds of women in the screening group were detectedwith stage I–II disease, more than two thirds ofunscreened women with clinically detected ovariancancer presented with stage III–IV ovarian cancer.There was also a substage shift within stage III in thescreening group with 50% of women being detectedwith stage IIIA or IIIB disease as opposed to only 16%in women without screening. Finally, no woman inthe screening group had or developed stage IV ovar-ian cancer during this trial, whereas 19.6% of un-screened women with ovarian cancer had stage IVdisease at the time of clinical diagnosis.

Data from the present investigation indicate thatannual ultrasonographic screening of women at riskfor ovarian cancer caused a decrease in stage at

detection and site-specific ovarian cancer mortalitywhen compared with women from the same geo-graphic area who did not have screening. The ob-served 14.5% positive predictive value associated withtransvaginal ultrasonography screening is unaccept-ably low, and routine screening of the general popu-lation cannot be recommended at the present time.However, focused research concerning risk of malig-nancy in ultrasonographically identified ovarian tu-mors should result in more selective surgery and anincrease in positive predictive value to levels associ-ated with screening for other malignancies.

The importance of early detection in ovariancancer is evident in that more than two thirds ofasymptomatic women with ovarian cancer detectedby screening in the present trial had localized diseaseat the time of detection, and their 5-year survival ratewas nearly 90%. In the absence of early detection,most women will continue to present with advanced-stage disease for which the cost of treatment is highand the cure rate is low.

REFERENCES1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics. 2010. CA

Cancer J Clin 2011;61:133–4.

2. Goff BA, Mandel LS, Melancen CH, Munz HG. Frequency ofsymptoms of ovarian cancer in women presenting to primarycare clinics. JAMA 2004;291:2705–12.

3. Goff BA, Mandel LS, Drescher CW, Urban N, Gough S,Schurman KM, et al. Development of an ovarian cancersymptom index. Cancer 2007;109:221–7.

4. Jelovac D, Armstrong DK. Recent progress in the diagnosisand treatment of ovarian cancer. CA Cancer J Clin 2011;61:183–203.

5. van Nagell JR, Higgins RV, Donaldson ES, Gallion HH,Powell DE, Pavlik EJ, et al. Transvaginal sonography as ascreening method for ovarian cancer. A report of the first 1000cases screened. Cancer 1990;65:573–7.

6. van Nagell JR, DePriest PD, Ueland FR, DeSimone CP,Cooper AL, Pavlik EJ, et al. Ovarian cancer screening withannual transvaginal sonography: findings of 25,000 womenscreened. Cancer 2007;109:1887–96.

7. Pavlik EJ, DePriest PD, Gallion HH, Ueland FR, Reedy MB,Kryscio RJ, et al. Ovarian volume related to age. GynecolOncol 2000;77:410–2.

8. Ueland F, DePriest P, Pavlik E, Kryscio R, van Nagell JR.Preoperative differentiation of malignant from benign ovariantumors: the efficacy of morphology indexing and Doppler flowsonography. Gynecol Oncol 2003;91:46–50.

9. Kurman RJ, Shih IeM. The origin and pathogenesis of epithe-lial ovarian cancer: a proposed unifying theory. Am J SurgPathol 2010;34:433–43.

10. Jacobs I, Skates SJ, MacDonald N, Menon U, Rosenthal A,Prys Davies A. Outcome of a pilot randomised controlled trialof ovarian cancer screening. Lancet 1999;253:1207–10.

11. Sato S, Yokoyama Y, Sakamoto T, Futagami M, Saito Y.Usefulness of mass screening for ovarian carcinoma usingtransvaginal ultrasonography. Cancer 2000;89:582–8.


12. Buys S, Partridge E, Greene MH, Prorok PC, Reding D, RileyTL, et al. Ovarian cancer screening in the Prostate, Lung,Colorectal and Ovarian (PLCO) cancer screening trial: find-ings from the initial screen of a randomized trial. Am J ObstetGynecol 2005;193:1630–9.

13. Partridge E, Kreimer AR, Greenlee RT, Williams C, Xu JL,Church TR, et al. Results from four rounds of ovarian cancerscreening in a randomized trial. Obstet Gynecol 2009;113:775–82.

14. Buys S, Partridge E, Black A, Johnson CC, Lamerato L, Isaacs C,et al. Effect of screening on ovarian cancer mortality. TheProstate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screen-ing Randomized Controlled Trial. JAMA 2011;305:2295–303.

15. Havrilesky LJ, Sanders GD, Kulasingam S, Chino JP, BerchuckA, Marks JR, et al. Development of an ovarian cancer screen-ing decision model that incorporates disease heterogeneity:implications for potential mortality reduction. Cancer 2011;117:545–53.

16. Prorok PC. Evaluation of screening program for the earlydetection of cancer. Natl Cancer Inst Stat Textbk Monogr1984;51:267–328.

17. Ueland FR, DePriest P, DeSimone C, Pavlik EJ, Kryscio RJ,van Nagell JR Jr. The accuracy of examination under anesthe-sia and transvaginal sonography in evaluating ovarian size.Gynecol Oncol 2005;99:400–3.

18. Kosary CL. Cancers of the ovary. SEER survival monograph:cancer survival among adults: US SEER program 1988–2001.

National Institutes of Health Publication 7-6215. Bethesda(MD): National Institutes of Health; 2007. p. 133–44.

19. Yancik R. Ovarian cancer. Age contrasts in incidence, histol-ogy disease stage at diagnosis, mortality. Cancer 1993;71(suppl):517–23.

20. Quirk JT, Natarajan N, Mettlin CJ. Age-specific ovarian cancerincidence rate patterns in the United States. Gynecol Oncol2005;99:248–50.

21. Crum CR, Drapkin R, Miron A, Ince T, Muto M, Kindel-berger DW, et al. The distal fallopian tube: a new model forpelvic serous carcinogenesis. Curr Opin Obstet Gynecol 2007;19:3–9.

22. Modesitt SC, Pavlik EJ, Ueland FR, DePriest PD, Kryscio RJ,van Nagell JR Jr. Risk of malignancy in unilocular ovariancystic tumors less than 10 centimeters in diameter. ObstetGynecol 2003;102:594–9.

23. Saunders BA, Podzielinski I, Ware R, Goodrich S, DeSimoneC, Ueland FR, et al. Risk of malignancy in sonographicallyconfirmed septated cystic ovarian tumors. Gynecol Oncol2010;118:278–82.

24. Menon U, Gentry-Maharj A, Hallet R, Ryan A, Burnell M,Jacobs I, et al. Sensitivity and specificity of a multimodal andultrasound screening for ovarian cancer, and stage distribution ofdetected cancers: results of the prevalence screen of the UKCollaborative Trial of Ovarian Cancer Screening (UKCTOCS).Lancet Oncol 2009;10:327–40.


Long-Term Survival of Women With Epithelial Ovarian Cancer Detected by Ultrasonographic Screening

Documents

Transcript of Long-Term Survival of Women With Epithelial Ovarian Cancer Detected by Ultrasonographic Screening