|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 24 (August 20), 2005: pp. 5588-5596 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.05.097 Screening for Familial Ovarian Cancer: Failure of Current Protocols to Detect Ovarian Cancer at an Early Stage According to the International Federation of Gynecology and Obstetrics SystemFrom the Southeast of Scotland Clinical Genetic Services, Western General Hospital; Simpson Centre for Reproductive Health, Royal Infirmary of Edinburgh; Ultrasound Department, Simpson Centre for Reproductive Health, Royal Infirmary of Edinburgh, Edinburgh, Scotland; Academic Unit of Medical Genetics and Regional Genetics Service and Department of Radiology, St Mary's Hospital, Manchester; Wessex Clinical Genetic Services, Princess Anne Hospital, Southampton; University of St Andrews, Bute Medical Buildings, St Andrews; and Lothian Health Care NHS Trust, The Dean Terrace Centre, Edinburgh, United Kingdom Address reprint requests to Diane Stirling, Macmillan Nurse Specialist in Genetics, Southeast of Scotland Clinical Genetic Services, Western General Hospital, Crewe Rd, Edinburgh, EH4 2XU Scotland; e-mail: diane.stirling{at}luht.scot.nhs.uk
PURPOSE: To assess the effectiveness of annual ovarian cancer screening (transvaginal ultrasound and serum CA-125 estimation) in detecting presymptomatic ovarian cancer in women at increased genetic risk. PATIENTS AND METHODS: A cohort of 1,110 women at increased risk of ovarian cancer were screened between January 1991 and March 2004; 553 were moderate-risk individuals (4% to 10% lifetime risk) and 557 were high-risk individuals (> 10% lifetime risk). Outcome measurements include the number and stage of screen-detected cancers, the number and stage of cancers not detected at screening, the number of equivocal screening results requiring recall/repetition, and the number of women undergoing surgery for benign disease. RESULTS: Thirteen epithelial ovarian malignancies (12 invasive and one borderline), developed in the cohort. Ten tumors were detected at screening: three International Federation of Gynecology and Obstetrics (FIGO) stage I (including borderline), two stage II, four stage III, and one stage IV. Of the three cancers not detected by screening, two were stage III and one was stage IV; 29 women underwent diagnostic surgery but were found not to have ovarian cancer. CONCLUSION: Annual surveillance by transvaginal ultrasound scanning and serum CA-125 measurement in women at increased familial risk of ovarian cancer is ineffective in detecting tumors at a sufficiently early stage to influence prognosis. With a positive predictive value of 17% and a sensitivity of less than 50%, the performance of ultrasound does not satisfy the WHO screening standards. In addition, the combined protocol has a particularly high false-positive rate in premenopausal women, leading to unnecessary surgical intervention.
Advances in the understanding of the genetics of ovarian cancer have led to an increased public and professional awareness of risks associated with a family history of the disease. Demand for ovarian cancer screening has increased, yet there is no clear evidence of its effectiveness in groups at increased risk.1-3 Primary carcinoma of the ovary is the fourth most common cause of death as a result of cancer in the United Kingdom, with 4,000 deaths each year.2,4 Advances in therapy have had little impact on this figure. Although etiology of the disease is poorly understood, both genetic and epidemiologic factors have been identified. A positive family history of ovarian cancer is the strongest known risk factor, whereas associations with ovulation pattern, nulliparity, and older age at first childbirth have also been reported.1-3,5 Prognosis in ovarian cancer is dependent on the International Federation of Gynecology and Obstetrics (FIGO) stage at initial presentation. The 5-year survival rate decreases from 85% in stage I to 10% in stage IV.6,7 The lack of specific symptoms in the early stages means that the majority of cases present late, hence the interest in presymptomatic screening, especially in women at increased risk. Epithelial tumors account for more than 90% of all ovarian cancers, and the lifetime risk is estimated to be between 1% and 1.5%. This risk increases with the number of affected relatives and the closeness of relationship. The lifetime risk estimate for women with one first-degree relative is 2% to 5%, and when there is more than one close relative affected, the risk increases to 10% to 23%.1,2,8 Highly penetrant cancer-predisposing genes account for 5% to 10% of all epithelial ovarian cancers.2,5,8 Those identified to date include BRCA1 and BRCA2, associated with breast and ovarian cancer, and the mismatch repair genes associated with hereditary nonpolyposis colon cancer. Germline mutations in these two classes of genes can confer an increased risk of ovarian cancer of up to 40% and 12%, respectively.9-11 In most instances, increased genetic risk is recognized first from family history alone and mutation screening may follow. For women at increased risk, management options include ovarian screening and prophylactic surgery.12 This study examined the efficacy of annual ovarian screening.
Three cancer genetics centers participated: Edinburgh, Manchester, and Southampton (United Kingdom). All women presenting to one of these centers, between January 1991 and March 2004, had their family history documented, assessed, and confirmed, as far as practicable, by clinical genetic services. Women fulfilling one of the family history increased-risk criteria (Table 1) were offered screening. Risk categories 1 to 3 represent moderate risk (4% to 10% lifetime) and categories 4 to 7 represent high risk (> 10% lifetime).
Women in both categories were offered annual ovarian cancer screening; however, genetic testing of BRCA1 and BRCA2 was only available to families meeting the high-risk criteria (in which there was a living affected relative). In several instances, women diagnosed during the study were offered genetic testing after diagnosis. Annual screening included transvaginal ultrasound (TVU) and serum CA-125 estimations (in Manchester before 1999, CA-125 was measured only if there was an abnormality on TVU). Cutoff values for CA-125 varied between centers and even with time in the same center, as new assays were developed, but generally were between 15 and 35 U/mL. Wherever possible, ultrasound scans were performed transvaginally. Ovarian volume was noted along with any morphologic abnormalities. Cysts were noted to be unilocular (simple), multilocular, or complex, and the wall was noted to be smooth or irregular with or without septations. The volume of the cyst was also noted. In premenopausal women, any cyst larger than 2.5 cm was reviewed in 6 weeks. Patients with multilocular or complex cysts were referred to a gynecologist for opinion, usually resulting in laparoscopy. Any patient with an enlarging cyst that did not resolve on sequential scans was also referred to a gynecologist. The methodology for this study was approved by the relevant local regional ethics committee at each center, and written informed consent was obtained from the participants.
In total, 1,110 women were screened; 1,048 had ultrasound scans and 760 women had serum CA-125 measurements. Table 2 records the sensitivity, specificity, and positive and negative predictive values for both modalities based on the following findings.
Ultrasound A total of 3,701 annual scans were performed on 1,048 women; 508 women were at moderate risk and 540 women were at high risk. A total of 614 of the women were premenopausal, 100 were perimenopausal, 140 were postmenopausal, and 194 did not have their menopausal status recorded. Equivocal findings meant that 9.8% of the annual scans had to be repeated once and 1.5% had to be repeated twice. Table 3 shows an additional breakdown of the ultrasound results by center and genetic risk. Ultrasound abnormality was noted in six of the 13 ovarian cancers; one of these was a borderline tumor and all but one of the affected women were postmenopausal. Scanning failed to demonstrate an abnormality in seven ovarian cancers; four of the women had functional ovaries. In 29 women, ultrasound demonstrated abnormalities that led to diagnostic surgery but proved not to be cancer. At least 23 (79%) of these women had functional ovaries. Figure 1 depicts these outcomes as true positives, false negatives, and false positives.
CA-125 A total of 2,617 annual CA-125 measurements were recorded for 760 women; 365 were moderate risk and 395 were high risk. A total of 401 of the women were premenopausal, 74 were perimenopausal, 106 were postmenopausal, and 179 did not have their menopausal status recorded. Results above the normal range meant that 5.5% of the annual CA-125 estimations had to be repeated once and 1.5% had to be repeated twice. Table 4 lists an additional breakdown of the CA-125 results by center and genetic risk. Eleven of the 13 women who developed ovarian cancer had undergone CA-125 screening. Levels were increased in nine of the ovarian cancers; one of these was a borderline tumor. Three of the affected women were premenopausal and six were postmenopausal. Five had an abnormality on ultrasound in addition to the increased CA-125. In two patients with ovarian cancer (one premenopausal and one postmenopausal) the serum CA-125 was within the normal range. Ultrasound detected one of these and the other was discovered at prophylactic surgery. In five patients (four premenopausal) elevated CA-125 levels, confirmed on repeat testing, led to diagnostic surgery but no cancer was present. Figure 2 depicts these outcome measurements as true positives, false negatives, and false positives.
Ovarian Cancers Thirteen epithelial ovarian cancers developed in the 1,100 women. Details are listed in Tables 5 and 6. All but one occurred in women in the high-risk category (they were considered to be at high risk before their own diagnosis): six had previously suffered breast cancer, seven were identified as pathogenic mutation carriers (six BRCA1, one BRCA2), and one was from a BRCA1 family but had not been tested. Mean age at diagnosis was 53 years (range, 39 to 64 years). Ten cancers (including the borderline tumor) were detected by screening: three FIGO stage I (including borderline), two stage II, four stage III, and one stage IV. One cancer was detected by ultrasound alone, four were detected by CA-125 alone, and five were detected by both (three at the first screening and seven at annual follow-up).
Three cancers were not detected by screening. One was discovered at prophylactic salpingo-oophorectomy 2 months after a normal CA-125 and TVU, and two presented symptomatically, one 12 months and one 4 months after the patients' most recent normal screening (Table 6).
Surgery for Benign Disease
The desire to improve the outcome for women at increased risk of ovarian cancer by providing regular screening is understandable, particularly in view of the promising results from corresponding programs for familial breast cancer.13,14 Various studies have examined the efficacy of ovarian screening using TVU, serum CA-125 estimation, color Doppler, pelvic examination, or a combination of these methods. These studies have shown diversity in selection criteria, some selecting women with only one affected close relative who may therefore have a lower risk—the low incidence of ovarian cancer in some studies may reflect this. The largest and most recent studies are listed in Table 8. 15-22 Three studies used risk criteria similar to those used in the present study, selecting women with two or more first- or second-degree relatives affected with ovarian cancer or another associated cancer.15,18,21,22
The largest and the most recent study15 screened 1,117 women, either every 6 months or annually (depending on family history). All six cancers that developed in this study were detected by screening: three of the cancers were stage I (two were borderline tumors), one was stage II, and two were stage III. Dørum et al21 screened 845 women. Sixteen cancers developed in the study period. It was not clear exactly how many were detected at screening. Ten women had increased serum CA-125 estimations but TVU results were not reported. An earlier publication of their results stated that seven of nine cancers were detected at screening, all with abnormal TVUs and four with increased serum CA-125 levels.22 Laframoise et al18 screened 1,342 women once every 6 months. Only one cancer developed during the study period. This was a stage I cancer detected by TVU. Overall, these studies have not provided consistent evidence that screening is effective in detecting ovarian cancer at an early stage in high-risk groups. In contrast to most of these reports, our study has demonstrated cancers mainly at follow-up rather than in the first (prevalent) round. This probably reflects the prolonged period of regular surveillance undergone by our study population, and provides direct evidence of the inability of current screening modalities to detect ovarian cancer early in its evolution. Previous studies have included borderline ovarian cancers in their assessment of effectiveness. In the present study, the single example of a borderline cancer was found in a member of a family with BRCA1 mutation. When the mutation was identified and the ovarian tumor patient opted for testing, it was established that she was a noncarrier. It is fairly clear that borderline tumors are a separate entity from invasive epithelial cancer with low rates of progression to invasive disease, and there is good evidence that borderline tumors are not part of the BRCA1 or BRCA2 spectrum.23,24 Given that, by definition, borderline tumors are of low malignant potential, they may be present for many years and therefore form an undue proportion of those detected in any screening program, particularly in the first screening round, although their detection is of little benefit to the patient.8 Including such cases will artificially boost the impression of detection at early stage within screening programs. The lack of such tumors in our program as a whole is evidence that they are not a significant factor in high-risk screening. If borderline tumors are excluded from previous studies, only one of four (25%) ovarian cancers were detected at stage I in the Kings study,15 only one of four (25%) were detected in the study by Karlan et al,20 and two of 12 (16%) were detected in the Norwegian study.21 Although nine (75%) of the invasive ovarian cancers in our study were detected by screening, the objective of screening (ie, detection of cancer at an early stage to reduce mortality) was not achieved. Excluding the borderline tumor, only two (16.7%) of the ovarian cancers that occurred were detected at stage I. The cancers that were not detected by screening were all FIGO stage III and IV. This is consistent with the combined evidence of previous studies reviewed by Hogg et al,25 who concluded that most ovarian cancers in women at high genetic risk are high-grade serous cancers and these are detected infrequently by screening at an early stage. Our cohort has been subdivided according to prior assessment of risk level and in most previous studies all of our cohort would have been classified as high risk. It is noteworthy that nine of the 13 cancers occurred in families with identified BRCA1 mutations, and that only one of the others was a member of a moderate-risk family. In contrast, equivocal TVU or CA-125 results, requiring recall and repeat testing, were as frequent in the moderate- as the high-risk cohort. Both ultrasound appearances of the ovaries and serum CA-125 levels are more difficult to interpret in premenopausal women26 and, as anticipated, most equivocal and false-positive results, including those that led to diagnostic surgery, occurred in pre- or perimenopausal patients. Conversely, eight of the 12 cancers were diagnosed in postmenopausal women. Thus, young (premenopausal) women were particularly poorly served by the surveillance program, and they experience most of the negative consequences (anxiety through equivocal results, unnecessary surgical intervention) but gain nothing in terms of early diagnosis of cancer. Few studies have been able to present sensitivity and specificity levels for ovarian screening because follow-up has not been undertaken to determine whether the women who had negative screening subsequently developed ovarian cancer. In our cohort it was not possible to calculate the combined sensitivity and specificity of ultrasound and CA-125 because not all of the women had both. In Manchester, women originally were offered ultrasound only; in Edinburgh, it was noted that some women seemed to opt out of ultrasound in the early days of the screening program. However, if we look at the modalities separately, CA-125 performed better than ultrasound, with a positive predictive value of 63% compared with 17%. All but one of the invasive ovarian malignancies detected by CA-125 were at stage II or greater, so although CA-125 was detecting cancer before symptoms developed, it was not detecting them before spread from the ovaries. Even allowing for wide CIs, the performance of ultrasound does not satisfy the WHO principles for screening.27 The alternative to screening is prophylactic salpingo-oophorectomy. This is not a trivial procedure and carries a complication rate estimated at approximately 1%.26 Nevertheless, two large studies have shown that it is highly effective in reducing risk of subsequent cancer of epithelial ovarian type, despite the recognized occurrence of primary peritoneal tumors in carriers of BRCA1 or BRCA2 mutations.28,29 Furthermore, if undertaken premenopausally, it has been found to reduce the risk of subsequent breast cancer by approximately 50%, and this protection does not seem to be abrogated by administering hormone replacement therapy after surgery to control menopausal symptoms.28,29 Whether the procedure is justified in women who have, at worst, a 50% chance of carrying a high-risk mutation (where a mutation has yet to be identified in the family or where the patient does not wish to be tested) remains a moot point; however, one serious consideration must be the frequency with which an occult carcinoma is discovered at prophylactic surgery.23,28,29 Among women attending the centers participating in this study, one patient had a stage III carcinoma of ovary discovered at prophylactic oophorectomy 2 months after a normal screening, whereas another patient, who opted for oophorectomy rather than screening, had a primary epithelial cancer of the fallopian tube with microscopic spread to both ovaries. Similar experiences have been reported anecdotally by most clinics providing services for familial ovarian cancer. Undoubtedly prophylactic oophorectomy is the only intervention proven, as yet, to reduce ovarian cancer death in high-risk populations,28,29 but multicenter studies have shown that many women are reluctant to contemplate this procedure30 in part, at least, because of unwarranted confidence in the efficacy of ovarian screening programs.31 In conclusion, annual surveillance by TVU scanning and serum CA-125 measurement in women at increased familial risk of ovarian cancer is ineffective in detecting ovarian malignancy before spread from the ovaries. The implication must be that annual surveillance has no impact on morbidity or mortality unless the relationship between stage at presentation and prognosis differs for familial and sporadic ovarian cancer. Current evidence suggests that outcome in BRCA1/BRCA2 ovarian malignancy may be equivalent or better than that in non-BRCA1/BRCA2,32,33 suggesting that high stage may not equate to early death. However, overall survival was still low in two pathology-based samples of Ashkenazi Jewish women.34,35 Indeed, four women had already died within 3 years of diagnosis in our own study. Given our findings that most ovarian cancers in women at high genetic risk have elevated CA-125 at presentation but that they can present at a relatively advanced stage within 12 months of a negative screening, there is a case for evaluating surveillance (within observational studies) by more frequent CA-125 measurement in women who choose not to have their ovaries removed. However, they must be made aware of the limitations of screening so that the decisions they make are fully informed.
The authors indicated no potential conflicts of interest.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Bell R, Petticrew M, Luengo S, et al: Screening for ovarian cancer: A systematic review. Health Technol Assess 2:1-84, 1998 2. Chappius PO, Foulkes WD: Overview of the clinical genetics of ovarian cancer, in Morrison PJ, Hodgson SV, Haites NE (eds): Familial Breast and Ovarian Cancer: Genetics, Screening and Management. Cambridge, United Kingdom, Cambridge University Press, 2002, pp 43-72 3. Scottish Intercollegiate Guidelines Network (SIGN): Epithelial Ovarian Cancer. Edinburgh, Scotland, SIGN (75th edition), 2003 4. Office for National Statistics. Cancer Statistics: Registrations. Series MBI No. 24. London, United Kingdom, Her Majesty’s Stationery Office, 1997 5. Lynch HT, Bewtra C, Wells IC, et al: Hereditary ovarian cancer, in Lynch HT, Kullander S (eds): Cancer Genetics in Women. Boca Raton, CRC Press, 1987, 49-97 6. Kristensen GB, Trope C: Epithelial ovarian cancer. Lancet 349:113-117, 1997[CrossRef][Medline] 7. Friedlander ML: Prognostic factors in ovarian cancer. Semin Oncol 25:305-314, 1998[Medline] 8. Stratton JF, Pharoah P, Smith SK, et al: A systematic review and meta-analysis of family history and risk associated with ovarian cancer. Br J Obstet Gynaecol 105:493-499, 1998[Medline]
9. Thompson D, Easton DF: Cancer incidence in BRCA1 mutation carriers. J Natl Cancer Inst 94:1358-1365, 2002 10. Ford D, Easton DF, Stratton M, et al: Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families: The Breast Cancer Linkage Consortium. Am J Hum Genet 62:676-689, 1998[CrossRef][Medline] 11. Aarnio M, Sankila R, Pukkala E, et al: Cancer risk in mutation carriers of DNA-mismatch-repair genes. Int J Cancer 81:214-218, 1999[CrossRef][Medline] 12. Brinkmann D, Jacobs I: Management of familial ovarian cancer, in Morrison PJ, Hodgson SV, Haites NE (eds): Familial Breast and Ovarian Cancer: Genetics, Screening and Management. Cambridge, United Kingdom, Cambridge University Press, 2002, pp 275-285 13. Macmillan RD: Screening women with a family history of breast cancer: Results from the British familial breast cancer group. Eur J Surg Oncol 26:149-152, 2000[CrossRef][Medline] 14. Moller P, Borg A, Evans DG, et al: Survival in prospectively ascertained familial breast cancer: Analysis of a series stratified by tumour characteristics, BRCA mutations and oophorectomy. Int J Cancer 101:555-559, 2002[CrossRef][Medline] 15. Tailor A, Bourne TH, Campbell S, et al: Results from an ultrasound-based familial ovarian cancer screening clinic: A 10-year observational study. Ultrasound Obstet Gynecol 21:378-385, 2003[CrossRef][Medline]
16. Leide A, Karlan BY, Baldwin RL, et al: Cancer incidence in a population of Jewish women at risk of ovarian cancer. J Clin Oncol 20:1570-1577, 2002
17. Scheuer L, Kauff N, Robinson M, et al: Outcome of preventative surgery and screening for breast and ovarian cancer in BRCA mutation carriers. J Clin Oncol 20:1260-1269, 2002 18. Laframboise S, Nedelcu R, Murphy J, et al: Use of CA125 and ultrasound in high-risk women. Int J Gynecol Cancer 12:86-91, 2002[CrossRef][Medline] 19. Taylor L, Schwartz PE: Cancer screening in a high risk population: A clinical trial. Ultrasound Med Biol 27:461-466, 2001[CrossRef][Medline] 20. Karlan BY, Baldwin RL, Lopez-Luevanos E, et al: Peritoneal serous papillary carcinoma, a phenotypic variant of familial ovarian cancer: Implications for ovarian cancer screening. Am J Obstet Gynecol 180:917-928, 1999[CrossRef][Medline] 21. Dørum A, Heimdal K, Lovslett K, et al: Prospectively detected cancer in familial breast/ovarian cancer screening. Acta Obstet Gynecol Scand 78:906-911, 1999[CrossRef][Medline] 22. Dørum A, Kristensen GB, Abeler VM, et al: Early detection of familial ovarian cancer. Eur J Cancer 32:1645-1651, 1996[CrossRef] 23. Stratton JF, Thompson D, Bobrow L, et al: The genetic epidemiology of early-onset epithelial ovarian cancer: A population-based study. Am J Hum Genet 65:1725-1732, 1999[CrossRef][Medline] 24. Werness BA, Ramus SJ, Whittemore AS, et al: Histopathology of familial ovarian tumors in women from families with and without germline BRCA1 mutations. Hum Pathol 31:1420-1424, 2000[CrossRef][Medline]
25. Hogg R, Freidlander M: Biology of epithelial ovarian cancer: Implications for screening women at high genetic risk. J Clin Oncol 22:1315-1327, 2004 26. Rufford B, Menon U, Jacobs I: Screening for familial ovarian cancer, in Morrison PJ, Hodgson SV, Haites NE (eds): Familial Breast and Ovarian Cancer: Genetics, Screening and Management. Cambridge, United Kingdom, Cambridge University Press, 2002, pp 220-233 27. Wilson JMG, Junger G. Principle and practice of screening for disease. Public Health Papers 34. Geneva, Switzerland, World Health Organisation, 1968
28. Kauff ND, Satagopan JM, Robson ME, et al: Risk-reducing salpingo-oophorectomy in women with a BRCA1 or BRCA2 mutation. N Engl J Med 346:1609-1615, 2002
29. Rebbeck TR, Lynch HT, Neuhausen SL, et al: Prophylactic oophorectomy in carriers of BRCA1 or BRCA2 mutations. N Engl J Med 346:1616-1621, 2002 30. Gynecology Oncology Group (USA). Protocol GOG-0199. Philadelphia, PA, Gynecology Oncology Group, 2003 31. Sheppard R, Fry A, Rush R, et al: Women at risk of ovarian cancer: Attitudes towards and expectations of the familial ovarian cancer clinic. Fam Cancer 1:31-37, 2001[CrossRef][Medline]
32. Boyd J, Sonoda Y, Federici MG, et al: Clinicopathologic features of BRCA-linked and sporadic ovarian cancer. JAMA 283:2260-2265, 2000
33. McGuire V, Whittemore AS, Norris R, et al: Survival in epithelial ovarian cancer patients with prior breast cancer. Am J Epidemiol 152:528-532, 2000 34. Cass I, Baldwin RL, Varkey T, et al: Improved survival in women with BRCA-associated ovarian cancer. Cancer 97:2187-2195, 2003[CrossRef][Medline]
35. Ben David Y, Chetrit A, Hirsh-Yechezkel G, et al: Effect of BRCA mutations on the length of survival in epithelial tumors. J Clin Oncol 20:463-466, 2002 Submitted May 18, 2004; accepted January 19, 2005.
Related Editorial
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
