|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 24 (August 20), 2005: pp. 5443-5445 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.03.909
Screening for Familial Ovarian Cancer: The Need for Well-Designed Prospective StudiesDepartment of Gynaecological Oncology, Institute for Women's Health, University College London, London, United Kingdom The issue of ovarian cancer screening (OCS) among women with a strong family history or genetic predisposition has a tendency to generate more heat than light. The report by Stirling et al1 continues that trend. The authors report a cohort of 1,110 women screened in three centers from 1991 to 2004, with CA-125 and/or ultrasound (U/S). Unfortunately, the article does not provide information about compliance with screening, the screen interval actually achieved, the detailed screening protocols used for U/S and CA-125 (which apparently varied over time and possibly among centers), the interval between screen-positive results and surgical intervention, or the method of follow-up. Despite these caveats and the small number of cancers (n = 13), the authors were sufficiently confident in their results to state that "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." They may be correct, but it is not possible to reach this conclusion from the data they have presented. The impression left after reading this article and the other literature on this topic is that there is a need for well-designed, rigorously conducted, and properly funded prospective trials of OCS among women at high risk. The identification of the BRCA1 and BRCA2 genes was an important step in refining risk assessment in families with a high incidence of breast and ovarian cancer (OC). However, the challenge of managing the risk of cancer remained. One option for OC is prophylactic surgery to remove the ovaries and tubes.2 This can be performed laparoscopically in the majority of cases with low morbidity, and is an effective method for prevention provided the ovaries are entirely removed and an occult malignancy is excluded.3,4 Oophorectomy does not eliminate primary peritoneal cancer but seems to reduce the risk of breast cancer. Although oophorectomy is an effective preventative measure, many women at increased risk of OC do not wish to take this step for a range of reasons. The alternative to oophorectomy for women at risk is OCS with the objective of intervening sufficiently early in the natural history of the disease to reduce mortality. There currently are two methods available for OCS: U/S to measure ovarian morphology, size, and blood flow, and measurement of the serum tumor marker CA-125.5 Given the markedly increased incidence of OC in high-risk families, it may have been anticipated that this would be the ideal group to include in definitive clinical trials. Unfortunately, that has not proven to be the case. The widespread provision of screening on a clinical basis for the high-risk group has made it difficult to implement rigorous prospective research protocols in this population. There are data available about the performance of OCS in the high-risk group from more than 10 reports of screening using CA-125 and/or U/S involving more than 6,000 women.6-16 Unfortunately, criteria for interpreting the test results vary from study to study, the screening protocols are not always clearly reported, and the completeness of follow-up is variable, making it difficult to reach firm or sound conclusions. Hogg and Friedlander17 reviewed the literature in 2004 and reported the detection of 38 primary invasive cancers, of which 14 (38%) were stage I or II. However, they included seven patients with OC reported twice in separate publications. When a correction is made for this, 14 of 31 patients detected were stage I or II (45%). They also noted 15 "interval" patient cases but did not relate this to the screening interval. When corrected for patients whose disease occurred within the screening interval, rather than later, the number of interval OCs in the publications reviewed was only two, giving an apparent sensitivity for detection of OC at 1-year follow-up of 94% (31 of 33). However, this is subject to the important caveat of incomplete reporting of follow-up. Hogg and Friedlander's key concern was that OCS might be detecting good prognostic histologic subtypes at an early stage but not the poor-prognostic group. They noted the lack of reliable data in the high-risk population to address this issue, and advocated prospective trials in women at increased genetic risk. The article by Stirling et al1 illustrates the challenges and limitations of many studies in this field. The findings do not seem to have been acquired from prospective implementation of a well-defined screening protocol. The repeat rates for CA-125 and U/S varied between centers from 1% to 13%, suggesting major differences in the way in which screens were performed or interpreted. The criteria for interpretation of CA-125 changed over time, different assays were used, and in one center, CA-125 was not included as a primary screen for more than half of the study duration. The preferred route for U/S was transvaginal, but the proportion of abdominal and transvaginal scans was not provided. This information is critical given the better resolution achieved by the transvaginal approach. Although the intended screening frequency was annual, no information is provided about the compliance with screening and the actual screening interval may have been longer. A total of 1,110 women had screenings during what seems to have been a 13-year period (1991 to 2004). Assuming steady recruitment, one would expect an average of at least six annual screenings per woman, giving a total of more than 6,000 annual screenings; however, only 3,701 scans and 2,617 CA-125 tests were performed.
With regard to the false-positive rate of OCS, Stirling et al1 believed that U/S was unsatisfactory because of its positive predictive value (PPV). The minimum acceptable PPV in OCS is generally regarded as The stage distribution of screening-detected cases is difficult to interpret without information about the interval between the screen result and surgical intervention. If this interval were prolonged, it would minimize the impact of screening on stage distribution. It is also noteworthy that until 1999, the Manchester center performed a CA-125 only if the U/S was abnormal, so there was no opportunity for CA-125 to detect patient cases independently at an early stage. The fact that all five patient cases detected in Manchester had just one CA-125 result suggests that these were the only CA-125 estimations and followed an abnormal U/S. Indeed, seven of 10 screening-detected patient cases had just one CA-125, suggesting that this, in effect, was a prevalence screening with CA-125. It is also unclear why there was no earlier intervention in the three patients with serial elevated CA-125 levels. Several of these factors would tend to minimize any stage shift, but nevertheless, 50% of the 10 screening-detected patient cases were stage I and II. A greater impact may be possible with well-defined and rigorously applied approaches to interpretation of U/S and CA-125 (such as the risk of ovarian cancer algorithm),18 and the potential of the field of serum proteomics.19 Certainly, it is not possible to conclude from this report that OCS cannot achieve a stage shift. There is a need for rigorous prospective studies of screening in the high-risk group using well-defined screening protocols, with high compliance and thorough follow-up. There are two multicenter, prospective, nonrandomized studies of OCS in the high-risk population in progress. In the United States, the Cancer Genetics Network study has involved more than 2,200 participants for 3 years in a study evaluating CA-125 measurements every 3 months using the risk of ovarian cancer algorithm.18 In parallel, the UK Familial Cancer Screening Study, after a slow start, now has more than 1,500 participants being screened annually with CA-125 and U/S, with serial serum samples collected every 4 months. Ultimately, the gold standard for assessing the impact of OCS is a randomized controlled trial (RCT) with mortality as an end point, but this has not been believed to be feasible or acceptable in the high-risk population. The only RCT of OCS performed to date was in the general population and demonstrated an improvement in survival in the screening arm.20 Larger RCTs to document the impact of screening on mortality are underway in the general population. The UK Collaborative Trial of Ovarian Cancer Screening (www.ukctocs.org), involving 200,000 women age 50 to 74 years from the general population, has nearly finished recruiting. OCS is also a component of the Prostate, Lung, Colon, and Ovary study being conducted by the National Institutes of Health, which involves 74,000 women.21 For now, women at high risk have to make decisions about managing their risk without the benefit of data from optimally performed trials. Women who have completed their families and are seeking a high degree of certainty will often opt for prophylactic salpingo-oophorectomy. Screening remains an important option for women who wish to defer oophorectomy for reasons of fertility or age, or who have a strong preference to retain their ovaries. Some women who understand the uncertainties of OCS still prefer this approach to undergoing a surgical procedure. So, how should women at high risk be counseled about the pros and cons of screening? Women considering screening should be informed about a number of key points. First, there is no definitive evidence about whether screening will (or will not) reduce mortality from OC. Until the large randomized trials in progress in the general population are complete, the situation will remain unclear. Even then the applicability of general population data to the high-risk group will be subject to debate, given the possibility of differences in the tumor biology of cancers in the high-risk group. Second, women should be informed that neither CA-125 nor U/S will detect all occurrences of cancer, and they should be told unequivocally that screening will miss some occurrences of OC as well as most occurrences of primary peritoneal cancer. Third, all women considering screening should be informed of the significant false-positive rate (higher in premenopausal women) for both U/S and CA-125, and the possibility that this can lead to anxiety and sometimes unnecessary surgery. Depending on the screening modality and protocol, up to eight women will undergo surgery for each woman with a screening-detected OC, and a small proportion will experience surgical complications. Finally, there is one positive aspect of screening that frequently is not mentioned. Screening can provide a degree of short-term reassurance, which women attending for OCS may value. In the report by Stirling et al,1 there were only three cancers in the 1,071 women who had negative screening results, compared with 10 in the 39 women with abnormal results and 13 in the entire population. The relative risk of OC for women with normal screening results compared with the entire population was 0.24 [(3/1,071)/(13/1,110)]. This is consistent with data from general population screening.22 Women can be told that although a negative screening does not exclude OC, it is associated with a reduced likelihood of a diagnosis of OC in the interval between screenings. In summary, assessing the impact of OCS is challenging and will not be satisfactorily achieved by retrospective analysis of clinical practice. The results of carefully designed prospective trials of OCS in the high-risk group and large RCTs underway in the general population are awaited. The information from these trials will shed light on the ability of screening to alter the natural history of OC and to reduce mortality from the disease. Author's Disclosures of Potential Conflicts of Interest The author indicated no potential conflicts of interest. REFERENCES
1. Stirling D, Evans DGR, Pichert G, et al: 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 system. J Clin Oncol 23:5588-5596, 2005
2. van Roosmalen MS, Verhoef LC, Stalmeier PF, et al: Decision analysis of prophylactic surgery or screening for BRCA1 mutation carriers: A more prominent role for oophorectomy. J Clin Oncol 20:2092-2100, 2002
3. 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
4. Rebbeck TR, Lynch HT, Neuhausen SL, et al: Prophylactic oophorectomy in carriers of BRCA1 or BRCA2 mutations. N Engl J Med 346:1616-1622, 2002
5. Jacobs IJ, Menon U: Progress and challenges in screening for early detection of ovarian cancer. Mol Cell Proteomics 3:355-366, 2004 6. Bourne TH, Campbell S, Reynolds K, et al: The potential role of serum CA 125 in an ultrasound-based screening program for familial ovarian cancer. Gynecol Oncol 52:379-385, 1994[CrossRef][Medline] 7. Weiner Z, Beck D, Shteiner M, et al: Screening for ovarian cancer in women with breast cancer with transvaginal sonography and color flow imaging. J Ultrasound Med 12:387-393, 1993[Abstract] 8. Muto MG, Cramer DW, Brown DL, et al: Screening for ovarian cancer: The preliminary experience of a familial ovarian cancer center. Gynecol Oncol 51:12-20, 1993[CrossRef][Medline] 9. Schwartz PE, Chambers JT, Taylor KJ: Early detection and screening for ovarian cancer. J Cell Biochem 23:233-237, 1995 10. Belinson JL, Okin C, Casey G, et al: The familial ovarian cancer registry: Progress report. Cleve Clin J Med 62:129-134, 1995[Medline] 11. Menkiszak J, Jakubowska A, Gronwald J, et al: Hereditary ovarian cancer: Summary of 5 years of experience [in Polish]. Ginekol Pol 69:283-287, 1998[Medline] 12. Karlan BY, Raffel LJ, Crvenkovic G, et al: A multidisciplinary approach to the early detection of ovarian carcinoma: Rationale, protocol design, and early results. Am J Obstet Gynecol 169:494-501, 1993[Medline] 13. 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] 14. Dorum A, Kristensen GB, Abeler VM, et al: Early detection of familial ovarian cancer. Eur J Cancer 32A:1645-1651, 1996[CrossRef] 15. Dorum 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]
16. Scheuer L, Kauff N, Robson M, et al: Outcome of preventive surgery and screening for breast and ovarian cancer in BRCA mutation carriers. J Clin Oncol 20:1260-1268, 2002
17. Hogg R, Friedlander M: Biology of epithelial ovarian cancer: Implications for screening women at high genetic risk. J Clin Oncol 22:1315-1327, 2004 18. Skates SJ, Menon U, MacDonald N, et al: Calculation of the Risk of ovarian cancer from serial CA-125 values for preclinical detection in postmenopausal women. J Clin Oncol 21:206-210, 2003 (10 suppl) 19. Petricoin EF, Ardekani AM, Hitt BA, et al: Use of proteomic patterns in serum to identify ovarian cancer. Lancet 359:572-577, 2002[CrossRef][Medline] 20. Jacobs IJ, Skates SJ, Macdonald N, et al: Screening for ovarian cancer: A pilot randomized control trial. Lancet 353:1207-1210, 1999[CrossRef][Medline] 21. Prorok PC, Andriole GL, Bresalier RS, et al: Design of the Prostate, Lung, Colorectal and Ovarian (PLCO) cancer screening trial. Control Clin Trials 21:273S-309S, 2000[CrossRef][Medline]
22. Jacobs IJ, Skates S, Davies AP, et al: Risk of diagnosis of ovarian cancer after raised serum CA 125 concentration: A prospective cohort study. BMJ 313:1355-1358, 1996 Related Article
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
|
10% (ie, no more than 10 surgical procedures for each case of OC detected). As Stirling et al showed, this is more easily achievable in the high-risk population than in general population screening given the increased incidence of OC. The PPV for U/S was 17% and the overall PPV achieved was 25% (10 of 39). Surprisingly, only five patients with benign disease proceeded to surgery on the basis of an elevated CA-125, producing a PPV for CA-125 of 63%, which is unusually high in the context of OCS. This raises the possibility that there was a reluctance to intervene on the basis of CA-125 elevation, which may have contributed to delays in intervention and late-stage distribution at diagnosis. The number of OCs reported by Stirling et al was too small to yield a reliable estimate of sensitivity, as reflected in the wide confidence limits (19% to 73% for U/S and 59% to 100% for CA-125). 
