|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Status of Tumor Markers in Ovarian Cancer ScreeningFrom the University of Texas M.D. Anderson Cancer Center, Houston, TX. Address reprint requests to Robert C. Bast, Jr, MD, University of Texas M.D. Anderson Cancer Center, Box 355, 1515 Holcombe Blvd, Houston, TX 77030; email: rbast{at}mdanderson.org.
One of the most promising approaches to managment of ovarian cancer is early detection. Stage I ovarian cancer can be cured with currently available therapy in more than 90% of patients. However, fewer than 25% of ovarian cancers are currently detected in stage I. Detection of a greater fraction of cancers at an early stage might improve clinical outcome. Given a prevalence of one patient with ovarian cancer among 2,500 asymptomatic postmenopausal women in the general population, a successful screening strategy must have a sensitivity of more than 75% and a specificity of more than 99.6% to achieve a positive predictive value of 10%. Approaches to screening include transvaginal sonography, serum markers, and two-stage strategies that use alterations in serum markers to prompt sonographic examination. Among the serum markers, CA-125 has been studied most extensively. Isolated values of CA-125 lack adequate sensitivity or specificity, but when monitored over time, serial CA-125 values can achieve a specificity of 99.6%. However, sensitivity is limited and CA-125 may only be expressed by 80% of early-stage cancers. Multiple markers may exhibit greater specificity when studied over time. To combine multiple markers, more sophisticated mathematical analysis will be required. At present, screening women at conventional risk should be restricted to clinical trials. In the future, however, screening for ovarian cancer may reduce the morbidity and mortality of this disease.
THIS YEAR in the United States approximately 23,300 women will be diagnosed as having ovarian cancer and 13,900 will die from this malignancy.1 Despite advances in the management of advanced ovarian cancer, 70% to 80% of patients will ultimately succumb to disease that is diagnosed in late stages. When ovarian cancer is diagnosed in stage I, more than 90% of patients can be cured with conventional surgery and chemotherapy. At present, however, only 25% of ovarian cancers are detected in stage I. Detection of a greater fraction of ovarian cancers in early stage might significantly affect survival.
The ultimate success of any screening strategy depends on the clinical biology of the disease. For effective screening of ovarian cancer, most tumors should arise from single clones of cells in the ovary rather than from multiple foci throughout the abdominal cavity. Our group and others2–4 have shown that more than 90% of sporadic ovarian cancers are clonal. When primary ovarian masses and peritoneal metastases have been compared, the same X chromosome is inactivated, the same patterns of loss of heterozygosity are present, and identical p53 mutations are found at both sites in more than 90% of patients. Screening also depends on the assumption that advanced metastatic disease arises from clinically detectable stage I lesions. From five to seven genetic alterations may be required to transform normal ovarian epithelial cells into a cancer. Given the heterogeneous behavior of ovarian cancer, neoplasms that arise in different patients might exhibit different genotypic alterations with different patterns of oncogene activation and suppressor gene loss. Cancers that are currently diagnosed in stage I might exhibit a distinct pattern of genetic abnormalities that permits growth and invasion but not metastasis. Stage III disease might exhibit a completely different genotype and phenotype. Conversely, recent results from expression array analysis have demonstrated a similar pattern of abnormalities in poorly differentiated stage I and stage III cancers,5 consistent with the possibility that stage I ovarian cancer is, in fact, the precursor of advanced disease. Finally, a sufficient interval must exist between the development of early-stage cancer and metastasis to permit screening at practical intervals. Skates and Singer6 developed a stochastic model for annual screening to detect ovarian cancer. Their analysis indicates that stage shifts (largely from stage III to stage I) achieved by effective screening could produce 3.4 ± 0.1 years of life saved per patient, provided the disease remained in stage I for 9 months and the interval from tumor inception to clinical detection was at least 1.3 years. The benefit of screening depended critically on the mean duration the disease was in stage I and the coefficient of variation of this mean. To estimate the duration of preclinical disease, Skates et al7 analyzed CA-125 values from 28 patients with ovarian cancer that had been detected during a longitudinal screening trial conducted with 22,000 women in the United Kingdom.8 An exponential increase in CA-125 with the growth of ovarian cancer was assumed, and serial CA-125 values were fitted using a longitudinal change point model to estimate the interval from tumor inception to clinical detection. The mean estimated duration of preclinical ovarian cancer was 1.9 ± 0.4 years.
In addition to biologic requirements for effective screening, early detection of ovarian cancer requires a high sensitivity and a particularly high specificity to attain an acceptable positive predictive value. Given the prevalence of one patient with epithelial ovarian cancer among 2,500 postmenopausal women in the United States, a screening strategy must have sensitivity greater than 75% and specificity greater than 99.6% to attain a positive predictive value of 10%; that is, 10 laparotomies for each patient with ovarian cancer detected. Several approaches have been evaluated for detecting epithelial ovarian cancer, including ultrasonography, serum markers, and a two-stage strategy in which increasing serum markers prompt sonography.
Early studies of sonography used transabdominal ultrasonography (TAU). Campbell et al9 studied 5,479 women with TAU and performed 326 operations to detect five ovarian cancers, achieving a positive predictive value of only 1.5%. Of interest, however, was that all five patients had stage I disease. Transvaginal sonography (TVS) can provide a more precise image of the ovary. Subsequent trials of TVS have been conducted in the United Kingdom, United States, and Japan.10–12 If the three major studies are considered together, approximately 66,620 women have been screened, prompting 565 operations to detect 45 ovarian cancers, 34 of which were invasive. Approximately 35 of the 45 borderline and invasive cancers (78%) were in stage I. Studies from the University of Kentucky11 achieved a positive predictive value of 9.9%, nearly equaling the goal of 10 operations for each patient with ovarian cancer detected. Overall, specificity for the major trials is at the margin of that required to achieve a positive predictive value of 10%. In addition, the sensitivity of TVS for detecting stage I ovarian cancer may not exceed 90%, particularly when screening prevalent disease.
The use of serum markers for early detection has largely focused on CA-125, a high-molecular-weight mucin (MUC 16) that was initially detected with a homologous double-determinant radioimmunoassay.13 The OC-125 murine monoclonal antibody14 was initially used to bind CA-125 antigen from serum on a bead. Because multiple epitopes are present on each CA-125 molecule, the same OC-125 antibody labeled with iodine-125 could be used to detect identical determinants on CA-125 molecules that had been bound. Overall, the CA-125 assay exhibits a sensitivity of 50% to 60% for stage I disease.15–17 Antigen levels can increase exponentially 10 to 21 months before diagnosis.7,18–21 Specificity of CA-125 is inadequate for screening, particularly in a premenopausal population in which endometriosis, adenomyosis, and retrograde menstruation can produce false-positive elevations of antigen levels. Specificity can, however, be improved by combining CA-125 with ultrasonography in a two-stage strategy and by sequential monitoring of CA-125 values over time. To date, the largest trial to use CA-125 was conducted by Jacobs et al8 in the United Kingdom. Postmenopausal women older than 45 years were randomly assigned to a control group (10,977 patients) or a screened group (10,985 patients). Three annual screenings were performed with CA-125. If CA-125 levels were greater than 30 U/mL, TAU was performed. When TAU results were abnormal, surgery was undertaken. Among 10,985 women screened, 29 operations were performed to detect six cancers, providing a positive predictive value of 21%. During 7 years of follow-up, 10 more cancers were diagnosed in the screened group. During the same intervals, 21 ovarian cancers were diagnosed in the control group. Median survival in the screened group (72.9 months) was significantly greater (P = .0112) than that in the control group (41.8 months). Subsequent improvements in the CA-125 assay facilitated monitoring CA-125 levels over time with less day-to-day variation. The M11 antibody developed by O’Brien et al22 detects a distinct epitope on the CA-125 (MUC 16) molecule. In the CA-125 II assay, CA-125 antigen is trapped on a bead with the M11 antibody. Bound CA-125 antigen is then detected with iodine-125–labeled OC-125, producing an assay that has a day-to-day coefficient of variation of less than 5%. The CA-125 II assay was used to analyze specimens from a study conducted in Stockholm by Einhorn et al.20 Patients with ovarian cancer exhibited progressively increasing CA-125 levels, whereas patients with benign gynecologic conditions, nongynecologic disease, or no detectable illness had levels that remained constant over time, even when elevated. Skates et al23 analyzed changes in CA-125 over time, studying linear regression of the logs for CA-125 II values. When the slopes and intercepts where plotted for patients with ovarian cancer and for healthy individuals, the two groups could be distinguished with a specificity of 99.7% and an apparent sensitivity of 83%, yielding a positive predictive value of 16%. In a prospective trial, Rosenthal and Jacobs24 randomly assigned 10,000 volunteers older than 50 years to a screening group (5,046 patients) and a control group (4,954 patients). A risk of ovarian cancer (ROC) algorithm was used and 101 patients were found to be at sufficiently high risk to undergo TVS. Seventeen patients had abnormal findings that prompted operations that detected four patients with ovarian cancer, yielding a positive predictive value of 20%. Among the four ovarian cancers detected, one was borderline stage IA and three were invasive, with two in stage IC and one in stage II. At present, a trial is under way in the United Kingdom that will include 200,000 postmenopausal women who will be randomly assigned to three groups. A control group (100,000 patients) will be followed up with conventional pelvic examinations, a second group (50,000 patients) will undergo annual TVS, and a third group (50,000 patients) will have CA-125 levels determined at least annually. On the basis of the ROC algorithm, patients in the third group will be referred for TVS, surgery, or both. Women will be screened for 3 years and followed up for 7 years. This trial may demonstrate more definitively the feasibility of screening for ovarian cancer as well as the effect of early detection on survival. Use of the ROC algorithm may improve on the sensitivity anticipated with single values of CA-125. CA-125 levels are greater than 35 U/mL in 50% to 60% of patients with stage I ovarian cancers at the time of conventional diagnosis. The ROC algorithm could potentially detect disease when the CA-125 level was less than 35 U/mL, providing sensitivity in excess of 60%. CA-125 cannot be detected, however, in tissue sections from 20% of ovarian cancers. As a consequence, the sensitivity of a CA-125–based screening strategy cannot exceed 80%. Greater sensitivity might be achieved with multiple markers provided that specificity is not compromised.
During the last two decades, a large number of serum tumor markers have been evaluated for their ability to detect early-stage epithelial ovarian cancer (Table 1  ). Our group has evaluated a Lewis X mucin determinant (OVX1) and a cytokine macrophage colony-stimulating factor (M-CSF) for their ability to detect stage I ovarian cancer and to complement CA-125.25,26 Among 89 serum samples obtained from patients with stage I ovarian cancer before surgery,16,17 CA-125 was greater than 35 U/mL in 69%. A combination of CA-125, OVX1, and M-CSF detected 84% of early-stage cancers and specificity was decreased from 99% to 84%. From an update of a recent review,27 29 different serum tumor markers have been evaluated in combination with CA-125 and were reported to increase sensitivity and specificity. Markers generally have been analyzed only two or three at a time. When used in combination with CA-125, sensitivity generally has been increased by 5% to 10% with multiple markers at the cost of a substantial decrease in specificity. David Fishman in the Early Detection Research Network and Nicole Urban of the University of Washington Ovarian Specific Program of Research Excellence (Seattle, WA) have each circulated panels of serum samples to determine the role of multiple serum markers in detecting ovarian cancer. Our laboratory is undertaking immunohistochemical analysis of multiple marker expression in ovarian cancer sections that have been selected for low or absent expression of CA-125 to identify antigens that would complement CA-125 at a tissue level.
Several novel markers recently have been reported. These have been discovered through multiple approaches. Laboratory investigation of the biology of the epidermal growth factor receptor has observed that decreased expression of a 110-kd fragment of epidermal growth factor receptor is found in serum samples from ovarian cancer patients.28 Murine monoclonal antibodies have also been prepared against mesothelin, a molecule related to megakaryocyte potentiating factor.29 On the basis of receiver operating curves, mesothelin has been shown to complement CA-125 and to improve detection of ovarian cancer when the two tests are used in combination. In addition, mesothelin shows promise as a marker for mesothelioma (I. Hellstrom, personal communication and oral presentations, July 2002). Lipid markers may also prove useful.30 Lysophosphatidic acid levels are modestly but significantly increased in most patients with stage I ovarian cancer, and elevated lysophosphatidic acid levels have been reported to identify patients with normal CA-125 levels (G. Mills, personal communication, January 2003). Use of gene expression array analysis has identified a number of novel markers, including HE4,31,32 prostasin,33 and osteopontin.34 Yousef and Diamandis35 have cloned and cataloged the human kallikrein genes, some of which also have been detected by expression arrays. Molecular and biochemical approaches have been used to identify kallikrein 6 and 10 as potential serum markers.36,37 Recent reports have used surface-enhanced laser desorption and ionization (SELDI) to detect novel patterns of low-molecular-weight moieties in serum samples from patients with ovarian cancer. This proteomic technique has been reported to yield 100% sensitivity and 95% specificity with a positive predictive value of 94%.38 Some of the most useful SELDI data focus on low-molecular-weight moieties with charge-to-mass ratios of less than 2,500. Such moieties include not only oligopeptides but also lipids and carbohydrates. Multiple algorithms have been generated with different data sets, and it will be important to define which peaks are consistently predictive of early-stage disease. SELDI–time-of-flight or matrix-assisted laser desorption ionization mass spectrometry–time-of-flight analysis might also be used to identify markers for which conventional immunoassays could be developed. To facilitate the combination of multiple markers or multiple peaks, mathematical analysis will be critical. Several approaches to mathematical modeling seem promising. Neural network analysis has been applied to the combination of multiple markers. Zhang et al39,40 used an artificial neural network (ANN) to distinguish malignant from benign pelvic masses and to detect early-stage epithelial ovarian cancer. If we assume that multiple markers would be used as the first phase of a two-stage screening strategy, specificity might be set at 98% so that no more than 2% of women screened would undergo a second procedure (eg, TVS). At a fixed specificity of 98%, the ANN (including CA-125 II, CA-72–4, CA-15–3, and M-CSF) exhibited 72% sensitivity for early-stage disease, whereas CA-125 II alone had a sensitivity of 48%. A second approach, pioneered by Skates et al,41 used mixtures of multivariate normal distributions to analyze data at a fixed specificity of 98%. A combination of CA-125 II, CA-15–3, CA-72–4, and M-CSF produced a sensitivity of 75% for early-stage disease, resulting in a similar outcome to that attained with ANN.
At present, screening for ovarian cancer must be regarded as a work in progress. Routine screening is not recommended for women at conventional risk of developing ovarian cancer. Screening for women at conventional or increased risk should be conducted in the context of a clinical trial. Randomized trials will require 10 years or more to complete and may provide false-negative outcomes. The Prostate, Lung, Colon, and Ovary (PLCO) trial was designed primarily to detect prostate, lung, and colon cancer. As a consequence, it is underpowered to detect improved survival in women with ovarian cancer and may not achieve statistical significance. The Bart’s 3 trial in the United Kingdom and the PLCO trial in the United States leave therapy at the discretion of local physicians, who may not perform effective cytoreductive surgery or use consistent chemotherapy. Hereditary ovarian cancer may or may not be amenable to early detection using serum markers. Hereditary cancers are more frequently polyclonal. Primary peritoneal disease can still develop in a fraction of women who have undergone prophylactic oophorectomy.42 In several cases, extremely short intervals have been observed among a normal CA-125 value, a normal TVS, and the presentation of advanced ovarian or peritoneal cancer.43 Screening of high-risk patients must be undertaken in younger women. In this setting, serum markers such as CA-125 can have a higher fraction of false-positive levels than that observed in postmenopausal populations. Success in detecting sporadic ovarian cancers will depend on the biology of the disease. As discussed herein, we do not yet know how often stage III and IV disease evolves from clinically detectable stage I cancer. It remains to be determined whether there is a reasonable average interval of disease in potentially detectable stage I disease and what the coefficient of variation will be around this average. Finally, we do not know how often benign tumors evolve into invasive ovarian cancers. In the years ahead, it will be important to identify an optimal first step in a two-stage screening strategy.27 Improved imaging techniques might be developed. Multiple marker levels could be measured over time. SELDI may prove to be a superior technology. Each of these approaches may be enhanced by inclusion of risk factors in addition to age. We also will need to identify an optimal second-stage confirmatory strategy. This might include improved imaging with three-dimensional power Doppler,44 the so-called ovarian Papanicolaou smear,27 or a confirmatory panel of serum markers. The optimal interval for cost-effective screening must be determined. Finally, we must consider the possibility that ovarian cancer screening might be bundled with screening for other cancers to make screening more convenient and cost effective.
DR. CANNISTRA: What are your minimal criteria for being interested in a new serum marker in this disease? DR. BAST: One would seek markers that detect cases missed by CA-125 and that retain high specificity. Alternatively, a single novel marker that exhibited very high sensitivity for early-stage disease would also be of interest. If LPA really recognizes 80% or 90% of stage I disease prospectively, as some preliminary data suggest [G. B. Mills, personal communication], this marker would be of great interest provided that the specificity was also very high. DR. SKATES: I would add that the complementarity means coming at no cost to specificity. You’d complement the sensitivity, but you’d want to fix the specificity at the same mark that you normally use for CA-125. Therefore, if there’s 98% specificity for CA-125, work out how to combine CA-125 and the new marker, work out how to fix the specificity, and then show that the sensitivity has been increased beyond what you get just with CA-125, particularly, for early-stage disease. There’s always a caveat, though, that preoperative serum from early-stage disease is precisely the group of patients who you don’t need to detect in a screening program because they’re already going to be clinically identified as early stage. DR. BAST: We’ve just profiled 44 ovarian cancers with Affimetrix gene expression arrays and are using these data to identify coordinate elevation of different potential serum markers [Proc Am Assoc Cancer Res, 44:573, 2003 (abstr 2520)]. There are some ESTs that mark CA-125, although interestingly, they’re not as frequently and dramatically elevated at the message levels than we would have expected. It should be possible to look at which of the markers are coordinately elevated. DR. SKATES: The PLCO trial started out as a 16-year trial [Control Clin Trials 21:273S–309S, 2000]. The idea was to have 6 years of screening with a minimum follow-up of 10 years. It recently got extended to another 5 years, so it’s now a 21-year trial. There are two reasons for that. One is that 75,000 women were randomized to two arms. With that number and the low incidence of the disease, you need lots of screening years to have an impact and sufficient numbers to differentiate chance differences from real differences. However, the issue it raises is are we fixated so much on ovarian cancer mortality as the endpoint that by the time we see that endpoint what we started off evaluating, the CA-125, is no longer relevant. Should we be looking at endpoints that are shorter than that when it comes to ovarian cancer? I understand that in prostate cancer there is a question as to whether or not the increased distribution of early-stage disease that was observed is really reflective of any prostate cancer mortality decrease in the future. There doesn’t seem to be an excess of ovarian cancers detected in the screening arm compared to the control arm. In fact, if you look at the numbers, there were 16 detected cancers in the screened arm versus 20 in the control arm. So there doesn’t seem to be an overdiagnosis of indolent long-term ovarian cancer diseases in the randomized trial. Because of that, I don’t think overdiagnosis for ovarian cancer is a problem like it is for prostate cancer. Therefore, I think detection of early-stage disease with a sufficient positive predictive value may, in fact, for ovarian cancer at least, be a reasonable endpoint that would be seen much earlier than ovarian cancer mortality. DR. BOOKMAN: One of the critical issues is how many advanced tumors have a defined early stage that persists for a period of time. For example, knowing the structure of the ovary and knowing there’s a surface epithelium, it’s clear that that tumor can detach and implant and spread without ever invading. There’s no basement membrane or structure that keeps it confined to the ovary. We know that there’s a lot of dynamic activity within the peritoneal cavity. If something really is a surface lesion, it may have a different biology than something that develops within the ovary itself, like a cyst, that’s bound by stroma and basement membrane and has to physically invade and get out before it can spread. DR. BAST: The critical question is, how often does cancer arise from the surface of the ovary and how often do cancers develop within cysts? Dr. Jeff Boyd spoke last week at the International Gynecologic Cancer Society meeting in Seoul and updated his studies of ovaries removed prophylactically from women with BRCA-1 and BRCA-2 mutations at Memorial Sloan Kettering Cancer Center. Nineteen of 20 early-stage ovarian cancers that they identified were in cysts. It was also interesting that they were able to show LOH for the other BRCA-1 allele and p53 mutations in at least half of those cases. Our early work at Duke with Andy Berchuck suggested that p53 was a good marker for metastases. Of cancers that are in stage I at the time of detection, p53 was rarely mutated [J Natl Cancer Inst 84:1793–1798, 1992]. Cancers in stages IC, II, and III had overexpression of p53 in 40% or 50% of cases. If p53 is a marker of metastatic potential, about half of these early-stage cancers are ready to metastasize, even though they are quite small. There are likely to be genes in addition to p53 that are very important for that process. Mutation of p53 may be necessary but not sufficient. DR. SKATES: We looked at how CA-125 changed over time and looked at the inflection point until the time at which it was clinically detected. That had an average duration of 1.9 years. But the real question is what fraction of that time was it in stage I disease. The only way I could infer that was to see what the stage shift is from a large screening trial. Then by back inverting that in the stage shift that you get from how regularly you were screened, you could work out what the average duration of stage I disease is. However, until you do the large study, it’s hard to make that back inference. DR. CANNISTRA: Do you feel that a randomized study, using ovarian cancer mortality as the endpoint, will be necessary in the evaluation of screening modalities in this disease? DR. SKATES: If you had a randomized trial between a control group and a screen group, and you saw in the control group 80% of the cancers were in late-stage disease and 20% early stage, and in the screen group, you saw 80% early stage and 20% late stage, that would be convincing evidence. If you found that with no more than, for example, five operations per ovarian cancer, that would be fairly convincing evidence that there’s value to ovarian cancer screening, as long as there weren’t many more ovarian cancers in the screen group than in the control group. DR. CANNISTRA: In this example, are you potentially missing a useful screening strategy because you’re asking too much of the screening test? DR. SKATES: I’m giving you an extreme example where I can say this would be convincing to me even though the endpoint is not ovarian cancer mortality. I’m not saying that that’s the bar for that particular endpoint, but I’m thinking that most people would be convinced by that extreme. Eventually, you can get into the debate as to where the bar should be, but the first question of the debate is which bar is it. Is it the ovarian cancer mortality bar or is it the early-stage detection with PPV?
1. Jemal A, Thomas A, Murray T, Thun M: Cancer statistics, 2002. CA Cancer J Clin 52:23–47, 2002
2. Jacobs IJ, Kohler MF, Wiseman RW, et al: Clonal origin of epithelial ovarian cancer: Analysis by loss of heterozygosity, p53, mutation and chromosome inactivation. J Natl Cancer Inst 84:1793–1798, 1992 3. Mok CH, Tsao SW, Knapp RC, et al: Unifocal origin of advanced human epithelial ovarian cancer. Cancer Res 18:5119–5122, 1992 4. Li S, Han H, Resnik E, et al: Advanced ovarian carcinoma: Molecular evidence of unifocal origin. Gynecol Oncol 51:21–25, 1993[CrossRef][Medline]
5. Shridhar V, Lee J, Pandita A, et al: Genetic analysis of early versus late stage ovarian tumors. Cancer Res 61:5895–5904, 2001 6. Skates SJ, Singer DE: Quantifying the potential benefit of CA 125 screening for ovarian cancer. J Clin Epidemiol 44:365–380, 1991[CrossRef][Medline] 7. Skates SJ, Jacobs IJ, MacDonald N, et al: Estimated duration of pre-clinical ovarian cancer from longitudinal CA125 levels. Proc Am Assoc Cancer Res 43: 1999 (abstr 288) 8. Jacobs I, Stabile I, Bridges J, et al: Multimodal approach to screening for ovarian cancer. Lancet 1:268–271, 1999[CrossRef] 9. Campbell S, Bhan V, Royston P, et al: Transabdominal ultrasound screening for early ovarian cancer. Br Med J 299:1363–1367, 1989 10. Bourne TH, Campbell S, Reynolds KM, et al: Screening for early familial ovarian cancer with transvaginal ultrasonography and colour blood flow imaging. Br Med J 306:1025–1029, 1993 11. Van Nagell JR, Depriest PD, Reedy MB, et al: The efficacy of transvaginal sonographic screening in asymptomatic women at risk for ovarian cancer. Gynecol Oncol 77:350–356, 2000[CrossRef][Medline] 12. Sato S, Yokoyama Y, Sakamota T, et al: Usefulness of mass screening for ovarian carcinoma using transvaginal ultrasonography. Cancer 89:582–588, 2001 13. Bast RC Jr, Klug TL, St John E, et al: A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 309:883–887, 1983[Abstract] 14. Bast RC Jr, Feeney M, Lazarus H, et al: Reactivity of a monoclonal antibody with human ovarian carcinoma. J Clin Invest 681:1331–1337, 1981 15. Jacobs I, Bast RC Jr: CA125 tumour-associated antigen: A review of the literature. Hum Reprod 4:1–12, 1989
16. Woolas RP, Xu FJ, Jacobs IJ, et al: Elevation of multiple serum markers in patients with stage I ovarian cancer. J Natl Cancer Inst 85:1748–1751, 1993 17. Van Haaften-Day C, Xu F, Yu Y, et al: OVX1, CA 125 and M-CSF as tumor markers for surface epithelial-stromal tumors of the ovary: A critical appraisal. Cancer 92:2837–2844, 2001[CrossRef][Medline] 18. Bast RC Jr, Siegal FP, Runowicz C, et al: Elevation of serum CA 125 prior to diagnosis of an epithelial ovarian carcinoma. Gynecol Oncol 22:115–120, 1985[CrossRef][Medline] 19. Zurawski VR, Knapp RC, Einhorn N, et al: An initial analysis of preoperative serum CA 125 levels in patients with early stage ovarian carcinoma. Gynecol Oncol 30:7–14, 1988[CrossRef][Medline] 20. Einhorn N, Sjovall K, Knapp RC, et al: Prospective evaluation of serum CA 125 levels for early detection of ovarian cancer. Obstet Gynecol 80:14–18, 1992[Medline]
21. Jacobs IJ, Skates S, Davies AP, et al: Risk of diagnosis of ovarian cancer after raised serum CA 125 concentration: A prospective cohort study. Br Med J 313:1355–1358, 1996 22. O’Brien TJ, Raymond LM, Bannon GA, et al: New monoclonal antibodies identify the glycoprotein carrying the CA125 epitope. Am J Obstet Gynecol 165:1957–1964, 1991 23. Skates SJ, Xu F-J, Yu Y-H, et al: Toward an optimal algorithm for ovarian cancer screening with longitudinal tumor markers. Cancer 76:2004–2010, 1995[CrossRef][Medline] 24. Rosenthal AN, Jacobs IJ: The role of CA 125 in screening for ovarian cancer. Int J Biol Markers 13:216–220, 1998[Medline]
25. Xu F-J, Yu Y-A, Daly L, et al: OVX1 radioimmunoassay complements CA-125 for predicting the presence of residual ovarian carcinoma at second-look surgical surveillance procedures. J Clin Oncol 11:1506–1510, 1993 26. Xu FJ, Ramakrishnan S, Daly L, et al: Increased serum levels of macrophage colony-stimulating factor in ovarian cancer. Am J Obstet Gynecol 165:1356–1362, 1991[Medline] 27. Bast RC Jr, Urban N, Shridhar V, et al: Early detection of ovarian cancer: Promise and reality, in Stack MS, Fishman DA (eds): Ovarian Cancer. Boston, MA, Kluwer, 2002, pp 61–97
28. Baron AT, Lafky JM, Boardman CH, et al: Serum sErbB1 and epidermal growth factor levels as tumor biomarkers in women with stage III or IV epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev 8:129–137, 1999
29. Scholler N, Fu N, Yang Y, et al: Soluble member(s) of the mesothelin/megakaryocyte potentiating factor family are detectable in sera from patients with ovarian carcinoma. Proc Natl Acad Sci U S A 96:11531–11536, 1999
30. Xu Y, Shen Z, Wiper DW, et al: Lysophosphatidic acid as a potential biomarker for ovarian and other gynecologic cancers. J Am Med Assoc 280:719–723, 1998 31. Schummer M, Ng WV, Bumgarner RE, et al: Comparative hybridization of an array of 21,500 ovarian cDNAs for the discovery of genes overexpressed in ovarian carcinomas. Gene 238:375–385, 1999[CrossRef][Medline]
32. Hough CD, Sherman-Baust CA, Pizer ES, et al: Large-scale serial analysis of gene expression reveals genes differentially expressed in ovarian cancer. Cancer Res 60:6281–6287, 2000
33. Mok SC, Chao J, Skates S, et al: Prostasin, a potential serum marker for ovarian cancer: Identification through microarray technology. J Natl Cancer Inst 93:1458–1464, 2001
34. Kim JH, Skates SJ, Uede T, et al: Osteopontin as a potential diagnostic for ovarian cancer. J Am Med Assoc 287:1671–1679, 2002 35. Yousef GM, Diamandis EP: Expanded human tissue kallikrein family: A novel panel of cancer biomarkers. Tumour Biol 23:185–192, 2002[CrossRef][Medline] 36. Diamandis EP, Yousef GM, Soosaipillai AR, et al: Human kallikrein 6 (zyme/protease M/neurosin): A new serum biomarker of ovarian carcinoma. Clin Biochem 33:579–583, 2000[CrossRef][Medline] 37. Luo LY, Bunting P, Scorilas A, et al: Human kallikrein 10: A novel tumor marker for ovarian carcinoma? Clin Chim Acta 306:111–118, 2001[CrossRef][Medline] 38. Petricoin EF, Ardekani AM, Hitt BA, et al: Use of proteomic patterns in serum to identify ovarian cancers. Lancet 359:572–577, 2002[CrossRef][Medline] 39. Zhang H, Xu F, Yu Y, et al: Combination of multiple serum markers using an artificial neural network to improve specificity on discriminating malignant from benign pelvic masses. Gynecol Oncol 73:56–61, 1999[CrossRef][Medline] 40. Zhang Z, Xu F, Yu Y, et al: Use of multiple markers to detect stage I epithelial ovarian cancers: Neural network analysis improves performance. Proc Am Soc Clin Oncol 20:219a, 2001 (abstr 873) 41. Skates SJ, Zhang Z, Bast RC: High sensitivity and specificity for detection of ovarian cancer using mixtures of multivariate normal distribution to combine information from CA125, CA15.3, CA72.4 & MCSF. Proc Am Soc Clin Oncol 21:205a, 2002 (abstr 816)
42. 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 43. 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] 44. Cohen LS, Escobar PF, Scharm C, et al: Three-dimensional power Doppler ultrasound improves the diagnostic accuracy for ovarian cancer prediction. Gynecol Oncol 84:352–353, 2002[CrossRef][Medline] Submitted January 10, 2003; accepted March 17, 2003.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
