|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2007.11.9594 on August 13 2007 © 2007 American Society of Clinical Oncology.
How Do You Distinguish a Malignant Pelvic Mass From a Benign Pelvic Mass? Imaging, Biomarkers, or None of the Above
Program in Women's Oncology, Department of Obstetrics and Gynecology, Women and Infants' Hospital, Brown University, Providence, RI
Department of Experimental Therapeutics, The University of Texas M.D. Anderson Cancer Center, Houston, TX The 125th attempt to develop a monoclonal antibody against ovarian cancer in a laboratory at the Dana-Farber Cancer Institute (Boston, MA) led to the discovery of one of the most widely used tumor markers, CA-125.1 Over the past two decades, the value of CA-125 has been tested in monitoring response to treatment, detecting recurrent disease, and screening for early-stage ovarian cancer.2,3 CA-125 has proven useful in determining prognosis and response to treatment for women undergoing chemotherapy.4,5 Persistent elevation of CA-125 has detected residual disease after primary treatment with high specificity. A rise in serum CA-125 levels often precedes clinical symptoms or imaging detection of recurrent disease by 3 to 6 months.6,7 While the use of a "good marker" to track a "bad disease" in these settings has been questioned, information provided by CA-125 provides additional time for patients to receive and to benefit from the many currently active drugs. As treatment of persistent and recurrent disease becomes more effective in the era of targeted therapy, CA-125 and other biomarkers should become of even greater value in patient management. While a single value of CA-125 lacks the specificity and sensitivity required for early detection, greater specificity has been attained by measuring CA-125 over time and by combining CA-125 with ultrasonography. Statistical analysis, such as the Risk of Ovarian Cancer Algorithm (ROCA model) developed by Skates et al,8 can examine changes in CA-125 levels over time. The United Kingdom Collaborative Trial on Ovarian Cancer Screening has completed accrual of 200,000 women, testing the value of a rising CA-125 to trigger transvaginal sonography in postmenopausal women at average risk for ovarian cancer.9 If this trial succeeds, there is clearly room for improvement, as 20% of ovarian cancers fail to express significant amounts of CA-125. A greater fraction of patients with early-stage disease might be detected with a panel of biomarkers that complement or replace CA-125. CA-125 has also been tested for the ability to distinguish malignant from benign pelvic masses. The ability to predict whether a tumor is malignant or benign before surgery is important. Approximately 20% of women will develop an ovarian cyst or pelvic mass in their lifetime and many of these women will undergo unnecessary surgery. Conversely, if surgery is indicated, the location in which the surgery is performed is critical. Despite modest gains in the fraction of women cured, optimal management of malignant ovarian cancers has significantly improved survival over the past 30 years. With the development and application of cytoreductive surgery, primary platinum-taxane chemotherapy and active second-line agents, the 5-year survival rate for ovarian cancer patients has improved from 20% to 50%.10,11 Several different reports have shown that patients treated by gynecologic oncologists will more often undergo full surgical staging or an optimal cytoreductive operation.12-14 More recently, multiple studies have shown ovarian cancer patients treated in tertiary care centers with multidisciplinary teams specializing in the management of ovarian cancer have fewer complications and longer survival rates.12,15-17 The triage of women to centers of excellence in the management of ovarian malignancy is critical for optimal treatment. Tests that permit the identification of women at high risk for the presence of a malignancy will be instrumental for the triage of patients to appropriate centers. Despite a number of trials examining CA-125, with or without the use of sonography, in women with a pelvic mass, it has become increasingly clear that no one modality will be sufficient to predict accurately the presence of an ovarian malignancy. Many different tumor markers have been analyzed, but none has achieved the sensitivity or specificity to be clinically useful as an individual test.18 Recently, the addition of HE-4 to CA-125, without the use of ultrasound, has increased the sensitivity of CA-125 by 22% at a specificity of 90%.19 Jacobs et al20 have used the combination of ultrasound, CA-125, and menopausal status to create a risk of malignancy index (RMI), achieving a sensitivity of 85% with a specificity of 97% for predicting the presence of ovarian cancers in women with pelvic masses. The contribution of CA-125 to the RMI was critical for assigning masses to the malignant category, whereas sonography was particularly important in identifying benign disease. One of the advantages of the RMI is its simplicity, making it appropriate for everyday clinical use. DePriest et al21 developed a morphology index utilizing the architectural features of ovarian neoplasms on pelvic ultrasound to predict a probability of malignancy without the use of CA-125. The morphology index achieved a positive predictive value of up to 0.45. Although many of the morphologic features and measurements that comprise the morphology index are routinely performed and reported in standard pelvic ultrasounds, many of the variables are often not reported or measured, therefore limiting its clinical utility in community practice. Perhaps for this reason, the morphology index is not frequently used to calculate a probability of malignancy. Another approach in assessing the risk for malignancy is to identify many independent variables associated with malignancy and to create an algorithm that uses each independent variable. In a prior article, Timmerman et al22 reported data collected by the International Ovarian Tumor Analysis Group (IOTAG). More than 50 defined variables were recorded and analyzed for 1,066 patients. A statistical analysis identified 12 independent risk factors for malignancy (age, personal history of ovarian cancer, maximum diameter of the lesion, and maximum diameter of the solid component, presence of ascites, Doppler blood flow, a purely solid lesion, irregular internal cyst wall, increase color score, hormone therapy, and pain with ultrasound). A logistic regression model was created using each of these variables, called the M1 model. The M1 model, with a probability value of .01, achieved a sensitivity of 93% and specificity of 77% in the IOTAG cohort. When receiver operating characteristic (ROC) curves for the RMI and M1 were calculated using the same cohort, the M1 model exhibited a statistically significant improvement in area under the curve (AUC). In this issue of the Journal of Clinical Oncology, Timmerman et al23 added CA-125 levels as a variable to the M1 model. The AUC of the ROC curves were compared with and without CA-125, as well as with CA-125 alone and with the RMI. The cohort was examined as a whole and then subdivided into premenopausal and postmenopausal groups. Analysis of the premenopausal group revealed no statistical difference in the AUC of the ROC curves when CA-125 was added to M1. This finding is not surprising because CA-125 tends to be elevated by many of the benign gynecologic conditions that present in this age group and is less frequently elevated in mucinous cancers and in borderline tumors that occur in younger women. Also, the prevalence of ovarian malignancy in the younger population is much lower than that in the postmenopausal group, making it more difficult to detect a significant difference with a limited number of premenopausal cases. A tumor marker, such as HE-4, that is elevated by ovarian malignancies and not as frequently elevated in benign conditions, which are commonly found in younger women, might add value to the M1 analysis in the premenopausal group. Given the heterogeneity of ovarian cancers in different women, a combination of multiple markers may provide maximum additive value to the M1 algorithm. The analysis of the postmenopausal group revealed that the addition of CA-125 to the M1 algorithm increased the AUC of the ROC curve over that of M1 alone, although the difference did not achieve statistical significance, with a P value of .0779. Of interest is the observation that the addition of CA-125 substantially improved the sensitivity for detecting malignant masses at levels of specificity approaching 90% in a portion of the ROC curve that would impact clinical decisions (Fig 3). Lack of overall improvement when CA-125 was added to ultrasonography is in contrast to reports by Jacobs et al,20 in which CA-125 made a considerable contribution to the RMI. The multiple ultrasound parameters employed in the M1 algorithm reduce the contribution that CA-125 adds to the algorithm when compared with the RMI that uses fewer ultrasound measurements. The M1 algorithm is heavily weighted with ultrasound parameters and the authors suggest that the presence of ascites, the maximum diameter of the solid component, and the maximum diameter of the lesion make CA-125 redundant in their model. Many of the ultrasound measurements and morphologic findings recorded as part of the IOTAG study are often not routinely reported in everyday clinical practice. In addition, many of the measurements or morphologic findings, particularly those obtained with Doppler, can vary from operator to operator, permitting the introduction of operator error and subjective data into the M1 equation. These issues may render the M1 model technically challenging for use in the general clinical setting. The authors feel the sonographic techniques included in the M1 model are ones that should be able to be achieved by any competent sonographer. However, in order for a test to be successful, it needs to be reproducible, technically simple, and applicable in everyday clinical situations across all spectra of health care. For instance, a test that is only reproducible in the hands of the best technologist at a tertiary care center may be of very little clinical value in the community. Complex algorithms that employ many different architectural features of an ovarian cyst or mass may or may not be clinically useful in the real world. A prospective multicenter evaluation of the M1 model will need to be performed to evaluate its utility in the general population. So where do we go from here? Neither imaging nor CA-125 alone may provide optimal tests to distinguish malignant from benign pelvic masses with optimal sensitivity or specificity. For triage of patients for care by specialists capable of thorough surgical exploration and cytoreductive surgery, sensitivity and positive predictive value for malignancy will be critical. Conversely, to delay or avoid surgery based on diagnostic tests, specificity and negative predictive value will be critical. At the extremes of sensitivity and specificity there is still room for improvement in M1. Computed tomography (CT), magnetic resonance imagine, and positron emission tomography-CT may help to stratify these patients into high and low risk groups. However, the cost of CT, magnetic resonance imaging, and positron emission tomography-CT make these modalities too expensive to utilize for all women with a pelvic mass or ovarian cyst. Likewise, serum tumor markers may not achieve the sensitivity or specificity needed to accurately identify patients that harbor a malignancy or have a benign neoplasm. More realistically, the employment of multiple biomarker assays for the stratification of patients into low-, moderate-, and high-risk groups will enable the economical use of imaging technology to be employed. Patients in a low-risk group may be observed over time and observed with serial tumor marker assays whereas patients at high risk for malignancy may be triaged directly to institutions with expertise in ovarian cancer. The patients falling into an intermediate-risk group may then undergo further risk stratification with advanced imaging technology. The most reliable model for the prediction of malignancy will be one that is easily reproducible and combines the use of objective results, such as multiple marker analysis combined with imaging that is available in both the tertiary and community settings. The important work performed by the authors and the IOTAG group provides a step forward in identifying algorithms that can be used to predict the presence of malignancy in women with a pelvic mass. The combination of the risk indicators identified in the M1 analysis with multiple marker models or proteomic patterns may have additive value for distinguishing benign from malignant ovarian masses. If reliable methods can be developed, the question remains whether clinicians in the community will utilize them and whether a larger fraction of patients likely to have ovarian cancer will be referred to surgeons with the appropriate expertise. At present, fewer than 50% of women with ovarian cancer in the United States receive care for ovarian cancer by or in collaboration with a gynecologic oncologist.12,24 Multiple barriers must be addressed, including patient preference to remain with her local practitioner, lack of health insurance in women younger than 65, transportation, and competing financial incentives. Access to information for women diagnosed with a pelvic mass, as well as programs of education for health professionals, will be required. Given the expense of managing recurrent ovarian cancer, successful primary treatment is cost effective and third party payers might take a more active role in encouraging appropriate testing and triage of patients with a pelvic mass. AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following authors or their immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: Richard G. Moore, Fujirebio Diagnostics (C); Robert C. Bast Jr, Ciphergen Incorporated (C), Fujirebio Diagnostics Incorporated (U) Stock Ownership: None Honoraria: None Research Funding: Robert C. Bast Jr, Fujirebio Diagnostics Incorporated, Ciphergen Incorporated Expert Testimony: None Other Remuneration: Robert C. Bast Jr, Fujirebio Diagnostic Incorporated AUTHOR CONTRIBUTIONS Conception and design: Richard G. Moore, Robert C. Bast Jr Data analysis and interpretation: Richard G. Moore Manuscript writing: Richard G. Moore, Robert C. Bast Jr Final approval of manuscript: Richard G. Moore, Robert C. Bast Jr NOTES published online ahead of print at www.jco.org on August 13, 2007. REFERENCES 1. Bast RC Jr, Feeney M, Lazarus H, et al: Reactivity of a monoclonal antibody with human ovarian carcinoma. J Clin Invest 68:1331-1337, 1981[Medline] 2. 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] 3. Jacobs I, Davies AP, Bridges J, et al: Prevalence screening for ovarian cancer in postmenopausal women by CA 125 measurement and ultrasonography. BMJ 306:1030-1034, 1993[Medline] 4. Hawkins RE, Roberts K, Wiltshaw E, et al: The prognostic significance of the half-life of serum CA 125 in patients responding to chemotherapy for epithelial ovarian carcinoma. Br J Obstet Gynaecol 96:1395-1399, 1989[Medline] 5. Rustin GJ, Nelstrop AE, McClean P, et al: Defining response of ovarian carcinoma to initial chemotherapy according to serum CA 125. J Clin Oncol 14:1545-1551, 1996 6. Gadducci A, Cosio S, Zola P, et al: Surveillance procedures for patients treated for epithelial ovarian cancer: A review of the literature. Int J Gynecol Cancer 17:21-31, 2007[CrossRef][Medline] 7. Sugiyama T, Nishida T, Komai K, et al: Comparison of CA 125 assays with abdominopelvic computed tomography and transvaginal ultrasound in monitoring of ovarian cancer. Int J Gynaecol Obstet 54:251-256, 1996[CrossRef][Medline] 8. 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 (suppl)[CrossRef][Medline] 9. Jacobs IJ, Menon U: Progress and challenges in screening for early detection of ovarian cancer. Mol Cell Proteomics 3:355-366, 2004 10. Goodman MT, Howe HL: Descriptive epidemiology of ovarian cancer in the United States, 1992-1997. Cancer 97:2615-2630, 2003 (suppl)[CrossRef][Medline] 11. Chan JK, Cheung MK, Husain A, et al: Patterns and progress in ovarian cancer over 14 years. Obstet Gynecol 108:521-528, 2006 12. Earle CC, Schrag D, Neville BA, et al: Effect of surgeon specialty on processes of care and outcomes for ovarian cancer patients. J Natl Cancer Inst 98:172-180, 2006 13. McGowan L, Lesher LP, Norris HJ, et al: Misstaging of ovarian cancer. Obstet Gynecol 65:568-572, 1985[Medline] 14. Olaitan A, Weeks J, Mocroft A, et al: The surgical management of women with ovarian cancer in the south west of England. Br J Cancer 85:1824-1830, 2001[CrossRef][Medline] 15. Paulsen T, Kjaerheim K, Kaern J, et al: Improved short-term survival for advanced ovarian, tubal, and peritoneal cancer patients operated at teaching hospitals. Int J Gynecol Cancer 16:11-17, 2006 (suppl 1)[CrossRef][Medline] 16. Engelen MJ, Kos HE, Willemse PH, et al: Surgery by consultant gynecologic oncologists improves survival in patients with ovarian carcinoma. Cancer 106:589-598, 2006[CrossRef][Medline] 17. Giede KC, Kieser K, Dodge J, et al: Who should operate on patients with ovarian cancer? An evidence-based review. Gynecol Oncol 99:447-461, 2005[CrossRef][Medline] 18. Bast RC Jr, Urban N, Shridhar V, et al: Early detection of ovarian cancer: Promise and reality. Cancer Treat Res 107:61-97, 2002[Medline] 19. Moore RG, Brown AK, Miller CM, et al: A novel multiple biomarker assay for the detection of ovarian carcinoma. J Clin Oncol 24:261s, 2006 (abstr 5023) 20. Jacobs I, Oram D, Fairbanks J, et al: A risk of malignancy index incorporating CA 125, ultrasound and menopausal status for the accurate preoperative diagnosis of ovarian cancer. Br J Obstet Gynaecol 97:922-929, 1990[Medline] 21. DePriest PD, Shenson D, Fried A, et al: A morphology index based on sonographic findings in ovarian cancer. Gynecol Oncol 51:7-11, 1993[CrossRef][Medline] 22. Timmerman D, Testa AC, Bourne T, et al: Logistic regression model to distinguish between the benign and malignant adnexal mass before surgery: A multicenter study by the International Ovarian Tumor Analysis Group. J Clin Oncol 23:8794-8801, 2005 23. Timmerman D, Van Claster B, Jurkovic D, et al: Inclusion of CA-125 does not improve mathematical models developed to distinguish between benign and malignant adnexal tumors. J Clin Oncol 25:4194-4200, 2007 24. Carney ME, Lancaster JM, Ford C, et al: A population-based study of patterns of care for ovarian cancer: Who is seen by a gynecologic oncologist and who is not? Gynecol Oncol 84:36-42, 2002[CrossRef][Medline] This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
