Will the cutting-edge technology of protein profiling revolutionize medicine in general and early ovarian cancer detection in particular?
|Jump to:||Choose article section... Identifying women at increased risk Early detection methods Ultrasound is effective in detecting advanced-stage disease Plasma and serum markers Ovarian Pap test Conclusions Key points|
Will the cutting-edge technology of protein profiling revolutionize medicine in general and early ovarian cancer detection in particular?
Some 23,300 American women will be diagnosed with ovarian cancer this year and 14,000 will die from the disease. Even though this cancer is curable if detected early on, it remains the leading cause of death among all gynecologic cancers in the United States and is the fifth most common female cancer. Moreover, its incidence has been steadily rising over the past 10 years, with a woman's overall lifetime risk now up to 1.8%.1
Despite improvements in median survival through surgical advances and new chemotherapeutic regimens, the overall 5-year survival rate for women with stage III/IV epithelial ovarian carcinoma (EOC) has barely budged from a dismal 12% over the past 30 years. In contrast, women diagnosed with disease confined to the ovary (stage IA-IB), require less morbid surgical intervention, may not need adjuvant chemotherapy, have an improved quality of life, and most importantly, nearly nine of every 10 will survive for 5 years.2,3
Thereforeshort of an effective therapythe early detection of early-stage EOC is the best way to improve survival. The challenge is to develop highly sensitive and specific methods that can be applied to the general population for the detection of early-stage ovarian cancer.
It's our goal in this article to discuss the development of new tumor marker tests with potential higher accuracy than the current serum biomarker CA125, new detection tests, and experimental technologies currently being evaluated by the National Cancer Institute as part of the National Ovarian Cancer Early Detection Program (Table 1). If we are correct, the use of genomics and proteomicsthe study of protein shape, function, and patterns of expressionscould revolutionize medicine by identifying aberrant gene and protein expression, permitting us to identify premalignant or malignant cells on the molecular levelbefore physical changes can be seen under a microscope.
National Ovarian Cancer Coalition
National Ovarian Cancer Early Detection Program
Ovarian Cancer National Alliance
The American Board of Genetic Counseling
The Gynecological Cancer Foundation-Women's Cancer Network
Before we can make a dent in the morbidity and mortality rates associated with ovarian cancer, we need to identify those women at increased risk. Epidemiologic factors associated with EOC are nulliparity, a personal history of breast cancer, a family history of ovarian or breast cancer, or both, as well as a diagnosis of an inherited malignancy syndrome.2,4,5 Women with one first-degree relative with ovarian cancer have a 4% to 7% lifetime chance of developing the disease by age 70 as compared to the general female population, whose risk is 1.8%.6
An inherited susceptibility gene is to blame for approximately 5% to 10% of all EOC cases. Clues in the family history that suggest hereditary susceptibility include: (1) two or more women with ovarian or breast cancer or both, especially if diagnosed premenopausally; (2) women who have separate diagnoses of both breast and ovarian cancers; (3) women who have had breast cancer in both breasts; (4) male relatives who have had breast cancer in addition to a female relative with breast or ovarian cancer; (5) women with ovarian cancer at any age who are of Ashkenazi Jewish ancestry.7
To date, researchers have identified BRCA1 and BRCA2, the two breast and ovarian cancer susceptibility genes. They've found that Jewish women of Eastern European ancestry, (a.k.a. Ashkenazi Jews), carry the BRCA1 (primarily 185delAG, 5382insC) and BRCA2 (primarily 6174delT) mutations at much higher rates (2%) than the general population. Roughly one in 40 Ashkenazi Jews are carriers of the BRCA1 and BRCA2 mutations, which may increase their chance of developing ovarian cancer up to 44% by age 70.4-6,8 Figure 1 depicts differences between sporadic, familial, and hereditary cancer pedigrees.
For decades clinicians have sought a means of detecting early-stage ovarian cancer, yet to date no technology has proven effective. A suitable screening test should have both high sensitivity (a positive test in an individual with the disease) and high specificity (a negative test in an individual without the disease).9 Specificity is a major concern in ovarian cancer screening because most women who test positive will require both diagnostic imaging and ultimately surgical intervention. For example, a test with 98% specificity would result in 50 false-positive procedures for every case of ovarian cancer detected on screening of postmenopausal women. A screening test for this population requires a 99.6% specificity to yield a positive predictive value of 10%, although lower specificity may be acceptable in those women appropriately identified to be within the high-risk population.
U/S remains the best diagnostic imaging tool to evaluate the adnexa. Although it has proven useful in detecting advanced-stage ovarian cancer in asymptomatic women, its value in detecting early-stage disease has yet to be realized.
Many studies have assessed U/S's usefulness and limitations for identifying stage I EOC in asymptomatic women. One group of investigators found that among high-risk women, the sensitivity of U/S for detecting stage I EOC was 25%, as compared to 67% for low-risk women.10 Other researchers found that U/S detected a total of 11 stage I tumors, five EOCs, and six granulosa cell and borderline tumors in a sample of women in the general population.11 Excluding granulosa and borderline tumors (which are usually detected when confined to the ovary), the sensitivity for detecting stage I EOC was 31%.
It is a clinical reality that a negative U/S examination may be falsely reassuring, as women continue to develop advanced-stage ovarian cancer within 6 to 12 months of a normal scan. A major limitation of transvaginal architectural screening is that ovarian cancers can arise from normal-sized ovaries that appear structurally normal, despite the present advances in diagnostic imaging technology.
The most recent of the rapid technological advances in diagnostic U/S over the last decade are three-dimensional transvaginal gray-scale volume imaging (3D TVS) and three-dimensional transvaginal power Doppler imaging (PD3D TVS). Initial studies suggest that these new technologies improve upon the diagnostic accuracy of two-dimensional transvaginal gray-scale imaging (2D TVS) in differentiating benign from malignant adnexal pathology.1-3 One reported advantage of 3D TVS using surface rendering is improved visualization of the internal architecture of adnexal masses containing cystic components. The addition of PD3D TVS allows the sonologist to thoroughly examine the complex adnexal mass for abnormal vascularity in three distinct planes.
We recently reported on the use of new 3D techniques to improve the diagnostic accuracy of 2D TVS in distinguishing benign complex adnexal masses from ovarian carcinoma.12 The objective of our study was to determine if 3D power Doppler U/S improves the specificity for ovarian cancer detection as compared with 2D U/S.
We reported on 71 women who underwent surgical exploration for a complex adnexal mass40 premenopausal women ranging in age from 22 to 53, with an average age of 32 yearsand 31 postmenopausal (age range 5280), with an average age of 59 years. Via histopathologic prediction, we correctly identified eight out of 11 endometriotic cysts, 10/13 cystic teratomas, 4/15 cystadenomas, and 3/10 cystadenofibromas. 2D TVS imaging identified 40 masses as suspicious for cancer and 3D TVS with surface rendering did not change this number.
Using this technique, we correctly identified all 14 women with an adnexal malignancy, yielding a sensitivity, specificity, and positive predictive value of 100%, 54%, and 35%, respectively, for gray-scale imaging. The addition of 3DPD significantly narrowed the suspect group from 40 to 28 patients. This reflects 12 complex masses that were correctly reclassified as benign rather than malignant because of a negative power Doppler examination. No malignant masses in this series had a negative power Doppler exam, which yielded a sensitivity, specificity, and positive predictive value of 100%, 75%, and 50%, respectively, for the combined modalities. Further, in a sample of high-risk women, the addition of 3D Doppler improved the ability to distinguish benign from malignant changes and was particularly useful in differentiating adenofibromas and cystic teratomas from borderline and malignant tumors.12,13 However, even the addition of 3DPD imaging has yet to improve our ability to identify ovarian cancers in normal-sized ovaries.
By the time clinicians usually diagnose ovarian cancer, cancer cells have already invaded and metastasized outside of the ovary. Early-stage disease often presents with few if any symptoms, but through understanding the genetic, molecular, and biochemical processes of carcinogenesis, invasion, and metastasis, researchers have identified novel genes, proteins, and lipids associated with ovarian cancer development and progression.
Ovarian cancers accumulate genetic aberrations that affect cell-cycle control, apoptosis, adhesion, angiogenesis, transmembrane signaling, DNA repair, and genomic stability. Specific genetic aberrations found in ovarian cancers include amplification and/or overexpression of various proteins, such as ErbB2 oncoprotein and phosphoinositide 3-kinase (PIK3CA).14-17 Those regions of recurrent abnormality may encode genes that play a part in ovarian cancer progression when differentially expressed, as a result of abnormal copy number or mutation.14-20
CA125. Investigators have studied a variety of ovarian tumor markers, focusing the most attention on the commonly used serum biomarker CA125, an ovarian cancer cell surface-associated protein that is expressed in 80% of nonmucinous EOCs.2 CA125 is shed from the cell surface of the fallopian tubes, endometrium, endocervix, peritoneum, pleura, pericardium, and bronchus. Little, if any, CA125 can be detected on normal ovarian epithelium, although the antigen is sometimes found on the ovary inside inclusion cysts, benign papillary excrescences, and when the epithelium undergoes tubal metaplasia.
The most valuable clinical applications of CA125 are for monitoring the disease status of women with metastatic gynecologic cancers, predicting residual disease before it's suspected clinically, and attempting to distinguish benign from malignant masses preoperatively. Overall, more than 80% of women with advanced-stage ovarian cancer will have an elevated CA125 (greater than 35 µ/mL), yet the test will detect early-stage asymptomatic ovarian cancer less than 50% of the time.21
CA125 is not specific for ovarian cancer. Elevated levels have been found in women with endometrial, fallopian tube, and pancreatic cancers, as well as gastric, breast, lung, and colon cancers. Among the benign gynecologic conditions that can also falsely elevate CA125 levels are pregnancy, pelvic inflammatory disease, endometriosis, fibroids, benign ovarian cysts, and menstruation.22 Nongynecologic conditions associated with elevated CA125 levels are pancreatitis, cirrhosis, colitis, peritonitis, peritoneal tuberculosis, radiation therapy, intraperitoneal chemotherapy, and postsurgical inflammation.
A recent NIH Consensus Statement did not recommend CA125 as a screening test in a general or high-risk population because an elevated value accurately detects malignancy in less than 3% of women.23,24 However, an elevated CA125 level in a postmenopausal woman presenting with a complex adnexal mass is a different story. Consider that suspicious and refer the patient to a gynecologic oncologist. Despite these limitations, CA125 has been proven to be the most useful tumor marker currently developed.
Proteomics. Far and away the most promising tool for identifying unique protein signatures in the serum of women with ovarian cancer appears to be proteomicsas mentioned earlier, the study of protein shape, function, and patterns of expressions. In preliminary studies, this technology can distinguish malignant from benign tumors of the ovary, using sophisticated analytical tools. Low-molecular-weight serum protein profiling may reflect the pathologic state of organs, thus aiding in early cancer detection.
Two types of mass spectroscopya technique that sorts proteins and other molecules based on their weight and electrical chargecan profile proteins in this range. These are matrix-assisted laser desorption and ionization time-of-flight (MALDI-TOF) and the new technology of surface-enhanced laser desorption and ionization time-of-flight (SELDI-TOF).5-8 Bioinformatics has been employed to study physiological outcomes and cluster gene microarray transcript profiles.9-12
However, uncovering changes in complex serum protein mass spectra patterns requires higher order analysis. We have linked SELDI-TOF spectral analysis with a high-order analytical approach to define the ideal discriminatory proteomic pattern. We use SELDI-TOF to profile proteins for a distinct pattern. These profiles contain thousands of data points that artificial intelligence-based systems learn through repeated profile analysis. This results in the ability to discriminate a handful of proteins, among thousands, that could be used to distinguish between cancerous and noncancerous conditions).25,26 We anticipate that this blood test we've developed will become clinically available this summer.
Petricoin and colleagues recently reported proteomic pattern results of a 116-sample study of women with and without ovarian cancer.25 Not only did they correctly identify all women (n=50) with ovarian cancerincluding all women with stage I cancers (n=18)but they also accurately classified as noncancerous 63 of 66 noncancerous samples. Subsequent testing using newer discriminatory protein patterns were shown to have 100% sensitivity and 100% specificity.25,26 A study examining protein patterns among prostate cancer cases showed similar results.
LPA. A phospholipid cited as a potential serum marker for early detection of EOC, lysophosphatidic acid (LPA) is a normal constituent of serum that is produced and released by activated platelets during coagulation. LPA plays a critical role in tumor metastasis by stimulating proliferation, invasion, migration, motility, inhibition of apoptosis, as well as chemoresistance. When comparing LPA and CA125 blood levels, researchers detected elevated plasma LPA levels in nine out of 10 patients with stage I, 24/24 patients with stage II, III, IV, and 14/14 with recurrent ovarian cancer. Most patients with other gynecologic cancers also had elevated LPA levels.
In contrast, they detected elevated plasma LPA levels in a minority of healthy controls (5/48), and in patients with benign gynecologic disease (4/18). CA125 levels were elevated in 28 of 47 women with ovarian cancer, including two of nine women with stage I disease. In preliminary studies, LPA levels were 95% sensitive and 89% specific.27 Although this is a significant improvement over the CA125 test, further improvement is necessary before LPA can be used clinically.
Serum p110 ErbB1 levels. Overexpression of ErbB1, ErbB2 and ErbB3 (members of a growth-factor receptor family) is common in human ovarian carcinoma-derived cell lines and tumors and is thought to play a critical role in the cause and progression of tumors.28,29 Researchers have developed an acridinium-linked immunosorbent assay (ALISA) to detect soluble ERbB1 in human body fluids. They found that serum p110 sErbB1 levels were significantly lower in women with stage I to IV EOC shortly after cytoreductive-staging laparotomy, compared to levels in healthy controls or in women with benign pelvic disease. This observation has led some to suggest that p110 ErbB1 levels may provide important diagnostic and/or prognostic information for managing women with EOC.
Osteopontin. This acidic, calcium-binding glycophosphoprotein is found in all body fluids and in components of extracellular matrix (ECM). Investigators have found that osteopontin blood levels are significantly higher in ovarian cancer cases, when compared to healthy controls. The specificity of the samples was 80.4%, while the sensitivity was 80.4% in early-stage disease and 85.4% in later-stage disease.30
Prostasin. Although this serine protease is most abundant in the prostate gland, it has also been isolated as a potential biomarker for ovarian cancer.31 Levels of serum prostasin in archived samples from 64 patients with ovarian cancer (55% stage III or IV) were almost double those of 137 controls. Twenty-four controls had other gynecologic cancers, 42 had benign gynecologic diseases, and 71 women had no known gynecologic diseases. In 14 of the 16 ovarian cancer patients with both pre-op and post-op serum samples, prostasin levels declined significantly after surgery.
Periostin. Although periostin is a secreted protein expressed in normal tissues (the stomach, aorta, placenta, uterus, and breast), it is not expressed in normal ovarian tissue. Researchers have shown that EOC cells secrete perisotin and that malignant ovarian ascites also contain high levels of periostin. This protein may play an important role in the adhesion and migration of ovarian epithelial cells.32
As we know, the cervical Pap test identifies premalignant changes on the cervix and provides an effective way to detect cervical cancer early on. Likewise, the ability to identify premalignant lesions on the ovary could be of significant value in detecting early-stage ovarian cancer.
Cyto- and histopathologists evaluated cytological data from ovarian surface epithelium in a blinded manner. A total of 120 ovaries were removed from 60 women undergoing surgical exploration because of gynecologic manifestations (90 ovaries were benign and 30 were malignant). An expert gynecologic pathologist and an expert cytopathologist independently evaluated all specimens. In each of the 120 samples, the final surgical pathology and the ovarian Pap cytopathology report matched. The preliminary report indicated that ovarian cytology could discern malignant ovarian epithelium from normal.33 The application of the molecular taxonomy of cancer can easily be applied to cells that appear cytologically normal but that have aberrant expression of genes, proteins, and methylation consistent with malignancy. Preliminary studies demonstrate that aberrant gene expression may help identify those cells that are malignant despite normal phenotypic appearance.
The future of early detection of ovarian cancer will depend on our ability to use new technologies to identify minute changes in plasma/serum, gene/protein expression, lipid ratios, as well as on ovarian architecture and vascularity (Figure 2). We must clinically apply the biochemical, genetic, and molecular aspects of ovarian carcinogenesis, invasion, and metastasis to bring about change in the toll of morbidity and mortality from epithelial ovarian cancer. The National Cancer Institute (NCI) and the Food and Drug Administration are committed to developing early detection methods, effective chemoprevention, and ovarian-specific therapies. The application of new technologies, such as SELDI-TOF, affords an opportunity to challenge scientific paradigms, and has already led to the identification of biologically relevant lipids, proteins, gene mutations, aberrant DNA methylation, and specific low-molecular-weight proteins that are biologically relevant to the disease process. All of the markers mentioned in this article are under formal NCI evaluation studies to determine if their use will be justified in clinical practice. We anticipate that the validation of these new detection tests and technologies will significantly improve women's health care and quality of life.
1. Jemal A, Thomas A, Murray T, et al. Cancer statistics, 2002. CA Cancer J Clin. 2002;52:23-47.
2. Ozols RF, Rubin SC, Thomas GM, et al. Epithelial Ovarian Cancer. In: Hoskins WJ, Perez CA, Young RC, eds. Principles and Practice of Gynecologic Oncology. Philadelphia, Pa: Lippincott Williams & Wilkins; 2000:981-1058.
3. American Cancer Society. Cancer facts and figures, 2002. Atlanta, Ga: American Cancer Society, 2002.
4. 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. 1998;62:676-689.
5. Moslehi R, Chu W, Karlan B, et al. BRCA1 and BRCA2 mutation analysis of 208 Ashkenazi Jewish women with ovarian cancer. Am J Hum Genet. 1998;66:1259-1272.
6. Stratton JF, Pharoah P, Smith SK, et al. A systematic review and meta-analysis of family history and risk of ovarian cancer. Br J Obstet Gynaecol. 1998;105:493-499.
7. American Medical Association. Gene testing for breast and ovarian cancer. Chicago, IL: American Medical Association, 1997.
8. Ford D, Easton DF, Bishop DT, et al. Risks of cancer in BRCA1-mutation carriers. Breast Cancer Linkage Consortium. Lancet. 1994;343:692-695.
9. Cronin KA, Weed DL, Connor RJ, et al. Case-control studies of cancer screening: theory and practice. J Natl Cancer Inst. 1998;90:498-504.
10. Bell R, Petticrew M, Sheldon T. The performance of screening tests for ovarian cancer: results of a systemic review. Br J Obstet Gynacol. 1998;105:1136-1147.
11. van Nagell JR Jr, DePriest PD, Reedy MB, et al. The efficacy of transvaginal sonographic screening in asymptomatic women at risk for ovarian cancer. Gynecol Oncol. 2000;77:350-356.
12. Cohen LS, Escobar PF, Scharm C, et al. Three-dimensional power Doppler ultrasound improves the diagnostic accuracy for ovarian cancer prediction. Gynecol Oncol. 2001;82:40-48.
13. Fishman DA, Cohen LS. Is transvaginal ultrasound effective for screening asymptomatic women for the detection of early stage epithelial ovarian carcinoma? Gynecol Oncol. 2000;77:347-349.
14. Slamon DJ, Godolphin W, Jones LA, et al. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science. 1989;244:707-712.
15. Shayesteh L, Lu Y, Kuo WL, et al. PIK3CA is implicated as an oncogene in ovarian cancer. Nat Genet. 1999;21:99-102.
16. Fei R, Shaoyang L. Combination antigene therapy targeting c-myc and c-erbB(2) in the ovarian cancer COC(1) cell line. Gynecol Oncol. 2002;85:40-44.
17. Brooks DJ, Woodward S, Thompson FH, et al. Expression of the zinc finger gene EVI-1 in ovarian and other cancers. Br J Cancer. 1996;74:1518-1525.
18. Friedlander ML. Prognostic factors in ovarian cancer. Semin Oncol. 1998;25:305-314.
19. Gray JW, Chin K, Waldman F. A molecular cytogenic view of chromosomal heterogeneity in solid tumors. In: Mihich E, Hartwell L, eds. Genomic Instability and Immortality in Cancer. New York, NY: Plenum Press; 1997:3-32.
20. Umayahara K, Cheneviex-Trench G, Daneshvar L, et al. Molecular genetic studies. In: Sharp F, Blackett T, Berek J, eds. Ovarian Cancer. Oxford, England: ISIS Medical Media; 1998:17-24.
21. Berek JS, Bast RC Jr. Ovarian cancer screening: The use of serial complementary tumor markers to improve sensitivity and specificity for early detection. Cancer. 1995;76:2092-2096.
22. Evans AC Jr, Berchuck A. Tumor markers. In: Hoskins WJ, Perez CA, Young RC, eds. Principles and Practice of Gynecologic Oncology, 2nd ed. Philadelphia, Pa: Lippincott-Raven; 1997:177-196.
23. Schwartz PE, Chambers JT, Taylor KJ. Early detection and screening for ovarian cancer. J Cell Biochem Suppl. 1995;23:233-237.
24. Rosenthal A, Jacobs I. Ovarian cancer screening. Semin Oncol. 1998;25:315-325.
25. Petricoin EF, Zoon KC, Kohn EC, et al. Clinical proteomics: translating benchside promise into bedside reality. Nat Rev Drug Discov. 2002;1:683-695.
26. Petricoin EF, Ardekani AM, Hitt BA, et al. Use of proteomic patterns in serum to identify ovarian cancer. Lancet. 2002;359:572-577.
27. Xu Y, Shen Z, Wiper DW, et al. Lysophosphatidic acid as a potential biomarker for ovarian and other gynecologic cancers. JAMA. 1998;280:719-723.
28. Baron AT, Cora EM, Lafky JM, Fishman DA, et al Soluble Epidermal Growth Factor Receptor (sEGFR/sErbB1) as a potential risk, screening, and diagnostic serum biomarker of epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2003;12:103-113.
29. Maihle NJ, Baron AT, Barrette BA, et al. EGF/ErbB receptor family in ovarian cancer. In: Stack MS, Fishman DA, eds. Ovarian Cancer. Boston, Mass: Kluwer Academic Publishers; 2002:247-258.
30. Kim JH, Skates SJ, Uede T, et al. Osteopontin as a potential diagnostic biomarker for ovarian cancer. JAMA. 2002;287:1671-1679.
31. Mok SC, Chao J, Skates S, et al. Prostasin, a potential serum marker for ovarian cancer: Identification through microarray technology. J Natl Cancer Inst. 2001; 93:1458-1464.
32. Gillan L, Matei D, Fishman DA, et al. Periostin secreted by epithelial ovarian carcinoma is a ligand for alpha(V) beta(3) and alpha(V) beta(5) integrins and promotes cell motility. Cancer Res. 2002;62:5358-5364.
33. Fishman DA, Bozorgi K. The scientific basis of early detection of epithelial ovarian cancer. In: Stack MS, Fishman DA, eds. Ovarian Cancer. Boston, Mass: Kluwer Academic Publishers; 2002:3-28.
David Fishman. Will high-tech tests revolutionize the early detection of ovarian Ca?.
Jun. 2, 2003;48:44-56.