• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Oct 14, 2003; 100(21): 12343–12348.
Published online Oct 1, 2003. doi:  10.1073/pnas.2033602100
PMCID: PMC218760
Medical Sciences

Identification of biomarkers for ovarian cancer using strong anion-exchange ProteinChips: Potential use in diagnosis and prognosis

Abstract

One hundred eighty-four serum samples from patients with ovarian cancer (n = 109), patients with benign tumors (n = 19), and healthy donors (n = 56) were analyzed on strong anion-exchange surfaces using surface-enhanced laser desorption/ionization time-of-flight mass spectrometry technology. Univariate and multivariate statistical analyses applied to protein-profiling data obtained from 140 training serum samples identified three biomarker protein panels. The first panel of five candidate protein biomarkers, termed the screening biomarker panel, effectively diagnosed benign and malignant ovarian neoplasia [95.7% sensitivity, 82.6% specificity, 89.2% accuracy, and receiver operating characteristic (ROC) area under the curve of 0.94]. The other two panels, consisting of five and four candidate protein biomarkers each, effectively distinguished between benign and malignant ovarian neoplasia and were therefore referred to as validation biomarker panel I (81.5% sensitivity, 94.9% specificity, 88.2% accuracy, and ROC = 0.94) and validation biomarker panel II (72.8% sensitivity, 94.9% specificity, 83.9% accuracy, and ROC = 0.90). The three ovarian cancer biomarker protein panels correctly diagnosed 41 of the 44 blinded test samples: 21 of 22 malignant ovarian neoplasias [10 of 11 early-stage ovarian cancer (I/II) and 11 of 11 advanced-stage ovarian cancer (III/IV)], 6 of 6 low malignant potential, 5 of the 6 benign tumors, and 9 of 10 normal patient samples. In conclusion, we have discovered three ovarian cancer biomarker protein panels that, when used together, effectively distinguished serum samples from healthy controls and patients with either benign or malignant ovarian neoplasia.

Of the gynecologic malignancies, ovarian cancer has the highest mortality rate. Ovarian cancer often eludes the clinician because of the lack of early symptoms and signs. Hence, ovarian cancer tends to present at a late clinical stage in >85% of patients and is often followed by the emergence and outgrowth of chemotherapy-resistant disease in these patients after conventional primary cytoreductive surgery and induction chemotherapy (15). The American Cancer Society reported that >23,000 women were diagnosed with ovarian cancer in the United States in 2002, and 60% of those diagnosed, ≈14,000, are projected to die of their disease (4). More women die from ovarian cancer than from all other gynecologic malignancies combined (4, 5). However, the 5-year survival rate for patients diagnosed with early-stage disease is often >90%, but it is <20% for advanced-stage disease, underscoring the importance of early detection (15).

The diagnostic and prognostic tumor biomarkers in use today are not adequate in distinguishing benign from malignant ovarian neoplasia and cannot differentiate among the various histological and clinically aggressive forms of ovarian cancer (6). The most commonly used biomarker for clinical screening and prognosis in patients with ovarian cancer is ovarian cancer antigen 125 (CA125). Serum CA125 levels are elevated in ≈80% of patients with advanced-stage epithelial ovarian cancer but are increased in only 50–60% of patients with early-stage disease (79). Serum CA125 levels may be falsely elevated in women with any i.p. pathology resulting in irritation of the serosa of the peritoneum or pericardium, uterine fibroids, renal disorders, and normal menses (79). Moreover, serum CA125 levels do not predict the outcome of cytoreductive surgery in patients with advanced epithelial ovarian cancer (10).

Ciphergen Biosystems (Fremont, CA) has developed the ProteinChip technology coupled with surface-enhanced laser desorption/ionization time-of-f light mass spectrometry (SELDI-TOF-MS) to facilitate protein profiling of complex biological mixtures (1114). In SELDI-TOF-MS analysis, a nitrogen laser desorbs the protein/energy-absorbing molecule mixture from the array surface, enabling the detection of the proteins captured by the array. The efficacy of the SELDI-TOF-MS technology for the discovery of cancer protein markers in serum has recently been demonstrated (15, 16).

In this article, we report the identification of three panels of biomarker proteins that can be used in the diagnosis of early-stage ovarian cancer. The biomarker panels not only permitted the distinction of patients with ovarian neoplasia (benign or malignant) from normal subjects, but they also allowed the identification of patients with early-stage (stage I/II) ovarian cancer from those patients with benign ovarian tumors or normal individuals. Finally, in a blind test, the biomarker panels developed from this study distinguished diseased from healthy patients.

Materials and Methods

Materials. We obtained serum samples from healthy individuals (n = 56), patients with ovarian cancer (n = 109), and patients with benign tumors (n = 19) through the Gynecological Oncology Group and Cooperative Human Tissue Network. Serum samples had been collected preoperatively from patients with malignant and benign ovarian tumors. Sample numbers used for profiling (training group) and validation (test group) are listed according to histopathology in Table 1, and the stage and grade of tumors from patients with ovarian cancer are listed in Table 2.

Table 1.
Sample number used in training and test groups according to histopathology
Table 2.
Stage vs. grade of adenocarcinoma samples

Preparation of Serum Samples for SELDI Analysis. Two different dilutions (1:4 and 1:25) of serum samples were processed on strong anion-exchange (SAX2) chips according to the manufacturer's protocols (Ciphergen Biosystems). Briefly, the array spots were preactivated with binding buffer (1× PBS/0.1% Triton X-100, pH 7.5) at room temperature for 15 min in a humidifying chamber. Each serum sample was first diluted 1:2 or 1:5 with 9 M urea/2% Chaps/50 mM Tris·HCl, pH 9.0, and was further diluted 1:2 or 1:5, respectively, in binding buffer. Three microliters of each diluted sample was spotted onto preactivated SAX2 protein array chips and incubated in a humidity chamber for 30 min at room temperature. The chips were washed twice with binding buffer and once with HPLC H2O, and then air-dried. The chips were then sequentially treated with sinapinic acid (3,5-dimethoxy-4-hydroxycinnamic acid), first with 0.6 μl of a 100% saturated solution followed by 0.8 μl of a 50% saturated solution. The sinapinic acid solution was 50% acetonitrile and 0.5% trifluoroacetic acid. The chips were analyzed with the Ciphergen ProteinChip Reader (model PBSII). Each dilution was analyzed separately to confirm reproducibility in identifying the differentially expressed proteins.

Ciphergen ProteinChip SELDI-TOF-MS Analysis. The arrays were analyzed with the Ciphergen ProteinChip Reader (model PBSII). The mass spectra of proteins were generated by using an average of 65 laser shots at a laser intensity of 230–280 arbitrary units. For data acquisition of low molecular weight proteins, the detection size range was between 2 and 18 kDa, with a maximum size of 25 kDa. The laser was focused at 10 kDa. The detector sensitivity was set at 8, and the laser intensity was set at 230 for the 1:4 and 250 for the 1:25 dilution. For the high molecular weight proteins, the detection size range was between 20 and 150 kDa, with a maximum size of 250 kDa. The laser was focused at 85 kDa. The detector sensitivity was set at 9, and the laser intensity was set at 260 for the 1:4 dilution and 280 for the 1:25 dilution. The mass-to-charge ratio (m/z) of each of the proteins captured on the array surface was determined according to externally calibrated standards (Ciphergen Biosystems): bovine insulin (5,733.6 Da), human ubiquitin (8,564.8 Da), bovine cytochrome c (12,230.9 Da), bovine superoxide dismutase (15,591.4 Da), bovine β-lactoglobulin A (18,363.3 Da), horseradish peroxidase (43,240 Da), BSA (66,410 Da), and chicken conalbumin (77,490 Da).

Statistical Analysis. The data were analyzed with proteinchip data analysis software version 3.0 (Ciphergen Biosystems). For each comparison, the raw intensity data were normalized by using the total ion current of all profiles in the groups compared. The peak intensities were normalized to the total ion current of m/z between 3,000 and 25,000 Da for the low molecular weight range and between 4,000 and 250,000 Da for the high molecular weight range. The test group (n = 44) was normalized to all 140 training samples before using their intensities against the statistically derived intensity cutoffs. The Biomarker Wizard application (nonparametric calculations; Ciphergen Biosystems) was used to compile all spectra and autodetect quantified mass peaks. Peak labeling was completed by using second-pass peak selection with 0.3% of the mass window, and estimated peaks were added. Sample statistics were performed on groups of profiles (normal vs. benign/cancer and normal/benign vs. cancer). Protein differences (fold changes) were calculated among the various groups. A protein was considered differentially expressed in the ovarian cancer groups if, when compared with the normal group, statistically significant differences in its intensity were observed (P ≤ 0.01) in both the 1:4 and the 1:25 dilution analysis. Using the intensities derived from the 1:25 dilution analysis, we performed univariate comparisons of marker intensity summary statistics of representative markers using statistical analysis software (sas, Version 8.0, SAS Institute, Cary, NC). For each marker, we used t tests and Wilcoxon rank sum tests to compare the mean and median standardized intensities, respectively, between the normal and cancer groups and to determine their corresponding P values (nonparametric for medians). The sas program was used to determine the “best” intensity cutoff for each marker at either highest accuracy or when sensitivity equals specificity. Receiver operating characteristic (ROC) curves [plot of sensitivity vs. (1-specificity) for each possible cutoff] were generated for proteins with low P values, and the highest individual diagnostic power was calculated by using sas. We performed multivariate logistic regression analysis on our biomarkers using sas. The program was used to analyze various combinations of markers giving predictive scores for each panel tested. This predictive score is a sum of the individually weighted marker intensities.

Results

Identification of Differentially Expressed Ovarian Cancer-Associated Proteins in Serum. Using SELDI-TOF-MS proteomics technology (Ciphergen Biosystems), we identified proteins differentially expressed between serum from healthy individuals (n = 46) and serum from patients with ovarian cancer (benign, n = 13; low malignant potential (LMP), n = 14; and malignant, n = 67). A representative pseudogel view of specific candidate ovarian cancer tumor markers and a spectral overlay of candidate markers from healthy vs. diseased individuals are shown in Fig. 1. We performed statistical analysis on potential biomarkers with the lowest P values (≤0.01). The P values were generated with nonparametric tests from both the biomarker wizard application (Ciphergen Biosystems) and the sas program. Sensitivity, specificity, overall accuracy, and ROC area values were computed for each biomarker using the sas program.

Fig. 1.
Differentially expressed ovarian cancer-associated serum proteins. (A) Detection of differentially expressed ovarian cancer-associated serum proteins. A representative pseudogel view of SELDI-TOF-MS analysis of serum samples, processed on a SAX2 chip ...

Neoplasia Biomarkers and Malignant Neoplasia Biomarkers. We first considered proteins that are differentially expressed in healthy individuals vs. patients with any ovarian tumor, including benign or malignant. Under the conditions tested, 10 biomarkers were identified (4.1kDa, 4.4kDa, 7.7kDa, 12.9kDa, 15.1kDa, 15.9kDa, 18.9kDa, 23.0kDa, 30.1kDa, and 53.5kDa) and classified as “neoplasia biomarkers” (markers best at identifying benign and malignant samples). All markers, except marker 12.9kDa, showed increased expression in patients with ovarian cancer. At highest accuracy, the individual markers in the neoplasia biomarkers group had ROC area values ranging from 0.711 to 0.833, sensitivities from 60.6% to 84.0%, specificities from 52.2% to 89.1%, and accuracies from 67.1% to 78.5%. We next examined proteins that are differentially expressed in healthy individuals and individuals with benign tumors vs. patients with malignant ovarian tumors. Under the conditions tested, we identified 13 proteins (3.1kDa, 4.5kDa, 5.1kDa, 7.8kDa, 8.2kDa, 13.9kDa, 16.9kDa, 18.6kDa, 21.0kDa, 28.0kDa, 79.0kDa, 93.0kDa, and 106.7kDa) and grouped them as “malignant neoplasia biomarkers” (markers best at identifying malignant samples). Markers 3.1kDa, 4.5kDa, 5.1kDa, 7.8kDa, 8.2kDa, 16.9kDa, and 18.6kDa showed increased expression in patients with ovarian cancer, whereas markers 13.9kDa, 21.0kDa, 28.0kDa, 79.0kDa, 93.0kDa, and 106.7kDa showed decreased expression in patients with ovarian cancer. The individual proteins in the malignant biomarkers group had values for ROC area ranging from 0.617 to 0.851, sensitivities from 48.1% to 81.5%, specificities from 66.1% to 88.1%, and accuracies from 61.3% to 79.3%.

Biomarker Panels. We further performed multivariate analysis on both the neoplasia and malignant neoplasia groups, separately, to identify panels of biomarkers that will diagnose ovarian neoplasm (benign or malignant) or distinguish between benign and malignant ovarian tumors. The sas program, through a process of statistical logistic backward elimination to avoid “overfitting” bias, produced panels with the least redundant markers. Thus, these markers were excluded from the panel. Multivariate analysis of the 10 candidate ovarian neoplasia biomarkers resulted in an ovarian cancer screening biomarker panel (SBP) of five markers (4.4kDa, 15.9kDa, 18.9kDa, 23.0kDa, and 30.1kDa) with a collective ROC area (0.94; Fig. 2) higher than the best individual ovarian neoplasia diagnostic biomarker (0.83; 15.1 kDa). The sensitivity, specificity, and overall accuracy for the SBP were 95.7%, 82.6%, and 89.2%, respectively (Table 3 and Fig. 2B). Similarly, multivariate analyses of the 13 malignant neoplasia biomarkers yielded two independent panels: validation biomarker panel (VBP) I of five markers (3.1kDa, 13.9kDa, 21.0kDa, 79.0kDa, and 106.7kDa) and VBP II of four markers (5.1kDa, 16.9kDa, 28.0kDa, and 93.0kDa) with ROC area values (0.94 and 0.90; Table 3 and Fig. 3A and C) higher than the best individual malignant ovarian neoplasia diagnostic biomarker (0.85; 79.0kDa). The sensitivity, specificity, and overall accuracy values for the VBPs were 81.5%, 94.9%, and 88.2% for panel I and 72.8%, 94.9%, and 83.9% for panel II, when maximizing overall accuracy (Table 3 and Fig. 3 B and D).

Fig. 2.
ROC curves and plot of sensitivity and specificity for the ovarian cancer screening panel. ROC curve analysis was based on 140 patients to compare the diagnostic performance of five neoplasia biomarkers making up the screening panel (4.4kDa, 15.9kDa, ...
Fig. 3.
ROC curves and plot of sensitivity and specificity for the ovarian cancer validation panels I and II. Five malignant neoplasia biomarkers making up the validation panel I (3.1kDa, 13.9kDa, 21.0kDa, 79.0kDa, and 106.7kDa) and four markers representing ...
Table 3.
Statistical summary of biomarker panels

Screening and Validation of Test Samples. We tested the SBP and the VBPs on 44 blinded test samples [10 normal, 6 benign, 6 LMP, and 22 invasive ovarian cancers, and 11 early-stage (I/II) and 11 advanced-stage (III/IV) carcinomas]. For the ovarian cancer SBP, the test sensitivity (number benign and cancer correctly labeled positive/number true positive) at highest accuracy was 91.2%, the test specificity (number normal correctly labeled negative/number true negative) was 80%, and test accuracy was 85.6%. For early-stage (I/II), the SBP using the highest accuracy cutoff resulted in a sensitivity of 100% (identified 11 of 11), advanced-stage (III/IV) sensitivity of 100% (identified 11 of 11), benign tumor sensitivity of 50% (3 of 6), and a nontumor specificity of 80% (identified 8 of 10 nondiseased). For the ovarian cancer VBP I, the overall test sensitivity, specificity, and accuracy were 71.4%, 100%, and 85.7%, respectively, and resulted in test sensitivity for earlystage (I/II) of 54.5% (identified 6 of 11), and for advanced-stage (III/IV) of 72.7% (identified 8 of 11), benign tumor specificity of 100% (identified 6 of 6), and nontumor specificity of 100% (identified 10 of 10). Finally, for the ovarian cancer VBP II, the overall test sensitivity, specificity, and accuracy were 89.3%, 56.3%, and 72.8%, respectively, and resulted in test sensitivity for early-stage (I/II) of 90.9% (identified 10 of 11) and 81.8% for advanced-stage (III/IV; identified 9 of 11), with benign tumor specificity of 50% (identified 3 of 6) and nontumor specificity of 60% (identified 6 of 10). Predictions made with all three panels together, by using their score thresholds at highest accuracy, correctly diagnosed 41 of 44 test samples: 21 of 22 malignant carcinomas [10 of 11 early-stage (I/II), 11 of 11 advanced-stage (III/IV)], 6 of 6 LMP, 5 of the 6 benign tumors, and 9 of 10 normal patient samples.

Discussion

One of the greatest challenges confronting the gynecologic oncologist is a way to positively impact the survival of ovarian cancer patients through the prevention and early detection of the disease. Recent advances in technologies for identifying proteins in complex mixtures have stimulated considerable interest in biomarker discovery research.

The 5-year survival for patients with early-stage ovarian cancer (stage I/II) approaches 95%, but it is <20% for advanced-stage disease (stages III and IV) (15). In >85% of patients with ovarian cancer, the disease is already advanced at the time of initial diagnosis. More often than not, after conventional primary cytoreductive surgery and induction chemotherapy, treatment is followed by the emergence and outgrowth of chemotherapy-resistant disease in these patients (15). Therefore, it is not surprising that ovarian cancer is the number one cause of death among all of the gynecological malignancies in recent years (4). Most recently, ovarian cancer cell lines, ovarian cancer tissue, and serum from patients with ovarian cancer have become the focus of gene-expression profiling (transcriptional profiling) or protein-profiling studies to identify candidate novel tumor biomarkers that may serve as indicators of early-stage disease (1741). Using the SAX2 ProteinChip array, we have identified groups of ovarian cancer-associated proteins (biomarker protein panels) that are differentially expressed in normal serum vs. the serum from patients with ovarian cancer. Our preliminary data also demonstrated that similar biomarkers are expressed at the tissue level (data not shown).

Our goal is not only to apply these biomarker panels for the early diagnosis of ovarian neoplasia (benign or malignant), but also to distinguish benign from malignant disease. Hence, the overall objective of this article is the creation of biomarker protein panels for the early detection of ovarian cancer, stage I and II disease. In the clinical setting, the biomarker protein panels will ultimately be used to screen patients for the presence of any ovarian neoplasm in conjunction with routine pelvic examination and ultrasound evaluation. Once the diagnosis of an ovarian neoplasm is made, then the other biomarker panels will be used to differentiate the presence of benign from malignant disease.

We recreated the “clinical setting” in the laboratory to test the biomarker panels. The three biomarker protein panels were blindly tested against a set of separate serum samples, including those from normal subjects and from patients with benign or malignant tumors. Used together, the three ovarian cancer biomarker protein panels correctly diagnosed 21 of 22 invasive ovarian cancers [10 of 11 early-stage (I/II) and 11 of 11 advanced-stage (III/IV) carcinomas], 6 of the 6 tumors of LMP, 5 of the 6 benign tumors, and 9 of 10 normal samples. Hence, by using these biomarker protein panels, we were able to correctly identify patients with early-stage (stage I/II) ovarian cancer with high sensitivity, specificity, overall accuracy, and ROC values (screening panel = 95.7% sensitivity, 82.6% specificity, 89.2% accuracy, and 0.94 ROC; validation panel I = 81.5% sensitivity, 94.9% specificity, 88.2% accuracy, and 0.94 ROC; validation panel II = 72.8% sensitivity, 94.9% specificity, 83.9% accuracy, and 0.90 ROC).

Serum CA125 levels are the standard to which new screening markers for ovarian cancer have been compared (6, 42). However, we were not able to identify a biomarker that correlated in size to CA125. We speculated that CA125 may not bind the SAX2 chips used in our studies or that its large size is not conducive to ionization and flight. With a cutoff of 30–35 units/ml, serum CA125 has been shown to have a sensitivity of only 50–60% when its specificity is approaching 99% for early-stage disease in apparently healthy postmenopausal women (6, 42). Mok et al. (43) have reported CA125 sensitivity of 64.9% at a fixed specificity of 94%. Rai et al. (35) reported that the recommended cutoff of 35 units/ml for CA125 resulted in 65.6% sensitivity and 97.2% specificity, whereas a cutoff of 18.5 units/ml resulted in 81.3% sensitivity and 94.4% specificity.

Similar to using protein biomarker panels, combining individual markers has been used by other investigators as one strategy to increase the overall ovarian cancer detection rate (16, 44). According to Mok et al. (43), the combination of serum CA125 and prostasin resulted in 92% sensitivity and 94% specificity, compared with a sensitivity of 51.4% and specificity of 94% of prostasin alone in the same set of samples. Rai et al. (35) reported that the logistic regression of their two markers (60- and 79kDa) resulted in a sensitivity of only 59.5% and a specificity of 94.7%; however, when they combined their markers with CA125, they observed a sensitivity of 93.8% and specificity of 94.4%. Concurrent measurement of CA125 and soluble interleukin 2 receptor α in sera identified 100% of ovarian carcinoma stage I/II, with a 9% false positive rate (45). We are currently analyzing the CA125 levels in our study samples by ELISA to determine whether the combination of serum CA125 with our protein biomarkers will further increase the already relatively high observed values for sensitivity, specificity, overall accuracy, and ROC associated with our biomarker panels.

We are presently purifying the proteins comprising the three biomarker panels. Several investigators have reported the successful purification and identification of markers using tryptic peptide mapping (35) or amino acid sequencing (46). We are purifying by fractionation and SDS/PAGE, followed by tryptic digestion of each panel biomarker. If antibodies directed against the specific proteins are available, we will perform immunoprecipitation and analyze the immunoprecipitated protein on SAX2 ProteinChips as well as analyze the depleted serum to determine whether the biomarker has been removed from the profile. If purified proteins, isolated from mammalian cells, are available, we will compare pure protein profiles to the patient serum protein profiles. The comparison will determine whether the biomarker peak aligns with the pure protein peak. Low abundance and novel biomarkers may require additional steps in purification.

In conclusion, we have discovered three ovarian cancer biomarker protein panels that, when used together, significantly improved the early detection of ovarian neoplasia (benign and malignant) in our patient population and effectively distinguished serum samples from patients with benign vs. malignant ovarian neoplasia. Hence, using serum biomarker proteins, we were able to correctly identify patients with early-stage (stage I/II) ovarian cancer with high sensitivity, specificity, overall accuracy, and ROC values. Larger-scale studies addressing the efficacy of these and other markers, used either individually or in combination, for detecting early-stage ovarian cancer will be essential. The multiple protein differences observed between and within the ovarian cancer and the nonovarian cancer control groups in our study enhanced the role of protein profiling as a potential diagnostic and prognostic approach. Although these markers distinguish patients with ovarian tumors from normal patients, additional studies are required to compare the protein profiles obtained from ovarian cancer serum samples to protein profiles obtained from patients with other cancers and autoimmune disorders. Thus, we will determine which of our protein biomarkers are ovarian cancer-specific by comparing ovarian cancer profiles to profiles from other cancers (e.g., breast, colon, and endometrial) and to profiles from other disorders (e.g., inflammatory bowel disease, arthritis, and autoimmune diseases).

Acknowledgments

We thank Dr. Stella Redpath for assistance with SELDI-TOF-MS technology, Dr. Jeffery Gornbein for assistance with statistical analysis, and Amy Shah and Joe Gamboa for assistance with this article. Dr. Farias-Eisner is a National Institutes of Health Women's Reproductive Health Research Scholar (through Grant HD01281:03, which has partially funded this work). The Joan English Fund for Women's Cancer Research (ID no. 95213) and the Liz Tilberis Scholars Program (Grant UCLA0101) supported this work.

Notes

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: CA125, ovarian cancer antigen 125; LMP, low malignant potential; ROC, receiver operating characteristic; SBP, screening biomarker panel; SAX2, strong anion-exchange; SELDI-TOF-MS, surface-enhanced laser desorption/ionization time-of-flight MS; VBP, validation biomarker panel.

References

1. Farias-Eisner, R., Teng, F., Oliveira, M., Leuchter, R., Karlan, B., Lagasse, L. D. & Berek, J. S. (1994) Gynecol. Oncol. 55, 108–110. [PubMed]
2. Farias-Eisner, R., Kim, Y. B. & Berek, J. S. (1994) Semin. Surg. Oncol. 10, 268–275. [PubMed]
3. Morice, P., Brehier-Ollive, D., Rey, A., Atallah, D., Lhomme, C., Pautier, P., Pomel, C., Camatte, S., Duvillard, P. & Castaigne, D. (2003) Ann. Oncol. 14, 74–77. [PubMed]
4. Greenlee, R. T., Hill-Harmon, M. B., Murray, T. & Thun, M. (2001) CA Cancer J. Clin. 51, 15–36. [PubMed]
5. Coleman, M. P., Esteve, J., Damiecki, P., Arslan, A. & Renard, H. (1993) IARC Sci. Publ. 121, 1–806. [PubMed]
6. Bast, R. C., Jr., Urban, N., Shridhar, V., Smith, D., Zhang, Z., Skates, S., Lu, K., Liu, J., Fishman, D. & Mills, G. (2002) Cancer Treat. Res. 107, 61–97. [PubMed]
7. Bast, R. C., Jr., Klug, T. L., St. John, E., Jenison, E., Niloff, J. M., Lazarus, H., Berkowitz, R. S., Leavitt, T., Griffiths, C. T., Parker, L., et al. (1983) N. Engl. J. Med. 309, 883–887. [PubMed]
8. Menon, U. & Jacobs, I. (2000) Curr. Opin. Obstet. Gynecol. 12, 39–42. [PubMed]
9. Cohen, L. S., Escobar, P. F., Scharm, C., Glimco, B. & Fishman, D. A. (2001) Gynecol. Oncol. 82, 40–48. [PubMed]
10. Memarzadeh, S., Lee, S. B., Berek, J. S. & Farias-Eisner, R. (2003) Int. J. Gynecol. Cancer 13, 120–124. [PubMed]
11. Rubin, R. B. & Merchant, M. (2000) Am. Clin. Lab. 19, 28–29. [PubMed]
12. Weinberger, S. R., Dalmasso, E. A. & Fung, E. T. (2001) Curr. Opin. Chem. Biol. 6, 86–91.
13. Fung, E. T., Thulasiraman, V., Weinberger, S. R. & Dalmasso, E. A. (2001) Curr. Opin. Biotechnol. 12, 65–69. [PubMed]
14. Issaq, H. J., Veenstra, T. D., Conrads, T. P. & Felschow, D. (2002) Biochem. Biophys. Res. Commun. 292, 587–592. [PubMed]
15. Wright, G. L., Jr., Cazares, L. H., Leung, S.-M., Nasim, S., Adam, B.-L., Yip, T.-T., Schellhammer, P. F., Gong, L. & Vlahou, A. (1999) Prostate Cancer Prostatic Dis. 2, 264–276. [PubMed]
16. Li, J., Zhang, Z., Rosenzweig, J., Wang, Y. Y. & Chan, D. W. (2002) Clin. Chem. 48, 1296–1304. [PubMed]
17. Xu, Y., Shen, Z. Z., Wiper, D. W., Wu, M. Z., Morton, R. E., Elson, P., Kennedy, A. W., Belinson, J., Markman, M. & Casey, G. (1998) J. Am. Med. Assoc. 280, 719–723. [PubMed]
18. Scholler, N., Fu, N., Yang, Y., Ye, Z., Goodman, G. E. & Hellstrom, K. E. (1999) Proc. Natl. Acad. Sci. USA 96, 11531–11536. [PMC free article] [PubMed]
19. Martoglio, A. M., Tom, B. D., Starkey, M., Corps, A. N., Charnock-Jones, D. S. & Smith, S. K. (2000) Mol. Med. 6, 750–765. [PMC free article] [PubMed]
20. Meinhold-Heerlein, I., Stenner-Liewen, F., Liewen, H., Kitada, S., Krajewska, M., Krajewski, S., Zapata, J. M., Monks, A., Scudiero, D. A., Bauknecht, T. & Reed, J. C. (2001) Am. J. Pathol. 158, 1335–1344. [PMC free article] [PubMed]
21. Hough, C. D., Cho, K. R., Zonderman, A. B., Schwartz, D. R. & Morin, P. J. (2001) Cancer Res. 61, 3869–3876. [PubMed]
22. van Haaften-Day, C., Shen, Y., Xu, F., Yu, Y., Berchuck, A., Havrilesky, L. J., de Bruijn, H. W., van der Zee, A. G., Bast, R. C., Jr., & Hacker, N. F. (2001) Cancer 92, 2837–2844. [PubMed]
23. Welsh, J. B., Zarrinkar, P. P., Sapinoso, L. M., Kern, S. G., Behling, C. A., Monk, B. J., Lockhart, D. J., Burger, R. A. & Hampton, G. M. (2001) Proc. Natl. Acad. Sci. USA 98, 1176–1181. [PMC free article] [PubMed]
24. Wong, K. K., Cheng, R. S. & Mok, S. C. (2001) BioTechniques 30, 670–675. [PubMed]
25. Mills, G. B., Bast, R. C., Jr., & Srivastava, S. (2001) J. Natl. Cancer Inst. 93, 1437–1439. [PubMed]
26. Xu, S., Mou, H., Lu, G., Zhu, C., Yang, Z., Gao, Y., Lou, H., Liu, X., Cheng, Y. & Yang, W. (2002) Chin. Med. J. (Engl. Ed.) 115, 36–41. [PubMed]
27. Mills, A. P., Jr. (2002) Trends Biotechnol. 20, 137–140. [PubMed]
28. Alaiya, A. A., Franzen, B., Hagman, A., Dysvik, B., Roblick, U. J., Becker, S., Moberger, B., Auer, G. & Linder, S. (2002) Int. J. Cancer 98, 895–899. [PubMed]
29. Jones, M. B., Krutzsch, H., Shu, H., Zhao, Y., Liotta, L. A., Kohn, E. C. & Petricoin, E. F., III (2002) Proteomics 2, 76–84. [PubMed]
30. Wu, X., Li, H., Kang, L., Li, L., Wang, W. & Shan, B. (2002) Gynecol. Oncol. 84, 126–134. [PubMed]
31. Diamandis, E., Okuri, A., Mitsui, S., Luo, L.-Y., Soosaipillai, A., Grass, L., Nakamura, T., Howarth, D. & Yamaguchi, N. (2002) Cancer Res. 62, 295–300. [PubMed]
32. Kristiansen, G., Denkert, C., Schluns, K., Dahl, E., Pilarsky, C. & Hauptmann, S. (2002) Am. J. Pathol. 161, 1215–1221. [PMC free article] [PubMed]
33. Fishman, D. A. & Bozorgi, K. (2002) Cancer Treat. Res. 107, 3–28. [PubMed]
34. Smith, D. I. (2002) Cytometry 47, 60–62. [PubMed]
35. Rai, A., Zhang, Z., Rosenzweig, J., Shih, L., Pham, T., Fung, E., Sokoll, L. & Chan, D. (2002) Arch. Pathol. Lab. Med. 126, 1518–1526. [PubMed]
36. Kim, J. H., Skates, S. J., Uede, T., Wong, K. K., Schorge, J. O., Feltmate, C. M., Berkowitz, R. S., Cramer, D. W. & Mok, S. C. (2002) J. Am. Med. Assoc. 287, 1671–1679. [PubMed]
37. Petricoin, E. F., Ardekani, A. M., Hitt, B. A., Levine, P. J., Fusaro, V. A., Steinberg, S. M., Mills, G. B., Simone, C., Fishman, D. A., Kohn, E. C. & Liotta, L. A. (2002) Lancet 359, 572–577. [PubMed]
38. Yousef, G. M., Polymeris, M. E., Yacoub, G. M., Scorilas, A., Soosaipillai, A., Popalis, C., Fracchioli, S., Katsaros, D. & Diamandis, E. P. (2003) Cancer Res. 63, 2223–2227. [PubMed]
39. Diamandis, E. P., Scorilas, A., Fracchioli, S., Van Gramberen, M., De Bruijn, H., Henrik, A., Soosaipillai, A., Grass, L., Yousef, G. M., Stenman, U. H., et al. (2003) J. Clin. Oncol. 21, 1035–1043. [PubMed]
40. Lou, L. Y., Katsaros, D., Scorilas, A., Fracchioli, S., Bellino, R., van Gramberen, M., De Bruijn, H., Henrik, A., Stenman, U. H., Massobrio, M., et al. (2003) Cancer Res. 63, 807–811. [PubMed]
41. Welsh, J. B., Sapinoso, L. M., Kern, S. G., Brown, D. A., Liu, T., Bauskin, A. R., Ward, R. L., Hawkins, N. J., Quinn, D. I., Russell, P. J., et al. (2003) Proc. Natl. Acad. Sci. USA 100, 3410–3415. [PMC free article] [PubMed]
42. Bast, R. C., Jr., Xu, F. J., Yu, Y. H., Barnhill, S., Zhang, Z. & Mills, G. B. (1998) Int. J. Biol. Markers 13, 179–187. [PubMed]
43. Mok, S. C., Chao, J., Skates, S., Wong, K., Yiu, G. K., Muto, M. G., Berkowitz, R. S. & Cramer, D. W. (2001) J. Natl. Cancer Inst. 93, 1458–1464. [PubMed]
44. Vlahou, A., Schellhammer, P. F., Mendrinos, S., Patel, K., Kondylis, F. I., Gong, L., Nasim, S. & Wright, G. L., Jr. (2001) Am. J. Pathol. 158, 1491–1502. [PMC free article] [PubMed]
45. Sedlaczek, P., Frydecka, I., Gabrys, M., van Dalen, A., Einarsson, R. & Harlozinska, A. (2002) Cancer 95, 1886–1893. [PubMed]
46. Klade, C. S., Voss, T., Krysteck, E., Ahorn, H., Zatloukal, K., Pummer, K. & Adolf, G. R. (2001) Proteomics 1, 890–898. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...