Serum folate receptor alpha (FRA) as a biomarker for ...

Serum Folate Receptor Alpha (FRA) as a Biomarker for Ovarian Cancer: Implications for Diagnosis, Prognosis and Predicting its Local Tumor Expression

Akira Kurosaki1,3, Kosei Hasegawa1,3*,Tomomi Kato2, Kenji Abe4, Tatsuya Hanaoka1,3, Akiko Miyara3, Daniel J O’Shannessy5, Elizabeth B Somers5, Masanori Yasuda2, Tetsuo Sekino4 and Keiichi Fujiwara1, 3

Departments of 1Gynecologic Oncology and 2Pathology, Saitama Medical University International Medical Center, 3

Gynecologic Oncology Translational Research Unit, Project Research Division,

Research Center for Genomic Medicine, Saitama Medical University 4

Department of Research and Development, EIDIA Co.Ltd.

5

Department of Translation Medicine and Diagnostics, Morphotek, Inc.

Short title: Serum FRA as a Biomarker for Ovarian Cancer

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as an ‘Accepted Article’, doi: 10.1002/ijc.29937 This article is protected by copyright. All rights reserved.

Page 3 of 39

International Journal of Cancer Serum FRA as a Biomarker for Ovarian Cancer

*Address correspondence to: Kosei Hasegawa,M.D.,Ph.D. Department of Gynecologic Oncology, Saitama Medical University International Medical Center, 1397-1 Yamane, Hidaka-shi, Saitama 350-1298, Japan Phone: +81-42-984-4111 Fax: +81-42-984-4743 E-mail: koseih@saitama-med,ac,jp

Key words: ovarian cancer, folate receptor alpha, serum biomarker, FRA, diagnostic Abbreviations: FRA, folate receptor alpha; sFRA, serum folate receptor alpha; FFPE, formalin fixed paraffin emended; MAbs, monoclonal antibodies; ELISA, Enzyme-Linked ImmunoSorbent Assay; ROC, receiver operator characteristics; EOC, epithelial ovarian cancer; progression free survival, PFS;

Category: Research Articles

[2] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.


Brief description: Folate receptor alpha (FRA) is considered an attractive therapeutic target for ovarian cancer. We prospectively investigated the levels of soluble FRA in the serum of patients suspected of having malignant ovarian tumors. We found serum FRA (sFRA) to be a specific diagnostic marker for ovarian cancer and further demonstrated a correlation between serum FRA and local tumor FRA expression, which suggests potential application for this non-invasive test in the area of FRA-targeted therapies.


Page 4 of 39

Page 5 of 39


Abstract Folate receptor alpha (FRA) is a GPI-anchored glycoprotein and encoded by the FOLR1 gene. High expression of FRA is observed in specific malignant tumors of epithelial origin, including ovarian cancer, but exhibits very limited normal tissue expression, making it as an attractive target for the ovarian cancer therapy. FRA is known to shed from the cell surface into the circulation which allows for its measurement in the serum of patients. Recently, methods to detect the soluble form of FRA have been developed and serum FRA (sFRA) is considered a highly promising biomarker for ovarian cancer. We prospectively investigated the levels of sFRA in patients clinically suspected of having malignant ovarian tumors. A total of 231 patients were enrolled in this study and analyzed for sFRA as well as tumor expression of FRA by immunohistochemistry. High sFRA was predominantly observed in epithelial ovarian cancer patients, but not in patients with benign or borderline gynecological disease or metastatic ovarian tumors from advanced colorectal cancers. Levels of sFRA was highly correlated to clinical stage, tumor grade and histological type and demonstrated superior accuracy for the detection of ovarian cancer than did serum CA125. High sFRA was significantly associated with shorter progression-free survival in both early and advanced ovarian cancer patients. Finally, tumor FRA expression status was strongly [4] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.


correlated with sFRA levels. Taken together these data suggest sFRA might be a useful non-invasive serum biomarkers for future clinical trials assessing FRA-targeted therapy.


Page 6 of 39

Page 7 of 39


Introduction Epithelial ovarian cancer (EOC) is the most lethal gynecologic malignancy. Because of difficulties in detection, diagnosis and treatment, the overall survival rate of ovarian cancer patients is still poor. It is therefore necessary to overcome these difficulties. A majority of cases (61%) of EOC are diagnosed at an advanced stage, with a corresponding 5-year survival rate of only 27%1. The fact that survival of patients with stages I–II disease ranges from 60%–90%, depending on tumor grade, suggests the potential for a high cure rate with earlier disease detection2 3. Since physical symptoms are absent in early stages of ovarian cancer, efforts are being made to develop assays for blood or tissue biomarkers. Although the serum CA125, an ovarian cell surface glycoprotein of unknown biological significance, is elevated in 80% of patients with advanced epithelial ovarian cancer, this marker has a positive predictive value of only 10% in early stage disease3. One of the limitations for CA125 is its relatively low specificity. Elevated CA125 is often associated with various non-malignant conditions, such as pregnancy, endometriosis, adenomyosis, uterine fibroids, pelvic inflammatory disease, menstruation, and benign ovarian cysts. Abnormal CA125 is also associated with other malignant conditions such as pancreatic, breast, lung, gastric, and colorectal cancers especially when they are associated with peritoneal spreading4 5 6 7 8. Although [6] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.


CA125 is helpful in the follow-up of the patient’s chemotherapy and in detecting early relapse in patients with already known ovarian cancers, it is not useful for identifying early stage disease or differential diagnosis of the above-mentioned diseases3. Folate receptor alpha (FRA) is a 38-40 kDa molecule, with high affinity for folic acid and its derivatives 9. FRA is anchored to cell membranes through a glycosylphosphatidylinositol moiety and transports folates into the cell via an endocytic process.

Elevated expression of FRA has been observed in various types of cancers,

including ovarian, uterine, endometrial, lung, breast carcinoma, high grade osteosarcoma, and pleural mesothelioma3 10 11 12 13 14 15 16 . FRA expression in normal tissues is restricted to the apical surfaces of polarized epithelial cells where it is inaccessible to circulating cytotoxic drugs17. Hence, there is much interest to use FRA as a target for tumor-specific killing using various types of anticancer therapeutics. FRA is currently under investigation as a diagnostic and therapeutic target both clinically and preclinically17, 18 19 20 21 22 23 24. There are two basic strategies to targeting FRA with therapeutic intent: the first is based on anti-folate receptor antibodies such as farletuzumab 18 19 or the antibody-drug conjugate IMGN853 25, and the second is based on folate-chemotherapy conjugates such as vintafolide/etarfolatide17 26 27. Both strategies have been investigated in Phase III clinical trials. [7] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.

Page 8 of 39

Page 9 of 39


In addition to its cell surface localization, FRA can be shed from the cell surface into the blood stream as a soluble form 17 28.

O’Shannessy et al have developed

electrochemiluminescent assays for detecting the soluble form of FRA and reported that sFRA is a potential biomarker to discriminate between serous ovarian cancers and normal controls10. Further study of the prognostic/diagnostic value of sFRA biomarker in a prospective clinical setting with precise patient information is warranted. In the current study, we prospectively investigated the levels of soluble FRA in the serum of patients who were clinically suspected to have malignant ovarian tumors prior to surgery. We evaluated the sFRA as a diagnostic marker for ovarian cancer and compared it to CA125. Further, we performed immunohistochemical staining for tumor FRA expression to investigate the relationship between sFRA and tumor FRA status to evaluate the possible application of this non-invasive test in future clinical studies of FRA-targeted therapies.



Materials and Methods Patients, sera and tumor specimens This was a prospective study reviewed and approved by the Institutional Review Board of Saitama Medical University International Medical Center. Written and oral informed consent was obtained from all participants. Two hundred and thirty-one (231) women, who were clinically suspected as having borderline to malignant ovarian tumors prior to surgery at Saitama Medical University International Medical Center from December 2010 to March 2013, were enrolled in this study. All patients received surgical operations. The mean age of the patients was 54.8 years with a range of 20 to 83 years. A consort diagram of all patients is shown in Figure 1. One hundred and thirty four (58.0%), 40 (17.3%) and 43 (18.6%) patients were pathologically diagnosed as malignant, borderline and benign ovarian tumors, respectively. Four patients were diagnosed as ovarian metastasis of a primary colorectal cancer. Of 134 malignant ovarian tumors, 128 tumors were diagnosed as epithelial ovarian cancer. There were 61 (47.7%) serous, 5 (3.9%) mucinous, 23 (18%) endometrioid and 34 (26.6%) clear cell carcinomas. The epithelial ovarian cancer cases consisted of 39 (30.5%) stage I, 22 (17.2%) stage II, 54 (42.2%) stage III and 13 (10.2%) stage IV. Formalin fixed paraffin emended (FFPE) and frozen tissue specimens were collected at [9] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.

Page 10 of 39

Page 11 of 39


the time of surgery, and stored at room temperature and -80 , respectively. Prior to surgery, serum was collected from all patients, and stored at -80

until use. Serum

CA125 was measured in the same serum sample in our central clinical laboratory. Clinical and histopathological information was obtained from clinical charts and pathology reports. Clinical stage of the tumor was reviewed based on the International Federation of Gynecology and Obstetrics (FIGO) staging system.

Histologic type and

grade were reviewed by pathologists experienced in the diagnosis of gynecologic oncology.

Measurement of sFRA We measured sFRA using a lab-developed two-step enzyme linked immunosorbent assay (ELISA). Two different human FRA specific mouse monoclonal antibodies 29 were used as capture and detection for the ELISA. Pre-diluted (1:9) 100 µL of the patients’ serum, calibrators or controls was added in duplicate and incubated at 4°C for eighteen-hours. After the plates were washed five times, 100 µL of HRP labeled antibody was added, and the plates were incubated one hour at room temperature. After the plates were washed five times, the color reaction was developed by the addition of 100 µL of TMB Substrate (KPL, Kirkegaard and Perry Laboratories) to each well and [10] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.


incubation at room temperature for 15 min. After stopping the reaction with 100 µL of 1mol/L hydrochloric acid solution, plates were measured spectrophotometrically at 450 nm and corrected for the absorbance at 620 nm. Two serum pools prepared by addition of recombinant FRA to level of 1400 and 2800 pg/mL were used as daily assay controls. The ELISA had a dynamic range of 250 to 8000 pg/mL and a signal-to-noise ratio of 3.4 for the lowest calibrator (250 pg/mL). The assay was shown to be linear with good spike-recovery of 99.5 and 103.8% on two separate serum samples. The intra-assay reproducibility using three human serum samples demonstrated a coefficient of variation of 2.3, 2.6 and 3.8 %.

Immunohistochemical staining We investigated the tumor FRA expression on FFPE specimens using a specific antibody NCL-L-FRalpha (clone BN3.2; Leica Biosystems, UK) on a total of 164 epithelial ovarian cancers, including borderline malignancies. Immunohistochemical staining was performed on full section slides using a BenchMark XT automated slide stainer and an iView DAB detection kit according to manufacturer’s instructions (Ventana Medical Systems Inc., Tucson, AZ). Sections were deparaffinized, hydrated and antigen retrieval with Cell Conditioning Solution 1 (Ventana Medical Systems Inc.), [11] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.

Page 12 of 39

Page 13 of 39


followed by the mouse monoclonal anti-FRA antibody (1:40 dilution) incubation for 30 minutes. Positive staining was defined as the presence of membranous or membranous with cytoplasmic stain in tumor cells. Slides were scored as < 1% positive tumor cells staining as negative, 1-25% positive tumor cells staining as weak (1+), 25-75% staining as moderate (2+), >75% staining as strong (3+). The scoring of FRA staining was evaluated in a blinded fashion by two independent observers. Any discrepancies were resolved by joint review over a double-headed microscope.

Statistical analysis Differences between the groups of patients were assessed by one-way ANOVA (with Tukey’s post hoc analysis), Student’s t test, and Mann-Whitney U-test. The receiver operator characteristic (ROC) curves display the trade-off between sensitivity and specificity for serum FRA differentiating between groups of patients. Cut-off values for sFRA and CA125 were calculated with Youden’s Index30, the maximum value of [Sensitivity + Specificity-1], calculated from all points of an ROC curve, and used as a criterion for selecting the optimum cut-off point. Survival was assessed using the Kaplan-Meier method and log-rank test for significance of differences between groups. Statistical analysis were performed using GraphPad Prism 6 (GraphPad Software, San [12] John Wiley & Sons, Inc. This article is protected by copyright. All rights reserved.


Diego, California, USA). All reported p-values were two-sided, and a value of p