Introduction
Epithelial ovarian cancer (EOC) is a highly malignant gynecological neoplasia with an incidence of 12/100 000 women [
1], and this rate has only slightly decreased in the last 80 years. While women of any age are at risk for this malignancy, postmenopausal women have a higher incidence. For example, 90% of women who suffer from EOC are older than 40 years of age, and the greatest number are 55 years or older. Moreover, due to the anatomic position of the ovaries, pelvic malignancies can remain obscured. In addition, a lack of symptoms until the advanced stages of tumor development results in an increased rate of metastasis at the time of diagnosis. Correspondingly, 80% of the ovarian neoplasias diagnosed are stage III-IV according to International Federation of Obstetrics and Gynecology (FIGO) criteria. While FIGO stage I EOC has a relatively high five-year survival rate (> 90%), survival rates markedly drop for patients with stage III-IV EOC (25–30%) [
2,
3]. Therefore, it is crucial to diagnose EOC as early as possible. Accordingly, the identification of serum biomarkers to detect EOC would represent an important and valuable advance for the monitoring and treatment of EOC progression.
Algorithms and triage protocols designed to evaluate potential cases of ovarian cancer in their early stages are currently limited, and rely on pelvic sonography and CA-125 determination [
4]. Moreover, the sensitivity and specificity of these approaches range from 70–80%. [
5]. Regarding CA-125, its levels are elevated in less than 50% of EOC cases, and it is undetectable in another 20% of EOC cases. In addition, high serum levels of CA-125 are also associated with benign gynecological diseases (e.g., cysts, endometriosis, etc.) [
6]. In 2008, Moore and colleagues identified human epididymis protein 4 (HE4) as a biomarker for ovarian cancer [
7]. Based on these findings, a risk of ovarian malignancy algorithm (ROMA) was developed, and is currently used to predict the presence of malignant ovarian cancer using a combination of CA-125 levels, HE4 expression, and menopausal status. In particular, the combination of HE4 and CA-125 in the ROMA has been associated with a higher sensitivity than any single biomarker [
8].
14-3-3 zeta is an important regulatory protein, which mediates intracellular signaling pathways by interfering with approximately 100 cellular proteins, including oncogenes and protooncogenes. Recently, two independent research groups, Waldemarson et al. [
9] and He et al
.
[
10], advocated 14-3-3 zeta as a potential biomarker for EOC. In addition, Kobayashi et al. [
11] recently demonstrated that 14-3-3 zeta protein is present in malignant ascites of patients with EOC, and is secreted by ascetic monocytes and macrophage. However, while the role of 14-3-3 zeta protein as an intracellular adaptor protein has been widely investigated, the function of the secreted protein is unclear. Therefore, the goal of the current pilot study was to assess the potential for 14-3-3 zeta protein to serve as a biomarker for monitoring patients with FIGO stage II-III EOC that undergo chemotherapy.
Discussion
Several studies have demonstrated the limitations associated with depending on any single tumor marker for the detection of EOC. Initially, CA-125 was widely used. However, other malignant and benign diseases also express CA-125, thereby limiting its reliability as a tumor marker. In particular, CA-125 has a high false-positive rate among women with benign gynecological conditions such as endometriosis [
14], and a low sensitivity in identifying patients with early-stage ovarian cancer [
15]. Accordingly, when EOC is diagnosed, 80% of cases are in an advanced stage of disease (e.g., FIGO III-IV) [
16]. To improve the specificity and sensitivity of an ovarian cancer diagnosis, additional tumor markers have been investigated. One novel tumor marker is HE4, which contains two whey acid protein (WAP) domains and eight cysteine residues that constitute a four-disulphide bond core [
17]. HE4 localizes to human chromosome 20q12-13.1 and its expression significantly increases during malignant transformation. However, HE4 is expressed in normal tissues as well, and therefore, is not tumor specific. Correspondingly, it has been hypothesized that the function of HE4 is related to both spermiotelcosis (a protease inhibitor involved in sperm maturation) and natural immunity, although the mechanistic details of HE4 functions remain to be clarified [
18]. As a tumor marker for the early detection of ovarian cancer, Moore et al. reported a sensitivity of 72.9% and a specificity of 95% for HE4 [
7]. Moreover, when both HE4 and CA-125 were detected, the sensitivity increased to 76.4%. Therefore, the detection of more than one biomarker resulted in a 33.1% increase in the sensitivity of CA-125, and a 3.5% increase in HE4 sensitivity [
7].
In the present study, ROMA values provided a PI based on the pre- or postmenopausal status of a patient, and the presence and levels of biomarkers CA-125 and HE4. As such, this PI relies on an accurate determination of serum levels of HE4 and CA-125. Moreover, in a recent study, the ROMA was found to be more effective in predicting ovarian cancer than the widely used risk of malignancy index (RMI), which employs ultrasound findings, CA-125 concentrations, and menopausal status [
19]. Furthermore, when the specificity was set to 75%, the RMI had a sensitivity of 84.6%. For the same specificity, the sensitivity of the ROMA was significantly higher (94.3%). Although biomarker concentrations can be assayed by various methods (e.g., ELISA, chemiluminescent microparticle immunoassay), a recent study conducted by Ruggeri et al. demonstrated that chemiluminescent immunoassays are more adequate and more reproducible than commercially available ELISA kits that are characterized by interassay imprecision percentages (CV%) ranging from 6.8-10.3%, compared to < 4% for ECLIA [
20]. The results of the present study are consistent with these findings, and they further support the use of the ECLIA method for routine determinations of CA-125 and HE4 levels. Furthermore, the deviation in accuracy for ELISA versus ECLIA can be attributed to the fully automated format of ECLIA, while ELISAs are manual assays that also require testing samples in duplicate.
14-3-3 zeta protein plays an important role in several different biological mechanisms. For example, it has been reported to be an adaptor protein for intracellular signaling since it contains tandem repeats of phosphoserine motifs that have the capacity to bind upstream and downstream signaling molecules [
21‐
24]. 14-3-3 zeta protein also facilitate cell migration by forming a ternary complex with integrin alpha-4 and paxillin [
23]. However, 14-3-3 zeta also has potential roles in cancerogenesis, based on its ability to bind NF-kappa B, beta-catenin, and Bcl-2, and to augment cancer cell proliferation [
25]. Furthermore, 14-3-3 zeta protein has been shown to block activation of p38 mitogen-activated protein kinase (MAPK), thereby mediating an anti-apoptotic mechanism [
26]. Numerous investigations have also suggested that 14-3-3 zeta protein is a key molecule in the malignant pathological processes of several malignancies, including oral, esophageal, lung, and breast cancers, as well as B cell lymphoma. Recently, He et al. reported that 14-3-3 zeta protein represents a candidate biomarker and a metastasis-promoting factor in ovarian cancer based on a serum proteomic analysis of a nude mouse xenograft model containing SKOV-3 cells and a mass spectrometry [liquid chromatography-tandem mass spectrometry (LC-MS/MS)] analysis to identify metastasis-related serum proteins [
10]. Significantly higher expression of 14-3-3 zeta was detected in EOC patients than in patients with benign gynecological diseases. Furthermore, compared to CA-125, serum levels of 14-3-3 zeta protein was significantly upregulated when microscopic peritoneal metastasis was present, or when bilateral ovaries were involved. Accordingly, the authors suggested that 14-3-3 zeta protein may be a useful tool in differentiating FIGO stage Ib and Ic ovarian cancers from stage Ia ovarian cancers in the clinic [
10]. However, the results of the present study are not consistent with these findings. For example, significant differences in the serum levels of 14-3-3 zeta protein was not detected in healthy menopausal women versus patients with advanced stage EOC. Furthermore, significant changes in serum levels of 14-3-3 zeta protein was not detected during the six consecutive cycles of chemotherapy treatment that were administered (Figure
2/G), although CT scans and CA-125 and HE4 levels unambiguously indicated the efficacy of the treatment. A possible explanation for these results is the insufficient number of patients enrolled in the current study. Thus, future studies should include a larger cohort in order to identify statistically significant changes. It is also possible that serum proteins may undergo degradation, even when stored at −80°C. In particular, it may be that 14-3-3 zeta is an unstable protein that needs to be assayed shortly after collection. Furthermore, an intriguing possibility is that 14-3-3 zeta may bind proteins activated by chemotherapeutic agents, or present as a result of chemotherapy, thereby obscuring detection of 14-3-3 zeta protein in serum. In the future s large-scale clinical investigation is necessary to evaluate the efficacy of 14-3-3 zeta protein, and to determine the sensitivity and the specificity of this biomarker comparing it to CA-125 and to HE4.
In conclusion, determination of CA-125 and HE4 serum levels for the ROMA represents a useful tool for the prediction of chemotherapy efficacy for EOC patients. However, based on our current findings, levels of 14-3-3 zeta protein were not found to reliably correlate with the clinical behavior of EOC, and therefore we question if it would be a useful biomarker for this disease.
Acknowledgments
This study was funded by OTKA K104960 (Szereday, L). We thank the laboratory assistants in the Department of Medical Microbiology and Immunology and the Institute of Laboratory Medicine for technical assistance. We also thank Jozsef Bodis, M.D., Dsc, Gabor L. Kovacs, M.D., Dsc, and Istvan Drozgyik, M.D., PhD for their helpful comments and discussion.
Competing interests
The authors have no conflict of interest with the present study to report.
Authors’ contributions
IH has made substantial contributions to the conception and the design of this study, he also contributed significantly by collecting ovarian cancer samples. Furthermore he is responsible for all corrections and finalizations made to the manuscript.PG participated in the study design and coordination.TK along with LS and BP performed all the necessary immunoassays and biomarker measurements. AK performed the radiological assesments of all patients and is responsible for the interpretation within the manuscript of the radiological findings.NF performed all the statistical analysis necessary. BF participated in the design of the study, collection of samples and in addition,he has also been involved in drafting the manuscript. All authors read and approved the final manuscript and have given final approval of the version to be published.