Diagnostic value of serum glutathione peroxidase 3 levels in patients with lung cancer

Abstract

Background

We selected glutathione peroxidase 3 (GPx3) as a specific candidate that is down regulated in patients with lung cancer. In this study, we examined the diagnostic value of serum GPx3, which is an extracellular protein and readily detectable in blood.

Methods

We collected serum samples from 342 patients with lung cancer and 126 controls (normal healthy people and patients with benign diseases or other malignancies). We measured serum GPx3 levels using the enzyme-linked immunosorbent assay.

Results

Mean serum GPx3 levels were significantly lower in the patient group compared with the control group (10.1 ± 5.0 μg/mL vs. 13.0 ± 5.8 μg/mL, P < 0.001). In addition, mean serum GPx3 levels tended to be lower in the patients without metastasis compared with those with metastasis (9.6 ± 4.5 μg/mL vs. 10.7 ± 5.7 μg/mL, P = 0.051). Furthermore, mean serum GPx3 levels had a significant difference according to initial treatments (P < 0.001). In other words, mean serum GPx3 levels were significantly lower in the surgery group (8.2 ± 4.1 μg/mL) compared with the concurrent chemoradiotherapy (11.5 ± 4.6 μg/mL, P < 0.001), chemotherapy (10.7 ± 5.6 μg/mL, P < 0.001), and supportive care groups (10.9 ± 4.8 μg/mL, P = 0.002).

Conclusion

Our results showed that serum GPx3 levels were significantly lower in the patients who underwent surgery, which indicates that the serum may have diagnostic value in patients at an operable stage of lung cancer, rather than those at a locally advanced or metastatic stage.

Introduction

Lung cancer has been a leading cause of cancer-related death worldwide in both men and women.^1–3 Most lung cancer patients present with advanced disease, for whom there is a limited availability of standard treatment modality. Therefore, the overall five-year survival rate amounts to 10–15%. In an effort to improve the survival rate in patients with lung cancer, both basic and clinical studies have been conducted. Thus, attempts have been made to develop reliable biological and molecular markers for the early detection, diagnosis, and prediction of lung cancer.^4–9

Glutathione peroxidase 3 (GPx3) is one of glutathione (GSH) peroxides that protect cells from oxidative damage generated by hydrogen peroxide (H₂O₂) and oxidizing GSH.¹⁰ The GPx families are expressed in nearly all mammalian tissues that generate H₂O₂. Most of them are intercellular proteins, but GPx3 is an extracellular secreted protein that is abundantly present in plasma.¹¹

Recent studies have shown that the down-regulation of GPx3 is closely associated with the progression of many types of cancer.^12–14 In addition, GPx3 methylation or deletion, which lowers the degree of GPx3 expression, are prevalently seen in such cancers.^15,16 GPx3 has selenocysteine that is encoded by UGA opal codon;¹⁷ selenium (Se) plays a role in converting GPx3 to a full-length and functional protein. It has been reported that whole blood and serum levels of Se and the degree of glutathione peroxidase (GSH-Px) activity are both significantly lower in patients with lung cancer compared with normal healthy controls.¹⁸ Moreover, the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) profile study (http://www.ncbi.nlm.nih.gov/geo/) showed that the GPx3 mRNA level was significantly lower in patients with lung cancer. A recent study has also shown that GPx3 methylation may have implications in chemotherapy response and clinical outcome in patients with cancer.¹⁵

Given the above background, we have selected GPx3 as a specific candidate that is down regulated in patients with lung cancer and conducted this study to examine the diagnostic value of serum GPx3 levels on the enzyme-linked immunosorbent assay (ELISA) in 342 patients with lung cancer.

Material and methods

Serum sample collection

In the current study, we collected serum samples from 342 patients with lung cancer before the initiation of treatment and 126 controls (normal healthy people and patients with benign diseases or other malignancies). The biospecimen and data were provided by the Korea Biobank Network and collected between February 2009 and February 2012. Blood samples were collected in BD Vacutainer SS Plus Blood Collection Tubes (BD Biosciences, USA). To collect serum, samples were centrifuged using Rotina 380R centrifuge (Hettich, Germany) at 3000 rpm at 4°C for 20 minutes. The samples were then stored at −180°C until further laboratory procedure. The Institutional Review Board (IRB) of Chonnam National University Hwasun Hospital approved the study (IRB approval number: HCRI 12045-3). Informed consent was waived because of the retrospective nature of the study.

Enzyme-linked immunosorbent assay (ELISA)

The serum levels of GPx3 were determined by a commercially available ELISA (enzyme-linked immunosorbent assay) kit (AdipoGen, Inc., Seoul, Korea). Briefly, each 96-well plate was coated with a polyclonal antibody specific for human GPx3. Each serum sample was diluted at a ratio of 1:250 with a diluent buffer, 100 μL of which was added to the wells. This was followed by a one hour incubation at 37°C. After the plate was washed three times, it was added with 100 μL of detection antibody. This was followed by another one hour incubation at 37°C. The plate was then washed three times again, and it was added with 100 μL of detection antibody. This was followed by further one hour incubation at 37°C. After the plate was washed five times, it was added with 100 μL of substrate solution. This was followed by a 20-minute incubation at room temperature. The plate was then read at a wavelength of 450 nm using the VERSAmax microplate reader (Molecular Devices, Sunnyvale, CA). Data analysis was performed using SOFTMax Pro version 5 software (Molecular Devices, Sunnyvale, CA). All of the experimental procedures were performed in duplicate, and measurements were averaged.

Statistical analysis

All ELISA data was expressed as mean ± standard deviation (SD). We used the student's t-test and one-way analysis of variance (ANOVA) to analyze continuous variables. The area under curve (AUC) was estimated by receiver operating characteristics curve (ROC) analysis. All tests were two-sided, and a P-value of <0. 05 was considered statistically significant. Statistical analysis was performed using SPSS version 20.0 (SPSS Inc., Chicago, IL).

Results

Baseline demographics characteristics

Baseline and demographic characteristics are represented in Table 1. The median age of the 342 patients with lung cancer (n = 342) was 68 years (range, 25–87 years) and most of our clinical series of patients were male (n = 265, 77.5%) and smokers (n = 256, 74.9%). In our series, initial treatments included surgery (n = 113, 33.0%), radiation or concurrent chemoradiotherapy (CCRT, n = 81, 23.7%), chemotherapy (n = 106, 31.0%), and supportive care only (n = 42, 12.3%). Supportive care means that patients received medication to alleviate cancer-related symptoms in advanced or locally advanced stages.

Table 1.

Demographics of lung cancer patients (N = 342)

Characteristics	Value (%)
Age (years)
Median	68
Range	25–87
Gender
Male	265 (77.5)
Female	77 (22.5)
Smoking
Current smoker	98 (28.7)
Ex-smoker	158 (46.2)
Never smoker	86 (25.1)
Histology
Adenocarcinoma	156 (45.6)
Squamous cell carcinoma	136 (39.8)
Non-small cell lung cancer, not otherwise specified	19 (5.6)
Small cell lung cancer	31 (9.1)
Stage
IA/IB	44 (12.9)/29 (8.5)
IIA/IIB	14 (4.1)/18 (1.8)
IIIA/IIIB	62 (18.1)/43 (12.6)
IV	112 (32.7)
Limited disease	11 (3.2)
Extensive disease	20 (5.8)
Treatment type
Surgery	113 (33.0)
Radiation or concurrent chemoradiotherapy	81 (23.7)
Chemotherapy	106 (31.0)
Supportive care	42 (12.3)

The 126 controls comprised 58 (46%) men and 68 (54%) women. Their median age was 52 years (range, 13–83 years). The control group comprised 58 (46%) normal healthy people confirmed by a regular medical check-up, 28 (22%) patients with benign diseases (9 cases of benign hepatobiliary tumors, 8 cases of benign endocrinologic tumors, 3 cases of benign gastrointestinal tumors, 2 cases of benign respiratory tumors, and 5 cases of other benign tumors) and 40 (32%) other malignancies (9 cases of gastric cancer, 8 cases of thyroid cancer, 5 cases of colorectal cancer, 5 cases of hepatocellular cancer, 3 cases of breast cancer, 3 cases of uterine cervix cancer, 2 cases of ovarian cancer, 2 cases of esophageal cancer, 1 case of bile duct cancer, 1 case of endometrial cancer, and 1 case of gastrointestinal stromal tumor).

Serum GPx3 levels

Mean serum GPx3 levels were significantly lower in the patient group compared with the control group (10.1 ± 5.0 μg/mL vs. 13.0 ± 5.8 μg/mL, P < 0.001). In addition, mean serum GPx3 levels were also significantly lower in the patients with benign diseases or other malignancies compared with the normal healthy group (13.0 ± 5.5 μg/mL vs. 13.0 ± 6.1 μg/mL, P < 0. 001) (Fig 1). Based on ROC analysis, the AUC was 0.652 (95% confidence interval [CI], 0.595–0.710, P < 0.001). With a the cut-off value of 9.92 μg/mL, the diagnostic sensitivity and specificity of GPx3 were estimated to be 65.9% and 55.5%, respectively, in distinguishing lung cancer from the control (Fig 2). There was no significant difference on the ROC curve between non-small cell lung cancer (NSCLC) and small cell lung cancer (AUC 0.433, 95% CI, 0.330–0.535, P = 0.217).

Figure 1.

Serum GPx3 levels of the lung cancer and control groups. ELISA, enzyme-linked immunosorbent assay.

Figure 2.

Accuracy of serum GPx3 for detecting lung cancer at the time of diagnosis using receiver operating characteristic curve analysis.

The difference of serum GPx3 levels with histology and smoking status

Mean serum GPx3 levels were significantly lower in the patients with adenocarcinoma compared with those with squamous cell carcinoma (9.3 ± 4.6 μg/mL vs. 10.5 ± 4.9 μg/mL, P = 0.039) (Fig 3). However, there were no significant differences in mean serum GPx3 levels between NSCLC and small cell lung cancer (9.9 ± 4.9 μg/mL vs. 11.2 ± 5.4 μg/mL, P = 0.189). Moreover, mean serum GPx3 levels were 9.6 ± 4.8 μg/mL in the never smokers, 10.3 ± 5.0 μg/mL in the current smokers, and 10.0 ± 5.1 μg/mL in the ex-smokers. These results indicate that smoking status had no significant effect on the mean serum GPx3 levels (P = 0.637).

Figure 3.

The difference between serum GPx3 levels and the histologic subtypes of lung cancer (P = 0.058). ELISA, enzyme-linked immunosorbent assay, NSCLC non-small cell lung cancer, NOS, not otherwise specified.

The difference of serum GPx3 levels with the tumor node metastasis (TNM) stage and initial treatments

According to the tumor node metastasis (TNM) staging by the American Joint Committee on Cancer 7th edition, we classified our patients into Stage IA, IB, IIA, IIB, IIIA, IIIB, IV, limited disease, and extensive disease. This showed that there were no significant differences in mean serum GPx3 levels between the IA (7.0 ± 3.0 μg/mL), IB (9.2 ± 4.8 μg/mL), IIA (9.6 ± 3.9 μg/mL), IIB (12.0 ± 5.5 μg/mL), IIIA (11.2 ± 5.0 μg/mL), IIIB (10.3 ± 3.6 μg/mL), IV (10.2 ± 5.4 μg/mL), limited disease (8.5 ± 4.1 μg/mL), and extensive disease (12.6 ± 6.0 μg/mL).

With the stratification of the patients depending on the metastasis of lung cancer, mean serum GPx3 levels tended to be lower in the patients without metastasis compared with those with metastasis (9.6 ± 4.5 μg/mL vs. 10.7 ± 5.7 μg/mL, P = 0.051). In addition, mean serum GPx3 levels had a significant difference according to initial treatments (P < 0.001) (Fig 4). In other words, mean serum GPx3 levels were significantly lower in the surgery group (8.2 ± 4.1 μg/mL) compared with the CCRT (11.5 ± 4.6 μg/mL, P < 0.001), chemotherapy (10.7 ± 5.6 μg/mL, P < 0.001), and supportive care groups (10.9 ± 4.8 μg/mL, P = 0.002).

Figure 4.

The difference between serum GPx3 levels and the initial treatment modalities (P < 0.001). CCRT, concurrent chemoradiotherapy, ELISA, enzyme-linked immunosorbent assay.

Discussion

Our results showed that serum GPx3 levels were significantly lower in the patients undergoing surgery compared with those receiving other initial treatments. Although we found no significant difference between serum GPx3 levels and TNM stage because of the heterogeneous group of patients in our study, those who underwent surgery are representative of cases of early-stage NSCLC that is neither unresectable nor metastatic, therefore, serum GPx3 could be worth further study, especially in operable stage lung cancer.

It is known that serum Se levels and GSH-Px activity are significantly lower in patients with lung cancer compared with the normal healthy control.¹⁸ To date, studies have been conducted to examine the significance of the sequence of and copy number variation in the genes involved in the GSH pathway in patients with lung cancer and the relevant cell lines. It has been reported not only that the single nucleotide polymorphism (SNP) of human class mu glutathione S-transferase (GSTM4) is associated with the overall survival of patients treated with cisplatin, but also that the overexpression of GSTM4 is associated with increased cisplatin resistance.¹⁹ GPx1 is involved in encoding cellular GSH-Px; P198L, one of its variants, may contribute to a two-fold increase in the risk of lung cancer.²⁰ Although there is variability in the interaction between genes involved in the GSH pathway and a family or smoking history, such genes are associated with the risk of lung cancer in both younger and older individuals.²¹ To date, however, no studies have examined the correlation between GPx3 and the risk of lung cancer.

Recent studies have reported the involvement of GPx3 in the pathogenesis of human cancer, describing that the degree of mRNA and protein expression of GPx3 is relatively lower in patients with cancer.^12,13,22 Additionally, it has also been reported that GPx3 is involved in genomic integrity by inactivating reactive oxygen species (ROS) that are known as DNA-damaging agents and mediators of cancer chemotherapeutic response.²³ Its contribution of anticancer drug resistance has been demonstrated in ovarian clear cell adenocarcinoma.²⁴ Moreover, it has been reported that there is significant down-regulation of GPx3 expression or its complete silencing in human cancer cell lines derived from various types of cancer (head-and-neck, breast and bladder cancers, and melanoma). It is also known that the GPx3 promoter methylation in CpG islands is increased in such cancers. Of note, there was no GPx3 promoter methylation in lung cancer cell lines, showing that there was decreased or no expression of GPx3.¹⁵ This indicates that the pathogenesis of lung cancer might arise from the novel mechanism without epigenetic modification based on the down-regulation of GPx3.

Once secreted into the blood vessels, GPx3 detoxifies ROS before entering the cells from plasma. Its secreted form is readily detected from the plasma. Thus, it plays a unique role in the cellular antioxidant system. In addition, it has also been regarded as a biomarker candidate. Our results demonstrated that serum GPx3 levels were significantly lower in patients who underwent surgery compared with other treatments, indicating that GPx3 levels may be a biomarker that has diagnostic value in early stage lung cancer. However, it is unclear whether GPx3 has a prognostic value, despite the high level of GPx3 groups that were in a more advanced stage (CCRT, chemotherapy, and supportive care groups) and the low level of the GPx3 group was earlier stage (surgery group). If serum GPx3 was validated for early stage lung cancer by further research, this blood-based biomarker could be a more accessible screening method than conventional radiologic study, such as low dose chest computed tomography.

The main limitation of the study is the heterogeneous nature of the control group, including benign diseases and other malignancies. Because different cancer types and diseases may have different levels of serum GPx3, a more homogeneous control group is needed for validation, in spite of the presented healthy control data (general medical exam group). Our results may be limited by the differences of baseline characteristics (median age and gender) between the study and the control groups. In addition, the initial treatment modalities are indirect staging markers, therefore, they may be not enough to reflect TNM staging.

Conclusion

Our results showed that serum GPx3 levels were significantly lower in the patients who underwent surgery, thus indicating that GPx3 levels may have a diagnostic value in patients with an operable stage of lung cancer, rather than those with a locally-advanced or metastatic one. However, little is known about the mechanisms by which GPx3 is involved in tumor suppression. Further prospective and functional studies are, therefore, warranted to validate our results and to explore the role of GPx3 in suppressing the occurrence of tumors in the early stage of carcinogenesis and the relevant outcomes.

Disclosure

No authors report any conflict of interest.