Abstract
Background/Aim: To evaluate the diagnostic value of circulating CXC chemokines as biomarkers for non-small cell lung cancer and compare them against a standard panel of already existing cancer biomarkers. Materials and Methods: A total of 90 individuals were enrolled in the study. We analyzed 30 patients with stage IA-IIB carcinoma of the lung who underwent pulmonary resection, 30 patients with metastatic NSCLC, and 30 healthy volunteers. The biomarkers levels were measured in plasma blood samples, by ELISA and immunoassays. Results: The levels of circulating CXCL4, CXCL8, CXCL9, CXCL10 and CXCL11 were higher and those of circulating CXCL1 were lower in patients with early-stage NSCLC compared to metastatic NSCLC patients and controls (p<0.05). CXCL4, CXCL9 and CXCL11 were included in the panel that showed a sensitivity of 100% versus 60% for CEA, CA125 and CYFRA21-1 (p<0.001). Conclusion: Combination of CXCL4, CXCL9 and CXCL11 has a high diagnostic value.
Lung cancer is by far the leading cause of cancer-related mortality globally, with an estimated 1.3 million new cases diagnosed worldwide each year, accounting for nearly 12% of all cancers and an estimated 1.1 million deaths each year (1). A blood-based biomarker is an attractive, non-invasive modality that could complement technologies, such as helical CT scans, in facilitating early detection and treatment of lung cancer. There is currently no single clinical biomarker that meets the sensitivity and specificity criteria required for screening or stratification purposes (2). Other biomarkers, such as carcinoembryonic antigen (CEA) (3), carcinoma antigen 125 (CA125) and cytokeratin-19 fragment (CYFRA 21-1) have been used in clinical practice for many years (4, 5) and there exist many molecules currently undergoing validation processes.
Numerous targets have been identified for research in the field, one of which is a group of C-X-C motif (CXC) chemokines. In the last two decades, studies have demonstrated that CXC chemokines and chemokine receptors can directly enhance or inhibit tumor-associated angiogenesis, promote tumor-related immunity, and enhance organ-specific metastases (6, 7). Malignant cells from different cancer types have different profiles of chemokine ligand and receptor expressions. Unlike other cytokines, chemokine proteins can be detected at picogram levels in the blood, which makes them more attractive targets for investigation in lung cancer biomarker research. There exists strong scientific evidence demonstrating direct or indirect involvement of CXCL4, CXCL7, CXCL8 and CXCL10 in the lung cancer development process (8-11).
The aim of the present study was to evaluate the diagnostic value of circulating CXC chemokines: CXCL1, CXCL4, CXCL5, CXCL6, CXCL7, CXCL8, CXCL9, CXCL10, CXCL11 and CXCL12 as biomarkers in non-small cell lung cancer and compare them against a standard panel of cancer biomarkers including CEA, CA125 and CYFRA 21-1. We hypothesized that NSCLC patients could have a distinct CXC chemokine profile, and levels of circulating CXC chemokines are characteristic for discrete processes of carcinogenesis.
Materials and Methods
Patients. From June 2010 to June 2011 total of 90 individuals were enrolled in the study. The cohort consisted of 30 asymptomatic patients diagnosed with stage IA–IIB adenocarcinoma and squamous cell carcinoma of the lung who underwent pulmonary resection, 30 patients with metastatic NSCLC, and 30 healthy volunteers. Surgical patients underwent pulmonary resection at our Institution. Postoperative staging of patients was based on the seventh edition of international TNM classification for lung cancer. Histological classification was performed according to the World Health Organization histologic typing of lung tumors (12). The final cohort included 30 asymptomatic patients, diagnosed with stage IA-IIB NSCLC. Two age-, sex-and smoking history-matched control groups were established; 30 patients with metastatic NSCLC and 30 healthy blood donors. Patients' demographics and cancer characteristics are provided in Table I.
This prospective study was conducted in accordance with the ethical standards of the Helsinki Declaration of the World Medical Association and with full approval by the Pauls Stradins Clinical University Hospital Foundation Ethics committee. All patients' data were obtained after informed consent by the patient.
Measurement of plasma CXC chemokine concentrations. Double blood samples of subjects were drawn from the cubital vein into 5-ml vacutainer tubes using a standard technique. In the surgical group, blood samples were obtained prior to surgery. Tubes were centrifuged at room temperature for 10 min at 1,300 × g. Following centrifugation plasma was immediately archived at −70°C pending utilization. Prior to ELISA assay samples were de-frosted at room temperature and processed according to the ELISA kit manufacturer (Raybiotech, Norcross, GA, USA) protocol (CXCL1/growth regulated oncogene-alpha (GRO-α), CXCL4/platelet factor 4 (PF4), CXCL5/epithelial-derived neutrophil-activating peptide 78 (ENA78), CXCL6/granulocyte chemotactic protein-2 (GCP2), CXCL7/ neutrophil-activating protein-2 (NAP2), CXCL8/interleukin-8 (IL-8), CXCL9/monokine induced by gamma interferon (MIG), CXCL10/ interferon gamma-induced protein 10 (IP10), CXCL11/interferon-inducible T-cell alpha chemoattractant (I-TAC) and CXCL12/stromal cell-derived factor 1 beta (SDF-1). Measurements were obtained twice per each sample. Measurements outside the working range for any given assay were excluded and may result in minor fluctuations in the cohort numbers. The CXC chemokine concentrations in plasma were calculated from the standard curve.
CEA, CA125 and CYFRA 21-1 levels were measured in the plasma of early-stage NSCLC patients by immunoassay in a certified clinical laboratory (normal range=CEA <5 ng/ml, CA 125 <35 U/ml, CYFRA 21-1 <2.08 ng/ml).
Statistical methods. Prior to analysis, data normality was assessed using the Kolmogorov-Smirnov test that indicated necessity for non-parametric test application. Biomarker levels were statistically evaluated for median differences in concentration by the Kruskal-Wallis test and a p-value under 0.05 was considered significant. Sensitivity and specificity of CXC chemokine levels were calculated using receiver-operating characteristics (ROC) curve analysis. The area under the ROC curve (AUC) was measured to assess discriminatory power of test. Test results with an AUC less than 0.7 were considered not clinically-useful. Test results with an AUC more than 0.85 were considered of high clinical value. Based on AUC values, a panel of three most clinically significant CXC chemokines were selected, and compared with standard lung cancer biomarker panel consisting of CEA, CYFRA 21-1 and CA 125 in the early-stage NSCLC group. The associations of clinical variables with individual biomarker levels were assessed by the Spearman's correlation coefficient, values between −0.4 and 0.4 were considered non-significant.
Patients' demographics and tumor characteristics.
Results
Results of the present study indicating that ELISA kits provided reproducible results and the observed biomarker levels are presented in Table II.
We observed significant differences in levels of circulating CXCL1, CXCL4, CXCL8, CXCL9, CXCL10 and CXCL11 in patients with early-stage NSCLC compared to metastatic NSCLC patients and controls. Interestingly there was no difference in CXCL9, CXCL10 and CXCL11 levels between patients with metastatic NSCLC and healthy controls. Representative box-and-whisker plots of CXC chemokine levels are shown in Figure 1. No statistically significant correlation between CXC chemokine levels and clinical parameters, such as age, sex, smoking, T stage, N stage and histological sub-type of NSCLC was noted (p>0.05).
ROC curve analysis showed that CXCL1, CXCL4, CXCL8, CXCL9, CXCL10 and CXCL11 are statistically significantly, sensitive and specific with the area under the ROC curve more than 0.7 (Table III).
Levels of CXC chemokines in the plasma of patients.
ROC curve analysis.
Based on the AUC values, we selected the three CXC chemokines with the highest values – CXCL4 (AUC=0.91), CXCL9 (AUC=0.86) and CXCL11 (AUC=0.8) to include in our biomarker panel.
CXC chemokine levels were interpreted as positive and negative based on criterion value corresponding with the Youden index. CEA, CA125 and CYFRA 21-1 levels were also interpreted as positive and negative based on the given normal range. All three CXC chemokines were positive in 9 patients (30%), two chemokines were positive in 13 patients (43.3%), and a single biomarker was positive in 8 patients (26.7%). There were no negative cases in the CXC chemokine panel. In contrast, application of the standard biomarker panel (CEA, CA125 and CYFRA 21-1) showed that all three biomarkers were positive in 3 cases (10%), two biomarkers were positive in 2 cases (6.7%), and a single biomarker was positive in 13 patients (43.3%). There were 12 negative cases (40%). The sensitivity of the CXC chemokine panel was calculated as 100% versus 60% for CEA, CA125 and CYFRA 21-1 (p<0.001).
Discussion
To date, there has not been any validated biomarker to show adequate sensitivity, specificity, and reproducibility to be used for the early detection and treatment of lung cancer. Cancers have a complex chemokine network that influences the immune-cell infiltration of a tumor, as well as tumor cell growth, survival and migration, and angiogenesis. CXC levels might be elevated in different cancer types, but one would speculate a chemokine profile or molecular fingerprint specific to cancer type. Tumor-related proteins may be secreted into the peripheral circulation of patients with cancer and are detectable by protein analysis.
Our findings suggest the most sensitive and specific diagnostic biomarkers in the early-stage NSCLC patient group, were CXCL4, CXCL9, CXCL10 and CXCL11. Results were statistically significant despite the small patient cohort. Interestingly, all four chemokines belong to the sub-group of ELR- CXC chemokines. The NH2-terminus of the majority of the CXC chemokines contain three aminoacid residues (Glu-Leu-Arg: the “ELR” motif) preceding the first cysteine amino-acid residue of the primary structure of these cytokines (13, 14). The family members that contain the ELR motif (ELR+): CXCL1, CXCL2, CXCL3, CXCL5, CXCL6, CXCL7 and CXCL8 are potent promoters of angiogenesis (15). In contrast, members that lack the ELR motif (ELR−):CXCL4, CXCL9, CXCL10 and CXCL11 are potent inhibitors of angiogenesis. This difference suggests, on a structural and functional level, that members of the chemokine family are unique cytokines regarding their ability to behave in a disparate manner in the regulation of angiogenesis.
Multiple comparison box and whisker plots representing CXC chemokine levels – plot A - CXCL1 (p<0.01), plot B - CXCL4 (p<0.001), plot C - CXCL8 (p<0.05), plot D - CXCL9 (p<0.001), plot E - CXCL10 (p<0.01), and plot F - CXCL11 (p<0.01). (Circles and triangles represent outside and far out values).
There exists limited scientific evidence on the diagnostic value of CXC chemokines. Previous reports have shown that CXCL4 (16) and CXCL7 contribute in tumor growth and development as well as in tumor angiogenesis in lung cancer (17) and other cancer cell types (18, 19), but CXCL8 levels can also be indicative of lung cancer risk several years before diagnosis (20). Our results did not confirm the diagnostic value of CXCL1, CXCL5, CXCL7, and CXCL12.
Recent studies have indicated that CEA and CYFRA 21-1 are reliable serum tumor markers for the diagnosis of lung cancer in addition to CT scans when combined or used individually at twice the standard cut-off level in high-prevalence rate groups (21). Despite that fact, standard biomarkers including CEA, CA125 and CYFRA 21-1 that are still widely used in clinical practice (22), did not show high sensitivity and were negative in 40% of cases, when only single biomarkers were positive in 43% of cases. The panel of CXC chemokines was much more sensitive and reliable than the panel of standard biomarkers, but further larger multi-institutional studies are required for validation of our findings. CXC chemokine profile assessment may be effective in high-risk patients and patients undergoing diagnostic investigations for possible NSCLC.
A limitation of the study is reporting on a single-Centre experience and therefore on a relatively small number of patients. We found an increased expression of CXCL4, CXCL8, CXCL9, CXCL10 and CXCL11 in plasma specimens of early-stage NSCLC patients compared to control subjects. The findings also demonstrated that a combination of CXCL4, CXCL9 and CXCL11 has a better diagnostic value in NSCLC than each marker separately, as NSCLC may have a specific molecular fingerprint. Further studies are required to determine if these chemokines can be utilized in a blood-based biomarker panel for the diagnosis of lung cancer.
Acknowledgements
This work was partially supported by State Research program in Medicine and Grant from Latvian science council.
Footnotes
Conflicts of Interest
None.
- Received September 12, 2015.
- Revision received October 1, 2015.
- Accepted October 19, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.