Effects of EGFR, CK19, CK20 and Survinin Gene Expression on Radiotherapy Results in Patients with Locally Advanced Head and Neck Cancer

Approximately 60% of all head and neck cancer(HNC) cases are diagnosed as locally advanced stage disease (Forastiere et al., 2008; Cooper et al., 2009). The primary goal is to increase the survival rate and protect organ functionality in the treatment of locally advanced head and neck cancer(LAHNC). The reason for achieving different results by applying similar treatment methods to tumors localized at the same area, at the same stage and sharing similar histopathologic characteristics may be the existence of cells with undefined predictive features that are resistant to the current treatment. Comprehensive studies conducted in order to predict radiosensitivity and radioresistance revealed that EGFR expression indicates a bad prognosis and resistance to cytotoxic agents including Abstract


Introduction
Approximately 60% of all head and neck cancer(HNC) cases are diagnosed as locally advanced stage disease (Forastiere et al., 2008;Cooper et al., 2009). The primary goal is to increase the survival rate and protect organ functionality in the treatment of locally advanced head and neck cancer(LAHNC). The reason for achieving different results by applying similar treatment methods to tumors localized at the same area, at the same stage and sharing similar histopathologic characteristics may be the existence of cells with undefined predictive features that are resistant to the current treatment. Comprehensive studies conducted in order to predict radiosensitivity and radioresistance revealed that EGFR expression indicates a bad prognosis and resistance to cytotoxic agents including

Patients and samples
Twenty-six patients with locally advanced (Stage III-IV) HNC who were admitted to Uludağ University Medical Faculty Department of Radiation Oncology and underwent chemo-radiotherapy were investigated in terms of the parameters that play a role in OS and DFS.
All cases provided informed constant including information about genetic studies. EGFR, CK19 and CK20 gene expression was assessed on all tumor and normal tissue samples which were taken during surgery and storedin paraffin after histological and pathological examinations. Increased and normal expression of EGFR, CK19 and CK20 in tumor tissue was considered positive. Peripheral blood samples were examined for EGFR, CK19, CK20 and survinin expression via RT-PCR before RT and three months after RT; any value above 0 was considered positive for EGFR, CK19 and CK20, while values above 0.2 were considered positive for survinin expression.

RT-PCR analysis
Tumor and normal tissues embedded in paraffin were treated with BIOstic (MO BIO Laboratories, Carlsbad, CA) and washed with 100% ethanol. The RNeasy FFPE Kit (QIAGEN, Germantown, Maryland, USA) was used for RNA isolation from deparaffinized tissue samples. Lymphocyte isolation was performed by washing peripheral blood samples with red blood cell lysis buffer (Roche Diagnostics, Mannheim, Germany) and total RNA isolation was performed with the High Pure RNA Isolation Kit (Roche Diagnostics). The density and purity of the RNA samples was measured by NanoDrop 2000 spectrophotometry (Thermo Scientific, Wilmington, Delaware, USA). RNA samples with ~1.8as the OD260/ OD280 optical density and with a concentration of 200-400ng/µl were chosen. The RT2 First Strand Kit (SABioscience) was used for cDNA synthesis from extracted RNAs; 5 µg of RNA was used for each cDNA synthesis. cDNA synthesis consisted of two phases: in the first phase, a mixture of 2-8 µl RNA, 2 µl GE and6-0 µL H20 was placed in the thermocycler for 5 minutes at 42°C and afterwards placed on ice for 1 minute. In the second phase, a mixture of 4 µl BC3, 1 µl P2, 2 µl RE3 and3 µl H2O was added to the first mixture and kept at 42°C for 15 minutes and 95°C for 5 minutes (abbreviations: GE:5X gDNA Elimination Buffer, BC3: 5X Reverse Transcriptase Buffer 3, H 2 O: RNAse-free H2O, P2: Primer and External Control Mixture, RE3: Reverse Transcriptase Enzyme Mixture 3). mRNA levels of peripheral blood and tumor samples were measured via specific primers for EGFR, survinin, CK19 and CK20 by applying them to the extracted cDNA samples. A mixture of 1µl of the primers, 12.5 µl RT2 PCR Sybr Green Master Mix, 1µl cDNA and10.5µl ddH2O was distributed in a96-well plate. The PCR process took place in an LC 480 device. In the denaturation phase of the PCR, the sample was heldfor 10 minutes for onecycle at 95°C, and in the amplification phase the sample was held at 95°C for 15 seconds for 40 cycles (a single fluorescent reading was performed at 72°C) and held at 60°C for 1 minute. Constant reading was performed during the melting curve (95°C to 65°C). The ∆∆CT method was used for the comparison of normal and tumor tissue expression results.

Radiotherapy
In the RT plan, bilateral parallel areas were irradiated with 1.8-2Gy/fx dose with a 6 MV photon beam until 46 Gy and completed to 50 Gy following spinal cord protection. The protected side of the posterior neck was also completed to 50 Gy with the appropriate energy level electrons. In definitive RT/radio-chemotherapy following 50 Gy,the primary tumor and the surrounding 2 cm area was increased to 63-75 Gy (total dose was 60-67 Gy in postoperative patients) and left at 50 Gy if there was no lymph node involvement. In cases of involved lymph nodes <3 cm, an additional dose (boost) was applied and completed to 60-66 Gy. The anterior area (supraclavicular area) was left at 46 Gy.

Statistics
SPSS16.0 (Statistical Package for the Social Sciences, Chicago, IL) was used for the statistical analysis. The Shapiro-Wilk test was used for the assessment of a normal distribution. The Mann-Whitney U test was used for the comparison of two groups with a non-normal distribution. The Wilcoxon signed rank test was used in order to compare paired samples. Pearson correlation and Spearman correlation coefficients were used on the expression of all markers BR and AR in both peripheral blood and tumor tissue samples. Fisher's exact chi-square test was used in the assessment of categorical data and for the comparison of circulating gene expression in the peripheral blood BR and AR. Survival analysis of univariate analysis was performed with the log-rank test via the Kaplan-Meier method. Cox regression analysis was used in the assessment of prognostic factors that may affect OS and DFS in multivariate analysis. p<0.05 was considered statically significant in all analyses.

Results
The demographic, clinical and histopathological characteristics of the cases are summarized in Table 1. The mean follow-up period was 18 months (3-26 months). During the follow-up period, progression was observed in primary tumors of seven cases (five larynx and two other tumors) and metastases were observed in six cases (two larynx, one maxillary and three nasopharynx cases).
EGFR, CK19, CK20 and survinin expression detected by RT-PCR was as follows in blood samples taken before radiotherapy (BR): six cases (23.1%), four cases (15.4%), one case (3.8%) and one case (3.8%), respectively. One case was positive for EGFR, three cases were positive for CK19 and four cases were positive for CK20 three months after RT, while there were no cases positive for survinin expression. Tumor tissues of 17 cases were analyzed for expression and five cases (19.5%) were positive for EGFR, eight cases (30.8%) were positive for CK19 and one case (3.8%) was positive for CK20 expression. The Spearman correlation test was performed and there was no Asian Pacific Journal of Cancer Prevention, Vol 16, 2015 3025 DOI:http://dx.doi.org/10.7314/APJCP.2015.16.7.3023 EGFR, CK19, CK20 and Survinin Genes Expression and Radiotherapy for Locally Advanced Head and Neck Cancer statistically significant correlation between EGFR, CK19, CK20 and survinin expression measured in tumor tissue before and after radiotherapy. According to the Pearson correlation test, there were no statistically significant correlations between gene expression levels measured BR.
There was no AR EGFR expression, while there were five cases positive for EGFR in tumor tissues BR (p=0.068). The lack of statistical significance in this issue is attributed to the small size of the study group. There was a statistically significant difference between survinin levels measured BR and AR (p=0.028). Positivity for EGFR, CK19, CK20 and survinin expression in peripheral blood and tumor samples did not affect OS according to the log-rank analysis (Figure 1). AR CK19 and CK20 levels in peripheral blood were found to be correlated (p=0.049) ( Table 2). It was observed that 22 cases (84.6%) were alive and four cases (15.4%) resulted in mortality at the date of assessment in September 2012. Mean OS was calculated as 19 months (10-38) and mean DFS was calculated as 20 months (3-24). It was noted that the mortalities occurred due to LAHNC. There were no statistically significant clinical, pathological, molecular or treatment-related parameters that affected OS and DFS.
OS was not affected by AR chemotherapy (p=0.083), while a statistically significant difference was observed in the DFS rates of tumors localized in the larynx in patients with AR chemotherapy (p=0.013) (24 months and 21 months). A significant risk factor could not be defined for OS and DFS with the Cox regression analysis.

Discussion
The current standard treatment methods of LAHNC are surgery, chemo-radiotherapy and/or cetuximab (Lin et al., 2004). The different results achieved by applying similar treatment methods to the same stage and localized tumors has led to defining prognostic and predictive factors that may affect treatment methods. Thus, the comprehensiveness of the studies conducted in this area is widened. Unnecessary RT treatments will be prevented byidentifying RT resistant cases (Venook, 2005). Additionally, with defining some parameters that may predict metastases, more aggressive treatment  methods can be used at the beginning of therapy. There are several recent studies asserting that the existence of EGFR gene expression can be used as a marker for metastasis in several tumor types (Tunca et al., 2012). Furthermore, there are other studies indicating that CK19, CK20 and survinin gene expression levels are increased in metastatic tumor cells found in the peripheral blood (Zhao et al., 2011). There have been no prior studies assessing EGFR, CK19, CK20 and survinin gene expression levels in the Turkish population.
In this study, we investigated the effects of EGFR, CK19, CK20 and survinin gene expression in tumor and normal tissue samples taken BR and peripheral blood samples taken both BR and AR on the benefits obtained with radiotherapy. Only 15.3% of cases investigated for EGFR expression levels in tumor tissues were positive. 75% of EGFR expression positive cases responded to therapy. 54% of patients without EGFR expression underwent therapy. Hence, the EGFR mutation was observed in only a small portion of patients achieving different results with treatment than in patients without EGFR expression. This gave rise to the thought that there may be EGFR mutations in tumor cells in some cases. In order to make a reliable statistical assessment about the effects of radiotherapy combined with cetuximab in patients with EGFR expression positive HNC, the number of patients included in this study should be increased. Also, the existence of EGFR gene mutations must be assessed in EGFR expression negative patients. In the assessment of EGFR expression in peripheral blood samples taken both BR and AR, we observed EGFR expression in 18.5% of patients. There was no EGFR expression in the peripheral blood samples of AR patients. Thus, we can assert that the possibility of metastasis can be reduced with RT in patients with increased EGFR expression.
The expression of the cytokeratin (CK) family is related to metastatic potential in several cancer types (Yang et al., 2010;Tunca et al., 2012;Shimada et al., 2012). In two of three studies assessing the effects of CK19 gene expression on HNC, it was stated that the expression of CK19 in bone marrow cells is related to relapse and metastasis (Andratschke et al., 2003;Wollenberg et al., 2004). Zhang et al. (2007) also indicated that increased CK expression in locally advanced OSCC cases is related to tumor relapse and decreased survival. CK19 expression was observed in 15.3% of patients in peripheral blood samples BR and it was found to be down regulated in all of these cases. Furthermore, CK19 expression was observed in the tumor tissues of two cases with laryngeal cancer that developed metastases. According to our findings, CK19 expression in the peripheral blood can be reduced by RT, and CK19 positivity may be related to metastasis risk in patients with laryngeal cancer. However, in order to make a reliable and statistically significant conclusion, the patient number included in this study should be increased.
In a recent study by Tunca et al., it was indicated that CK20 expression in the peripheral blood is related to metastasis and relapse in patients with breast cancer (Tunca et al., 2012). Toyoshima et al. stated that increased CK20 expression in peripheral blood is related to tumor diameter (Toyoshima et al., 2009). CK20 expression was observed 3.8% of our cases and this proportion was found to increase AR. Thus, we assert that CK20 expression cannot be inhibited by RT. Furthermore, when assessing EGFR, CK19 and CK20 expression levels together, AR positivity of CK19 and CK20 was found to be related by Pearson correlation analysis (p=0.049). Thus, it can be stated that CK19 andCK20 positivity continues even after RT in patients with CK19, CK20 and EGFR positive tumor tissue and peripheral blood samples BR.
It is a well-known fact that survinin is an anti-apoptotic gene that is expressed in several cancer types, including HNC, and is related to tumor aggressiveness and drug resistance (Khan et al., 2010;Zhu et al., 2011). In a study by Khan et al. (2010) it was indicated that increased survinin expression in patients with HNC is correlated with clinical stage, metastasis and OS, and plays role in RT resistance. In a study assessing the relationships of 70 different proteins between before and after radiotherapy via ELISA in HNC cases, it was found that only survinin could be used as a marker for radiotherapy sensitivity in HNC patients (Farnebo et al., 2011). We also found that survinin expression levels taken BR and AR were correlated (p=0.028). These results indicate potential resistance to RT in patients with increased survinin expression in the peripheral blood.
As a conclusion, in order to increase the reliability of our study, the number of patients included in the study must be increased, and our findings should be confirmed with advanced protein analysis. Nevertheless, we can assert that HNC patients with positive gene expression for any of EGFR, CK19, CK20 and survinin get less benefit from RT.
In our study, EGFR, CK19, CK20 and survinin expression was not correlated with OS when they were assessed independently. Other studies conducted in order to assess the relationship between the metastatic potential of HNC and EGFR, CK19 andCK20 expression have been inadequate. In order to determine the potential benefits of RT in patients with HNC, new studies are necessary to examineother metastasis markers together with the genes investigated here both BR and AR.