Radiosensitivity-Related Genes and Clinical Characteristics of Nasopharyngeal Carcinoma

Materials and Methods Clinicopathological data of 185 patients with NPC treated at Nanfang Hospital of Southern Medical University between January 2013 and December 2014 were retrospectively analyzed. SPSS statistical software was used to analyze the clinicopathological data related to radiotherapy efficacy. Three patients who achieved complete remission and three with disease progression after CRT were selected. Differentially expressed genes (DEGs) were screened via mRNA microarray analysis of primary diagnostic endoscopy specimens. Results The peripheral blood leukocyte count, platelet count, and EBV-DNA copy number in NPC patients who were resistant to radiotherapy were higher than those in NPC patients who were sensitive to radiotherapy. The RobustRankAggreg (RRA) analysis method identified 392 DEGs, and the 66 most closely related genes among the DEGs were identified from the PPI network. Conclusion The results of this study indicate that screening for DEGs and pathways in NPC using integrated in silico analyses can help identify a series of genetic and clinical signatures for NPC patients treated with neoadjuvant chemotherapy followed by concurrent chemoradiotherapy.


Introduction
Nasopharyngeal carcinoma (NPC) is a highly malignant tumor originating from the epithelium; this disease shows a specific ethnic and geographical distribution [1]. NPC is prevalent among South and East Asian populations [2]. Lymph node and distant metastasis can occur early during disease progression without any obvious symptoms. More than 70% of NPC patients have locally advanced disease upon confirmation of their diagnosis. Improvements in treatment methods and implementation of comprehensive treatment strategies have substantially increased the 5-year survival rate of early-stage NPC patients to approximately 85% at present [3,4]. Radiotherapy is an essential component of the treatment of nondisseminated disease with curative intent, and its application in conjunction with chemotherapy prolongs survival [5]. Because NPC is sensitive to radiotherapy and the nasopharynx has a unique anatomical position, the comprehensive treatment of NPC is mainly radical radiotherapy [6]. Although local control of NPC has improved significantly due to advances in radiotherapy and comprehensive treatments, some patients do not benefit from radiotherapy due to the radiation resistance caused by local recurrence and distant metastasis [7]. Moreover, a poor response to chemoradiotherapy is related to substantial adverse effects and high medical costs.
In the era of personalized medicine, individualized radiotherapy guided by biomarkers and/or combination therapy has begun to emerge [8]. Many studies have evaluated whether incorporating other clinical factors and molecular biomarkers into the current anatomical staging system can better predict survival because this system is insufficient in predicting the prognosis or therapeutic effect [9]. This type of research mainly focuses on the specific genetic characteristics relating to the diagnosis, prognosis, or prediction of treatment response [10]. Therefore, establishing a model to predict the survival outcome of patients with locally advanced NPC will help identify the patients who would benefit most from intensity-modulated radiation therapy (IMRT). The aim of this study is to explore the relevant clinical factors or sensitivity predictors of chemoradiotherapy, which will help guide the selection of individualized treatment options for NPC patients, improve the curative effect, and avoid ineffective or excessive treatment.

Patient and Specimen Selection
2.1.1. Patient Selection. Clinicopathological data of 185 patients with NPC from Nanfang Hospital of Southern Medical University between January 2013 and December 2014 were retrospectively analyzed. The inclusion criteria were as follows: (1) biopsy-proven NPC, (2) initial treatment with no previous antitumor therapies, and (3) stage I to IVb NPC according to the 2010 edition of the American Joint Committee on Cancer (AJCC) staging system. Other eligibility criteria included a Karnofsky score > 70, age 18-70 years old, and normal electrocardiography (ECG), complete blood count, and liver and renal function results. The baseline examination included nasopharynx and neck MRI, chest and upper abdomen enhanced CT, and whole-body bone scanning. The protocol was approved by the ethics committee of Nanfang Hospital NFEC-2018-013 and implemented in accordance with the principles of the Declaration of Helsinki. All participants provided informed consent.

Specimen Selection.
Three patients with complete remission and three with disease progression after CRT were selected ( Table 1). The principle of specimen selection was as follows. (1) Resistant group: three months after the end of radiotherapy and chemotherapy, no obvious regression of the tumor was confirmed by imaging, and the pathology confirmed tumor residue. (2) Sensitive group: three months after the end of radiotherapy and chemotherapy, the tumor responded completely or partially to the treatment as confirmed by imaging and pathology. The endoscopic specimens obtained from the first diagnosis were used for the mRNA microarray.

Therapeutic Method.
(1) Chemotherapy: stage I patients received concurrent radiotherapy and chemotherapy while stage II-IVb patients received induction chemotherapy (IC) and concurrent chemotherapy. IC consisted of paclitaxel +fluorouracil (TP) or cisplatin+fluorouracil (PF) regimens. Two cycles of chemotherapy were repeated every three weeks, followed by concurrent radiotherapy (RT) and cisplatin administration (40 mg/m 2 ) weekly on weeks 7 through 14.
(2) Radiotherapy: IMRT was used in radical radiotherapy, and the target area was defined according to the ICRU [11] 71 and 83 [12] recommendations and the international consensus guidelines [13]. Reductions in nasopharyngeal and neck masses were examined by MRI or CT three months after treatment. RECIST 1.1 criteria were used to evaluate treatment efficacy [14]. A complete response (CR) was defined by the disappearance of all lesions, with nodes measuring <10 mm and the EBV-DNA copy number reduced to a normal level. An increase of ≥20% from the nadir or baseline was defined as progressive disease (PD). A decrease of at least 30% in the sum of diameters of target lesions in

Characteristics of Clinical
Data. The sensitive group included 124 patients with an average age of 46.1 years, and the resistant group included 61 patients with an average age of 46.1 years. No significant differences in gender ratio, age distribution, smoking history, or drinking history between the two groups were found. The T staging and N staging of the resistant group were significantly higher than those of the sensitive group (p = 0:001, 0.005, respectively). No significant difference was found in the probability of metastasis between the two groups before treatment (p = 0:577). The results showed that the proportion of differentiated tumors in the resistant group was significantly higher than that in the sensitive group (p = 0:01) ( Table 2).

Analysis of Hematological Examination and Treatment
Plan before Chemoradiotherapy. The level of EBV DNA in the resistant group was significantly higher than that in No noticeable difference was observed in the proportion of patients receiving induction chemotherapy between the two groups. A total of 104 patients in the sensitive group received concurrent chemotherapy during radiotherapy, as did 41 patients in the resistant group. No significant difference was found in the cumulative dose of cisplatin between the two groups (sensitive group 138:62 ± 98:533 mg, resistant group 113:28 ± 105:13 mg, p = 0:110), and no significant difference was found in the number of patients with a cumulative dose of cisplatin > 200 mg/m 2 between the two groups. A total of 107 patients received adjuvant chemotherapy, and no significant difference was found in the proportion of adjuvant chemotherapy and the cumulative dose of cisplatin between the two groups. All patients were assessed for nasopharyngeal lesion regression within 1 year after treatment. The response rate of radiotherapy and chemotherapy in the sensitive group was 98.4% (122/124). In the resistant group, 49.2% (30/61) of the lesions shrunk after the treatment. Among them, 83.3% (25/30) of the patients had an in situ recurrence of nasopharyngeal tumors within 2 years, and 13.3% (4/30) had an in situ recurrence of nasopharyngeal tumors within 3 years. Two patients were reexamined in the external hospital after treatment, and the case data of the other hospital indicated that the nasopharyngeal tumor had recurred. However, the imaging data after treatment could not be provided, and the curative effect was not evaluated. Twenty-eight patients experienced distant metastasis during observation, including 13 cases of metastasis in the sensitive group and 15 cases in the resistant group. The statistical analysis showed that the proportion of distant metastases was higher in the resistant group than in the sensitive group, and this difference was significant (p = 0:016) ( Table 3).  Figure 1. The microarray results identified 392 differentially expressed genes between the two groups: 92 downregulated genes and 300 upregulated genes. Figure 2 shows the differential expression of multiple genes between the two groups included in the microarray. Figure 3 shows the cluster heat map of the top 200 DEGs.  Figure 1: Standardization of gene expression. The blue bar represents the data before normalization, and the red bar represents the normalized data. The abscissa represents each sample, and the ordinate represents the quantity of expression. 4 BioMed Research International ( Figure 4). Table 4 shows the results of the gene ontology enrichment analysis of DEGs in NPC. In the biological process group, the DEGs were mainly concentrated in pathways relating to the immune response, blood microparticles, negative regulation of cell proliferation, positive regulation of early endosome to late endosome transport, establishment of epithelial cell apical/basal polarity, inflammatory response, T cell receptor signaling pathway, regulation of cytokine secretion, MyD88-dependent Toll-like receptor signaling pathway, positive regulation of T cell proliferation, defense response to TAF11  LOC283485  INPP5E  PPFIBP2  ATP6AP1  ISCA2  KIF21B  C1orf54  OLFM2  IFFO2  RAB21  CLEC1A  SNX10  PPP1R15B  CYP19A1  DDIT4  ALG8  GPR84  COG3  TIFA  PDE4B  HIST1H2AC  ZIC2  LRRC57  NPHP3  CCDC17  ZNF202  THEM5  PTPRJ  CDKN3  TLR10  ZBED5−AS1  PCDH18  MS4A7  CTC−338M12.4  LOC150005  LINC00472  ZNF300P1  PTPRZ1  SPDL1  LOC643072  IGFBPL1  HSF2  TRIM45  ZNF304  CWC15  ULK3  NPM3  COMMD3  TXNL4A  MMEL1  GEMIN6   BioMed Research International composition group, the DEGs were mainly enriched in the apical part of the cell, microvillus, membrane-tomembrane docking, Golgi cisterna membrane, endoplasmic reticulum, intracellular environment, membrane, endocytic vesicle lumen, filopodium, filopodium assembly, nucleoplasm, Toll-like receptor 1-Toll-like receptor 2 protein complex, microvillus membrane, extracellular matrix organization, ruffle membrane, Golgi-associated vesicle membrane, and TCR signalosome. These results show that most DEGs are significantly enriched in protein binding, the nucleoplasm, the membrane, the intracellular environment, and the endoplasmic reticulum.   Table 5 and Figure 5 show the most significant enrichments of DEGs from the KEGG analysis. The signaling pathways of DEGs were mainly enriched in the PI3K-Akt signaling pathway, tuberculosis, Epstein-Barr virus infection, phagosomes, cytokine-cytokine receptor interactions, inflammatory bowel disease (IBD), legionellosis, influenza A, leishmaniasis, and antigen processing and presentation.

Analysis of DEGs in NPC Using a PPI Network.
Using the STRING database [16] (http://string-db.org) to construct a protein-protein interaction (PPI) network, we obtained 392 DEGs, including 300 upregulated genes and 92 downregulated genes. After pruning away the orphaned and loosely connected nodes, an interactome network of DEGs was constructed, as shown in Figure 6. The 66 hub genes, including 55 upregulated genes and 11 downregulated genes, showing the most significant interaction are listed in Table 6.

Discussion
NPC is a malignant tumor that is very sensitive to radiation, and its sensitivity can differ depending upon the degree of tumor differentiation. A lower degree of differentiation indicates a higher sensitivity to radioactivity. In China, especially in south China, the most common pathological type is the nonkeratinized undifferentiated type (WHO type II) [17], which is very sensitive to radiation. However, many studies have shown that radiotherapy can not only kill the tumors but also change the expression level of many genes and proteins [18,19]. These changes can reduce the sensitivity of the tumor to radiation and thus lead to radiation resistance. Radiotherapy resistance is the main cause of failure in the treatment of nasopharyngeal carcinoma [20]. Therefore, exploring the molecular mechanism of radiotherapy resistance of nasopharyngeal carcinoma is very important in improving the effects of radiotherapy and improving the prognosis of patients with nasopharyngeal carcinoma.
Among the 185 patients included, 61 patients with NPC had poor curative effects or local recurrence in the short term. Comparing the differentiation of tumors in the two groups, the proportion of differentiation types in the resistant group was higher than that in the sensitive group (14.3% vs. 3.3% in the resistant group, p = 0:01). This result again suggests that more differentiated tumors are less sensitive to radiotherapy. Epstein-Barr virus infection is very common in NPC patients. Mutirangura [21] in 1997 and LO [22] in 1999 found that the positive rate of EBV-DNA and the copy number were significantly higher in the sera of NPC patients than in the sera of healthy controls. Subsequent studies confirmed that the positive rate and level of EBV-DNA detection in nasopharyngeal cancer patients with recurrence or metastasis (the median quantitative concentration of EBV in the recurrence or metastasis group was 32350 copies/ml) were significantly higher than those in patients who achieved clinical remission. In the sensitive group, the median concentration of EBV-DNA was 0 copies/ml (p = 0:01) [23]. During follow-up, the researchers found that plasma EBV-DNA levels were elevated in patients with recurrent nasopharyngeal cancer approximately six months before the onset of clinical symptoms. Similar to our results, the EBV level in the blood of the nasopharyngeal-cancer-resistant group was significantly higher than that of the sensitive group: the median concentration of EBV in the resistant group was 142,400 copies/ml compared with 20,800 copies/ml in the sensitive group (p = 0:028). Therefore, we speculated on whether EBV infection and the EBV-DNA copy number are associated with radiosensitivity in NPC.
According to the current theory regarding the molecular biology of cancer, the radiosensitivity of cancer cells may be  STAT6, HLA-DQB1, IL21R, TLR2, IL2RG, NFKB1, IL1A   Tuberculosis  hsa05152  9  2.85E-05 0.002826  HLA-DQB1, IL10RA, ATP6AP1, TLR1, TLR2, NFKB1 8 BioMed Research International regulated by a complex network. Differences in any link in such a system, such as a mutation or a difference in expression in a single gene, may affect radiosensitivity. At present, although the research on radiosensitivity has been extensive, most studies have examined only one or a limited number of genes and their expression products. To further understand the molecular mechanism of radiosensitivity of cancer cells, simultaneously detecting the expression of several genes related to this network is necessary. Gene chip technology provides a semiquantitative analysis of a large number of genes at the whole-genome level. It can compare differences in gene expression among different samples at the same time, reveal new genes, and analyze gene interaction networks through a clustering analysis and functional enrichment. Therefore, gene chip technology is widely used in research on a variety of cancers, such as cholangiocarcinoma [24], colorectal cancer [25], breast cancer [26], and pancreatic cancer [27]. However, the identification of DEGs and hub genes will help us better understand the molecular mechanism of NPC progression and consequently develop more biomarkers, which will be helpful for the study of the early diagnosis and treatment of NPC [28,29]. However, the gene research in NPC mainly focuses on the occurrence, development, recurrence, and metastasis of NPC [30,31]. Few similar studies on radiation resistance have been published.
In this study, we used the gene expression profile technology of an mRNA microarray to obtain a large number of DEGs and then used molecular biology information technology for data processing to identify molecular markers that can predict the efficacy of CRT. The results identified 392  BioMed Research International DEGs, including 300 upregulated genes and 92 downregulated genes. The DEGs in NPC were analyzed by GO functional annotation, which showed that most DEGs are significantly enriched in protein binding, the nucleoplasm, membranes, intracellular environment, and the endoplasmic reticulum. Thus, the differential genes may be mainly related to cells. At present, many studies have shown that the radiosensitivity is closely related to the cell cycle [32].
The signaling pathways of DEGs determined by the KEGG signal pathway analysis were mainly enriched in the PI3K-Akt signaling pathway, tuberculosis, Epstein-Barr virus infection, phagosomes, cytokine-cytokine receptor interactions, inflammatory bowel disease (IBD), legionellosis, influenza A, leishmaniasis, and antigen processing and presentation. The PI3K-Akt signaling pathway is closely related to tumorigenesis and cancer progression. Many studies have been published on the relationship between the PI3K-Akt signaling pathway and nasopharyngeal carcinoma. They include, for example, how microRNA-29 targets FGF2 and inhibits the proliferation, migration, and invasion of nasopharyngeal carcinoma cells via the PI3K/AKT signaling pathway; CHL1 suppresses tumor growth and metastasis in nasopharyngeal carcinoma by repressing the PI3K/AKT signaling pathway via interaction with Integrin β1 and Merlin [33]; and microRNA-29 targets FGF2 and inhibits the proliferation, migration, and invasion of nasopharyngeal carcinoma cells via the PI3K/AKT signaling pathway [34]. Therefore, the PI3K/Akt signaling pathway plays an important role in the occurrence, development, recurrence, and metastasis of nasopharyngeal carcinoma. In this study, we Figure 6: PPI network. Notes: The more proteins interact with each other, the larger the dot is, indicating that the more central the network is, the more critical and important the role is. Red indicates upregulated genes, and green indicates downregulated genes. Abbreviation: PPI: protein-protein interaction. Using the STRING database to construct a PPI network, we obtained 66 hub genes, including 55 upregulated genes and 11 downregulated genes. The mechanisms and functions of these genes and their roles in the radiotherapy resistance of NPC should be further studied.
For a long time, research on improving the effect of radiotherapy has been focused on cancer cells; however, in recent years, research has focused on the tumor microenvironment (TME). As a result, most of the classical radiobiology dogma fails to consider the effect of radiotherapy on the TME, and the reaction of radiotherapy to the TME may be very important for the success or failure of the treatment. Therefore, the attempt to combine radiotherapy with new biological targeted therapy is usually based on the potential to enhance cancer cell death induced by radiotherapy rather than on the potential to enhance radiosensitization by influencing TME. Many strategies have been proposed to overcome the radiation resistance of tumor cells, but little research has been conducted on the TME-mediated mechanisms of radiation resistance and how to circumvent these mechanisms [35]. In recent years, radiation therapy and the immune microenvironment have become popular research topics [36,37]. In this study, the results of the gene chip analysis showed that the genes differentially sensitive and insensitive to CRT in NPC were associated with the immune microenvironment, including the immune response, T cell receptor signaling pathway, positive regulation of T cell proliferation, positive regulation of the innate immune response, Toll-like receptor TLR1:TLR2 signaling pathway, TCR signalosomes, and intracellular pH reduction. However, how these different genes affect the efficacy of preoperative chemoradiotherapy is unclear; further basic scientific and clinical research studies are needed.

Conclusion
The peripheral blood leukocyte count, platelet count, and EBV-DNA copy number were higher in patients with NPC who were resistant to radiotherapy than in those who were sensitive to radiotherapy. The results of this study show that bioinformatics analyses of gene chip data can help us to identify and screen a series of gene characteristics related to NPC sensitivity to radiotherapy and chemotherapy. The genes involved in the mechanism of radiosensitivity in NPC patients are closely related to platelet aggregation, inflammatory factors, and EBV infection pathways, which will be of great clinical value in the future. Monitoring and controlling the cytokines related to inflammation and immunity can prevent and delay the occurrence and development of tumors to a certain extent. Additional follow-up studies should explore the relevant factors or predictive indicators of chemoradiotherapy sensitivity, which will help guide the selection of individualized treatment options for NPC patients, improve the curative effect, and avoid ineffective or excessive treatment. Some limitations to this study exist. We analyzed a series of differential genes of radiotherapy resistance in nasopharyngeal carcinoma; however, due to time and financial constraints, we did not conduct subsequent functional verification. Further molecular biological experiments are required to confirm the function of the identified genes associated with NPC.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
We declare no competing interests.