|Year : 2018 | Volume
| Issue : 7 | Page : 1525-1534
Expression profile analysis identifies a two-gene signature for prediction of head and neck squamous cell carcinoma patient survival
Xue Xu1, Mengzhi Li2, Jun Hu3, Zheng Chen4, Jinyu Yu5, Yan Dong6, Chengtao Sun6, Junqing Han4
1 Department of Gerontology, The Second Hosipital of Shandong University, Jinan, PR China
2 Department of Surgical Oncology, Taian City Central Hospital, Tai'an, PR China
3 Department of Radiology, Affiliated Hospital of Shandong Academy of Traditional Chinese Medicine, Jinan, PR China
4 Department of Tumor Research and Therapy Center, Shandong Provincial Hospital Affiliated to Shandong University, Shandong University, Jinan, China
5 Department of Renal Cancer and Melanoma, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Peking University Cancer Hospital and Institute, Beijing, PR China
6 Department of Radiotherapy, Shandong Provincial Hospital Affiliated to Shandong University, Shandong University, Jinan, China
|Date of Web Publication||19-Dec-2018|
324 Jingwu Weiqi Road, Jinan, Shandong 250021
324 Jingwu Weiqi Road, Jinan, Shandong 250021
Source of Support: None, Conflict of Interest: None
Aim: The aim of this study is to identify a gene prognostic signature for the head-and-neck squamous cell carcinoma (HNSCC). HNSCC is one of the most common malignancies worldwide; however, the molecular mechanisms underlying the malignancy are unclear.
Materials and Methods: We analyzed the gene expression profiles of GSE2379, GSE53819, and GSE59102 derived from the gene expression omnibus, and the cancer genome atlas (TCGA) HNSC databases. The R software was used to identify the differentially expressed genes (DEGs) between HNSCC tissues and normal controls. Gene ontology, Kyoto Encyclopedia of Genes and Genomes pathway, protein-protein interactions network, and survival analyses of common DEGs were also performed.
Results: A total of 52 upregulated and 31 downregulated DEGs were identified. Functional analyses demonstrated that these DEGs were mainly enriched in extracellular matrix-receptor interaction, focal adhesion, tyrosine metabolism, and cytokine-cytokine receptor interaction. According to the survival analyses, PLAU and SERPINE1 could predict the overall survival of HNSCC patients from the TCGA cohort. Multivariable Cox regression analyses showed that the PLAU and SERPINE1 were independent prognostic factors for HNSCC patients. The prediction power of this two-gene signature was evaluated through receiver operating characteristic curve analysis and achieved a better prognostic value than PLAU (area under curve 0.613 [95% confidence interval 0.569–0.656] vs. 0.577 [0.533–0.621]; P = 0.008) or SERPINE1 (0.613 [0.569–0.656] vs. 0.586 [0.541–0.629]; P = 0.043) when considered alone.
Conclusions: The study has identified a set of novel genes and pathways that play significant roles in the carcinogenesis and progression of HNSCC. This two-gene signature may prove to be a useful therapeutic target for HNSCC.
Keywords: Gene ontology, head-and-neck squamous cell carcinoma, protein-protein interactions network, signature, survival analysis
|How to cite this article:|
Xu X, Li M, Hu J, Chen Z, Yu J, Dong Y, Sun C, Han J. Expression profile analysis identifies a two-gene signature for prediction of head and neck squamous cell carcinoma patient survival. J Can Res Ther 2018;14:1525-34
|How to cite this URL:|
Xu X, Li M, Hu J, Chen Z, Yu J, Dong Y, Sun C, Han J. Expression profile analysis identifies a two-gene signature for prediction of head and neck squamous cell carcinoma patient survival. J Can Res Ther [serial online] 2018 [cited 2019 Oct 17];14:1525-34. Available from: http://www.cancerjournal.net/text.asp?2018/14/7/1525/247724
| > Introduction|| |
Head-and-neck squamous cell carcinoma (HNSCC) is one of the most common malignant tumors originating from the mucosa of oral and nasal cavities, oropharynx, hypopharynx, larynx, or nasopharynx. It is among the sixth leading cancers worldwide and accounts for more than 600,000 cases annually.,, HNSCC is a heterogeneous disease with more aggressive tumor biology and worse clinical outcomes. Despite multidisciplinary treatments such as surgical resection and chemoradiotherapy, the 5-year survival rate of HNSCC remains 40%–50%.,, Therefore, better understanding of the genetic and molecular characteristics in the tumorigenesis and progress of HNSCC will lead to the development of novel treatments and promote the prognosis of patients with HNSCC.
High-throughput gene expression profiling including microarrays or RNA-Seq technologies is widely used to identify differentially expressed genes (DEGs) between tumor and normal tissues.,, Some researchers performed gene expression profiling studies to identify several DEGs in HNSCC., However, the functional interactions among the genes were not considered and their findings were limited to small-scale clinical samples with heterogeneity. Integrating and re-analyzing published sequence data might overcome these disadvantages. Gene expression omnibus (GEO) and the cancer genome Atlas More Details (TCGA) provide an opportunity to make integrative analyses with the large-scale clinical sample to identify the significance of the gene signature for prognosis prediction in HNSCC.
In the current study, we integrated gene expression profiling (GSE2379, GSE53819, and GSE59102) datasets and HNSCC RNA-Seq and matched clinical information from a large cohort of HNSCC patients obtained from the TCGA database, to develop a two-gene panel prognostic risk prediction model.
| > Materials and Methods|| |
Gene expression omnibus and the cancer genome atlas datasets
The GSE2379, GSE53819, and GSE59102 datasets, which compare the gene expression profiles of normal controls with hypopharyngeal cancer, nasopharyngeal carcinoma, and laryngeal squamous cell carcinoma (LSCC), respectively, were downloaded from the NCBI GEO database.
The microarray data of GSE2379, which contained 34 hypopharyngeal cancer tissue samples and four normal tissues was based on GPL91 (Affymetrix Human Genome U95A Array, Affymetrix Inc., Santa Clara, CA, USA) and GPL8300 (Affymetrix Human Genome U95 Version 2 Array, Affymetrix). The microarray data of GSE53819 and GSE59102 were both based on GPL6480 (Agilent-014850 Whole Human Genome Microarray 4 × 44K G4112F (Probe Name version), Agilent Technologies, Palo Alto, CA, USA). The GSE53819 dataset included 18 nasopharyngeal carcinoma primary tumors and 18 non-cancerous nasopharyngeal tissues. As for GSE59102, 29 LSCC samples and 13 margin samples were included.
High-throughput data of level 3 RNA-Seq diagnosed with HNSCC were downloaded from TCGA. A total of 562 HNSCC patients with detailed follow-up time were included for subsequent analysis.
Data preprocessing and differential expression analysis
The downloaded GEO data were preprocessed by background correction and transformation from probe level to gene symbol through R language, followed by normalization. The limma R package was subsequently used for the calculation of differential expression of genes between HNSCC samples and noncancerous tissues. The edgeR package was used for DEG screening from TCGA data. Adjusted P < 0.01 and (logFC) >1 were chosen as the cutoff criteria. The Venn diagram and heatmaps were generated using the FunRich software.
Gene ontology enrichment analysis
Gene ontology (GO) enrichment analysis was performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID). Information regarding the enriched biological processes (BPs), cellular components (CCs), and molecular functions (MFs) was obtained to analyze the overlapping DEG functions. A value of P < 0.05 was defined as the cutoff criterion for significant function.
Kyoto encyclopedia of genes and genomes pathway analysis
Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analyses were performed to investigate the functions and processes of the common DEGs among GSE2379, GSE53819, GSE59102, and TCGA-derived datasets, using online tools such as DAVID. The cutoff value for significant pathway involvement was set as P < 0.05.
Protein-protein interaction network construction
The search tool for the retrieval of interacting genes database was used to analyze the protein-protein interaction (PPI) of common DEGs. The Cytoscape software was employed for visualizing the PPI networks of common DEGs. The hub genes were also identified using the molecular complex detection (MCODE) plugin.
Prognostic signatures generation and prediction
The association between the expression of identified hub genes and overall survival (OS) (2, 5, and 10 years) in HNSCC patients was assessed using HNSCC data from the TCGA cohort. The expression data of hub genes and OS of patients with HNSCC from the TCGA cohort were downloaded from the UCSC Xena browser. In addition, the OS data of 121 patients who received radiotherapy were also analyzed.
Furthermore, a prognostic potential of two-gene panel was constructed combining the PLAU and SERPINE1 to compare its prognostic validity with the PLAU model separately or the SERPINE1 model with a receiver operating characteristic (ROC) analysis.
The Illumina 450 K methylation array datasets were downloaded from the UCSC Xena browser of TCGA (HNSCC). The association between PLAU and SERPINE1 methylation status and OS (2, 5 and 10 years) in HNSCC patients were also examined using the UCSC Xena browser.
All statistical analyses were performed using the Statistical Package for the Social Science (SPSS) software, Version 20 (SPSS Inc., Chicago, IL, USA). All statistical tests were two-sided and values of P < 0.05 were considered statistically significant.
Validation of the diagnostic effectiveness of the prognostic genes
The ROC curve was used to assess the diagnostic effectiveness of the identified prognostic genes in HNSCC patients based on the GSE2379, GSE53819, GSE59102, and HNSCC datasets from TCGA.
| > Results|| |
The differentially expressed genes among GSE2379, GSE53819, GSE59102, and the cancer genome atlas
As shown in the Venn diagram, 472 (260 upregulated and 212 downregulated), 2078 (792 upregulated and 1286 downregulated), 2772 (1295 upregulated and 1477 downregulated), and 9203 (5831 upregulated and 3372 downregulated) DEGs were identified from the GSE2379, GSE53819, GSE59102, and TCGA datasets, respectively. In total, 52 upregulated and 31 downregulated genes were found to be significantly differentially expressed in these four independent cohorts [Figure 1]a and [Figure 1]b. [Table 1] and [Table 2] show the upregulated and downregulated DEGs found be common among the four cohorts analyzed.
|Figure 1: Venn diagram and heatmaps of differentially expressed genes among GSE2379, GSE53819, GSE59102, and the cancer genome atlas.(a and b) Commonly upregulated and down regulated differentially expressed genes.(c-f) Heatmaps of common differentially expressed genes between the head-and-neck squamous cell carcinoma tissues and normal controls in GSE2379, GSE53819, GSE59102, and the cancer genome atlas|
Click here to view
|Table 1: The overlapped differentially expressed genes identified among GSE2379, The Cancer Genome Atlas, GSE53819 and GSE59102|
Click here to view
|Table 2: The overlapped differentially expressed genes identified among GSE2379, The Cancer Genome Atlas, GSE53819 and GSE59102|
Click here to view
The heatmaps were generated using the FunRich software, based on the expression levels of common DEGs in the GSE2379, GSE53819, GSE59102, and TCGA cohorts [Figure 1]c, [Figure 1]d, [Figure 1]e, [Figure 1]f. As shown in [Figure 1]c, [Figure 1]d, [Figure 1]e, [Figure 1]f, each row in the heatmaps represents a gene and each column represents a biological sample. The color indicates the expression levels of genes between HNSCC and normal tissues.
Gene ontology and Kyoto encyclopedia of genes and genomes pathway enrichment analysis
The DEGs were classified into three functional groups as follows: MF group, BP group, and CC group. In the BP group, calcium ion homeostasis, extracellular matrix (ECM) organization, cell adhesion, biological adhesion, and blood vessel development were the commonly enriched categories [Figure 2]a. In the CC group, the common DEGs were mainly enriched in the extracellular region, extracellular matrix, collagen, basement membrane, and actin cytoskeleton [Figure 2]b. For the CC ontology, enriched categories among common DEGs were correlated with cytokine activity, metalloendopeptidase activity, structural molecule activity, actin binding, and growth factor activity [Figure 2]c. KEGG pathway enrichment analysis showed that the common DEGs were mainly enriched in ECM-receptor interaction, focal adhesion, tyrosine metabolism, and cytokine-cytokine receptor interaction pathways [Figure 2]d.
|Figure 2: Significantly enriched gene ontology and Kyoto encyclopedia of genes and genomes pathway terms of differentially expressed genes in the head-and-neck squamous cell carcinoma. (a-c) were significantly enriched biological processes, molecular function, and cellular component of common differentially expressed genes in the head-and-neck squamous cell carcinoma. (d) Significantly enriched Kyoto encyclopedia of genes and genomes pathways of common differentially expressed genes in the head-and-neck squamous cell carcinoma|
Click here to view
Key candidate genes identification with differentially expressed genes protein-protein interaction network
In case of the PPI network analysis, 78 (50 upregulated and 28 downregulated) of the 83 commonly altered DEGs were filtered into the PPI network complex, which contained 78 nodes and 364 edges [Figure 3]a. The central node genes (more than 10 connections or interactions) were also identified [Figure 3]e. The most significant 10 nodes, including PLAUR, MMP1, PLAU, COL4A2, SPP1, SERPINE1, MMP3, ILB1, CXCL12, and COL1A1, were identified as hub genes using the MCODE plugin. All hub genes were upregulated, except for CXCL12.
|Figure 3: (a) Seventy-eight differentially expressed genes were filtered into the differentially expressed genes protein-protein interaction network. (b and c) The most significant two modules were identified. (d) The pathways of common differentially expressed genes in module 1 and 2 are mainly enriched in extracellular matrix-receptor interaction, focal adhesion, and cytokine-cytokine receptor interaction. (e) The identified central node genes|
Click here to view
The entire PPI network was analyzed using the MCODE plugin. The most significant two modules, module 1 and module 2, consisting of 19 nodes and 10 nodes, respectively, were chosen for further pathway analyses [Figure 3]b and [Figure 3]c.
KEGG pathway enrichment analyses revealed that the genes in module 1 and 2 were mainly enriched in ECM-receptor interaction, focal adhesion, and cytokine-cytokine receptor interaction [Figure 3]d.
Identification of prognostic signature
Survival analyses performed in HNSCC data from TCGA showed that PLAU and SERPINE1 were significantly related to the survival time of patients with HNSCC [Figure 4]. Further, the Multivariate Cox regression analysis showed that PLAU and SERPINE1 might be an independent prognostic indicator for HNSCC [P = 0.027 and P = 0.034, respectively, [Table 3]. Furthermore, the log–rank test for OS performed on HNSCC data obtained from TGCA showed that high PLAU and SERPINE1 expression levels were significantly associated with an unfavorable 2 years OS [P = 0.009 and P = 0.006, [Figure 5]a and [Figure 5]d, 5 years OS [P = 0.0005 and P = 0.0017, [Figure 5]b and [Figure 5]e, and 10 years OS [P = 0.0029 and P = 0.0032, [Figure 5]c and [Figure 5]f.
|Figure 4: The Kaplan–Meier curves of the top 10 hub genes in the head-and-neck squamous cell carcinoma. The Kaplan–Meier curves of overall survival in the head-and-neck squamous cell carcinoma patients with high or low PLAUR (a), MMP1 (b), PLAU (c), COL4A2 (d), SPP1 (e), SERPINE1 (f), MMP3 (g), ILB1 (h), CXCL12 (i) and COL1A1 (j) expression in the cancer genome atlas head-and-neck squamous cell carcinoma|
Click here to view
|Table 3: Multivariate cox analysis of potential prognostic factors for patients with head and neck squamous cell carcinoma|
Click here to view
|Figure 5: The Kaplan–Meier curves of 2 years overall survival, 5 years overall survival, and 10 years overall survival in the head-and-neck squamous cell carcinoma patients with high or low PLAU (a-c) and SERPINE1 (d-f) expression in the cancer genome atlas cohort|
Click here to view
Since the elevated expression of PLAU and SERPINE1 was associated with poor OS in HNSCC patients, it was hypothesized that their methylation status might be associated with OS. By analyzing the DNA methylation data, survival analysis of HNSCC based on the UCSC Xena browser of TCGA revealed that hypomethylated PLAU was significantly associated with an unfavorable 2 years OS [P = 0.04664, [Figure 6]a and 5 years OS [P = 0.01977, [Figure 6]b, but not with 10 years OS [P = 0.1050, [Figure 6]c. However, there was no correlation between the methylation status of SERPINE1 and OS of HNSCC patients from the TCGA cohort [Figure 6]d, [Figure 6]e, [Figure 6]f.
|Figure 6: Hypomethylated PLAU is an indicator of poor overall survival in the cancer genome atlas head-and-neck squamous cell carcinoma patients. The Kaplan–Meier curves of 2 years overall survival, 5 years overall survival, and 10 years overall survival in the head-and-neck squamous cell carcinoma patients with hypermethylated or hypomethylated PLAU (a-c) and SERPINE1 (d-f) in the cancer genome atlas cohort|
Click here to view
We also analyzed the clinical data of 121 patients treated with radiotherapy. Survival analyses showed that elevated expression of PLAU and SERPINE1 was associated with poor OS [P = 0.0256 and 0.0005, [Figure 7]a and [Figure 7]b.
|Figure 7: Survival analysis of patients who received radiotherapy with high or low PLAU and SERPINE1 in the cancer genome atlas (a and b). Comparing the prognostic value of age, sex, tumor-node-metastasis, PLAU, and SERPINE1 (c). PLAU and SERPINE1 combined panel achieved higher prognostic prediction power (d)|
Click here to view
PLAU and SERPINE1 combined panel is a more sensitive predictive tool
We performed an ROC analysis to compare the prognostic value between age, sex, tumor-node-metastasis (TNM), PLAU expression, and SERPINE1 expression. PLAU and SERPINE1 had similar prognostic values [P = 0.957, [Figure 7]c, which were significantly better than those of age, sex, or TNM [P = 0.005, 0.040, and 0.040, [Figure 7]c. To develop a more sensitive predictive tool, we constructed a two-gene panel combining PLAU and SERPINE1 based on the cohort of TCGA HNSCC. A two-gene panel had a better prognostic value than PLAU or SERPINE1 expression considered independently (area under curve [AUC] 0.613 [95% confidence interval [CI] 0.569–0.656] vs. 0.577 [0.533–0.621], P = 0.008; AUC 0.613 [95% CI 0.569–0.656] vs. 0.586 [0.541–0.629], P = 0.043. [Figure 7]d).
Validation of the diagnostic effectiveness of the PLAU and SERPINE1
The common DEGs of PLAU and SERPINE1 based on the GSE2379, GSE53819, GSE59102, and HNSCC datasets from TCGA were sent for ROC analysis. All these two hub genes showed high diagnostic values to distinguish HNSCC from non-cancerous tissues with AUC >0.90, except for SERPINE1 in TCGA [Figure 8] and [Table 4].
|Figure 8: Validation of receiver operating characteristic analyses of PLAU and SERPINE1 in the head-and-neck squamous cell carcinoma patients. The ROC of PLAU or SERPINE1 in GSE2379 (a, e), TCGA HNSC (b, f), GSE53819 (c, g) and GSE59102 (d, h). The sensitivity curve is colored red, and the identify line is colored blue. The X-axis shows the false-positive rate, which has been shown as “1-Specificity.” The Y-axis indicates the true positive rate, which has been shown as “Sensitivity”|
Click here to view
|Table 4: The area under the curve of PLAU and SERPINE1 based on GSE2379, The Cancer Genome Atlas, GSE53819 and GSE59102 dataset|
Click here to view
| > Discussion|| |
In HNSCC, TNM staging and tumor grade were considered as prognostic predictors for survival. However, the poor clinical outcome of HNSCC is still a therapeutic challenge due to the limitations of the current staging system. Identifying markers for precise prognostic prediction for patients with HNSCC will thus address this challenge. In recent years, with the revolutionary development of next-generation sequencing, advances have been made that greatly contribute to our understanding of the molecular landscape of HNSCC initiation and progress.,, Identification of predictive biomarkers based on gene expression data is critical for increasing the probability of treatment success rate and achieving better clinical outcomes.
GEO and TCGA provide large-scale genome sequencing to improve the understanding of the molecular mechanism in HNSCC by using integrated bioinformatics analysis, especially TCGA with complete follow-up of clinical data of large cohort in HNSCC patients. Some studies based on the GEO or TCGA data had demonstrated some molecular predictors of several cancer types including HNSCC,,, non-small-cell lung cancer, colorectal cancer,, gastric adenocarcinoma, and breast cancer. This strategy of integrated genomic analysis provides an opportunity to discover and validate the underlying molecular mechanism driving tumorigenesis and progression in multiple cancers.
In the current study, based on 3 GEO datasets and TCGA HNSCC, 10 out of 83 common DEGs were identified as hub genes. According to the Kaplan–Meier survival analysis and Multivariate Cox regression analysis, PLAU and SERPINE1 were further selected, whose high expression were significantly associated with worse OS. PLAU, also known as uPA, is involved in the physiological and pathological processes of ECM transition, cell migration, invasion, cell signaling, cell proliferation, and apoptosis. It is reported that PLAU was directly linked to aging processes and age-related disease. Compared with normal tissue, PLAU was abnormally expressed in HNSCC tissue, however, without correlation to survival analysis in HNSCC patients., The sequences of SERPINE1, also known as PAI-1, is involved in the invasion, metastasis, and the apoptosis of multiple tumor cells and has been known as a poor prognostic factor in several common tumors,,,, except for HNSCC. Furthermore, PLAU and SERPINE1 were found to be potential novel biomarkers for radioresistance in patients suffering from HNSCC. It is noteworthy that there are very few reports on the involvement of PLAU and SERPINE1 in HNSCC; however, these two genes may provide new leads in the experimental research and clinical study of this disease.
In addition, KEGG pathway enrichment analysis revealed that the 83 DEGs (common to the four cohorts analyzed) are significantly enriched in ECM-receptor interaction, focal adhesion and cytokine-cytokine receptor interaction signaling pathways. These enriched pathways also revealed the molecular mechanism of the occurrence and development of HNSCC, which can offer new insights into the molecular mechanism of the tumorigenesis and progression of HNSCC.
Biomarker panels for prognostic prediction would be of high value and will be better than single markers. The combined prediction power of PLAU and SERPINE1 was evaluated based on ROC curve and two-gene signature that showed a better prognostic prediction (AUC 0.613, 95% CI 0.569–0.656, P < 0.05) than the PLAU or SERPINE1 model considered alone in the HNSCC cohort. Univariable and Multivariable Cox regression analyses showed that the PLAU and SERPINE1 were independent prognostic factors for patients with HNSCC.
| > Conclusion|| |
In conclusion, this study identified a set of novel genes and pathways that play significant roles in the carcinogenesis and progression of HNSCC. PLAU and SERPINE1 could predict overall survival as independent prognostic factors in HNSCC patients from TCGA cohort. The prediction power of this two-gene signature achieved a better prognostic value than did the PLAU or SERPINE1 alone.
Financial support and sponsorship
This work was supported by grants from the Shandong Provincial Natural Science Foundation, China (ZR2017PH057).
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Marur S, Forastiere AA. Head and neck squamous cell carcinoma: Update on epidemiology, diagnosis, and treatment. Mayo Clin Proc 2016;91:386-96.
Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A, et al.
Global cancer statistics, 2012. CA Cancer J Clin 2015;65:87-108.
Bhattacharjee A, Giri S, Roy M, Chakraborty A. Correlation of serum lactate dehydrogenase and alkaline phosphatase in different histological grades of head and neck squamous cell carcinoma and premalignant lesions. J Cancer Res Ther 2018;14:934-40.
Pakfetrat A, Dalirsani Z, Hashemy SI, Ghazi A, Mostaan LV, Anvari K, et al.
Evaluation of serum levels of oxidized and reduced glutathione and total antioxidant capacity in patients with head and neck squamous cell carcinoma. J Cancer Res Ther 2018;14:428-31.
Leemans CR, Braakhuis BJ, Brakenhoff RH. The molecular biology of head and neck cancer. Nat Rev Cancer 2011;11:9-22.
Vermorken JB, Mesia R, Rivera F, Remenar E, Kawecki A, Rottey S, et al.
Platinum-based chemotherapy plus cetuximab in head and neck cancer. N Engl J Med 2008;359:1116-27.
Bonner JA, Harari PM, Giralt J, Cohen RB, Jones CU, Sur RK, et al.
Radiotherapy plus cetuximab for locoregionally advanced head and neck cancer: 5-year survival data from a phase 3 randomised trial, and relation between cetuximab-induced rash and survival. Lancet Oncol 2010;11:21-8.
Tallari RV, Singh OP, Yogi V, Yadav S. Five versus ten fractions per week radiotherapy in locally advanced head and neck cancer. J Cancer Res Ther 2017;13:224-9.
Jardin F. Next generation sequencing and the management of diffuse large B-cell lymphoma: From whole exome analysis to targeted therapy. Discov Med 2014;18:51-65.
Guo Y, Bao Y, Ma M, Yang W. Identification of key candidate genes and pathways in colorectal cancer by integrated bioinformatical analysis. Int J Mol Sci 2017;18. pii: E722.
Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA Jr., Kinzler KW. Cancer genome landscapes. Science 2013;339:1546-58.
Ginos MA, Page GP, Michalowicz BS, Patel KJ, Volker SE, Pambuccian SE, et al.
Identification of a gene expression signature associated with recurrent disease in squamous cell carcinoma of the head and neck. Cancer Res 2004;64:55-63.
Ye H, Yu T, Temam S, Ziober BL, Wang J, Schwartz JL, et al.
Transcriptomic dissection of tongue squamous cell carcinoma. BMC Genomics 2008;9:69.
Pathan M, Keerthikumar S, Ang CS, Gangoda L, Quek CY, Williamson NA, et al.
FunRich: An open access standalone functional enrichment and interaction network analysis tool. Proteomics 2015;15:2597-601.
Sacco AG, Cohen EE. Current treatment options for recurrent or metastatic head and neck squamous cell carcinoma. J Clin Oncol 2015;33:3305-13.
Patel SG, Shah JP. TNM staging of cancers of the head and neck: Striving for uniformity among diversity. CA Cancer J Clin 2005;55:242-58.
Rizzo G, Black M, Mymryk JS, Barrett JW, Nichols AC. Defining the genomic landscape of head and neck cancers through next-generation sequencing. Oral Dis 2015;21:e11-24.
Leemans CR, Snijders PJF, Brakenhoff RH. The molecular landscape of head and neck cancer. Nat Rev Cancer 2018;18:269-82.
de Bree R, Senft A, Coca-Pelaz A, Kowalski L, Lopez F, Mendenhall W, et al.
Detection of distant metastases in head and neck cancer: Changing landscape. Adv Ther 2018;35:161-72.
Cao W, Liu JN, Liu Z, Wang X, Han ZG, Ji T, et al.
A three-lncRNA signature derived from the atlas of ncRNA in cancer (TANRIC) database predicts the survival of patients with head and neck squamous cell carcinoma. Oral Oncol 2017;65:94-101.
Sanati N, Iancu OD, Wu G, Jacobs JE, McWeeney SK. Network-based predictors of progression in head and neck squamous cell carcinoma. Front Genet 2018;9:183.
Zhou C, Ye M, Ni S, Li Q, Ye D, Li J, et al.
DNA methylation biomarkers for head and neck squamous cell carcinoma. Epigenetics 2018;13:398-409.
Dhont L, Pintilie M, Kaufman E, Navab R, Tam S, Burny A, et al.
Helicase-like transcription factor expression is associated with a poor prognosis in non-small-cell lung cancer (NSCLC). BMC Cancer 2018;18:429.
Lee H, Flaherty P, Ji HP. Systematic genomic identification of colorectal cancer genes delineating advanced from early clinical stage and metastasis. BMC Med Genomics 2013;6:54.
Tan P, Yeoh KG. Genetics and molecular pathogenesis of gastric adenocarcinoma. Gastroenterology 2015;149:1153-62.e3.
Zhang L, Huang Y, Zhuo W, Zhu Y, Zhu B, Chen Z, et al.
Identification and characterization of biomarkers and their functions for lapatinib-resistant breast cancer. Med Oncol 2017;34:89.
Cardoso AL, Fernandes A, Aguilar-Pimentel JA, de Angelis MH, Guedes JR, Brito MA, et al.
Towards frailty biomarkers: Candidates from genes and pathways regulated in aging and age-related diseases. Ageing Res Rev 2018;47:214-77.
Pinsky M, Rauch M, Abbas A, Sharabi-Nov A, Tamir S, Gutman R, et al.
Long-lived weight-reduced αMUPA mice show higher and longer maternal-dependent postnatal leptin surge. PLoS One 2017;12:e0188658.
Lallemant B, Evrard A, Combescure C, Chapuis H, Chambon G, Raynal C, et al.
Clinical relevance of nine transcriptional molecular markers for the diagnosis of head and neck squamous cell carcinoma in tissue and saliva rinse. BMC Cancer 2009;9:370.
Sepiashvili L, Hui A, Ignatchenko V, Shi W, Su S, Xu W, et al.
Potentially novel candidate biomarkers for head and neck squamous cell carcinoma identified using an integrated cell line-based discovery strategy. Mol Cell Proteomics 2012;11:1404-15.
Kwaan HC, Wang J, Svoboda K, Declerck PJ. Plasminogen activator inhibitor 1 may promote tumour growth through inhibition of apoptosis. Br J Cancer 2000;82:1702-8.
Liu X, Wu J, Zhang D, Bing Z, Tian J, Ni M, et al.
Identification of potential key genes associated with the pathogenesis and prognosis of gastric cancer based on integrated bioinformatics analysis. Front Genet 2018;9:265.
Wu DM, Wang S, Wen X, Han XR, Wang YJ, Fan SH, et al.
MircoRNA-1275 promotes proliferation, invasion and migration of glioma cells via SERPINE1. J Cell Mol Med 2018;22:4963-74.
Fang H, Placencio VR, DeClerck YA. Protumorigenic activity of plasminogen activator inhibitor-1 through an antiapoptotic function. J Natl Cancer Inst 2012;104:1470-84.
Jerhammar F, Ceder R, Garvin S, Grénman R, Grafström RC, Roberg K, et al.
Fibronectin 1 is a potential biomarker for radioresistance in head and neck squamous cell carcinoma. Cancer Biol Ther 2010;10:1244-51.
Cheng WY, Ou Yang TH, Anastassiou D. Development of a prognostic model for breast cancer survival in an open challenge environment. Sci Transl Med 2013;5:181ra50.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8]
[Table 1], [Table 2], [Table 3], [Table 4]