Collagen: A possible prediction mark for gastric cancer

Matrix metalloproteinase‑1 expression is regulated by HIF‑1‑dependent and epigenetic mechanisms and serves a tumor‑suppressive role in gastric cancer progression

The matrix metalloproteinase (MMP) family is associated with degradation of the extracellular matrix and is known to promote cancer invasion. The present study aimed to investigate the biological role of MMP-1 in gastric cancer cells and analyze the association between MMP-1 expression and the clinical outcomes of gastric cancer patients. In the present study, hypoxia accelerated invasion, accompanied by elevated MMP-1 expression in the gastric cancer cell line 58As9. Additionally, hypoxia-inducible factor-1α (HIF-1α) knock- down in 58As9 cells reduced MMP-1 expression under hypoxic conditions. Treatment with 5-aza-2-deoxycytidine and trichostatin A restored MMP-1 expression in the MMP-1-deficient cell lines MKN45 and MKN74. These results indicated that MMP-1 expression was controlled by both HIF-1α-dependent and epigenetic mechanisms in gastric cancer cell lines. In addition, MMP-1 knockdown impaired the hypoxia-induced invasiveness of 58As9 cells, implicating MMP-1 in the elevated invasion. By contrast, knockdown enhanced the proliferative ability of 58As9 cells, whereby expression of cell cycle-related genes was subsequently altered. In nude mouse models, the knockdown accelerated the growth of xenograft tumor and the development of peritoneal dissemination. In an immunohistochemical study using 161 surgically resected cancer tissues, the Ki67 score was significantly higher in the group with low MMP-1 expression (P<0.001). Disease-free survival (DFS) and disease-specific survival (DSS) were both significantly reduced in patients with low MMP-1 expression (log-rank test; DFS: P=0.005; DSS: P=0.022). Multivariate analysis demonstrated that MMP-1 expression was an independent prognostic factor for DFS and DSS [DFS: HR=2.11 (1.22–3.92) P=0.005, DSS: HR=2.90 (1.23–8.50) P=0.012]. In conclusion, the present study indicated that MMP-1 may serve as a tumor-suppressive factor that inhibits gastric cancer progression, although it promoted invasion in vitro.

Identification of genes and pathways related to breast cancer metastasis in an integrated cohort

BACKGROUND

Breast cancer is the most common malignant disease in women. Metastasis is the most common cause of death from this cancer. Screening genes related to breast cancer metastasis may help elucidate the mechanisms governing metastasis and identify molecular targets for antimetastatic therapy. The development of advanced algorithms enables us to perform cross-study analysis to improve the robustness of the results.

MATERIALS AND METHODS

Ten data sets meeting our criteria for differential expression analyses were obtained from the Gene Expression Omnibus (GEO) database. Among these data sets, five based on the same platform were formed into a large cohort using the XPN algorithm. Differentially expressed genes (DEGs) associated with breast cancer metastasis were identified using the differential expression via distance synthesis (DEDS) algorithm. A cross-platform method was employed to verify these DEGs in all ten selected data sets. The top 50 validated DEGs are represented with heat maps. Based on the validated DEGs, Gene Ontology (GO) functional and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed. Protein interaction (PPI) networks were constructed to further illustrate the direct and indirect associations among the DEGs. Survival analysis was performed to explore whether these genes can affect breast cancer patient prognosis.

RESULTS

A total of 817 DEGs were identified using the DEDS algorithm. Of these DEGs, 450 genes were validated by the second algorithm. Enriched KEGG pathway terms demonstrated that these 450 DEGs may be involved in the cell cycle and oocyte meiosis in addition to their functions in ECM-receptor interaction and protein digestion and absorption. PPI network analysis for the proteins encoded by the DEGs indicated that these genes may be primarily involved in the cell cycle and extracellular matrix. In particular, several genes played roles in multiple signalling pathways and were related to patient survival. These genes were also observed to be targetable in the CTD2 database.

CONCLUSIONS

Our study analysed multiple cross-platform data sets using two different algorithms, helping elucidate the molecular mechanisms and identify several potential therapeutic targets of metastatic breast cancer. In addition, several genes exhibited promise for applications in targeted therapy against metastasis in future research.

Identification of Potential Biomarkers Associated with Prognosis in Gastric Cancer via Bioinformatics Analysis

Background

Gastric cancer (GC) is one of the leading causes of cancer-related mortality worldwide. We aimed to identify differentially expressed genes (DEGs) and their potential mechanisms associated with the prognosis of GC patients.

Material/Methods

This study was based on gene profiling information for 37 paired samples of GC and adjacent normal tissues from the GSE118916, GSE79973, and GSE19826 datasets in the Gene Expression Omnibus database. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were used to investigate the biological role of the DEGs. The protein–protein interaction (PPI) network was constructed by Cytoscape, and the Kaplan-Meier plotter was used for prognostic analysis.

Results

We identified 119 DEGs, including 21 upregulated and 98 downregulated genes, in GC. The 21 upregulated genes were mainly enriched in extracellular matrix-receptor interaction, focal adhesion, and transforming growth factor-β signaling, while the 98 downregulated genes were significantly associated with gastric acid secretion, retinol metabolism, and metabolism of xenobiotics by cytochrome P450. Thirty hub DEGs were obtained for further analysis. Twenty-five of the 30 hub DEGs were significantly associated with the prognosis of GC, and 21 of the 25 hub DEGs showed consistent expression trends within the 3 profile datasets. KEGG reanalysis of these 21 hub DEGs showed that COL1A1, COL1A2, COL2A1, COL11A1, THBS2, and SPP1 were mainly enriched in the extracellular matrix-receptor interaction pathways.

Conclusions

We identified 6 genes that were significantly related to the prognosis of GC patients. These genes and pathways could serve as potential prognostic markers and be used to develop treatments for GC patients.

Overexpression of COL24A1 in Hepatocellular Carcinoma Predicts Poor Prognosis: A Study Based on Multiple Databases, Clinical Samples and Cell Lines

Purpose

Hepatocellular carcinoma (HCC) ranks as the second leading cause of cancer-related deaths worldwide and its incidence tends to increase globally. Current biomarkers can hardly reflect diagnostic significance and therapeutic efficiency accurately. The role of collagen 24A1 (COL24A1) in HCC, as a member of the collagen family, remains unclear and needs to be proved.

Methods

HCC-associated RNA sequencing datasets were obtained from The Cancer Genome Atlas (TCGA), International Cancer Genome Consortium (ICGC) and Oncomine databases. R language was used to screen and analyze the differential expression genes (DEGs). Gene Ontology (GO) pathway analyses of DEGs were carried out using g:profiler. Then, the STRING database and cytoscape software were applied to estimate the correlations between DEGs. GEPIA and HCCDB database were used to verify and analyze these results. Finally, qPCR and immunohistochemistry staining on clinical samples and HCC cell lines were conducted to further verify the above results.

Results

The expression of COL24A1 in HCC tissues was significantly higher than that in normal hepatic (NH) tissues. The analysis of TCGA and Oncomine database samples proved that COL24A1 expression was connected with tumor grade, tumor stage and pathologic stage. Meanwhile, the overall survival (OS) curve also validated that high expression of COL24A1 indicated poor prognosis. And the above results have been verified in our clinical samples and HCC cell lines. Furthermore, there is a significant increase of the expression of COL24A1 in various tumors. COL24A1, together with other interactive proteins, has a significant effect on the prognosis of HCC.

Conclusion

Overall, COL24A1 may be an oncogene in the development of HCC. Also, the overexpression of COL24A1 in HCC may help diagnose HCC and evaluate prognosis.

Alteration and prognostic values of collagen gene expression in patients with gastric cancer under different treatments

Collagen (COL) genes participate in tumor extracellular matrix (ECM)-receptor interactions and focal adhesion pathways, which play a crucial role in tumor invasion and metastasis. The prognostic value of COL genes has been shown for several malignancies. In the present study, we analyzed multiple microarray datasets using the Oncomine database to identify alterations of COL genes in gastric cancer (GC). Gene expression levels were analyzed by quantitative real-time polymerase chain reaction (qRT-PCR) and immunohistochemistry (IHC) in GC tissues and matched adjacent tissues. The prognostic value of differentially expressed COL genes in GC was evaluated by Kaplan-Meier survival analysis based on the complete mRNA transcriptomics data from The Cancer Genome Atlas (TCGA). We found that seven COL genes (COL1A2, COL4A1, COL4A2, COL6A1, COL6A2, COL6A3, and COL11A1) were elevated in GC. Among them, stepwise multivariate Cox regression was applied, and it was determined that COL4A1 and COL4A2 were signature and independent prognostic biomarkers in GC patients with obviously different overall survival (OS). High expression of COL4A1, COL4A2, COL6A1, COL6A2, and COL6A3 was correlated with poorer prognosis of GC patients treated by surgery only, while higher expression of COL4A1 and COL11A1 correlated with poorer survival of patients treated by 5-fluorouracil-based adjuvant therapy. Our results indicate that overexpression of COL genes might be utilized as novel prognostic markers for GC and assist with therapy selection.

Construction of a circular RNA-microRNA-messengerRNA regulatory network in stomach adenocarcinoma

OBJECTIVES

Circular RNAs (circRNAs) can interact with microRNAs (miRNAs) to regulate gene expression in cancer cells. However, the roles of competitive endogenous RNA (ceRNA) networks consisting of differentially expressed circRNAs (DECs), miRNAs, and messenger RNAs (mRNAs) in stomach adenocarcinoma (STAD) remain unclear. This study was performed to explore novel regulatory networks in STAD.

METHODS

The circRNA expression profiles, as well as miRNA and mRNA sequence data of STAD, were retrieved from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA), respectively. Candidates were identified to construct a network through a comprehensive bioinformatics strategy. The expression of hub-genes identified by protein-protein interactions (PPI) was validated by quantitative reverse transcription (RT) polymerase chain reaction.

RESULTS

A total of 51 DECs were identified in the GSE83521 and GSE89143 datasets of GEO. A total of 11 448 differentially expressed mRNAs (DEMs) and 458 differentially expressed miRNAs (DEMIs) were obtained by RNA sequencing of TCGA-STAD. Prediction by using five online databases (Cancer-Specific CircRNA, CircInteractome, miRTarBase, miRDB, and TargetScan) resulted in the selection of 6 DECs, 6 DEMIs, and 36 DEMs to establish a circRNA-miRNA-mRNA regulatory network based on the interactions of circRNA-miRNA and miRNA-mRNA. Through PPI analysis, four hub-genes (COL10A1, COL5A2, COL4A1, and COL3A1) were discovered. Moreover, overexpressions of COL10A1, COL5A1, and COL4A1 were associated with a poor overall survival rate of patients with STAD. On the basis of TNM staging, we found that the expressions of COL10A1, COL5A2, and COL3A1 in T2, T3, and T4 was significantly higher than in T1. Hub-genes expressions were validated in STAD tissues and cell lines.

CONCLUSIONS

Our study provides a novel perspective on the regulatory mechanism of STAD involving ceRNAs including DECs, DEMIs, and DEMs.

MiR-129-5p suppresses gastric cancer cell invasion and proliferation by inhibiting COL1A1

Gastric cancer (GC) is one of the most lethal cancers worldwide. In this study, we aimed to explore the role of miR-129-5p, a newly identified miR-129 member, in GC cells as well as the potential mechanism of action. The results of reverse transcription - qualitative polymerase chain reaction (RT-qPCR) and Western Blot showed that miR-129 was downregulated in GC cells compared with normal ones. Using MTT, colony formation, wound healing assay, and a Transwell assay, we evaluated the proliferation, migration, and invasion abilities of transfected cells, and confirmed miR-129-5p as a tumor suppressor in GC. After a microarray analysis comparing different gene expressions in miR-129-5p transfected SGC-7901 cells, COL1A1 was selected for biggest fold-change and potential target of miR-129-5p predicted by TargetScan. Measured by RT-qPCR and Western blot, COL1A1 turned out to be upregulated in GC tissues and cells. We further confirmed the targeting relationship between miR-129-5p and COL1A1 by dual luciferase assay. By manipulating the expression of COL1A1 in SGC-7901 cells, cell proliferation, migration, and invasion were examined and the tumor-promoting function of COL1A1 was validated. Moreover, co-transfection of miR-129-5p mimics and COL1A1 attenuated the tumor-promoting effects induced by a single-transfection of COL1A1, and miR-129-5p inhibitor counteracted the tumor-suppressing effects of COL1A1 siRNA. Collectively, the data demonstrate the important functions of the miR-129-5p-COL1A1 axis in GC: miR-129-5p suppresses GC cell proliferation, migration, and invasion, by selectively inhibiting COL1A1. This study provides new therapeutic targets for the clinical treatment of GC.

CXCL12 methylation-mediated epigenetic regulation of gene expression in papillary thyroid carcinoma

Papillary thyroid carcinoma (PTC) is the most common type of thyroid cancer, and its incidence rate is rapidly growing. It is necessary to understand the pathogenesis of PTC to develop effective diagnosis methods. Promoter methylation has been recognized to contribute to the alterations in gene expression observed in tumorigenesis. Our RNA-seq data identified 1191 differentially expressed mRNAs and 147 differentially expressed lncRNAs in PTC. Next, promoter methylation of these genes was detected by reduced representation bisulfite sequencing (RRBS) technology and comprehensively analyzed to identify differential methylation. In total, 14 genes (13 mRNAs and 1 lncRNA), in which methylation was intimately involved in regulating gene expression, were proposed as novel diagnostic biomarkers. To gain insights into the relationships among these 14 genes, a core co-function network was constructed based on co-expression, co-function and co-methylation data. Notably, CXCL12 was identified as an essential gene in the network that was closely connected with the other genes. These data suggested that CXCL12 down-regulation in PTC may be caused by promoter hypermethylation. Our study was the first to perform an RRBS analysis for PTC and suggested that CXCL12 may contribute to PTC development by methylation-mediated epigenetic regulation of gene expression.

Transcriptional Regulatory Network Analysis for Gastric Cancer Based on mRNA Microarray

We aimed to screen the differential expressed genes (DEGs) and transcriptional factors (TFs) related to gastric cancer. GSE19826 microarray data downloaded from Gene Expression Omnibus was used to identify the differentially expressed genes (DEGs) and PPI network of DEGs were constructed by the Retrieval of Interacting Genes database. Pathway enrichment analysis of DEGs were performed by Gene Set Enrichment Analysis. Then, the transcriptional regulatory network was constructed based on TRANSFAC database. Finally, regulatory impact factor (RIF) of TF was calculated. We identified 446 DEGs including 209 up- and 237 down-regulated genes. These DEGs were mainly significantly enriched in 5 pathways including ECM receptor interaction (p = 0.013899), spliceosome (p = 0.025591), bladder cancer (p = 0.026316), focal adhesion (p = 0.047809) and WNT signaling pathway (p = 0.048077). PPI network with 247 nodes and 913 edges were constructed and COL5A2 was the hub node. Transcriptional regulatory network with 6 differently expressed TFs, 58 non-differently expressed TFs, 44 DEGs and 735 non-DEGs was constructed. Finally, top 5 TFs including CRX, TFAP4, NKX2-1, MYB and RARG with higher Z_RIF were screened. The identified DEGs such as COL5A2 and TOP2A, and TFs including EGR2, FOXM1, NKX2-1 and TFAP4 might be the critical genes and TFs for gastric cancer.

Identification of key genes associated with gastric cancer based on DNA microarray data

The present study aimed to identify genes with a differential pattern of expression in gastric cancer (GC), and to find novel molecular biomarkers for GC diagnosis and therapeutic treatment. The gene expression profile of GSE19826, including 12 GC samples and 15 normal controls, was downloaded from the Gene Expression Omnibus database. Differentially-expressed genes (DEGs) were screened in the GC samples compared with the normal controls. Two-way hierarchical clustering of DEGs was performed to distinguish the normal controls from the GC samples. The co-expression coefficient was analyzed among the DEGs using the data from COXPRESdb. The gene co-expression network was constructed based on the DEGs using Cytoscape software, and modules in the network were analyzed by ClusterOne and Bingo. Furthermore, enrichment analysis of the DEGs in the modules was performed using the Database for Annotation, Visualization and Integrated Discovery. In total, 596 DEGs in the GC samples and 57 co-expression gene pairs were identified. A total of 7 genes were enriched in the same module, for which the function was phosphate transport and which was annotated to participate in the extracellular matrix-receptor interaction pathway. These genes were collagen, type VI, α3 (COL6A3), COL1A2, COL1A1, COL5A2, thrombospondin 2, COL11A1 and COL5A1. Overall, the present study identified several biomarkers for GC using the gene expression profiling of human GC samples. The COL family is a promising prognostic marker for GC. Gene expression products represent candidate biomarkers endowed with great potential for the early screening and therapy of GC patients.

A systems biology approach to detect key pathways and interaction networks in gastric cancer on the basis of microarray analysis

The aim of the present study was to explore key molecular pathways contributing to gastric cancer (GC) and to construct an interaction network between significant pathways and potential biomarkers. Publicly available gene expression profiles of GSE29272 for GC, and data for the corresponding normal tissue, were downloaded from Gene Expression Omnibus. Pre‑processing and differential analysis were performed with R statistical software packages, and a number of differentially expressed genes (DEGs) were obtained. A functional enrichment analysis was performed for all the DEGs with a BiNGO plug‑in in Cytoscape. Their correlation was analyzed in order to construct a network. The modularity analysis and pathway identification operations were used to identify graph clusters and associated pathways. The underlying molecular mechanisms involving these DEGs were also assessed by data mining. A total of 249 DEGs, which were markedly upregulated and downregulated, were identified. The extracellular region contained the most significantly over‑represented functional terms, with respect to upregulated and downregulated genes, and the closest topological matches were identified for taste transduction and regulation of autophagy. In addition, extracellular matrix‑receptor interactions were identified as the most relevant pathway associated with the progression of GC. The genes for fibronectin 1, secreted phosphoprotein 1, collagen type 4 variant α‑1/2 and thrombospondin 1, which are involved in the pathways, may be considered as potential therapeutic targets for GC. A series of associations between candidate genes and key pathways were also identified for GC, and their correlation may provide novel insights into the pathogenesis of GC.

Prognostic value of type XXII and XXIV collagen mRNA expression in head and neck cancer patients

Squamous cell carcinoma of the head and neck (HNSCC) is the sixth most common type of cancer, affecting ~500,000 individuals worldwide annually. Collagen is the major constituent of the extracellular matrix component in tumors and plays a crucial role in tumor development. The aim of this study was to determine the mRNA expression of type XXI, XXII, XXIII and XXIV α1 collagen (COL21A1, COL22A1, COL23A1 and COL24A1, respectively) in head and neck squamous cell carcinoma (HNSCC) and investigate its correlation with disease progression. This study investigated the mRNA expression levels of COL21A1, COL22A1, COL23A1 and COL24A1 in 70 HNSCC primary samples and 44 matched pairs of tumor and adjacent normal mucosal tissues using quantitative polymerase chain reaction (qPCR). Expression data were compared to the clinicopathological variables in order to determine the correlation between expression and disease progression. Our results demonstrated that the mRNA levels of COL22A1 and COL24A1 were significantly higher in HNSCC tissues compared to those in the corresponding normal tissues from the same individuals (n=44; P<0.001 and P=0.019, respectively). The COL22A1 mRNA levels were found to be significantly associated with lymph node metastasis (P=0.018) and pathological stage (P=0.024), whereas the COL24A1 mRNA levels were significantly associated with tumor size (P=0.045). The high expression levels of COL22A1 and COL24A1 mRNA were statistically correlated with a decrease in disease-free survival (DFS) (log-rank test, P<0.001). The results of the multivariate logistic regression analysis revealed that high expression levels of the COL22A1 and COL24A1 gene pair were associated with a high odds ratio for recurrence of 14.62 (95% confidence interval: 2.77-77.26; P=0.002). Therefore, the upregulation of COL22A1 and COL24A1 mRNA may play a critical role in the progression of HNSCC and provide useful information as a prognostic predictor for HNSCC patients.

Converting a microarray signature into a diagnostic test: a trial of custom 74 gene array for clarification and prediction the prognosis of gastric cancer

Background

Gastric cancer (GC) is associated with high mortality rates and an unfavorable prognosis at advanced stages. In addition, there are no effective methods for diagnosing gastric cancer at an early stage or for predicting the outcome for the purpose of selecting patient-specific treatment options. Therefore, it is important to investigate new methods for GC diagnosis.

Methodology/Principal Findings

To facilitate its use in a diagnostic setting, a group of 74 genes with diagnostic and prognostic information was translated into a customized microarray containing a reduced set of 1,042 probes suitable for high throughput processing. In this report, we demonstrate for the first time that the custom mini-array can be used as a reliable diagnostic tool in gastric cancer. With an AUC value of 0.565 (95% CI 0.305-0.825) indicating a perfect test, the sensitivity and specificity of diagnosis from the ROC curve were calculated to be 70% and 80%, respectively.

Conclusions/Significance

The data clearly demonstrate the reproducibility and robustness of the small custom-made microarray. The array is an excellent tool for classifying and predicting the outcome of disease in gastric cancer patients.

MRI T2 hypointensity of metastatic brain tumors from gastric and colonic cancers

BACKGROUND

Metastatic brain tumors from gastric and colon cancers are frequently revealed by hypointensity on T2-weighted magnetic resonance images (MRIs). However, the reason for this T2 hypointensity has yet to be clarified. We hypothesize that it is due to collagen deposition within the tissues.

METHODS

Seven metastatic brain tumors, from 3 gastric cancers and 4 colon cancers were investigated. The degree of hypointensity of these tumors in T2-weighted images was quantitatively assessed as the ratio of gray-scale densities of tumor to brain using ImageJ. The result was compared with the amount of collagen in the resected specimens, which was quantified by ImageJ analysis software, utilizing the colour deconvolution method following Azan-Mallory staining. The degree of hypointensity was also compared with the ratio of viable epithelial component area/whole tissue area. Additionally, collagen distribution was studied by immunohistochemical staining.

RESULTS

There was a clear negative correlation between intensity in T2-weighted images of these metastatic tumors and the amount of collagen they contained (R (2) = 0.766). However, there was no significant correlation between the T2 intensity and the ratio of viable epithelial component. Immunohistochemical analysis revealed that collagen types I, III, VII, X, and XI were expressed in the epithelial components and types IV, V, and VI were expressed in the stromal areas of the metastatic tumors. Collagen deposition was observed not only in stromal fibrous areas, but also in cytoplasmic areas in these metastatic tumors.

CONCLUSIONS

Hypointensity of metastatic brain tumors arising from gastric and colonic cancers may be due to the accumulation of collagen in the tissues.

Identification of significant pathways in gastric cancer based on protein-protein interaction networks and cluster analysis

Gastric cancer is one of the most common and lethal cancers worldwide. However, despite its clinical importance, the regulatory mechanisms involved in the aggressiveness of this cancer are still poorly understood. A better understanding of the biology, genetics and molecular mechanisms of gastric cancer would be useful in developing novel targeted approaches for treating this disease. In this study we used protein-protein interaction networks and cluster analysis to comprehensively investigate the cellular pathways involved in gastric cancer. A primary immunodeficiency pathway, focal adhesion, ECM-receptor interactions and the metabolism of xenobiotics by cytochrome P450 were identified as four important pathways associated with the progression of gastric cancer. The genes in these pathways, e.g., ZAP70, IGLL1, CD79A, COL6A3, COL3A1, COL1A1, CYP2C18 and CYP2C9, may be considered as potential therapeutic targets for gastric cancer.

Spectral Clustering of Microarray Data Elucidates the Roles of Microenvironment Remodeling and Immune Responses in Survival of Head and Neck Squamous Cell Carcinoma

Purpose

To identify functionally related prognostic gene sets for head and neck squamous cell carcinoma (HNSCC) by unsupervised statistical analysis of microarray data.

Patients and Methods

Microarray analysis was performed on 14 normal oral epithelium and 71 HNSCCs from patients with outcome data. Spectral clustering (SC) analysis of the data set identified multiple vectors representing distinct aspects of gene expression heterogeneity between samples. Gene ontology (GO) analysis of vector gene lists identified gene sets significantly enriched within defined biologic pathways. The prognostic significance of these was established by Cox survival analysis.

Results

The most influential SC vectors were V2 and V3. V2 separated normal from tumor samples. GO analysis of V2 gene lists identified pathways with heterogeneous expression between HNSCCs, notably focal adhesion (FA)/extracellular matrix remodeling and cytokine-cytokine receptor (CR) interactions. Similar analysis of V3 gene lists identified further heterogeneity in CR pathways. V2CR genes represent an innate immune response, whereas high expression of V3CR genes represented an adaptive immune response that was not dependent on human papillomavirus status. Survival analysis demonstrated that the FA gene set was prognostic of poor outcome, whereas classification for adaptive immune response by the CR gene set was prognostic of good outcome. A combined FA&CR model dramatically exceeded the performance of current clinical classifiers (P < .001 in our cohort and, importantly, P = .007 in an independent cohort of 60 HNSCCs).

Conclusion

The application of SC and GO algorithms to HNSCC microarray data identified gene sets highly significant for predicting patient outcome. Further large-scale studies will establish the usefulness of these gene sets in the clinical management of HNSCC.