Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Feb 12;150(2):91.
doi: 10.1007/s00432-024-05606-8.

An angiogenesis-associated gene-based signature predicting prognosis and immunotherapy efficacy of head and neck squamous cell carcinoma patients

Bangjie Chen #  1   2 , Yanxun Han #  2 , Shuyan Sheng #  2 , Jianyi Deng #  2 , Emely Vasquez  3 , Vicky Yau  4 , Muzi Meng  5   6 , Chenyu Sun  7 , Tao Wang  8 , Yu Wang  8 , Mengfei Sheng  1   9 , Tiangang Wu  1 , Xinyi Wang  2 , Yuchen Liu  2 , Ning Lin  10 , Lei Zhang  11 , Wei Shao  12   13
Affiliations

An angiogenesis-associated gene-based signature predicting prognosis and immunotherapy efficacy of head and neck squamous cell carcinoma patients

Bangjie Chen et al. J Cancer Res Clin Oncol. .

Abstract

Objectives: To develop a model that can assist in the diagnosis and prediction of prognosis for head and neck squamous cell carcinoma (HNSCC).

Materials and methods: Data from TCGA and GEO databases were used to generate normalized gene expression data. Consensus Cluster Plus was used for cluster analysis and the relationship between angiogenesis-associated gene (AAG) expression patterns, clinical characteristics and survival was examined. Support vector machine (SVM) and least absolute shrinkage and selection operator (LASSO) analyzes and multiple logistic regression analyzes were performed to determine the diagnostic model, and a prognostic nomogram was constructed using univariate and multivariate Cox regression analyses. ESTIMATE, XCELL, TIMER, QUANTISEQ, MCPCOUNTER, EPIC, CIBERSORT-ABS, CIBERSORT algorithms were used to assess the immune microenvironment of HNSCC patients. In addition, gene set enrichment analysis, treatment sensitivity analysis, and AAGs mutation studies were performed. Finally, we also performed immunohistochemistry (IHC) staining in the tissue samples.

Results: We classified HNSCC patients into subtypes based on differences in AAG expression from TCGA and GEO databases. There are differences in clinical features, TME, and immune-related gene expression between two subgroups. We constructed a HNSCC diagnostic model based on nine AAGs, which has good sensitivity and specificity. After further screening, we constructed a prognostic risk signature for HNSCC based on six AAGs. The constructed risk score had a good independent prognostic significance, and it was further constructed into a prognostic nomogram together with age and stage. Different prognostic risk groups have differences in immune microenvironment, drug sensitivity, gene enrichment and gene mutation.

Conclusion: We have constructed a diagnostic and prognostic model for HNSCC based on AAG, which has good performance. The constructed prognostic risk score is closely related to tumor immune microenvironment and immunotherapy response.

Keywords: Angiogenesis-associated gene; Diagnosis; HNSCC; Immunotherapy; Prognostic signature.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1
Fig. 1
Consensus clustering analysis of HNSCC patients based on AAGs expression patterns. A Consensus matrix heatmap defining two clusters (k = 2) and their correlation area. B PCA analysis indicating an obvious difference in transcriptomes between the two subgroups. C Survival analysis of different subtypes. D Differences in clinicopathologic characteristics and expression levels of AAGs between the two distinct subgroups
Fig. 2
Fig. 2
HNSCC clusters based on angiogenesis-related DEGs. A, B GO and KEGG enrichment analysis of DEGs in two HNSCC clusters; C Two HNSCC clusters Correlation between group and TME score; D ssGSEA calculated the abundance of 23 infiltrating immune cell types in two HNSCC clusters; E CIBERSORT calculated the correlation of 22 immune cells and TME scores between cluster 1 and cluster 2 patients Difference (*p < 0.05; **p < 0.01; ***p < 0.001)
Fig. 3
Fig. 3
Analysis of pathway enrichment and TME differences between two different clusters. A GSVA analysis between two different clusters; B Expression difference analysis of immune checkpoints in cluster 1 and cluster 2; C Expression difference analysis of HLA molecules in two HNSCC clusters (*p < 0.05; **p < 0.01; ***p < 0.001)
Fig. 4
Fig. 4
Construction and validation of HNSCC diagnostic signature. A, B Multivariate least absolute shrinkage A and selection operator (LASSO) B Regression analysis; C Support vector machine analysis; D The intersection of the genes screened by SVM and multivariate lasso analysis was used to construct the diagnostic signature; E Sensitivity and specificity of the ROC curve based on the TCGA dataset; F Sensitivity and specificity of the ROC curve derived based on the GSE14520 dataset
Fig. 5
Fig. 5
Construction of HNSCC prognostic signature. A Forest plot showed that the prognosis related AAGs were preliminarily obtained by univariate regression analysis; B Forest plot showed that the AAGs used to construct the prognosis model was further determined by multivariable regression analysis; C, D Scatter plot of risk score distribution and patient survival status in the training and validation cohorts, respectively; E, F Ranking plot and median value of risk score in the training and validation cohorts, respectively; G, H Kaplan–Meier analysis comparing OS between high- and low-risk groups in the training and validation cohorts, respectively; I, J ROC curves reflecting the sensitivity and specificity of the signature for predicting 1-year, 3-year and 5-year survival in the training and validation cohorts, respectively; K, L Expression pattern comparison of the six selected prognostic genes used to construct the signature between the high- and low-risk groups in the training and validation cohorts, respectively
Fig. 6
Fig. 6
Construction and validation of a nomogram. A Construction of the nomogram. B 1-, 3-, and 5-year ROC curves of the nomogram in the entire cohort. C–E ROC curves of the nomogram and each independent prognostic factor for 1-, 3-, and 5-year OS, respectively. F–H DCA curves of the nomogram and each independent prognostic factor for 1-, 3-, and 5-year OS, respectively
Fig. 7
Fig. 7
Gene set enrichment analysis identifies biological pathways and processes associated with risk scores within the TCGA cohort, while multiple algorithms determined levels of multiple immune cell infiltration associated with risk scores. A GSEA analysis of the high-risk group based on the GO database; B GSEA analysis of the high-risk group based on the KEGG database; C GSEA analysis of low-risk groups based on GO database; D GSEA analysis of low-risk group based on KEGG database; E, F The relationship between AAG risk score and immune cell infiltration in HNSCC samples, using Timer, CiberSort, Xcell, QuantieQ, MCPCounter, EPIC, and CiberSor algorithms
Fig. 8
Fig. 8
Association between AAG-based risk scores and immunotherapy sensitivity. A IPS, IPS-PD-1 blocker, IPS-CTLA-4 blocker, IPS-CTLA-4 and PD-1 blocker scores between high- and low-risk groups; B TIDE scores between high- and low-risk groups, respectively; C MSI scores between high- and low-risk groups; D T cell dysfunction scores between high- and low-risk groups; E T cell exclusion scores between high- and low-risk groups; F Correlation between AAG score and mDNAsi stemness index; G Correlation between AAG score and mRNAsi stemness index; H, I The mutational profiles of HNSCC patients between different risk groups. (*** p < 0.001)
Fig. 9
Fig. 9
Comparison of IC50 of multiple anti-tumor drugs in high-risk and low-risk groups
Fig. 10
Fig. 10
The IHC staining results showed that the expression levels of OLR1, PDGFA, S100A4, MSX1, and APOH in HNSCC tissues were higher than those in adjacent normal tissues, except for SERPIAN5, which was lower in HNSCC tissues than in adjacent normal tissues
Fig. 11
Fig. 11
The western blotting and RT-qPCR results showed that the expression levels of OLR1, PDGFA, S100A4, MSX1, and APOH in HNSCC tissues were higher than those in adjacent normal tissues, except for SERPIAN5, which was lower in HNSCC tissues than in adjacent normal tissues
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Al-Mterin MA, Murshed K, Alsalman A, Abu-Dayeh A, Elkord E. Associations of different immune checkpoints-expressing CD4(+) Treg/ T cell subsets with disease-free survival in colorectal cancer patients. BMC Cancer. 2022;22:601. doi: 10.1186/s12885-022-09710-1. - DOI - PMC - PubMed
    1. Bader JE, Voss K, Rathmell JC. Targeting metabolism to improve the tumor microenvironment for cancer immunotherapy. Mol Cell. 2020;78:1019–1033. doi: 10.1016/j.molcel.2020.05.034. - DOI - PMC - PubMed
    1. Bejarano L, Jordāo MJC, Joyce JA. Therapeutic targeting of the tumor microenvironment. Cancer Discov. 2021;11:933–959. doi: 10.1158/2159-8290.Cd-20-1808. - DOI - PubMed
    1. Butkiewicz D, Gdowicz-Klosok A, Krzesniak M, Rutkowski T, Krzywon A, Cortez AJ, Dominczyk I, Skladowski K. Association of genetic variants in ANGPT/TEK and VEGF/VEGFR with progression and survival in head and neck squamous cell carcinoma treated with radiotherapy or radiochemotherapy. Cancers (basel) 2020 doi: 10.3390/cancers12061506. - DOI - PMC - PubMed
    1. Chalabi M, Fanchi LF, Dijkstra KK, Van den Berg JG, Aalbers AG, Sikorska K, Lopez-Yurda M, Grootscholten C, Beets GL, Snaebjornsson P, et al. Neoadjuvant immunotherapy leads to pathological responses in MMR-proficient and MMR-deficient early-stage colon cancers. Nat Med. 2020;26:566–576. doi: 10.1038/s41591-020-0805-8. - DOI - PubMed