Development and validation of radiomic signatures of head and neck squamous cell carcinoma molecular features and subtypes

Chao, Huang; Cintra, Murilo Bicudo; Brennan, Kevin; Zhou, Mu; Colevas, A. Dimitrios; Fischbein, Nancy J.; Zhu, Shankuan; Gevaert, Olivier

doi:10.1016/j.ebiom.2019.06.034

Cited by 80 publications

(86 citation statements)

References 67 publications

(95 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In H&N squamous cell carcinoma, four subtypes of cancers were identified using expression profiling techniques 56 . Our study suggests that these subtypes, and potentially others, could be directly identified using baseline 18 F-FDG PET images currently acquired as part of the diagnostic management of HNC patients, as it has been shown similarly from CT scans 55 . Indeed, according to the values of the PET radiomic features, it would be possible to position a patient on a specific configuration in the identified modules, and from that, infer potentially altered biological pathways.…”

Section: Discussionsupporting

confidence: 52%

“…This could likely reflect tumor heterogeneity, which is quite clear at the molecular level in H&N 54 . Such an ability to differentiate the patients according to radiomics, together with inferring on the activity of molecular pathways, could have a strong impact on the quality of the prediction of the response to treatment and the recurrence-free survival 54,55 . Using this heterogeneity as a basis for a more personalized treatment is a very appealing perspective that holds a high potential benefit for patients 54 .…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Transcriptomics in cancer revealed by Positron Emission Tomography radiomics

Tixier

Cheze-Le-Rest

Schick

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Metabolic images from Positron Emission Tomography (PET) are used routinely for diagnosis, follow-up or treatment planning purposes of cancer patients. in this study we aimed at determining if radiomic features extracted from 18 F-Fluoro Deoxy Glucose (FDG) PET images could mirror tumor transcriptomics. In this study we analyzed 45 patients with locally advanced head and neck cancer (H&N) that underwent FDG-PET scans at the time of diagnosis and transcriptome analysis using RnAs from both cancer and healthy tissues on microarrays. Association between pet radiomics and transcriptomics was carried out with the Genomica software and a functional annotation was used to associate PET radiomics, gene expression and altered biological pathways. We identified relationships between PET radiomics and genes involved in cell-cycle, disease, DNA repair, extracellular matrix organization, immune system, metabolism or signal transduction pathways, according to the Reactome classification. Our results suggest that these FDG PET radiomic features could be used to infer tissue gene expression and cellular pathway activity in H&n cancers. these observations strengthen the value of radiomics as a promising approach to personalize treatments through targeting tumor-specific molecular processes. 18 F-FDG Positron emission tomography (PET) imaging is largely used for diagnostic purposes in several cancer types, allowing accurate disease staging 1,2 , but it is also gaining ground for therapy applications, including monitoring treatment response and planning in the field of external beam radiotherapy 3. Although visual interpretation may be sufficient for diagnosis, a (semi)quantitative analysis of PET data for image guided therapy applications is most frequently necessary. The simplest parameter usually extracted from PET images used in clinical practice is the maximum standardized uptake value (SUV max) corresponding to the highest single voxel intensity within a region of interest. On the one hand, SUV max has often been reported as a biomarker with potential for improving overall patient management, including the prediction of response to therapy and survival in several cancers 4-6. On the other hand, numerous studies have shown that the predictive/prognostic value of SUV max can be limited, particularly when using images to characterize tumors on baseline PET images 7-9. For this reason, there has been an increasing interest in extracting additional parameters from 18 F-FDG PET images with the objective of more fully characterizing the entire tumor uptake. Most studies initially focused on the delineation of the metabolically active tumor volume (MATV) and associated SUV measurements (such as the mean value or total lesion glycolysis) 10,11 , while more recent studies have concentrated on parameters characterizing the shape or activity distribution within the tumor. These parameters, today known as radiomic features 14 , include 1 st-order image features such as (cumulative) intensity histograms, geometrical tumor shape descriptors...

show abstract

Section: Discussionsupporting

confidence: 52%

Section: Discussionmentioning

confidence: 99%

Transcriptomics in cancer revealed by Positron Emission Tomography radiomics

Tixier

Cheze-Le-Rest

Schick

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…22,[46][47][48][49] Identification of optimal classification methods and the best radiomic features for prognostic analyses in head and neck cancer patients may indeed broaden the scope of radiomics in precision oncology. [48][49][50][51][52] In this study we aim to investigate the potential value of radiomic features extracted from gross tumor volumes (GTVs) delineated on diagnostic CE-CT images of oropharyngeal cancer patients as potential biomarkers of HPV status.…”

Section: Introductionmentioning

confidence: 99%

Application of radiomics for the prediction of HPV status for patients with head and neck cancers

et al. 2020

View full text Add to dashboard Cite

Purpose To perform radiomic analysis of primary tumors extracted from pretreatment contrast‐enhanced computed tomography (CE‐CT) images for patients with oropharyngeal cancers to identify discriminant features and construct an optimal classifier for the characterization and prediction of human papilloma virus (HPV) status. Materials and methods One hundred and eighty seven patients with oropharyngeal cancers with known HPV status (confirmed by immunohistochemistry—p16 protein testing) were retrospectively studied as follows: Group A: 95 patients (19HPV− and 76HPV+) from the MICAII grand challenge. Group B: 92 patients (52HPV− and 40HPV+) from our institution. Radiomic features (172) were extracted from pretreatment diagnostic CE‐CT images of the gross tumor volume (GTV). Levene and Kolmogorov–Smirnov's tests with absolute biserial correlation (>0.48) were used to identify the discriminant features between the HPV+ and HPV− groups. The discriminant features were used to train and test eight different classifiers. Area under receiver operating characteristic (AUC), positive predictive and negative predictive values (PPV and NPV, respectively) were used to evaluate the performance of the classifiers. Principal component analysis (PCA) was applied on the discriminant feature set and seven PCs were used to train and test a generalized linear model (GLM) classifier. Results Among 172 radiomic features only 12 radiomic features (from 3 categories) were significantly different (P < 0.05, |BSC| > 0.48) between the HPV+ and HPV− groups. Among the eight classifiers trained and applied for prediction of HPV status, the GLM showed the highest performance for each discriminant feature and the combined 12 features: AUC/PPV/NPV = 0.878/0.834/0.811. The GLM high prediction power was AUC/PPV/NPV = 0.849/0.731/0.788 and AUC/PPV/NPV = 0.869/0.807/0.870 for unseen test datasets for groups A and B, respectively. After eliminating the correlation among discriminant features by applying PCA analysis, the performance of the GLM was improved by 3.3%, 2.2%, and 1.8% for AUC, PPV, and NPV, respectively. Conclusions Results imply that GTV's for HPV+ patients exhibit higher intensities, smaller lesion size, greater sphericity/roundness, and higher spatial intensity‐variation/heterogeneity. Results are suggestive that radiomic features primarily associated with the spatial arrangement and morphological appearance of the tumor on contrast‐enhanced diagnostic CT datasets may be potentially used for classification of HPV status.

show abstract

“…The patients' pronunciation, vision, hearing, breathing, and social communication become deeply affected by these diseases, with their quality of life becoming greatly reduced in addition to a heavy social burden [26][27][28]. In recent years, many diagnostic molecular markers related to HNSCC have been identi ed [29][30][31][32], but it is di cult to achieve accurate early detection, which may be the most important reason for the high mortality of patients with HNSCC. Therefore, there is an urgent need to develop an effective means of early detection, diagnosis, and treatment to improve the treatment of HNSCC.…”

Section: Discussionmentioning

confidence: 99%

Co-expression Network Identification and Clinical Prognostic Evaluation of Hub Genes in Head and Neck Squamous Cell Carcinoma

Feng¹,

Han²,

Yu³

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: According to statistics, even with active treatment, the recurrence rate of HNSCC is at 40%-50%. Head and neck cancer remains a challenge for otolaryngologists. Therefore, the identification of new biomarkers is an urgent need for the diagnosis, treatment, and prognosis of malignant tumors of the head and neck. Methods: In this study, transcriptome data from The Cancer Genome Atlas (TCGA) database and the Gene Expression Omnibus (GEO) database were used to identify differentially expressed genes (DEGs). Weighted gene co-expression network analysis (WGCNA) was performed to identify gene modules and hub genes related to head and neck squamous cell carcinoma (HNSCC). Protei-Protei interaction(PPI) network and Cytoscape software were used to analyze the protein interaction network. HNSCC clinical data from the TCGA and Gene Expression Profile Interactive Analysis 2 databases were used to analyze the survival rate of hub genes, and the correlation between hub genes and tumor stage was further analyzed.Results: A total of 2836 and 570 DEGs were identified from the TCGA expression data and GEO gene chip datasets, respectively. We found that the green module had the highest correlation with HNSCC. A total of 15 hub genes were also identified. In the Human Protein Atlas database, we found that thioredoxin reductase 1 (TXNRD1) was overexpressed in HNSCC tumors compared with normal tissues at the transcriptional level. Survival analysis also suggested that TXNRD1 was a poor prognostic factor for HNSCC.Conclusion: Our results indicate that TXNRD1 is very likely to be identified as a potential biomarker and target for HNSCC. However, further research is required to fully reveal its role in HNSCC pathogenesis as well as its value as a prognostic biomarker.

show abstract

Development and validation of radiomic signatures of head and neck squamous cell carcinoma molecular features and subtypes

Cited by 80 publications

References 67 publications

Transcriptomics in cancer revealed by Positron Emission Tomography radiomics

Transcriptomics in cancer revealed by Positron Emission Tomography radiomics

Application of radiomics for the prediction of HPV status for patients with head and neck cancers

Co-expression Network Identification and Clinical Prognostic Evaluation of Hub Genes in Head and Neck Squamous Cell Carcinoma

Contact Info

Product

Resources

About