Development of a prediction model for radiotherapy response among patients with head and neck squamous cell carcinoma based on the tumor immune microenvironment and hypoxia signature

• Published in: Cancer medicine (Malden, MA) MA), 2022-12, Vol.11 (23), p.4673-4687
• Main Authors: Zhu, Guang‐Li, Yang, Kai‐Bin, Xu, Cheng, Feng, Rui‐Jia, Li, Wen‐Fei, Ma, Jun
• Format: Article
• Language: English
• Subjects: Algorithms; Cancer therapies; Chemotherapy; DNA damage; Gene expression; gene signature; Genomes; Head & neck cancer; Head and neck carcinoma; Head and Neck Neoplasms - genetics; Head and Neck Neoplasms - radiotherapy; head and neck squamous cell carcinoma; Humans; Hypoxia; Hypoxia - genetics; immune; Immune system; Immunotherapy; Medical prognosis; Metastases; Metastasis; Microenvironments; Penicillins; Prediction models; Prognosis; Radiation Oncology; Radiation therapy; Radiosensitivity; Signal transduction; Squamous cell carcinoma; Squamous Cell Carcinoma of Head and Neck - genetics; Squamous Cell Carcinoma of Head and Neck - radiotherapy; Transcriptomics; Tumor Microenvironment - genetics; Tumors
• ISSN: 2045-7634; 2045-7634
• DOI: 10.1002/cam4.4791

Introduction The immune system and hypoxia are major factors influencing radiosensitivity in patients with different cancer types. This study aimed at developing a model to predict radiotherapy response in patients with head and neck squamous cell carcinoma (HNSCC) based on the tumor immune microenvironment and hypoxia signature. Materials and Methods We first evaluated the hypoxia status and tumor immune microenvironment in the Cancer Genome Atlas (TCGA) cohort by using transcriptomic data. Differentially expressed genes (DEGs) were identified between the “high immunity and low hypoxia” and “low immunity and high hypoxia” groups and those DEGs significantly associated with disease‐specific survival in the univariate Cox regression analysis were selected as the prognostic DEGs. We selected the immune hypoxia–related genes (IHRGs) by intersecting prognostic DEGs with immune and hypoxia gene sets. We used the IHRGs to train a multivariate Cox regression model in the TCGA cohort, based on which we calculated the IHRG prognostic index (IHRGPI) for each patient and validated its efficacy in predicting radiotherapy response in the Gene Expression Omnibus cohorts. Furthermore, we explored potential mechanisms and effective combinational treatment strategies for different IHRGPI groups. Results Five IHRGs were used to construct the IHRGPI, which was used to dichotomize the cohorts. The patients with lower IHRGPI showed a better radiotherapy response across different cohorts and endpoints, including overall survival, progression‐free survival, and recurrence‐free survival (p