Radiogenomic predictors of treatment response in head and neck squamous cell carcinoma

University Of Kansas Medical Center, Kansas City KS

Investigators

Linked publications & trials

Abstract

Despite recent advances in treatment of head and neck squamous cell carcinoma (HNSCC), five-year overall survival (OS) has remained poor due to high rates of disease progression after treatment. A lack of reliable prognostic biomarkers limits ability to predict disease progression in non-human papillomavirus (HPV) mediated HNSCC. Among candidates for salvage surgery in the setting of disease progression after primary chemoradiotherapy, 5-year OS is estimated at just 16%. Radiomics, involving analysis of large numbers of quantitative tumor imaging features, has been used to develop image-based prognostic models. Delta radiomics incorporates differences between pre- and post-treatment images, thereby capturing quantitative measures of treatment response. To address critical need for prognostic biomarkers for patients with HNSCC, we aim to develop and validate radiogenomic models to predict 2-year overall and progression-free survival using tumor-derived genomic biomarkers and delta radiomic features derived from pre- and post-treatment contrast-enhanced computed tomography. Prior studies developed highly accurate models of cancer treatment response using delta radiomics, but not in patients with locoregionally advanced HNSCC. Few studies have evaluated reliable and concordant genomic and radiomic prognostic phenotypes in HNSCC. We hypothesize that a multi-‘omic model integrating radiogenomic features will accurately predict 2-year overall and progression-free survival. We aim to 1) develop a fully annotated, patient-centered, clinical outcome and radiographic dataset with associated tumor samples for patients with locoregionally advanced HPV-negative HNSCC treated with definitive chemoradiotherapy and 2) develop radiogenomic models to predict 2-year OS and PFS using tumor-derived genomic biomarkers and delta radiomic features derived from pre- and post-treatment contrast-enhanced computed tomography. We will engage the Quantitative ‘Omics Core to extract radiomic features and explore radiogenomic phenotypes associated with 2-year OS and PFS. We will use the Biobanking and Biomarker Validation core to generate a HNSCC patient-derived tissue microarray to identify mutations and quantify gene expression relevant to epidermal growth factor receptor, MAPK-associated protein kinase 2 (MK2) and hypoxia pathways. With the Patient and Community Engagement core, we will ensure that patient priorities are consistently represented in the design. Long-term goal is to support future trials aimed at early identification of patients at high-risk of disease progression who are likely to benefit from precision treatment approaches.