Combined Imaging and RNA Analyses to Predict Head and Neck Cancer Recurrence

Washington University, Saint Louis MO

Investigators

Linked publications, trials & patents

Abstract

Title: Combined Imaging and RNA Analyses to Predict Head and Neck Cancer Recurrence ABSTRACT Head and neck squamous cell carcinomas (HNSCCs) encompass a diverse group of tumors that generally are aggressive in their biological behavior. Recurrence is the most common form of treatment failure for HNSCC patients receiving standard therapy. Approximately 50% of HNSCC cases will develop recurrence, and the 5- year survival rate for recurrent patients is only 16~36%. Early prediction of HNSCC recurrence is one of the most challenging yet important tasks for stratifying HNSCC patients and supporting personalized treatment strategies to improve patient care. We and others have shown that human ribonucleic acids (RNAs) and human papillomavirus (HPV) RNAs are promising biomarkers and play critical regulatory roles in HNSCC. Radiologic imaging biomarkers derived from PET, CT, and MR imaging data have shown promise in stratifying patients with favorable and unfavorable prediction for treatment response. Their non-invasive characteristics also allow for convenient and longitudinal monitoring of tumor progression and heterogeneous response during the treatment course. Histologic images provide key information about microscopic structure of cells and tissues of organisms. Recent reports and our preliminary studies have shown that histologic imaging biomarkers, can aid in clinical decision-making by identifying metastases, subtyping and grading tumors, and predicting treatment failures. Clinicopathologic biomarkers show prognostic values through retrospective studies. Still, many HNSCC patients have recurred tumors despite favorable prediction by these biomarkers. The major goal of this study is to develop a comprehensive and robust computational model for early prediction of HNSCC treatment failures leading to tumor recurrence. We will integrate our recently developed advanced learning-based techniques to build prognostic models using about 1,200 patient cases collected from two institutions. The prognostic model will form a solid basis for individualized care of HNSCC patients based on predicted treatment outcomes. Moreover, our work is expected to discover the correlations among multimodal data, leading to dynamic patient stratification to support adaptive treatment strategies.