Project 2: Predicting Treatment Responses Using Single Cell RNA Sequencing and Bioengineered Patient-derived Organotypic Models of HNC

PROJECT SUMMARY - PROJECT 2
Our current inability to accurately predict treatment outcomes for head and neck cancer (HNC) patients
represents a major challenge for clinicians and undoubtedly contributes to both the poor overall and progression
free survival rates in advanced disease. Currently no predictive biomarkers are used clinically for definitive
therapies, thus, there is a compelling need to develop new biomarkers and tools to improve clinical decision
making. Functional biomarkers that can provide multiple orthogonal endpoints are well suited to reporting on a
complex and dynamic environment such as the tumor microenvironment (TME). For this reason, we intend to
investigate HNC biomarkers both directly in tumor samples and in a bioengineered patient-specific model to
create a novel suite of endpoints. We will use state of the art single cell RNA sequencing (scCITE-seq), protein
expression signatures from tumor tissue microarrays (TMA’s) and analysis of circulating tumor cells (CTC’s). We
will utilize this multi-omic, patient-specific data set to identify and validate signatures of treatment efficacy and
stratify patient outcomes. We will then test the feasibility of using patient specific bioengineered models to inform
patient care in a clinical pilot study. The bioengineered model of the HNC TME is made entirely of cells derived
from the same patient tumor sample, from the same patient cohort used for CITEseq, TMA and CTC analysis.
These microscale patient-specific (built from the individual patients own cells) bioengineered models recapitulate
the TME architecture, containing a HNC epithelial spheroid surrounded by a matrix containing fibroblasts and
immune cells and flanked by blood and lymphatic microvessels. Our specific aims are: 1) Evaluate the ability of
HNC patient-specific bioengineered models to predict treatment efficacy, where we will build patient-specific
bioengineered models for 22 HNC patients (representing HPV-positive and HPV-negative disease and patients
treated with primary surgery with (chemo)radiation or primary chemoradiation) and treat them with the same
treatment the patient receives. Metrics of treatment success in the models will be correlated with actual patient
outcomes including progression free survival. 2) Identify HNC biomarkers using scCITE-seq and TMA, where
we will perform scCITE-seq and will correlate gene expression and cell populations with patient outcomes to
investigate existing putative biomarkers and identify additional novel biomarkers. Biomarkers will be further
investigated in a TMA and in CTC’s. 3) Use of bioengineered models to inform dose de-escalation in a clinical
pilot study, where tissue will be acquired from surgery from 24 HPV+ HNC patients and used for patient-specific
bioengineered model creation. Models will be treated to determine the radiosensitivity of a patient’s tumor and
to stratify intermediate risk patients between 50 or 60 Gy treatment groups. Primary endpoints will focus on
feasibility of model integration with secondary endpoints including local control. The successful completion of
these aims will provide powerful new tools for the stratification of HNC patients and improved clinical decision
making to help inform the most effective treatment selections for individual HNC patients in the future.

Grant Number: 5P50CA278595-09
NIH Institute/Center: NIH
Principal Investigator: David Beebe