An evolutionary study of head and neck cancer before and following radiotherapy

BACKGROUND Head and Neck cancer (HNC) is the seventh most common cancer globally, affecting ~660,000 patients per year worldwide, with a 50% mortality rate. Both incidence and mortality have risen in recent years, in both high and lower/middle income countries. The main treatment options are surgery followed by adjuvant radiotherapy or definitive concurrent chemoradiation (CRT).

However, around half of patients relapse following CRT and require neck salvage surgery.

Despite recent advances in immunotherapy, treatment decisions are still generally made on the basis of HPV status and stage, without consideration of molecular factors.

AIMS • Define the evolutionary trajectories of HNCs in the absence of treatment, to provide a baseline against which to investigate the effects of treatment • Characterise copy number aberrations by applying shallow WGS to recurrent tumours from 48 patients and to matched primary samples from 19 patients • Grow epithelioids from up to 10 primary and 5 recurrent oropharyngeal tumours, and characterise them using spatial transcriptomics and shallow WGS. • Identify cell-cell interactions between cell types • Classify tumours according to their evolutionary responses to radiotherapy METHODS In Aim 1, we will analyse Whole Genome Sequencing data from previous consortia studies including TCGA, ICGC and the 100,000 genomes project using methods previously developed in the Wedge group, including Battenberg, DPClust and Plackett-Luce timing models.

In Aim 2 we will use QSNASeq to call copy number aberrations and CNSignatures to infer signatures from copy number calls.

Epithelioids represent enduring, self-sustaining 3D tissue cultures, directly derived from tumours and their associated normal tissue.

In Aim 3 we will establish Epithelioids from Head and Neck cancers, using methods previously developed in the Antoran lab. In Aim 4, cell-cell interactions will be identified using network based methods such as NetworkX.

In Aim 5, evolutionary trajectories will be inferred using methods previously developed in the Wedge lab such as Plackett-Luce timing models.

HOW THE RESULTS OF THIS RESEARCH WILL BE USED The construction of evolutionary trajectories enables the prediction of future tumour development at an early stage.

Over the medium term, this will enable clinicians to make treatment decisions from a knowledge base informed by likely response of an individual tumour to radiotherapy.

This study represents an exemplar that may be applied to cancers in other tissue types and subject to a range of treatments.