SIEVERT: Studying tumour EVolution during External beam Radiation Therapy

Background: Radioresistance is a significant cause of morbidity and mortality in patients diagnosed with cancer.

Whilst some malignancies exhibit an intrinsic resistance to radiotherapy, others acquire or strengthen mechanisms to counteract radiotherapy during fractionated treatment.

Within these tumours, subclones that demonstrate a survival advantage to irradiation may emerge de novo or through positive selection. The genomic and transcriptional events underpinning this evolution are poorly understood.

Aims: We will capitalise on relevant experience within and outside of RadNet in Cambridge and UCL to repurpose two novel technologies (Cytosponge and EPIgenetic expression interference from Cell tumour DNA (EPICseq)) to study radiotherapy-related clonal evolution at the genomic and transcriptomic level by direct tumour sampling and through analyses of circulating tumour DNA (ctDNA) in oesophageal (OAC) and lung (LUAC) adenocarcinomas.

Methods: We will adapt the Cytosponge - initially designed for the screening and surveillance of patients with Barrett's oesophagus - to repeatedly sample OAC tumour tissue at sequential timepoints during chemoradiotherapy (CRT) in 20 OAC patients.

The impact of CRT on tumour clonality and the underpinning genomic and transcriptomic evolution will be mapped. ctDNA will be extracted from paired blood samples, from which changes in tumour gene expression will be inferred by deep sequencing around transcriptional start sites in order to determine promoter fragment entropy using EPICseq; the accuracy of which will be determined by correlation with Cytosponge-sampled tumour tissue.

Alongside, in vitro analyses using 2D LUAC cell lines and 3D OAC patient-derived organoids will assess the ability of a custom gene panel, the expression of which is inferred by ctDNA-based EPICseq analysis, to identify radiotherapy-treated samples.

As a further proof of principle, the same approach will be used to predict radiation response using an existing biobank of sequential bloods from 50 radiation-treated LUAC patients and the 20 OAC-treated, Cytosponge-sampled, patients.

How the results of this research will be used: This research will provide a fundamentally greater understanding of how and when to radiosensitise tumours.

Crucially, it will also provide a proof-of-principle for the real-time blood-based analysis of the tumour transcriptome in patients receiving radiotherapy.

This provides a platform on which translation-rich trials can be built that (1) develop adaptive and personalised radiosensitisation strategies that specifically target each tumour's radioresistance mechanism, (2) develop early stop/go signatures for radiotherapy and (3) use radiotherapy as a novel tool to elicit druggable signalling pathways; studies we will catalyse by establishing a relevant consortium within RadNet.