Developing biomarker-driven treatments for oesophagogastric cancer

Abstract

Introduction Despite recent advances in both genetic characterisation and development of novel targeted agents the outlook for advanced oesophagogastric (OG) cancers remain poor and there is a need for improved biomarker-driven treatment approaches. The work presented in this thesis details the design, set-up and oversight of the iMYC trial, a biomarker-driven phase 2 clinical trial in advanced OG cancer incorporating a novel prospective screening programme, as well as results from the application of differing methodologies of biomarker identification including single gene analysis and whole genome sequencing techniques, applied to both tumour and circulating tumour (ct)DNA. Methods iMYC is an open label, phase II non-randomised trial assessing the efficacy of ibrutinib monotherapy in advanced pre-treated MYC and/or HER2 amplified OG cancer. As part of the trial a screening programme utilising a fluorescent in-situ hybridisation (FISH) assay for assessment of tumour MYC amplification was instituted, and a further digital droplet (dd)PCR assay developed and applied to both tumour and ctDNA. Low coverage whole genome sequencing (LcWGS) was applied to ctDNA extracted from blood samples taken as part of a prior translational research study. Samples were taken at both baseline and on progression after comparable first-line systemic chemotherapy in advanced OG adenocarcinoma patients. Analysis of somatic copy number alterations (SCNAs), genomic instability and evolution of genomic changes over the course of chemotherapy was assessed. Results One hundred and thirty five archival tumour specimens have undergone successful FISH analysis as part of the iMYC trial, with 23% displaying evidence of MYC amplification, most commonly in the presence of polysomy. Inter-tumour heterogeneity was observed, with the percentage of cancer cells harbouring MYC amplification ranging widely between samples (median 51%, range 11-94%). Intra-tumoural clonal diversity of MYC amplification was also observed, with a significant degree of variance in amplification ratios (Bartlett's test for equal variance p<0.001), and an association between greater variance in MYC amplification and improved outcome with first-line chemotherapy. The ddPCR assay was most accurate in quantifying MYC amplification in tumour-derived DNA from cases with a high proportion (>70%) of amplified cells within the tumour specimen, but was not reliable in samples containing a low proportion of amplified cells or in ctDNA from this cohort of patients. To date eight patients have undergone treatment within the main component of the trial. Thirty samples underwent LcWGS analysis of ctDNA at a baseline time-point. In 23/30 (76.7%) cases an analysable SCNA profile was found. The presence of liver metastases, primary tumour in-situ or of oesophageal or junctional tumour location predicted for a high circulating tumour DNA fraction, and concordance was seen with prior targeted tumour sequencing results as well as additional further amplification events identified. SCNA profiles changed during chemotherapy, indicating that cancer cell populations evolved during treatment, however no recurrent SCNA changes were acquired at progression. Conclusions The iMYC trial represents the first attempt to prospectively identify and target MYC amplifications in OG cancer. MYC amplifications have been found to be a commonly occurring event, and heterogeneity of amplification patterns has been observed. The sensitivity of a novel ddPCR assay appears limited to tumour samples displaying a high degree of MYC amplification only. The successful analysis of SCNA profiles in ctDNA, as well as the detection of clinically significant amplification events, suggest that LcWGS may be a promising technology to assess the genomic landscapes of metastatic OG cancers. Tracking the evolution of OGA cancer cell populations in ctDNA is feasible during chemotherapy and the observation of genetic evolution warrants investigation in larger series and with higher resolution techniques to reveal potential genetic predictors of response and drivers of chemotherapy resistance. The presence of liver metastasis is a potential biomarker for the selection of patients with high ctDNA content for such studies.