PD-0017 - Druggable genetic dependencies for molecularly defined subgroups of oesophageal cancer identified from high-throughput functional profiling

Date 27 June 2014
Event World GI 2014
Session Poster discussion session III – Endoscopy
Topics Oesophageal Cancer
Translational Research
Presenter Irene Chong
Citation Annals of Oncology (2014) 25 (suppl_2): ii5-ii13. 10.1093/annonc/mdu164
Authors I. Chong1, J. Campbell2, I. Bajrami2, R. Brough2, J. Frankum2, C. Lord2, A. Ashworth2
  • 1The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, London/UK
  • 2The Institute of Cancer Research, London/UK



Recent DNA sequencing and copy number analysis of oesophageal cancer has delineated the first draft of the mutational landscape of this disease. The development of large scale RNA interference (RNAi) interventions, resulting in selective silencing of specific proteins, has the potential to pinpoint druggable candidate genetic dependencies relevant to defined molecular subtypes of oesophageal cancer. This study aims to identify candidate genetic dependencies and therapeutic targets for oesophageal cancers with specific genetic defects by high-throughput functional viability profiling of oesophageal cancer cell line models.


To define molecular subgroups of oesophageal cancer, a list of genes currently known to be recurrently mutated, amplified or deleted, were annotated using data from recently published genome-wide molecular profiling studies. The presence or absence of these genetic aberrations were evaluated in 17 commercially available oesophageal cell line models by annotating the genotype from publically available exome sequencing and SNP data (Cancer Cell Line Encyclopedia and COSMIC cell line project) as well as from array comparative hybridization data generated in-house for each cell line. To maximize the opportunity for defining therapeutic targets for oesophageal cancers with specific genetic defects, the 17 oesophageal cell lines were transfected with a 384 well-plate arrayed siRNA library targeting 714 kinases and kinase-related genes. Plate-centred viability data from each screen was standardized using a Z-score statistic.


As proof of concept, we used supervised hierarchical clustering of the viability data to identify gene dependencies in ERBB2 amplified cell lines. Targeting of ERBB2 itself was selectively lethal in ERBB2 amplified cell lines (P = 0.037), supporting the hypothesis that ERBB2 amplification can lead to an oncogene addiction effect. Furthermore, ERBB2 amplified cell lines were also sensitive to silencing of other members of the MAPK family which signal downstream from ERBB2. Silencing of VEGFR-2 in CCND1 amplified cell lines resulted in the most significant effect on cell viability (P < 0.0001). Monoclonal antibodies targeting VEGFR-2, such as ramucirumab, have been evaluated in unselected patient populations with advanced oesophagogastric cancer, and it is likely that evaluation of CCND1 amplification as a candidate biomarker within these clinical studies would be worthwhile. Silencing of the Bruton agammaglobulinaemia tyrosine kinase gene (BTK) resulted in selective lethality in MYC amplified oesophageal cell lines (P = 0.008). The potential of BTK small molecule inhibition as a novel therapeutic strategy in MYC amplified oesophageal cancer warrants further evaluation, as the use of ibrutinib, which selectively targets BTK, has been shown to result in durable single-agent efficacy in other cancers such as chemo-refractory mantle-cell lymphoma and chronic lymphocytic leukaemia.


The analysis of functional viability profiles of oesophageal cell lines generated from high-throughput siRNA screening has identified candidate druggable genetic dependencies relevant for particular subgroups of oesophageal cancer. Further in vivo validation and clinical correlation are required to consolidate these results.