• Cancer
• gastric cancer
• oesophageal cancer
• cell death

Elucidation of the molcular pathogenesis of cancer has provided targets for novel rationally designed targeted anticancer drugs.1 Potential targets are those molecules upon which the growth and maintenance of tumours depends and, in this regard, a key breakthrough was the recognition that tumours in individual patients can become ‘addicted’ to continued signalling from individual molecules. In such cases, developing targeted drugs against these molecules—especially receptor tyrosine kinases (RTK)—has proved very successful.2 After just over a decade of clinical research, the treatment paradigm that has emerged is the use of a predictive biomarker assay which identifies that the growth of a particular patient's tumour is dependent upon activation of a molecule and then the use of an agent against this target in that patient.1 2 This paradigm has recently been successfully applied to gastric cancer with the demonstration of the effectiveness of trastuzumab, a monoclonal antibody against HER2, in HER2-positive tumours.3 Typically, minority subgroups of human cancers (based on anatomical and histopathological classification) express the predictive biomarker for a particular targeted agent and, as these subgroups are identified, a new molecular classification of human cancer is emerging that is intimately linked to available therapeutic agents. Between 7% and 17% of gastric cancers have either high expression of the HER2 protein (assessed by immunohistochemistry, 3+ score) or HER2 gene amplification, which are the clinically useful predictive biomarker assays for HER2 growth dependence and potential benefit from trastuzumab, and this is fairly typical of the proportional size of biomarker-defined subgroups that have emerged.1–3

Accordingly, a key aim of translational cancer research is to develop the investigational methodology that allows identification of addiction to a particular molecule. Gene amplification appears to be a reliable …