Andrew's research project brings together several cutting-edge technologies: combining single cell sequencing and CRISPR gene editing with neural networks to establish a computational model that predicts therapeutic response to combination drug therapies. Each year there are about 750 cases of Acute Lymphoblastic Leukaemia (ALL), predominantly in those of age 0-4, where cure rates are high. However, in older patients and children who relapse, survival is dismal, leading to around 240 deaths annually. Glucocorticoid Receptor (GR) activation by dexamethasone (Dex) has been one of the mainstay therapies for ALL for over three decades. The caveat is drug resistance: 60% of adult patients are unresponsive to therapy at relapse and is therefore a critical problem for survival.
Combinatorial therapies have the potential to bypass resistance. Screening human cancer cells for drug response has been successful in predicting single therapeutic options from a patient's genotype but does not resolve the challenge of identifying particularly efficacious combinations of drugs. Andrew proposes an alternative strategy: first, establishing how cancer cells respond to combination therapies by experimentally targeting combinations of genes in a high-throughput manner; and secondly, integrating the results through machine learning. By predicting therapies that bypass resistance, Andrew's work will play a key role in improving the current five-year survival rate in adults which, at only 40%, is considerably short of public targets, including CRUK's own 75% 10-year survival goal.