Estimation of the heterogeneity of a tumour with a single sample and its impact on survival
Stage de 8 semaines au Bart's cancer institute de la Queen Mary University de Londres, sous la direction de Trevor Graham et de Pierre Martinez.
Abstract Although progress has been made in the research of treatment, cancer is still the first cause of death in France. It is an evolutionary disease, caused by mutations : a cell acquires a mutation and passes it on to its daughter cells. Then some of these cells can acquire additionnal mutations, making the tumour heterogeneous. The heterogeneity of a tumour could have an impact on prognosis, or on the way the cancer is treated. Moreover, it is very complicated to get more than one sample of a tumour. Therefore, we wanted to find a method to estimate the diversity of a tumour (here the mutational divergence) with only one sample, and see if this estimation correlates with the prognosis. A random sampling approach was used to estimate the mutational divergence. The score obtained was compared to the real value, and the survival probabilities of patients with high and low scores were studied. It appeared that the random sampling method is a good proxy, as it correlates with the tumour diversity. However, it had no impact on survival; So, either the method is not efficient and needs to be improved, or one sample is not enough to get information about a whole tumour.