I have pinched this title from a slide shown by Charlie Swanton at the First London Cancer Breast Study Day earlier this week! He was one of about a dozen speakers and his session was on a personalised approach to breast cancer treatment. Back to last year's theme of stratified medicine.
He was making the point that predicting drug response and the whole issue of multi-drug resistance are some of the challenges of the personalised approach. The genomics revolution makes it possible to make treatment or trial eligibility decisions fairly quickly based on sequencing from a single biopsy sample. But there is a question as to whether this is the best approach.
Looking at mutations found in samples from primary and secondary tumours shows that while some are ubiquitous, many are not; in fact, the majority are not. Intra-tumour heterogeneity means that drug response can't always be predicted from the sequencing of a single sample.
All this, he suggested, means that the traditional model of linear evolution of cancer is not helpful. Instead, tumours and their evolution are more like trees with branches going off in different directions and then themselves branching. Some tumour-trees have long trunks, while in others the branching starts much lower down. This goes some way to explaining why sometimes we don't see the expected response to a particular drug.
Another reason why stratified medicine is so important.