Despite survival gains that have been made with targeted anti-cancer medicines, patients ultimately relapse due to drug resistant disease. It is widely understood that this is due to the presence of drug resistant cells present in the tumour – a concept acknowledged since the 1970s [Nowell 1976]. Acquired genetic lability is required for cancer cells to express a phenotype that can escape both normal, homeostatically controlled, tissue turnover as well as evade immune surveillance. Therefore, it is no surprise that some cancer cells will gain further advantage via drug resistance. The onset of this resistance will dictate the time to progression and ultimately death. It would therefore be beneficial to anticipate the evolution of resistance in treated tumours to inform the optimal treatment regimen, optimal sequence of treatments, combination strategies and to prioritise mutations to target with new medicines,
In this talk current knowledge of resistance kinetics in the clinic based upon observations and model-based analyses will be reviewed. The question of whether drug resistance is innate or acquired on treatment will be discussed as well as evidence for both processes. Using models of resistance kinetics, the relationship between resistant disease, PFS and OS can be demonstrated. These models can also be used to understand the optimal regimen to control resistant disease. However, an ongoing challenge is how to interrogate clinical data that represent the “patient journey” through multiple lines of therapy. This is important so that the influence treatment history has on the duration of response to subsequent treatment options can be understood.
Moving back to animal models of cancer, a review of modelling of xenografted tumour experiments reveals that similar resistance kinetics are observed. This suggests that modelling assumptions of the relative fitness of drug resistant vs sensitive cells can be tested along with modelling the impact of spatially constrained solid tumour growth. An example of using in vitro data for different NSCLC EGFR driven mutants to predict clonal selection in vivo will be used to demonstrate these concepts. Thus, these nonclinical in vitro and in vivo systems, coupled with mathematical modelling, could prove to be useful tools for investigating clonal evolution.
Nowell, P. C. (1976). The clonal evolution of tumour cell populations. Science, 194(4260), 23–8. https://doi.org/10.1126/science.959840