Predicting response to radiotherapy: evolutions and revolutions.

PURPOSE: To review the many changes which have occurred in the past decades in the field of predicting outcome after radiotherapy from biological characteristics of the tumour or normal tissue. This review will also describe the present state of the art and emerging trends for the future.
CONCLUSIONS: From using explanted cells, glass electrodes, exogenous proliferation and hypoxia tracers, and others, there has been a move towards monitoring expression and mutation of genes. Initially this was possible for just one or a few genes, but methods are now available which allow genome-wide monitoring at either the DNA or RNA level. The potential advantage of this evolution is not only to predict but also to understand potential causes of failure, allowing more rational and effective interventions. Comparative genomic hybridisation, mRNA expression profiling, microRNA profiling and promoter methylation profiling have all shown promise in finding signatures correlating with outcome, including after treatment involving radiotherapy. Expected trends for the future are: more signatures relevant to radiotherapy will be discovered; signatures will be refined and reduced to their essentials, such that genome-wide screening will give way to tailored signatures, quantifiable by routine non-array technology; more focus will be on assays predicting which pathway-specific radiosensitising drugs will be effective (exploiting tumour weaknesses); more signatures will be subjected to validation in randomised trials; and proteomics, DNA sequencing and imaging methods will play progressively increasing roles.