Predictive markers for normal tissue reactions: fantasy or reality?

Interpatient heterogeneity in normal tissue reactions varies considerably, yet the genetic determinants and the molecular mechanisms of therapeutic radiation sensitivity remain poorly understood. Predictive assays and markers for normal tissue reactions are still in their infancy, although some progress has been made, particularly, for predicting late toxicity. For instance the T-lymphocyte radiation-induced apoptosis assay was shown to significantly predict differences in late toxicity between individuals and an 18 gene classifier based on radiation-induced expression in subcutaneous fibroblasts has also been identified that differentiated between patients with a high and low risk of radiation-induced fibrosis. However, the technical set-up for gene expression measurements means that this latter assay is unlikely to be introduced soon into a routine clinical setting but has importantly allowed the identification of genes that are involved in the fibrotic process. Serum markers have also been identified that show potential for the prediction of patients who will develop acute and late pulmonary toxicity. Few genetic predictive markers for normal tissue reaction have been identified and validated. Many of the single nucleotide polymorphism association studies have been limited by size and the inclusion of subjects with different kinds of radiation morbidity. International collaboration to assemble well-defined cohorts and technological progress should mean that the identification and validation of such markers using candidate gene approaches and whole genome association studies, which have been successful in other research areas, will make rapid progress.