All anticancer molecularly targeted agents on the market today have been approved with one or no companion diagnostic based on a specific genomic molecular alteration. These drugs have followed the same clinical development than chemotherapeutic agents and have been developed in selected tumor types and histologies. Now, some molecular alterations have been described across different tumor types, although with variable prevalence and functional impact. The latter raises the question of whether treatment decision should be mainly based on molecular biology, independently of tumor location and histology. This approach refers to what is commonly named personalized medicine and can today be addressed in clinical trials, since major advances in high throughput technologies allow depicting most druggable molecular alterations for an affordable cost in a timeframe that is compatible with clinical practice. Several studies have been initiated that aim at personalizing medicine in oncology. They include molecular screening programs, as well as personalized medicine trials that can be divided in two categories: (I) stratified clinical trials according to either molecular alterations or tumor types; and (II) algorithm-testing trials evaluating a treatment algorithm instead of drugs efficacy. Multiple challenges are associated with personalized medicine trials, but the main one remains our ability to predict drug efficacy based on molecular alterations. It is expected that taking into account several molecular alterations for the prediction of drug efficacy using systems biology approaches will improve patients' outcome. Bioinformatics research will be an important factor of future progression in this emerging field.
Keywords: Algorithms; biomarkers; clinical trials; high throughput technologies; personalized medicine.