|| Checking for direct PDF access through Ovid
Low productivity in the pharmaceutical industry positions personalized medicine and companion diagnostics as an attractive approach. The oncology community is enthusiastic and optimistic that advances in molecular diagnostics have the potential to materialize the dream of personalized cancer therapy with the aim to: help the drug industry enhance the probability of success and, probably, accelerate time to market; help the diagnostics industry develop diagnostic tests utilizing precious human samples; and support accurate diagnosis and selection of the most efficacious and least toxic therapies. However, this spectacular road is not yet paved and remains facing the following biggest challenges: the need for better understanding of biology and limitations of preclinical models; pre-analytical variables that are likely to alter results; analytical discrepancies; and the wide gap between clinical/pharmaceutical and diagnostic understandings. This article explores these challenges, their possible impacts, and provides some suggestions for mitigation.