The article on theranostics[1] was lucid, to the point, and informative. As a diagnostic pathologist I am dealing with interpretation of new tests that determine the candidacy of patients for specific targeted therapies. The immuno-histochemical tests (i.e., Herceptest, EGFR PharmDX, and c-Kit) are not much of a problem to perform and interpret. The tissue-based molecular assays, on the other hand, not only require sufficient sample sizes (as you pointed out), but also require special handling to keep them ''molecular-friendly.'' Current tissue fixation and processing methods used in most pathology laboratories destroy RNA and other macromolecules that are essential for pharmacodiagnostics. Let's hope that we could bring the old fashioned (19th century) histomorphology to a higher standard and render patient biopsy material suitable for new and sophisticated molecular testing.