We are currently implementing methods that might be predictive of tumor response early in the treatment regimen and comparing their respective value: use of smart contrast agent directed to cell death, diffusion-weighted MRI (cellularity), spectroscopy of choline (membrane turnover), 18F-FLT PET (cell proliferation). A Dynamic Nuclear Polarization (DNP, “Hypersense”) system has been installed in the laboratory and allows the study of metabolic fluxes using 13C-MRS. We are looking to the value of fumarate/malate and pyruvate/lactate as biomarker of response to anti-cancer treatment. Hyperpolarized substrates are also used for the stratification of tumors that may benefit from innovative therapies that modulate the metabolism of cancer cells. This multi-modal strategy significantly contributes to the identification of early non-invasive imaging markers of tumor response to combined targeted therapies in the transition towards individualized cancer therapy, with a special focus on the resistance to first line therapy in advanced breast cancer, and in advanced melanoma, with the ultimate goal of sparing patient’s cycles of futile therapy, and possibly allow them to move to other, possibly experimental therapies. Another illustrative example of our ongoing studies is the use of these imaging biomarkers to evaluate the efficacy of anti-cancer strategies such as dose painting and dose escalation in radiation therapy.