The accumulation of amyloid beta (Abeta) in Alzheimer's disease is caused by an imbalance of production and clearance, which leads to increased soluble Abeta species and extracellular plaque formation in the brain. Multiple Abeta-lowering therapies are currently in development: an important goal is to characterize the molecular mechanisms of action and effects on physiological processing of Abeta, as well as other amyloid precursor protein (APP) metabolites, in models which approximate human Abeta physiology. To this end, we report the translation of the human in vivo stable-isotope-labeling kinetics (SILK) method to a rhesus monkey cisterna magna ported (CMP) nonhuman primate model, and use the model to test the mechanisms of action of a gamma-secretase inhibitor (GSI). A major concern of inhibiting the enzymes which produce Abeta (beta- and gamma-secretase) is that precursors of Abeta may accumulate and cause a rapid increase in Abeta production when enzyme inhibition discontinues. In this study, the GSI MK-0752 was administered to conscious CMP rhesus monkeys in conjunction with in vivo stable-isotope-labeling, and dose-dependently reduced newly generated CNS Abeta. In contrast to systemic Abeta metabolism, CNS Abeta production was not increased after the GSI was cleared. These results indicate that most of the CNS APP was metabolized to products other than Abeta, including C-terminal truncated forms of Abeta: 1-14, 1-15 and 1-16; this demonstrates an alternative degradation pathway for CNS amyloid precursor protein during gamma-secretase inhibition.