Although definitive proof of a causative link between aggregation of amyloid peptides (Aβ) and Alzheimer’s disease (AD) remains elusive, much effort has been devoted to devising means of lowering Aβ levels. The ability to assess the effectiveness of such therapies in the clinic has so far been hampered by the lack of a method to determine drug effects on Aβ production or clearance from the human CNS. Aβ is produced by the action of β- and γ-secretases on amyloid precursor protein (APP) and is degraded by a number of enzymes, including insulin-degrading enzyme (IDE) and neprilysin. Inhibitors of both β- and γ-secretase have been developed as potentially disease-modifying treatments, although it is unclear whether increased production, reduced clearance, or a combination of both is responsible for elevated levels of Aβ ?in AD patients. Writing in the Annals of Neurology, researchers at Washington University School of Medicine have now described the use of a recently developed technique known as stable isotope-linked kinetics (SILK) to determine the effect of a γ-secretase inhibitor, 
LY450139 (semagacestat), on production of Aβ. In a double-blind study, 20 healthy volunteers were assigned to receive varying doses of LY450139 or placebo (n = 5 per group). The volunteers also received an intravenous infusion of a labelled form of the amino acid leucine which, over the course of a couple of hours, became incorporated into newly synthesised proteins, including Aβ. By periodically sampling cerebrospinal fluid (CSF), the team were able to monitor the proportion of labelled Aβ to give a measure of the rate of production. CSF sampling was continued after the labelled leucine infusion was switched off to provide a measure of the rate of Aβ clearance. The results suggested that LY450139 caused a dose-dependent decrease in Aβ production, with an 84% reduction at the highest dose (280mg). There were no differences in Aβ clearance between the placebo and drug treated groups. The SILK procedure takes 36 hours but offers the potential for clearer interpretation of drug effects since analysis of measurements of unlabelled Aβ in CSF is confounded by natural fluctuations in levels.
Ongoing clinical studies are looking at the effect of LY450139 on cognitive function and biochemical and brain imaging biomarkers in AD patients.