Introduction: Previously, development of novel brain radiotracers has largely relied on simple screening tools. Improved selection methods at the early stages of radiotracer discovery and an increased understanding of the relationships between in vitro physicochemical and in vivo radiotracer properties are needed. We investigated if high performance liquid chromatography (HPLC) methodologies could provide criteria for lead candidate selection by comparing HPLC measurements with radiotracer properties in humans.
Methods: Ten molecules, previously used as radiotracers in humans, were analysed to obtain the following measures: partition coefficient (Log P); permeability (P,); percentage of plasma protein binding (%PPB); and membrane partition coefficient (K(m)). Relationships between brain entry measurements (Log P, P %PPB) and in vivo brain percentage injected dose (%ID); and K specific binding in vivo (BP(ND)) were investigated. Log P values obtained using in silico packages and flask methods were compared with Log P values obtained using HPLC.
Results: The modelled associations with %ID were stronger for %PPB (r(2)=0.65) and P, (r(2)=0.77) than for Log P (r(2)=0.47) while 86% of BPND variance was explained by K. Log P values were variable dependant on the methodology used.
Conclusions: Log P should not be relied upon as a predictor of blood-brain barrier penetration during brain radiotracer discovery. HPLC measurements of permeability, %PPB and membrane interactions may be potentially useful in predicting in vivo performance and hence allow evaluation and ranking of compound libraries for the selection of lead radiotracer candidates at early stages of radiotracer discovery. (C) 2012 Elsevier Inc. All rights reserved.