Conceptualisation of diagnostic agents: from empirical in vivo screening to rational in vitro predictive parameters

Abstract

A new rational conceptualisation protocol in new isotopic diagnostic agents has been designed to avoid systematic empirical in vivo screening. This protocol is based on multiple regression analysis, in order to determine the pharmacokinetic model capable of explaining in the best possible way the in vivo behaviour of molecules injected into an organism. Nine technetium complexes (99mTc-L) were synthesised from aminothiol ligand vectors. These complexes were characterised in terms of their physical, chemical and biological in vitro properties, i.e. lipophilicity (P), free fraction unbound to plasmatic proteins (Fup), the fraction unbound to blood cells (FuCb), and membrane adsorption fraction (Fad), an original factor assessed in vitro. Thus, two pharmacokinetic models were tested. The first takes into account the parameters classically used in pharmacokinetics (P, Fup, FuCb), and the second, in parallel with the first, includes the membrane adsorption rate (polar/apolar/polar membrane model). According to the phenomenon of tissular distribution, the explanatory power of the second model, including the Fad, is radically greater than that of the classical model. The comparative adjusted coefficient of determination (R2) of the model, including the Fad versus R2 of the first model, are heart (91%/6%), liver (89%/47%), spleen (64%/44%), lung (78%/26%), kidney (70%/33%) and brain (73%/8%), respectively. In addition, as for the myocardium, the membrane adsorption factor seems to be the only predictive factor of the affinity between the myocardium and neutral or cationic molecules. This refutes the generally held view that only cationic molecules could have an affinity with the heart.