An in vitro model was developed to evaluate the in vivo stability of lanthanide polyaminocarboxylate complexes. The ligand-to-metal ratios for the chelates EDTA, CDTA, DTPA, MA-DTPA (monoamide-DTPA) and DOTA with the lanthanides lanthanum, samarium, and lutetium were optimized to achieve > or = 98% complexation yield for the resultant radiolanthanide complexes. The exchange of the radiolanthanides from their EDTA, CDTA, DTPA, MA-DTPA and DOTA complexes with Ca(2+) was determined by in vitro adsorption and in vitro column studies using hydroxyapatite (HA), an in vitro bone model. In vitro serum stability of these radiolanthanide complexes was used as an additional indicator of in vivo stability, although the mechanism of instability in serum will be different than with bone. The in vitro studies were consistent with the expected findings that the smallest lanthanide (Lu) formed the most stable complexes. In vivo studies were done to validate the in vitro model. Biodistribution studies in normal CF-1 mice showed that in vivo stability of the complex (i.e., the more lanthanide remaining in complex form) could be assessed by a combination of the urinary, bone and liver uptake. For example, biodistribution studies demonstrate that high urinary excretion correlated with complex stability, while high liver plus bone uptake correlated with complex instability. The urinary excretion of the EDTA complexes decreased from (177)Lu to (140)La indicating a loss in stability in the direction of (140)La, consistent with the in vitro studies. The more stable a lanthanide complex is, the lower its exchange with HA in vitro will be, and the lower its combined bone plus liver uptake and higher its urinary excretion will be in vivo. This investigation indicates that the in vivo stability can be determined by a screening method that measures the degree of exchange from the lanthanide chelate with hydroxyapatite (HA) and its serum stability.