Investigating the relevance of saturable binding in radioimmunotherapy using pharmacokinetic modeling

1152 Objectives In radioimmunotherapy, the therapeutic activity is often calculated based on pretherapeutic measurements to achieve a suitable radiation absorbed dose. Equivalence of therapeutic and pretherapeutic biodistribution is assumed. However, in case saturation of antibody binding sites plays an important role, this assumption might not be justified. Different amounts of administered antibody and residual antibody may lead to a substantially different therapeutic biodistribution. In this study, this conjectured effect for radioimmunotherapy with 90Y-labelled anti-CD66 antibody is investigated. Methods Time activity serum data of 10 patients with acute leukaemia were fitted to a set of 8 two-compartment models, which differed with respect to linear or non-linear binding and global or local fitting of the model parameters. The best model was chosen according to the corrected Akaike information criterion (AICc). To validate the model prediction power, simulations of therapeutic serum concentration were compared to additional therapeutic measurements. Results Two non-linear models were most supported by the data (sum of probabilities for non-linear models >99.9%). The empirically most supported model accurately predicted (18.3±2.8) h the measured therapeutic residence times (18.4±3.8) h. Calculated residence times of free and bound antibody during therapy differed by a factor of (2.1±0.2) and (0.54±0.09) compared to dosimetry. Conclusions The assumption of equivalent biodistribution between dosimetry and therapy using 90Y-labeled anti-CD66 antibody is not justified. The presented method can also be employed to validate other systems where saturation of binding sites might alter the treatment effect due to a different therapeutic biodistribution. Research Support The authors gratefully acknowledge a grant by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation, grant no. GL 236/7-2)