Anti-TfR bispecific antibody(Alector)

In this manuscript, we present a translational physiologically-based pharmacokinetic (PBPK) model to characterize receptor-mediated transcytosis (RMT) of anti-transferrin receptor (TfR) monoclonal antibodies (mAbs) in the central nervous system (CNS). The model accounts for the state-of-the-art knowledge of the brain's anatomy and physiology, and physiological parameters were fixed according to different species. By estimating a few parameters associated with the TfR concentration, the TfR turnover, and the internalization rate, the model simultaneously characterizes plasma, whole brain, interstitial fluid (ISF), and cerebrospinal fluid (CSF) PK of unbound and bound anti-TfR mAbs with different binding affinities in mice, rats, and monkeys obtained from various literature sources within a threefold prediction error. The final PBPK model was validated using external anti-TfR mAb PK data in mice and monkeys with different affinities and doses. The simulation reasonably predicted plasma and brain PK of monovalent/bivalent anti-TfR mAbs within a threefold prediction error and characterized a bell-shaped relationship between the brain ISF/plasma AUC ratio and the K_D value. Although further refinements of the PBPK model and clinical validation are required, this PBPK model may provide physiologically-based translation of CNS disposition of anti-TfR mAbs by accounting for the physiological difference of the endogenous RMT system among different species. The PBPK model may also guide selection of other endogenous receptors, lead optimization, and clinical development of novel CNS-targeted mAbs.

Using a doxorubicin-resistant subline (K562/ADM) of human leukaemia K562 cells (Tsuruo et al, 1986), the effect of immunoliposomes that targeted a cellular transferrin receptor (TFR) was examined by neutralization of doxorubicin (DOX) resistance. OKT9-CIL, prepared by conjugation of DOX-encapsulated liposome with an anti-TFR monoclonal antibody, OKT9 (Aisenberg and Wilkes, 1980), showed similar binding to both K562 and K562/ADM. Although an 80-fold higher sensitivity to free DOX on cell growth inhibition in K562 than in K562/ADM was found, the difference was clearly diminished after OKT9-CIL treatment through the increased sensitivity of K562/ADM. The cellular DOX level 30 min after the exposure of free DOX was 45-fold lower in K562/ADM than in K562, whereas nearly equivalent DOX levels were detected in K562 and K562/ADM after OKT9-CIL treatment. In addition, DOX in K562/ADM in the free DOX treatment was efficiently excreted by 54% within 120 min of incubation, whereas almost all DOX supplied by OKT9-CIL remained uncleared. Fluorescence microscopic observation showed that OKT9-CIL was internalized into juxtanuclear vesicles in K562/ADM cells. These results suggest that OKT9-CIL has a potency to accumulate DOX, resulting in augmentation of DOX cytotoxicity in DOX-resistant tumour cells.