Preclinical assessment of KAL-ABP target engagement and efficacy using pet, MRI and CSF biomarkers

We report preclinical study results of a novel anti-amyloid agent (KAL-ABP). KAL-ABP includes an Aβ oligomer binding protein (ABP) fused to “carrier” module that allows the delivery of the active compound through the blood-brain barrier (KAL-ABP-BBB). KAL-ABP has been reported to selectively bind Aβ1-42 oligomer in vitro and ex vivo and prevent its toxicity. Here, using in vivo longitudinal biomarker acquisitions and McGill-R-Thy1-APP transgenic rat model (Tg), we further replicate and extend these results by quantifying KAL-ABP-BBB cerebrospinal fluid concentrations; target engagement via changes in amyloid load with [18F]NAV4694 positron emission tomography scans (NAV-PET); therapeutic efficacy via changes in hippocampal volumetry (MRI-HV) and vasculature. A total of 17 rats (wild type (WT-ABP), N=4; Tg-ABP, N=9; Tg-SAL, N=4). They had an initial dose of either KAL-ABP-BBB 30mg/Kg/IV followed by 15mg/Kg/IV/week, or placebo on the subsequent 4 weeks. 60-minute NAV-PET dynamic scan was acquired concomitantly with a bolus tail vein injection of the radiotracer after a 9-minute transmission scan. The images were reconstructed using Maximum a posteriori algorithm with scatter, dead time, and decay correction. Non-displaceable binding potential (BPND) was generated using cerebellar grey matter as a reference region based on simplified reference tissue model (SRTM). 60-minute scan via Fast Imaging with Steady-state Precession sequence with TE/TR of 2.5/5.0ms was used to generate structural images. Then, MRI-HV (mm3) was acquired via manual segmentation. Susceptibility MRI were acquired using echo planar acquisitions for vasculature. For image processing and statistical analysis, MINC tools, VoxelStats, R, and Prism7 were used. Linear Mixed effect regression model (corrected by sex, age, weight) and simple t-test were used to estimate the target engagement and therapeutic efficacy, respectively. KAL-ABP-BBB reached CSF concentrations of 300 ng/mL. KAL-ABP-BBB intervention reduced global [18F]NAV4694 BPND(27%; p=0.0442) and rectified baseline genotype observation (Tg vs Wt; p=0.0087). KAL-ABP-BBB intervention increased MRI-HV (8%; p=0.0099). No vascular abnormality was observed during KAL-ABP-BBB intervention. This study provides strong evidence that KAL-ABP-BBB crosses the BBB, safely lowers the amyloid burden and mitigates hippocampal atrophy in vivo. Overall, these results invite further clinical study in humans.