SAR228810: an antibody for protofibrillar amyloid ß peptide designed to reduce the risk of amyloid-related imaging abnormalities (ARIA).

BACKGROUND: Anti-amyloid ß (Aß) immunotherapy represents a major area of drug development for Alzheimer's disease (AD). However, Aß peptide adopts multiple conformations and the pathological forms to be specifically targeted have not been identified. Aß immunotherapy-related vasogenic edema has also been severely dose limiting for antibodies with effector functions binding vascular amyloid such as bapineuzumab. These two factors might have contributed to the limited efficacy demonstrated so far in clinical studies. METHODS: To address these limitations, we have engineered SAR228810, a humanized monoclonal antibody (mAb) with limited Fc effector functions that binds specifically to soluble protofibrillar and fibrillar forms of Aß peptide and we tested it together with its murine precursor SAR255952 in vitro and in vivo. RESULTS: Unlike gantenerumab and BAN2401, SAR228810 and SAR255952 do not bind to Aß monomers, low molecular weight Aß oligomers or, in human brain sections, to Aß diffuse deposits which are not specific of AD pathology. Both antibodies prevent Aß42 oligomer neurotoxicity in primary neuronal cultures. In vivo, SAR255952, a mouse aglycosylated IgG1, dose-dependently prevented brain amyloid plaque formation and plaque-related inflammation with a minimal active dose of 3 mg/kg/week by the intraperitoneal route. No increase in plasma Aß levels was observed with SAR255952 treatment, in line with its lack of affinity for monomeric Aß. The effects of SAR255952 translated into synaptic functional improvement in ex-vivo hippocampal slices. Brain penetration and decoration of cerebral amyloid plaques was documented in live animals and postmortem. SAR255952 (up to 50 mg/kg/week intravenously) did not increase brain microhemorrhages and/or microscopic changes in meningeal and cerebral arteries in old APPSL mice while 3D6, the murine version of bapineuzumab, did. In immunotolerized mice, the clinical candidate SAR228810 demonstrated the same level of efficacy as the murine SAR255952. CONCLUSION: Based on the improved efficacy/safety profile in non-clinical models of SAR228810, a first-in-man single and multiple dose administration clinical study has been initiated in AD patients.

NF-L in cerebrospinal fluid and serum is a biomarker of neuronal damage in an inducible mouse model of neurodegeneration.

Accumulation of neurofilaments (NFs), the major constituents of the neuronal cytoskeleton, is a distinctive feature of neurological diseases and several studies have shown that soluble NFs can be detected in the cerebrospinal fluid (CSF) of patients with neurological diseases, such as multiple sclerosis and frontotemporal dementia. Here we have used an inducible transgenic mouse model of neurodegeneration, CamKII-TetOp25 mice, to evaluate whether NF-L levels in CSF or blood can be used as a biochemical biomarker of neurodegeneration. Induction of p25 transgene brain expression led to increase in CSF and serum NF-L levels that correlated with ongoing neurodegeneration. Switching off p25 prevented further increases in both CSF and serum NF-L levels and concomitantly stopped the progression of neurodegeneration. The levels of CSF NF-L detected in p25 mice are about 4-fold higher than the CSF levels detected in patients with chronic neurodegenerative diseases, such as symptomatic FTD (bvFTD). In addition, our data indicate that the NF-L detected in CSF is most likely a cleaved form of NF-L. These results suggest that CSF and serum NF-L are of interest to be further explored as potential translational dynamic biomarkers of neurodegeneration or as pharmacodynamics biomarkers at least in preclinical animal studies.