P1‐106: A HEAD‐TO‐HEAD COMPARISON OF LEAD CLINICAL ANTI‐Aβ ANTIBODIES

Le résumé fourni par la source

Approaches using monoclonal antibodies (mAbs) to target the amyloid β-protein (Aβ) constitute the largest and most advanced therapeutic effort to treat Alzheimer's disease (AD). Despite generally good outcomes in preclinical mouse models, anti-Aβ immunotherapy has yielded limited success in humans. While many reasons could explain poor translation of pre-clinical leads into viable therapies, the most pressing issue remains the lack of knowledge about whether trial antibodies engage the toxic forms of Aβ in human brain. Here, we report the development of an unbiased, medium-to-high throughput in vitro assay that combines the use of Aβ-rich human (AD) brain extracts and human iPSC-derived neurons to assess the relative ability of 5 clinical lead mAbs to engage with toxic Aβ. Aqueous extracts of temporal cortices from mild AD patients were prepared as described (Wang et al. 2017 J Neurosci) and a portion immunodepleted of Aβ using the novel pan anti-Aβ antiserum, S97. Human iPSCs were infected with appropriate vectors to drive neurogenin 2 differentiation and grown in 96-well plates at 5000 cells per well (Hong et al. 2018 Acta Neuropath). On day 21 post-induction, AD brain extracts (+/- immunodepletion) were added to neurons +/- test mAbs. Neurons were imaged every 2 hours for 4 days using an IncuCyte live-cell imaging system, and neurite number and density quantified. Five clinical mAbs, an in-house aggregate-specific mAb, 1C22, and a non-Aβ antibody (Avastin) were compared blind to mAb identity. Soluble AD brain extracts caused a time-, dose, and Aβ-dependent decrease in neurite length and branch points (but not cell bodies), whereas extracts immunodepleted of Aβ had no effect. When tested at a single concentration (1.5 mg/ml), comparator mAbs provided varying degrees of protection with 1C22 ∼SAR228810 >aducanumab ∼bapinezumab >gantenerumab >BAN2401. Control mAb (Avastin) afforded no protection. Dose curves of the best 3 mAbs confirmed this order of relative efficacy. While no single assay can predict the absolute utility of an anti-Aβ antibody when administered to humans, the novel quantitative paradigm described here, if widely adopted, would enable important objective preclinical comparisons of new anti-Aβ antibodies and current lead antibodies in human trials.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.