STATE OF THE ART AT AAIC 2018

Johanne Kaplan, PhD, Chief Development Officer

This year’s Alzheimer’s Association International Conference (AAIC) in Chicago brought together over 5,900 attendees from 68 countries around the globe for a total of 2,875 presentations on dementia science.  Two major themes generated a lot of excitement: 1) the potential to impact the course of Alzheimer’s disease (AD) by targeting toxic amyloid-beta oligomers (AβO), and 2) the emerging wealth of data on the use of biomarkers to assess disease status and track the impact of treatment.

The causative role of soluble AβO in the pathogenesis of AD and the importance of neutralizing AβO toxicity represented a common theme across multiple oral presentations and posters.  In particular, a presentation of the recent clinical results obtained with BAN2401 by Eisai/Biogen was eagerly awaited.  BAN2401 is a humanized IgG1 monoclonal antibody reported to have selectivity for soluble aggregates of Aβ (also referred to as protofibrils) without binding to monomers.  Among clinical stage antibodies, this represents to date the best binding profile toward the goal of more selective neutralization of AβO.  In front of a packed audience, Dr. Lynn Kramer, Eisai’s Chief Medical Officer for neurology, described the topline results of an 18-month global Phase 2 study involving 856 subjects.  BAN2401 produced a dose-dependent reduction in clinical decline compared to placebo as measured by ADCOMS (a novel composite of 12 elements from 3 different cognitive assessment scales), with the top dose of 10 mg/kg bi-weekly significantly reducing decline by 35 % (p=0.027) and 30% (p=0.034) at 12 and 18 months, respectively.  The antibody also produced a statistically significant reduction in PET tracer uptake for brain amyloid (plaque) across all doses tested versus placebo. Treatment was overall well-tolerated with infusion-related reactions and amyloid related imaging abnormalities (ARIA) as the most common adverse events. The incidence of brain edema (ARIA-E) was around 10% at the highest dose of 10 mg/kg monthly or bi-weekly.  These results are very encouraging although the use of a new scoring system (ADCOMS) and, more importantly, the fact that ARIA-prone APOE4 carriers were excluded from the top 10 mg/kg bi-weekly dose group after the start of the study (at the request of European regulators), are being viewed as potentially confounding factors.

Overall, the BAN2401 results provide strong support for targeting of AβO as a therapeutic strategy for AD while at the same time leaving room for improvement.  The occurrence of ARIA-E at the top dose (which would presumably be even higher if APOE4 patients had not been excluded partway through the trial), along with the observed plaque clearance, suggest that BAN2401 likely still interacts with plaque fibrils, taking antibody away from toxic AβO and causing dose-limiting toxicity.  As outlined in ProMIS’ poster presentation entitled Humanized PMN310 shows enhanced therapeutic potential by binding toxic low molecular weight Aβ oligomers while avoiding ARIA-related binding to Aβ deposits in AD patient brains”, ProMIS Neurosciences lead clinical candidate circumvents these issues.  Humanized PMN310 (huPMN310) shows unrivaled selectivity for native toxic AβO derived from human AD brains, with no binding to Aβ monomers or plaque.  In addition, the choice of a non-inflammatory IgG4 isotype for humanization provides another layer of safety which altogether should allow for the administration of higher and more therapeutically effective doses of antibody.  

Another topic of high interest centered around advancements in the use of biomarkers fueled by the convergence of emerging results from large longitudinal studies (e.g. DANI, DIAN, Swedish BioFINDER) together with recent FDA guidance encouraging the use of cerebrospinal fluid (CSF) and imaging biomarkers in clinical testing.  The usefulness of the core CSF biomarkers Aβ, t-Tau and p-Tau to assess and track the status of AD patients is still in evidence along with more novel biomarkers.  In particular, CSF levels of neurofilament light chain (Nfl), neurogranin/SNAP25 and YLK-40 were reported to be among the most promising markers of neurodegeneration, synaptic damage and neuroinflammation, respectively.  In various studies, these markers were found to be elevated well before the occurrence of symptoms and were predictors of cognitive decline. As the base of knowledge continues to expand, the inclusion of CSF biomarkers in clinical trials will grow in importance as a tool for patient selection and assessment of therapeutic efficacy.  

ProMIS’ clinical candidate, huPMN310, is on track to enter Phase 1 clinical testing in the second half of 2019 and will include biomarker assessment as an integral part of the study design.