Last year, we shared Phase 2a results from our Alzheimer’s program at CTAD in Madrid. With the full paper now published, we unpack what we tested, what we learned, and how it helps advance our search for better treatments.

12/9/25, 00:00

Molecular glue

Alzheimer’s disrupts the delicate calcium balance inside neurons, which undermines synapses and eventually accelerates hallmark pathology like amyloid-beta plaques and tau tangles.

Septin filaments help organize cell membranes and regulate calcium entry. In Alzheimer’s, their function is impaired, contributing to neurotoxic calcium levels. Our small-molecule approach acts as a molecular glue to restore septin filament integrity and normalize calcium handling, an idea that showed neuroprotection and cognitive rescue in preclinical models (previously published in Science).

Phase 2a tests

We ran a 28-day, randomized, double-blind, placebo-controlled Phase 2a in people with mild-to-moderate Alzheimer’s disease. Primary aims were safety and tolerability. Exploratory readouts covered pharmacokinetics (including CSF exposure), CSF biomarkers (tau, synaptic and inflammatory markers), quantitative EEG, and computerized plus standard cognitive tests. Participants received placebo or ascending oral doses of our first-generation compund REM127.

We observed:

• On-target CNS exposure: CSF concentrations exceeded the in-vitro septin affinity at all active dose levels, consistent with saturated target engagement.

• Core pathology: CSF p-tau181 and total tau decreased over 28 days, aligning with the hypothesis that restoring septin function can mitigate processes upstream of tau dysregulation.

• Circuits and synapses: EEG alpha-1 power strengthened by week two and week four, CSF dopamine increased versus placebo, and select synaptic markers shifted relative to baseline.

• Cognition: Participants on active therapy responded faster on a visual memory task and improved on orientation (ADAS-Cog subtest). A pre-specified multivariate readout combining delayed picture recognition + alpha-1 EEG + CSF p-tau181 strongly favored active treatment.

Read together, these signals—pathology, electrophysiology, neurochemistry, cognition—tell a coherent, mechanism-anchored story over just 28 days.

Halted early

Despite of these encouraging pharmacodynamics, the trial was halted prematurely because of dose-dependent liver enzyme elevations (ALT/AST), which did normalize after stopping treatment. The take-home for us was straightforward: the pharmacology looks promising, the therapeutic window of this first molecule did not.

Follow-up work helped us to understand which properties of the molecule lead to its accumulation and dose-limiting tolerability. A backup molecule without these liabilities was ready for further development. Our strategic pivot to a second-generation candidate with safer PK and hepatic profile, while preserving septin-restoring activity, puts us back on the road towards an effective, oral Alzheimer’s disease drug.

Second generation in the clinic

As announced earlier this year, we initiated a placebo-controlled Phase 1 study of REM392, our second-generation septin modulator, in healthy volunteers. The study evaluates safety, tolerability, and pharmacokinetics and is designed to lay the groundwork for a Phase 2 trial in patients.

Like its predecessor, REM392 targets septin filament function to help restore neuronal calcium balance, but with an optimized safety profile and PK.

Results in context

The convergent changes we observed—reductions in CSF p-tau, increases in EEG power, higher CSF dopamine, and improvements on memory measures—are mechanistically coherent with our mode of action. They are consistent with pharmacological restoration of septin filaments leading to normalization of neuronal calcium handling and, in turn, improved circuit function.

At the same time, this was a small, safety-oriented study that was terminated early and not powered for definitive efficacy conclusions. Over a four-week window, it is neither surprising nor concerning that not every biomarker shifted; these findings should be interpreted as exploratory signals that warrant confirmation in larger and longer trials.

Crucially, the study yielded decision-grade insight. It delineated the contexts in which the biological mechanism appears to be engaged and clarified priorities for medicinal-chemistry optimization, providing a productive foundation for advancing the second-generation candidate and designing subsequent clinical trials.

Read the paper