DAOIB plus AO

Chang Gung Memorial Hospital in Taiwan is running a Phase 2 trial (NCT06467539) testing DAOIB - a D-amino acid oxidase inhibitor, almost certainly sodium benzoate based on the Lane group's prior decade of work - combined with an unnamed antioxidant in 80 patients with early dementia [1]. The bet: boost NMDA glutamate signaling and quench oxidative stress simultaneously, both of which go sideways early in Alzheimer's disease. This is academic work from the Hsien-Yuan Lane group, which has run prior placebo-controlled trials of sodium benzoate showing ADAS-Cog improvement in early-phase AD [2] and in behavioral symptoms of dementia [3]. No pharma sponsor, no FDA designations, and the active comparator is DAOIB plus placebo - meaning a positive readout tells you whether adding the antioxidant helps, not whether DAOIB itself works versus untreated baseline. The reason to track this trial is as a read-across for Takeda's luvadaxistat, the only drug-like DAAO inhibitor in commercial development [4]. Watch as mechanism validation, not as an investable event in itself.

Phase 2, recruiting, n=80, single academic sponsor (Chang Gung Memorial Hospital) [1]. No FDA breakthrough, fast track, or orphan designations - this is a Taiwan investigator-initiated trial, not on the FDA's radar. Sodium benzoate, the most likely identity behind 'DAOIB', is already GRAS (Generally Recognized As Safe, the FDA designation for food additives with established human safety) as a food preservative, which makes regulatory friction low but commercial appeal essentially zero - there is nothing to patent on the active molecule. Primary endpoint is change from baseline on the ADAS-Cog (the standard Alzheimer cognitive scale) at weeks 8, 16, and 24 [1]. Readout timing is uncertain; with 80-patient enrollment from a single Taiwan site and a 24-week treatment period, a 2027 top-line is plausible but the sponsor has issued no public completion guidance. We log this as '2027' rather than a specific quarter because there is no basis for quarter-level precision. The combination intervention has no associated company filings; ABVC Biopharma and related public companies that have historically licensed Chang Gung neurology assets show no disclosed connection to this specific protocol in SEC EDGAR filings reviewed through Q1 2026 [5].

Glutamate is the brain's main 'go' signal. NMDA receptors - one of the gates that glutamate flips open - need a second key to fire: a co-agonist called D-serine. The enzyme D-amino acid oxidase (DAAO) chews up D-serine. Block DAAO and you get more D-serine, stronger NMDA signaling, and theoretically better synaptic plasticity, which is the cellular machinery behind learning and memory [6]. Sodium benzoate is the cheap, oral DAAO inhibitor that crosses the blood-brain barrier well enough to matter. The antioxidant addition targets a separate problem: Alzheimer's brains accumulate oxidative damage early, and that damage selectively kills the same glutamatergic neurons NMDA enhancement is trying to wake up. The protocol does not publicly disclose the antioxidant's identity. Lane group prior work has paired DAAO inhibition with N-acetylcysteine (NAC) and with other off-the-shelf antioxidants; NAC is the most plausible candidate because it also modulates the cysteine-glutamate antiporter and so has a glutamatergic-side rationale, but this is inference, not confirmation. The combination is biologically coherent - boost the signal, protect the wire. The catch: NMDA-enhancement has a graveyard. Memantine, the only NMDA-active AD drug approved, is an antagonist, not an agonist, used in moderate-to-severe disease. Multiple agonist-side approaches have failed in mild AD over the past 20 years. The Lane group's prior placebo-controlled Phase 2 in early-phase AD (n=60, sodium benzoate 250-750 mg/day, 24 weeks) reported ADAS-Cog improvement versus placebo [2], but the field broadly has not accepted DAAO inhibition as a validated AD mechanism on the strength of a single small trial.

Single-site randomized trial at Chang Gung Memorial Hospital, n=80, with DAOIB plus antioxidant versus DAOIB plus placebo [1]. The control choice is the biggest design flaw: both arms get DAOIB, so a positive result only proves the antioxidant adds something on top - it does not isolate whether DAOIB does anything against placebo alone. For a target with this much skepticism, that is a meaningful gap. Primary endpoint is ADAS-Cog change at 8, 16, and 24 weeks. Twenty-four weeks is short for an Alzheimer's cognitive readout - the amyloid-clearing antibodies needed 18 months to show CDR-SB separation in CLARITY-AD and TRAILBLAZER-ALZ2, and those drugs have vastly larger biological footprints than a small-molecule co-agonist enhancer. Patient population is described as 'early dementia,' which can span MCI and mild AD. The public ClinicalTrials.gov record does not specify amyloid PET or CSF biomarker stratification for inclusion; if the protocol does not biomarker-stratify for amyloid status, the response signal could be diluted by mechanistically heterogeneous patients. Enrollment of 80 across a single site is doable but slow. No interim analysis is publicly disclosed.

Our model gives this drug a 4% chance of eventual approval. That starts from the historical approval rate for Phase 2 drugs in this area-about 24%-then adjusts based on ten facts about the trial and its sponsor. The estimate is pulled down mainly by heavier-than-usual blinding, the sponsor's thin or weak approval record, weak earlier-phase results, and the use of a randomized design. The remaining facts are close to average for this stage and don't move the number much either way.

Efficacy risk is the dominant concern. ADAS-Cog at 24 weeks in 80 patients has limited power to detect modest cognitive effects, and the 'early dementia' label without specified amyloid biomarker stratification risks pooling responders and non-responders. The active-control design means even a clean win only tells you the antioxidant matters - it does not validate DAOIB. Safety risk is low; sodium benzoate has a long human exposure history at food-preservative doses, though gram-scale therapeutic doses can cause GI complaints and theoretical hyperammonemia in vulnerable patients. If the antioxidant is NAC, its safety profile is also well established with mild GI as the main signal. Execution risk is moderate: single-site academic trials in Taiwan have completed before, but enrollment pace is not public and there is no interim disclosure mechanism. Commercial risk is the killer. Sodium benzoate is unpatentable. The antioxidant is likely an off-the-shelf compound. Even a clean positive readout has no obvious path to FDA filing without a sponsor willing to fund a 1,000+ patient registrational trial against the lecanemab/donanemab-era standard of care, which now includes amyloid-clearing antibodies in eligible early AD patients.

Watch, don't trade. This is mechanism-validation work, not a commercial event. The single most important reason to track NCT06467539 is as a read-across for Takeda's luvadaxistat (TAK-831), the only drug-like DAAO inhibitor in active commercial development [4]. Luvadaxistat ran a Phase 2 (INTERACT) in cognitive impairment associated with schizophrenia that missed its primary symptom endpoint but signaled on cognitive measures; Takeda has continued development in CIAS rather than AD. A positive DAOIB+AO readout would meaningfully update the prior on DAAO inhibition as a viable cognitive-symptom mechanism, which spills into how investors should value the luvadaxistat program. A negative or noisy readout adds to the pile of failed NMDA-enhancement attempts in AD and makes any DAAO-for-cognition pitch harder to sell. The trial itself has no public stock attached - it will not move a name on the day of readout. Check back when results post, most likely 2027 based on enrollment timing and 24-week follow-up. If the trial reports a clean separation, that is the signal worth flagging. If it reports a wash, the DAAO-for-AD thesis takes another hit and the field can stop revisiting it.