Luvadaxistat Fails to Improve Schizophrenia Cognition in Phase 2 Trial

The Search for Glutamatergic Solutions to Cognitive Impairment

Cognitive deficits represent a core clinical dimension of schizophrenia that strongly predicts a patient's functional trajectory and quality of life [1, 2]. While conventional antipsychotics effectively modulate dopamine pathways to manage psychosis, they offer minimal relief for the neurocognitive impairments that often persist despite symptomatic remission [3, 4]. This therapeutic ceiling has shifted research focus toward N-methyl-D-aspartate receptors (a type of glutamate receptor essential for synaptic plasticity and memory formation), which appear to be underactive in the schizophrenic brain [5, 6]. One targeted approach involves inhibiting D-amino acid oxidase (an enzyme responsible for degrading D-serine, a necessary co-agonist for glutamate receptor activation) to boost synaptic signaling [7, 8, 9]. Recent clinical investigations, such as the ERUDITE phase 2 trial of 203 adults, found that the selective inhibitor luvadaxistat failed to significantly improve Brief Assessment of Cognition in Schizophrenia (a standardized battery of cognitive tests) composite scores at doses of 20 mg (p = 0.75) or 50 mg (p = 0.69) [10]. Conversely, a 2025 meta-analysis of five trials involving 530 participants indicated that D-amino acid oxidase inhibitors as a class may still yield a standardized mean difference of 0.359 (p = 0.017) in cognitive function, suggesting that efficacy may vary significantly by specific compound or patient subgroup [7].

ERUDITE Trial Design and Patient Demographics

The ERUDITE study (NCT05182476) was designed as a phase 2, randomized, double-blind, placebo-controlled, parallel-group trial to evaluate the efficacy of luvadaxistat in addressing cognitive deficits. Luvadaxistat functions as a selective D-amino acid oxidase inhibitor, targeting the enzyme responsible for the metabolic breakdown of D-serine. Because D-serine acts as a critical co-agonist for N-methyl-D-aspartate receptors (a class of glutamate receptors involved in synaptic plasticity), inhibiting this enzyme is intended to increase D-serine availability and enhance glutamatergic signaling in the central nervous system. For practicing psychiatrists, finding an effective adjunctive agent could mean the difference between a patient requiring assisted living versus achieving independent functioning. To test this, the trial specifically enrolled adults with schizophrenia who were already receiving stable background antipsychotic therapy, ensuring that the experimental treatment was assessed as an add-on intervention rather than a monotherapy. To minimize the impact of the placebo effect and establish a stable baseline, all participants first underwent a 2-week placebo run-in period. Following this phase, the researchers randomized participants in a 2:1:1 ratio to one of three study arms: a placebo group, a group receiving luvadaxistat 20 mg once daily, or a group receiving luvadaxistat 50 mg once daily. This randomization resulted in a total of 203 participants, with 101 assigned to placebo, 50 to the 20 mg dose, and 52 to the 50 mg dose. The active treatment period lasted 12 weeks, during which clinicians monitored the participants for changes in cognitive performance and the incidence of treatment-emergent adverse events to determine if the compound could safely provide a therapeutic benefit beyond standard antipsychotic care.

Primary Cognitive Outcomes and Statistical Results

The ERUDITE trial achieved a high rate of protocol adherence, with 178 participants (87.7%) completing the full 12-week treatment period. The researchers established the primary endpoint as the change from baseline to Day 98 in the Brief Assessment of Cognition in Schizophrenia (BACS) composite score, a standardized clinical battery used to measure core cognitive domains such as executive function, verbal memory, and processing speed. Statistical analysis of this primary endpoint revealed that luvadaxistat failed to meet its objective for either dosage group. For participants receiving luvadaxistat 20 mg, there was no significant improvement in BACS composite scores compared to the placebo group at Day 98, demonstrating a least-squares mean difference of -0.7 (95% CI: -2.8, 1.4; p = 0.75). Similarly, the luvadaxistat 50 mg cohort did not show a significant benefit over placebo, yielding a least-squares mean difference of -0.5 (95% CI: -2.7, 1.6; p = 0.69). For clinicians hoping to prescribe a targeted cognitive enhancer, these data indicate that adding this specific D-amino acid oxidase inhibitor to standard antipsychotic regimens did not result in statistically significant changes in cognitive performance or functioning within this patient population.

Safety Profile and Adverse Event Incidence

The researchers monitored safety outcomes by tracking the incidence of treatment-emergent adverse events, defined as any unfavorable medical occurrences that emerge or worsen after the initiation of the study drug. In the ERUDITE trial, treatment-emergent adverse events were reported in approximately one-third of participants across all groups, suggesting a consistent safety profile regardless of the intervention. Specifically, the rates of adverse events were comparable between the luvadaxistat and placebo groups, indicating that the D-amino acid oxidase inhibitor did not significantly increase the burden of side effects compared to the control arm. When assessing the clinical impact of these events, the investigators found that most treatment-emergent adverse events were mild or moderate in severity, typically not requiring discontinuation of the study medication or intensive medical intervention. This characterization of the side effect profile is critical for clinicians managing schizophrenia, as treatment adherence is highly sensitive to the tolerability of adjunctive therapies. Ultimately, no safety concerns were identified during the study, providing evidence that while luvadaxistat did not meet its primary cognitive efficacy endpoints, it was generally well-tolerated at both the 20 mg and 50 mg daily dosages.

References

1. McCutcheon RA, Keefe RS, McGuire P. Cognitive impairment in schizophrenia: aetiology, pathophysiology, and treatment. Molecular Psychiatry. 2023. doi:10.1038/s41380-023-01949-9

2. Veselinović T, Neuner I. Progress and Pitfalls in Developing Agents to Treat Neurocognitive Deficits Associated with Schizophrenia. CNS Drugs. 2022. doi:10.1007/s40263-022-00935-z

3. Meyer JM, Correll CU. Increased Metabolic Potential, Efficacy, and Safety of Emerging Treatments in Schizophrenia. CNS Drugs. 2023. doi:10.1007/s40263-023-01022-7

4. Spark DL, Fornito A, Langmead CJ, Stewart GD. Beyond antipsychotics: a twenty-first century update for preclinical development of schizophrenia therapeutics. Translational Psychiatry. 2022. doi:10.1038/s41398-022-01904-2

5. Stănculete MF, Căpățînă O. Glutamate-Based Therapeutic Strategies for Schizophrenia: Emerging Approaches Beyond Dopamine. International Journal of Molecular Sciences. 2025. doi:10.3390/ijms26094331

6. Tumdam R, Hussein Y, Garin-Shkolnik T, Stern S. NMDA Receptors in Neurodevelopmental Disorders: Pathophysiology and Disease Models. International Journal of Molecular Sciences. 2024. doi:10.3390/ijms252212366

7. Chang C, Hsia Y, Liu W, Lee J, Lin C, Lane H. Symptomatic and cognitive effects of D-amino acid oxidase inhibitors in patients with schizophrenia: a meta-analysis of double-blind randomized controlled trials. Schizophrenia. 2025. doi:10.1038/s41537-025-00604-2

8. Fradley R, Goetghebeur P, Miller DJ, et al. Luvadaxistat: A Novel Potent and Selective d-Amino Acid Oxidase Inhibitor Improves Cognitive and Social Deficits in Rodent Models for Schizophrenia. Neurochemical Research. 2023. doi:10.1007/s11064-023-03956-2

9. Wang J, Guo Y, Han W, Zhang H, Jiang P. D-serine: A Multitalented Neuromodulator in Brain Function, Systemic Homeostasis, and Disease. Cellular and Molecular Neurobiology. 2026. doi:10.1007/s10571-026-01696-9

10. Khin NA, Fan RH, Ge T, et al. A phase 2 randomized controlled trial of luvadaxistat in treatment of adults with cognitive impairment associated with schizophrenia: results from the ERUDITE study.. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology. 2026. doi:10.1038/s41386-026-02410-5