Antipsychotic medication (AP) is primarily effective in treating positive psychotic symptoms such as delusions, hallucinations, and disorganized thinking, which are cardinal symptoms of schizophrenia. However, AP does not ameliorate the accompanying negative symptoms (e.g., anhedonia, avolition, and social withdrawal), and cognitive deficits, which are highly disabling and predictive of patients’ long-term prognosis (1–3). Persistent negative symptoms have also shown to be a critical predictor for future treatment resistance (4). A growing body of evidence support that a shortening of the interval between onset of psychosis and initiation of an efficient intervention lead to better long-term outcome (5–7).

The pathophysiology of psychosis is complex and involves multiple neurotransmitters, e.g., serotoninergic- (8), GABAergic, and glutamatergic (9) systems. Nevertheless, modulation (particularly antagonism) of striatal dopamine D2 receptors (D2R) remain the key common denominator of all licensed antipsychotics (10). Even so, around 30% of patients with psychosis display inadequate response to antidopaminergic treatment (11, 12). Mounting evidence from post- mortem-, brain imaging-, genetic-, and pharmacologic challenge studies indicate glutamatergic dysregulation, specifically of the N-methyl-D-aspartate receptor (NMDAR), as part of the pathophysiology of psychosis and schizophrenia (13–15). Through excitotoxicity, glutamate dysregulation may underlie structural degeneration, e.g., in hippocampus, which is a key finding in patients with schizophrenia (16–19).

Using 3T magnetic resonance spectroscopy (MRS), we recently reported increased thalamic glutamate levels in antipsychotic-naïve patients with first episode psychosis (FEP) compared to healthy controls (HC). Importantly, high thalamic glutamate levels predicted poor treatment response (9). Other studies have also associated glutamate levels with treatment outcome (20–22). Nevertheless, cross-sectional studies of glutamate levels between patients with FEP and HC have shown inconsistent results, (21, 23, 24) likely reflecting differences in glutamate levels across brain regions and in patients’ variable previous antipsychotic exposure, as well as the complexity of the disorder, with individual differences regarding the neurobiological findings Studies have also shown that thalamic glutamate levels are heritable, genetically related to psychosis (25), and that in individuals at ultra-high risk of psychosis (UHR), thalamic glutamate levels are related to attention and level of functioning (26, 27). Moreover, we previously demonstrated that thalamic and anterior cingulate cortex (ACC) levels of glutamate are associated with symptomology and cognition, specifically spatial working memory, set-shifting and sustained attention both in UHR and FEP (9, 26, 28, 29).

Cutting-edge 7T MRS can non-invasively determine brain metabolite levels of e.g., γ-aminobutyric acid (GABA), N-acetylaspartate, N-acetylaspartyl and importantly, 7T allows for discrimination of glutamate concentrations from glutathione and glutamine (22, 30, 31). Moreover, multiple voxels/brain regions can be investigated with 7T. In the context of glutamate and psychosis, the thalamus, ACC, hippocampus, dorsolateral prefrontal cortex (DLPFC), and basal ganglia are of particular interest, e.g., (17, 29, 32–34).

Regarding brain structure, the use of structural 7T sequences offers increased sensitivity to quantify neurodegenerative processes such as volume loss e.g., by hippocampal subfield segmentation (35).

Also, high-resolution quantification of tissue parameters such as iron deposition using quantitative susceptibility mapping allow characterization of subtle changes in the “dopaminergic regions” of the basal ganglia, including striatum, pallidum, and substantia nigra (SN) (35). Recently, a promising non-invasive method for in vivo-investigating dopamine in neuropsychiatric illness has been developed. Neuromelanin-sensitive MRI (NM-MRI) purports to detect the content of neuromelanin (NM), a product of dopamine metabolism that accumulates in the SN. NM may be regarded as a proxy of structural and functional integrity of dopaminergic cells in SN (36). Findings indicate that NM-MRI signals correlate with NM concentration, dopamine levels in the striatum and severity of psychosis in schizophrenia, even without neurodegeneration (37).

NMDA-receptor modulation has been investigated in several preclinical studies (38–40). Interestingly, different NMDA-receptor compounds lead to various effects on molecular, anatomical, and behavioral levels. Specifically, memantine exerts opposite effects than ketamine on the genetic level, which may underlie the two drugs’ different clinical effects (38, 41).

Against this background, modulation of glutamatergic NMDAR has been investigated as a potential treatment target in patients with schizophrenia or other psychotic disorders. Memantine is a non-competitive NMDAR antagonist used for treatment of Alzheimer’s disease. Two recent meta-analyses (42, 43) and one systematic review (44) have concluded that adjunctive memantine to antipsychotics is safe and significantly improves negative symptoms in chronic medicated patients with schizophrenia. One of the meta-analyses also indicated a potential beneficial effect on global levels of cognition of patients treated with adjunctive memantine compared to the placebo group (42).

Memantine has been theorized to ameliorate progression of negative symptoms by counteracting excitotoxicity correlated to high glutamate levels in early stages of psychosis (45), but this has not yet been investigated in initially antipsychotic-naïve patients.

AMEND is an ongoing double-blinded clinical trial (Clinicaltrials.gov, ID: NCT04789915), where antipsychotic-naïve patients with first-episode psychosis will be randomly allocated to treatment with first-line antipsychotic compound with addition of memantine or placebo. Patients will be matched with HCs and all participants will undergo baseline and 12-weeks examinations (Figure 1).

We hypothesize that add-on memantine to amisulpride will be superior to amisulpride monotherapy in reducing negative symptoms, and that this effect will correlate with thalamic glutamate levels. Moreover, we anticipate that add-on memantine will restore regional white matter integrity and improve cognitive functioning.

By rational combination of two licensed, off-patent drugs, amisulpride and memantine, AMEND aims to optimize treatment of psychosis, especially the accompanying negative and cognitive symptoms, alongside with unraveling the signature of memantine response. If successful, AMEND will provide pivotal neurobiological evidence for future stratification of patients with first-episode psychosis before initial antipsychotic treatment.

There are few potential risks related to participating in the AMEND study. Side effects of memantine are relatively mild compared to antipsychotic medication. In previous memantine add-on RCTs in schizophrenia patients, the reported side effects did not differ from placebo (42) and chronic patients treated with memantine as add-on treatment had an improvement compared to placebo (79), why we assume that the added treatment also will improve patients with first-episode psychosis. The participants will have the advantages of a thorough examination program with more detailed psychopathological and neuropsychological testing than normally provided.

Magnetic resonance imaging procedures have not been associated with any health risk, however, the examination might cause discomfort for people with claustrophobia and anxiety. Participants will be followed carefully during the project and any potential discomfort will be attempted minimized by research staff. Great care will be taken by the research group to ensure that all examinations are carried out as smoothly and efficiently as possible, including breaks when needed. Scanning will be discontinued in cases of discomfort or anxiety, and communication between staff and participants will be possible during the whole scan.

Participants with head/neck tattoos or permanent make-up might experience warming and soreness of the painted area, and scanning will be interrupted should these symptoms arise. Incidental findings on MRI scans might lead to worrying and further clinical work-up. Blood samplings might cause discomfort for some people but will be taken by trained personal and under hygienic circumstances.

Participants can withdraw from the study at any time without explanation, and without any consequences for the following treatment.

Regarding the statistics, our power calculation and number of participants are not sized for all exploratory analyses, but our RCT design and previous pilot data, justify the rationale for conducting these tests. Admittedly, a mean 2-point reduction in PANSS negative symptoms may only represent a modest clinical improvement for the individual patient participating in this study. However, the RCT design in combination with 7T MRS data will provide strong neuroscientific evidence on how NMDARs may mediate effects on unresolved symptom domains. If successful, the combined clinical and neurobiological insights from AMEND will provide an important steppingstone, which will have major impact for further research, for stakeholders in the pharmaceutical industry, and for future patients with psychosis.

With the ambitious goal of moving clinical psychiatry toward pre-treatment stratification based on pathophysiological markers, AMEND will provide evidence for rational drug repurposing to optimize treatment outcome for patients with psychosis.

The trial has been approved by the Danish Health and Medicines Authority, the National Committee on Health Research Ethics, Danish National Committee on Biomedical Research, and the Danish Data Protection Agency. Trial participation presupposes oral and written informed consent of all participants. Memantine will be used off-label, and tolerability and clinical symptoms will be monitored closely. Personal integrity and privacy concerning participants will be protected by the Danish Law of Health. The study will be conducted in accordance with the Helsinki Declaration II and monitored according to Good Clinical Practice principles (ICH-GCP). AMEND is registered at EudraCT nr: 2021-001061-20 and ClinicalTrials.gov (NCT04789915). The trial can be subjected to quality audit. Research data and results (positive and negative), will be presented at national and international scientific meetings and conferences. Articles will be submitted to peer-reviewed journals. Authorships will adhere to the International Committee of Medical Journal Editors statement (ICMJE).²

The studies involving human participants were reviewed and approved by Danish Health and Medicines Authority, the National Committee on Health Research Ethics, Danish National Committee on Biomedical Research, and the Danish Data Protection Agency. The patients/participants provided their written informed consent to participate in this study.

BE conceived the study. KB, MN, BF, BG, and HS contributed to the study design was adjusted in discussions. AM, CL, VB, HL, DS, MS, AL, WS, and CP refined the clinical- and neuropsychiatric examination programs with respect to their respective fields of expertise. KS, OB, and BE drafted the manuscript. All authors approved the submitted version of the manuscript.

BE was part of the Advisory Board of Eli Lilly Denmark A/S, Janssen-Cilag, Lundbeck Pharma A/S, and Takeda Pharmaceutical Company Ltd; and has received lecture fees from Bristol-Myers Squibb, Boehringer Ingelheim, Otsuka Pharma Scandinavia AB, Eli Lilly Company, and Lundbeck Pharma A/S. HS has received honoraria as speaker from Sanofi Genzyme, Denmark and Novartis, Denmark, as consultant from Sanofi Genzyme, Denmark, and Lundbeck AS, Denmark, and as editor-in-chief (Neuroimage Clinical) and senior editor (NeuroImage) from Elsevier Publishers, Amsterdam, The Netherlands. He has received royalties as book editor from Springer Publishers, Stuttgart, Germany and from Gyldendal Publishers, Copenhagen, Denmark. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.