For many decades, the dopamine theory has been the most influential neurobiological concept in understanding psychosis (1) and it is argued that all currently used antipsychotics work through directly or indirectly changing dopaminergic neurotransmission (2). Although there is converging evidence from different lines of research for the involvement of dopamine in the emergence of psychotic symptoms (1), alterations in other neurotransmitter systems have also been demonstrated, and there is growing agreement that different neurobiological etiologies can underlie psychotic syndromes (3–5).

Firstly, a glutamate theory of schizophrenia has been established mainly based on the observation that antagonists at the glutamatergic NMDA receptor, such as ketamine, phencyclidine, or endogenously formed anti-NMDA antibodies, directly and reliably cause psychotic symptoms and genetic studies showing an increased risk of schizophrenia in individuals with gene variants affecting the glutamate system (6, 7). However, since clinical trials of glutamatergic agents developed based on this theory have yielded only mixed results, no primary glutamatergic drug is to date available for routine clinical use. Currently, studies on glutamatergic drugs focus on compounds targeting specific receptor subtypes as well as on subgroup analyses to identify patient groups with marked glutamatergic alterations and hence potentially increased response to glutamatergic interventions (6, 7).

Secondly, reduced neurotransmission at acetylcholine receptors has been proposed as a non-dopaminergic mechanism underlying schizophrenia symptoms. Acetylcholine receptors are generally divided into nicotinic and muscarinic receptors according to binding ligand, and both have been implicated in the etiology of schizophrenia. Based on the extremely high rate of nicotine use in patients with schizophrenia, which has been interpreted as self-medication of existing deficits, nicotinic acetylcholine receptors, particularly the α7-subtype, have initially become the focus of research interest (8, 9). An important line of research in this context involves sensory gating, i.e., the inhibition of electrophysiological brain responses to (mostly auditory) stimuli played at close temporal intervals after an initial stimulus. Deficits in sensory gating have been frequently reported in schizophrenia patients and linked to attentional deficits (8, 10, 11). Interestingly, agonists at nicotinic receptors (e.g., nicotine) were shown to improve these sensory gating deficits, as well as some of the cognitive deficits, in schizophrenia patients (9, 12, 13). Moreover, post-mortem studies confirmed a reduced density of α7—receptors in different brain regions in deceased schizophrenia patients. Both findings renewed interest in developing nicotinic drugs for schizophrenia, especially for its cognitive deficits, which are insufficiently ameliorated by conventional antipsychotics (8, 9). However, despite promising results in preclinical and phase 1/2 trials, phase 3 trials on nicotinic agonists in schizophrenia have been disappointing due to lack of efficacy and/or tolerability and no primarily nicotinic drug is approved for schizophrenia treatment (12, 13). Current research on nicotinic agents for schizophrenia hence focuses on finding optimal dosage regimes and investigating allosteric modulators (13).

A deficit at muscarinic acetylcholine receptors in at least a proportion of patients with schizophrenia is suggested, for example, by post-mortem studies showing reduced density of muscarinic receptors in patients, by imaging studies with radioligands showing reduced availability of muscarinic receptors, and by the ability of muscarinic antagonists such as scopolamine to elicit schizophrenia-like symptoms in healthy individuals (14, 15). Based on post-mortem studies, it was proposed that approximately one quarter of schizophrenia patients belongs to a subgroup referred to as muscarinic receptor deficit sub-group with schizophrenia (MRDS), which shows a massive and widespread reduction in the density of muscarinic receptors in the CNS (16–18). Based on that, agonists and positive allosteric modulators (PAM) at muscarinic receptors have been investigated as potential schizophrenia treatments [Table 1 for an overview, (22) for further discussion]. Intriguingly, recent successes were reported in phase 2 (20) and phase 3 (21) trials of xanomeline, an agonist at the muscarinic M1 and M4 cholinergic receptors with no direct dopaminergic effect. It should be noted that in these trials, xanomeline was administered combined with trospium, a peripheral muscarinic antagonist, in order to mitigate peripheral cholinergic side effects, but for reasons of simplicity, I will speak of xanomeline as the centrally active compound in the following.

These developments make it seem possible that in the near future, a second substance class of muscarinic antipsychotics may be established alongside current antidopaminergic antipsychotics. This would lend new urgent importance to the question of whether there is an “antimuscarinic subtype” in schizophrenia and, particularly, how to identify it in clinical settings since this subtype might show non-response to antidopaminergic medication and response to procholinergic treatment. This perspective is intended to present several potential in vivo markers of muscarinic deficit in schizophrenia that ideally in the future could be used for personalized therapy planning and help in deciding whether to initiate treatment primarily with a dopamine antagonist or a muscarinic acetylcholine agonist.

The following sections describe markers that could indicate a muscarinic deficit in patients with schizophrenia. A summary is provided in Table 2.

The current success of agonists at muscarinic acetylcholine receptors in the treatment of schizophrenia makes it seem possible that in the future a second substance class of procholinergic antipsychotics will be established alongside the primarily antidopaminergic antipsychotics. In this case, it would be important for individualized treatment to find specific predictors of response to procholinergic (compared with antidopaminergic) antipsychotics. Possible candidates for such predictors include treatment resistance to antidopaminergic treatment, the presence of severe cognitive impairment and/or visual hallucinations, reduced mismatch negativity, the presence of antimuscarinic antibodies, and reduced availability of muscarinic receptors on radionucleotide imaging.

Muscarinic acetylcholine receptors can be divided into 5 subtypes (M1–M5), with mainly agonists at M1 and M4 subtypes being considered candidates in schizophrenia treatment, whereas stimulation of M2 and M3 subtypes is related to gastrointestinal and other peripheral side effects. Animal research furthermore suggests that enforcing stimulation at M4 receptors through positive allosteric modulators (PAM) particularly improves proxy markers of positive schizophrenia symptoms such as hyperlocomotion or impaired prepulse inhibition induced by dopamine agonists or NMDA antagonists (28, 57). In contrast, PAM at M1 receptors have mainly been investigated for cognition-enhancing potentials across different neuropsychiatric disorders including schizophrenia. For instance, it could be shown that PAM at M1 receptors improve performance in tasks designed to assess cognitive ability (e.g. novel object recognition task) in schizophrenia mouse models (28, 57). While xanomeline is an M1 and M4 preferring but not fully specific muscarinic agonist, these results highlight that possibly in the future specific schizophrenia symptom clusters could be addressed with selective modulators. In this case, treatment could be guided by both predominant symptoms (positive vs. cognitive symptoms) and possibly by investigations with subtype-specific PET tracers (58).

A crucial question in attempts to find biomarkers for response to muscarinic agonists concerns the precise cause-effect relationship between muscarinic deficit and psychotic symptoms, specifically the question, if muscarinic deficiency causes psychosis through dopaminergic or dopamine-independent pathways. Obviously, the hope to identify distinct predictors for the response to antidopaminergic vs. procholinergic treatment rests on the assumption that there are dissociable etiologies behind psychotic syndromes (schematically, a hyperdopaminergic and a hypomuscarinergic etiology, respecetively). A variety of complex interactions between muscarinergic and dopaminergic neurotransmission have been established (59), leading to the hypothesis that muscarinic agonists ultimately work through modifying dopamine release (14). This hypothesis is supported for example by the finding that psychotic symptoms in M1 receptor knockout mice are related to increased dopamine levels in the nucleus accumbens (60). Conversely, substantial parts of the evidence summarized above like the existence of a distinguishable subgroup with marked deficiency in muscarinic receptors, the non-response to antidopaminergic drugs in a proportion of patients, and the discovery that non-responding patients show no alterations in the dopamine system according to PET studies speak for the existence of patient subgroup where psychotic symptoms are mainly caused by dopamine-independent mechanisms (3). Accordingly, it has been hypothesized that different subtypes of schizophrenia pathophysiologically involve a disturbance in the dopaminergic vs. cholinergic modulation of glutamate receptors, which ultimately causes the development of psychotic symptoms (4). This would indeed imply a primarily dopaminergic vs. a primarily cholinergic etiology, and thus the hope that (a proportion of) patients with inadequate response to dopaminergic interventions could benefit from muscarinergic agents. Fittingly, a potential synergism of antidopaminergic and promuscarinic interventions is suggested by animal studies showing enhanced efficacy of classical antipsychotics by additional administration of M1 receptor PAM (61). Results from the ongoing ARISE trial, which is evaluating the efficacy of xanomeline-trospium as add-on therapy in 400 patients with inadequate response to antidopaminergic antipsychotics such as risperone, paliperone, and aripiprazole and is scheduled to be completed in March 2024 (62), may soon provide new insights into this debate: Adding a second antipsychotic to a first antipsychotic has generally not been shown to be effective in randomized trials (63) and is not a recommended strategy in the case of non-response (64), which may be because all current antipsychotics ultimately act through dopaminergic mechanisms and therefore no synergistic effect occurs when combining multiple agents. Hence, xanomeline proving effective as an add-on treatment would be (tentative) evidence that it indeed possesses a distinct non-dopaminergic mechanism of action.

The most rigorous way of testing the capacity of the suggested biomarkers to identify those schizophrenia patients who will likely benefit more from a procholinergic as compared to an antidopaminergic treatment would be to test their specific ability to predict individual outcome of a clinical trial of, for example xanomeline-trospium compared with amisulpride treatment. A specific predictor, as compared to a general prognostic factor, is a pretreatment variable (e.g., a quantification of the markers listed above), whose influence on the treatment outcome (e.g., on the reduction in a schizophrenia severity score such as the positive and negative syndrome scale) differs between treatments (65). For instance, it could be hypothesized that presence of antimuscarinic antibodies predicts good treatment outcome in the xanomeline arm, but bad outcome in the amisulpride arm of a trial. Statistically, this would be captured in a significant interaction effect between presence of antimuscarinic antibodies and treatment in a predictive model (65). A study testing such treatment-predictor interactions would be quite effortful, as all potential markers would need to be collected at baseline and the detection of treatment-by-variable interactions requires relatively large sample sizes (66). However, if successful, i.e., if one or more markers prove to be specific predictors of the outcome of procholinergic or antidopaminergic treatment, it would represent a real milestone in the individualized treatment of schizophrenia.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The author confirms being the sole contributor of this work and has approved it for publication.