Antiglutamatergic agents for obsessive-compulsive disorder: Where are we now and what are possible future prospects?

Abstract

Recent data suggest that obsessive-compulsive disorder (OCD) is driven by an imbalance among the habit learning system and the goal-directed system. The frontostriatal loop termed cortico-striatal-thalamo-cortical (CSTC) circuitry loop is involved in habits and their dysfunction plays an important role in OCD. Glutamatergic neurotransmission is the principal neurotransmitter implicated in the CSTC model of OCD. Hyperactivity in the CSTC loop implies a high level of glutamate in the cortical-striatal pathways as well as a dysregulation of GABAergic transmission, and could represent the pathophysiology of OCD. Moreover, the dysregulation of glutamate levels can lead to neurotoxicity, acting as a neuronal excitotoxin. The hypothesis of a role of neurotoxicity in the pathophysiology of OCD clinically correlates to the importance of an early intervention for patients. Indeed, some studies have shown that a reduction of duration of untreated illness is related to an earlier onset of remission. Although robust data supporting a progression of such brain changes are not available so far, an early intervention could help interrupt damage from neurotoxicity. Moreover, agents targeting glutamate neurotransmission may represent promising therapeutical option in OCD patients.

Key Words: Obsessive-compulsive disorder; Antiglutamatergic agents; Glutammate; Early intervention; Neurophysiopathology; Duration of untreated illness

Core Tip: In pathophysiology of obsessive-compulsive disorder (OCD), dysfunction of the cortico-striatal-thalamo-cortical (CSTC) loop could provoke an imbalance between goal-directed system and habit learning system. Glutamate is the principal neurotransmitter implicated in the CSTC model of OCD. Glutammate dysregulation and neurotoxicity seem to be correlated, thus, an early intervention and a reduction of duration of untreated illness appear central in treatment of OCD, as well as the use of glutamate-modulating drugs that could help to interrupt damage from neurotoxicity.

INTRODUCTION

Obsessive-compulsive disorder (OCD) is a severe and debilitating neuropsychiatric condition that affects 2.5%-3.0% of the general population[1,2].

OCD usually occurs in childhood or adolescence[3] and is associated with important distress, disability and suicidality[4]. Conventionally, OCD is driven by irrational beliefs (obsessions) considered as the product of a cognitive bias (including overestimation of threat, increased personal responsibility and thought-action fusion) and compulsions, which are rational avoidance responses triggered by these irrational beliefs that form a coping mechanism to neutralize anxiety and reduce the likelihood that fears will be realized. Preceding DSM-5[5], OCD was considered an anxiety disorder: Anxiety symptoms are indeed commonly expressed by patients[6]. Nevertheless, anxiety is not essential for OCD diagnosis and the condition was considered to be more similar to other disorders, therefore OCD was moved to a separated section of DSM-5[7].

First-line treatments include cognitive behavior therapy (CBT) and pharmacological treatment with serotonin reuptake inhibitors (SRIs)[8]. Unfortunately, around 40% of patients with OCD do not achieve remission of their symptoms with these therapies[6]. Even when an alternate selective serotonin re-uptake inhibitor or clomipramine treatment is given, or when an atypical antipsychotic is added, a considerable portion of refractory patients (30%) remain with the diagnosis, which is associated with serious social disability, patient’s and family’s suffering as well as a high suicide rate[9]. To understand the mechanisms underlying the development of OCD, a variety of hypothesis have been proposed, one of which is a dysfunction of the serotonergic and dopaminergic systems[10]. More recently, different studies focused their attention on the role of the glutamatergic system in OCD pathology[11]. Glutamate is the principal neurotransmitter involved in cortico-striatal-thalamic circuits (CSTC)[12] and, according to recent hypotheses (CSTC model of OCD), it would seem implicated in the pathogenesis of OCD. Moreover, there is consensus among experts on the importance of early intervention in OCD patients to reduce the duration of untreated illness (DUI) as well as to reduce the ‘toxic’ effect of an extended DUI in OCD[13]. The first aim of this mini review is an overview of the role of glutamate in CSTC models of OCD and the use of antiglutamatergic agents. Moreover we propose an intervention on the DUI in order to optimize fundamental time, due to the supposed toxic damages, and subsequently an early use of antiglutamatergic agents. Therefore, an early intervention could be both reduce the toxic effect of an extended DUI and, if necessary, encourage early use of antiglutamatergic agents.

WHAT DRIVES THE NEUROPHYSIOPATHOLOGY OF OCD?

As previously mentioned, conventionally OCD is driven by obsessions, while compulsions would be a rational avoidance response triggered by irrational beliefs. Patients report that, despite the repetitive and ritualistic nature of such behaviors, which are unproductive and frequently without any purpose, they are unable to discontinue them.

Recent data[14,15] suggest that OCD is driven by interference in the balance between the habit learning system and the goal-directed system. The habit learning system is based on historical information, regardless of past rewards, and it can lead to behavioural rigidity even in the face of rapid changes in the environment; on the other hand, the goal-directed system applies control over habits in light of changes, including changes in response to the evaluation of actions and outcomes. A neurocomputational study[16] found that OCD patients made choices mostly based on model-free (i.e., habit) rather than model-based (i.e., executive control) learning. It is difficult for patients with OCD to modulate their future behavior based on immediate feedback. Therefore, in light of these hypotheses, habit formation appears to be abnormal in patients affected by OCD. According to the contemporary habit hypothesis, compulsive behaviours would not be driven by irrational and intrusive thoughts, but would be the consequences of a deficit in the control over goal-directed actions, leading to amplified reliance on habitual thoughts.

Orbitofrontal and cingulate cortices and the caudate nucleus are involved in habit learning[17,18] and in goal-directed control[19-21]. The frontostriatal loop termed CSTC circuitry is involved in habits and there is consensus that dysfunction of these areas plays an important role in OCD[22-24]. Dysfunction of the CSTC could cause disruptions between the goal-directed system and habitual control[25], which would cause an overreliance on the latter. Neuroimaging studies have shown that dysfunction of these areas has been implicated in OCD, including structural abnormalities, altered brain activation and connectivity[22,26]. The hypothesis is that in patients with OCD, the hyperactivation or hyperconnection of CSTC leads to a uncontrolled positive feedback loop[24,27]. This would trigger the impulse to perform compulsions, which would in turn consolidate the habit of executing compulsions, increasing the need to perform them[10].

In line with this behavioural hypothesis, OCD has been removed from anxiety disorders in DSM-5 and placed into its own category of “obsessive-compulsive and related disorders”.

STATE OF ART ABOUT THE USE OF ANTIGLUTAMATERGIC AGENTS IN OCD

Traditionally, OCD medications have targeted serotonergic pathways. SRIs are commonly administered to treat patients with OCD and although the exact mechanism of their action remains elusive, SRIs are considered to exert their effects by influencing the CSTC[28]. However, the proportion of non-responder patients suggests a role of other neurotransmitter systems outside of the serotonergic in the pathophysiology of OCD[10].

Glutamate is the principal neurotransmitter implicated in the CSTC model of OCD[12]. In the neuropathophysiology of OCD, a dysregulation of the glutamatergic signal within the cortico-striatal circuitry has been suggested, which would lead to a reduced concentration of glutamate in the anterior cingulate cortex on the one hand, and overactivity of glutamatergic signalling in the striatum and orbitofrontal cortex on the other[29]. Hyperactivity in the CSTC loop implies a high level of glutamate in the cortical-striatal pathways and a dysregulation of GABAergic transmission that could represent the pathophysiology of OCD[10]. Furthermore, both ionotropic and metabotropic glutamate receptors are located in the candidate brain circuits of OCD and have been virtually associated with every form of learning in the brain, including habit learning[14,30]. Additionally, several studies showed an increase in glutamate in the cerebrospinal fluid of OCD patients compared to controls[1,31,32]. Genetics studies have also found an association between glutamate genes and OCD[33-35]. A polymorphism encoding for N-methyl-D-aspartic acid receptor (NMDAR) has been associated with OCD in families[33].

In light of a possible role of glutamatergic signal dysregulation in OCD, it is possible to find in glutamatergic drugs candidates for the augmentation strategy of OCD’s therapy. Glutamate-modulating drugs have shown promise as potential therapeutic agents in other psychiatric disorders with a high comorbidity with OCD such as depression, bipolar disorder and suicide[36-38]. Moreover, several studies suggested that neuromodulation techniques using noninvasive devices, such as transcranial magnetic stimulation (TMS), or invasive procedures, such as deep brain stimulation (DBS), could offer additional support for the CSTC model of OCD. Based on the results of a recent randomized controlled trials (RCT), the Food and Drug Administration approved deep TMS, for the treatment of OCD[39,40]. Also DBS, which involves implantation of electrodes that modulate specific brain function, is approved for treatment of refractory OCD[41]. The first trials on DBS for OCD were conducted in the 2003 by Gabriëls et al[42] showed that DBS targeted at striatal areas is effective and safe for patients with refractory OCD.

This background has motivated the interest in studying glutamate modulators in patients who are unresponsive to standard pharmacotherapeutic approaches. Several clinical studies were conducted to evaluate the effect of glutamate-modulating drugs in OCD, however, they differed significantly in terms of treatment resistance, comorbidities, age and gender of the patients[29].

In particular, some studies using memantine and riluzole reported promising benefits[43]. Other studies have explored the role of stimulants and nutritional supplements such as N-acetylcysteine (NAC) and glycine[44-46]. Glutamate-modulating RCT for refractory OCD are reported in Table 1. To be more precise we describe the principal studies reported in the literature:

Table 1 Glutamate-modulating randomized controlled trials for refractory obsessive-compulsive disorder.

Ref.	Drugs	Adjunctive/monotherapy	Subjects	Duration & dose	Results
Haghighi et al[48], 2013	Memantine	Adjunctive to SRIs	40 pt	12 wk. Dose: 5-10 mg/d	Significant reduction. Y-BOCS score
Ghaleiha et al[49], 2013	Memantine	Adjunctive to fluvoxamine	42 pt	8 wk. Dose: 20 mg/d	Significant reduction. Y-BOCS score
Modarresi et al[50], 2018	Memantine	Adjunctive to SRIs	32 pt	12 wk. Dose: 20 mg/d	Significant reduction. Y-BOCS score
Farnia et al[51], 2018	Memantine	Adjunctive to SRIs	99 pt	8 wk. Dose: 10 mg/d	No significant reduction. Y-BOCS score
Naderi et al[56] , 2019	Amantadine	Adjunctive to SRIs	100 pt	12 wk. Dose: 100 mg/d	Significant reduction. Y-BOCS score
Grant et al[59], 2014	Riluzole	Adjunctive to SRIs	60 pt	12 wk. Final dose: 100 mg/d	No significant reduction. Y-BOCS score
Pittenger et al[58], 2015	Riluzole	Adjunctive to SRIs	38 pt	12 wk. Final dose: 100 mg/d	No significant reduction. Y-BOCS score
Emamzadehfar et al[60], 2016	Riluzole	Adjunctive to fluvoxamine	50 pt	10 wk. Final dose: 100 mg/d	Significant reduction. Y-BOCS score
Rodriguez et al[61], 2013	Ketamine	Monotherapy	15 pt	Intravenous infusions dose: 0.5 mg/kg	Significant reduction. Y-BOCS score
Greenberg et al[46], 2009	Glycine	Adjunctive pharmacological or psychotherapeutic treatment	24 pt	12 wk. Dose: 60 mg/d	No significant reduction. Y-BOCS score
Afshar et al[44], 2012	N-acetylcysteine	Adjunctive to SRIs	48 pt	12 wk. Dose: 2.4 g/d	Significant reduction. Y-BOCS score
Sarris et al[68], 2015	N-acetylcysteine	Adjunctive to SRIs	44 pt	16 wk. Dose: 3 g/d	No significant reduction. Y-BOCS score
Paydary et al[45], 2016	N-acetylcysteine	Adjunctive to fluvoxamine	44 pt	10 wk. Dose: 2 g/d	Significant reduction. Y-BOCS score
Costa et al[69], 2017	N-acetylcysteine	Adjunctive to SRIs	40 pt	16 wk. Dose: 3 g/d	No significant reduction. Y-BOCS score
Ghanizadeh et al[70], 2017	N-acetylcysteine	Adjunctive to SRIs	34 pt	10 wk. Dose: 2.4 g/d	Significant reduction. Y-BOCS score
Li et al[71], 2020	N-acetylcysteine	Adjunctive to SRIs	11 pt	12 wk. Dose: 2.7 g/d	Significant reduction. Y-BOCS score
Esalatmanesh et al[75], 2016	Minocycline	Adjunctive to fluvoxamine	102 pt	10 wk. Dose: 200 mg/d	Significant reduction. Y-BOCS score
Kushner et al[78], 2007	D-cycloserine	Adjunctive to CBT	32 pt	125 mg	Significant reduction. Y-BOCS score
Wilhelm et al[79], 2008	D-cycloserine	Adjunctive to CBT	23 pt	100 mg	Significant reduction. Y-BOCS score
Farrell et al[80], 2013	D-cycloserine	Adjunctive to CBT	17 pt	25-50 mg	Significant reduction. Y-BOCS score
Andersson et al[81], 2015	D-cycloserine	Adjunctive to CBT	128 pt	12 wk. Dose: 50 mg	Significant reduction. Y-BOCS score
Mataix-Cols et al[82], 2014	D-cycloserine	Adjunctive to CBT	27 pt	50 mg	Significant reduction. Y-BOCS score
Storch et al[83], 2007	D-cycloserine	Adjunctive to CBT	24 pt	250 mg	No significant reduction. Y-BOCS score
Storch et al[85], 2010	D-cycloserine	Adjunctive to CBT	30 pt	25 mg	No significant reduction. Y-BOCS score
Storch et al[84], 2016	D-cycloserine	Adjunctive to CBT	142 pt	50 mg	No significant reduction. Y-BOCS score

RCT: Randomized controlled trials; OCD: Obsessive-compulsive disorder; SRIs: Serotonin reuptake inhibitors; CBT: Cognitive behaviour therapy; pt: Patients; Y-BOCS: Yale-Brown Obsessive Compulsive Scale.

Memantine

Memantine, a low-to-moderate affinity noncompetitive NMDAR antagonist[47] currently approved for the treatment of Alzheimer’s disease, is used as an augmentative agent to SRIs in the treatment of moderate to severe OCD. Four RCTs[48-51] were conducted using memantine vs placebo as an adjunctive treatment to SRIs[48,50,51] or fluvoxamine[49]. In three of these, conducted on 40, 42 and 32 patients, respectively[48-50], the adjunction of memantine (in the target dose of 5-10 mg/d for 12 wk[48] or 20 mg/d for 8 wk[49] or 12 wk[50] showed a reduction in scores at the Yale-Brown Obsessive Compulsive Scale (Y-BOCS). In the last RCT[51] conducted on 99 patients, the use of memantine (in the target dose of 10 mg/d for 8 wk) vs placebo and vs gabapentin did not lead to the reduction of Y-BOCS scores.

A recent review on the subject conducted by Marinova et al[29] concluded that memantine is the compound that most consistently showed a positive effect as an augmentation therapy in OCD. However, a critical letter[52], exposed some concern on this conclusion mainly due to some methodological gaps and because all four RCTs were conducted in the same geographic area. Andrade[52] concluded that, in the light of these limitations, it is still premature to recommend the use of memantine as an OCD augmentation therapy and that more studies are needed.

Amantadine

Amantadine, a weak noncompetitive NMDAR antagonist originally used in the treatment and prophylaxis of viral infections, has been used in the treatment of some neurological conditions including Parkinson’s disease, dementia, multiple sclerosis, cocaine withdrawal, pain and depression[53-55]. In one RCT, conducted on 100 patients diagnosed with moderate to severe OCD[56] the use of amantadine vs placebo as an adjunctive treatment to SRIs, showed a significant effect on Y-BOCS compared to placebo.

Riluzole

Riluzole acts through voltage-gated ion channels modulating the outflow of glutamate and enhancing the reuptake of extracellular glutamate. It is approved for the treatment of amyotrophic lateral sclerosis and some data report benefit in the use of riluzole for depression and anxiety[57]. Three RCTs were conducted using riluzole vs placebo adjunctive to SRIs[58,59] or fluvoxamine[60]. Only one study, conducted on 50 patients[60], using riluzole at a final dose of 100 mg/d for 10 wk, showed a reduction in the Y-BOCS score; while the other two, conducted on 38[58] and 60 patients[59], respectively, both using riluzole at a final dose of 100 mg/d for 12 wk, did not lead to the reduction in the Y-BOCS scores.

Ketamine

Ketamine, a noncompetitive NMDA receptor antagonist, is approved for the treatment of depression. An RCT by Rodriguez and colleagues[61], conducted on 15 drug-free adults with OCD (intravenous infusions of ketamine 0.5 mg/kg vs placebo) reported significant improvement of OCD symptoms in patients that received ketamine vs placebo. This study was the first RCT that proved the effect of antiglutamatergic agents on OCD symptoms without the presence of an SRIs.

Glycine

Glycine is an amino acid, NMDA glutamate receptor agonist. Only one RCT[46] conducted on a sample of 24 patients with OCD used glycine augmentation (60 g/d for 12 wk) vs placebo. This study had a high dropout rate due to the adverse effects of glycine. However, nearly significant improvements in Y-BOCS scores were observed in evaluable patients.

Nicotine

Some studies have also suggested that nicotine acts by promoting glutamatergic transmission as well as by stabilizing the glutamatergic hyperactivity of the loop running from the orbitofrontal cortex to the cyngulate gyrus, the striatum and the thalamus[62-65]. Nicotine acts on the desensitisation of nAChRs by increasing both glutamatergic transmission and, transiently, GABAergic transmission. In particular, nAChR on GABA neurons are more desensitized then those on the glutamatergic system. This would lead to a shift toward excitation, reducing inhibitory inputs[65]. According to these data and with the complexity of OCD’s pathophysiology, nicotine can be considered as a possible add-on therapeutic option in OCD resistant subjects, due to its peculiar mechanism of action. Only one study[66] was conducted on 5 patients using nicotine. Patients who received nicotine reported a reduction in Y-BOCS scores, especially in the compulsive score. A recent systematic review on the efficacy of nicotine administration suggested the need to test nicotine use in OCD in a large RCT. However, such suggestion raises ethical issues related to nicotine administration for its potential addiction effect[67].

NAC

NAC is a glutamate-modulating drug acting through the inhibition of the synaptic glutamate release. Five RCTs were conducted on adults[44,45,68-70] and one on children[71] using NAC vs placebo as an adjunctive therapy to SRIs[44,68-71] or fluvoxamine[45]. In three studies, conducted on 48, 44, 34 and 11 patients, respectively[44,45,70,71], the use of NAC (in the dose of 2.4 g/d for 12 wk[44] or 2 g/d for 10 wk[45] or of 2.4 g/d for 10 wk[70] or 2.7 g/d for 12 wk[71] showed improvement in the severity of OCD symptoms. On the other hands, in two studies conducted on 44[68] and 40 patients[69], respectively, the use of NAC (in the dose of 3 g/d for 16 wk)[68,69] did not lead to a reduction of OCD symptoms.

Minocycline

Minocycline is an antibiotic that crosses the blood-brain barrier and has a neuroprotective effect that decreases glutamate-induced neurotoxicity[72]. It is used in some neurodegenerative diseases such as amyotrophic lateral sclerosis and Parkinson[73,74]. Only one RCT was conducted on 102 patients using minocycline (200 mg/d for 10 wk) vs placebo as an adjunctive treatment to fluvoxamine. The results showed a reduction in Y-BOCS scores in the group that received minocycline[75].

D-cycloserine

D-cycloserine is a partial agonist at the N-methyl-d-aspartate receptor in the amygdala and some data have suggested a role of this receptor in fear extinction. For this reason, it was hypothesized that d-cycloserine may increase the efficacy of exposure therapy, a component of CBT for anxiety and OCD[76,77].

Eight RCTs were conducted on OCD patients using d-cycloserine vs placebo as an augmentation treatment of CBT with exposure and response prevention. Among these RCTs, four studies, conducted on 32[78], 23[79], 17[80] and 128 patients[81], respectively, used d-cycloserine in doses of 25[82] to 125 mg[78] close to each CBT session, reporting a significant amelioration of OCD symptom. In the other four studies, conducted on 27[82], 24[83], 30[45] and 142[84] patients, respectively, used d-cycloserine in doses of 25[85] to 250 mg[84] close to each CBT session, but did not show any statistically significant improvement.

In light of such contrasting results, it is not yet possible to give a clear answer on the use of glutamate-modulating drugs in OCD.

Therefore, further randomized placebo-controlled trials in larger study populations are essential to draw definitive indications on the efficacy of the use of antiglutamatergic agents in OCD.

THE REASON WHY EARLY INTERVENTION COULD BE IMPORTANT FOR OCD

As we underline, glutamate neurotransmission plays a crucial role in the CSTC of OCD[12]. Furthermore, glutamate is also implicated in neuronal plasticity, learning, and memory[86]. Dysregulation of glutamate levels may bring glutamate receptor hyperactivity or even excitotoxicity in neurons and, in pathological situations, glutamate could lead to neurotoxicity by acting as a neuronal excitotoxin[87].

In support of this hypothesis, neuroimaging studies found changes in multiple sites of age-related brain structures suggesting that OCD is characterized by complex and nonlinear neurophysiopathological mechanisms[88,89].

The hypothesis of the role of neurotoxicity in the neurophysiopathology of OCD is clinically related to the importance of early intervention for patients. Indeed, some studies showed that a faster onset in the treatment of OCD correlates with an earlier onset of remission[90]. On the contrary, late intervention is associated with increased comorbidity, disability, reduced treatment response and remission[91-93].

The DUI for OCD is very high compared to other mental disorders[94]. Most cases of OCD arise in childhood or adolescence[92] and symptoms are often ignored[95] extending the time for diagnosis (from 2 to 10 year) and consequently for starting treatment[92,96,97] with average DUI of two to three years[98,99]. Although there is a paucity of data to support a progression of such brain changes so far, it is assumed that early intervention, even with the use of glutamate-modulating drugs, could help to stop neurotoxic damage[13]. A crucial point may be the switch from a time-dependent neurotransmitter dysfunction to an irreversible structural damage. This may explain both refractory to SSRI and the unresponsiveness to anti-glutamatergic agents.

Identifying and confirming the efficacy and safety of new therapeutic approaches for OCD could also be very useful for early intervention in these patients. An international group of expert clinicians and scientists with extensive OCD experience are documenting the negative impact associated with the delayed treatment on clinical outcomes, suggesting the importance of a greater emphasis on targeted early intervention for OCD patients as is already is the case for psychotic disorders. Early intervention is more promising for reducing chronicity as well as the economic and social burdens associated with OCD.

CONCLUSION

The hypothesis of a hyperactive CSTC loop that implies a high level of glutamate in the cortical-striatal pathways and a dysregulation of GABAergic transmission may represent the pathophysiology of OCD. Despite the effectiveness of SRIs in treating OCD, the treatment-resistant symptoms observed in 40% of patients present enormous clinical issues. The refractory to SRIs and the unresponsiveness to antiglutamatergic agents may be explained by a time-dependent neurotransmitter dysfunction, which may lead to an irreversible structural damage. There is a need to develop novel pharmacological strategies, which are not exclusively related to refractory subjects. If glutamate-neurotoxicity is extensively confirmed, the time from OCD onset will be the most important variable to take into consideration. In such a scenario, agents targeting glutamate neurotransmission may represent a promising treatment, especially in the case of a timely prescription.