Adjunctive treatment with evenamide, a glutamate modulator, is associated with clinically meaningful benefits in patients with treatment-resistant schizophrenia and inadequate response to antipsychotics: recent clinical findings from randomized trials

Abstract

Background:

Despite their efficacy in improving positive symptoms, first- and second-generation antipsychotics (FGAs and SGAs) fail to provide benefit to a large portion of patients with schizophrenia. Evenamide, uniquely acting at the site of dysfunction, normalizes hippocampal glutamatergic aberrant activity and reduces downstream hyperdopaminergic state, potentially alleviating also negative and cognitive symptoms, and providing benefit to patients with treatment-resistant schizophrenia (TRS) or inadequate response to antipsychotics (APs).

Objectives:

Evaluate the clinical benefits of glutamate modulation through evenamide as an add-on treatment to FGAs and SGAs in patients with TRS or inadequate response to APs.

Design:

Post-hoc analyses of results from two phase II/III clinical trials with evenamide add-on to antipsychotic treatment.

Methods:

Data from Study 014/015, an open-label, 1-year trial in patients with TRS, was analyzed to assess the proportion of patients no longer meeting the baseline severity criteria for TRS (modified Intent-to-Treat (mITT) population; Observed Cases (OC)) and the proportion of patients in the mITT population achieving remission. Data from Study 008A, a randomized, double-blind, placebo-controlled, 4-week trial in patients with schizophrenia not adequately benefiting from SGA (including clozapine), was analyzed to assess patients’ response according to the number of previously failed AP attempts (ITT population; mixed model repeated measures (MMRM) linear regression model). The effect of evenamide on social functioning and life engagement was evaluated in both studies (mITT population) using an MMRM linear regression model for Study 008A and an OC analysis for Study 014/015.

Results:

Patients with TRS receiving evenamide add-on for 1-year improved to such an extent that 55% of them no longer satisfied baseline TRS severity criteria, and approximately one-fourth achieved remission according to literature criteria (27.6% Lieberman et al., 1993; 25.0% Andreasen et al., 2005). In Study 008A, irrespective of the number of failed APs, a statistically significant drug-placebo difference was observed in patients with a single failed attempt (−4.9, p = 0.0122) and in those with 2 or more failed attempts (−2.36, p = 0.0311). Moreover, add-on treatment with evenamide in Study 014/015 was associated with a progressive improvement on the PANSS subdomains of social functioning and life engagement. In Study 008A as well, evenamide resulted in a greater effect, compared to placebo, on the same subdomains.

Conclusion:

Add-on treatment with evenamide to APs (including clozapine) was associated with clinically meaningful and progressive long-term benefits across numerous post-hoc analyses including improvements on patient’s daily life and functioning.

Plain language summary

Evenamide, a new drug under development, may help people with schizophrenia who do not respond to standard treatments, improving symptoms and daily functioning

Many standard medications for schizophrenia, called antipsychotics, help reduce “positive” symptoms (like hearing voices or having unusual thoughts), but they do not benefit everyone. Some patients still have issues with thinking, emotions, and one third of them do not gain any benefit from treatment, a condition called treatment-resistant schizophrenia (TRS).

What is evenamide

Evenamide is a novel drug under development that works differently from other antipsychotics. It helps normalize the neurotransmitter glutamate in a brain area called ‘hippocampus’ which in turn rebalances the levels of another key substance, dopamine. Because of this, evenamide might help not just the “positive” symptoms but also the “negative” ones (low motivation or flat mood).

What was studied

Researchers looked at two main clinical trials:

1. A one-year study in people with TRS, where they added evenamide to the regular antipsychotics patients were already using. They checked how many patients improved so much that they could not being defined TRS anymore, and how many achieved remission. 2. A four-week, placebo-controlled trial in patients who did not respond adequately to antipsychotics. Researchers looked at how previous failed treatments affected response to evenamide added to their standard of care. Researchers also looked at the effects of evenamide on patients’ social involvement and wellbeing.

What was found

1. In the one-year study, more than half of patients improved to the point that they no longer met criteria for TRS, and about one-fourth achieved remission. 2. In the four-week study, patients given evenamide improved more than those on placebo irrespective of how many failed prior treatments patients had gone through. In both studies, meaningful gains in social functioning and wellbeing in patients taking evenamide were observed.

Take-home message

These results suggest that evenamide may deliver long-term benefits, enhancing the wellbeing and social involvement.

Introduction

More than 40 first-and-second-generation antipsychotics (FGAs and SGAs) that have been approved worldwide for the treatment of schizophrenia in the last 75 years have led to an improvement of positive symptoms (hallucinations, delusions, hostility, conceptual disorganizations, etc.), but without any improvement of negative, cognitive symptoms, or functioning. These benefits derive from the blockade of Type-2 Serotonin (5HT2)/Type-2 Dopamine (D2) receptors, without the involvement of other neurotransmitters, such as glutamate, gamma Aminobutyric acid (GABA), etc., that have been shown to be abnormal in patients with schizophrenia and involved in the causation of treatment resistance to antipsychotics. ¹ Additionally, only clozapine has demonstrated efficacy in benefiting patients with treatment-resistant schizophrenia ² (TRS) and reducing suicidality (InterSePT study ³ ). These therapeutic limitations have prompted attempts to benefit patients with TRS through the use of antipsychotic polypharmacy, that is, the addition of one antipsychotic to another, which has been noted in 20%–40% ⁴ of cases in recent years, although no study with current antipsychotics has demonstrated efficacy as an add-on treatment. ⁵ In addition, polypharmacy may increase the prevalence of treatment-related side effects, ⁶ and it seems to be useful only in the management of hyperprolactinemia, by prescribing an additional antipsychotic with a weaker D2 blockade, for example, aripiprazole, which allows reducing the dose of the D2-blocking drug. Therefore, rather than using combinations of similar drugs, adding a medication with a different mode of action that addresses the underlying pathology in these patients, such as a glutamate modulator, may enhance therapeutic response without increasing side effects.

The well-established dopamine hypothesis of schizophrenia has lately given ground to the N-methyl-D-aspartate receptor (NMDAR) hypofunction hypothesis that suggests that hypofunction of NMDARs, expressed by cortical and hippocampal GABAergic neurons, plays a key role in the pathophysiology of schizophrenia. ⁷ The reduced activity of NMDARs on parvalbumin-positive GABAergic interneurons in the ventral hippocampus would result in disinhibition and consequent hyperexcitability of pyramidal neurons, leading to increased glutamate release and firing of dopaminergic neurons in the ventral tegmental area. ⁸ Because reduced recurrent inhibition also disrupts cortical processing, this hypothesis can explain the neural basis, not only of positive, but also of cognitive and negative symptoms. Consistently, insights from the MAM (methylazoxymethanol acetate) animal model ⁹ indicated that the striatal dopamine system itself is not the primary site of pathology but is likely dysregulated by hyperactivity of the hippocampus. ¹⁰ While traditional antipsychotics block dopamine receptors to compensate for dopaminergic system overactivity, they act several steps downstream from the site of dysfunction, leading to significant side effects, noncompliance, and ineffectiveness in treating negative and cognitive symptoms, particularly in TRS.

Numerous lines of evidence involving subjects with at-risk mental state for psychosis, patients experiencing their first episode of psychosis, and patients with TRS, support the involvement of glutamate abnormalities in the aetiopathogenesis of schizophrenia. For instance, a growing body of evidence indicates that patients with TRS have a normal striatal dopamine synthesis capacity (comparable to healthy volunteers), while patients considered as “responders”, that is, those who respond to D2 antagonists, exhibit increased capacity for dopamine synthesis.^11,12 However, paradoxically, glutamate levels are significantly higher in patients with TRS, compared to healthy volunteers, and higher compared to patients who respond to D2-blocking antipsychotics. ¹³

This accumulating evidence supports the belief that agents that regulate glutamate signaling may have therapeutic potential in treating the symptomatology of schizophrenia. The most common strategy for achieving this has been targeting NMDAR dysfunction through agents that modulate NMDAR activity. Molecules with this mechanism of action have been evaluated as add-on in clinical trials and, although altogether they provided statistically significant benefits in both treatment-refractory and non-treatment-refractory patients, these findings were mostly derived from small, phase II, monocentric trials and were rarely replicated in larger, phase III, international studies. ¹⁴ In addition, pomaglumetad methionil (LY2140023), a mGlu2/3 agonist, provided preliminary evidence of benefits when used as monotherapy in patients with schizophrenia, ¹⁵ however, replication of these findings in a larger trial was unsuccessful. ¹⁶

Evenamide is a new chemical entity devoid of any biological activity at over 150 CNS targets, except for voltage-gated sodium channels (VGSC), exerting its activity in a state-dependent manner with a higher affinity for the inactivated state of these channels. By selectively inhibiting aberrant VGSC activity, evenamide modulates sustained repetitive firing without impairing the normal neuronal excitability and normalizes excessive glutamate release without affecting its basal levels. ¹⁷ Evenamide was tested in a range of doses between 1.25 and 45 mg/kg po in a model of pre-pulse inhibition (PPI) deficit induced by NMDAR antagonists such as MK-801 (three studies), phencyclidine (PCP) (one study), and ketamine (two studies), ¹⁸ which are known to produce a hyper-glutamatergic activity that correlates with induced schizophrenia symptoms in both animals and humans. ¹⁹ Evenamide was effective in reversing the PPI deficits in all models, either as monotherapy or in combination with ineffective doses of clozapine. ²⁷ Unlike conventional antipsychotics, evenamide exerts its effects by normalizing hyperactive hippocampal neurons, thereby addressing the site of dysfunction without significant side effects, ²⁰ and also indirectly impacting ventral tegmental area function. Furthermore, the effects of evenamide persist beyond its presence in the brain, suggesting circuit-level plasticity changes (see figure 1F in Uliana et al. ²⁰ ) that contribute to its extended therapeutic action—an effect not observed with any other antipsychotic in the MAM model. ²⁰

The efficacy of evenamide as an add-on treatment has been recently evaluated in a 52-week, open-label study in TRS patients (Studies 014/015^21,22) and also in a randomized, double-blind, placebo-controlled study in patients with inadequate response to their current SGA (Study 008A ²³ ). Treatment with evenamide 30 mg bid in Study 008A was associated with statistically significant improvement, compared to placebo, in the primary efficacy endpoint (PANSS total score: LS mean difference (SE) (95% confidence interval) = −2.5 (0.90) [−4.3, −0.7], p = 0.006, Cohen’s d effect size = 0.33) and all the other key efficacy measures (Positive and Negative Syndrome Scale (PANSS) positive, PANSS negative, Clinical Global Impression Scale of Change/Severity (CGI-C/S)), in addition to a significant and clinically meaningful improvement achieved on the responder analyses for the PANSS total (⩾20% improvement from baseline) and CGI-C (at least “much improved”). ²³ These findings are supported by the results from Study 014/015 performed in patients with TRS, which indicate exceptional, clinically important, and increasing efficacy of evenamide (7.5–30 mg bid) in treatment-resistant patients observed up to and including 1 year across the key efficacy measures (i.e., PANSS total, CGI-S, mean rating on the CGI-C, as well as responder analyses for all the efficacy measures^21,22).

Evenamide has been well-tolerated at all doses administered to date, with a low incidence of treatment-emergent adverse events, the most common being somnolence, headache, insomnia, and nasopharyngitis; no adverse event with an incidence greater than 5% has ever been reported with evenamide. Similarly, the incidence of adverse events leading to study drug discontinuation and of serious adverse events was extremely low. Moreover, no pattern of clinically significant or notable findings was noted on any of the safety measures adopted in the studies, including EEG and seizure checklists, ECG, vital signs, laboratory tests, and physical, neurological, and standard eye examinations, confirming evenamide’s favorable safety profile.

The objective of the current publication is to present additional post-hoc analyses of randomized studies conducted to date investigating the clinically meaningful benefits of evenamide. These results further support the role of glutamatergic modulation in the treatment of patients with TRS and chronic schizophrenia not benefiting from their antipsychotic treatment.

Methods

This publication is based on post-hoc analyses of data from three randomized clinical studies conducted with evenamide, Studies 014/015 and 008A, in patients with schizophrenia. The methods of the primary analyses are described in the previous publications.^21–23 The CONSORT reporting guidelines have been followed to report results. ²⁴

All studies were approved by an Institutional Review Board/Independent Ethics Committee at each site and conducted in accordance with the Declaration of Helsinki. ²⁵ All patients provided written informed consent before enrollment in each of the trials.

Post hoc analysis 1: Study 014/015—Patients no longer meeting severity criteria for TRS

This analysis was performed to assess whether patients with TRS enrolled in Study 014/015, who must have satisfied pre-specified operational severity criteria at baseline, based on ratings of the PANSS and CGI-S, still met these criteria, or did not, at subsequent timepoints and at the study endpoint. Specifically, the severity criteria used to select TRS patients for the study at baseline were:

1. PANSS total score ⩾ 70,

2. CGI-S of moderately to severely ill (score of 4–6),

3. Total score ⩾ 20 on the combined total of the PANSS symptom items: P1 (Delusions), P2 (Conceptual Disorganization), P3 (Hallucinatory Behavior), P4 (Excitement), P6 (Suspiciousness/Persecution), P7 (Hostility), and G9 (Unusual Thought Content),

4. Score of 4 (moderate) or more on at least 2 of the PANSS core symptoms of psychosis: P2 (Conceptual Disorganization), P3 (Hallucinatory Behavior), P6 (Suspiciousness/Persecution), and G9 (Unusual Thought Content).

This analysis determined the proportion of patients in the modified Intent-to-Treat population (mITT = patients who received at least one dose of study medication, and had a baseline and at least one post-baseline efficacy assessment for the PANSS) who no longer met each of these criteria individually at three key timepoints of the study: week 6, 6-months (week 30), and 1-year (week 52). Moreover, the proportion of patients who no longer met any of the four criteria concomitantly was determined at the same timepoints. The analysis was performed using both the observed cases (OC) and the last observation carried forward (LOCF) approaches.

Post hoc analysis 2: Study 014/015—Patients meeting remission criteria

Two sets of remission criteria were used for this analysis (Lieberman et al., ²⁶ and Andreasen et al., ²⁷ ), which define remission as follows:

1. Lieberman et al. 1993: Patients were considered remitted if they had no rating greater than 3 (mild) on the following SADS-C + PD positive symptom items (Schedule for Affective Disorders and Schizophrenia—Change Version with Psychosis and Disorganization items): suspiciousness, severity of delusions, hallucinations, impaired understandability, bizarre behavior; a CGI-S of maximum “mildly ill” (i.e., score ⩽ 3); and a CGI-C of at least “much improved” (i.e., score ⩽ 2). This level of improvement was required to be maintained for at least 8 weeks;

2. Andreasen et al. 2005: The consensus definition of remission developed by the Remission in Schizophrenia Working Group consists of a maximum score of “mild” (3) on the following eight PANSS items: P1 (Delusions), P2 (Conceptual Disorganization), P3 (Hallucinatory Behavior), N1 (Blunted Affect), N4 (Passive/Apathetic Social Withdrawal), N6 (Lack of Spontaneity and Flow of Conversation), G5 (Mannerism and Posturing), G9 (Unusual Thought Content). This level of response needs to be maintained for at least 6 months.

For the Lieberman et al. criteria, adaptations of SADS-C + PD items were necessary in order to apply them to the current study, in which the PANSS was used. For this purpose, the following PANSS items were selected: P1 (Delusions, corresponding to the SADS item “severity of delusions”), P2 (Conceptual disorganization, replacing the SADS item “impaired understandability”), P3 (Hallucinatory Behavior, corresponding to the SADS item “hallucinations”), P6 (Suspiciousness, corresponding to the SADS item “suspiciousness”), and G5 (Mannerism and Posturing, replacing the SADS item “bizarre behavior”).

The analysis assessed the proportion of patients in the Study 014/015 mITT population who satisfied each of the remission criteria described above at some point during the treatment period of the study.

Post hoc analysis 3: Study 008A—Patients with 1 versus 2 or more failed antipsychotic attempts

This analysis was performed to evaluate the PANSS response of patients participating in Study 008A, that is, patients with chronic schizophrenia not responding adequately to a therapeutic dose of their SGA (including clozapine), who previously failed one versus two or more trials with antipsychotics. The current antipsychotic medication that the patient was taking concomitantly in the study was considered a failed medication, in addition to all prior antipsychotic medications tried in the past, regardless of the reason for discontinuing them. Multiple attempts with the same molecule at different doses were counted as a single failed attempt.

Two strata were identified (i.e., one attempt or two or more attempts of failed antipsychotics) and statistically tested using a stratified Wilcoxon test. Once strata were found significant (p < 0.05), follow-up mixed model analyses were performed for eligible subjects of the Intent-to-Treat (ITT = patients randomized in the study) population. A mixed model linear regression model was fitted with the number of antipsychotic attempts as strata, treatment, and treatment-by-antipsychotic attempts interaction as fixed effects, and baseline PANSS total score value as a covariate.

Post hoc analysis 4: Study 008A and 014/015—Effect of evenamide on social functioning and life engagement

In the absence of a scale that evaluated these domains directly, the effect of evenamide on social functioning and life engagement in Studies 008A and 014/015 was evaluated by analyzing the total score on two subsets of PANSS items that are suggestive of benefit in these two subdomains:

• PANSS items linked to social functioning have been proposed by Purnine et al. ²⁸ and include P3 (Hallucinatory Behavior), P6 (Suspiciousness/Persecution), N2 (Emotional Withdrawal), N4 (Passive/Apathetic Social Withdrawal), N7 (Stereotyped Thinking), and G16 (Active Social Avoidance);

• PANSS items linked to life engagement have been proposed by Ismail et al. ²⁹ and include N1 (Blunted Affect), N2 (Emotional Withdrawal), N3 (Poor Rapport), N4 (Passive/Apathetic Social Withdrawal), N5 (Difficulty in Abstract Thinking), N6 (Lack of Spontaneity and Flow of Conversation), G6 (Depression), G7 (Motor Retardation), G13 (Disturbance of Volition), G15 (Preoccupation), G16 (Active Social Avoidance).

For Study 008A, a mixed model repeated measures (MMRM) linear regression model was fitted with sex, stratification factor (Clozapine Y/N), treatment, pooled site, visit, and treatment-by-visit interaction as fixed effects and baseline value as covariate. The mean change from baseline was calculated at each visit (day 8, day 15, day 22, and day 29) for the evenamide 30 mg bid and placebo groups for the mITT population. In addition, the least square (LS) mean change from baseline was also calculated at the endpoint (day 29), and the LS mean difference between evenamide and placebo with the associated p-value was obtained.

For Study 014/015, the mean change from baseline at each of the three key timepoints in the study (week 6, week 30, and week 52) was calculated for the mITT population.

SAS software (SAS Institute Inc. version 9.4, Cary, NC, USA) was utilized for all the analyses.

Results

Study 014/015—Patients no longer meeting severity criteria for TRS

For each of the severity criteria used to select TRS patients for the study, an increasing proportion of patients treated with evenamide (at doses of 7.5, 15, and 30 mg bid) in Study 014/015 no longer met each criterion at the three timepoints evaluated: week 6, week 30, and week 52 (Figure 1). When the four criteria were considered altogether, 26.3% (OC) of patients improved to such an extent that they no longer satisfied the combined severity criteria already at week 6; the proportion further increased at the subsequent timepoints and was more than doubled after 1 year of treatment, with 55.0% of patients no longer being considered treatment-resistant, according to the severity of their illness at week 52.

Figure 1.

Study 014/015 proportion of patients no longer meeting baseline severity criteria for TRS at key timepoints, mitt population (OC).

Bar graphs show percentages of patients calculated as the number of patients no longer meeting criteria divided by the number of patients observed at each visit.

*P1 (Delusions), P2 (Conceptual Disorganization), P3 (Hallucinatory Behavior), P4 (Excitement), P6 (Suspiciousness/Persecution), P7 (Hostility), G9 (Unusual Thought Content).

^#P2 (Conceptual Disorganization), P3 (Hallucinatory Behavior), P6 (Suspiciousness/Persecution), G9 (Unusual Thought Content).

Similar results were obtained in the analysis using the LOCF approach (Table S1).

Study 014/015—Patients meeting remission criteria

The application of different sets of remission criteria proposed by Lieberman et al. ²⁶ and Andreasen et al. ²⁷ to long-term 1-year data from Study 014/015 yielded similar results, with 27.6% and 25.0% of patients treated with evenamide achieving remission according to each criterion, respectively (Table 1).

Table 1.

Study 014/015 proportion of patients meeting remission criteria defined in the literature, mITT population.

Method	Criteria	Maintenance requirement	n (%) of patients meeting remission criteria N = 156
Lieberman et al., 1993	P1, P2, P3, P6, G5 ⩽ 3 CGI-S maximum “mildly ill” CGI-C at least “much improved”	8 weeks	43 (27.6%)
Andreasen et al., 2005	P1, P2, P3, N1, N4, N6, G5, G9 ⩽ 3	24 weeks	39 (25.0%)

P1 (Delusions), P2 (Conceptual Disorganization), P3 (Hallucinatory Behavior), P6 (Suspiciousness/Persecution), N1 (Blunted Affect), N4 (Passive/Apathetic Social Withdrawal), N6 (Lack of Spontaneity and Flow of Conversation), G5 (Mannerism and Posturing), G9 (Unusual Thought Content).

mITT, modified intent-to-treat.

Study 008A—Patients with 1 versus 2 or more failed antipsychotic attempts

Almost 70% of the patients randomized in Study 008A had failed two or more antipsychotic trials (including the current one), although, as per the inclusion criteria, they did not have a documented diagnosis of TRS. The reasons for failing the prior antipsychotics include lack of response, inadequate response, lost response, and safety or tolerability problems. The remaining one-third of patients had tried only one antipsychotic, that is, the current medication taken during the study, without achieving an adequate response.

In the mixed model follow-up analysis that evaluated the mean change from baseline on the PANSS total score in the two groups, that is, patients who failed one antipsychotic attempt versus those who failed 2 or more, a significantly greater improvement (p < 0.05) was observed in evenamide-treated compared to placebo-treated patients in both groups (Table 2). Although a numerically greater improvement with evenamide was observed in the group of patients with ⩾2 failed attempts compared to the group with only one failed attempt, the placebo response was also larger, resulting in a smaller drug-placebo difference in the former group.

Table 2.

Study 008A: Change from baseline in PANSS total score at day 29 by number of failed antipsychotic attempts, ITT population.

AP attempts	Statistics	Evenamide 30 mg bid N = 127	Placebo N = 154	Evenamide—placebo p-value (95% CI)
1	n (%)	38 (29.9)	50 (32.5)	−4.09 0.0122 (−7.27, −0.90)
	Mean change (SD)	−8.9 (6.45)	−4.8 (6.39)
⩾2	n (%)	89 (70.1)	104 (67.5)	−2.36 0.0311 (−4.50, −0.22)
	Mean change (SD)	−10.5 (8.31)	−8.3 (7.71)

A mixed model linear regression model is fitted with No. AP attempts as Strata (AP_attempts 1/⩾2), treatment, and treatment-by-AP_attempts interaction as fixed effects, and baseline PANSS total score value as a covariate.

AP, antipsychotic; ITT, Intent-to-treat; N, total number of subjects in the ITT population at Day 29; n, number of patients; SD, standard deviation.

Study 008A and 014/015-Effect of evenamide on social functioning and life engagement

Study 008A

A gradually decreasing trend was observed in the social functioning (Figure 2(a)) and life engagement (Figure 2(b)) subdomains total scores throughout the 4-week treatment period in Study 008A, with a progressive separation of the evenamide group from the placebo group starting from the second (social functioning) and third (life engagement) week. At the final visit (day 29), add-on treatment with evenamide 30 mg bid resulted in a statistically significant (p = 0.012) greater improvement, compared to placebo, in the PANSS subdomain of social functioning, as well as a greater improvement compared to placebo on the life engagement subdomain, although the drug-placebo difference was borderline significant (p = 0.055; Table 3).

Figure 2.

Study 008A mean change from baseline at each visit on PANSS subdomains of social functioning and life engagement, mITT population. (a) Social functioning subdomain: P3 (Hallucinatory Behavior), P6 (Suspiciousness/Persecution), N2 (Emotional Withdrawal), N4 (Passive/Apathetic Social Withdrawal), N7 (Stereotyped Thinking), G16 (Active Social Avoidance). (b) Life engagement subdomain: N1 (Blunted Affect), N2 (Emotional Withdrawal), N3 (Poor Rapport), N4 (Passive/Apathetic Social Withdrawal), N5 (Difficulty in Abstract Thinking), N6 (Lack of Spontaneity and Flow of Conversation), G6 (Depression), G7 (Motor Retardation), G13 (Disturbance of Volition), G15 (Preoccupation), G16 (Active Social Avoidance).

Graphs show the mean change from baseline at each visit. Error bars represent standard error.

Table 3.

Study 008A mean change from baseline to endpoint (day 29) on PANSS subdomains of social functioning and life engagement, mITT population.

Subdomain	Statistic	Evenamide 30 mg bid N = 131	Placebo N = 156	Evenamide—placebo p-value (95% CI)
Social functioning	LS Mean change (SE)	−3.1 (0.23)	−2.4 (0.22)	−0.7 (0.28) 0.012 (−1.3, −0.2)
Life engagement	LS Mean change (SE)	−2.6 (0.31)	−1.8 (0.29)	−0.7 (0.38) 0.055 (−1.5, 0.0)

A MMRM linear regression model is fitted with sex, stratification factor (clozapine Y/N), treatment, pooled site, visit, and treatment-by-visit interaction as fixed effects and baseline value as covariate. Social functioning subscale = sum (P3 P6 N2 N4 N7 G16). Life engagement subscale = sum (N1 N2 N3 N4 N5 N6 G6 G7 G13 G15 G16).

CI, confidence interval; LS, least square; mITT, modified intent-to-treat; MMRM, mixed model repeated measures; N, total number of subjects in the mITT population; SE, standard error.

Study 014/015

An improvement of −2.6 points from baseline to week 6 was observed on the social functioning subdomain of the PANSS in patients with TRS treated with evenamide add-on in Study 014/015; such benefit further increased at subsequent timepoints up to 1 year (Figure 3(a)). Similarly, an improvement of −2.4 points at week 6, which was further increased up to 1 year, was noted on the life engagement subdomain (Figure 3(b)).

Figure 3.

Study 014/015 mean change from baseline at three key timepoints on PANSS subdomains of social functioning and life engagement, mITT population. (a) Social functioning subdomain: P3 (Hallucinatory Behavior), P6 (Suspiciousness/Persecution), N2 (Emotional Withdrawal), N4 (Passive/Apathetic Social Withdrawal), N7 (Stereotyped Thinking), G16 (Active Social Avoidance). (b) Life engagement subdomain: N1 (Blunted Affect), N2 (Emotional Withdrawal), N3 (Poor Rapport), N4 (Passive/Apathetic Social Withdrawal), N5 (Difficulty in Abstract Thinking), N6 (Lack of Spontaneity and Flow of Conversation), G6 (Depression), G7 (Motor Retardation), G13 (Disturbance of Volition), G15 (Preoccupation), G16 (Active Social Avoidance).

Graphs show the mean change from baseline at key timepoints. Error bars represent standard error.

Discussion

Dopaminergic dysregulation, long regarded as the principal abnormality in schizophrenia, may actually result from abnormalities in the glutamatergic system, which have been noted especially in the hippocampus. ¹⁰ Evenamide, by modulating glutamate in the hippocampus (site of dysfunction), has the potential to treat the symptoms of schizophrenia more broadly, addressing also those poorly managed by the currently available antipsychotics. Indeed, recent experiments in the MAM animal model offer new evidence of evenamide’s efficacy in ameliorating negative symptoms and cognitive deficits, and inducing neuroplasticity. ²⁰

Previous attempts to prove the benefits of add-on glutamatergic modulation have not been successful in other clinical trials. ²¹ A potential explanation for these failures, where a new compound was added to an existing antipsychotic, might be that many patients who are poor responders may actually be noncompliant with the medication: failure to measure plasma levels during screening and baseline may lead to inclusion of patients who are non-compliant rather than poor responders (see Anand et al., ²³ ). If randomized, the data associated with these subjects may increase the placebo response at the endpoint, in particular if these subjects’ compliance with their background antipsychotic dosing improves because of their participation in the clinical trial.

Clinical trials conducted to date with evenamide have demonstrated its efficacy and safety as an add-on treatment to antipsychotics.^21–23 The finding that evenamide add-on was associated with statistically significant improvement on the primary and secondary efficacy measures establishes the efficacy of evenamide as an antipsychotic. Furthermore, the statistically significant and clinically meaningful improvement (⩾20% improvement on the PANSS, “much improved” rating on the CGI-C) in patients with inadequate response to their antipsychotic medication indicated the therapeutic benefit of evenamide as an add-on treatment. Although the clozapine group in Study 008A accounted only for 15% of the total study population, the benefit observed in poorly responding patients who were taking evenamide in addition to clozapine was very similar to that noted in patients taking evenamide in combination with other SGAs.

Notably, the results from Study 008A indicated a statistically significant improvement not only of positive, but also of negative symptoms (PANSS Negative Subscale), similar to the results in the MAM model and supporting the hypothesis that by targeting the site of dysfunction, that is, the overactive hippocampus, evenamide can effectively control negative symptoms, in addition to positive psychotic symptoms. This is further supported by the post-hoc analysis presented in this publication, which showed a positive and statistically significant effect of evenamide on a selection of PANSS items that are suggestive of benefit on the patients’ functioning in social life. A similar analysis on the impact on life engagement revealed a greater improvement on this PANSS subdomain for patients receiving evenamide added to their SGA compared to those being treated solely with their background antipsychotic. The authors recognize that benefits on life engagement, social functioning, and overall well-being need to be evaluated in longer placebo-controlled trials, and that these preliminary findings would need to be replicated in future trials with validated scales, for example, Personal and Social Performance, Quality of Life Enjoyment and Satisfaction Questionnaire scales. However, the trend observed over the 4-week period of Study 008A is promising, as it seems to suggest a continued positive impact of evenamide on social functioning and life engagement, which may further increase over a longer timeframe. Results in the 014/015 study demonstrating continuing improvement in the Level of Functioning scale further indicate the long-term (1-year) benefits of evenamide on functioning. ³²

The possibility of observing the effect of evenamide treatment over 1-year in Study 014/015 allowed exploration of whether patients achieved remission, according to two definitions requiring the maintenance of a certain level of symptom severity for a period of time (8 weeks or 24 weeks). Although the proportion of patients treated with evenamide in Study 014/015 who achieved remission could not be compared to a placebo arm, the finding that ∼25% met the two different definitions of remission described in the literature is promising. This result is especially meaningful in the TRS population, where substantial improvement is seldom observed. Achievement of remission is hard to evaluate in clinical trials due to the requirement of observing symptom remission over a pre-defined timeframe. However, remission rates were evaluated in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study, ³⁰ which involved a study population with PANSS and CGI-S baseline scores and other baseline characteristics similar to those of Study 014/015. In the CATIE study, a remission rate of 11.7% was observed ³¹ according to Andreasen et al. criteria. ²⁷ This proportion, despite being observed in non-TRS patients, is lower compared to the one noted in TRS patients in Study 014/015. It is even more remarkable that, in Study 014/015, 50% of patients treated with evenamide experienced such a level of improvement that, at the end of the study, they could no longer be defined as treatment-resistant based on the global severity of illness and the severity of their core symptoms.

Although Study 008A was not conducted in patients with TRS, a post-hoc analysis revealed that a subgroup of the global population could be regarded as TRS-like based on the number of failed attempts of treatment with antipsychotics. Patients were stratified by 1 versus 2 or more failed attempts (including the current medication to which they were not responding), according to Kane et al., ² criteria, irrespective of whether the failure was due to lack of efficacy or tolerability, or both. Results of the PANSS mean change from baseline indicated statistically significant efficacy in both groups, suggesting that evenamide might be effective, compared to the standard of care, in treatment-resistant patients.

Glutamate modulation through evenamide was associated with a favorable safety profile devoid of typical adverse events of 5-HT2/D2-blocking agents.^21–23 All clinical trials conducted with evenamide to date confirmed the absence of any pattern of safety abnormalities on the multiple safety measures used (e.g., vitals signs, laboratory tests, ECG, and EEG), the low incidence of adverse events and serious adverse events, and extremely low adverse dropout rate.^21–23 This favorable safety profile is likely shared by other glutamate-modulating antipsychotics, as underlined by Goh et al. ¹⁴ and Downing et al. ¹⁶

Strengths

Study 014/015 was 1-year in duration, allowing the evaluation of the long-term effects of evenamide. Study 008A was a randomized, double-blind, placebo-controlled trial that evaluated the improvement on evenamide 30 mg bid as add-on compared to placebo in patients with chronic schizophrenia not responding to their current SGA, and was conducted in 11 countries, 45 sites, in Latin America, Asia, and Europe.

Limitations

Study 014/015 was an open-label trial that lacked a control group and involved patients predominantly from a single country; in addition, patients enrolled, despite being diagnosed with TRS, were not allowed to take clozapine as background treatment. The short duration of Study 008A (4 weeks) prevented the observation of a further improvement and maintenance of the benefit over a longer timeframe.

Path forward

A phase III, international program including two randomized, double-blind, placebo-controlled trials (Study 023 ENIGMA-TRS 1 and Study 022 ENIGMA-TRS 2) has been initiated to assess the efficacy and safety of two fixed doses of evenamide (15 and 30 mg bid) in patients with documented TRS (according to TRRIP criteria ³² ). These studies have been designed to address the major limitations of the two antecedent studies presented in the above paragraph. Both ENIGMA-TRS studies will have the primary efficacy endpoint set at 12 weeks. In the ENIGMA-TRS 1 study, patients will receive treatment in a double-blind, placebo-controlled fashion for a total period of 1 year, with long-term efficacy endpoints at weeks 26 and 52. Additionally, these studies present other unique features: the prospective confirmation of treatment resistance during the screening period as per TRRIP criteria, ³² the measurement of background antipsychotic plasma levels to exclude cases of noncompliance rather than non-response, and the involvement of an international eligibility assessment committee that will confirm the diagnosis of TRS at screening. Positive results from the currently ongoing ENIGMA-TRS program would confirm the benefit of glutamatergic modulation by evenamide as an add-on to Standard of Care in patients living with TRS.

Conclusion

The results from the presented analyses provide strong evidence that the benefit obtained with evenamide as an add-on treatment to antipsychotic(s) is clinically meaningful and that modulation of aberrant glutamatergic activity in the hippocampus can address unmet needs in patients with inadequate or no response to antipsychotics.