Recent and on-going trials for the treatment of levodopa-induced dyskinesia: a review of the clinical trial databases

ABSTRACT

L-3,4-dihydroxyphenylalanine (Levodopa)-induced dyskinesia remains a condition for which there are few therapeutic options available. Fortunately, the past 5 years have seen the completion of several clinical trials, some of which yielded positive and encouraging results. Here, we performed a review of the clinical trials that were completed or for which outcomes were disclosed within the past 5 years. Promising results were obtained in Phase II trials with the serotonin type 1A (5-HT_1A) agonist befiradol, the dopamine D₃ receptor antagonist mesdopetam and the phosphodiesterase inhibitor CPL500036. In contrast, the metabotropic glutamate 4 (mGlu₄) receptor negative allosteric modulator foliglurax and JM-010 (a combination of the 5-HT_1A partial agonist buspirone and the and the 5-HT type 1B and 1D [5-HT_1B/1D] agonist zolmitriptan) did not meet their endpoints in Phase II studies. Lastly, robot-assisted Repetitive Transcranial Magnetic Stimulation (rTMS) of the pre-supplementary motor area may be a promising non-pharmacological approach to alleviate dyskinesia.

1. Introduction

L-3,4-dihydroxyphenylalanine (levodopa)-induced dyskinesia is a common side-effect from long-term therapy with levodopa in patients with Parkinson’s disease (PD), affecting up to 50% of patients in the 5 years after initiating therapy [1] and almost 95% after 15 years of administration [2]. Dyskinesia is characterized by abnormal, involuntary movements such as chorea, athetosis, dystonia, and myoclonus which can involve the arms, the legs, or the entire body [3]. The pathophysiology of dyskinesia is complex to say the least, but in simplified terms, may stem from an interplay between the loss of dopamine-producing neurons in the substantia nigra (presynaptic neurons losing the ability to store and release dopamine) and the ensuing hypersensitivity of dopamine receptors in the striatum [4,5]. Dyskinesia is related to the dosing of levodopa, classified into peak-dose dyskinesia, diphasic dyskinesia, and off-period dystonia (or wearing-off dyskinesia) [6]. Peak-dose dyskinesia, the most common type, occurs during the anti-parkinsonian effect of levodopa, i.e. ‘on-time’, which starts from 30 to 90 min after taking the dose, when the plasma levels are the highest. Diphasic dyskinesia two phases occur after taking the initial dose during the switching on process and when the effect starts to wear-off. Off-period dystonia occurs when the plasma levels are the lowest, most commonly in the morning before the first dose [7]. Several studies have dedicated efforts to understand modifiable (e.g. dosage and treatment duration), as well as non-modifiable risk factors (e.g. female sex and younger age) to gain a better understanding of this complex process [8,9]. Dyskinesia hinders the quality of life of patients [10,11]. The treatment of dyskinesia is a delicate balance, with clinicians often finding themselves in a difficult position, where decreasing levodopa dose could alleviate some of the patient’s dyskinesia, but at the expense of reemergence of parkinsonism [12]. By hindering optimization of levodopa treatment, dyskinesia may result in some patients being under-treated.

A thorough discussion on the optimal management of dyskinesia is complex and beyond the scope of this article, as several factors such as disease duration and severity, type of dyskinesia, and medication regimen must be factored in. Briefly, peak-dose dyskinesia can be managed with smaller, more frequent doses of immediate-acting levodopa formulations, and by reducing or discontinuing adjunct therapies such as monoamine oxidase (MAO) type B (MAO-B) or catechol-O-methyltransferase (COMT) inhibitors if used, in addition to adding amantadine [13,14]. Diphasic dyskinesia may be addressed by keeping levodopa plasma levels beyond the dyskinetic threshold, in practice by increasing the dopaminergic load, although this may lead to peak dose dyskinesia [15]. However, practically speaking, treatment options for levodopa-induced dyskinesia are limited, with only amantadine extended-release (ADS-5102) being approved by the United States Food and Drug Administration (FDA) [16], following the successful completion of two Phase III trials [17,18]. However, amantadine extended-release is not efficacious in all patients, and cognitive and psychiatric adverse effects may at times occur, in addition to edema and constipation [19]. In addition, tolerance to the anti-dyskinetic effect of amantadine, at least the immediate-release form, was previously reported [20]. Whereas the atypical antipsychotic clozapine is regarded as efficacious by the International Parkinson and Movement Disorder Society [14], the risk of agranulocytosis limits its use in practice [21]. The discovery and development of new drugs for the treatment of dyskinesia therefore remains an unmet need. Here, we seek to summarize and discuss clinical trials with anti-dyskinetic molecules or non-pharmacological approaches that have been conducted or for which results were disclosed in the past 5 years.

2. Methodology

The majority of the trials were extracted from ClinicalTrials.gov database and European Union Drug Regulating Authorities Clinical Trials (EudraCT) database. In both databases, “levodopa-induced dyskinesia” was used as a search term to yield the results of trials. The clinical trials selected for inclusion in this article are the ones that are on-going or that were completed, terminated or published in the peer-reviewed literature during the past 5 years. Studies initiated in the past 5 years whose status is “unknown” are also included. However, studies with unpublished/undisclosed or negative results will not be discussed as deeply as those with positive results and compounds moving forward in the development pipeline.

3. Results

The molecules presented here have been categorized according to their primary mechanism of action. The drug listed, along with their mechanism(s) of action, a summary of the studies cited and references, are presented in Table 1.

Table 1.

Summary of recent clinical trials that evaluated potential treatments for dyskinesia.

drug/intervention (Phase, N)	Mechanism of action	Main outcome(s) and measure(s)	Design	Status	Highlighted results	References
serotonergic drugs
befiradol Phase IIa, N = 27	5-HT1A full agonist	safety and tolerability of befiradol, titrated up to a maximum of 2 mg/day efficacy at reducing dyskinesia (UDysRS, PD home dyskinesia diary) effect on parkinsonism (UPDRS part III and combined parts I-IV)	randomised, parallel, quadruple-blind, placebo-controlled	completed	safety and tolerability endpoints met	NCT05148884 [26]
buspirone Phase III, N = 99	5-HT1A partial agonist	between-group comparison of changes in UDysRS scores comparison of MDS-UPDRS parts III-IV	randomised, parallel, double-blind, placebo-controlled	completed	no result posted	NCT02617017
buspirone Phase I, N = 6		force-plate measurements of dyskinesia UDysRS total score adverse event monitoring/frequency	randomised, cross-over, quadruple-blind, placebo-controlled	terminated (low enrollment)	no result posted	NCT02589340
DSP-9632P Phase I, N = 7	5-HT1A partial agonist	evaluate the amount of striatal dopamine release derived from levodopa following single transdermal administration of DSP-9632P using [11C]-raclopride PET UDysRS MDS-UPDRS	randomised, double-blind, placebo-controlled, crossover study	completed	no results posted	NCT05435729
JM-010 Phase II, N = 30	5-HT1A partial agonist and 5-HT1B/1D agonist	change in dyskinesia severity as assessed by the AIMS UPDRS part III	randomised, double-blind, placebo-controlled, 2-way cross-over	completed	reduction of dyskinesia without deleterious effect on parkinsonism	NCT02439203 [31]
glutamatergic drugs
AV-101 Phase II, N = 20	antagonist of the glycine-binding site on NMDA receptors	UDysRS part III AUC and peak score MDS-UPDRS part III	randomised, cross-over, double-blind, placebo-controlled	unknown	no result posted	NCT04147949
dipraglurant Phase IIb/III, N = 140	mGlu5 negative allosteric modulator	change from baseline on the UDysRS total score	randomised, placebo-controlled, double-blind	terminated (low enrollment due to the pandemic)	no results posted	NCT04857359
dipraglurant Phase II/III, N = 140		long-term safety and tolerability evaluate the continued use of dipraglurant from baseline score on the MDS-UPDRS	single group, open-label	terminated (low enrollment due to the pandemic)	no result posted	NCT05116813
foliglurax Phase II, N = undisclosed	mGlu4 positive allosteric modulator	change from baseline in UDysRS total score	multi-centre, randomised, double-blind, placebo-controlled, parallel-arm	withdrawn (business reasons)	no results posted	NCT03331848
foliglurax Phase II, N = 157		change from baseline to end of treatment period in the daily awake off time (Hauser diary entries)	randomised, parallel, triple-blind, placebo-controlled	completed	no change from baseline to day 28 in off time (primary endpoint) or dyskinesia (secondary endpoint) compared to placebo	NCT03162874 [42]
ketamine Phase II, N = 30	NMDA antagonist	change in UDysRS total score	randomised, double-blind, prospective, active placebo- controlled	suspended	no results posted	NCT04912115
dopaminergic drugs
mesdopetam Phase IIb, N = 156	D3 dopamine receptor antagonist	change in duration of on-time without troublesome dyskinesia (24-h patient home diaries) UDysRS	randomised, double-blind, placebo-controlled study	completed	primary endpoint not met, but significant effect on UDysRS	NCT04435431
phosphodiesterase drugs
CPL500036 Phase II, N = 108	PDE10 inhibitor	change from baseline in total UDysRS score at week 4 compared to placebo	randomised, parallel, double-blind, placebo-controlled	completed	all endpoints were met	NCT05297201
lenrispodun Phase II, N = 132	PDE1B inhibitor	on-time without troublesome dyskinesia (Hauser diary)	randomised, parallel, quadruple-blind, placebo-controlled	on-going	no result posted	NCT05766813
sigma receptor drug
pridopidine Phase II, N = 23	sigma-1 receptor agonist	change from baseline in dyskinesia (UDysRS)	randomised, parallel, quadruple-blind, placebo-controlled	terminated (pandemic)	no results posted	NCT03922711
non-pharmacological interventions
robot-assisted repetitive transcranial magnetic stimulation of the pre-supplementory motor area Phase N/A, N = 17	modulation of cortical excitability with magnetic fields	levodopa-induced change in task-related regional neural activity as indexed by the blood oxygen level dependent signal parkinsonism (UPDRS part III) time to dyskinesia onset (UDysRS)	randomised, cross-over, single-blind, sham-controlled	completed	reduction of dyskinesia severity, delay of dyskinesia onset, no effect on levodopa anti-parkinsonian action	NCT03354455 [56]
transcranial direct current stimulation of the primary motor cortex Phase N/A, N = 32	direct application of weak, low, current to the head with at least one cephalic electrode	change in the total score of UDysRS part III	randomised, parallel, single-blind, sham-controlled	recruiting	no result posted	NCT05752240
transcranial static magnetic field stimulation Phase II, N = 50	modulation of underlying brain areas via placement of a magnet over the scalp	change from baseline of the maximal dyskinesia severity within 90 min after levodopa intake (UDysRS) effect on parkinsonism (MDS-UDPRS part III)	randomised, parallel, double-blind, sham-controlled	completed	no difference between active treatment and sham group	NCT02657681 [57]

5-HT: serotonin; 5-HT1A: serotonin type 1A receptor; 5-HT1B/1D: serotonin type 1B/1D receptors; AIMS: Abnormal Involuntary Movement Scale; AUC: area under the curve; MDS: Movement Disorder Society; mGlu4: metabotropic glutamate type 4 receptor; mGlu5: metabotropic glutamate type 5 receptor; PD: Parkinson’s disease; PDE: phosphodiesterase; PET: positron emission tomography; UDysRS: Unified Dyskinesia Rating Scale; UPDRS: Unified Parkinson’s Disease Rating Scale.

3.1. Serotonergic drugs

The drugs reviewed here act as agonists at serotonin (5-HT) type 1A (5-HT_1A) and/or 5-HT type 1B/1D (5-HT_1B/1D) receptors. The mechanisms whereby 5-HT_1A agonism might alleviate dyskinesia are complex, but likely involve modulation of dopamine and glutamate in the striatum. Thus, in studies performed in the 6-hydroxydopamine (6-OHDA)-lesioned rat, activation of 5-HT_1A receptors reduced dopamine [22] and glutamate [23] release in the striatum. In comparison, the mechanisms underlying a potential anti-dyskinetic effect of 5-HT_1B/_1D agonists remain largely unexplored. However, the activation of 5-HT_1B receptors does not seem to modulate striatal dopamine release [22] but may interact with dopamine D₁ receptors via the p11 adaptor protein [24], which may play a role in reduction of dyskinesia.

3.1.1. Befiradol

Befiradol (NLX-112, F13640), is a highly selective full 5-HT_1A agonist [25]. NCT05148884 evaluated the safety, tolerability, and preliminary efficacy of befiradol in PD patients with moderate to severe dyskinesia. The study was a randomized, double-blind, placebo-controlled, two-arm, Phase IIa trial with 27 patients. Befiradol was up-titrated to either 2 mg/day (1 mg twice daily) or to the highest tolerated dose less than 2 mg/day over 4 weeks, maintained at the tolerated dose for an additional 2 weeks, and then down-titrated over 2 weeks. The primary outcomes evaluated the safety and tolerability of befiradol compared to placebo during 8 weeks of daily treatment and determined its adverse event profile, notably changes from baseline in electrocardiogram (ECG), vital signs, physical examination, suicidal ideation, and behavior. Secondary endpoints included measure of dyskinesia severity using the Unified Dyskinesia Rating Scale (UDysRS), duration of on-time and quality of on-time via patient diaries, Unified Parkinson’s Disease Rating Scale (UPDRS) part III and UPDRS parts I–IV combined, dyskinesia assessed via a wearable device, as well as Clinical Global Impression (CGI). Befiradol met the primary endpoint of safety and tolerability. Befiradol also reduced dyskinesia compared to baseline. Compared to placebo, befiradol reduced UDysRS part III score. Befiradol also reduced parkinsonism [26].

3.1.2. Buspirone

Buspirone is a partial 5-HT_1A agonist [27] that has been extensively studied for the treatment of dyskinesia in PD, with inconsistent results depending on the dose and a tendency to hinder the anti-parkinsonian effect of levodopa at higher dose (reviewed in [28]). Two studies have been completed/terminated in the past 5 years. NCT02617017, was a multi-center, international, randomized, placebo-controlled, double-blind Phase III trial, which sought to validate the serotonergic hypothesis of dyskinesia in PD. The primary outcome measured the between-group comparison of changes in UDysRS scores between baseline and week 12. Results have yet to be posted.

NCT02589340 was a Phase I trial that evaluated the efficacy of buspirone in combination with amantadine at reducing dyskinesia in PD patients. It consisted of a randomized, placebo-controlled, double-blind, two-period cross-over study. The study was terminated and results have not yet been posted.

3.1.3. DSP-9632P

NCT05435729 was conducted to evaluate the dopamine release derived from levodopa in the brain after administration of DSP-9632P (transdermal tandospirone, a partial 5-HT_1A agonist [29]), as well as to evaluate its safety and tolerability in PD patients with dyskinesia. The study has been completed but no results are posted. It should be noted that, in the past, tandospirone was administered to PD patients and tended to interfere with the anti-parkinsonian action of levodopa [30].

3.1.4. JM-010

NCT02439203 evaluated the effect of JM-010, a combination of buspirone and the 5-HT type 1B and 1D (5-HT_1B/5-HT_1D) agonist zolmitriptan [31]. It was a Phase II, randomized, double-blind, placebo-controlled, 2-way crossover study in PD patients with moderate to severe dyskinesia. It aimed to assess the efficacy, safety, tolerability, and pharmacokinetics of JM-010. The primary efficacy outcome measure was investigator-rated change in dyskinesia severity as assessed by the Abnormal Involuntary Movements Scale (AIMS) for 7 days. JM-010 significantly alleviated dyskinesia. JM-010 was well tolerated and no serious adverse events were reported.

NCT04377945 was a double-blind, placebo-controlled, randomized, multi-center Phase II trial that assessed the effects of various doses of JM-010 and its individual components. The study primarily sought to measure the effect of JM-010 on dyskinesia severity with the UDysRS after 12 weeks. The study was terminated for business reasons, following failure of NCT03956979, another Phase II trial that investigated the effect of JM-010 on dyskinesia over 12 weeks. Although the results of this last study were not published in the peer-reviewed literature, the sponsor disclosed the lack of efficacy compared to placebo regarding the primary endpoint.

3.2. Glutamatergic drugs

3.2.1. AV-101

AV-101 (L-4-chlorokynurenine) is an investigational prodrug of 7-chlorokynurenic acid, which antagonizes the glycine-binding site on the N-methyl-D-aspartate (NMDA) glutamatergic receptor [32]. Glycine and D-serine are co-agonists at the NMDA receptor and are required for its activation [33]. As glutamatergic activity in the striatum is regarded as a core tenet of the pathophysiology of levodopa-induced dyskinesia [34], a therapy that would diminish the activity of NMDA receptors could in theory reduce dyskinesia severity.

NCT04147949 was designed to evaluate the efficacy and safety of AV-101 in dyskinetic PD patients and consisted of a randomized, double-blind, placebo-controlled, cross-over, proof-of-concept Phase II study. The primary outcome was to assess dyskinesia and PD motor symptoms of the subjects using UDysRS Part III peak score and Part III area under the curve for 14 days. The treatments, 1,440 mg of AV-101 capsule or placebo, were intended to be administered orally twice daily for 14 days. The status of the study is unknown, and no development has been reported recently. Of note, AV-101 now seems to be under development for major depressive disorder and neuropathic pain, suggesting that it may not be moved forward for the treatment of levodopa-induced dyskinesia.

3.2.2. Dipraglurant

Antagonism of metabotropic glutamate (mGlu) type 5 (mGlu₅) receptors reduces the excessive glutamatergic transmission associated with the dyskinetic state [35]. In addition, mGlu₅ receptors were shown to form heteromers with dopamine D₁ receptors, which mediates non-canonical dopamine signaling in the parkinsonian state [36]. Antagonizing mGlu₅ receptors can therefore also regulate dopaminergic transmission.

After a successful Phase II trial [37], the mGlu₅ receptor negative allosteric modulator dipraglurant (ADX48621) [38] entered NCT04857359, a Phase IIb/III trial that was meant to be followed by an open-label extension phase, NCT05116813. Both studies were terminated due to slow recruitment and challenges faced during the pandemic. Dipraglurant has since been repurposed for brain injury recovery post-stroke and traumatic brain injury and is ready to start a Phase II study for this indication [39].

3.2.3. Foliglurax

Foliglurax (PXT002331) is a positive allosteric modulator of mGlu type 4 (mGlu₄) receptors [40]. The precise mechanism whereby mGlu₄ activation reduces dyskinesia is uncertain but may involve regulation of glutamate release at the cortico-striatal synapse [41].

Our research identified two clinical trials that aimed to evaluate the efficacy, safety, and tolerability of foliglurax at reducing dyskinesia in subjects with PD. NCT03331848 sought to assess the efficacy, safety, and tolerability of 8-week oral treatment with foliglurax in reducing dyskinesia and motor fluctuations in patients with PD. The study was a multi-center, randomized, double-blind, placebo-controlled, parallel-arm, proof-of-concept Phase IIa trial with estimated 86 patients with PD experiencing motor complications of levodopa. The study arms were placebo oral capsules or foliglurax 20 mg twice daily for 8 weeks. The intended primary outcome was a change from baseline in UDysRS total score in subjects with PD experiencing dyskinesia over 26 days. The study, however, was withdrawn before recruitment for business reasons.

NCT03162874 was a randomized, double-blind, placebo-controlled, parallel-arm, proof-of-concept Phase IIa trial with 157 enrolled PD patients with motor fluctuations and dyskinesia. Patients were administered foliglurax orally twice daily at the doses of 10 mg, 30 mg, or placebo for 28 days [42]. The primary outcome measured in this study was the change from baseline to end of treatment period in the daily awake off-time based on subject Hauser diary entries for 28 days. The published results showed apparent dose-dependent decreases in daily awake off-time duration. The improvement from baseline to day 28 in off-time (primary endpoint) and dyskinesia (secondary endpoint) did not reach significance when compared to placebo for either dose of foliglurax. Reasons for failure remain speculative, but may entail inadequate brain occupancy by the compound, as criticism was raised about the radioligand that was utilized to measure central target engagement [43]. As for its safety profile, it was generally safe and no relevant safety signals were present. The development of foliglurax has since been terminated.

3.2.4. Ketamine

Ketamine is a dissociative anesthetic that binds to several receptors but is generally considered to primarily antagonize NMDA receptors [44], and this dampening of glutamatergic transmission may have led to its testing as an anti-dyskinetic agent [45]. A Phase II clinical trial was designed to assess the efficacy of low-dose intra-venous ketamine infusions in reducing dyskinesia in PD patients (NCT04912115). It was a double-blind, randomized, parallel-group design with two arms. Participants were randomly assigned to receive either the investigational drug, ketamine, or an active placebo, midazolam, which induces mild sedation to prevent bias due to the sedative effects of ketamine. The primary outcome measure changes in the UDysRS total score from baseline to week 8. The study was suspended and no results have been reported.

3.3. Dopaminergic drugs

3.3.1. Mesdopetam

Mesdopetam (IRL790, IPN60170) is a dopamine D₃ antagonist built on an agonist chemical structure with 10-fold greater affinity for D₃ over D₂ receptors [46]. NCT04435431 was a Phase IIb randomized, double-blind, and placebo-controlled trial involving 156 participants who were assigned to one of four groups, receiving either 2.5 mg, 5.0 mg, or 7.5 mg of mesdopetam twice daily, or placebo. The primary endpoint was the change in the duration of daily on-time without troublesome dyskinesia, as recorded in 24-h patient diaries. Results showed that mesdopetam did not lead to a statistically significant improvement in this primary outcome compared to placebo. However, mesdopetam did show significant anti-dyskinetic effects on the secondary endpoint, the UDysRS, particularly at the highest dose of 7.5 mg, with positive results observed at four (p = 0.045), eight (p = 0.004), and 12 (p = 0.026) weeks. These effects were achieved without impairing motor function, which was a critical aspect of the trial. The safety profile of mesdopetam was comparable to that of placebo, with adverse events occurring in 56.9% of the mesdopetam-treated subjects versus 46.2% in the placebo group. The most common adverse events were related to nervous system disorders, although these were slightly less frequent in the mesdopetam group than in the placebo group. Additionally, a small percentage of patients treated with mesdopetam reported decreased mobility during the first month of treatment, but this issue did not persist in the subsequent months [47]. A Phase III trial assessing the anti-dyskinetic effect of mesdopetam 7.5 mg twice daily is now being planned.

3.4. Phosphodiesterase drugs

3.4.1. CPL500036

CPL500036 (CPL’36) is a potent and selective phosphodiesterase 10 (PDE10) inhibitor [48]. PDE10 is abundant within the striatum, where it regulates intra-cellular levels of cyclic nucleotides and modulates the activity of striatal projection neurons [49]. NCT05297201 aimed to determine potential anti-dyskinetic, safety, tolerability, and pharmacokinetic properties of low and high doses of CPL500036 in PD patients. This double-blind, randomized, placebo-controlled, parallel-group, dose ranging Phase II trial compared two doses of the drug with placebo and sought to recruit 108 patients. The primary outcome of the study was to assess the effect of low and high doses of CPL500036 compared to placebo, on the reduction of dyskinesia based on changes from baseline in total UDysRS score at week 4. Patients were administered 20 mg or 40 mg of CPL500036, or placebo once daily for a 28-day treatment period. The study was completed earlier this year, and the company announced that all endpoints were met, with results exceeding expectations [50].

3.4.2. Lenrispodun

Lenrispodun (ITI-214) is a potent phosphodiesterase 1B (PDE1B) inhibitor [51]. PDE1B is encountered in the striatum [52], where it participates in the regulation of dopaminergic transmission [53]. NCT05766813 is an active Phase II trial designed to evaluate the efficacy and safety of lenrispodun as an adjunctive therapy in the treatment of motor fluctuations in patients with PD, in which dyskinesia is a secondary endpoint. This is a multi-center, randomized, double-blind, placebo-controlled study. The trial involves two parallel groups: one receiving lenrispodun 30 mg daily and the other receiving placebo. The study aims to determine whether lenrispodun can extend periods of good motor function without troublesome dyskinesia, as recorded in patient diaries (Hauser Diary). The trial began in March 2023 and is expected to be completed by October 2025.

3.5. Sigma receptor drug

3.5.1. Pridopidine

Pridopidine (ACR16, PL-101, TV-7820) is a sigma-1 receptor agonist [54] that also exhibits affinity for several other pharmacological targets, notably 5-HT_1A, 5-HT type 2A (5-HT_2A) and dopamine D2 receptors [55]. The precise mechanism of action of activation of sigma-1 receptors in dyskinesia is unknown, and it was suggested that affinity for the targets mentioned above might play a role in a potential anti-dyskinetic effect [55]. NCT03922711 was a multi-center, randomized, three-arm, parallel-group, double-blind, placebo-controlled Phase II trial. The study aimed to measure changes in dyskinesia using the UDysRS over a planned treatment period of 16 weeks. Unfortunately, the study was terminated early due to the COVID-19 pandemic, which significantly impacted its ability to reach its primary and secondary endpoints as originally planned. At the time of termination, only 23 participants had been enrolled. The primary endpoint analysis was adjusted to consider data available up to Week 8/10 for those participants who were on the study drug at that time. Pridopidine is now under regulatory review by the European Medicines Agency for potential approval for Huntington’s disease and remains under development for and amyotrophic lateral sclerosis.

3.6. Non-pharmacological interventions

3.6.1. Repetitive transcranial magnetic stimulation

NCT03354455 assessed the effect of robot-assisted Repetitive Transcranial Magnetic Stimulation (rTMS) of the pre-supplementary motor area on dyskinesia in PD [56]. Seventeen patients received either rTMS or sham procedure in a randomized within-subject counterbalanced cross-over design, to assess levodopa-induced change in task-related regional neural activity, PD symptoms, and time to onset of dyskinesia after intake of levodopa. The intervention resulted in a decrease in dyskinesia severity by 25% at the group level with a significant reduction of median UDysRS score by 3.0 points, and it showed a delay in onset of dyskinesia, without adversely affecting the anti-parkinsonian action of levodopa, assessed via the UPDRS part III score.

3.6.2. Transcranial direct current stimulation

NCT05752240 is a randomized single-blind parallel assignment trial that seeks to perform transcranial Direct Current Stimulation (tDCS) vs sham stimulation of the primary motor cortex over 10 sessions. The primary endpoint is the change in the total score of part III of the UDysRS. The status of the trial is unknown as of late April 2025.

3.6.3. Transcranial static magnetic field stimulation

NCT02657681 was a Phase II randomized double-blind parallel assignment study that investigated transcranial Static Magnetic Field Stimulation (tSMS) of the motor cortex compared to sham stimulation in PD patients with dyskinesia [57]. Fifty patients were enrolled, 25 in the active group and 25 in the sham group. Device efficacy was assessed by evaluating dyskinesia severity using UDysRS scores. Although the approach was safe, tSMS failed to be superior to sham stimulation.

4. Discussion

Despite the effectiveness of levodopa at alleviating PD motor symptoms, dyskinesia emerges as a common adverse effect, severely impacting patients’ quality of life and limiting therapeutic dosing. A tool commonly used in clinical trials to rate dyskinesia severity is the UDysRS, which comprises four parts, encompassing both patient perception and objective assessment. This scale has been clinimetrically validated and was found to have acceptable levels of internal consistency, as well as high inter- and intra-rater reliability [58]. However, a criticism of that scale is that it remains unclear what a ‘clinically important’ change might be, and the impact of statistical difference may not necessarily translate into less impairment or better quality of life for patients. Accordingly, efforts have been made to define what changes obtained in Phase II trials might lead to success in Phase III trials [59–61]. This limitation of that scale should be kept in mind when interpreting the results of the various studies discussed above. We also note that few studies employed wearable sensors, which could provide additional measurement of dyskinesia.

The past 5 years have seen several molecules being assessed, many of which led to promising results. It is particularly exciting to see novel ways of modulating old targets (e.g., befiradol and mesdopetam), as well as compounds modulating new targets (e.g., CPL500036, lenrispodun, pridopidine). However, the development of some molecules for the treatment of dyskinesia has been the unfortunate victim of the pandemic (e.g., dipraglurant, pridopidine), while other drugs failed to show efficacy (e.g., foliglurax and JM-010). Lastly, some drugs have seemingly been repurposed for indications unrelated to dyskinesia or PD (e.g., AV-101 and pridopidine). It is quite unfortunate that the results of the clinical trial with AV-101 have not been released and will perhaps never be disclosed, as the mechanism of action of the compound is novel and enticing. It is encouraging that befiradol, mesdopetam, and CPL500036 met their safety and efficacy endpoints at the Phase II level, with a Phase III trial being planned with mesdopetam.

Regarding befiradol, the molecule was efficacious in two non-human primate models of PD [62,63], and there is extensive pre-clinical evidence supporting an anti-dyskinetic effect of 5-HT_1A activation, however at times at the price of reducing the anti-parkinsonian effect of levodopa [64–68]. The 5-HT_1A agonists tested so far in the clinic, i.e. buspirone [69,70], sarizotan (which was admittedly not selective [71]) [72,73] and tandospirone [30] did not all succeed at alleviating dyskinesia, but all led to deterioration of parkinsonism. It will therefore be interesting to see whether DSP-9632P ends up being efficacious and well tolerated. In contrast, befiradol had a positive effect on parkinsonism, indicating that deterioration of parkinsonian disability may not be a liability of this molecule.

In the case of mesdopetam, its unique structural and binding properties might allow it to safely antagonize D₃ and possibly D₂ receptors without deleterious effect on parkinsonism, a risk inherent to this mechanism of action in PD. Indeed, in the past, molecules with D₂ antagonistic effects such as olanzapine [74,75], ziprasidone [76], risperidone [77,78] or haloperidol [79], have all led to an exacerbation of parkinsonian disability. Blocking dopamine D₂ and D₃ receptors without hindering the anti-parkinsonian action of levodopa would be a breakthrough and would potentially lead to a reappraisal of the way dopamine receptors can be interacted with in PD.

We are very excited by CPL500036, which has met its efficacy endpoints at the Phase II level. Inhibition of PDE10 is a new clinical approach for the treatment of dyskinesia. PDE10 is highly expressed in the striatum [80], where it modulates the activity of the direct and indirect pathways [48]. According to the company’s internal release, CPL500036 exceeded their expectations [50]; whether a Phase III trial will be initiated remains unknown, but based upon the results, the compound and the approach in general, appear extremely promising for the treatment of dyskinesia. Similarly, it will be fascinating to discover whether the PDE1B inhibitor lenrispodun is also efficacious at reducing dyskinesia, as PDE inhibitors targeting two different enzymes could potentially advance to Phase III testing.

Regarding the transcranial stimulation paradigms discussed above, their efficacy remains to be demonstrated on a broader scale, yet the concept is interesting. Indeed, the putative mechanism is to modulate cortical excitability so that a balance can be found between excess of movement (dyskinesia) and a paucity of movement (parkinsonism). It is encouraging to notice that different brain areas have been modulated, using varied protocols; perhaps it is only a matter of time before the right protocol and brain region are discovered. These stimulation paradigms have the advantage of not requiring medication intake and would thereby minimize polypharmacy.

5. Future perspectives

Here, we have provided a summary of the on-going or recently completed clinical trials for the treatment of levodopa-induced dyskinesia in PD. We are optimistic for the future of the treatment of dyskinesia, as a few agents have proved to be safe, well tolerated and efficacious at the Phase II level. However, we also must remain realistic, as it is unclear if these successful compounds will enter Phase III testing, and if they will eventually be efficacious in hypothetical Phase III trials.

What is interesting as well is that the molecules discussed above encompass novel ways to modulate old targets, as well as ways to interact with new targets. Thus, perhaps what was needed all along to effectively and safely activate 5-HT_1A receptors was a full and highly selective 5-HT_1A agonist such as befiradol, while partial agonists and less selective drugs were meant to be either ineffective or poorly tolerated? Similarly, perhaps an antagonist with agonist-like binding properties such as mesdopetam is the way to safely and effectively target dopamine receptors? Results of Phase II studies would so far suggest so, but definite answers will come from Phase III trials. Regarding new pharmacological targets, we are enthusiastic about the potential of PDE inhibitors such as CPL500036 and lenrispodun, although we acknowledge that this is a new target, and that more studies are required to better understand the mechanisms underlying the efficacy of the approach, and to confirm the positive results obtained with the molecule.

We would like to end this article with a question to which we do not pretend to have the answer. Dyskinesia is an incredibly complex phenomenon, with a wealth of neurotransmitters involved [6,81]. Amantadine [45] and clozapine [82], the only two drugs regarded as efficacious for the treatment of dyskinesia by the International Parkinson and Movement Disorder Society [14], bind to several receptors, and it is unclear which one(s) underlies their efficacy. We make this observation, as we noted that all drugs listed above harbor selectivity for a specific target. Whereas it may be feasible to alleviate dyskinesia with pharmacologically selective molecules, will it be possible to eradicate it with a target-selective approach? Time will tell.

Funding Statement

PH has received payments from Neurodiem, AbbVie, Throughline Strategies, Sanford Burnham Prebys, Consynance Therapeutics, Sunovion, and adMare BioInnovations and is a Scientific Advisor to Talon Pharmaceuticals.

Article highlights

• There are only few available options for the treatment of levodopa-induced dyskinesia.

• Pharmacological targets recently evaluated in clinical trials include 5-HT_1A ± 5-HT_1B/1D receptors, mGlu₄ receptors, mGlu₅ receptors, NMDA receptors, dopamine D₃ receptors, sigma-1 receptors, as well as PDE10 and PDE1B.

• Targets for which drug candidates remain under development include 5-HT_1A receptors (befiradol), dopamine D₃ receptors (mesdopetam), as well as PDE10 (CPL500036).

• Recent non-pharmacological approaches under investigation include rTMS, tDCS, and tSMS.

Author contributions

Jawad Al-Kassmy, Mohammed Alsalmi, Woojin Kang, and Michael Palayew performed the literature research and wrote the first draft. Philippe Huot reviewed the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).