Adenosine 2A Receptor Antagonists for the Treatment of Motor Symptoms in Parkinson's Disease

Abstract

Background

Treatment of motor fluctuations in Parkinson's disease (PD) remains an unmet challenge. Adenosine 2A (A_2A) receptors are located along the indirect pathway and represent a potential target to enhance l‐3,4‐dihydroxyphenylalanine (l‐DOPA) antiparkinsonian action.

Methods

This article summarizes the preclinical and clinical literature on A_2A antagonists in PD, with a specific focus on their effect on off time, on time, and dyskinesia.

Findings

Several A_2A receptor antagonists have been tested in preclinical studies and clinical trials. In preclinical studies, A_2A antagonists enhanced l‐DOPA antiparkinsonian action without exacerbating dyskinesia, but A_2A antagonists were generally administered in combination with a subthreshold dose of l‐DOPA, which is different to the paradigms used in clinical trials, where A_2A antagonists were usually added to an optimal antiparkinsonian regimen. In clinical settings, A_2A antagonists generally reduced duration of off time, by as much as 25% in some studies. The effect of on time duration is less clear, and in a few studies an exacerbation of dyskinesia was reported. Two A_2A antagonists have been tested in phase III settings: istradefylline and preladenant. Istradefylline was effective in two phase III trials, but ineffective in another; the drug has been commercially available in Japan since 2013. In contrast, preladenant was ineffective in a phase III trial and the drug was discontinued. A phase III study with tozadenant will begin in 2015; the drug was effective at reducing off time in a phase IIb study. Other A_2A antagonists are in development at the preclinical and early clinical levels.

Dopamine replacement with l‐3,4‐dihydroxyphenylalanine (l‐DOPA) is the most effective therapy to alleviate the manifestations of Parkinson's disease (PD).1 Unfortunately, chronic dopaminergic therapy leads to dyskinesia and motor fluctuations in the majority of patients.2 Dyskinesia and motor fluctuations affect virtually all patients with advanced disease.3 Treatment of motor fluctuations is challenging because it requires a fine tuning between extending duration of antiparkinsonian benefit while avoiding increasing dyskinesia. The clinically available classes of agents, catechol‐O‐methyltransferase (COMT) and monoamine oxidase B (MAO‐B) inhibitors, reduce off time duration by approximately 1 hour daily, a ≈20% reduction.4, 5, 6, 7 However, daily off time is not abolished with COMT and MAO‐B inhibitors, and agents from these categories may also exacerbate dyskinesia severity. The quest for drugs that would further address motor fluctuations without worsening dyskinesia therefore remains ongoing.

Over the past decade, blockade of adenosine 2A (A_2A) receptors has been assessed as a potential way to address motor fluctuations, and several A_2A antagonists have been studied for that indication. A_2A receptors are abundant within the striatum, where they are localized within two neuronal populations and exert different physiological effects. A_2A receptors are encountered on striatal projection neurons forming the indirect pathway, where they form functional heterocomplexes with dopamine D_2, 8 metabotropic glutamate 5 (mGlu5),9 and cannabinoid 1 (CB₁)10 receptors. A_2A receptors are also encountered at the presynaptic level on corticostriatal projection terminals, where they oligomerize with adenosine A₁ receptors and modulate glutamate release8, 11 (Fig. 1). Adenosine is increasingly recognized as an important neuromodulator of the basal ganglia, where it exerts its effects through both synaptic and nonsynaptic transmission mechanisms.12

Figure 1.

Localization of A_2A receptors within the striatum. A_2A receptors are expressed on cortical projection neurons contacting striatal projection neurons of the direct pathway, where they form dimers with adenosine 1 (A₁) receptors. A_2A receptors are also encountered on striatal projection neurons of the indirect pathway, where they form dimers with dopamine D₂, mGlu₅, and CB₁ receptors. Modified from a previous work.9

This review article summarizes the literature on A_2A antagonists in PD, with a specific focus on their effect on motor function. A summary of the efficacy of the molecules listed is presented in Tables 1 (preclinical studies) and 2 (clinical trials).

Table 1.

Selective A_2A antagonists in animal models of PD

Drug	Animal Model	Effect on Motor Parameters
Istradefylline	6‐OHDA‐lesioned rat	No dyskinesia development when administered de novo No dyskinesia elicited when administered to primed animals No exacerbation of dyskinesia when administered with l‐DOPA
	MPTP‐lesioned primate	Antiparkinsonian action without eliciting dyskinesia in primed animals Enhances l‐DOPA antiparkinsonian action in the marmoset, but not in the macaque No exacerbation of dyskinesia when administered with l‐DOPA
Preladenant	MPTP‐lesioned primate	Antiparkinsonian action without eliciting dyskinesia in primed animals Enhances antiparkinsonian action of low dose l‐DOPA without eliciting dyskinesia
ST‐1535	MPTP‐lesioned primate	Increases motor activity as monotherapy Enhances the antiparkinsonian action of a subthreshold l‐DOPA dose

Only drugs that have been also tested in clinical settings are presented in the table.

Table 2.

Selective A_2A antagonists in clinical trials

Drug	Effect on Motor Parameters
Istradefylline	Phase II Reduced off time duration Inconsistent effect on on time duration May increase frequency of dyskinesia Possible tachyphylaxis to antiparkinsonian action when used as monotherapy Phase III Reduced off time duration in 2 studies, but had no effect on off time duration in another one Improvement of UPDRS Part III score in the onstate
PBF‐509	Unknown; phase I clinical trials completed
Preladenant	Phase II Reduced off time duration Extended on time duration May increase frequency of dyskinesia Phase III Not superior to placebo to reduce off time duration
ST‐1535	Unknown; phase I clinical trial completed
ST‐4206	Unknown; phase I clinical trial in progress
V81444	Unknown; phase Ib/II clinical trial in progress
Tozadenant	Phase IIb Reduced off time duration Extended on time duration Extended duration of on time without troublesome dyskinesia
Vipadenant	Phase II Reduced off time duration Extended on time duration, mostly on time without troublesome dyskinesia

Methods and Findings

A_2A Receptors in PD

Very few studies have assessed the fate of A_2A receptors in PD and motor complications. Striatal A_2A receptor levels are not altered after 6‐OHDA lesion of the medial forebrain bundle in mice, compared to the unlesioned side.13 A postmortem study performed in PD patients found increased A_2A receptors in the putamen of dyskinetic PD patients and increased A_2A levels in the globus pallidus pars externa of PD patients, regardless of their dyskinetic status.14 A PET study found increased A_2A receptor levels in both the caudate and putamen of dyskinetic patients, compared to nondyskinetic patients and healthy individuals.15 Whether these increases occur at the pre‐ and/or postsynaptic level(s) is unknown. The clinical impact of these increases is uncertain, but suggests that A_2A receptors may be involved in dyskinesia, although it will be critical to determine whether the increases are pre‐ and/or postsynaptic before making such assumptions. Thus, pre‐ and postsynaptic A_2A receptors could thus exert different effects on parkinsonian disability and dyskinesia severity. This anatomical fingerprint is unknown for most of the drugs that will be discussed here, although it appears that istradefylline interacts primarily with postsynaptic A_2A receptors, whereas SCH‐442,416 antagonizes mostly presynaptic receptors.16

Nonselective A_2A Antagonists

Caffeine

Caffeine is a nonselective A₁ and A₂ antagonist.17 Caffeine may act as an inverse agonist at A_2A receptors.18 Caffeine also inhibits nonselectively the enzyme phosphodiesterase.17

In an early double‐blind, placebo‐controlled, crossover clinical trial performed in 5 PD patients, caffeine (800–1,400 mg daily in 3 intakes) did not lead to any enhancement of l‐DOPA or piribedil antiparkinsonian action. Patients reported a ≈20% increase of dyskinesia severity.19 In another early double‐blind, placebo‐controlled, crossover study performed in 6 PD patients, caffeine (1,000 mg daily in 4 intakes) did not enhance the antiparkinsonian action of bromocriptine.20

A randomized, placebo‐controlled, pharmacokinetic/pharmacodynamic study showed that caffeine 200 mg 15 minutes before l‐DOPA/carbidopa 250/25 mg reduced time to maximal plasma l‐DOPA concentration,21 perhaps further contributing to the adjunct effect of caffeine on l‐DOPA antiparkinsonian action.

Another study evaluated the effect of caffeine (100 mg daily) on freezing of gait. In this nonrandomized, single‐blind, placebo‐controlled, clinical trial, caffeine modestly improved freezing of gait, but the effect was not sustained.22 In an open‐label 6‐week study conducted in 25 PD patients, caffeine 200 mg twice‐daily (BID) was found to be well tolerated and to have a beneficial effect on the UPDRS Part III.23 In a randomized, double‐blind, placebo‐controlled 6‐week study where 200 mg BID caffeine was administered, UPDRS part III score was modestly reduced. Rigidity and bradykinesia were the two most improved parameters; dyskinesia severity was not affected.24

In a study that retrospectively analyzed caffeine consumption of patients enrolled in the Comparison of the Agonist Pramipexole with Levodopa on Motor Complications of Parkinson's Disease (CALM‐PD) and CALM Cohort extension studies, patients taking >206 mg of caffeine daily tended to present less dyskinesia than patients taking lower caffeine doses, although this was not significant.25

A phase III study assessing the effects of caffeine (200 mg BID) on PD motor symptoms will soon start enrolling patients (NCT01738178).

Several clinical studies (not all cited here) have suggested that caffeine might exert a neuroprotective effect and reduce the risk of developing PD, especially in males and postmenopausal females not under hormone replacement therapy,26, 27, 28, 29, 30 although once PD has developed, caffeine does not seem to slow the rate of disease progression.31

There is thus great variability in terms of methodology employed, patient assessment, and dose of caffeine employed in the studies cited above. However, high‐dose caffeine may exacerbate dyskinesia, whereas its effect on parkinsonism appears modest, if at all. The upcoming phase III study assessing the effect of caffeine on PD motor symptoms will provide highly valuable information on the antiparkinsonian effectiveness of caffeine.

Theophylline

Theophylline is a methylxanthine that, as with caffeine, antagonizes both A₁ and A₂ receptors and is a nonselective phosphodiesterase inhibitor.32, 33 In the 6‐OHDA‐lesioned rat, theophylline increased motor activity, alone and in combination with l‐DOPA.34 Abnormal involuntary movements (AIMs) were not assessed in this study.

In an unblinded dose‐ascending study, theophylline, up to 600 mg daily, produced a subjective improvement of parkinsonism in 10 patients with early PD and 10 with advanced PD. Theophylline also extended duration of on time in patients with advanced PD.35 In an unblinded study, the addition of theophylline 150 mg daily to l‐DOPA improved UPDRS Part III score by ≈30% over 12 weeks.36 Such findings were not replicated in a randomized, double‐blind, crossover, placebo‐controlled trial conducted in 10 PD patients where theophylline 300 mg daily was administered.37 In an earlier open‐label trial, intravenous (IV) infusion of aminophylline, of which the active ingredient is theophylline, did not enhance l‐DOPA antiparkinsonian action.38

Based on this small number of studies and the methodology employed, theophylline may have a slight antiparkinsonian effect, if any, when administered in combination with l‐DOPA.

Theobromine

Theobromine is a methylxanthine contained in cocoa. Theobromine nonselectively antagonizes A₁ and A₂ receptors, less potently than caffeine,39 and is a nonselective phosphodiesterase inhibitor.40 In a study, 200 g of 80% dark chocolate was administered to PD patients and, compared to cocoa‐free white chocolate, did not produce any change on parkinsonian disability, assessed by a blinded investigator who performed the UPDRS Part III scale 1 hour after intake.41 Interestingly, though, PD patients tend to eat more chocolate than normal individuals.42 A study assessing plasma levels of chocolate constituents and their effects on motor and nonmotor symptoms of PD has recently started (NCT02275884).

Selective A_2A Receptor Antagonists

Istradefylline

Istradefylline (KW‐6002) is a relatively selective reversible A_2A antagonist. Depending on the species and brain areas studied, its affinity for A_2A receptors ranges from 2.2 to 65 nM, whereas it varies between 120 to >1,000 nM for A₁, 150 nM for A_2B, >1,000 nM for A₃, and >10,000 nM for the adenosine transporter.43

Preclinical Studies

In a study conducted in the 6‐OHDA‐lesioned rat, de novo administration of istradefylline to animals previously unexposed to l‐DOPA did not lead to development of AIMs; de novo administration of istradefylline and l‐DOPA did not lead to development of more‐severe AIMs, compared to l‐DOPA alone. Administration of istradefylline as monotherapy to rats previously exposed to l‐DOPA and exhibiting AIMs with each l‐DOPA intake did not elicit AIMs, and, when administered with l‐DOPA, istradefylline did not exacerbate AIMs. The antiparkinsonian action of l‐DOPA, assessed by the rotarod performance, was not enhanced by istradefylline,44 possibly because the rotarod test can be affected by AIMs and rotational behavior.45

In the MPTP‐lesioned common marmoset, istradefylline as monotherapy reversed parkinsonism to an extent similar to l‐DOPA, without eliciting dyskinesia, in previously primed animals.46 Istradefylline also increased motor activity in parkinsonian marmosets.47 When istradefylline was administered to MPTP‐lesioned common marmosets primed with l‐DOPA to exhibit dyskinesia, it enhanced the antiparkinsonian action of the D₁ agonist, SKF‐80,723, the D₂ agonist, quinpirole, and l‐DOPA. Interestingly, additional reversal of parkinsonian disability was greater after combination with quinpirole and l‐DOPA than with SKF‐80,723, which suggests that dual stimulation of the indirect pathway with a D₂ agonist and an A_2A antagonist may provide greater antiparkinsonian benefit than stimulation of the direct pathway with a D₁ agonist and stimulation of the indirect pathway with an A_2A antagonist. Combining istradefylline with l‐DOPA did not exacerbate dyskinesia.48 In the parkinsonian marmoset, adding istradefylline to a low dose of l‐DOPA resulted in a reduction of dyskinesia severity, after chronic treatment, whereas the antiparkinsonian benefit was maintained49; such a strategy has seldom been tested in clinical studies, where istradefylline or other A_2A antagonists are usually added to optimal antiparkinsonian medication of patients experiencing motor fluctuations.

In the MPTP‐lesioned macaque, istradefylline reduced parkinsonism in l‐DOPA‐naïve animals. De novo coadministration of istradefylline with the dopamine agonist, apomorphine, for 3 weeks did not lead to dyskinesia development, whereas animals treated de novo with apomorphine as monotherapy began experiencing dyskinesia after 10 to 12 days of dopaminergic therapy.50 In MPTP‐lesioned macaques previously exposed to l‐DOPA, istradefylline alone reversed parkinsonian disability without eliciting dyskinesia. When administered with l‐DOPA, it did not result in greater antiparkinsonian action or more‐severe dyskinesia.51

Clinical Trials

In a phase II clinical trial, istradefylline (40 and 80 mg daily) failed to enhance the antiparkinsonian action of an optimal dose of l‐DOPA administered IV or to exert an antiparkinsonian effect as monotherapy; however, it potentiated the antiparkinsonian action of a suboptimal dose of IV l‐DOPA and extended its duration of action while eliciting significantly less dyskinesia than optimal l‐DOPA infusion.52

In another phase II clinical trial, istradefylline (20 and 40 mg daily) was added to the usual dopaminergic therapy and significantly reduced off time duration, but duration of on time with dyskinesia was increased; whether this was on time with troublesome or nontroublesome dyskinesia was not mentioned, but dyskinesia severity was not exacerbated.53 In another phase II trial, istradefylline (20 and 60 mg daily) reduced off time duration by ≈20%. Neither duration of on time nor duration of on time without troublesome dyskinesia was significantly increased, but it is interesting to note that duration of on time with troublesome dyskinesia was double in istradefylline 20 mg daily treatment compared to placebo (not significant).54 In another phase II study, istradefylline 40 mg daily was compared to placebo. Istradefylline reduced off time duration by ≈10% and increased duration of on time with dyskinesia more than placebo, although neither on time with troublesome nor on time with nontroublesome dyskinesia was extended. In contrast, on time without troublesome dyskinesia was significantly increased, compared to placebo. Occurrence of dyskinesia was more frequent in patients treated with istradefylline than placebo.55 In another phase II trial, istradefylline 20 and 40 mg daily both reduced off time duration by ≈20% and improved UPDRS Part III score while in the on state. Dyskinesia was reported by 3 patients in the placebo group, 10 in the istradefylline 20 mg group, and 8 in the istradefylline 40 mg group (groups were of similar size, with ≈120 patients in each).56

Istradefylline, 20, 40, or 60 mg daily, was administered as an adjunct to dopaminergic therapy for 52 weeks in an open‐label extension study whose primary endpoint was long‐term assessment of safety and tolerability. Reduction of off time was sustained, suggesting that there is no tachyphylaxis occurring upon A_2A receptor blockade. As many as 24% of patients reported dyskinesia or dyskinesia exacerbation.57

At phase III level, the efficacy of istradefylline 20 mg daily was compared to placebo. Istradefylline reduced off time duration by ≈15%, but had no significant effect on any on time parameters. Approximately 22% of istradefylline‐treated patients reported dyskinesia, compared to 12% in the placebo group.58 In another phase III trial, istradefylline 20 and 40 mg daily was effective at reducing off time duration, although 3 times as many patients experienced dyskinesia in the istradefylline groups, regardless of the dose, when compared to placebo; duration of on time with troublesome dyskinesia was not increased, whereas on time without dyskinesia was extended in the istradefylline 40 mg group.59

In contrast, istradefylline, 10, 20, or 40 mg daily, failed to reduce off time duration in a placebo‐controlled phase III study encompassing 610 patients. UDPRS Part III score in the on state was, however, significantly improved. Dyskinesia occurrence was similar in all groups, although there was a dose‐dependent occurrence of hallucinations. The placebo effect in this study was greater than that in the studies cited above, which may explain the lack of statistical significance for the off time parameter.60 Last, another phase III study was performed, where istradefylline was compared to entacapone, but the results of this trial have not been published (6002‐EU‐007, NCT00199394).

In these phase II and III studies, istradefylline was added to patients' antiparkinsonian regimen, which was not reduced, in comparison to some of the studies conducted in the nonhuman primate and the IV l‐DOPA trial cited above. Istradefylline is generally effective to enhance l‐DOPA antiparkinsonian benefit when assessed as off time reduction (≈10%–20% reduction of off time duration), but its overall effect on dyskinesia is unclear. It thus extended duration of on time without troublesome dyskinesia in most studies, but more patients on istradefylline experienced dyskinesia compared to placebo. Thus, although not universal, it seems that adding istradefylline to a therapeutic regimen in patients with motor fluctuations carries the risk of eliciting or exacerbating dyskinesia, at least in a subset of patients. Such a scenario was not encountered at the preclinical level when istradefylline was administered with subthreshold doses of l‐DOPA, but this has not been assessed in clinical settings, with the exception of the study where istradefylline was administered with a subtherapeutic dose of IV l‐DOPA. However, the potential of istradefylline to elicit or exacerbate dyskinesia should be mitigated by the fact that, whereas dyskinesia may have been reported by patients taking istradefylline, its severity was not assessed with rating scales and it should not be inferred that istradefylline exacerbates dyskinesia based on the literature cited above. Moreover, the reduction of off time afforded by istradefylline is not accompanied by an increase of duration of on time with troublesome dyskinesia, indicating that dyskinesia does not impact on patients' quality of life. Speculatively, the fact that dyskinesia was more reported by patients taking istradefylline might be owing to its low selectivity for A_2A over A₁ receptors or by the fact that it preferentially acts on postsynaptic receptors, as mentioned above.

In a phase II study where istradefylline 40 mg daily was administered as monotherapy, there was a significant improvement of UPDRS Part III score at week 2, but this was not sustained. Reasons for this apparent tachyphylaxis might involve relatively early disease, insufficient study power, or too low dose.61 It should be mentioned that this potential tachyphylaxis has not been encountered in preclinical studies when A_2A antagonists were administered as monotherapy or when istradefylline and other A_2A antagonists were administered over longer periods in combination with l‐DOPA. Nevertheless, it does not provide strong arguments in favor of the use of A_2A antagonists as l‐DOPA‐sparing agents in early PD, and further studies should be performed to address the potential of early therapy with istradefylline and A_2A antagonists as an l‐DOPA‐sparing strategy.

Istradefylline appears to be generally well tolerated by PD patients. In these studies, besides dyskinesia, the most commonly reported adverse events were nausea and constipation. Other adverse events included dizziness, diarrhoea, insomnia, and psychiatric complaints, including hallucinations.

Three studies with istradefylline in PD are currently active. One of these is an observational study that assesses the effect of istradefylline on nonmotor symptoms (NCT02073981), another one assesses the effect of mild hepatic insufficiency on istradefylline pharmacokinetics (NCT02256033, Phase I), whereas the third one evaluates the effect of istradefylline (20 and 40 mg daily) in patients with mild‐to‐moderate PD taking l‐DOPA (NCT01968031; phase III). Since March 2013, istradefylline (Nouriast; manufactured by Kyowa Hakko Kirin Co, Ltd) has been approved for manufacturing and marketing in Japan and became the first selective A_2A antagonist commercially available to treat motor fluctuations. The indication in Japan is to address the wearing‐off phenomenon in PD patients under l‐DOPA therapy. The drug is marketed as 20‐mg tablets and recommended daily dose is 20 or 40 mg (http://www.kyowa-kirin.com/news_releases/2013/e20130325_04.html). In the company's 2014 third‐quarter financial summary, it was indicated that Nouriast sales were progressing steadily (http://www.evaluategroup.com/View/69891--1002-modData/product/nouriast\l&&_ViewArgs=%7b%22_EntityType%22%3a0%2c%22_Parameters%22%3a%7b%22_ContextData%22%3a%22%7b%5c%22isEPVantage%5c%22%3afalse%2c%5c%22percentage%5c%22%3a-1%2c%5c%22searchWords%5c%22%3a%5c%22%5c%22%2c%5c%22sectionID%5c%22%3a%5c%22%5c%22%2c%5c%22storyID%5c%22%3a), suggesting that istradefylline is increasingly employed in Japan to treat motor fluctuations.

PBF‐509

PBF‐509 is a selective A_2A antagonist (http://www.palobiofarma.com/) currently under development for PD. Very little is known about its pharmacology and efficacy at the preclinical level, but two phase I trials have been completed recently (NCT01691924, NCT02111330).

Preladenant

Preladenant (SCH‐420,814/MK‐3814) is a selective A_2A antagonist that exhibits ≈1,000‐fold selectivity over other adenosine receptors.62 Preladenant is the most selective A_2A antagonist that has been tested in clinical trials thus far.

In the MPTP‐lesioned macaque, preladenant as monotherapy reversed parkinsonian disability without eliciting dyskinesia in primed animals. When added to a low dose of l‐DOPA, it enhanced its antiparkinsonian action, without exacerbating dyskinesia.63

In a phase II study, preladenant 1, 2, 5, and 10 mg BID was added to the usual antiparkinsonian regimen and compared to placebo. Preladenant (5 and 10 mg) reduced off time duration by ≈25% to 30% and increased on time duration by ≈10% to 15%. UPDRS Part III score in the on state was not improved. Dyskinesia was reported by 13% of patients in the placebo group, 9% in the preladenant 5 mg BID group, and 13% in the preladenant 10 mg BID group.64 In an open‐label 36‐week extension of this study, patients received preladenant 5 mg BID. The benefit on off time reduction and on time extension was maintained, but as many as 33% of patients experienced dyskinesia.65 The researchers suggested several factors that might account for the higher prevalence of dyskinesia in the open‐label phase, such as longer disease duration or increase in l‐DOPA dose, but an effect of preladenant could not be ruled out either. A study where l‐DOPA doses would have remained constant, but preladenant doses increased might have provided answers to this question. In these studies, adverse events most commonly reported included constipation, somnolence, as well as, paradoxically, diarrhoea and insomnia.

However, preladenant 2, 5, and 10 mg BID was not superior to placebo to reduce off time duration in two phase III studies.66 It is noteworthy that, in one of these phase III studies, rasagiline 1 mg once‐daily was used as an active comparator and was not superior to placebo either at reducing off time duration,66 which makes this apparent “failure” of preladenant to reach its primary endpoint difficult to interpret. However, the drug development program has been discontinued (http://www.mercknewsroom.com/press-release/research-and-development-news/merck-provides-update-phase-iii-clinical-program-prelade).

ST‐1535 and ST‐4206

ST‐1535 is a relatively selective A_2A antagonist with a Ki of 9.2 to 52 nM at A_2A receptors and 85 nM at A₁ receptors, without significant affinity for other receptors.67 In the 6‐OHDA‐lesioned rat, ST‐1535 enhanced l‐DOPA‐induced rotational behavior.68, 69 ST‐1535 was also administered to MPTP‐lesioned marmosets and potentiated the antiparkinsonian action of a subthreshold dose of l‐DOPA, in addition to providing a mild increase in motor activity as monotherapy.70 A phase I clinical trial with ST‐1535 has been performed (http://www.sigma-tau.it/fasidisviluppo.asp), and the compound was well tolerated.71

ST‐4206 is a metabolite of ST‐1535 that displays a Ki of 12 nM for A_2A receptors and 200 nM for A₁ receptors.72 ST‐4206 is presently at the phase I development stage (http://www.sigma-tau.it/principaliprogettiinsviluppo.asp).

Tozadenant

Tozadenant (SYN‐115) is a selective A_2A antagonist, with a Ki of 4.9 to 11.5 nM for the A_2A receptor and 1,320 nM for the A₁ receptor in human.73, 74 To the best of our knowledge, no study reporting the effect of tozadenant in animal models of PD has been published.

In a randomized, double‐blind, placebo‐controlled perfusion MRI phase II study performed in 21 PD patients, tozadenant 20 and 60 mg BID reduced thalamic cerebral blood flow, which is consistent with reduced activity along the indirect pathway75 and would argue for A_2A selectivity at these doses, which are lower than those demonstrated as clinically effective. Hence, in a randomized, double‐blind, placebo‐controlled phase IIb trial, tozadenant was administered in combination with l‐DOPA. Tozadenant 120, 180, and 240 mg BID each reduced daily off time duration by ≈25%. Both tozadenant 120 and 240 mg BID increased on time duration, but only tozadenant 120 mg BID increased duration of on time without troublesome dyskinesia. There was a dose‐dependent increase in patients reporting dyskinesia, with 20% in the tozadenant 180 and 240 mg BID reporting this adverse effect.76 It is important to mention that dyskinesia severity was not rated on a scale and that an exacerbation of dyskinesia cannot be inferred from such data, especially given that, during the extra on time obtained with tozadenant, dyskinesia was not disabling. Moreover, this tozadenant study suggests that there may be a ceiling to the therapeutic effect of A_2A receptor blockade, after which no further benefit is conferred, but adverse effects (e.g., dyskinesia) become more prominent. Dopamine‐related adverse events (e.g., constipation, dizziness, and nausea) tended to occur in a dose‐related fashion.

A phase III study assessing the efficacy of tozadenant as an adjunct to l‐DOPA for patients with motor fluctuations is planned for 2015 (http://www.ucb.com/investors/UCBtomorrow/tozadenant).

V81444

V81444 (BIIB34) is an A_2A antagonist currently under development for PD (http://www.vernalis.com/development-2/ncepipeline/2014-02-13-01-54-32/v81444#description). To our knowledge, the pharmacological profile of V81444 has not been disclosed, nor has its efficacy at the preclinical level. A phase I study in healthy male volunteers77 and a PET study to assess A_2A occupancy levels78 have been conducted. A phase Ib/II study has been initiated recently (http://www.vernalis.com/media-centre/latest-releases/2013-releases/659-vernalis-initiatesphase-ibii-proof-of-concept-study-with-v81444).

Vipadenant

Vipadenant (V2006/BIIB014) is a selective A_2A antagonist with a Ki of 1.3 nM for this receptor, compared to 68, 63, and 1,005 nM for A₁, A_2B, and A₃ receptors, respectively.79 In a study assessing the effect of de novo administration of A_2A antagonists, caffeine and SCH‐412,348 did not induce dyskinesia, whereas istradefylline and vipadenant induced mild dyskinesia that was not statistically different to vehicle. When administered to rats previously exposed to l‐DOPA and exhibiting dyskinesia, none of these A_2A antagonists elicited abnormal movements that were significantly different to placebo, although caffeine and vipadenant triggered mild dyskinesia.80

Vipadenant (30 and 100 mg daily) was assessed in phase II clinical trials81, 82 and tolerability issues were raised; indeed, as many as 41% of vipadenant‐treated patients experienced adverse effects. Vipadenant reduced off time duration and extended on time, mostly on time without troublesome dyskinesia. Because these studies were published as abstracts, few details are available. The vipadenant development program has been discontinued because of preclinical toxicology studies (http://www.vernalis.com/media-centre/latest-releases/584-vernalis-announces-a2A-receptor-antagonist-programme-for-parkinsons-disease-continues-withnext-generation-compound).

Other Selective A_2A Receptor Antagonists

Other A_2A antagonists have been tested in the 6‐OHDA‐lesioned rat, including ANR‐94,83, 84 KF‐17,837,85 SCH‐412,348,62 SCH‐58,261,86, 87 and SCH‐BT2.88 All of these drugs potentiated l‐DOPA‐induced rotational behaviors. SCH‐412,348 reversed parkinsonian disability in the MitoPark mouse.89

Studies conducted in the MPTP‐lesioned mouse suggested that caffeine,90 istradefylline,91 KF‐18,445,91 preladenant,92 ST‐1535,93 and theophylline90 could exert neuroprotective and anti‐inflammatory effects. Although they will not be reviewed in detail here, these data are in agreement with lower incidence of PD in coffee drinkers and suggest that A_2A blockade by caffeine may play a role in reducing the risk of PD. Potential neuroprotective mechanisms include reduction of glutamate excitotoxicity94 and reduction of neuroinflammation.95

Conclusions

Several A_2A antagonists have been tested in idiopathic PD and animal models. At the preclinical level, A_2A antagonists have shown effectiveness to enhance l‐DOPA antiparkinsonian action, without eliciting dyskinesia in primed animals. Studies in the MPTP‐lesioned primate have also suggested that monotherapy with an A_2A antagonist in early PD alleviates parkinsonian disability and does not lead to dyskinesia development. Most studies performed in animal models have, however, used an experimental paradigm different to the one that was employed in the clinic (i.e., administration of an A_2A antagonist with a subtherapeutic dose of l‐DOPA). This l‐DOPA‐sparing paradigm might be difficult to tolerate by patients and has seldom been tested in the clinic, where A_2A antagonists were generally administered with an optimal antiparkinsonian regimen to patients with motor fluctuations. For this reason (primarily methodological), the positive results obtained in preclinical settings may not translate successfully at the clinical level. Whereas the A_2A antagonists, istradefylline, preladenant, and tozadenant, have been effective at reducing off time duration in phase II settings, the efficacy of istradefylline has been variable at the phase III level, whereas preladenant did not demonstrate efficacy over placebo as an adjunct to l‐DOPA. No other A_2A antagonist has been tested at the phase III level. The effect of adding A_2A antagonists to a therapeutic regimen of l‐DOPA on dyskinesia is unclear, but the literature reviewed here suggests that dyskinesia may be more frequent with these agents. That being said, it is critical to state that, although dyskinesia may be more frequently experienced by patients taking an A_2A antagonist, its severity was not objectively rated using validated scales in the studies cited here and that the extra on time afforded by antagonizing A_2A receptors was mostly good‐quality on time, where dyskinesia was not interfering with patients' motor function. It is also noteworthy that, when efficacious, A_2A receptor blockade appears to enhance antiparkinsonian benefit to an extent similar to that of currently available therapies with COMT and MAO‐B inhibitors,5, 6 although this has been assessed as a primary endpoint in only two clinical trials, in one of which neither preladenant nor rasagiline were superior to placebo,66 whereas the other one, comparing istradefylline to entacapone, has not been published (6002‐EU‐007, NCT00199394). However, the dyskinesigenic potential of A_2A antagonists may prove to be more favorable than that of COMT and MAO‐B inhibitors and may provide arguments to use A_2A antagonists in patients with motor fluctuations and dyskinesia. In summary, whereas A_2A antagonists appear as a promising, well‐tolerated, new class of therapeutic agents for the treatment of motor symptoms in PD, further studies are needed to better define their optimal use in advanced PD. At the moment, the current indication of using A_2A antagonists in PD is in advanced disease, when patients experience motor fluctuations (istradefylline, in Japan). Most studies have assessed the effectiveness of A_2A antagonists as adjunct therapy to address wearing‐off and enhance l‐DOPA antiparkinsonian action, and the sole study that assessed the efficacy of istradefylline as monotherapy in early PD found tachyphylaxis, thus not supporting A_2A blockade as a L‐DOPA‐sparing agent in early PD. However, further studies are needed to explore the potential benefit of A_2A antagonists for this indication. Moreover, their effectiveness for a breadth of nonmotor features, such as wakefulness, motivation, and mood, might also be worth exploring.

Author Roles

(1) Research Project: A. Conception, B. Organization, C. Execution; (2) Manuscript: A. Writing of the First Draft, B. Review and Critique.

E.P.: 1B, 2A, 2B

P.H.: 1A, 1B, 1C, 2A, 2B

Disclosures

Funding Sources and Conflicts of Interest: The authors report no sources of funding and no conflicts of interest.

Financial Disclosures for previous 12 months: P.H. has received payments from Philippe Huot MD Inc.