Abnormalities in the thalamo- and cortico-striatal glutamatergic transmission have long been implicated in the pathogenesis of LIDs. The thalamostriatal system consists of multiple neural systems that originate from a wide variety of thalamic nuclei and terminate in functionally segregated striatal territories where ionotropic (NMDA and AMPA) and metabotropic receptors are densely present. Moreover, abnormal phosphorylation and synaptic redistribution of several NMDA subunit receptors seem to play an important role in the expression of LIDs, and have been found in the dyskinetic state.119 An increase in the binding of these receptors has been reported in patients with LIDs,120 leading to a state of hypersensitivity to glutamatergic action in the dopaminergic neurons in the substantia nigra.121 Subsequently, glutamate-mediated neurodegeneration ensues, facilitating the expression of LIDs after L-dopa treatment.
Despite existing data at the preclinical level showing the antidyskinetic effect of several NMDA modulators, the results from clinical trials using NMDA antagonists for the treatment of LIDs have been disappointing. Therefore, none of the agents with anti-NMDA activity has been approved by the FDA for the treatment of LIDs.122 The story is different, however, for amantadine.
Amantadine, a low-affinity, noncompetitive NMDA receptor antagonist has been considered the most effective drug used in reduction of dyskinesia without worsening of motor performance. Luginger et al.128 studied the effect of amantadine in patients with LIDs in a 5-week, double-blind crossover trial. Dyskinesia severity assessed following oral L-dopa challenges and by self-scoring dyskinesia diaries was reduced by approximately 50% after amantadine treatment compared with baseline or placebo. Similarly, dyskinesia assessments on the Unified Parkinson’s Disease Rating Scale (UPDRS) part IV also revealed significant improvement after treatment with amantadine. The magnitude of the L-dopa motor response to oral challenges was not different after amantadine or placebo treatment, and there was no significant reduction of daily off time when patients received active treatment.123 Another double-blind, placebo-controlled, crossover study of 18 advanced PD patients attempted to determine the effects of amantadine on LIDs and motor fluctuations in PD.129 In the 14 patients completing this trial, amantadine significantly reduced dyskinesia severity by 60% compared to placebo, without altering the antiparkinsonian effect of L-dopa. Motor fluctuations occurring with patients’ regular oral L-dopa regimen also improved according to UPDRS and patient-kept diaries124 There have been a few more double-blind, placebo-controlled studies showing improvement of dyskinesia after amantadine administration. Snow et al.130 showed a 24% reduction without any influence on the severity of “on” period parkinsonism. On the other hand Da Silva-Junior et al.125 demonstrated that the duration of LIDs was reduced by amantadine despite the fact that the dyskinesia scores were not changed, and that amantadine significantly decreased UPDRS scores, showing that this substance can improve the subjective experience of dyskinesias.
Most of these efficacy studies had active treatment periods of only 2 to 4 weeks and data on the long-term antidyskinetic effect of amantadine are limited and contradictory. A double-blind study was designed to assess the duration of the antidyskinetic effect of amantadine on LID over a period of 12 months. Forty patients treated for 7.5 (S.D. 2.2) years with levodopa 729.3 (S.D. 199.4) mg/day and DA agonists, having peak-dose or diphasic dyskinesia with or without pain, were assessed. Twenty patients received amantadine (300 mg) and 20 received a placebo. After 15 days of amantadine treatment there was a reduction by 45% in the total dyskinesia scores; however, all patients receiving amantadine were withdrawn in 3–8 months, inducing a rebound with increase of dyskinesia by 10%–20% in 11 patients.126 One year after completion of an acute, double-blind, placebo-controlled, crossover study, 17 of the original 18 patients returned for reevaluation (13 remained on amantadine) of motor symptoms and duration of antidyskinetic effect of amantadine in advanced PD. Results showed that one year after initiation of amantadine cotherapy, its antidyskinetic effect was similar in magnitude (56% reduction in dyskinesia compared with 60% 1 year earlier). Motor complications occurring with the patients’ regular oral L-dopa regimen also remained improved according to the UPDRS part IV scores.127 A randomized placebo-controlled parallel group study was performed to assess the long-term antidyskinetic effect of amantadine in 32 PD patients on stable amantadine therapy for LID over at least one year.128 The mean duration of amantadine treatment in all patients was 4.8 (2.9) years. Patients were switched in a double-blind manner to amantadine or placebo and followed for 3 weeks. In order to determine whether early treatment with amantadine would delay the development of LIDs, patients with the clinical diagnosis of PD with no history of anti-PD drugs at first visit and subsequently treated with L-dopa for at least 5 years were divided in two groups. Group A received amantadine prior to L-dopa use while group B did not. Amantadine treatment prior to the use of L-dopa did not delay onset of LID nor reduce the incidence of LID.129 Dyskinesia duration and intensity assessed in these intervals showed a significant increase of UPDRS part IV scores in patients treated with placebo compared with no significant change in patients staying on amantadine. In the AMANDYSK trial, 57 amantadine-treated patients were randomly switched to placebo and the washout of amantadine was associated with marked worsening of LIDs in a median time of 7 days.130
Currently, there are two ongoing multicenter, randomized, double-blind, placebo-controlled study to evaluate the tolerability and efficacy of each of three dose levels of extended-release formulations of amantadine (Adamas and Osmotica formulations), dosed once daily for the treatment of LIDs in subjects with PD. The use of this medication is expected to improve safety and tolerability of amantadine via the stabilization of its plasma concentrations throughout the day.131
Other agents that block NMDA receptors (remacemide132 and riluzole,133 both noncompetitive NMDA receptor antagonists) have not been proven to be of benefit for the management of LIDs in PD patients. However, another antagonist, dextromethorphan, was given to PD patients with motor fluctuations in a double-blind, placebo-controlled, crossover study, showing that average and maximum dyskinesia scores improved by >50%, without compromising the antiparkinsonian response magnitude or duration of L-dopa.134 There is an ongoing clinical trial to evaluate the efficacy, safety, and tolerability of the use of a new compound containing 45 mg dextromethorphan and 10 mg quinidine (a potent inhibitor of CYP2D6 which in turn metabolizes dextromethorphan rapidly and extensively) compared with placebo for the treatment of LIDs in PD patients.135
The effect of memantine on cardinal symptoms of PD on patients with LIDs was studied in a crossover design by randomizing subjects to memantine or placebo. Results showed that memantine may improve parkinsonian symptoms independently of dopaminergic drugs but without improvement of LIDs.136 A randomized, double-blind study giving memantine versus placebo analyzed motor and dyskinesia scores and their axial subscores, demonstrating that memantine treatment was associated with lower axial motor symptoms and dyskinesia scores.137 Varanese et al.143 described the effects of memantine in 3 patients with motor complications and their follow-up evaluations at 1–5 years, suggesting that memantine may be a potentially effective drug with a low side effect profile for the treatment of LIDs in PD, particularly when other alternative agents such as amantadine are contraindicated. Unexpectedly, Vidal et al.138 reported improvement of LIDs in 2 cases of individuals diagnosed with PD dementia who were started on memantine, given for cognitive improvement, but whose most impressive clinical outcome was the unanticipated improvement in dyskinesias. During the last few years, a number of studies have tested selective NMDA antagonists in the treatment of LIDs, though without robust effect.139–141
A number of preclinical studies have highlighted the involvement of metabotropic glutamate receptors (mGluRs) in the pathophysiology of LIDs.142 Antagonists of mGluR5 reduce overactivity of NMDA receptors and resulting overexcitability, both important factors in the expression of LIDs. The effect of the mGluR5 antagonist [(2-methyl-1,3-thiazol-4-yl)ethynyl]pyridine (MTEP) was analyzed in an experimental model of PD with LID in 6-OHDA-lesioned rats, demonstrating that mGluR5 antagonists may be useful for the symptomatic treatment of LIDs.143 Other mGluR5 inhibitors, such as AFQ056 (mavoglurant) and ADX48621 (dipraglurant) are currently in clinical trials in the treatment of LIDs.144, 145
A model of MPTP-induced PD monkeys treated with L-dopa was used to investigate the acute effects of the novel mGluR5 antagonist AFQ056 (mavoglurant) on motor behavior and dyskinesias, showing LIDs reduction without affecting the beneficial motor effects of L-dopa.146 Berg et al.152 assessed the efficacy, safety, and tolerability of AFQ056 in PD patients with LIDs in two randomized, double-blind, placebo-controlled studies with moderate to severe LID (study 1) and severe LID (study 2) on stable dopaminergic therapy to those randomized to placebo. A double-blind, placebo-controlled study of patients with PD and moderate to severe LIDs who were receiving stable L-dopa/antiparkinsonian treatment and were not currently receiving amantadine, were randomized to receive either different doses of AFQ056 or placebo.147 Patients randomized to AFQ056 200 mg daily administered in 2 doses had a significant improvement in dyskinesia at week 12 compared to those randomized to placebo in a dose-response relationship. The most common adverse events were dizziness, hallucinations, fatigue, nasopharyngitis, diarrhea, and insomnia. Currently, there is an ongoing, open-label study to generate long-term safety, tolerability, and efficacy data for AFQ056 in patients who have participated in and completed any AFQ056 phase II study in PD-LID.148
Dipraglurant (ADX48621), a novel mGluR5 negative allosteric modulator, was evaluated for the safety, tolerability, and antidyskinetic profile in PD patients with LID.149 Seventy-six patients with moderate to severe LID were randomized in a double-blind placebo-controlled study. Dipraglurant significantly reduced dyskinesia without affecting L-dopa efficacy. There were no significant changes in safety monitoring parameters. Furthermore, dipraglurant increased daily “on” time without dyskinesia in weeks 1, 2, 3, and 4 and reduced daily “off” time at week 4.
There is evidence regarding the contribution of AMPA receptors to the pathogenesis of parkinsonian signs and LIDs. Konitsiotis et al.156 compared the ability of an AMPA agonist (CX516) and a noncompetitive AMPA antagonist (LY300164) to alter parkinsonian symptoms and LID in MPTP-lesioned monkeys. CX516 alone, and when combined with low-dose L-dopa, did not affect motor activity but did induce dyskinesia. Moreover, following injection of the higher doses of L-dopa, it increased LID by up to 52%. LY300164 (talampanel) potentiated the motor activating effects of L-dopa and at the same time decreased LID by up to 40%.150 Several clinical studies have evaluated the effects of talampanel in PD and LID; however, no data have been published.122
Perampanel, a novel, noncompetitive, selective AMPA-receptor antagonist demonstrated evidence of efficacy in reducing motor symptoms in animal models of PD. One study assessed the safety and efficacy of different doses of perampanel versus placebo for treatment of “wearing off” motor fluctuations in patients with PD.151 There were no significant changes in dyskinesia or cognitive function when perampanel was compared to placebo. Two multicenter randomized, double-blind, placebo-controlled, parallel group phase III studies assessed the efficacy and safety of adjunctive perampanel in patients with PD and motor fluctuations. In neither of these studies was perampanel superior to placebo in improving motor symptoms or LIDs.152 Another study aimed to assess the efficacy and tolerability of perampanel in L-dopa-treated patients with moderately severe PD and motor fluctuations using an active comparator study design.153 Although perampanel was generally well tolerated, it was not superior to placebo on any efficacy end point measures including duration and severity of dyskinesias.
Studies in the MPTP-lesioned primate model of PD have demonstrated that alpha2-adrenergic receptor antagonists can alleviate LID.154–157 Fipamezole is an alpha2-antagonist which has a high affinity at human alpha(2A), alpha(2B), and alpha(2C) receptors. It also has a moderate affinity at histamine H1 and H3 receptors and serotonin (5-HT) transporter. In the MPTP-lesioned marmoset, fipamezole significantly reduced LID without compromising the antiparkinsonian action of L-dopa.154 The duration of action of the combination of L-dopa and fipamezole was 66% greater than that of L-dopa alone.
A proof-of-concept study in 10 PD patients with LIDs showed beneficial effects at single doses of 60 and 90 mg of fipamezole in a double-blind, randomized, placebo-controlled, dose-escalating 28-day study conducted at 25 centers in the United States and seven centers in India.155 There was no statistically significant difference in suppressing LID between fipamezole and placebo. However, because of inhomogeneity recognized between U.S. and Indian study populations, a prespecified subgroup analysis of the U.S. population showed that fipamezole at 90 mg reduced LIDs. Fipamezole was associated with an acceptable profile of adverse effects, including mild, transient elevation of blood pressure.
A study of the selective and potent alpha2-adrenergic receptor antagonist, idazoxan, showed that the drug reduces LIDs in the MPTP-lesioned marmoset model of PD.162 Furthermore, the coadministration of idazoxan with L-dopa more than doubled its duration of action as an antiparkinsonian drug compared to that seen with L-dopa alone. However, idazoxan as a monotherapy displayed no antiparkinsonian effect. Another experimental study in MPTP-lesioned monkeys demonstrated that idazoxan alone increased locomotor activity and improved the disability score with virtually no dyskinesias in 50% of the animals.156 In combination with L-dopa, idazoxan did not impair the antiparkinsonian response but significantly reduced dyskinesias in all animals up to 65% and delayed their onset. A pilot, randomized, placebo-controlled study in PD patients with LID assessed the effects of single oral doses of idazoxan on motor parkinsonian disability and LID. The severity of LID improved after 20 mg idazoxan pretreatment, while there was no concomitant deterioration in the antiparkinsonian response to L-dopa.157 However, the same dosage in another trial proved to be ineffective and caused adverse effects.158 An in vivo microdialysis analysis in 6-OHDA-lesioned rats with LIDs showed that the decrease of L-dopa-derived extracellular DA levels in the lesioned striatum in dyskinetic rats significantly contributed to the antidyskinetic effect of idazoxan.159
Preladenant is one of the most selective A2A-receptor antagonists, with a robust in vivo activity.167 In 6-OHDA-lesioned rats, daily administration of preladenant showed inhibition of L-dopa-induced behavioral sensitization, which suggested that this drug had the potential to reduce the risk of LIDs. Subsequently, the effects of preladenant in MPTP-lesioned primate models of PD with LID were investigated showing evidence of improvement of motor ability without induction of LIDs.160 Hauser et al.161 aimed to assess the efficacy and safety of preladenant in patients with PD and motor fluctuations who were receiving L-dopa and other antiparkinsonian drugs. This phase 2, randomized, placebo-controlled, double-blind trial enrolled 253 patients at 44 sites in 15 countries. At week 12, the mean daily off time (primary outcome measure) was significantly reduced compared to placebo in patients on 5 mg and 10 mg of preladenant. The most common adverse events were worsening of parkinsonism, somnolence, dyskinesia, nausea, constipation, and insomnia. Although the use of preladenant twice daily was clinically useful to reduce off time, it failed to significantly improve LID. The 36-week, open-label phase, follow-up study of preladenant 5 mg twice a day as an L-dopa adjunct in 140 subjects with fluctuating PD reported treatment-emergent adverse events by ≥15% of subjects, including dyskinesia (33%) and constipation (19%). Preladenant treatment also provided “off” time reductions and “on” time increases, but at the expense of increased dyskinesia rates.162
Nicotinic receptors have been implicated in the expression of LID in PD, although solid scientific support is lacking. Nicotine and nicotinic agonists increase dopamine release from mesolimbic and nigrostriatal neurons in vitro and in vivo. There is ample evidence from in vitro experiments on synaptosome preparations for the facilitation of basal dopamine release by presynaptic nicotinic receptors on striatal dopaminergic terminals causing excitation of dopaminergic SNc neurons as well as a modulatory influence.163
Nicotine, a nonspecific agonist of nicotinic receptors, is very effective in preventing the occurrence of LIDs and reducing established LIDs in several parkinsonian animal models. However, preclinical results are limited and should be interpreted with caution.122
5. Other pharmacological treatments
Clozapine was the first “atypical” antipsychotic, and appears to be free of extrapyramidal side effects despite its D2 receptor blocking action. It has effects on a large number of neurotransmitter systems so that interpretation of its modes of action is open to much speculation. In addition to having antitremor activity in PD, it has been reported in a double-blind, placebo-controlled trial of 50 subjects, to reduce dyskinesias, reduce on time with dyskinesias without increasing off time.164 The mean dose used was 39 mg/d. Prospective, open-label studies have also supported the beneficial effect of clozapine on dyskinesias, without reducing on time or worsening motor function, with one study using only 30 mg/d,165 whereas another used doses that were generally between 100 and 200 mg/d, with one subject on 400 mg/d.166 The higher doses produced marked somnolence, offsetting the benefit.
Other treatment options that have been studied include partial dopamine agonists (aripiprazole,167 pardoprunox168, 169), monoamine oxidase-B inhibitors (selegiline,170 rasagiline,171 safinamide172), cannabidiol extract,173 5-HT agonists, and anticonvulsants.118 However, their use remains limited, as the great majority have failed to confirm safety, tolerability, or effectiveness. Numerous ongoing studies, involving dopaminergic and nondopaminergic pathways, are attempting to provide further information about these parameters, constituting innovative steps toward novel therapeutic approaches toward this disabling complication of L-dopa.
It is beyond the scope of this chapter to comprehensively review the effects of neurosurgical interventions in the treatment of LIDs in PD. Both pallidotomy and pallidal deep brain stimulation (DBS) are probably the most effective neurosurgical treatment, because they significantly improve all of the LIDs, including off-period dystonia, without reduction of L-dopa dosage. Subthalamic (STN) DBS has no direct therapeutic effects on LID, but can substantially improve this iatrogenic symptom as a result of decreased L-dopa dosage.174 Pallidotomy and globus pallidus internus (GPi) DBS or subthalamotomy and STN DBS have assumed a pivotal role in surgical treatment of PD. Interventions in either the STN or the GPi seem to be similar in controlling most of the other motor aspects of PD; nonetheless, GPi surgery seems to induce a more particular and direct effect on dyskinesia, while the antidyskinetic effect of STN interventions is mostly dependent on a reduction of dopaminergic drug dosages.175
The first double-blind, crossover study evaluating the results of GPi versus STN DBS in PD showed that both procedures significantly reduced dyskinesia (by 58% with STN DBS and 66% with GPi DBS).176 Furthermore, the average medication dosage, measured in L-dopa equivalents, was decreased significantly more with STN DBS. Another study with a longer follow-up, mean 48.5 months, showed 64% mean improvement in LID after this period of observation.177 Eleven PD patients undergoing GPi DBS were followed for 5 years showing that, despite a decline on the motor benefit for the off-period scores after 3 years, the improvement in LID was sustained.178
Long-term studies of bilateral STN DBS in patients with advanced PD have demonstrated the sustained efficacy of this therapy over time. A 5-year prospective study of 49 consecutive patients treated with STN DBS noted improvement of LIDs with dyskinesia disability and duration at 5 years being improved by 58% and 71%, respectively, in comparison with baseline.179 Similar benefits with respect to dyskinesia were observed in a 5-year follow-up of 37 PD patients treated with stimulation of the STN.180 Finally, a comprehensive metaanalysis of 921 patients who underwent STN DBS between 1993 and 2004 noted an average reduction in LID following surgery of 69.1%.181
In conclusion, although STN and GPi procedures have different mechanisms of action, both are effective treatment strategies to control LIDs. GPi interventions may have a more immediate effect, independent of reduction of L-dopa daily dosage; although the tendency is to adopt STN DBS, as this procedure also brings marginally better improvements in off-period motor scores than GPI DBS.182