Safety and Efficacy of Istradefylline in Parkinson’s Disease Patients With and Without Preexisting Dyskinesia: Pooled Analysis of 8 Randomized Controlled Trials
Article information
Abstract
Objective
To evaluate the efficacy of istradefylline in Parkinson’s disease patients experiencing motor fluctuations with and without dyskinesia and characterize potential predictors for treatment-emergent dyskinesia with istradefylline.
Methods
Pooled analysis of 8 phase 2b/3 trials of istradefylline (20 or 40 mg/day) versus placebo.
Results
Data from 2,719 patients, 56% of whom presented with baseline dyskinesia, were analyzed post hoc. The presence of baseline dyskinesia did not affect the mean decrease in “OFF” time with dyskinesia, increase in “ON” time without troublesome dyskinesia, or improvement in the Unified Parkinson’s Disease Rating Scale motor score associated with istradefylline treatment. Dyskinesia was reported in 17% of patients receiving istradefylline, with higher rates for women (21%), patients with a BMI <18.5 kg/m2 (22%), and patients receiving catechol-o-methyltransferase inhibitors plus dopamine agonists (22%) and monoamine oxidase B inhibitors (25%).
Conclusion
Istradefylline treatment resulted in greater reductions in total “OFF” hours/day and increases in “ON” time without troublesome dyskinesia than did placebo, regardless of the presence or absence of preexisting dyskinesia.
INTRODUCTION
In people with Parkinson’s disease (PD), adenosine A2A receptors are overexpressed in spiny projection neurons projecting from the striatum to the external globus pallidus [1], with marked overexpression in patients exhibiting dyskinesia. Istradefylline is an adenosine A2A receptor antagonist that selectively reduces PD-related overactivity of the indirect striatal output pathway, thereby restoring the balance between the outputs of the indirect and direct basal ganglia pathways and reducing motor fluctuations [2]. In a pooled analysis of eight phase 2b/3 randomized controlled trials, 12 weeks of treatment with istradefylline (20 and 40 mg/day) reduced the “OFF” time with dyskinesia by 0.38 to 0.45 hours, increased the “ON” time without troublesome dyskinesia by 0.33 to 0.40 hours, and improved the Unified Parkinson’s Disease Rating Scale (UPDRS) Part III (motor function) score relative to that of the placebo [3]. However, dyskinesia was reported as the most frequent adverse event (AE), occurring at a higher rate in patients receiving istradefylline (16%–18% for istradefylline vs. 10% for placebo) and more frequently among patients who had indicated preexisting dyskinesia than in patients without preexisting dyskinesia [3].
Although whether the marked overexpression of A2A receptors is a cause or effect of dyskinesia (and/or dopaminergic treatment) is still uncertain, it has been proposed that functional changes in adenosine signaling occur in patients with preexisting dyskinesia [4]. This post hoc analysis of pooled data from eight phase 2b/3 randomized controlled trials aimed to further investigate the effect of the presence of preexisting dyskinesia on the efficacy of istradefylline and the effects of patient demographics and concomitant medications on dyskinesia reported as an AE during the studies.
MATERIALS & METHODS
Study design
Data from eight similarly designed 12- or 16-week phase 2b/3 randomized clinical trials (NCT00199394, NCT00199407, NCT00199420, NCT00455507, NCT00456586, NCT00456794, NCT00955526, and NCT01968031) were pooled as previously described [3]. Details of each of the separate studies have been described previously; all studies were performed in compliance with the Declaration of Helsinki, with all sites having obtained local or central Institutional Review Board approval and all patients providing written informed consent. This post hoc analysis was limited to adult patients receiving 20 and 40 mg istradefylline (doses approved for use in clinical practice) or placebo [5,6].
Patients
Briefly, eligible patients were ≥20 years old (≥30 years in 6 studies) with a diagnosis of PD7 and a modified Hoehn and Yahr stage of 2–4 in the “OFF” state (“ON” state in 6002-014). All patients had received levodopa therapy for ≥1 year prior to study entry, were on a stable levodopa dose for ≥4 weeks prior to randomization, and were experiencing an average of ≥2–3 hours OFF-time/day at baseline [3]. The levodopa daily dosage (LDD) could be reduced in patients experiencing AEs (6002-US-005, 6002-US-006, and 6002-US-013), levodopa-related AEs (6002-US-018, 6002-0608, and 6002-009), or dopamine-related AEs (6002-EU-007) but could not be increased above the baseline LDD. Patients continued concomitant treatment with other approved anti-PD medications at stable dosages.
Outcomes
The time spent in the “OFF” state or ON-time without troublesome dyskinesia (sum of ON-time without dyskinesia plus ON-time with nontroublesome dyskinesia) was self-reported in patient-completed 24-hour PD diaries [8], and UPDRS Part III scores in the “ON” state were assessed by the investigator [3]. The presence of preexisting dyskinesia was determined using diary records from the week preceding the baseline visit. Patients also reported the presence of dyskinesia as a treatment-emergent adverse event (TEAE) during the studies.
Statistical analysis
The data are presented descriptively. Heterogeneity of preexisting dyskinesia between treatment groups was evaluated by the type III test for interaction (preexisting dyskinesia by treatment group) for continuous outcomes and the Cochran–Mantel–Haenszel test for discrete outcomes. Exposure-adjusted incidence rates were calculated as the number of events/total subject years of exposure to istradefylline or placebo.
RESULTS
Baseline characteristics and demographics
Patient disposition and pooled baseline characteristics for the 848 patients treated with istradefylline 20 mg/day, 879 treated with 40 mg/day, and 992 treated with placebo have been previously reported [3]. Supplementary Figure 1 (in the online-only Data Supplement) summarizes the pooling of the eight studies. Across all treatment groups, 1,515 (56%) patients were categorized as experiencing dyskinesia at study baseline. Most baseline characteristics were balanced between patients with and without preexisting dyskinesia, although patients reporting dyskinesia at baseline had a slightly greater mean time since PD diagnosis and greater LDD at study entry (Supplementary Table 1 in the online-only Data Supplement).
Effect of the presence or absence of preexisting dyskinesia on istradefylline efficacy
As shown in Figure 1, compared with placebo, treatment with istradefylline at 20 mg/day and 40 mg/day resulted in a greater reduction from baseline to week 12 in terms of the mean total “OFF” hours/day and an increase in the mean total “ON” time without troublesome dyskinesia, regardless of the presence or absence of preexisting dyskinesia (nominal p=0.849 and p=0.494 for the interaction with preexisting dyskinetic status, respectively). Similarly, the presence or absence of preexisting dyskinesia had no effect on the greater reduction in UPDRS Part III (ON) scores from baseline observed with istradefylline than with placebo treatment (nominal p=0.167 for interaction between treatment group and preexisting dyskinesia status).
Effectiveness of istradefylline in the presence and absence of pre-existing dyskinesia. Change from baseline (mean±SE) in total OFF-time (A), ON-without troublesome dyskinesia (B), and UPDRS Part III motor scores (C) at 12 weeks/early termination. SE, standard error; UPDRS, Unified Parkinson’s Disease Rating Scale.
Dyskinesia reported as a TEAE
There were no apparent differences in the rates of dyskinesia reported between patients treated with 20 mg/day and those treated with 40 mg/day. As previously reported, most dyskinesia events are mild or moderate, and the rates of new-onset dyskinesia in patients without preexisting dyskinesia are low and similar for patients receiving placebo (3.5%), 20 mg/day of istradefylline (4.4%), and 40 mg/day of istradefylline (8.4%) [3]. The exposure-adjusted incidence rate (incidence per subject-year) of dyskinesia was 0.83 patient-years for patients receiving 20 mg/day of istradefylline and 0.86 patient-years for patients receiving 40 mg/day of istradefylline, whereas it was 0.44 patient-years for patients receiving the placebo.
Dyskinesia was reported as a TEAE more frequently in patients receiving an LDD >400 mg (dyskinesia was reported in 10%–11% of patients receiving <400 mg/day of levodopa at baseline, in 14%–17% of patients receiving 400–800mg/day levodopa, and in 22%–24% of patients receiving >800 mg/day of levodopa at baseline). The reported dyskinesia rates were lower in patients receiving istradefylline+levodopa only (13%) and in those receiving istradefylline+levodopa+one other PD medication at baseline than in patients receiving multiple PD medications in combination (Figure 2). In contrast, the number of PD medications used in combination did not appear to affect the rate of dyskinesia reported as a TEAE in placebo-treated patients.
Dyskinesia reported as a TEAE stratified by baseline characteristics or concomitant PD medications. BMI, body mass index; LDD, levodopa daily dose; COMT, catechol-o-methyltransferase; MAO-B, monoamine oxidase B; TEAE, treatment-emergent adverse event; PD, Parkinson's disease.
Body mass index (BMI) generally did not affect reported dyskinesia rates in placebo-treated patients, with the exception of slightly higher reported rates in placebo-treated patients with a BMI ≥30 kg/m2. Dyskinesia was reported as an AE in 17% of patients receiving istradefylline 20 or 40 mg/day, with similar reported rates among patients with BMIs of 18.5–24.9 kg/m2, 25–29.9 kg/m2, and ≥30 kg/m2 and slightly higher rates among istradefylline-treated patients with a BMI <18.5 kg/m2 (20/93 [22%]) (Figure 2). In patients receiving istradefylline, the percentage of reported dyskinesia was increased in women compared to men (21% [159/743] vs. 14% [140/1,022], respectively).
DISCUSSION
Findings from this post hoc analysis of pooled data suggest that the presence of preexisting dyskinesia did not negatively impact the efficacy of istradefylline in reducing OFF-time, increasing ON-time without troublesome dyskinesia, or improving UPDRS motor scores when it was added to existing treatment regimens. Female patients, those with a low BMI, and patients treated with polypharmacy were more likely to report dyskinesia as an AE with the initiation of istradefylline therapy.
Dyskinesia is a well-known side effect of dopaminergic treatments, and once priming for dyskinesia has occurred, it can become treatment limiting, hindering the effective control of movement by dopaminergic medications [9]. Preclinical studies in animal models of PD [10,11] suggest that the administration of A2A antagonists, including istradefylline, potentiates the symptomatic efficacy of levodopa but does not adversely affect the appearance or expression of dyskinesia in animals that are already primed for their expression [12]. Our data show the translation of these preclinical observations to the clinic, where the existence of preexisting dyskinesia did not affect the beneficial effects of istradefylline on UPDRS Part III motor scores. The higher rates of dyskinesia reported as an AE in patients with preexisting dyskinesia (seen in this pooled population) than in those without dyskinesia [3] have also been confirmed by postmarketing evaluations in Japan [13].
TEAEs of dyskinesia were mostly mild or moderate, with no observed difference in severity reported by patients receiving a placebo or istradefylline (20 mg/day and 40 mg/day). Although the exposure-adjusted incidence of dyskinesia was greater in patients receiving istradefylline than in those receiving the placebo, there was no apparent difference observed between istradefylline 20 mg/day and 40 mg/day, suggesting that there may be little to no increase in the risk of developing dyskinesia in patients who would benefit from the higher dose of istradefylline. Nevertheless, it is difficult to differentiate dyskinesia reported as a TEAE of istradefylline from that occurring as a result of levodopa/carbidopa treatment.
While the pathophysiology of priming for dyskinesia remains elusive, the cumulative levodopa dose has been a consistent predictor of dyskinesia emergence in other studies [14,15], and in accordance with this hypothesis, we found that dyskinesia was reported less frequently in patients receiving <400 mg daily of levodopa, particularly compared with patients treated with >800 mg daily of levodopa. Similarly, female sex and low BMI (<18.5 kg/m2) have been reported as factors predictive of dyskinesia, and we also observed slightly higher rates of dyskinesia in women than in men and patients with higher BMIs. Dyskinesia was reported more frequently in patients treated with multiple adjunct combinations, including catechol-o-methyltransferase (COMT) inhibitors, dopamine agonists, and monoamine oxidase B (MAOB) inhibitors, than in those receiving istradefylline and levodopa alone or with only one other concomitant medication. Conversely, reported rates of dyskinesia were relatively low for patients treated with either amantadine or a dopamine agonist. These observations are consistent with the robust antidyskinetic effect of amantadine [16,17] and the lower risk of developing dyskinesia with dopamine agonist monotherapy.
The strengths of this analysis include its sample size. The ability of patients to continue with their current adjunct medications reflects routine practice. However, we acknowledge several limitations, not least the post hoc nature of the analysis and the smaller sample sizes of patients receiving COMT inhibitors and MAO-B inhibitors in combination with istradefylline and levodopa, which limit interpretation of the results. Patients with PD often experience anosognosia for dyskinesia [18]. However, peak-dose dyskinesia was used only as an inclusion criterion in study NCT01968031, whereas the presence of dyskinesia was captured only in other studies that used patient diaries and the UPDRS.
It has been suggested that istradefylline is a good option for patients whose dopaminergic treatment options may be limited by dyskinesia symptoms [19]. The results of this analysis support this rationale and demonstrate that istradefylline may be an effective adjunctive treatment in patients, independent of baseline dyskinesia or dyskinesia status, including in patients who are already taking multiple adjunctive treatments to manage their PD symptoms.
Supplementary Materials
The online-only Data Supplement is available with this article at https://doi.org/10.14802/jmd.25047.
Baseline characteristics and demographics
Patient disposition for patients treated with istradefylline 20 mg/day or 40 mg/day or placebo across the 8 randomized trials.
Notes
Conflicts of Interest
The authors have no financial conflicts of interest.
Funding Statement
This work was supported by Kyowa Kirin, Inc.
Acknowledgments
The authors and investigators would like to thank the patients and caregivers who participated in these studies. Medical writing assistance was provided by Anita Chadha-Patel PhD (ACP Clinical Communications), Michelle Jones, PhD, MWC and Anthony DiLauro, PhD, of MedVal Scientific Information Services, LLC, which were funded by Kyowa Kirin, Inc.
Author Contributions
Conceptualization: Nobutaka Hattori, Lawrence Elmer, Stuart H. Isaacson, Rajesh Pahwa, Olivier Rascol, Kapil Sethi, Fabrizio Stocchi, Lia Kostiuk. Formal analysis: Yu Nakajima. Investigation: Nobutaka Hattori, Lawrence Elmer, Stuart H. Isaacson, Rajesh Pahwa, Olivier Rascol, Fabrizio Stocchi. Funding acquisition: Hannah Cummings. Methodology: Yu Nakajima. Supervision: Hannah Cummings, Lia Kostiuk. Writing—original draft: Hannah Cummings, Lia Kostiuk. Writing—review & editing: Nobutaka Hattori, Lawrence Elmer, Stuart H. Isaacson, Rajesh Pahwa, Olivier Rascol, Kapil Sethi, Fabrizio Stocchi, Yu Nakajima.