FormalPara Key Summary Points

Why carry out the study?

Levodopa is the mainstay of Parkinson’s disease (PD) treatment; however, high doses are a risk factor for motor complications.

Istradefylline (IST) is an A2A receptor antagonist used as adjunctive treatment to levodopa in PD patients with off episodes; however, its impact on the increase in levodopa dosing is unclear.

In this study, we hypothesized that IST would modulate increases in the levodopa dose and result in less levodopa use overall.

What was learned from this study?

IST therapy had a significant effect on modulating levodopa dose increases, with lower overall levodopa use compared with no IST therapy.

IST therapy significantly improved device-evaluated motor activities to 36 weeks; no effect was seen in the group without IST therapy.

IST therapy adjunctive to levodopa may provide greater benefit compared with increased levodopa doses alone.

Introduction

Parkinson’s disease (PD) is a neurodegenerative disorder associated with motor dysfunction, including akinesia, resting tremor, and rigidity, as a result of dopaminergic neuronal degeneration [1]. Levodopa is the mainstay treatment for motor dysfunction in PD, by supplementing decreased dopamine levels. However, levodopa treatment is usually associated with motor fluctuations, which present a problem for PD patients [2]. Furthermore, the medium spiny neurons in the striatum connect not only with dopaminergic neurons but also with cortical glutamatergic and cholinergic interneurons. Non-dopaminergic modulation is therefore also a useful therapeutic approach for patients with PD. In this context, several non-dopaminergic drugs, including anti-cholinergic drugs, amantadine [3], zonisamide [4], and istradefylline (IST) [5], have been developed for the treatment of PD.

The adenosine A2A receptor predominantly localizes to the striatum and notably modulates the indirect pathway, which is important for the control of voluntary movement [6]. Patients with PD have hyperactivation of this indirect pathway, leading to reduced voluntary movements and bradykinesia [7]. The adenosine A2A receptor antagonist IST can improve parkinsonism by normalizing basal ganglia function [8], and is currently indicated as adjunctive treatment to levodopa/decarboxylase inhibitors in patients with PD with wearing-off/off episodes [5].

Previous nonclinical and clinical results also indicated that adjunctive IST may have long-lasting anti-Parkinsonian effects without the need for increased levodopa doses [9,10,11]. However, the ability of IST to prevent the elevation of levodopa dose in the treatment of patients with PD experiencing wearing-off remains unclear. To address this question, we conducted a multicenter, open-label, randomized, parallel-group controlled study (ISTRA ADJUST PD) [12]. The objective of this trial was to investigate the effect of adjunctive IST on the cumulative dose of medications containing levodopa in PD patients experiencing wearing-off. Additionally, we employed a triaxial accelerometry monitoring system to reduce the risk of bias, especially performance bias, in the open-label assessment of the effectiveness of IST in relation to the daily movements of patients. The efficacy and safety of IST were also evaluated.

Methods

Details of the methods can be found in the previously published study protocol [12].

Study Design

A 37-week, multicenter, open-label, randomized, parallel-group controlled study was conducted as the ISTRA ADJUST PD study from February 2019 to May 2022 at 20 sites in Japan (with the registration period starting from May 2019 lasting until November 2020). Most observations were carried out at 4- or 12-week intervals throughout the study period.

Standard Protocol Approvals, Registrations, and Patient Consent

We carried out the present study in compliance with the Japan Clinical Trials Act and all related national and international guidelines for human trials, and with the Declaration of Helsinki. The Juntendo University Certified Review Board (J18-006) reviewed and approved the study protocol (Version 6.0, 1 November 2021) and all other study documentation (approval number CRB3180012). This trial was registered with the Japan Registry of Clinical Trials (jRCTs031180248).

All patients provided written informed consent before enrollment. We anonymized all patient data so that no personal information was associated with the wearable devices.

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

The study protocol has been published previously [12], and the statistical analysis plan file has been uploaded as a supplemental file (Data S1).

Inclusion and Exclusion Criteria

Inclusion Criteria

Eligible patients met the following criteria: receiving levodopa-containing medications ≥ three times daily (total daily dose of levodopa, 300–400 mg); experiencing wearing-off; age, 30–84 years; PD diagnosis in accord with the International Parkinson and Movement Disorder Society criteria; stage ≤ 3 modified Hoehn & Yahr scale (mH&Y) (on); and provision of written informed consent. Patients concomitantly receiving anti-PD drugs other than levodopa (e.g., dopamine agonists, catechol-O-methyltransferase inhibitors, and monoamine oxidase type B inhibitors) were also included in this study [12].

Exclusion Criteria

Patients were excluded if they met any of the following criteria: prior treatment with IST; taking any study drug ≤ 4 months prior to enrollment; having dementia or a Mini-Mental State Examination (MMSE) Japanese ver. score ≤ 23; previous neurosurgery for PD such as stereotactic surgery, deep brain stimulation, gamma knife; ongoing or prospective treatment with levodopa/carbidopa hydrate enteral suspension; initiation of PD treatment or any changes in regimen ≤ 4 weeks prior to enrollment; taking strong CYP3A4 inhibitors such as itraconazole and clarithromycin ≤ 14 days prior to enrollment; moderate or severe hepatic disorder; lactation or pregnancy; and the investigator’s discretion for ineligibility.

Randomization and Masking

Patients were randomized in a 1:1 ratio to the IST group (20 mg/day, escalating to 40 mg/day) or control group (without IST treatment). Randomization was performed centrally using a minimization method by computer allocation, with stratification according to age (< 60 or ≥ 60 years), presence or absence of dyskinesia, and levodopa equivalent dose (< 400 or ≥ 400 mg/day). Participants and study investigators were not blinded to treatment allocation.

Procedures

Patients in the IST group received orally administered IST (20 mg tablet, once daily) starting at week 0. The IST daily dose was increased to 40 mg at week 1 if the patient tolerated the treatment well and sustained motor symptoms. If treatment with 40 mg IST was not tolerated, dose reduction was allowed. The dosage of levodopa-containing drugs was increased by 50 mg/day at week 0 in patients in the control group. The attending physician then assessed whether the dose needed to be increased by 50 mg/day every 4 weeks, based on the following criteria: an increase of 50 mg/day was indicated if the clinical global impression of severity (CGI-S) scale was ≥ 4 on the observation day. If an intolerable adverse drug reaction (ADR) occurred because of the dose increase, the dose of medication containing levodopa was reduced at the discretion of the physician. The dose or dosing regimens of other adjunctive anti-PD drugs were not changed ≤ 4 weeks prior to enrollment and throughout the 37-week treatment period. The dose of a specific drug was reduced in the event of intolerable ADRs causally related to that drug.

Outcomes

The primary endpoint was the cumulative additional dose of medications containing levodopa throughout the treatment period of 37 weeks as the area under the curve (AUC) in patients in the IST group compared with that in the control group.

Secondary efficacy endpoints were comparisons of additional levodopa doses on each observation day in weeks 4–36 in the IST and control groups; the number of days until the first dose increase after week 4; CGI-S score; patient global impression of severity (PGI-S) score; mH&Y staging scale (on/off) score; Movement Disorder Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) [13] Part I score (non-motor experiences of daily living), Part II score (motor experiences of daily living), Part III score (motor examination), and Part IV score (motor complications); and Parkinson’s Disease Questionnaire-39 (PDQ-39).

During the study, patients wore a wristband-type triaxial accelerometry system (UW-301BT, Hitachi Systems, Ltd., Tokyo, Japan) on their non-dominant wrist for 7 days every 12 weeks. Data on movement frequency and intensity, and gait (step count, pace, and laterality) were obtained from the wearable device. Correlations between CGI-S and other outcomes such as MDS-UPDRS Parts I, II, III, and IV, PDQ-39, PGI-S, mH&Y, and device data were evaluated to confirm whether decisions on the dose escalation differed based on CGI-S and other outcomes.

Secondary safety endpoints were adverse events and ADRs classified based on System Organ Class and Preferred Terms defined in the Medical Dictionary for Regulatory Activities (Japanese edition, version 24.1). Correlations between the motor information obtained using the wearable device and the MDS-UPDRS Part II or III were evaluated as exploratory endpoints.

Statistical Analysis

To the best of our knowledge, no previous studies have investigated the effect of IST treatment on the flexibility of the doses of medications containing levodopa. In the present study, we therefore utilized real-world data from the medical claims database (Medical Data Vision Co., Ltd., Tokyo, Japan) to estimate the minimum between-group difference by simulating cumulative levodopa doses in PD patients (non-IST treatment) using Mann–Whitney U tests with 50 patients in each group, ensuring a power of 80% at the two-sided significance level of 5%. Under these conditions, a between-group difference of approximately 21.3% over 9 months (270 days) was estimated as the AUC for cumulative additional levodopa doses between groups (data on file; Kyowa Kirin Co., Ltd., Tokyo, Japan). It was assumed that IST might reduce the levodopa dose escalation of medications containing levodopa by at least 20% on the basis of the mean additional levodopa dose over 9 months (approximately 265 mg) in a previous study [14], and recommendation from the Expert Medical Advisory board for this study based on the clinical experiences in Japan. Based on the above, we set the sample size at 111 patients, assuming a 10% dropout rate and to ensure that 50 patients were evaluable in each group in the efficacy analyses.

The efficacy and safety analysis sets included all patients except those who withdrew consent before the start of the observation period (week 0), were withdrawn by the study investigator, or those in the IST group who did not initiate IST administration.

The Mann–Whitney U test was used for between-group comparisons for the primary and secondary endpoints. For the primary endpoint, we calculated the cumulative additional dose (AUC throughout the 37-week treatment period) as the total dose (daily dose × number of days) added to the dose of medications containing levodopa (300–400 mg/day) at randomization. Wilcoxon’s signed-rank test was used to evaluate changes from baseline. The log-rank test and Cox proportional hazards model were used to compare the number of days from the start of the observation (week 0) to the time of levodopa dose increase between the groups, and time-to-event curves were prepared using the Kaplan–Meier method. The correlation of the secondary endpoint scores was examined using Spearman’s correlation coefficient.

A two-sided 5% significance level was set for between-group comparisons, and two-sided 95% confidence intervals were calculated. We made no imputation for missing data. A detailed statistical analysis plan was prepared before the database was finalized and locked. In this study, SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) was used for statistical analyses.

Results

Patients

A total of 115 patients were enrolled between May 2019 and November 2020 (Fig. 1). One patient withdrew from the trial before the baseline assessment, and 114 patients were thus randomized (participation rate 99%) to the IST or control group (57 patients per group). Five patients were excluded from the IST group because they did not start IST treatment, withdrew consent, or were withdrawn by the study investigator, and four patients were excluded from the control group because they did not meet the inclusion criteria or met the exclusion criteria, withdrawal of consent, or were withdrawn by the study investigator. Therefore, 52 and 53 patients in the IST group and control group, respectively, were included in the efficacy and safety analyses. During the study period, three patients in the IST group failed to increase the dosage of IST to 40 mg, four patients in the IST group dropped out of the study because of withdrawal of consent, adverse events, or withdrawal by the study investigator, and five patients in the control group dropped out because of withdrawal of consent. Forty-eight patients in each group thus completed the 37-week trial. Overall, there were no significant differences in baseline demographic and clinical characteristics between the IST group and control group (Table 1).

Fig. 1
figure 1

Disposition of the study participants (CONSORT flow diagram). IST istradefylline

Full size image
Table 1 Patient characteristics
Full size table

Outcomes

Primary Outcome

The mean (± standard deviation [SD]) cumulative additional levodopa dose as the AUC for 37 weeks was lower in the IST group (3229.8 ± 5692.7 mg) compared with that in the control group (15,056.6 ± 5187.1 mg) (p < 0.0001).

Secondary Outcomes

The levodopa dose increase was significantly reduced over 36 weeks in the IST group compared with the control group (p < 0.0001). The mean (± SD) add-on levodopa dose was significantly lower in the IST group (25.0 ± 40.2 mg/day) than in the control group (73.6 ± 43.9 mg/day) at week 36 (p < 0.0001) (Fig. 2).

Fig. 2
figure 2

Additional levodopa doses in the IST and control groups. Data shown as mean ± SD. *p < 0.0001, Mann–Whitney U test (vs. control). Add-on dose at week 36 (mean ± SD): IST group = 25.0 ± 40.2 mg/day; control group = 73.6 ± 43.9 mg/day (p < 0.0001, Mann–Whitney U test). IST istradefylline, SD standard deviation

Full size image

The percentage of patients who added levodopa after 4 weeks tended to be lower in the IST group than in the control group, but there was no significant difference between the two groups (Fig. 3). The levodopa dose was increased in 19 patients in the IST group (36.5%) and 25 patients (47.2%) in the control group at week 36. The mean (± SD) number of days until the first levodopa dose increase after week 4 tended to be longer in the IST group (202.1 ± 87.6) than in the control group (191.8 ± 88.8).

Fig. 3
figure 3

Number of days until first levodopa dose increase after week 4. Vertical axis indicates percentage of patients who first increased the levodopa dose after week 4. Data shown as mean ± SD. Number of days until first levodopa dose increase: IST group = 202.1 ± 87.6 days; control group = 191.8 ± 88.8 days (p = 0.2912, log-rank test; p = 0.2946, Cox proportional hazards model). IST istradefylline, SD standard deviation

Full size image

The CGI-S score, as an indicator of levodopa dose increase, decreased at all time points after week 4 (vs. week 0) in both the IST and control groups (p < 0.05), with no significant difference between the groups (Table S1). There were no significant differences in the scores for other secondary outcomes between the two groups during the study.

MDS-UPDRS Part I only improved in the IST group, whereas MDS-UPDRS Parts III and IV improved in both groups from baseline (week 0) to week 36 (all p < 0.05) (Table 2). MDS-UPDRS Part II, mH&Y (on/off), PDQ-39, and PGI-S showed no significant improvement at week 36 compared with those at baseline in either group (Table 2, Table S1). However, temporary but significant improvements were observed for mH&Y (off), PDQ-39, or PGI-S in the IST group and MDS-UPDRS Part II in the control group (p < 0.05, Table 2 and Table S1).

Table 2 Changes in MDS-UPDRS Parts I–IV, mH&Y (on, off), and PDQ-39 scores from weeks 0 to 12, 24, and 36
Full size table

Motor activity indicators, including motion frequency and intensity measured using the triaxial accelerometry system, improved significantly only in the IST group at week 36 compared with that at baseline (all p < 0.05, Table 3). Furthermore, light exercise intensity (≥ 1.5, < 3 metabolic equivalents) time increased significantly, whereas low exercise intensity (< 1.5 metabolic equivalents) time decreased significantly (both p < 0.05, Table 3). CGI-S, as an indicator of levodopa dose increase, correlated with PGI-S and MDS-UPDRS Part III (Table S2). Most of the motor activity parameters from wearable devices were correlated with MDS-UPDRS Part III and moderately correlated with MDS-UPDRS Part II (Table S3). However, the number of steps per day, and duration of moderate or more-intense exercise (≥ 3 metabolic equivalents) were not correlated with MDS-UPDRS Part II (Table S3).

Table 3 Change in motor activity from weeks 0 to 36 measured by a wearable device (triaxial accelerometry system)
Full size table

ADRs were observed in 15 patients (28.8%) and seven patients (13.2%) in the IST and control groups, respectively (Table 4), and one patient discontinued the study in the IST group because of dyskinesia. The most common ADRs were dyskinesia, somnolence, and nausea (5.8% each) in the IST group, and nausea (3.8%) in the control group (Table 4). There were no serious ADRs in either group. Overall, there was no clear difference between the groups in the type of ADRs.

Table 4 Adverse drug reactions
Full size table

Classification of Evidence

This study provides level II evidence that IST treatment is effective in the adjustment of levodopa dose escalation with less cumulative levodopa use in PD patients with wearing-off.

Discussion

This study demonstrated a significant reduction in the cumulative additional levodopa dose throughout the 37-week treatment period in patients with PD receiving IST. There was no notable difference in the backgrounds of participants treated with and without IST, indicating appropriate randomization. MDS-UPDRS Part I and device-evaluated motor activities were significantly improved from baseline to 36 weeks only in the IST group. We also confirmed that CGI-S correlated with PGI-S and MDS-UPDRS Part III in both groups, suggesting that it was appropriate to judge increased levodopa dose based on CGI-S. Although IST suppressed levodopa dose escalation in PD patients, ADRs were more common in the IST group than in the control group. However, there were no serious ADRs in either group and no difference in dropout rates between the groups, indicating that IST was well tolerated for the treatment of PD.

The current study showed that the cumulative additional levodopa dose was significantly reduced by IST throughout the 37-week treatment period. The increase in daily levodopa dose from baseline to 36 weeks was 25.0 ± 40.2 mg in the IST group and 73.6 ± 43.9 mg in the control group. Given that the first levodopa administration at week 0 was additional 50 mg in the control group, IST may have a similar impact on preventing escalation of levodopa dose as an additional 50 mg/day of levodopa. There was no significant difference between the two groups in the number of days until the first levodopa dose increase after week 4. However, fewer patients in the IST group had a first levodopa dose increase before week 36 (IST: 36.5% vs. control: 47.2%) and the largest difference was found at week 28 (IST: 25.3% vs. control: 38.2%). The observation periods in most previous studies were approximately 12 weeks [8], and the long-term efficacy of IST for parkinsonism was thus unclear. In contrast, the present results indicated that IST may not only be useful for treating parkinsonism, but may also reduce the need for increased levodopa doses for long periods. Watts et al. investigated whether the addition of ropinirole prolonged-release delayed the onset of levodopa-induced dyskinesia, compared with levodopa administration alone [14]. Their study design was similar to the current study, but the additional levodopa dose until 6–9 months was 245 mg, which was approximately three times that in the present study. This suggests that the parkinsonism experienced by the participants in our study might have been too mild to require an increased dose of levodopa. Further studies are thus needed to confirm whether IST can prevent increases in levodopa dosage in patients with advanced PD.

Several clinical endpoints were similarly improved in both groups compared with those at baseline, and no endpoints deteriorated. Investigator objective scores, including MDS-UPDRS Part III and CGI-S, both decreased significantly during the trial period in both groups. Notably, the MDS-UPDRS Part III fell by > 3.25 in the IST group, which was considered a clinically meaningful improvement [15]. Furthermore, the mH&Y (off) was significantly improved in the IST group at weeks 12 and 24.

Previous studies showed improvements in MDS-UPDRS Part II following treatment with 40 mg IST, but most only revealed improved on-phase motor function and off-time reduction [16, 17]. IST is a non-dopaminergic drug and may thus have potential effectiveness against levodopa-resistant symptoms. The present study also evaluated patient-reported outcomes, which reflect patient quality of life (QoL) [18, 19]. The PGI-S improved significantly at weeks 4, 16, 20, and 24, and the MDS-UPDRS Part I improved at weeks 12 and 36 in the IST group. Previous studies showed that IST improved non-motor symptoms such as daytime sleepiness, apathy, and fatigue, which are difficult to treat with levodopa [20,21,22]. These effects may contribute to the improvement of MDS-UPDRS Part I. Although MDS-UPDRS Part II, which includes QoL-related items [23], was ameliorated in the control group at weeks 12 and 24, PDQ-39 decreased in the IST group at week 12. Overall, patients in the IST group reported more improved patient-reported outcome items than those in the control group, suggesting that the administration of IST may provide better QoL in patients with PD.

This study was designed as an open-label trial and as such was subject to critical performance bias. We therefore analyzed data from a triaxial accelerometry monitoring system, which did not include evaluator subjectivity, to overcome this issue and confirmed a correlation between CGI-S and accelerometry parameters, such as frequency of motion while awake and intensity of daily exercise. These parameters improved significantly from baseline to week 36 in the IST group and were correlated with MDS-UPDRS Part III. IST may thus have a greater effect on motor performance than levodopa administration alone for 36 weeks. Furthermore, the improvement in motor activity parameters from wearable devices may reflect this amelioration in off-time. Walk pitch and walk balance decreased in the IST group; however, these parameters were not correlated with CGI-S or MDS-UPDRS Part III, suggesting that deterioration of these parameters might have minimal influence on the patient’s activities of daily life. A previous single-arm, open-label, prospective, multicenter study revealed that the administration of IST ameliorated the gait-related total scores of MDS-UPDRS Part III from baseline, with significant improvements in gait, gait freezing, and postural stability [24]. In this context, the decrease in walk pitch may reflect the improvement of small steps. IST might be effective on gait disturbance, which is sometimes difficult to treat with levodopa.

This study had several limitations. First, this was an open-label study with performance bias, and the increase in levodopa dose may have been affected by the participants’ judgment. If a patient disliked the idea of increasing their drug treatment, the administration of levodopa may have been suppressed. In the present study, there was no improvement in accelerometry parameters in the control group. This might be associated with a hesitation to increase the levodopa dose. However, the participation by many centers reduces investigator bias. Second, the diagnosis of PD was based on clinical features rather than pathology, and some patients with atypical parkinsonism might thus have been included. However, there was no notable difference in the backgrounds of the two groups, suggesting that this did not influence the analysis. Third, we did not validate the association between the accelerometry parameters and motor symptoms in PD patients. However, the accelerometry parameters were correlated with CGI-S and MDS-UPDRS Part III scores, indicating that the system may reflect motor performance in patients with PD. Fourth, we used CGI as a criterion for levodopa dose escalation based on methods such as those described by Watts et al. [14], who determined the optimal dose while observing the patient's condition. Finally, we set the research period as 37 weeks because this period was long enough for levodopa dose escalation to occur based on a report by Watts et al. [14]. Furthermore, levodopa dose escalation without changing other drugs over a long period could lead to problems, such as the onset of levodopa-induced ADRs. However, an observation period of 37 weeks might not provide adequate time to identify alterations in the necessary levodopa dose [12]. Therefore, long-term studies will be needed in the future. Despite these limitations, we believe that the current results show that IST administration can reduce levodopa dose escalation in patients with PD.

Conclusion

Treatment with IST effectively reduced levodopa dose escalation in patients with PD, resulting in less cumulative levodopa use during the study period. Data from a wearable accelerometry device suggested that IST resulted in greater objective improvements in motor function than increased levodopa dose. This study clarified the effectiveness of adjunctive IST compared with increased doses of levodopa in PD patients with wearing-off.