← Back

PREMANDYSK

Effect on Dyskinesia of the Early Combination of Amantadine to Levodopa-Therapy in Parkinson's Disease: A Randomized, Placebo-Controlled Study (PREMANDYSK)

Year of Publication: 2026

Authors: Rascol O, Ory-Magne F, Meissner WG, ..., on behalf of the PREMANDYSK Study Group

Journal: Movement Disorders

Citation: Rascol O, et al. Effect on Dyskinesia of the Early Combination of Amantadine to Levodopa-Therapy in Parkinson's Disease: A Randomized, Placebo-Controlled Study (PREMANDYSK). Mov Disord. 2026;41(3):729. DOI: 10.1002/mds.70120

Link: https://doi.org/10.1002/mds.70120

Clinical Question

Does starting amantadine early — before dyskinesia develops — cut the risk of levodopa-induced dyskinesia in Parkinson's disease?

Bottom Line

Early adjunctive amantadine-IR (200 mg/day) halved the 18-month incidence of levodopa-induced dyskinesia in early PD; the benefit disappeared after discontinuation, indicating a symptomatic rather than disease-modifying mechanism, but the magnitude of prevention justifies considering amantadine-IR earlier in the disease course for at-risk patients.

        Major Points
        Amantadine-IR 200 mg/day halved 18-month LID incidence: 11% vs 22% in placebo (P=0.025; adjusted OR 0.43, 95% CI 0.19–0.98)
Time to onset of LID was significantly delayed in the amantadine-IR group (Kaplan-Meier log-rank P=0.042)
Amantadine-IR group required 70 mg/day less levodopa dose increase over 18 months (95% CI 21–119 mg, P=0.005), suggesting an antiparkinsonian symptomatic benefit
No significant difference in LID rates at end of delayed-start phase (all on amantadine-IR; 12% vs 20%, P=0.13) or after 1-month washout (16% vs 22%, P=0.23), arguing against a long-lasting disease-modifying mechanism
Exploratory benefits observed during Period 1 on freezing of gait, fatigue, and quality of life — warrant further dedicated investigation
Safety profile consistent with prior amantadine-IR reports; AEs were the main driver of dropout (14 amantadine vs 11 placebo)

      

Design

Study Type: Randomized, placebo-controlled, double-blind, parallel-arm, multicenter trial with three sequential phases: 18-month placebo-controlled (Period 1), 3-month delayed-start all-active (Period 2), 1-month washout (Period 3)

Randomization: 1

Blinding: Double-blind (participants, investigators, and site personnel blinded until end of study)

Allocation: Parallel group

Enrollment Period: June 1, 2012 to February 26, 2019

Follow-up Duration: 22 months total (18 + 3 + 1 months)

Centers: 15

Countries: France

Sample Size: 207

Analyzed: 207

Analysis: Intention-to-treat (ITT) for Period 1 primary and exploratory outcomes; per-protocol (PP) for Period 2 and 3 secondary outcomes; full analysis set (FAS) sensitivity analysis for primary outcome

Power Calculation: 80% power to detect a difference between 25% (placebo, based on prior data) and 10% (amantadine-IR) LID incidence at 18 months; one-sided type I error 5%; 184 participants needed (92/group); 202 recruited allowing 10% dropout

Registration: NCT01538329

Inclusion Criteria

Age ≥35 years
Diagnosis of Parkinson's disease
Receiving levodopa (levodopa/carbidopa or levodopa/benserazide) for ≥2 months but <1 year at dose ≥150 mg/day
Free from levodopa-associated motor complications (no motor fluctuations, no dyskinesia) per PD specialist assessment
Stable antiparkinsonian treatment for ≥1 month before randomization

Exclusion Criteria

Atypical or secondary parkinsonism
Dementia (MMSE score <26)
Behavioral disorder (score ≥3 on ECMP scale)
Treatment with antipsychotics
Prior treatment with amantadine-IR

Arms

Field	Amantadine-IR	Control
N	99	108
Intervention	Amantadine immediate-release 200 mg/day (two 100 mg capsules: one in the morning, one at midday) as add-on to levodopa for 18 months (Period 1); all participants then received amantadine-IR in Period 2 (3 months) and placebo in Period 3 (1 month)	Matched placebo capsules twice daily as add-on to levodopa for 18 months (Period 1); all participants then received amantadine-IR in Period 2 (3 months) and placebo in Period 3 (1 month)
Duration	22 months	22 months

Outcomes

Outcome	Type	Control	Intervention	HR / OR / RR	P-value
Presence or absence of peak-dose levodopa-induced dyskinesia (LID) at 18 months (end of Period 1), assessed by expert neurologist judgment and confirmed by MDS-UPDRS item IV-1 (score >0)	Primary	22% (24/108)	11% (11/99)	0.43	0.025 (primary unilateral chi-squared); 0.045 (logistic regression adjusted)
	Secondary
	Secondary				0.13 (NS)
	Secondary				0.23 (NS)
	Secondary
	Safety
	Safety
AEs were the leading cause of discontinuation; detailed AE tables not available in provided text excerpt					Adverse
Dropouts due to AE (Period 1)	Adverse

Subgroup Analysis

Exploratory outcomes showed mild but significant positive effects of amantadine-IR on freezing of gait, fatigue (Parkinson Fatigue Scale), and quality of life (PDQ-8) during Period 1; motor and non-motor fluctuations also assessed exploratorily via MDS-UPDRS IV-3 and EFNM scale

Criticisms

One-sided (unilateral) chi-squared test used for the primary outcome is less conservative than a two-sided test, lowering the threshold for declaring significance
Delayed-start and washout analyses did not reach statistical significance, leaving the potential for a disease-modifying effect unresolved
No adjustment for multiple comparisons applied across numerous exploratory outcomes, increasing risk of false-positive findings
Long enrollment period (2012–2019) may reflect evolving standard-of-care practices that complicate interpretation
Full baseline demographics (age, sex, disease duration) not available in the provided text excerpt

Funding

Toulouse University Hospital; French Ministry of Health (PHRC No. 11 253 01)

Based on: PREMANDYSK (Movement Disorders, 2026)

Authors: Rascol O, Ory-Magne F, Meissner WG, ..., on behalf of the PREMANDYSK Study Group

Content summarized and formatted by NeuroTrials.ai.

PREMANDYSK

(2026)

Objective

To investigate whether early adjunctive amantadine immediate-release (IR) 200 mg/day reduces the incidence of levodopa-induced peak-dose dyskinesia in early Parkinson's disease patients without existing motor complications.

Study Summary

• Amantadine-IR halved dyskinesia incidence at 18 months vs placebo: 11% vs 22% (P=0.025); adjusted OR 0.43 (95% CI 0.19–0.98, P=0.045)
• Amantadine-IR group required 70 mg/day less levodopa dose increase over 18 months (95% CI 21–119 mg, P=0.005)
• Time to dyskinesia onset favored amantadine-IR (log-rank P=0.042)
• No significant difference in dyskinesia rates at end of delayed-start phase (12% vs 20%, P=0.13) or after 1-month washout (16% vs 22%, P=0.23), suggesting symptomatic rather than disease-modifying effect
• Exploratory benefits on freezing of gait, fatigue, and quality of life were observed

Intervention

Amantadine immediate-release (IR) 200 mg/day (two 100 mg capsules: one morning, one midday) as add-on to stable levodopa for 18 months, followed by a 3-month delayed-start phase (all participants on amantadine-IR) and 1-month washout

Inclusion Criteria

Age ≥35, PD diagnosis, on levodopa ≥150 mg/day for ≥2 months but <1 year, free from motor complications (no LID, no motor fluctuations), stable treatment ≥1 month before randomization

Study Design

Arms: Amantadine-IR 200 mg/day (n=99) vs Placebo (n=108)

Patients per Arm: 99 vs 108

Outcome

• Primary (18 months): LID incidence 11% (amantadine-IR) vs 22% (placebo), P=0.025; OR 0.43 (95% CI 0.19–0.98)
• Time to LID onset: log-rank P=0.042 favoring amantadine-IR
• Levodopa dose increase 70 mg/day less in amantadine-IR arm (P=0.005)
• LID at delayed-start end (Period 2): 12% vs 20%, P=0.13 (NS)
• LID after 1-month washout (Period 3): 16% vs 22%, P=0.23 (NS)
• Exploratory benefits on freezing of gait, fatigue, and quality of life (Period 1)

Bottom Line

Major Points

Amantadine-IR 200 mg/day halved 18-month LID incidence: 11% vs 22% in placebo (P=0.025; adjusted OR 0.43, 95% CI 0.19–0.98)
Time to onset of LID was significantly delayed in the amantadine-IR group (Kaplan-Meier log-rank P=0.042)
Amantadine-IR group required 70 mg/day less levodopa dose increase over 18 months (95% CI 21–119 mg, P=0.005), suggesting an antiparkinsonian symptomatic benefit
No significant difference in LID rates at end of delayed-start phase (all on amantadine-IR; 12% vs 20%, P=0.13) or after 1-month washout (16% vs 22%, P=0.23), arguing against a long-lasting disease-modifying mechanism
Exploratory benefits observed during Period 1 on freezing of gait, fatigue, and quality of life — warrant further dedicated investigation
Safety profile consistent with prior amantadine-IR reports; AEs were the main driver of dropout (14 amantadine vs 11 placebo)

Study Design

Study Type: Randomized, placebo-controlled, double-blind, parallel-arm, multicenter trial with three sequential phases: 18-month placebo-controlled (Period 1), 3-month delayed-start all-active (Period 2), 1-month washout (Period 3)
Randomization: Yes
Blinding: Double-blind (participants, investigators, and site personnel blinded until end of study)
Sample Size: 207
Follow-up: 22 months total (18 + 3 + 1 months)
Centers: 15
Countries: France

Primary Outcome

Definition: Presence or absence of peak-dose levodopa-induced dyskinesia (LID) at 18 months (end of Period 1), assessed by expert neurologist judgment and confirmed by MDS-UPDRS item IV-1 (score >0)

Control	Intervention	HR/OR	P-value
22% (24/108)	11% (11/99)	0.43 (0.19 to 0.98)	0.025 (primary unilateral chi-squared); 0.045 (logistic regression adjusted)

Limitations & Criticisms

One-sided (unilateral) chi-squared test used for the primary outcome is less conservative than a two-sided test, lowering the threshold for declaring significance
Delayed-start and washout analyses did not reach statistical significance, leaving the potential for a disease-modifying effect unresolved
No adjustment for multiple comparisons applied across numerous exploratory outcomes, increasing risk of false-positive findings
Long enrollment period (2012–2019) may reflect evolving standard-of-care practices that complicate interpretation
Full baseline demographics (age, sex, disease duration) not available in the provided text excerpt

Citation

Rascol O, et al. Effect on Dyskinesia of the Early Combination of Amantadine to Levodopa-Therapy in Parkinson's Disease: A Randomized, Placebo-Controlled Study (PREMANDYSK). Mov Disord. 2026;41(3):729. DOI: 10.1002/mds.70120

View Original Publication →

PREMANDYSK

Effect on Dyskinesia of the Early Combination of Amantadine to Levodopa-Therapy in Parkinson's Disease: A Randomized, Placebo-Controlled Study (PREMANDYSK)

Clinical Question

Bottom Line

Major Points

Design

Inclusion Criteria

Exclusion Criteria

Arms

Outcomes

Subgroup Analysis

Criticisms

Funding

Ask about PREMANDYSK