Prasinezumab slows motor progression in rapidly progressing early-stage Parkinson's disease

Pathological α-synuclein aggregation is a hallmark of Parkinson’s disease (PD) and is implicated in disease progression. Prasinezumab is a monoclonal antibody targeting aggregated α-synuclein and was evaluated in early-stage PD in the PASADENA phase 2 trial. Although the primary endpoint (change from baseline to week 52 in MDS-UPDRS sum of Parts I+II+III) was not met, prasinezumab-treated participants showed less progression on the clinician-rated motor signs scale (MDS-UPDRS Part III) compared with placebo. Prior work and observational cohorts (e.g., PPMI) indicate that over a 1-year period early PD patients exhibit meaningful decline in Part III but minimal changes in Parts I and II. The authors hypothesized that prasinezumab’s effect would be more detectable within subpopulations characterized by faster motor progression, thereby increasing signal-to-noise and revealing potential treatment effects. The study pre-specified multiple subpopulations based on factors associated with faster progression (e.g., MAO-B inhibitor use, Hoehn and Yahr stage 2, diffuse malignant phenotype) and evaluated whether motor progression slowing was greater in these groups.

The paper references prior evidence on α-synuclein’s role in PD pathogenesis and progression. It cites the PASADENA phase 2 trial’s primary negative outcome as well as a separate phase 2 trial of another anti-aggregated α-synuclein monoclonal antibody (cinpanemab), which was negative on multiple endpoints. Observational data from PPMI and the De Novo Parkinson study indicate that over one year early PD patients show clinically meaningful worsening on MDS-UPDRS Part III but minimal change on Parts I and II, supporting the rationale to focus on motor signs. These prior observations frame the expectation that detecting disease-modifying effects requires sufficient progression on the endpoint of interest.

Design and participants: PASADENA Part 1 was a multicenter, multinational, randomized, double-blind, placebo-controlled phase 2 study in early-stage PD (NCT03100149). A total of 443 individuals were screened and 316 enrolled, randomized 1:1:1 to placebo, prasinezumab 1,500 mg, or prasinezumab 4,500 mg administered intravenously every 4 weeks for 52 weeks. Randomization was stratified by age (<60 vs ≥60 years), sex at birth, and MAO-B inhibitor use at baseline (yes vs no). Except for stable MAO-B inhibitor use (allowed at baseline), other symptomatic PD medications (e.g., levodopa, dopamine agonists) were not allowed at baseline, and their use during the double-blind period was discouraged unless necessary; in such cases, a pre-symptomatic-treatment visit captured MDS-UPDRS scores. Key inclusion criteria: idiopathic PD (age 40–80), bradykinesia plus resting tremor or rigidity, DaT-SPECT consistent with PD, treatment-naive or on stable MAO-B inhibitor ≥90 days. Key exclusions: atypical parkinsonism, certain familial PD genes (Parkin, PINK1, DJ-1), MMSE ≤25, use of certain PD meds beyond allowed thresholds, and prior prasinezumab exposure. Ethics: Conducted per Declaration of Helsinki and GCP with IRB/ethics approvals; written informed consent obtained. Endpoints and subpopulations: Secondary endpoints analyzed included MDS-UPDRS Parts I, II, and III. Subpopulation analyses were pre-specified; groups required ≥20% of modified intention-to-treat (mITT) participants per subgroup to be included. Primary prespecified subpopulations included MAO-B use at baseline (yes/no), Hoehn and Yahr stage (2 vs 1), REM sleep behavior disorder (RBD) status (RBDSQ ≥5 vs <5), and data-driven phenotypes (diffuse malignant vs nondiffuse malignant). Exploratory prespecified subpopulations included age at baseline (≥60 vs <60), sex, disease duration (>12 vs ≤12 months), age at diagnosis (≥60 vs <60), and motor subphenotypes (akinetic-rigid vs tremor-dominant; PIGD vs tremor-dominant). Two additional primary subpopulations (α-synuclein skin staining; DaT-SPECT SBR ipsilateral putamen <0.6 vs ≥0.6) did not meet the ≥20% threshold and were not analyzed further. Motor phenotype derivations: Tremor and PIGD scores were computed from specific MDS-UPDRS items; classification thresholds were applied to define tremor-dominant, PIGD, or intermediate. Akinetic-rigid phenotype was defined using averaged bradykinesia, rigidity, and axial items with tremor/akinetic-rigid score ratios for classification. Data-driven “diffuse malignant” phenotype required motor score >75th percentile plus ≥1 nonmotor score >75th percentile (autonomic dysfunction, sleep problems, or cognitive impairment) or all three nonmotor scores >75th percentile; others were classified as nondiffuse malignant. Statistical analysis: Analyses were performed in the mITT population using mixed models for repeated measures (MMRM). Two estimand strategies were used per ICH E9(R1): (1) Hypothetical strategy excludes data after start/increase of symptomatic therapy or MAO-B dose change and estimates treatment effects via the model covariance structure; (2) Treatment policy strategy includes all data regardless of symptomatic therapy, with analyses in OFF (12 h levodopa withdrawal) and ON states. Covariates included stratification factors (age group, sex, MAO-B use) and DaT-SPECT SBR (contralateral putamen). For analyses within a subpopulation defined by a covariate, that covariate was removed. Prasinezumab dose groups (1,500 and 4,500 mg) were pooled as no dose-response had been observed previously. Relative reduction (RR) was calculated as the difference in estimated mean change (prasinezumab minus placebo) divided by the placebo mean change. Sample size targeted ~100 per arm to detect a 3-point difference in MDS-UPDRS I+II+III with ~80% power at two-sided alpha 20%.

• Enrollment and arms: 316 participants randomized (placebo n=105; prasinezumab 1,500 mg n=105; prasinezumab 4,500 mg n=106). Groups were generally comparable at baseline.
• Overall placebo progression: Under the hypothetical strategy, mean (s.e.) change from baseline to week 52 in MDS-UPDRS Part III for the overall placebo group was 5.57 (0.90) points; Parts II and I changes were 2.75 (0.37) and 0.77 (0.30) points, respectively.
• Faster progression in selected placebo subgroups (hypothetical strategy): MAO-B at baseline 6.82 vs treatment-naive 5.04; Hoehn and Yahr stage 2: 6.34 vs stage 1: 2.17; age ≥60: 7.04 vs <60: 3.83; PIGD phenotype: 8.40 vs tremor-dominant: 4.70.
• Primary prespecified subpopulations (MDS-UPDRS Part III, hypothetical strategy; adjusted mean difference prasinezumab pooled vs placebo, 80% CI, RR): • MAO-B inhibitors at baseline (n=115): −2.66 (−4.87, −0.45); RR −39.0%. • Treatment-naive (n=201): −0.87 (−2.69, 0.94); RR −17.3%. • Hoehn and Yahr stage 2 (n=238): −2.55 (−4.19, −0.90); RR −40.2%. • Hoehn and Yahr stage 1 (n=78): +3.14 (0.32, 5.95); relative change +144.7%. • RBDSQ ≥5 (n=85): −2.76 (−5.78, 0.25); RR −35.6%. • RBDSQ <5 (n=230): −1.03 (−2.63, 0.57); RR −20.7%. • Diffuse malignant (n=59): −7.86 (−12.90, −2.82); RR −64.0%. • Nondiffuse malignant (n=257): −0.77 (−2.20, 0.66); RR −16.2%.
• MAO-B subgroup, treatment policy strategy (Part III): OFF state difference −2.60 (80% CI −4.51, −0.70); RR −54.3%. ON state difference −2.60 (−4.57, −0.63); RR −62.2%.
• No significant differences between prasinezumab and placebo on MDS-UPDRS Parts I or II across strategies and subpopulations.
• Dose groups (1,500 mg vs 4,500 mg) showed comparable results; pooled analyses reported.
• Sensitivity analyses including additional baseline covariates yielded the same pattern of results.

The exploratory analyses support the hypothesis that prasinezumab’s effect on motor sign progression (MDS-UPDRS Part III) is more apparent in subgroups characterized by faster progression (e.g., MAO-B users at baseline, Hoehn and Yahr stage 2, diffuse malignant phenotype). These findings align with the concept that detecting disease modification requires sufficient progression on the endpoint. The relationship between larger placebo-group worsening and larger relative reductions with prasinezumab suggests that longer observation periods may be required to detect effects in slowly progressing populations. The absence of differences on patient-reported motor and nonmotor scales (MDS-UPDRS Parts II and I) over 1 year, despite clinician-rated motor changes, is consistent with prior cohorts and may indicate that motor signs progress earlier or that self-reported measures change more slowly; longer trials may be necessary to assess effects on these domains.

This exploratory subgroup analysis of the PASADENA phase 2 trial suggests that prasinezumab may slow the progression of clinician-rated motor signs over 1 year in early-stage PD participants with more rapidly progressing disease. The effect size appeared larger in subgroups with greater placebo progression (e.g., diffuse malignant phenotype). Given the post hoc nature and limited duration, these findings require confirmation in additional randomized clinical trials, potentially with longer follow-up and enrichment for rapidly progressing phenotypes.

• Post hoc exploratory analysis increases risk of type I error and limits causal inference; no adjustment for multiple comparisons was applied, and 80% CIs were reported.
• Two prespecified primary subpopulations (α-synuclein skin staining and very abnormal DaT-SPECT SBR) were not analyzed due to <20% representation; α-synuclein skin staining may be confounded by technical preprocessing.
• Trial duration was 52 weeks; slowly progressing subgroups may require longer follow-up to detect effects, especially on MDS-UPDRS Parts I and II.
• Potential baseline imbalances across subgroups (e.g., disease duration, DaT-SPECT SBR, stage distribution) could confound subgroup effects despite covariate adjustments.
• Dose-response was not established; analyses pooled doses.
• Generalizability may be limited to early-stage, largely untreated or MAO-B–treated PD populations within trial inclusion criteria.