TEMPO-3
(2026)Objective
To evaluate the efficacy, safety, and tolerability of tavapadon as adjunctive therapy to oral levodopa in adults with Parkinson disease experiencing motor fluctuations.
Study Summary
• Daily off-time was significantly reduced with tavapadon vs placebo (−1.88 vs −0.93 hours; difference, −0.94 hours; 95% CI, −1.48 to −0.41; P <.001)
• Most adverse events were nonserious (93.2%) and mild to moderate in severity; common AEs with tavapadon included nausea (14.3%), dyskinesia (10.0%), and dizziness (7.6%)
Intervention
Tavapadon 5-15 mg once daily (flexible-dose) vs placebo, both adjunctive to stable oral levodopa for 27 weeks
Inclusion Criteria
Adults aged 40-80 years with Parkinson disease (UK Brain Bank criteria), modified Hoehn & Yahr score 2-3 in on-state, ≥2.5 hours daily off-time, stable levodopa dose ≥400 mg/day (≥3-4 doses daily), good levodopa response
Study Design
Arms: Tavapadon 5-15 mg once daily vs Placebo
Patients per Arm: 252 (tavapadon), 255 (placebo)
Outcome
• Key secondary: Tavapadon reduced daily off-time by 0.94 hours vs placebo (P <.001; Cohen d = 0.35)
• MDS-UPDRS Part II improved by 1.2 points with tavapadon vs placebo (P = .02); Part III improved by 2.4 points (P = .01)
Bottom Line
In adults with Parkinson disease experiencing motor fluctuations on levodopa, adjunctive tavapadon (5-15 mg once daily) significantly increased daily good-on-time by 1.1 hours (P <.001) and reduced off-time by 0.94 hours (P <.001) compared with placebo, with a favorable safety profile characterized by mostly mild-to-moderate adverse events.
Major Points
- Tavapadon significantly increased daily on-time without troublesome dyskinesia (good-on-time) by 1.10 hours vs placebo (1.70 vs 0.60 hours; P <.001; Cohen d = 0.40)
- Daily off-time was significantly reduced with tavapadon vs placebo (−1.88 vs −0.93 hours; difference −0.94 hours; P <.001; Cohen d = 0.35)
- The increase in good-on-time was primarily driven by increased on-time without dyskinesia
- MDS-UPDRS Part II (activities of daily living) improved by 1.2 points with tavapadon vs placebo (nominal P = .02)
- MDS-UPDRS Part III (motor function) improved by 2.4 points with tavapadon vs placebo (nominal P = .01)
- Most adverse events (93.2%) were nonserious and mild-to-moderate in severity
- Common AEs with tavapadon (≥5%) were nausea (14.3%), dyskinesia (10.0%), and dizziness (7.6%)
- Discontinuation due to AEs was higher with tavapadon (17.1%) vs placebo (9.1%)
- No notable differences between groups in rates of impulse control disorders, somnolence, or measured orthostatic hypotension
Study Design
- Study Type
- Phase 3, multicenter, double-blind, placebo-controlled, parallel-group randomized clinical trial
- Randomization
- Yes
- Blinding
- Double-blind (participants, investigators, and all blinded sponsor personnel)
- Sample Size
- 507
- Follow-up
- 27 weeks treatment period plus 4-week safety follow-up (total ~35 weeks)
- Centers
- 148
- Countries
- 14 countries
Primary Outcome
Definition: Change from baseline to week 26 in total daily on-time without troublesome dyskinesia (good-on-time) based on 2-day average of Hauser diary
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 0.60 hours increase (LSM) | 1.70 hours increase (LSM) | - (0.60 to 1.70) | <.001 |
Limitations & Criticisms
- Study population was predominantly White (96.8%), limiting generalizability to other racial and ethnic groups
- 27-week treatment duration may not fully capture long-term safety and efficacy, particularly for adverse events that emerge with prolonged treatment (e.g., impulse control disorders)
- Higher discontinuation rate in tavapadon group (36.9%) vs placebo (19.2%), primarily due to adverse events
- No direct comparison with currently available D2/D3 dopamine agonists or other adjunctive therapies
- Trial conducted during COVID-19 pandemic, which may have affected enrollment times and discontinuation rates
- No significant differences observed in quality of life measures (PDQ-39, EQ-5D-5L)
- MDS-UPDRS assessments conducted 2-3 hours after last levodopa dose may not fully capture peak effects
- Secondary outcome analyses not corrected for multiplicity; P-values should be interpreted as nominal
Citation
JAMA Neurol. Published online March 20, 2026. doi:10.1001/jamaneurol.2026.0577