Abstract
Background: Adenosine A2A receptor antagonists—istradefylline, preladenant, and tozadenant—are adjunctive therapies for motor fluctuations in Parkinson’s disease (PD). However, the comparative efficacy and safety of different agents and doses remain unclear. Methods: We conducted pairwise and dose-stratified network meta-analyses (NMA) of randomized controlled trials (RCTs) evaluating A2A antagonists regarding OFF time, ON time with troublesome dyskinesia, Unified Parkinson’s Disease Rating Scale Part III (UPDRS III) scores, and adverse events (AEs). Random-effects models using restricted maximum likelihood (REML) accounted for within- and between-study variability. Heterogeneity and inconsistency were assessed using τ2, I2, and Q statistics. Results: Pairwise meta-analysis showed significant reductions in OFF time versus placebo (mean difference [MD] = −0.35; 95% confidence interval [CI]: [−0.52, −0.18]), with benefits for istradefylline (MD = −0.29; 95% CI: [−0.57, −0.01]) and preladenant (MD = −0.36; 95% CI: [−0.56, −0.15]); tozadenant was ineffective. NMA of OFF time (11 RCTs, 45 comparisons) found modest benefits for istradefylline 20 mg (MD = −0.33; 95% CI: [−0.66, 0.01]) and 40 mg (MD = −0.47; 95% CI: [−0.83, −0.11]) with moderate-certainty evidence; preladenant and tozadenant showed low or very low certainty. For ON time with troublesome dyskinesia (6 studies, n = 1439), istradefylline 40 mg significantly increased ON time (MD = 0.61; 95% CI: 0.14–1.07), while 20 mg showed a positive trend; preladenant was ineffective. UPDRS III scores (8 RCTs, 31 comparisons) showed no significant improvements, with substantial heterogeneity (τ2 = 2.107; I2 = 85.0%). In safety analyses (13 studies, n = 5364), istradefylline 20 mg (OR = 1.34; 95% CI: 1.04–1.72) and preladenant 10 mg (OR = 1.37; 95% CI: 1.05–1.80) increased AEs; other doses were similar to placebo. The network heterogeneity was low (I2 = 10.1%). Conclusions: Istradefylline 40 mg provided modest improvements in the OFF and ON times with moderate-certainty evidence; preladenant and tozadenant showed limited or uncertain efficacy. No agents significantly improved UPDRS III scores, and some doses increased AEs. Evidence is constrained by heterogeneity, imprecision, and methodological limitations. High-quality, large RCTs are needed to clarify dose–response effects and confirm the clinical role of A2A antagonists in PD.