Factors predictive of the development of Levodopa‐induced dyskinesia and wearing‐off in Parkinson's disease

The Stalevo Reduction in Dyskinesia Evaluation in Parkinson's Disease (STRIDE-PD) study compared the initiation of levodopa (l-dopa) therapy with l-dopa/carbidopa (LC) versus l-dopa/carbidopa/entacapone (LCE) in patients with Parkinson's disease. In the current study, the STRIDE-PD study population was investigated to determine the effect of l-dopa dose and other risk factors on the development of dyskinesia and wearing-off. Patients were randomized to receive LCE (n=373) or LC (n=372). Blinded assessments for dyskinesia and wearing-off were performed at 3-month intervals for the 134- to 208-week duration of the study. The patients were divided into 4 dose groups based on nominal l-dopa dose at the time of onset of dyskinesia (or at study conclusion if there was no dyskinesia): group 1, <400 mg/day (n=157); group 2, 400 mg/day (n=310); group 3, 401 to 600 mg/day (n=201); and group 4, >600 mg/day (n=77). Similar analyses were performed with respect to wearing-off and any motor complication. The times to onset and frequency of dyskinesia, wearing-off, or any motor complication were compared using the log-rank test (overall trend test) and a Cox proportional hazards model (pairwise comparisons). A stepwise Cox proportional hazards model was used to screen predictive factors in a multivariate analysis. The risk of developing dyskinesia and wearing-off increased in an l-dopa dose-dependent manner (P<0.001 for both). Analyses using l-dopa equivalent doses produced comparable results. Factors that were predictive of dyskinesia, in rank order, were: young age at onset, higher l-dopa dose, low body weight, North American geographic region, LCE treatment group, female gender, and more severe Unified Parkinson's Disease Rating Scale (UPDRS) Part II. Multivariate analyses identified similar predictors for wearing-off but included baseline UPDRS Part III and excluded weight and treatment allocation. The risk of developing dyskinesia or wearing-off was closely linked to l-dopa dose. The current results suggest that physicians should use the lowest dose of l-dopa that provides satisfactory clinical control to minimize the risk of both dyskinesia and wearing-off.