Abstract
PURPOSE: To identify factors affecting valbenazine efficacy and change in tardive dyskinesia (TD) symptoms after valbenazine discontinuation using post hoc analysis of the J-KINECT study evaluating valbenazine efficacy and safety in patients with TD. METHODS: J-KINECT comprised a 6-week, placebo-controlled, double-blind period; a 42-week valbenazine extension period; and a 4-week follow-up period. Predictors for valbenazine efficacy were analyzed using multiple regression analysis for change in Abnormal Involuntary Movement Scale (AIMS) total score and logistic regression analysis for ≥50% improvement in AIMS total score, from baseline to week 6. For factors predicting changes in TD symptoms after valbenazine discontinuation, multiple regression analysis examined the change in AIMS total score from weeks 48 to 52. RESULTS: Factors associated with valbenazine efficacy (decrease in AIMS total score at week 6) were baseline AIMS total score and interaction terms of valbenazine dose and baseline AIMS total score (adjusted R2 =0.318). Valbenazine dose was significantly associated with ≥50% improvement in AIMS total score from baseline to week 6 [odds ratio (95% confidence interval)=1.02 (1.01, 1.03), P <0.001]. Factors associated with exacerbation of TD symptoms (increase in AIMS total score) 4 weeks after valbenazine discontinuation were baseline anticholinergic use and atypical antipsychotic use (adjusted R2 =0.182). CONCLUSIONS: Although multiple regression analysis did not identify strong predictors of valbenazine efficacy and change in TD symptoms after valbenazine discontinuation, valbenazine dose was associated with meaningful improvement in TD. Rather than indicating efficacy in patients with specific factors, the data suggest a possible dose-dependent trend in improvement with valbenazine.