Predictors of real-world response in adults treated with botulinumtoxin-A for upper limb spasticity

Background: To explore predictors of treatment response in adults receiving botulinumtoxin-A (BoNT-A) for upper limb spasticity (ULS) and to describe the association between response and concomitant medication, professional physical therapy, and quality of life (QoL). Methods: This was a secondary analysis of the longitudinal Upper Limb International Spasticity study (ULIS-III) data set. Eligible patients were adults receiving/initiating BoNT-A treatment for ULS in clinical practice. The treatment response was a goal attainment scaling (GAS) T score change of ≥10 points from baseline to follow-up for each cycle. Concomitant medication was recorded. Time spent on professional physical therapy and self-management was recorded using the Upper Limb Focal Spasticity Therapy Recording (ULSTR) schedule. QoL was assessed using the EQ-5D-5L questionnaire. Baseline patient and clinical characteristics were described. Factors predictive of treatment response were explored using stepwise multivariable regression modeling (linear or logistic as appropriate for the outcome) and backward elimination (α=0.1). Results: Overall, 828/981 patients had response data; 71.9% responded to BoNT-A therapy. Potential predictors of response identified were the use of injection guidance techniques ( P =0.001), female sex ( P =0.031), and abobotulinumtoxin-A therapy (vs. onabotulinumtoxin-A; P <0.001). Compared with nonresponders, responders had marginally lower use of concomitant medication, received marginally more professional physical therapy and (subgroup only) had improved QoL. Age, duration/distribution, severity of impairment, and previous BoNT-A had no predictive value. Conclusions: Future analyses of ULIS-III data will aim to validate these findings and further explore determinants of treatment response and the interplay between care modalities.