Objectives: Botulinum toxin A (BoNT-A) can enable achievement of treatment goals in upper limb spasticity (ULS). However, real-world data are limited. We examine health and economic outcomes of BoNT-A in ULS in the real-world.

Methods: ULIS III (NCT02454803) was a multicenter, non-interventional, prospective, 2-year study in ULS on abobotulinumtoxinA (aboBoNT-A), onabotulinumtoxinA (onaBoNT-A), or incobotulinumtoxinA (incoBoNT-A). Patients changing BoNT-A during follow-up were excluded from this analysis. Response was defined as ≥10-point increase in cumulated GAS-T at the last visit versus baseline. Improvement ≥the minimal clinically important difference (MCID of 0.074) in quality of life (QoL) was assessed with EQ-5D-5L (US, Australia). Annual treatment costs in 2021€ were estimated for France, Italy, Portugal. Multivariate logistic regression of response and linear regression of annual treatment costs evaluated the effect of BoNT-A on health and economic outcomes.

Results: 828 patients provided cumulated GAS-T score data (555 aboBoNT-A, 196 onaBoNT-A, 77 incoBoNT-A). Response rates with onaBoNT-A over the study period were significantly lower than those with aboBoNT-A (adjusted OR 0.47[0.33-0.68]) and no significant difference was found between aboBoNT-A and incoBoNT-A (p=0.864). A favorable response to BoNT-A was associated with higher likelihood of clinically significant improvement in QoL; improvement in excess of the MCID was observed in 46% of responders versus 30% of non-responders at last study visit (versus baseline). A significant reduction in annual treatment costs was observed between aboBoNT-A and incoBoNT-A in France (adjusted cost difference -€592.54 [p<0.001] with vial sharing, -€541.09 [p<0.001] without, N=88), with no significant difference in Italy or Portugal; aboBoNT-A and onaBoNT-A costs were not significantly different in three countries.

Conclusions: These real-life data suggest aboBoNT-A can be cost-effective in ULS. At similar treatment costs, aboBoNT-A was associated with a higher responder rate than onaBoNT-A, which may be linked to improved QoL. AboBoNT-A was also cost-saving versus incoBoNT-A in France.