Adverse Event Profiles of High Dose Botulinum Toxin Injections for Spasticity

Background: Botulinum toxin injections are effective in relieving focal spasticity resulting from upper motor neuron injuries. Doses approved in the United States for onabotulinumtoxinA and incobotulinumtoxinA are up to 400 units, yet higher doses are often used. Objective: To determine differences in risk of adverse events when using higher (>600 units) as compared to lower doses within clinically applicable categories; the difference in adverse events between types of botulinum toxin-A, and any association of the injection of cervical muscles with increased risk for adverse events. Design and Setting: Retrospective analysis of injections performed over a 3-year period at a freestanding rehabilitation hospital network. Participants: Persons with spasticity or dystonia undergoing ona- and/or incobotulinumtoxinA injections. Interventions: Not applicable. Main Outcome Measures: Adverse events for injections were divided into the three dose ranges (≤400 units, 401-600 units, or > 600 units). Results: 889 injections in 342 patients met inclusion criteria with 65% ≤400 units, 21% 401-600 units, and 14% >600 units. Adverse events were not significantly increased in doses of 401-600 units relative to ≤400 units (OR 0.97, 95% CI 0.31, 2.98). Doses of toxin over 600 units were associated with significantly increased relative risk of adverse events (OR 2.98, 95% CI 1.12, 8.13). There were no significant differences between adverse event rates for onabotulinumtoxinA or incobotulinumtoxinA (P >.99). Inclusion of cervical muscles in isolation did significantly increase the risk of adverse events (OR 4.21, 95% CI 1.15, 15.46). Conclusion: Risk for adverse events were not significantly increased in doses of ona- and/or incobotulinumtoxinA up to 600 units, suggesting that the current 400 units upper bound of approved dose may need to be reexamined. Doses above 600 units were found to increase the rate of adverse effects and clinical benefit versus risk should be taken into account. Level of Evidence: III.