Adverse events following botulinum toxin type A treatment in children with cerebral palsy

SIR–Naidu et al.1 examine the relationship between botulinum toxin type A (BoNT-A) dose, Gross Motor Function Classification System (GMFCS) level, and subsequent adverse events necessitating emergency department consultation or antibiotic prescription. Using logistic regression, they concluded that higher BoNT-A doses were associated with increased odds of systemic and respiratory complications independent of GMFCS level. They further theorized that some deaths following BoNT-A treatment in children with compromised oropharyngeal function may be causally related to the BoNT-A injections and now withhold or limit its use in children at GMFCS levels IV and V. However, in Western Australia we continue to treat children in all GMFCS levels with BoNT-A, with fully informed parental/guardian consent, on the basis of our own audits. The Cerebral Palsy Mobility Service (CPMS) at Princess Margaret Hospital for Children has been the sole Western Australian provider of BoNT-A treatment for children with cerebral palsy (CP) since 1995. We have audited the CPMS database which records dose, dilution, and muscle selection for all BoNT-A treatments and any adverse events, which are identified at the follow-up appointment 4 to 6 weeks after treatment. By December 2007, 2362 BoNT-A treatments for spasticity had occurred in 539 children, 434 of whom had CP. Of the 374 children with CP and a recorded GMFCS level, 44.6% (n=167) were at level I; 25.4% (n=39) level II; 10.4% (n=95) level III, 9.9% (n=37) level IV, and 9.6% (n=36) were at level V. Botox (Allergan, Irvine, CA, USA) maximum dose of 16U/kg or 300U total was used before 2007 and a maximum of 20U/kg or 450U thereafter. General anaesthetic was used for 42% of injections and primarily midazolam/EMLA (lidocaine+prilocaine) for the remainder. Fifteen adverse events were considered possibly attributable to BoNT-A in 14 children with CP. One child (GMFCS level II) had episodes of generalized weakness following the 5th and 9th series of BoNT-A injections. Another child (GMFCS level III) experienced generalized weakness and two were reported with excessive temporary local weakness. One child (GMFCS level I) reported incontinence. The remaining events were a rash in five children, vomiting in one, and localized pain in three; no event occurred in children in GMFCS level V. All resolved spontaneously and the majority (10/14) went on to have further BoNT-A treatments. Following concerns raised internationally regarding the safety of BoNT-A in children with severe impairment, a comprehensive retrospective chart review was conducted of children at GMFCS levels IV and V receiving BoNT-A from May 2006 to May 2008. Data were collected concerning BoNT-A dosage, dilution, muscle selection, comorbidities, plans for re-injection, and adverse events occurring in the month before and the month after injection, including emergency department presentation; respiratory tract infections; weakness; dysphagia and incontinence (almost all children were permanently incontinent). Sixty-four children (mean age 6y 8mo, SD 2y 4mo, range 1–19y) received 150 lower limb treatments with BoNT-A in this period. The mean Botox dose was 11.96U/kg (SD 3.43, range 2–20U/kg) injected under general anaesthetic for 44%, the remainder split between midazolam/EMLA and fentanyl/EMLA. In order of frequency, the goals of BoNT-A treatment were to assist with standing, seating, supine positioning and hygiene, hip management in very young children, and lastly, pain. Forty-two patients (66%) returned for repeat BoNT-A treatment. Of the 22 that did not return, 10 went to surgery, one required intrathecal baclofen, four moved on to adult services, and in seven BoNT-A was no longer indicated. Following the 150 treatments, six adverse events were identified, only two of which we attribute to BoNT-A, one of localized and one of generalized weakness. Both children were high functioning at GMFCS level IV without significant comorbidities and recovered spontaneously. The other four adverse events were lower respiratory tract infections in children with significant comorbidities, one of which followed BoNT-A combined with orthopaedic surgery during which the child aspirated while under anaesthetic. This respiratory infection rate of 2.7% per month (4/150) after BoNT-A compares very favourably with 7.3% in the month before injections (see Table I) and with 31% over 6 months reported in a neurologically impaired population with oromotor dysfunction.2 The 63% reduction in respiratory tract infection rate in the month after injection seen in our population is similar to the 56% reduction in hospitalizations for respiratory issues reported following BoNT-A treatments for drooling.3 In conclusion, we are not convinced that BoNT-A treatment confers any additional risk to that already present as a result of the comorbidities that frequently occur in children with severe CP. We continue to use BoNT-A in children with CP at GMFCS levels IV and V in Western Australia, documenting treatment outcomes and side effects.