Safety of Botulinum Toxin Type A forChildren With Nonambulatory CerebralPalsyPriya Edwards, MBBS, FRACP, FAFRMa,b, Leanne Sakzewski, PhD, BOccThyb, Lisa Copeland, MBBS, FRACP, FAFRMa,b,Laura Gascoigne-Pees, BA (Hons)a,b, Kim McLennan, MBBS, FRACP, FAFRMa,b, Megan Thorley, MScMed, BScOTa,b,Megan Kentish, BSc PTa,b, Robert Ware, PhDc,d, Roslyn N. Boyd, PhD, MSc (Physiotherapy)b

abstractOBJECTIVE: To determine safety of intramuscular botulinum toxin A (BoNT-A) injections to reducespasticity and improve care and comfort of nonambulatory children with cerebral palsy (CP).

METHODS: Nonambulatory children with CP were randomly allocated to receive either BoNT-A(n = 23) or sham procedure (n = 18) in Cycle 1. In Cycle 2, the BoNT-A group received a secondepisode of BoNT-A (n = 20) and sham group received their first episode of BoNT-A (n = 17). Apediatric rehabilitation specialist masked to group allocation graded each adverse event (AE)according to system, severity (mild, moderate, serious, sentinel) and causality (unlikely/unrelated; possible; probable/definite).

RESULTS: There was no difference for all moderate/serious AEs between the BoNT-A and sham/control groups in either Cycle 1 (incident rate ratio = 1.30, 95% confidence interval =0.43–4.00; P = .64) or Cycle 2 (incident rate ratio = 0.72, 95% confidence interval = 0.30–1.75;P = .47). In Cycle 2, 1 serious, 3 moderate (single-episode group), and 24 mild (single-episodegroup n = 10; 2 episode group n = 14) AEs were probably/definitely related to BoNT-A.

CONCLUSIONS: Children receiving BoNT-A were at no greater risk of moderate/serious AEscompared with a sham control procedure. There was no increased risk of moderate/seriousAEs between one and two episodes of BoNT-A.

WHAT’S KNOWN ON THIS SUBJECT: Children withmarked cerebral palsy (CP) are considered atgreater risk of adverse events (AEs) afterintramuscular injections of BoNT-A. To date therehas been no randomized controlled trialexamining safety of intramuscular BoNT-Ainjections in children with marked CP.

WHAT THIS STUDY ADDS: Children withnonambulatory CP had no greater risk ofmoderate or serious AEs after intramuscularinjections of BoNT-A compared with a sham/control group. There was no greater risk of AEsfor children receiving 2 compared with 1 episodeof BoNT-A.

aQueensland Paediatric Rehabilitation Service, Lady Cilento Children’s Hospital, Brisbane Australia; bQueenslandCerebral Palsy and Rehabilitation Research Centre, School of Medicine, The University of Queensland, cCentre forChildren’s Health Research, Children’s Health Queensland, and dSchool of Population Health, The University ofQueensland, Brisbane, Australia

Dr Edwards made substantive contributions to the conception and design of the study, acquisitionand interpretation of data, and revision of the article and wrote the initial draft of the article; DrSakzewski contributed to the conception, design, data analysis and interpretation, and revision ofthe article; Drs Copeland and McLennan made substantive contributions to the conception anddesign of the study and acquisition and interpretation of data; Ms Gascoigne-Pees and Ms Thorleycontributed to the conception, design, data analysis and interpretation, and revision of the articleand approval of the final manuscript; Ms Kentish contributed to conception and design, acquisitionof funding, acquisition of data, and revision of the final manuscript; Dr Ware contributed to dataanalysis and interpretation, biostatistical advice, and revision of the article; Ms Boyd contributed tothe conception, design, data interpretation, and revision of the article; and all authors approved ofthe final manuscript as submitted.

This trial has been registered with the ACTRN Register (http://isrctn.org) (identifierACTRN12609000360213).

www.pediatrics.org/cgi/doi/10.1542/peds.2015-0749

DOI: 10.1542/peds.2015-0749

Accepted for publication Aug 19, 2015

PEDIATRICS Volume 136, number 5, November 2015 ARTICLE by guest on September 21, 2020www.aappublications.org/newsDownloaded from

Botulinum toxin A (BoNT-A) isregularly used for goal-directed focalmanagement of spasticity anddystonia for children with cerebralpalsy (CP).1,2 Intramuscular BoNT-Ainjections decrease spasticity locallyby interfering with cholinergictransmission at the neuromuscularjunction.3 In a double-blindrandomized controlled trial (RCT),BoNT-A has been shown to reducepain4 and improve care and comfortgoals for nonambulatory childrenwith CP.5 In 2008, the safety of BoNT-A for children with CP was called intoquestion by the US ConsumerAdvocacy Group Public Citizen, whichpetitioned the US Federal DrugAdministration to increase labelwarnings for commercially availableBoNT-A products.6 At this time, 9deaths had been reported in childrenwith CP under 16 years of age afterintramuscular injections of BoNT-A.7

Concerns had been raised after BoNT-A injections in children withneuromuscular conditions andsubsequent adverse events (AEs). Thetemporal association resulted ina black box warning appearing onlabels of BoNT-A (Dysport andMyobloc [Botox Allergan PLC, Dublin,Ireland]) of the risk of spread ofBoNT-A beyond the injection site,with associated risks of dysphagia,aspiration, pneumonia, and death.7

Gaining a clear understanding of theincidence of AEs in children with CPafter BoNT-A is complex. Because ofthe heterogenous nature of CP,consideration must be given tocomorbidities and level ofimpairment. A population-basedstudy demonstrated a 10-foldincrease in comorbidities for childrenclassified as Gross Motor FunctionClassification System (GMFCS) LevelV compared with Level I.8 Seventy-three percent of total reportedcomorbidities occurred in the GMFCSV group despite representing only34% of the population.8 Aretrospective audit of AEs after BoNT-A showed an increase in the numberof AEs according to GMFCS level. The

authors articulated concerns aboutadministration of BoNT-A to childrenwith preexisting respiratory andswallowing difficulties.9 Twosystematic safety reviews highlightedvariability in reporting of AEssecondary to a lack of consistentdefinitions, reporting methodology,indications, preparations, dosing,dilution, sedation, and injectiontechniques.10,11 Our recentrandomized, double-blinded sham/controlled trial supported efficacy ofintramuscular injections of BoNT-Afor nonambulatory children with CP(GMFCS IV and V) to improve comfortand reduce carer burden.5 Childrenwho received BoNT-A compared withthe sham procedure had significantlyand clinically greater gains inperformance of and satisfaction withcare and comfort goals (CanadianOccupational Performance Measure;mean difference of 2.2, 95%confidence interval [CI] 0.8–3.5; P =.02). Children receiving BoNT-A hadsignificantly less pain compared withbaseline and improved reportedquality of life.5 This study nowinvestigates whether intramuscularinjections of BoNT-A increased therisk of AEs for children with CPclassified as GMFCS IV or V comparedwith a sham/control group andwhether there was an increase in AEswith repeated BoNT-A injections.

METHODS

Design and Procedure

This study was conducted ata pediatric tertiary referral center forchildren with CP in Brisbane,Australia, between March 2009 andAugust 2011. This study wasundertaken in 2 cycles. Cycle 1 wasa double-blind, randomized sham/control trial. Children wererandomized using concealedallocation to either receiveintramuscular BoNT-A injections orsham procedure.12 In Cycle 2, allchildren received BoNT-A treatment.Participants, parents and assessorswere blinded to group allocation

which was not revealed until aftercompletion of Cycle 2. Full ethicalapproval and trial registration wereobtained and study methodology isreported in detail elsewhere.12

Participants

Recruitment was undertaken througha statewide tertiary referral service.Eligibility was confirmed bya pediatric rehabilitation specialist.Children were eligible for inclusion ifthey had (1) a confirmed diagnosis ofCP and were aged 2 to 16 years, (2)were GMFCS IV or V, (3) hadspasticity affecting comfort or carerburden. Children were ineligible ifthey had (1) body weight ,10 kg, (2)received intramuscular BoNT-Ainjections or orthopedic surgerywithin 6 months of study entry, or (3)change to spasticity medicationwithin 2 months of study entry.

Interventions

Assessment of hypertonicity andparent/caregiver care and comfortgoals were performed to determinemuscle groups for BoNT-A injectionbefore randomization in Cycle 1. Onentry to the trial, a comorbidityassessment was completed by 1 of 3rehabilitation specialists (P.E., K.M.,T.C.; Supplemental Appendix 1). Onthe date of injections, an assessmentof health status was completed by studynurses (Supplemental Appendix 1).

Injection Procedure

Topical analgesia (EMLA cream) wasapplied 1 hour before and intranasalfentanyl (1.5 mcg/kg of 300 mcg/mLsolution) was administered 10minutes before the procedure. Doserange of BoNT-A was 0.5 U to 4 UBotox (Allergan PLC) per muscle witha maximum mean dose of 12 U Botoxper kg/body weight (maximum totaldose 400 U). Dilution of Botox was100 U per mL for all injections.Blinding was maintained by the use ofa screen blocking view of theprocedure from the monitoring nurse,parent, and child. Families were notrequired to pay for any medication

896 EDWARDS et al by guest on September 21, 2020www.aappublications.org/newsDownloaded from

associated with the trial. All trialmedications were funded through theQueensland Pediatric RehabilitationService, Royal Children’s Hospital,Brisbane, Australia.

Sham Procedure

Topical analgesia (EMLA cream) wasapplied 1 hour before and intranasalsaline was administered 10 minutesbefore the sham procedure. A screenwas used to obscure vision from thesham injecting procedure.12 Thesham group did not receive anyBoNT-A injections in Cycle 1. All siteswere covered with adhesive dressingsand tincture of iodine to mask thesham injection site.

Reporting of AEs

An AE is defined as any unfavorableand unintended sign, symptom, ordisease temporally associated withthe use of a medication, whether ornot considered related to themedication.13 In this study, AEs wererecorded during the injectionprocedure and at 2, 4, and 16 weeksafter each procedure ona standardized questionnaire bya nurse (V.W., J.L.) masked to groupallocation (Supplemental Appendix2). A pediatric rehabilitationspecialist (P.E.) experienced in use ofBoNT-A and masked to groupallocation independently graded eachAE according to system, severity, andcausality. For causality, events weregraded as unrelated or unlikely/possibly, or as probably/definitelyrelated to the intervention procedure(defined in Fig 1).14 These wereattributed on the basis of patienthistory, onset or duration ofsymptoms, contact history of similarsymptoms, and consistency withprevious literature of common AEsthought to be related tointramuscular BoNT-A injections. Forseverity, events were graded as mild,moderate, serious, or sentinel(defined in Fig 1). An external safetyofficer was notified of any serious orsentinel events within 72 hours. Inthe case of a sentinel event, the study

would be suspended until reviewedby the independent safety officer withexpedited reporting to the ethicscommittee.

Sample Size

The sample size of 40 participantswas initially calculated for the parenttrial considering efficacy ofintramuscular injections of BoNT-Afor nonambulatory children with CPto improve comfort and reduce carerburden.12 When considering the sizeof the between-group differencedetectable for the primary outcomefor this study (moderate/seriousAEs), it was assumed that 10% ofBoNT-A participants would havea serious AE. With a = .05, there was80% power to detect a fivefolddifference in moderate/serious AErates between groups (ie, to detect aAE rate in the control group of #2%).

Statistical Analysis

Outcome variables were summarizedas means and SD if continuous and asfrequency (percentage) if categorical.The association between treatmentgroup and number of events in Cycles1 and 2 was investigated using Poissonregression with robust standard errors.Group (BoNT-A/control) was enteredas the main effect. The associationbetween health status risk factors(epilepsy, gastrostomy, hydrocephalus)and the incidence of total and relatedAEs were was investigated usingPoisson regression with robuststandard errors. Model assumptions

were tested with the deviancegoodness-of-fit statistic. Results arepresented as incident rate ratio (IRR)and 95% CI. Analysis was undertakenusing Stata statistical software V12.0(StataCorp, College Station, TX).

RESULTS

Study recruitment and follow-up forboth cycles of the study are reportedin Fig 2. Seventy-seven children wereeligible for inclusion. Forty-one wererandomized in Cycle 1 to the sham/control (n = 18) or BoNT-A group(n = 23). Thirty-seven childrencompleted Cycle 2 including 20children (87%) who received BoNT-Ain Cycle 1 with a repeat episode ofBoNT-A in Cycle 2. In the BoNT-Agroup, 1 child in Cycle 1 did notproceed to Cycle 2 because ofconcerns with AEs (patientidentification no. [ID5] floppy, novocalization, increased drooling:Supplemental Table 5), and 2 exitedthe study because of progression toorthopedic surgery (ID30, 38,Supplemental Table 5). In the shamgroup, 1 child exited the study beforeCycle 2 because of progression toorthopedic surgery (ID3,Supplemental Table 6). Seventeenchildren (94%) received their first in-trial episode of BoNT-A in Cycle 2.Efficacy of clinical outcomes in Cycle1 have been reported elsewhere.5

Safety Results in Cycle 1

All children in both groups completedfollow-up for Cycle 1. Patient

FIGURE 1Definitions of grading severity and causality of adverse events.

PEDIATRICS Volume 136, number 5, November 2015 897 by guest on September 21, 2020www.aappublications.org/newsDownloaded from

demographics, baseline characteristics,and comorbidities for each group arereported in Table 1. Groups weredemographically similar at baseline;however, the sham group had a greaterproportion of children with epilepsyand using oral medications forspasticity (Table 1). In Cycle 1, themean total dose of BoNT-A was 10.5 UBotox per kilogram body weight (SD62.5 U/kg) and the dilution wasstandardized at 100 U Botox per 1 mLnormal saline (Table 1).

Adverse events are summarized inTable 2 and reported individually inSupplemental Tables 5 and 6.Adverse events by severity aredisplayed in Fig 3.

The overall incidence of moderate/serious AEs was similar betweengroups (IRR = 1.30, 95% CI:0.43–4.00; P = .64). There was littledifference between groups inincidence of moderate/serious eventsthat were related to the BoNT-A/

fentanyl procedure (IRR = 0.78, 95%CI: 0.12–5.15; P = 0.80). There werea small number of serious AEs, 2 in 1child (,1%) in the sham group and 3in 3 children (13%) in the BoNT-Agroup. No serious AE wasindependently classified as beingeither definitely or probably relatedto the BoNT-A/fentanyl intervention.For moderate events, 4 AEs werereported in 3 children (17%) in thesham group and 7 in 5 children(22%) in the BoNT-A group.Moderate/serious events weresimilar between groups. Eventsclassified as having a definite/probable relationship to theintervention included 1 in the shamgroup (ID21: focal seizures on nightof procedure) and 2 in the BoNT-Agroup (ID30: weakness, gagging,vomiting, constipation, temperature,seizure; ID41: increased saliva at 1week and increased seizures at 2weeks, see Supplemental Tables 5and 6).

Overall, in Cycle 1, there were a totalof 33 mild, moderate, or serious AEsreported for 23 children (56% ofchildren) across both groups. Ofthese, 8 AEs occurred in 5 children(27%) in the sham/control group and25 AEs in 18 children (78%) in theBoNT-A group. Although childrenreceiving BoNT-A did not havea statistically significantly greaterincidence of mild, moderate, orserious AEs, they were 2.45 times aslikely to have any AE as a child in thesham group (IRR = 2.45, 95% CI:0.96–6.20; P = .06) (Table 3). Most ofthis difference is attributable toa large number of bruises related tothe injection procedure in the BoNT-Agroup that were classified as mildevents. There were 2 mild AEsreported in 2 children (11%) in thesham group and 15 mild AEs in 13children in the BoNT-A group (57%).Events classified as definitely/probably related to the interventionincluded 1 (ID29) in the sham groupand 10 (IDs 8, 15, 17, 18, 26, 28, 30,33, 37, 41: bruising, see SupplementalTables 5 and 6) in the BoNT-A group.

FIGURE 2AE profile according to Consolidated Standards of Reporting Trials Flow Diagram. ITB, intrathecalbaclofen; LL, lower limb; MACS, Manual Ability Classification System; MAS, Modified Ashworth Scale;MTS = Modified Tardieu Scale; UL, upper limb.

898 EDWARDS et al by guest on September 21, 2020www.aappublications.org/newsDownloaded from

Safety Results in Cycle 2

In Cycle 2, the mean total dose ofBoNT-A was 10.8 Botox perkilogram body weight (SD 62.4 U/kg) for the 2 BoNT-A episodegroup and 10.5 U/kg (SD 6 1.9U/kg)for the single-episode BoNT-A groupwith the dilution standardized at100 U Botox per 1 mL normal saline(Table 1). One child receiveda maximum dose of 14 U/kg,exceeding trial protocol.

The incidence of moderate/seriousAEs was similar between groups (IRR= 0.72, 95% CI: = 0.30–1.75; P = .47).There were 2 serious AEs in 2children (12%) in the single-episodeBoNT-A group and 4 serious AEs in 2children (10%) receiving repeatinjections. One serious AE (ID10) wasthought to be definitely/probablyrelated to the BoNT-A/ fentanylintervention from the single-episodegroup. There were 11 moderate AEs

in 8 children (47%) in the singleBoNT-A group and 7 moderate AEs in6 children (30%) in the repeat BoNT-A group. Of these events, 3 wereconsidered to be definitely/probablyrelated, all of which occurred in thesingle-injection group (ID7hallucinating, teary, upset withintiming of fentanyl; ID25 increaseddrooling 1–4 weeks; ID29 decreasedhead control, increased saliva, urinaryincontinence at 2–10 days; seeSupplemental Tables 5 and 6).

There were a total of 49 AEs in 27children (73%) across both groups.Of these, 23 AEs occurred in 12children (71%) in the single-episodeBoNT-A group (n = 17) and 26 AEs in15 children (75%) in the repeat-BoNT-A group (n = 20; IRR = 0.96,95% CI: 0.52–1.79; P = .90).

There were 10 mild AEs in 7 children(41%) in the single BoNT-A injectiongroup and 15 mild AEs in 13 children(65%) in the repeat-injection group.Eighteen mild AEs from both groups(75%) were definitely related to theintervention (bruising).

There was 1 child from the repeatBoNT-A episode group who hadrepeated serious AEs, 1 in Cycle 1 and3 in Cycle 2 (ID31). These AEs wereclassified as unlikely/unrelated to theintervention. One child from thesingle-injection group had repeatedmoderate AEs, 2 in Cycle 1 (sham/control group) and 1 in Cycle 2(ID32). These events were repeatedlower respiratory tract infectionsclassified as unlikely related to theintervention.

Association Between Health StatusRisk Factors and Incidence of AEs

Children with simple or complexepilepsy did not have significantlygreater risk of moderate/serious AEscompared with no epilepsy aftereither 1 or 2 injections of BoNT-A(Table 4). Similarly, the presence ofhydrocephalus posed no significantlygreater risk of moderate/serious AEsafter either 1 or 2 injections of BoNT-A. Children with a gastrostomy were

TABLE 1 Baseline Characteristics and Comorbidities of Study Participants

Characteristics Cycle 2 BoNT-A Group(n = 23)

Sham/Control + Cycle 1 BoNT-AGroup (n = 18)

DemographicAge y, mean (SD) 7 y, 1 mo (3 y, 7 mo) 7 y, 5 mo (3 y, 9 mo)Gender, male, n (%) 16 (70) 11 (61)

Weight, kg, mean (SD) 20.6 (8.5) 21 (7.6)ClassificationGMFCS IV, n (%) 3 (13) 0 (0)GMFCS V, n (%) 20 (87) 18 (100)MACS III, n (%) 0 (0) 1 (5)MACS IV, n (%) 6 (26) 2 (11)MACS V, n (%) 13 (57) 10 (56)

Too young for MACS classification, n (%) 4 (17) 5 (28)

Epilepsy, n (%) 15 (65) 15 (83)Complex epilepsy ($2 daily

anticonvulsants), n (%)11 (48) 10 (44)

Gastrostomy or nasogastric feeding, n (%) 15 (65) 12 (67)Bulbar/pseudobulbar palsy/swallowing

problems, n (%)18 (78) 12 (67)

Drooling, n (%) 19 (83) 15 (83)History of gastroesophageal reflux, n (%) 9 (39) 8 (44).5 diagnosed RTIs in past 12 mo, n (%) 7 (30) 7 (39)Carer-reported history chronic/frequent

pain, n (%)16 (70) 11 (61)

Use of daily oral medications to managespasticity, n (%)

13 (57) 16 (89)

Congenital heart disease, n (%) 1 (4) 0 (0)Inflammatory bowel disease, n (%) 1 (4) 0 (0)Diabetes mellitus, n (%) 1 (4) 0 (0)Asthma, n (%) 2 (9) 3 (17)Shunt-dependent hydrocephalus, n (%) 5 (22) 3 (17)

A priori proposed BoNT-A dose providedCycle 1 total dose u/kg, mean (SD) 10.5 (2.5) 9.9 (2.6)Cycle 1 LL u/kg, mean (SD) 10.3 (2.2)a 8 (3.5)a

Cycle 1 UL u/kg, mean (SD) 8.2 (3.7)a 7 (3)b

Cycle 2 Total dose u/kg, mean (SD) 10.8 (2.4) 10.5 (1.9)Cycle 2 LL u/kg, mean (SD) 9.6 (2.5)c 8.9 (3.2)d

Cycle 2 UL u/kg, mean (SD) 7.8 (4.2)e 7.2 (4.0)d

BoNT-A dose for sham/control group refers to BoNT-A received in Cycle 2 and no BoNT-A was administered to the controlgroup in Cycle 1. LL, lower limb; MACS, Manual Ability Classification System; RTIs, respiratory tract infections; UL, upperlimb.a n = 13.b n = 10.c n = 11.d n = 12.e n = 15.

PEDIATRICS Volume 136, number 5, November 2015 899 by guest on September 21, 2020www.aappublications.org/newsDownloaded from

1.9 and 5.4 times more likely to havea moderate/serious AE after 1 and 2injections of BoNT-A, respectively.

Longer-term Follow-up

No child died while participating inthe study. All children whoparticipated in the trial havesubsequently been followed upposttrial through our clinical service.Since study completion in August2011 to November 2014, 7 of the 41children have died. Two childrenwere in the single-episode group and4 were in the 2-episode group. Thesedeaths were all reviewed for anyrelationship to BoNT-A treatment andcomorbidities. One death hada temporal association with BoNT-A.This child received sham treatment inCycle 1. After Cycle 1, BoNT-Ainjections were administered at 6months post–follow-up withorthopedic surgery. SubsequentBoNT-A injections occurred at 20months after surgery and thenanother episode 7 months thereafter.This child (ID3) died 15 days after thethird posttrial episode of BoNT-A. Thedeath was associated with prolongedstatus epilepticus. The child was

known to have epilepsy. The medicalteam treating this child postulatedsubtherapeutic dosing ofanticonvulsants as causal. This deathwas reported to and reviewed by the

study external safety committee. Forthe 6 other children who diedpoststudy, causes of death includedrespiratory complications (n = 3);gastrointestinal tract hemorrhage

TABLE 2 AEs for Group Allocation, AE by System and Patient ID Number

Cycle 1 AE Type Sham/Control (n = 18) BoNT-A (n = 23)

Total: Events/Children, n

AE System and ID No. Total: Events/ Children, n AE System and ID No.

#4 wk Serious 0 2/2 S = ID5, N = ID31Moderate 3/3 R = ID32, 35; N = ID21 5/4 R = ID34; S = ID14, 30, 41; N = ID41Mild 2/2 S = ID36; G = ID29 15/13 R = ID9, 31, 38, N = ID1532, B = ID8, 15, 17, 18,

26, 28, 30, 33, 37, 41#4 mo Serious 2/1 R = ID36; G = ID36 1 R = ID27

Moderate 1 R = ID32 2/2 R = ID14; N = ID39Mild 0 0

Cycle 2 AE Type Sham/Control BoNT-A 31 (n = 17) BoNT-A 32 (BoNT-A Group) (n = 20)

Total: Events/Children, n

AE System and ID No. Total: Events/Children, n

AE System and ID No.

#4 wk Serious 2/2 N = ID10, 12 1 N = ID31Moderate 8/8 S = ID7, 25, 29; R = ID12, 23, 32; N = ID42;

G = ID214/4 S = ID41; R = ID9, 13; N = ID18

Mild 10/7 S = ID42; R = ID25; G = ID16; B = ID2, 7, 12, 16,20, 25, 42

14/12 R = ID17, 28, 31; B = ID8, 15, 17, 18, 24, 26, 31, 37,39, 40, 41

#4mo

Serious 0 3/2 S = ID9, G = ID31, E = ID31

Moderate 3/2 R = ID7, 25; N = ID7 3/3 R = ID17, 26; N = ID13Mild 0 1 R = ID33

B, bruising; E, renal; G, gastroenterological; N, neurologic; R, respiratory; S, systemic.

FIGURE 3AEs by relationship to BoNT-A and severity.

900 EDWARDS et al by guest on September 21, 2020www.aappublications.org/newsDownloaded from

(n = 1); complex cyanotic congenitalheart disease (n = 1), and renal failure(n = 1). There were no formalautopsies, and these were based onclinical judgment by their pediatricrehabilitation specialist andneurologist.

DISCUSSION

In the first cycle of this double-blindRCT of intramuscular injections ofBoNT-A compared with a sham/control group with independentevaluation of AE severity andcausality, there were no statisticallysignificant differences betweengroups in the rate of moderate/serious AEs for nonambulatorychildren with CP. Children in theBoNT-A group were almost 2.5 timesmore likely to have any AE; however,this difference is due to a greaternumber of mild AEs in the BoNT-Agroup. Furthermore, in Cycle 2 whencomparing 1 to 2 episodes of BoNT-Atreatment in trial conditions with thesame unbiased, blinded evaluation ofAE severity and causality, there wasa greater incidence of moderate/serious AEs in the single-injectiongroup considered to be related to theintervention. Results indicate thatthere was no added risk of havinga second episode of BoNT-A after a 6-month period (the mean number ofrelated moderate/serious AEs perchild was 0.13 after 1 injection and0.10 after 2 injections).

To our knowledge, this is the firstdouble-blind sham/controlled trial ofBoNT-A injections with independentevaluation of AE severity andattribution of causality innonambulatory children with CP.Verbal probability expressions havebeen reported extensively in theliterature and are often used indiscussions with patients becausethey are easily understood in terms ofrisk.15 Independent evaluation ofpretrial comorbidities for bothtreatment and control groups has notpreviously been undertakenprospectively in a double-blind trialof BoNT-A in CP. Prospectiveevaluation of AEs has been reportedusing clinical audits with similarreporting of significant comorbiditiesin children with CP classified GMFCSIV and V.16 In a retrospective safetyreview study, O’Flaherty reporteda higher incidence of health issues ina 1-month period preceding BoNT-Ainjections compared with 1 monthafter BoNT-A in children classifiedGMFCS IV and V (95/316 (30%) and56 of 255 children (22%).16 Althoughour sham/control group had a slightlygreater proportion of children withepilepsy and regular use of spasticitymedication at baseline, this differencewas not statistically different.

Consistent with other RCTs of BoNT-A,17 further follow-up of the sampleposttrial identified some childrenwho died. Significant morbidity andmortality are associated with

nonambulatory CP and comorbidepilepsy18; however, overall lifeexpectancy, particularly for childrenwith severe disabilities, has improvedover a 20-year period.19 Ina prospective cohort of 6277 childrenwith CP, those with severe CPrepresented 18% of the person-yearsbut 69% of deaths.19 Another cross-sectional survey of the health statusof children with CP classified GMFCSV (n = 122) who had a feeding tubefound that the group that used themost health care resources had thehighest rate of respiratory problemsand poorest Global Health scores.20 Aretrospective review of CP data fromCalifornia (513 children) reportedreduced mobility and feeding abilityas factors associated with increasedrisk for death.19 When reviewing thechildren who participated in thisstudy, clinicians believe that selectionof eligible patients with pain andcomfort goals identified the most frailand medically complex children. Thepresent study was a carefullyconducted RCT that studied a high-risk population and identified noadditional risk with BoNT-A andfentanyl compared with a controlgroup and no additional risk froma second episode of BoNT-Ainjections. A potential limitation forinterpretation of the study is that AEswere related to the combination ofBoNT-A and fentanyl, and it was notpossible to differentiate the relativecontribution of each on AEs. Fentanylis considered to be a safe andeffective analgesic21; however, itcannot be discounted that fentanylmay have contributed to AEs in thestudy.

One potential limitation of our studyis that it was powered to evaluateefficacy of BoNT-A to improve careand comfort goals in nonambulatorychildren with CP and may have beenunderpowered to determinedifferences in incidence of moderate/serious AEs between groups.However, the rate difference betweengroups was not large, with a pointestimate of between-group difference

TABLE 3 Differences Between Groups for the Total Number of AEs and the Number of Severe/Moderate AEs Definitely/Probably Related to the BoNT-A and Fentanyl Intervention inCycle 1 (RCT) and Cycle 2 (1 vs 2 BoNT-A Injections)

Control Treatment IRR (95% CI); P

Mean (SD) Mean (SD)

Cycle 1 n = 18 n = 23All events 0.44 (0.86) 1.09 (0.85) 2.45 (0.96–6.20); .06Moderate/serious events 0.33 (0.69) 0.43 (0.66) 1.30 (0.43–4.00); .64All related events 0.17 (0.38) 0.56 (0.73) 3.39 (1.05–10.91); .04Related moderate/serious events 0.11 (0.32) 0.09 (0.29) 0.78 (0.12–5.15); .80

Cycle 2 n = 17 n = 20All events 1.35 (1.37) 1.30 (1.22) 0.96 (0.52–1.79); .90Moderate/serious events 0.76 (0.90) 0.55 (0.89) 0.72 (0.30–1.75); .47All related events 0.71 (0.77) 0.55 (0.51) 0.78 (0.41–1.49); .45Related moderate/serious events 0.24 (0.44) 0 (0) NC

n/c, not calculable because there were no related moderate or severe events in treatment group.

PEDIATRICS Volume 136, number 5, November 2015 901 by guest on September 21, 2020www.aappublications.org/newsDownloaded from

for moderate/serious events in Cycle1 only 30% higher in the BoNT-Agroup than control group. When onlyrelated events were considered,participants in the control group hada similar incidence of moderate/serious AEs compared with BoNT-Aparticipants. These results suggestthat there may be only a relativelysmall difference in moderate andserious related events in both studycycles, although a larger trial toconfirm this would be beneficial. Ourinitial sample size calculations mayhave underestimated the actualnumber of moderate/serious AEs.This is because previous studies havenot reported AEs in the sameframework as this study (ie, not

reporting on a control group; notgrading by likely relationship tointervention). We observed a mean of0.58 events per child after their firstBoNT-A injection. Post hoc powercalculations indicated that there was80% power to detect a statisticallysignificant between-group differenceif the rate of moderate/serious AEs inthe control group was #0.17. Furtherexploration of the relationshipbetween comorbidities and increasedrisk of AEs suggested that thepresence of a gastrostomy mayincrease the risk of AEs after BoNT-A.These findings are in line withprevious reports of increased risk ofrespiratory AEs for children withpseudo-bulbar palsy17,22; however

our results should be viewed withcaution because of the small samplesize for this secondary analysis. Inaddition, a longer-term prospectivetrial with repeated episodes of BoNT-A in this patient group is required.

CP remains an incurable conditionand as such, treatments offered arefor symptom control. For childrenwith nonambulatory CP, consideredtreatment options are usually basedon reducing pain, improving healthstatus and quality of life, andreducing carer burden. A range ofgeneralized medical treatments (eg,oral medications, intrathecalbaclofen), focal targeted treatments(eg, orthopedic surgery), and therapyinterventions are regularly offered for

TABLE 4 Relationship Among Epilepsy, Feeding Status, Hydrocephalus for the Total Number of AEs and the Number of Severe/Moderate AEs Definitely/Probably Related to the Sham Injection (Sham Group in Cycle 1), BoNT-A and Fentanyl Intervention for 1 Injection (BoNT-A Group in Cycle 1 andSham Group in Cycle 2), and 2 injections (BoNT-A group in Cycle 2)

No. of Injections All Events Moderate/Serious Events Related Moderate/Serious Events

Mean (SD) IRR (95% CI); P Mean (SD) IRR (95% CI); P Mean (SD) IRR (95% CI); P

Sham only (n = 18)EpilepsyComplex (n = 10) 0.20 (0.42) 1.00 0.20 (0.42) 1.00 — —

Simple (n = 5) 0.40 (0.89) 2.00 (0.22–18.22); 0.54 0.40 (0.89) 2.00 (0.22–18.22); 0.54 — —

Absent (n = 3) 1.33 (1.53) 6.67 (1.25–35.67); 0.03 0.67 (1.15) 3.33 (0.42–26.76); 0.26 — —

HydrocephalusAbsent (n = 15) 0.40 (0.91) 1.00 0.33 (0.72) 1.00 — —

Present (n = 3) 0.67 (0.58) 1.67 (0.41–6.83); 0.48 0.33 (0.58) 1.00 (0.14–7.21); 1.00 — —

GastronomyAbsent (n = 6) 0.67 (1.21) 1.00 0.33 (0.81) 1.00 — —

Present (n = 12) 0.33 (0.65) 0.50 (0.09–2.87); 0.44 0.33 (0.65) 1.00 (0.12–8.49); 1.00 — —

1 injection (n = 40)EpilepsyComplex (n = 21) 1.38 (1.32) 1.00 0.67 (0.91) 1.00 0.33 (0.58) 1.00Simple (n = 8) 0.88 (0.64) 0.63 (0.34–1.19); 0.16 0.38 (0.52) 0.56 (0.19–1.65); 0.29 0.13 (0.35) 0.38 (0.05–2.76); 0.34Absent (n = 11) 1.09 (0.83) 0.79 (0.44–1.43); 0.44 0.55 (0.69) 0.82 (0.33–2.06); 0.67 0.27 (0.47) 0.82 (0.24–2.77); 0.75

HydrocephalusAbsent (n = 31) 1.26 (1.12) 1.00 0.61 (0.84) 1.00 0.29 (0.53) 1.00Present (n = 9) 1.00 (1.00) 0.79 (0.39–1.60); 0.52 0.44 (0.53) 0.73 (0.30–1.75); 0.48 0.22 (0.44) 0.77 (0.19–3.08); 0.71

GastronomyAbsent (n = 14) 0.71 (0.61) 1.00 0.36 (0.50) 1.00 0.21 (0.43) 1.00Present (n = 26) 1.46 (1.2) 2.05 (1.19–3.51); 0.009 0.69 (0.88) 1.94 (0.82–4.59); 0.13 0.31 (0.55) 1.44 (0.42–4.88); 0.56

2 injections (n = 20)EpilepsyComplex (n = 11) 1.18 (0.75) 1.00 0.55 (0.82) 1.00 0.09 (0.30) 1.00Simple (n = 3) 2.00 (2.65) 1.69 (0.46–6.25); 0.43 1.00 (1.73) 1.83 (0.29–11.75); 0.52 0 (0) NCAbsent (n = 6) 1.17 (1.17) 0.99 (0.43–2.28); 0.98 0.33 (0.52) 0.61 (0.14–2.61); 0.51 0.17 (0.41) 1.83 (0.13–26.06); 0.66

HydrocephalusAbsent (n = 16) 1.44 (1.31) 1.00 0.69 (0.95) 1.00 0.13 (0.34) 1.00Present (n = 4) 0.75 (0.50) 0.52 (0.25–1.08); 0.08 0 (0) NC 0 (0) NC

GastrostomyAbsent (n = 7) 1.14 (0.90) 1.00 0.14 (0.38) 1.00 0 (0) 1.00Present (n = 13) 1.38 (1.39) 1.21 (0.56–2.62); 0.63 0.77 (1.01) 5.38 (0.74–39.43); 0.10 0.15 (0.38) NC

Effect estimates presented as incidence rate ratio (IRR) and 95% confidence interval. NC, not calculable; —, no reported AEs and IRRs not calculated.

902 EDWARDS et al by guest on September 21, 2020www.aappublications.org/newsDownloaded from

symptom control targeting care andcomfort goals. Many of thesetreatments pose significant risk ofAEs but remain part of the clinicians’toolbox. As part of clinicians’obtaining a valid consent totreatment, families need to considerbalancing the risk and benefit. Theethics of withdrawing realtherapeutic options to a group withmultiple complexities also needscareful consideration. There is anincreased risk in the incidence of mildAEs; however, this may be consideredacceptable risk given the meaningfulchanges in care and comfort goals.Our recommendation is thatintramuscular injections of BoNT-Afor nonambulatory children with CPcan lead to meaningful changes incare and comfort goals but needs toconsider relative risk of moderate/serious AEs that may be associatedwith other comorbidities. Childrenwho have gastrostomies may be ata higher risk of serious AEs, butlonger-term prospective evaluation ofsafety with repeated episodes isneeded.

CONCLUSIONS

In a double-blind sham/controlledtrial and subsequent follow-up of 1versus 2 episodes of BoNT-A toaddress care and comfort goals, therewas no evidence of an increased riskof moderate/serious AEs innonambulatory children with CPreceiving intramuscular injections ofBoNT-A administered with fentanyl.There is an increased risk in theincidence of mild AEs; however, thismay be considered acceptable riskgiven the meaningful changes in careand comfort goals.

Children who have gastrostomies maybe at a higher risk of serious AEs, butlonger-term prospective evaluation ofsafety with repeated episodes isneeded.

Our recommendation is thatintramuscular injections of BoNT-Afor nonambulatory childrenwith CP can lead to meaningfulchanges in care and comfort goalsbut needs to consider the relativerisk of moderate/serious AEs that

may be associated with othercomorbidities.

ACKNOWLEDGMENTS

Special thanks to Dr Paul Kubler forhis role as external safety officer,Samantha Donaghey (occupationaltherapist) for project coordinationand conducting research assessments,and to Jayne Lindsley (Clinical NurseCoordinator) and Vicky Witherford(Clinical Nurse Coordinator) for theirnursing care in treatment and shamprocedures and adverse eventmonitoring.

ABBREVIATIONS

AE: adverse eventBoNT-A: botulinum toxin ACI: confidence intervalCP: cerebral palsyGMFCS: Gross Motor Function

Classification SystemID: identification numberIRR: incident rate ratioRCT: randomized controlled trial

Address correspondence to Dr Priya Edwards, Queensland Paediatric Rehabilitation Service, Lady Cilento Children’s Hospital, South Brisbane, Queensland, Australia

4101. E-mail: priya.edwards@health.qld.gov.au

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2015 by the American Academy of Pediatrics

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

FUNDING: This study was an investigator initiated and conducted trial by the Queensland Cerebral Palsy Health Service who received an unrestricted Educational

Grant (10268) through the Royal Children’s Hospital Foundation from Allergan PLC, Australia. RB was supported by a National Health and Medical Research Council

of Australia (NHMRC) Career Development Fellowship Level 2 (1037220) and LS by a NHMRC Early Career Research Fellowship (1090828).

POTENTIAL CONFLICT OF INTEREST: Dr Edwards, Dr Copeland, Dr McLennan, Ms Kentish, and Ms Thorley are involved in a public Paediatric Botulinum Toxin Program.

The other authors have indicated they have no potential conflicts of interest to disclose.

REFERENCES

1. Love SC, Novak I, Kentish M, et al;Cerebral Palsy Institute. Botulinum toxinassessment, intervention and after-carefor lower limb spasticity in children withcerebral palsy: international consensusstatement. Eur J Neurol. 2010;17(suppl2):9–37

2. Fehlings D, Novak I, Berweck S, Hoare B,Stott NS, Russo RN; Cerebral PalsyInstitute. Botulinum toxin assessment,intervention and follow-up for paediatric

upper limb hypertonicity: internationalconsensus statement. Eur J Neurol. 2010;17(suppl 2):38–56

3. Aoki KR. Pharmacology and immunologyof botulinum toxin serotypes. J Neurol.2001;248(suppl 1):3–10

4. Barwood S, Baillieu C, Boyd R, et al. Analgesiceffects of botulinum toxin A: a randomized,placebo-controlled clinical trial. Dev MedChild Neurol. 2000;42(2):116–121

5. Copeland L, Edwards P, Thorley M, et al.Botulinum toxin A for nonambulatorychildren with cerebral palsy: a doubleblind randomized controlled trial. JPediatr. 2014;165(1):140–146.e4

6. Administration FUSFaD. EarlyCommunication about an Ongoing SafetyReview of Botox and Botox Cosmetic(Botulinum toxin Type A) and Myobloc(Botulinum toxin Type B). Available at:http://www.fda.gov/Drugs/DrugSafety/

PEDIATRICS Volume 136, number 5, November 2015 903 by guest on September 21, 2020www.aappublications.org/newsDownloaded from

PostmarketDrugSafetyInformationforPatientsandProviders/DrugSafetyInformationforHeathcareProfessionals/ucm070366.htm.208. Accessed August 2,2008

7. Apkon SD, Cassidy D. Safetyconsiderations in the use of botulinumtoxins in children with cerebral palsy.PM R. 2010;2(4):282–284

8. Shevell MI, Dagenais L, Hall N; REPACQConsortium. Comorbidities in cerebralpalsy and their relationship toneurologic subtype and GMFCS level.Neurology. 2009;72(24):2090–2096

9. Naidu K, Smith K, Sheedy M, Adair B, YuX, Graham HK. Systemic adverse eventsfollowing botulinum toxin A therapy inchildren with cerebral palsy. Dev MedChild Neurol. 2010;52(2):139–144

10. Naumann M, Jankovic J. Safety ofbotulinum toxin type A: a systematicreview and meta-analysis. Curr Med ResOpin. 2004;20(7):981–990

11. Albavera-Hernández C, Rodríguez JM,Idrovo AJ. Safety of botulinum toxin typeA among children with spasticitysecondary to cerebral palsy: a systematicreview of randomized clinical trials. ClinRehabil. 2009;23(5):394–407

12. Thorley M, Donaghey S, Edwards P, et al.Evaluation of the effects of botulinumtoxin A injections when used to improveease of care and comfort in childrenwith cerebral palsy whom are non-ambulant: a double blind randomizedcontrolled trial. BMC Pediatr. 2012;12:120

13. Clinical safety data management:definitions and standards for expeditedreporting. Presented at the InternationalConference on Harmonisation ofTechnical Requirements for Registrationof Pharmaceuticals for Human Use.www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E2A/Step4/E2A_Guideline.pdf.; 1994.

14. Naranjo CA, Busto U, Sellers EM, et al. Amethod for estimating the probability ofadverse drug reactions. Clin PharmacolTher. 1981;30(2):239–245

15. Mazur DJ, Hickam DH. Patients’interpretations of probability terms. JGen Intern Med. 1991;6(3):237–240

16. O’Flaherty SJ, Janakan V, Morrow AM,Scheinberg AM, Waugh MC. Adverseevents and health status followingbotulinum toxin type A injections inchildren with cerebral palsy. Dev MedChild Neurol. 2011;53(2):125–130

17. Graham HK, Boyd R, Carlin JB, et al. Doesbotulinum toxin a combined with bracingprevent hip displacement in childrenwith cerebral palsy and “hips at risk”? Arandomized, controlled trial. J BoneJoint Surg Am. 2008;90(1):23–33

18. Harvey AS, Nolan T, Carlin JB. Community-based study of mortality in children withepilepsy. Epilepsia. 1993;34(4):597–603

19. Strauss D, Shavelle R, Reynolds R,Rosenbloom L, Day S. Survival incerebral palsy in the last 20 years: signsof improvement? Dev Med Child Neurol.2007;49(2):86–92

20. Liptak GS, O’Donnell M, Conaway M, et al.Health status of children with moderateto severe cerebral palsy. Dev Med ChildNeurol. 2001;43(6):364–370

21. Borland M, Jacobs I, King B, O’Brien D. Arandomized controlled trial comparingintranasal fentanyl to intravenousmorphine for managing acute pain inchildren in the emergency department.Ann Emerg Med. 2007;49(3):335–340

22. Graham HK, Aoki KR, Autti-Rämö I, et al.Recommendations for the use ofbotulinum toxin type A in themanagement of cerebral palsy. GaitPosture. 2000;11(1):67–79

904 EDWARDS et al by guest on September 21, 2020www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2015-0749 originally published online October 19, 2015; 2015;136;895Pediatrics 

McLennan, Megan Thorley, Megan Kentish, Robert Ware and Roslyn N. BoydPriya Edwards, Leanne Sakzewski, Lisa Copeland, Laura Gascoigne-Pees, Kim

PalsySafety of Botulinum Toxin Type A for Children With Nonambulatory Cerebral

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/136/5/895including high resolution figures, can be found at:

Referenceshttp://pediatrics.aappublications.org/content/136/5/895#BIBLThis article cites 20 articles, 1 of which you can access for free at:

Subspecialty Collections

subhttp://www.aappublications.org/cgi/collection/neurologic_disorders_Neurologic Disordershttp://www.aappublications.org/cgi/collection/neurology_subNeurologyhttp://www.aappublications.org/cgi/collection/disabilities_subChildren With Special Health Care Needsfollowing collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlin its entirety can be found online at: Information about reproducing this article in parts (figures, tables) or

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on September 21, 2020www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2015-0749 originally published online October 19, 2015; 2015;136;895Pediatrics 

McLennan, Megan Thorley, Megan Kentish, Robert Ware and Roslyn N. BoydPriya Edwards, Leanne Sakzewski, Lisa Copeland, Laura Gascoigne-Pees, Kim

PalsySafety of Botulinum Toxin Type A for Children With Nonambulatory Cerebral

http://pediatrics.aappublications.org/content/136/5/895located on the World Wide Web at:

The online version of this article, along with updated information and services, is

http://pediatrics.aappublications.org/content/suppl/2015/10/14/peds.2015-0749.DCSupplementalData Supplement at:

by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397. the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2015has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

by guest on September 21, 2020www.aappublications.org/newsDownloaded from