DOI: 10.1542/peds.2012-1317; originally published online March 18, 2013; 2013;131;661Pediatrics

Hubert Bland and Parviz HabibiCaroline Pao, Arvind R. Shah, Elizabeth Reus, Joseph Eliahoo, Fabiana Gordon,

Mina M. Chowdhury, Sheila A. McKenzie, Christopher C. Pearson, Siobhan Carr,Randomized Controlled Trial

Heliox Therapy in Bronchiolitis: Phase III Multicenter Double-Blind  

  http://pediatrics.aappublications.org/content/131/4/661.full.html

located on the World Wide Web at: The online version of this article, along with updated information and services, is

 

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2013 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

Heliox Therapy in Bronchiolitis: Phase III MulticenterDouble-Blind Randomized Controlled Trial

WHAT’S KNOWN ON THIS SUBJECT: Bronchiolitis, a leading causeof infant hospitalization, has few proven treatments. A few smallstudies have reported the beneficial effects of a mixture of 21%oxygen + 79% helium (Heliox). The 2010 Cochrane Review concludedthat additional large randomized controlled trials were needed todetermine the therapeutic role of Heliox in bronchiolitis.

WHAT THIS STUDY ADDS: The Bronchiolitis Randomized ControlledTrial Emergency-Assisted Therapy with Heliox—An Evaluation(BREATHE) trial is the largest multicenter randomized controlledtrial to date to investigate the efficacy of Heliox in acutebronchiolitis. The delivery method for Heliox therapy was found tobe crucial to its efficacy.

abstractBACKGROUND AND OBJECTIVE: Supportive care remains the mainstayof therapy in bronchiolitis. Earlier studies suggest that helium-oxygentherapy may be beneficial, but evidence is limited. We aimed to compareefficacy of 2 treatment gases, Heliox and Airox (21% oxygen 1 79%helium or nitrogen, respectively), on length of hospital treatment forbronchiolitis.

METHODS: This was a multicenter randomized blinded controlled trial of319 bronchiolitic infant subjects randomly assigned to either gas; 281 sub-jects completed the study (140 Heliox, 141 Airox), whose data was ana-lyzed. Treatment was delivered via facemask (nasal cannula, if thefacemask intolerant) 6 continuous positive airway pressure (CPAP).Severe bronchiolitics received CPAP from the start. Primary end pointwas length of treatment (LoT) required to alleviate hypoxia and respira-tory distress. Secondary end-points were proportion of subjects needingCPAP; CPAP (LoT); and change in respiratory distress score.

RESULTS: Analysis by intention to treat (all subjects); median LoT (inter-quartile range, days): Heliox 1.90 (1.08–3.17), Airox 1.87 (1.11–3.34), P =.41. Facemask tolerant subgroup: Heliox 1.46 (0.85–1.95), Airox 2.01 (0.93–2.86), P = .03. Nasal cannula subgroup: Heliox 2.51 (1.21–4.32), Airox 2.81(1.45–4.78), P = .53. Subgroup started on CPAP: Heliox 1.55 (1.38–2.01),Airox 2.26 (1.84–2.73), P = .02. Proportion of subjects needing CPAP:Heliox 17%, Airox 19%, O.R. 0.87 (0.47–1.60), P = .76. Heliox reduced re-spiratory distress score after 8 hours (mixed models estimate, 20.1298;P , .001). The effect was greater for facemask compared with nasalcannula (mixed models estimate, 0.093; P = .04).

CONCLUSIONS: Heliox therapy does not reduce LoT unless given via a tight-fitting facemask or CPAP. Nasal cannula heliox therapy is ineffective.Pediatrics 2013;131:661–669

AUTHORS: Mina M. Chowdhury, MB, ChB,a Sheila A.McKenzie, FRCP,b Christopher C. Pearson, FRACP,c SiobhanCarr, FRCPCH,b Caroline Pao, MRCP,b Arvind R. Shah,FRCPCH,c Elizabeth Reus, MSc,a Joseph Eliahoo, PhD,e

Fabiana Gordon, PhD,e Hubert Bland, MB, ChB,f and ParvizHabibi, PhD, FRCP, FRCPCHa

aDepartment of Pediatrics, Wright Fleming Institute, ImperialCollege, London, United Kingdom; bDepartment of Pediatrics,Royal London Hospital Whitechapel, London, United Kingdom;cDepartment of Pediatrics, Women’s and Children’s Hospital,North Adelaide, Australia; dDepartment of Pediatrics, NorthMiddlesex University Hospital, London, United Kingdom;eStatistical Advisory Service, Imperial College, London, UnitedKingdom; and fSurrey Clinical Research Centre, University ofSurrey, Guildford, United Kingdom

KEY WORDSrandomized controlled trial, heliox, bronchiolitis

ABBREVIATIONSAirox—mixture of 21% oxygen + 79% nitrogenBREATHE—Bronchiolitis Randomized Controlled Trial Emergency-Assisted Therapy with Heliox—An EvaluationCPAP—continuous positive airway pressureFiO2—fraction of inspired oxygenFM—facemaskHeliox—mixture of 21% oxygen + 79% heliumLoT—length of treatmentNC—nasal cannulaRCT—randomized controlled trialRSV—respiratory syncytial virusSpO2—percutaneous oxygen saturation

Dr Chowdhury and Dr Habibi designed the study and wrote thetrial protocol and manuscript; Dr McKenzie, Dr Bland, Dr Carr,Dr Pao, Dr Shah, and Dr Pearson contributed to the study designand writing of manuscript; Dr Habibi was the chief investigator;Dr Chowdhury was the trial coordinating investigator; Dr McKenzie,Dr Carr, Dr Pao, Dr Shah, and Dr Pearson were principal inves-tigators at collaborating centers; Dr Habibi, Dr Chowdhury, andMs Reus coordinated study conduct and data collection at thecenters and played a major role in supervision of the trialnurses involved in recruitment; and Dr Eliahoo and Dr Gordonperformed the statistical analyses and contributed to thewriting and approval of the final manuscript.

This trial has been registered at http://www.controlled-trials.com/ISRCTN18238432

www.pediatrics.org/cgi/doi/10.1542/peds.2012-1317

doi:10.1542/peds.2012-1317

Accepted for publication Dec 18, 2012

Address correspondence to Parviz Habibi, PhD, FRCP, FRCPCH,Reader and Consultant in Pediatric Intensive Care andRespiratory Medicine, Wright Fleming Institute, Imperial College,Norfolk Place, London W2 1PG, United Kingdom. E-mail: p.habibi@imperial.ac.uk

(Continued on last page)

PEDIATRICS Volume 131, Number 4, April 2013 661

ARTICLE

at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

Acute viral bronchiolitis is a leadingcause of infant hospitalization, witha rising incidence and health-economicburden in developed countries.1–3 Inthe United States, ∼75 000 respiratorysyncytial virus (RSV)-positive bronchio-litic infants are hospitalized each year.4,5

Stang et al3 estimated the annual costburden of this to be between US$365and $585 million. Although there aremany treatments, few have a strongevidence base or have demonstrateda reduced length of hospital stay orneed for respiratory support,6 with ox-ygen being the mainstay of therapy. Amixture of 21% oxygen + 79% helium(Heliox) is lighter than air or oxygen,promoting laminar flow in areas ofturbulence or airway narrowing andthus may improve respiratory distressand wheezing.7 Heliox also reduces re-spiratory system resistance, hasa higher binary diffusion coefficient forCO2 and O2 and therefore may enhancealveolar gas exchange and lung re-cruitment, and is an inert gas with anexcellent safety profile.8–10 Heliox maytherefore be a useful therapy in bron-chiolitis that is associated with smallairway inflammation causing increasedrespiratory system resistance and in-creased work of breathing. Since 1996when Paret et al11 successfully treateda bronchiolitic infant in respiratoryfailure with Heliox, 9 studies (combinedtotal of 172 infants) have investigatedHeliox treatment in bronchiolitis.7,9,12–18

Six of these studies7,9,13,16–18 reportedvarious clinical benefits, including im-provement of hypercapnea and re-spiratory distress. These studies hadsmall sample sizes, used different de-livery methods for Heliox, and were notalways blinded, and only 1 study16

assessed the effect of Heliox on length oftreatment (LoT) or hospital stay. Fur-thermore, regardless of clinical efficacy,heliox must be cost-effective. The cost ofa single bed day on a pediatric ward is$1947.19,20 Even reducing the hospitalLoT by 0.5 dayswould save $974,which is

equivalent to 14 cylinders of Heliox ata cost of $70 each.21 In our experience,only 3 to 5 cylinders per day are con-sumed during Heliox therapy, thus sup-porting our hypothesis that Heliox maybe cost-effective if it reduces duration ofhospital treatment. The 2010 CochraneReview concluded that additional largerandomized controlled trials (RCTs)were needed to investigate the deliverysystem for Heliox and determine itstherapeutic role in bronchiolitis.22 Wetherefore report the largest phase III RCT,called Bronchiolitis Randomized Con-trolled Trial Emergency-Assisted Therapywith Heliox—An Evaluation (BREATHE).Results of this work have been pre-viously published as an abstract.23

METHODS

Setting

A prospective, double-blind RCT wascarried out in the bronchiolitis seasonsduring 2005 to 2008 across 4 centers inthe United Kingdom and Australia.

Participants

Pediatricians in the emergency de-partments or pediatric wards of par-ticipating hospitals, independent ofthe BREATHE study, assessed infants(or 12-month corrected age if pre-mature). They clinically determined ifthe infants had a diagnosis of bron-chiolitis (history of upper respiratorytract infection followed by wheezing,coughing, breathing difficulty, or chestcrackles on auscultation) and if theyneeded hospitalization for respiratorydistress or hypoxia (percutaneousoxygen saturation [SpO2] , 93% inroom air). Exclusion criteria were asfollows: imminent intubation; SpO2 ,93% despite 15 L/minute O2 via non-rebreathing facemask (FM); trache-ostomy; participation in another studyin the previous 4 weeks; salbutamol,epinephrine, or ipratropium therapywithin 1 hour or systemic steroidswithin 4 hours of entry into the study;

and bronchiolitis readmission within24 hours of exit from BREATHE. Anindependent data committee moni-tored safe conduct throughout thestudy. The trial was registered in-ternationally and had independentethics committee approval.

Interventions

Heliox or a mixture of 21% oxygen + 79%nitrogen (Airox), labeled as gas A or B,was the treatment or control in-tervention with additional oxygen ti-trated via Y-connection tubing, resultingin 2 gas mixes: A or B 6 additional ox-ygen. Gas delivery was by a tight-fitting3-valve, nonrebreathing facemask (FM;1192; Intersurgical) or a nasal cannula(NC; BC 2745-20; Fisher & PaykelHealthcare) if the subject was FM in-tolerant. Gas A or B drove the continuouspositive airway pressure (CPAP) device(EME infant flow driver; CareFusion).

Outcome Measures

The primary end point was the total LoTto alleviate hypoxia (SpO2 $ 93% inroom air) and respiratory distress(minimal work of breathing). LoT wascalculated from the start to successfulstop of the trial gas, as defined by clin-ical stability (minimal work of breathingand SpO2 $ 93%) for 1 hour breathingroom air. Minimal work of breathingwas qualified as having a normal re-spiratory rate, no cyanosis, no nasalflaring, no tracheal tug or grunting, nohead bobbing, and no use of accessorymuscles except for mild intercostalrecessions. Secondary end points wereproportion of each treatment groupneeding CPAP and the change in re-spiratory distress over time measuredby the Modified Wood’s Clinical AsthmaScore (Table 1). We used a scoring toolto assess change in respiratory distressover time, similar to that used in pre-vious heliox studies.13,16 The scoring toolwas a modified version of the originalWood’s Clinical Asthma Score.24

662 CHOWDHURY et al at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

At each assessment period, the type ofdelivery (FM or NC) and FM tolerancewas prospectively recorded on the trialclinical report forms to allow analysisof LoT in the FM and NC subgroups. Wealso recorded the duration of CPAPtherapy (CPAP LoT) in a group of severebronchiolitic subjects who were com-menced on CPAP from the beginning oftreatment to compare the impact ofHeliox versus Airox on duration of CPAP.

Management of bronchiolitis was stan-dardized to a strict protocol across trialcenters: gas A or B6 oxygen humidifiedvia MR850 (Fisher & Paykel Healthcare),minimal handling, attention to hydrationstatus, and careful airway toilet/suction.Intravenous fluids were preferred overnasogastric feeding in subjects with se-vere respiratory distress. Bronchodilator,epinephrine, or steroid use constitutedtrial protocol violations. A team ofBREATHE trial nurses ensured adherenceto these standards and the trial protocol.

The BREATHE treatment protocol isshown in Fig 1. Subjects had FM therapyfor 30 minutes. If they were FM in-tolerant, the NC protocol was used. Ifsubjects were hypoxic despite FM/NC,the CPAP protocol was started. FM tol-erance was recorded at each assess-ment period, and FM tolerant wasstrictly defined as mask on all of thetime except for oronasal suction andfeeding. FM intolerance was defined asincreasing agitation, distress, shakinghead from side to side, and pulling theFM off the face for 30 minutes. NC was

started in subjects who remained FMintolerant for 30 minutes. The optimumflow rate for trial gas was based on ti-tration against oxygen to achieve SpO2$ 93% using the minimum flow of ad-ditional oxygen. For FM therapy, themaximum combined flow rate (gas A/B+ oxygen) was 10 L/minute and for NCtherapy it was 3 L/minute based onconsensus of practice.25 CPAP was star-ted if subjects were hypoxic (SpO2 ,93%) despite oxygen.4 L/minute via FMor .2 L/minute via NC. CPAP was dis-continued once subjectswereweaned to1 to 2 cmH2O pressure and were nolonger hypoxic in fraction of inspiredoxygen (FiO2) , 0.4 for 1 hour. CPAPfailure was defined as hypoxia (SpO2 ,93%) despite 9 cmH2O pressure and FiO20.6, whereupon subjects exited the trial.

Those subjects with severe bronchiolitisatpresentation,whorequiredimmediatecommencement of CPAP driven by gas Aor B, followed the CPAP protocol (Fig 1).

Sample Size

Sample size calculation (using nQueryAdvisor v4.0) was based on an unpairedt test with 80% power for a 2-sided a of5% to detect a 0.75-day LoT reduction. Abaseline mean LoT for bronchiolitis of2.7 days (SD= 2.3 days)was assumed.26,27

The calculation showed that 298 sub-jects would be needed to achieve areduction in total LoT by 0.75 days.Three-quarters of a day duration wasselected both from the point of viewof feasibility of recruitment and also

representing a clinically importantthreshold (eg, a 0.75-day LoT reductionin a center that treats 100 infants wouldsave 75 bed-days and was considereda positive impact on health economics).

Randomization and Blinding

After written informed consent wasobtained from parents, patients wereenrolled and allocated to Gas A or B bytelephone using computer-stratifiedblock-randomization. Parents/legal guard-ians and clinical/study personnel wereblinded to randomization sequence andallocation. Randomization codes re-mained secure until the end of the trial.For blinding, identical cylinders markedGas A or B and identical equipment andconnectionswereused. TheCPAPoxygendial was blanked off and FiO2 was reg-ulated by the LED display. The air inlet tothe CPAP device was modified to deliverGas A or B using identical connectors.

Statistical Methods

The primary endpoint (LoT) was ana-lysed blind, for all subjects, based onintention to treat, in order to determineany difference in LoT between treat-ment groups known as gas A and gas B.Subgroup LoT data was analyzed forFM, NC subjects and for CPAP subjectswho were severe enough to warrantCPAP from the start. The Mann-Whitneytest was used to compare LoT betweentreatment groups (as data wasskewed), with results summarized asmedians with interquartile ranges

TABLE 1 Modified Wood’s Clinical Asthma Score

SCORE 0 1 2Pulse oximetry SpO2 . 93%: Room Air SpO2 , 94%: Room Air SpO2 , 94%: 40% FiO2Accessory muscle use NONE of:

• Recessions• Head bobbing• Nasal flaring• Tracheal tug

ANY one of:• Recessions• Head bobbing• Nasal flaring• Tracheal tug

ALL four of:• Recessions• Head bobbing• Nasal flaring• Tracheal tug

Cyanosis No Yes – In room air Yes – In 40% FiO2Cerebral function Normal Depressed or Agitated ComaBreath sounds Normal Unequal Decreased or absentExpiratory wheezing None Yes

Maximum score = 11.

ARTICLE

PEDIATRICS Volume 131, Number 4, April 2013 663 at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

FIGURE 1BREATHE treatment protocol.

664 CHOWDHURY et al at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

(IQR). Fisher’s exact test was used tocompare proportions progressing toCPAP in the two groups. All analyseswere two tailed with an alpha level setat 0×05. STATA 10 and SPSS 17 wereused for analysis. Change in respira-tory distress over time measured byMWCAS was analyzed for all subjectsas well FM and RSV positive (RSV1)subgroups. Mixed Models methodologywas used since it takes into accountcorrelatedmeasures (gas type, gender,birth gestation, age, weight, virus sta-tus, temperature, heart rate, re-spiratory rate, SpO2). Due to the nature

of the data, the square root trans-formation of MWCAS was used as thedependant variable for the modelling.The modelling and estimation of theeffects of interest was carried out bythe PROCMIXED routine in SPSS version17, with a significance level set at 5%.

RESULTS

Participant Flow, Recruitment,Baseline Characteristics, andNumbers Analyzed

Infants presenting with any respira-tory signs or symptoms were screened

between the period of 2005 to 2008. Atotal of 361 patients with clinically di-agnosed bronchiolitis were consideredfor eligibility. Consent was obtained for319 subjectswhowere randomized andenrolled into the study. The consortflowchart (Fig 2) shows that 3 subjectswere excluded (2 withdrawal of con-sent and 1 screening failure); there-fore, 316 subjects were allocated totreatments. An additional 35 subjectswere excluded because of protocol vi-olation, consent withdrawal, screeningfailure, hospital clinician’s decision, orpremature disruption of therapy. Thus,

FIGURE 2Consort flowchart.

ARTICLE

PEDIATRICS Volume 131, Number 4, April 2013 665 at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

281 subjects (140 Heliox and 141 Airox)with similar baseline characteristics(Table 2) completed the study, and theirdata was analyzed. The results aresummarized in Table 3. There were nohospital readmissions for subjectswho had completed or exited the trial.

Length of Treatment

Analysis of data from all 281 subjectsshowed no difference in median LoTbetween treatment groups [Heliox 1.90days (interquartile range 1.08–3.17)compared to Airox 1.87 days (inter-quartile range 1.11–3.34), P 5 .41].However, LoTwas significantly reducedin favor of Heliox for FM-tolerant sub-jects [Heliox, 1.46 days (interquartilerange 0.85–1.95); Airox, 2.01 days

(interquartile range 0.93–2.86) P = .03].A more notable reduction in LoT wasseen in RSV+ subjects [Heliox, 1.31 days(interquartile range 0.61–1.91); Airox,2.18 days (interquartile range 1.40–2.95) P = .004]. There was no differencein LoT for NC subjects [Heliox 2.51 days(interquartile range1.21–4.32), Airox2.81 days (interquartile range 1.45–4.78) P = .53].

Averting Need for CPAP

Analysis of data from all 281 subjectsshowed no reduction in proportion ofcases progressing to CPAP [24 of 140Heliox subjects (17%) vs 27 of 141 Airoxsubjects (19%); odds ratio 0.87 (0.47–1.60), P 5 .76]. However, for FM toler-ant RSV1 subjects there was a 66%

reduction in proportions requiring CPAPin favor of Heliox, at borderline signifi-cance [3 of 27 Heliox subjects (11%) vs10 of 31 Airox subjects (32%); odds ratio0.26 (.07–1.02), P5 .76]

Impact on CPAP Efficacy

Heliox significantly reducedmedian LoTfor severe bronchiolitic subjects whowere started directly onto CPAP [Heliox1,55 days (interquartile range 1.38–2.01), Airox 2.26 days (interquartilerange 1.84–2.73); P = .02].

Impact on Respiratory Distress

Heliox reduced respiratory distress inall 281 subjects across all time pointsand statistically significant at 8 hoursonwards. MWCAS (mixed models esti-mate = 20.1298; 95% confidence in-terval 20.202 to 20.057, P , .001).Regardless of gas type, FM was moreeffective than NC (mixed models esti-mate5 0.093; 95% confidence interval0.005 to 0.181, P = .04).

Adverse Events

Six subjects required intubationfor different reasons. In one casethere was CPAP equipment malfunc-tion which precipitated emergency

TABLE 2 Baseline Characteristics of Subjects in the Study

Baseline Characteristics Units Heliox (N = 140) Airox (N = 141)

Gender male:female ratio 1:1.59 1:1.52Gestation at birth wk 39.0 (38.0240.0) 40.0 (38.0240.0)Age at presentation wk 10.90 (5.85225.50) 17.70 (6.80229.40)Weight at presentation kg 5.65 (4.3427.70) 5.70 (4.4027.70)NPA +ve at presentation N (% of cases) 111 (79.3%) 116 (82.3%)Temperature °C 37.0 (37.0238.0) 37.0 (37.0238.0)Heart rate beats per min 152.0 (136.02165.0) 148.5 (133.52168.0)Respiratory rate breaths per min 56.0 (44.0262.0) 53.0 (47.0263.5)SpO2 in room air % 92.0% (89.0295.0) 91.0% (89.0294.0)Modified Wood’s ClinicalAsthma Score

Maximum score of 11 3.0 (2.023.0) 3.0 (2.024.0)

Values are medians and IQRs. NPA, nasopharyngeal aspirate virus detection.

TABLE 3 Summary of BREATHE Findings

Impact on LoT Intervention N Mean LoT (95% Confidence Interval), days Median LoT (Interquartile Range), days P

All subjects Heliox 140 2.268 (1.993 to 2.544) 1.902 (1.08323.173) .41Airox 141 2.487 (2.180 to 2.794) 1.865 (1.11423.344)

NC (6CPAP) Heliox 40 2.952 (2.335 to 3.569) 2.505 (1.21024.315) .53Airox 47 3.296 (2.643 to 3.948) 2.810 (1.45024.780)

FM (6CPAP) Heliox 44 1.538 (1.234 to 1.841) 1.464 (0.85221.947) .03Airox 40 2.236 (1.744 to 2.728) 2.006 (0.92822.859)

FM (6CPAP), RSV+ Heliox 27 1.411 (0.982 to 1.841) 1.310 (0.60821.906) .004Airox 31 2.456 (1.868 to 3.044) 2.179 (1.39622.951)

CPAP LoT Heliox 9 1.619 (1.301 to 1.934) 1.552 (1.38022.013) .02Airox 12 2.380 (1.773 to 2.986) 2.258 (1.84422.727)

Proportion of cases needing CPAP Percent Odds Ratio (95% Confidence Interval)All cases Heliox 24/140 17% 0.87 .78

Airox 27/141 19% (0.4721.60)FM, RSV+ Heliox 3/27 11% 0.26 .07

Airox 10/31 32% (0.0721.02)

Impact on MWCAS Comparison Estimate of Fixed Effects (95% Confidence Interval)Results are Heliox effect relative

to Airox over timeAll patients 20.1298 (20.202 to 20.057) ,.001FM relative to NC 0.093 (0.005 to 0.181) .04

666 CHOWDHURY et al at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

intubation. Two subjects suffered re-current apnoeas immediately followingenrolment (i.e. screening failure). Onesubject due to receive CPAP was intu-bated and transferred out because nohigh dependency bed available whileone subject was intubated because ofdelay in high dependency bed avail-ability. One subject had a previouslyundetected patent ductus arteriosusand was therefore referred to thecardiologists.

DISCUSSION

Bronchiolitis is a leading cause of infanthospitalization worldwide yet, to date,oxygen and supportive care are themainstay of treatment for the vastmajority of patients. The excellent safetyprofile and unique physical properties ofHeliox position it as a potentially usefultherapy for acute bronchiolitis. TheBREATHE study found the deliverymethodof Heliox to be crucial to its efficacy: itwas only beneficial if given by a tight-fitting FM or by CPAP. Thus, if deliveredeffectively, Heliox therapy significantlyreduced length of hospital treatment atdifferent points of care (ward, PICU). TheBREATHE study was powered to detecta 0.75-day reduction in LoT. However,the statistically significant finding ofa 0.55-day reduction in LoT in FM-tolerantpatients is still a clinically useful find-ing, given the very significant pressureon hospital beds every winter imposedby bronchiolitis. Furthermore, given theneedto justify thehealth-economiccostofHeliox, the results of the subgroup anal-ysis support the rational use of Heliox inRSV+ patients who showed a greaterreduction in LoT (0.87 day).

Heliox rapidly reduced respiratory dis-tress, making it a potentially usefulstabilization tool in emergency care. Theeffect was greater when consideringRSV+ subjects (archetypical bronchiol-itis), consistent with Martinon-Torres’findings on his group of RSV+ subjects.16

However, the authors of that study found

Heliox shortened the length of hospitalstay. Because many other variables canaffect the length of hospital stay, wechose to measure total LoT as the pri-mary end point, because it is directlylinked to therapy.

We found Heliox conferred no benefitover oxygen when delivered by NC atflow rates of #3 L/minute. We believethe difference in efficacy between FMand NC is caused by several factors. Atight-fitting FM is used in clinical sit-uations when the highest FiO2 needs tobe delivered with minimal air entrain-ment. Delivering a high concentrationof helium must also be very importantfor Heliox therapy, and our previousfindings28,29 also provide support foruse of the nonrebreathing FM. The useof an NC is subject to significant airentrainment, and thus variable FiO2, asdemonstrated by Sung et al.30 The dis-advantages of the NC would be magni-fied when using Heliox, a much lightergas. Furthermore, in small infants,nasal prongs may also increase re-sistance to nasal airflow and work ofbreathing. It is not surprising thereforethat NC therapy with Heliox at conven-tional flow rates is ineffective.

The BREATHE study is the only RCT thusfar to investigate the rate of CPAP use inbronchiolitis. However, the observed66% reduction only reached borderlinestatistical significance, which may bebecauseoftherelativelysmallnumberofCPAP subjects (n = 58). Martinon-Torreset al17 in their Heliox-CPAP study dem-onstrated a reduction in work ofbreathing and improved CO2 clearance.The BREATHE study also enabled as-sessment of the impact of Heliox onCPAP efficacy. We analyzed data forsubjects who were started on CPAPfrom the beginning of the trial becausetheir pathophysiology had not yet beenaltered by previous therapy. We foundthat CPAP duration was significantlyreduced if Heliox was the driving gas forCPAP. The numbers of subjects totaled

only 21, so we cannot draw any strongconclusions. However, if the finding ofreduced treatment time was to be rep-licated in a larger CPAP study, thiswouldrepresent a significant health-economicbenefit for using Heliox to drive CPAP incases of severe bronchiolitis.

We did not select patients based onclinical severity but screened consec-utive infants presenting with any re-spiratory signs or symptoms and found361 patients with clinically diagnosedbronchiolitis. We could not rule out thepossibility that some patientsmay havehad asthma. However, any cases ofasthma-bronchiolitis overlap wouldhave been balanced out between the 2treatment groups through the processof randomization. Virus detection fromnasopharyngeal aspirate was carriedout by the hospitals on 281 of the en-rolled patients. We found 227 of 281(�80%) were virus positive, with RSVaccounting for the vast majority. Thecenters routinely assayed only for RSV,para-influenza, adenovirus, Flu A/B,and rhinovirus. Therefore, the 80% vi-rus positive rate (Table 2) was mostlikely an underestimate and we areconfident that our clinical selectioncriteria captured mostly viral bron-chiolitis.

The BREATHE study is the largest phaseIII, multicenter, double-blinded RCT ofHeliox in bronchiolitis. It attempted toresolve the challenges of blinding. Theuse of special hosingmaterial, identicalin appearance for Heliox and Airox en-sured that there was no difference insound generation that could havealerted the investigator to identify thestudy gas.

The BREATHE study linked efficacy to themode of delivery: CPAP or tight-fittingnonrebreathing FM. However, thelarge number of subjects treated by NCmeant that the positive findings in favorofHelioxwere limited toonly84subjectswho tolerated the FM. Although the useof the FM was a major limitation of this

ARTICLE

PEDIATRICS Volume 131, Number 4, April 2013 667 at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

study, we chose it as the delivery in-terface because it is readily available,and others13,16 had reported effectiveHeliox delivery by FM. Our previousresearch28,29 also found that a tight-fitting 3-valve, nonrebreathing FMwould deliver the highest concentra-tion of oxygen and helium. However,FMs are generally poorly tolerated inyoung children, and this was also ourexperience. To maximize FM compli-ance in their study, Martinón-Torreset al16 used swaddling to comfort-ably restrain subjects with the FM heldagainst the face by soft elasticizedtube netting applied over the head. Weused a bonnet to allow a tighter-fittingFM and help prevent the elastic bandfrom slipping down, used swaddling,and encouraged staff and mothers tomaximize compliance. Despite ourbest efforts, many subjects did nottolerate the FM, and it proved impos-sible to apply it continuously. Maskintolerance was highlighted after re-cruitment of the first few subjects intothe trial when we found nursing staffwishing to use an NC, the commonlyused mode of oxygen delivery forinfants. It was considered unethical towithhold a trial gas treatment if sub-jects had already started to improve.We therefore considered delivery oftrial gas via NC because Williamset al31 successfully used NC with Helioxin infants, and oxygen is already con-ventionally administered via NC. There-fore, NC was included as a protocolamendment for subjects who were FMintolerant.

We prospectively studied a cohort ofclinically diagnosed bronchiolitic in-fants, regardless of severity and viraletiology, enabling us to identify a treat-ment for bronchiolitis that can be usedacross different modalities (FM andCPAP) and points of care (emergencydepartment, ward, and PICU). Nonethe-less, this approach to recruitmentresulted in a smaller proportion of

severe cases, making it difficult to con-clusively study outcomes such as CPAP.The latter is very important becauseCPAP compared with standard treat-ment has been shown to improve ventila-tion, with a reduction in hypercapnea,32

with growing consensus that CPAP ther-apy prevents deterioration and need formechanical ventilation.

We did not collate data on length of staybut rather focused on length of treat-ment. The study was powered for thelatter because LoT was a clear anddefinitive endpoint in the disease pro-cess.

The BREATHE study was not powered toinvestigate intubation rates, whichwould also have been informative. Only8 of 316 (2.8%) needed intubation.Therefore, amuch larger sample size ora moderate number of a more selectedsevere group of patients would beneeded to investigate the impact ofHeliox on intubation rates. The BREATHEprotocol (part B, not yet carried out)had originally been designed to in-vestigate intubation rate in a moresevere subgroup of bronchiolitics (de-fined as already requiring CPAP). Abaselinefigure of 35%was assumed forthe bronchiolitis intubation rate (de-fined as CPAP failure) derived fromdatafor our PICU at St Mary’s hospital. Thepower calculation showed that a totalof 86 severe bronchiolitics on CPAPwould need to be recruited to detecta reduction in the intubation rate from35% to 10% with 80% power.

The BREATHE study has highlighted theneed to review our approach to re-spiratory care in general: use of FM hastolerance issues, and NC (at conven-tional flow rates) has limited efficacy.Although NC delivery of Heliox at flowrates ,3 L/minute was ineffective, wetheorize that at higher flow rates (tol-erated based on our anecdotal experi-ence), Heliox therapy by nasal cannulamay be effective. A well-designed RCTcomparing 3-valve nonrebreathing FM

versus high-flow NC therapy is neededto identify the optimal method of de-livery. Furthermore, a sufficiently pow-ered RCT of Heliox-driven versusconventional Airox-driven CPAP woulddetermine whether Heliox reduces theneed for mechanical ventilation.

In the meantime, the clinical practicerecommendations arising from theBREATHE study findings are as follows:

� Heliox therapy should be startedfor bronchiolitic infants who re-quire hospital admission for treat-ment of hypoxemia or respiratorydistress.

� If the use of Heliox therapy needsto be rationalized, it could be tar-geted to those who are RSV posi-tive.

� Heliox therapy should only be de-livered via a tight-fitting nonre-breathing FM or CPAP, as per theprotocol outlined in the BREATHEstudy.

CONCLUSIONS

The BREATHE study showed that thedeliverymethodofHeliox is critical to itsefficacy. Heliox is effective if deliveredvia a tight-fitting nonrebreathing FM orCPAP but not via a NC at conventionalflowrates.With effective delivery, Helioxreduces the LoT, alleviates respiratorydistress, improves CPAP efficacy, andmay reduce the need for CPAP. A moreacceptable patient interface for effec-tive delivery remains the challenge forindustry if Heliox is to be more widelyused in pediatric respiratory care.

ACKNOWLEDGMENTSWe thank the patients, families, andnursing, medical, and other healthcarestaff across all the hospitals, whosecooperation and support allowed usto undertake the BREATHE trial. We es-pecially thank the BREATHE team mem-berswhosemeticulousattention to trialconduct ensured the rigor of this study.

668 CHOWDHURY et al at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

REFERENCES

1. Nair H, Nokes DJ, Gessner BD, et al. Globalburden of acute lower respiratory infectionsdue to respiratory syncytial virus in youngchildren: a systematic review and meta-analysis. Lancet. 2010;375(9725):1545–1555

2. Shaw RJ, Djukanovic R, Tashkin DP, MillarAB, du Bois RM, Orr PA. The role of smallairways in lung disease. Respir Med. 2002;96(2):67–80

3. Stang P, Brandenburg N, Carter B. The eco-nomic burden of respiratory syncytial virus-associated bronchiolitis hospitalizations.Arch Pediatr Adolesc Med. 2001;155(1):95–96

4. Healthcare Cost and Utilization Project(HCUP). Patient and hospital characteristicsfor ICD-9-CM all-listed diagnosis code 466.11(RSV Bronchiolitis). Available at: http://hcupnet.ahrq.gov/. Accessed April 10, 2011

5. Shay DK, Holman RC, Newman RD, Liu LL, StoutJW, Anderson LJ. Bronchiolitis-associatedhospitalizations among US children, 1980-1996. JAMA. 1999;282(15):1440–1446

6. Bourke T, Shields M. Bronchiolitis. Clin Evid(Online). Accessed April 11, 2011. doi:pii: 0308

7. Cambonie G, Milési C, Fournier-Favre S,et al. Clinical effects of heliox administra-tion for acute bronchiolitis in younginfants. Chest. 2006;129(3):676–682

8. Kneyber MC, van Heerde M, Twisk JW, PlötzFB, Markhors DG. Heliox reduces re-spiratory system resistance in respiratorysyncytial virus induced respiratory failure.Crit Care. 2009;13(3):R71

9. Martinón-Torres F, Rodríguez-Núñez A,Martinón-Sánchez JM. Nasal continuouspositive airway pressure with heliox versusair oxygen in infants with acute bronchio-litis: a crossover study. Pediatrics. 2008;121(5). Available at: www.pediatrics.org/cgi/content/full/121/5/e1190

10. Nawab US, Touch SM, Irwin-Sherman T, et al.Heliox attenuates lung inflammation andstructural alterations in acute lung injury.Pediatr Pulmonol. 2005;40(6):524–532

11. Paret G, Dekel B, Vardi A, Szeinberg A, LotanD, Barzilay Z. Heliox in respiratory failuresecondary to bronchiolitis: a new therapy.Pediatr Pulmonol. 1996;22(5):322–323

12. Gross MF, Spear RM, Peterson BM. Helium-oxygen mixture does not improve gas ex-change in mechanically ventilated childrenwith bronchiolitis. Crit Care. 2000;4(3):188–192

13. Hollman G, Shen G, Zeng L, et al. Helium-oxygen improves Clinical Asthma Scores inchildren with acute bronchiolitis. Crit CareMed. 1998;26(10):1731–1736

14. Iglesias FC, Huidobro FB, Miguez NC, GuerreroSM, Vazquez LP, Maranon PR. Heliox-drivenbronchodilator nebulization in the treat-ment of infants with bronchiolitis. AnPediatr (Barc). 2009;70(1):40–44

15. Liet JM, Millotte B, Tucci M, et al; CanadianCritical Care Trials Group. Noninvasivetherapy with helium-oxygen for severebronchiolitis. J Pediatr. 2005;147(6):812–817

16. Martinón-Torres F, Rodríguez-Núñez A,Martinón-Sánchez JM. Heliox therapy ininfants with acute bronchiolitis. Pediatrics.2002;109(1):68–73

17. Martinón-Torres F, Rodríguez-Núñez A,Martinón-Sánchez JM. Nasal continuouspositive airway pressure with heliox ininfants with acute bronchiolitis. RespirMed. 2006;100(8):1458–1462

18. Mayordomo-Colunga J, Medina A, Rey C,Concha A, Los Arcos M, Menéndez S.Helmet-delivered continuous positive air-way pressure with heliox in respiratorysyncytial virus bronchiolitis. Acta Paediatr.2010;99(2):308–311

19. Edwards JD, Rivanis C, Kun SS, Caughey AB,Keens TG. Costs of hospitalized ventilator-dependent children: differences betweena ventilator ward and intensive care unit.Pediatr Pulmonol. 2011;46(4):356–361

20. Katz BZ, Lo J, Sorrentino M. Costs of Re-spiratory Syncytial Virus Infection at a Ter-tiary-Care Children’s Hospital. Pharm TherJ. 2011;28(5):343–345

21. Kass JE. Heliox redux. Chest. 2003;123(3):673–676

22. Liet JM, Ducruet T, Gupta V, Cambonie G.Heliox inhalation therapy for bronchiolitisin infants. Cochrane Database Syst Rev.2010; (4):CD006915

23. Chowdhury MMM, Pao C, Shah A, et al.Helium-oxygen gas mixtures in infants withacute bronchiolitis: results of a phase llldouble blind multi-centre RCT (BREATHEStudy). Am J Respir Crit Care Med. 2010;181:A2281. http://ajrccm.atsjournals.org/cgi/reprint/181/1_MeetingAbstracts/

24. Wood DW, Downes JJ, Lecks HI. A clinicalscoring system for the diagnosis of re-spiratory failure. Preliminary report onchildhood status asthmaticus. Am J DisChild. 1972;123(3):227–228

25. Frey B, Shann F. Oxygen administration ininfants. Arch Dis Child Fetal Neonatal Ed.2003;88(2):F84–F88

26. Kini NM, Robbins JM, Kirschbaum MS,Frisbee SJ, Kotagal UR; Child Health Ac-countability Initiative. Inpatient care foruncomplicated bronchiolitis: comparisonwith Milliman and Robertson guidelines.Arch Pediatr Adolesc Med. 2001;155(12):1323–1327

27. National Health Service. The informationcentre for health and social care. Hospitalepisode statistics for England (2003/04).Available at: http://www hesonline nhs uk.Accessed August 9, 2004

28. Chowdhury MMM, Ng C, Habibi P. Helioxtreatment in spontaneously breathingchildren. Proc Am Thorac Soc. 2006;3:A405

29. Chowdhury MMM, Ng C, Habibi P. Manage-ment of hypoxia: optimizing FiO2 in spon-taneously breathing children. Proc AmThorac Soc. 2006;3:A404

30. Sung V, Massie J, Hochmann MA, Carlin JB,Jamsen K, Robertson CF. Estimating in-spired oxygen concentration delivered bynasal prongs in children with bronchiolitis.J Paediatr Child Health. 2008;44(1–2):14–18

31. Williams J, Stewart K, Tobias JD, BerkenboschJW. Therapeutic benefits of helium-oxygendelivery to infants via nasal cannula. PediatrEmerg Care. 2004;20(9):574–578

32. Thia LP, McKenzie SA, Blyth TP, Minasian CC,Kozlowska WJ, Carr SB. Randomised con-trolled trial of nasal continuous positiveairways pressure (CPAP) in bronchiolitis.Arch Dis Child. 2008;93(1):45–47

(Continued from first page)

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

Copyright © 2013 by the American Academy of Pediatrics

FINANCIAL DISCLOSURE: Dr Bland was a Clinical Director of British Oxygen Company (BOC) Medical during the period of conduct of this trial; the other authorshave indicated they have no financial relationships relevant to this article to disclose.

FUNDING: Supported by a research grant from BOC Medical and equipment support from Fisher and Paykel Healthcare, EME (Electro-Medical Equipment)Carefusion, and Covidien Nellcor.

ARTICLE

PEDIATRICS Volume 131, Number 4, April 2013 669 at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from

DOI: 10.1542/peds.2012-1317; originally published online March 18, 2013; 2013;131;661Pediatrics

Hubert Bland and Parviz HabibiCaroline Pao, Arvind R. Shah, Elizabeth Reus, Joseph Eliahoo, Fabiana Gordon,

Mina M. Chowdhury, Sheila A. McKenzie, Christopher C. Pearson, Siobhan Carr,Randomized Controlled Trial

Heliox Therapy in Bronchiolitis: Phase III Multicenter Double-Blind  

ServicesUpdated Information &

mlhttp://pediatrics.aappublications.org/content/131/4/661.full.htincluding high resolution figures, can be found at:

References

ml#ref-list-1http://pediatrics.aappublications.org/content/131/4/661.full.htat:This article cites 29 articles, 4 of which can be accessed free

Citations

ml#related-urlshttp://pediatrics.aappublications.org/content/131/4/661.full.htThis article has been cited by 1 HighWire-hosted articles:

Subspecialty Collections

_tracthttp://pediatrics.aappublications.org/cgi/collection/respiratoryRespiratory Tractthe following collection(s):This article, along with others on similar topics, appears in

Permissions & Licensing

tmlhttp://pediatrics.aappublications.org/site/misc/Permissions.xhtables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2013 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

at Pontificia Universidad Catolica de Chile on April 25, 2013pediatrics.aappublications.orgDownloaded from