www.elsevier.com/locate/jcfJournal of Cystic Fibrosis 13 (2014) 106–110

Short Communication

Induced sputum compared to bronchoalveolar lavage in young,non-expectorating cystic fibrosis children☆

Hannah Blau a,b,⁎, Barry Linnane c, Rosemary Carzino d,e, Esta-Lee Tannenbaum d, Billy Skoric d,e,Philip J. Robinson d,e,f, Colin Robertson d,e,f, Sarath C. Ranganathan d,e,f

a Graub CF Center and Pulmonary Institute, Schneider Children's Medical Center of Israel, 14 Kaplan Street, Petah Tikva 49202, Israelb Sackler Faculty of Medicine, Tel Aviv University, Ramat Aviv 69978, Israel

c Cystic Fibrosis Unit, Paediatric Department, Mid-Western Regional Hospital, Dooradoyle, Limerick, Irelandd Department of Respiratory Medicine, Royal Children's Hospital, 50 Flemington Road, Parkville, Victoria 3052, Australia

e Infection and Immunity Theme, Murdoch Children's Research Institute, Royal Children's Hospital, 50 Flemington Road, Parkville, Victoria 3052, Australiaf Department of Paediatrics, University of Melbourne, Parkville, Victoria 3052, Australia

Received 21 February 2013; received in revised form 31 May 2013; accepted 31 May 2013Available online 24 June 2013

Abstract

Background: Induced sputum (IS) is feasible and safe in young CF children and is a readily accessible, non-invasive technique. However, it has notbeen compared to bronchoalveolar lavage (BAL), the gold standard for diagnosing lower airway infection.Methods:We compared bacterial yield from IS and BAL in 11 non-expectorating CF children, aged 3 to 7.4 years. IS samples were obtained in 10/11 cases.Results: Eight out of ten had the same predominant bacteria cultured from IS and BAL: Pseudomonas aeruginosa and Stenotrophomonasmaltophilia [1], Staphylococcus aureus [3], and upper respiratory tract flora [4]. In one, Serratia marcescens and Haemophilus parainfluenzaewere cultured from IS alone, whereas in one, non-group B Haemophilus influenzae was cultured from BAL alone.Conclusions: As proof of principle, IS samples showed good bacteriologic correlation with BAL. Larger studies are recommended to confirm IS asa clinically valuable tool and measure for early intervention studies in young CF children.© 2013 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.

Keywords: Cystic fibrosis; Bronchoalveolar lavage; Induced sputum; Infants

1. Introduction

Chronic infection and inflammation play a central role in thelife-shortening lung disease of cystic fibrosis (CF), beginning

Abbreviations: BAL, bronchoalveolar lavage; CF, cystic fibrosis; IS, inducedsputum; URTF, upper respiratory tract flora; P. aeruginosa, Pseudomonasaeruginosa; S. aureus, Staphylococcus aureus; H. influenzae, Haemophilusinfluenzae; S. marcescens, Serratia marcescens.☆ This work was performed at the Department of Respiratory Medicine, RoyalChildren's Hospital, Melbourne, Australia.⁎ Corresponding author at: Pulmonary Institute, Schneider Children's MedicalCenter of Israel, 14 Kaplan Street, Petah Tikva 49202, Israel. Tel.: +972504057141; fax: +972 3 9253147.

E-mail address: hblau@post.tau.ac.il (H. Blau).

1569-1993/$ -see front matter © 2013 European Cystic Fibrosis Society. Publishedhttp://dx.doi.org/10.1016/j.jcf.2013.05.013

in the first few months of life and often silent [1]. Severe,neutrophilic inflammation that leads to bronchiectasis and,eventually, respiratory failure and death, has been identified asearly as two months of age in association with early bacterialand viral infections [2].

Bronchoalveolar lavage (BAL) remains the “gold standard”for defining airway microbiology and inflammation in youngchildren [3], but has important limitations. It is invasive,requires anesthesia, consumes time and resources and is costly.Furthermore, IS can be more readily repeated within the routineCF clinic visit, to determine if, for example, Pseudomonasaeruginosa has been successfully eradicated following aninitial positive BAL or IS culture. BAL often samples onlyone or two segments of the lung, limiting detection of infection.

by Elsevier B.V. All rights reserved.

107H. Blau et al. / Journal of Cystic Fibrosis 13 (2014) 106–110

The Australian Respiratory Early Surveillance Team for CysticFibrosis (AREST CF) demonstrated that structural lung diseasebegins in the first few months of life [4] and before the medianage of 30.5 months when infection with P. aeruginosa wasdetected [5]. Infections with bacteria such as Staphylococcusaureus (S. aureus) and non-typeable Haemophilus influenzaeare more prevalent during early childhood. Focus in particular onS. aureus is increasing, as it is associated with poorer nutritionaloutcomes [6], decline in lung function [7] and bronchiectasis [8]as well as having an impact on CF inflammation and clinicaloutcomes [9].

Therefore, there is an urgent need for repeated, non-invasivemeasures of infection and inflammation for young childrenwith CF in order to impact on development of lung disease.Induced sputum (IS) using hypertonic saline inhalation inexpectorating children and adults is a safe, simple, non-invasiveand repeatable diagnostic tool for obtaining secretions pooledfrom the lower airways [9].

Lower respiratory sampling is required to distinguish lowerairway infection from upper airway colonization and thereforemore appropriately target treatment. We have demonstratedsafety and feasibility of IS previously in young children, with asuperior bacteriologic yield compared to cough swabs [10]. Todate, however, there have been no studies comparing IS to BALfor bacteriologic yield in young, non-expectorating children withCF.We therefore embarked on a pilot study to compare these twoprocedures, within a cohort of young non-expectorating childrenwith CF.

2. Methods

We enrolled non-expectorating children with a confirmeddiagnosis of CF, treated at the CF center at the Royal Children'sHospital, Melbourne, who were undergoing BAL, either aspart of an “Early Surveillance Program” that included annualassessments of BAL, CT and infant lung function [2,6–8], aspart of the Australasian Cystic Fibrosis Bronchoalveolar Lavage(ACFBAL) study [4], or if clinically indicated for microbiologicdiagnosis. This amalgamation of subjects included both stable

Table 1Patient induced sputum collection details.

Patient no. Age (y) Cough amount Cough quality a O2 sat. pre O

1 4.24 None Dry 99 992 4.19 Moderate Productive 96 953 4.04 None Dry 99 994 4.98 None Dry 98 975 5.3 Moderate Productive 99 986 5.08 Moderate Productive 98 967 5.19 None Dry 99 998 7.44 Mild Productive 98 989 4.03 Mild Productive 99 9910 3.02 Moderate Dry 98 98

Physiotherapy type: 1 = bubble PEP, 2 = trampoline, 3 = exercise under supervisioa Cough productivity was assessed subjectively by the physiotherapist prior to orb ‘Distressed’: crying and resisting somewhat.c ‘Very distressed’: crying and resisting a lot (as seen during taking blood tests in

and acutely sick children, making the study population quiteheterogenous. However, in all of these subjects, comparisonof BAL and IS yields is of major importance. As BAL isan invasive procedure, we chose children who were alreadyundergoing the procedure, as the subjects for this comparativepilot study. A sub-group of these children who lived in proximityto the hospital and agreed to come in for a separate visit in orderto undergo an IS procedure, was selected to have an IS specimentaken on a single occasion for microbiological study. Thesechildren were selected consecutively from the potential studypopulation.

Specimens of IS were taken in the CF clinic, 1–13 days afterthe BAL in order not to influence parameters of inflammationas described previously [11]. No change in antibiotic therapyoccurred between samples.

Children were pre-treated with two puffs of salbutamol,100 μg/puff, using a spacer device with mask, followed by aninhalation of 10 cc of 4.5% saline solution via an ultrasonicnebulizer for 10 min. Physiotherapy, adapted to the individual,was then performed to mobilize secretions (Table 1). Oropha-ryngeal suction was performed with a size 10 catheter connectedto a sterile mucus extractor to obtain a clearly visualizedspecimen of sputum. Sputum and BALwere cultured by standardprocedures in the clinical microbiology laboratory of the RoyalChildren's Hospital, using 104 cfu/mL to determine a positiveculture. Sputum clumps were dispersed mechanically, usingsterile glass beads [12].

The study was approved by the Royal Children's HospitalEthics in Human Research Committee (EHRC) ref. no. 25054and parents of participants signed informed consent forms ineach case.

3. Results

Eleven children, aged 3.0 to 7.4 years, participated in thestudy. The details of the IS procedure are shown in Table 1. Theentire procedure took between 20 and 45 min, includinginhalation of salbutamol and hypertonic saline, physiotherapy

2 sat. post Type physio Procedure time (min) Subjective experience

2 45 Tolerated fairly well3 30 Tolerated fairly well2 40 Tolerated well1, 2 45 Distressed b

2 45 Distressed b

2 35 Tolerated fairly well1 20 Tolerated fairly well1 30 Tolerated fairly well3 25 Tolerated well1, 3 30 Very distressed c

n.during physiotherapy.

some young children).

108 H. Blau et al. / Journal of Cystic Fibrosis 13 (2014) 106–110

and suction. The sputum suction maneuver lasted 1–2 min.There was no change in oxygen saturation during suction.

Of the 11 subjects participating, 10 succeeded in providing asample of induced sputum (Table 2). There was no relationshipbetween cough productivity (as described in Table 1) andmicrobiologic yield of the IS procedure, as patients with drycough often had IS cultures which were positive for bacteria. Ofthe 10 specimens collected, 8 had the same predominant bacteriacultured from both the IS and the BAL specimens with identicalbacteria sensitivity profiles. Table 2 also indicates whether thebacteria cultured were scant, moderate or profuse in quantity.Where present, upper respiratory tract flora (URTF) was scantin BAL and profuse in IS cultures. There was no correlationbetween BAL and IS regarding culture ofAspergillus orCandida.Candida was present in two IS specimens and one other BAL

Table 2Microbiology findings of BAL and induced sputum specimens.

Patientno.

Age(y)

BetweenBAL & IS (d)

Patientgroup

Microbiology

BAL

Microorganism Antimicrobial

1 4.24 13 AREST URTF + n/a2 4.19 12 AREST S. aureus +

mixed neg.bacilli +

S. aureus:pen-s, clindamvanco-s, erythrflucox-s, and te

3 4.04 5 AREST No bacterial growth n/a

4 4.98 6 ACFBAL P. aeruginosa +++S. maltophilia +++A. fumigatis +

P.aeruginosa:tobra-s, cipro-samik-s, piper-snorflox-s, tim-and aztreonamS. Maltophilia:cotrimoxazole-

5 5.3 8 ACFBAL URTF ++ n/a

6 5.08 5 ACFBAL S. aureus +++URTF ++

S.aureus: pen-rvanco-s, erythrand teicoplanin

7 5.19 9 ACFBAL URTF + n/a8 7.44 1 Clinical MRSA +++

C. glabrata +

S.aureus: pen-rvanco-s, erythrand teicoplanin

9 4.03 6 AREST URTF + n/a10 3.02 4 AREST H. influenza

(not type b)Ampicillin-r,augmentin-s,cefotaxime-s, a

Bacterial growth: + = scant, ++ = moderate, +++ = profuse; BAL = bronchoalveoEarly Surveillance Team for Cystic Fibrosis’, an early surveillance program studyClinical = BAL was done for clinical diagnostic purposes.d = days; URTF — upper respiratory tract flora; MRSA = methicillin resistant S. aSensitivity to antibiotics: pen = penicillin; vanco = vancomycin; erythro = erythromgentamicin; ceftaz = ceftazidime; amik = amikacin; piper = piperacillin; norflox = no

specimen whereas Aspergillus was found in one BAL but not inthe corresponding IS.

4. Discussion

In this pilot study we have confirmed the feasibility and safetyof IS in non-expectorating children with CF and compared resultsfrom IS with those from BAL. IS samples provided a similarmicrobiologic yield to BAL procedures as proof of principle,although the study was underpowered for formal statisticalanalysis.

Although we aimed to perform IS within a week of BAL,this was not always feasible for logistic reasons, as parentscould not bring the child in for earlier IS. Fortunately, antibiotictherapy remained the same in all cases during this period. It was

IS

susceptibility Microorganism Antimicrobial susceptibility

URTF +++ n/a

ycin-s,o-s,icoplanin-s

S. aureus +++ S. aureus:pen-r, clindamycin-s,vanco-s, erythro-s,flucox-s, and teicoplanin-s

S. marcescens +H. parainf ++URTF

S. marcescens:ampicillin-r,cephalothin-r,cefotaxime-s,gentamicin-s,ceftazidime-s,augmentin-r,cotrimoxazole-s,cefpodoximer-I, tobra-s and cipro-s

genta-s, ceftaz-s,, meropenem-s, cefipime-s,s, pip/tazobactam-s,-s

s

P. aeruginosa +S. maltophilia +

P.aeruginosa: genta-s, ceftaz-s,tobra-s, cipro-s, meropenem-s amik-s,piper-s, cefipime-s, norflox-s, tim-s,pip/tazobactam-s, and aztreonam-sS. Maltophilia:cotrimoxazole-s

URTF +C. albicans ++

n/a

, clindamycin-s,o-s, flucox-s,-s

S. aureus ++URTF +++C. albicans +

S.aureus: pen-r, clindamycin-s,vanco-s, erythro-s, flucox-s,and teicoplanin-s

URTF +++ n/a, clindamycin-r,o-r, flucox-r,-s

MRSA +++ S.aureus: pen-r, clindamycin-r,vanco-s, erythro-r, flucox-r,and teicoplanin-s

URTF +++ n/a

nd cotrimoxazole-r

URTF +++ n/a

lar lavage; IS = induced sputum; y = years; AREST = ‘Australian Respiratory; ACFBAL = the Australasian Cystic Fibrosis Bronchoalveolar Lavage study;

ureus.ycin; flucox = flucloxacillin; tobra = tobramycin; cipro = ciprofloxacin; genta =rfloxacillin; tim = timentin; and pip/tazobactam = piperacillin/tazobactam.

109H. Blau et al. / Journal of Cystic Fibrosis 13 (2014) 106–110

reassuring that the predominant bacteria cultured from BALand IS remained the same in 8 of 10 cases.

In older patients with CF, IS provides an accurate measure ofinfection and inflammation [9], with improvement demonstratedfollowing therapy with intravenous antibiotics [13]. In young CFchildren who are able to expectorate only following hypertonicsaline inhalation, IS yielded similar evidence for inflammationand infection as that from spontaneously expectorating children[14]. In this group, as well as in our previous study [10], themicrobiological yield increased compared to that with spontane-ously expectorated sputum or oropharyngeal cough swabs. Wetherefore did not include a comparison to oropharyngeal coughswabs, which would have made the experience more traumatic.

The higher prevalence and more profuse growth of URTFin IS compared to BAL suggests that IS is more likely to becontaminated by upper airway secretions than BAL specimens.However, as both types of sampling often contain upper airwayflora, we do not believe this negates the utility of IS.

Recently, the diversity of organisms that constitute the lowerrespiratory microbiota, including the identification of speciesfound commonly in the upper respiratory tract, has beendescribed [15]. Thus, in the present study, it is possible thatthese organisms are present in the lower respiratory secretionsand may not necessarily be indicative of contamination fromthe upper airways. Importantly, in one case, H. influenza wasidentified by BAL alone. This suggests that in children withsignificant clinical disease where IS does not identify abacteria, BAL should still be considered a diagnostic option,especially if there is no improvement with empiric antibiotictherapy.

The IS procedure can be distressful. Although most childrentolerated it fairly well (Table 1) and did not cry or show upset, onechild was severely distressed, crying and resisting the procedure.The main unpleasant element during IS is the actual suctionprocedure, which lasts only 1–2 min. This is similar to bloodtests and other common procedures, which are often distressing,and resisted physically by very young children. However, thisbrief distress should be balanced by the importance of thebacteriologic diagnosis enabling appropriate antibiotic therapy.Reliable detection of bacterial pathogens depends upon thefrequency of cultures obtained [16]. In our center (SCMCI), IS innon-expectorating infants with CF is now routinely performed,about 4–6 times per year, enabling early detection, treatment andascertainment of eradication of infections, which would not bepossible with BAL. Recently, we have developed interventionswhich make the IS much less stressful, by providing calmsurroundings and distraction for the child.

The validity and safety of IS for bacteriologic diagnosis ininfants have been demonstrated in other populations. Two largestudies in infants showed IS testing to be safe and useful fordiagnosing Pneumocystis jirovecii [17], and demonstrated ahigher yield for tuberculosis than gastric lavage [18]. Severalstudies have shown that even a 7% hypertonic saline inhalationsolution can be used safely as a therapeutic agent in infants[19,20].

The recent evidence for the role of S aureus in thedevelopment of lung disease in CF [6–8], means that improved

surveillance and treatment of infection with this organism isindicated. Better targeting of therapy is required in order toimprove the benefit: risk ratio of such treatment. The resultsfrom our pilot study suggest that the technique of IS may havesuch a role, especially in those centers where BAL is not feasible.Further studies are required to establish the role of IS to detectlower respiratory infection with S aureus, P. aeruginosa andother organisms in infants with CF and the reliability of IS culturefor bacteria and assay of inflammatory markers, both for clinicalpractice decisions and as surrogate outcome measures forinterventional studies in young children with CF.

References

[1] Konstan MW, Hilliard KA, Norvell TM, Berger M. Bronchoalveolarlavage findings in cystic fibrosis patients with stable, clinically mild lungdisease suggest ongoing infection and inflammation. Am J Respir CritCare Med 1994;150:448–54.

[2] Sly PD, Brennan S, Gangell C, de Klerk N, Murray CM, Mott L, et al.Lung disease at diagnosis in infants with cystic fibrosis detected bynewborn screening. Am J Respir Crit Care Med 2009;180:146–52.

[3] Stafler P, Davies JC, Balfour-Lynn IM, Rosenthal M, Bush A.Bronchoscopy in cystic fibrosis infants diagnosed by newborn screening.Pediatr Pulmonol 2011;46:696–700.

[4] Wainwright CE, Vidmar S, Armstrong DS, Byrnes CA, Carlin JB, Cheney J,et al. Effect of bronchoalveolar lavage-directed therapy on Pseudomonasaeruginosa infection and structural lung injury in children with cysticfibrosis: a randomized trial. JAMA 2011;306:163–71.

[5] Douglas TA, Brennan S, Gard S, Berry L, Gangell C, Stick SM, et al.Acquisition and eradication of P. aeruginosa in young children with cysticfibrosis. Eur Respir J 2009;33:305–11.

[6] Ranganathan SC, Parsons F, Gangell C, Brennan S, Stick SM, Sly PD.Evolution of pulmonary inflammation and nutritional status in infants andyoung children with cystic fibrosis. Thorax 2011;66:408–13.

[7] Pillarisetti N, Williamson E, Linnane B, Skoric B, Robertson CF,Robinson P, et al. Infection, inflammation and lung function decline ininfants with cystic fibrosis. Am J Respir Crit Care Med 2011;184:75–81.

[8] Stick SM, Brennan S, Murray C, Douglas T, von Ungern-Sternberg BS,Garratt LW, et al. Bronchiectasis in infants and preschool childrendiagnosed with cystic fibrosis after newborn screening. J Pediatr2009;155:623–8.

[9] Sagel SD, Sontag MK, Wagener JS, Kapser RK, Osberg I, Accurso FJ.Induced sputum inflammatory measures correlate with lung function inchildren with cystic fibrosis. J Pediatr 2002;141:811–7.

[10] Mussaffi H, Fireman EM, Mei-Zahav M, Prais D, Blau H. Induced sputumin the very young: a new key to infection and inflammation. Chest2008;133:176–82.

[11] Fahy JV, Wong H, Liu J, Boushey HA. Comparison of samples collectedby sputum induction and bronchoscopy from asthmatic and healthysubjects. Am J Respir Crit Care Med 1995;152:53–8.

[12] Pye A, Stockley RA, Hill SL. Simple method for quantifying viablebacterial numbers in sputum. J Clin Pathol 1995;48:719–24.

[13] Ordonez CL, Henig NR, Mayer-Hamblett N, Accurso FJ, Burns JL,Chmiel JF, et al. Inflammatory and microbiologic markers in inducedsputum after intravenous antibiotics in cystic fibrosis. Am J Respir CritCare Med 2003;168:1471–5.

[14] Ho SA, Ball R, Morrison LJ, Brownlee KG, Conway SP. Clinical valueof obtaining sputum and cough swab samples following inhaledhypertonic saline in children with cystic fibrosis. Pediatr Pulmonol2004;38:82–7.

[15] Charlson ES, Diamond JM, Bittinger K, Fitzgerald AS, Yadav A, Haas AR,et al. Topographical continuity of bacterial populations in the healthy humanrespiratory tract. Am J Respir Crit Care Med 2011;184:957–63.

[16] Ratjen F. Diagnosing and managing infection in CF. Paediatr Respir Rev2006;7S:S151–3.

110 H. Blau et al. / Journal of Cystic Fibrosis 13 (2014) 106–110

[17] Zar HJ, Tannenbaum E, Hanslo D, Hussey G. Sputum induction as adiagnostic tool for community-acquired pneumonia in infants and youngchildren from a high HIV prevalence area. Pediatr Pulmonol 2003;36:58–62.

[18] Zar HJ, Hanslo D, Apolles P, Swingler G, Hussey G. Induced sputumversus gastric lavage for microbiological confirmation of pulmonarytuberculosis in infants and young children: a prospective study. Lancet2005;365(9454):130–4.

[19] Subbarao P, Balkovec S, Solomon M, Ratjen F. Pilot study of safety andtolerability of inhaled hypertonic saline in infants with cystic fibrosis.Pediatr Pulmonol 2007;42:471–6.

[20] Rosenfeld M, Davis S, Brumback L, Daniel S, Rowbotham R, Johnson R,et al. Inhaled hypertonic saline in infants and toddlers with cystic fibrosis:short-term tolerability, adherence, and safety. Pediatr Pulmonol 2011;46:666–71.