Tracheobronchial Malacia and Stenosis in Children in Intensive Care: Bronchograms Help to Predict Thorax: First Published As 10.1136/Thx.54.6.511 on 1 June 1999

Thorax 1999;54:511–517 511 Tracheobronchial malacia and stenosis in children in intensive care: bronchograms help to predict Thorax: first published as 10.1136/thx.54.6.511 on 1 June 1999. Downloaded from oucome

Robert J Burden, Frank Shann, Warwick Butt, Michael Ditchﬁeld

Abstract chial tree1–3 and collapse of the airways that is Background—Severe tracheobronchial most apparent in expiration. The condition is malacia and stenosis are important causes usually self-limiting, but severe disease can of morbidity and mortality in children in result in prolonged treatment in intensive care intensive care, but little is known about and is often fatal. how best to diagnose these conditions or The tracheobronchial tree can be assessed determine their prognosis. with many diVerent modalities including non- Methods—The records of all 62 children in contrast fluoroscopy, tracheobronchography, whom one or both of these conditions had computed tomography, magnetic resonance been diagnosed by contrast cinetracheo- imaging, bronchoscopy, or a combination of all bronchography in our intensive care unit these methods. The diagnosis of tracheobron- in the period 1986–95 were studied. chomalacia, however, requires a dynamic Results—Seventy four per cent of the 62 assessment of the trachea and bronchi through- children had congenital heart disease; out a respiratory cycle with demonstration of none was a preterm baby with airways collapse of the airway in expiration. Broncho- disease associated with prolonged ventila- scopy is of limited value because the airway is tion. Fifteen of the children had airway splinted by the bronchoscope which reduces stenosis without malacia; three died be- the amount by which the airways collapse. In cause of the stenosis and two died from addition, small airways or airways beyond a other causes. Twenty eight of the 47 stenosis cannot be examined. Non-contrast children with malacia died; only eight fluoroscopy can satisfactorily assess the state of 4 children survived without developmental the trachea but it does not show the bronchi 56 or respiratory handicap. All children adequately. Spiral and ultrafast computed 6 needing ventilation for malacia for longer tomography and magnetic resonance imaging http://thorax.bmj.com/ than 14 consecutive days died if their with rapid acquisition sequences can show the bronchogram showed moderate or severe trachea and main bronchi, but these techniques malacia of either main bronchus (15 do not provide optimal information because the airways move in and out of the plane of cases), or malacia of any severity of both 7 bronchi (three additional cases); all chil- imaging during a respiratory cycle. In addi- dren needing ventilation for malacia for tion, they are time consuming and often longer than 21 consecutive days died if require increased sedation. Tracheobroncho- their bronchogram showed malacia of any graphy can rapidly and accurately assess the severity of the trachea or a main bronchus entire airway dynamically without significant on September 27, 2021 by guest. Protected copyright. morbidity, and it is a simple procedure to per- (three additional cases). These findings 58 were strongly associated with a fatal form in children who are already intubated. outcome (p<0.00005); they were present in Little is known about the natural history of 21 children (all of whom died) and absent tracheobronchomalacia in children in intensive in 26 (of whom seven died, six from care, or the severity and location of the airways non-respiratory causes). They had a posi- disease. We have reviewed our experience with tive predictive value for death of 100%, but tracheobronchial malacia and stenosis in children in our intensive care unit and analysed the the lower limit of the 95% confidence findings on cinetracheobronchography. Paediatric Intensive interval was 83.9% so up to 16% of patients Care Unit, Royal meeting the criteria might survive. Children’s Hospital, Methods Conclusion—In this series the findings on Following local ethical committee approval, we Parkville, Victoria contrast cinetracheobronchography com- 3052, Australia reviewed the records of 62 children admitted to R J Burden bined with the duration of ventilation pro- our intensive care unit in whom the diagnosis F Shann vided a useful guide to the prognosis of of either malacia or stenosis of the airways was W Butt children with tracheobronchomalacia. M Ditchfield made following bronchography in the 10 years The information provided by broncho- from January 1986 to December 1995. Chart scopy was less useful. Correspondence to: information was recorded systematically by a ( 1999; :511–517) Professor F Shann. Thorax 54 clinician (RJB) blinded to the results of bronchographic review. These data included Received 2 June 1998 Keywords: bronchography; tracheobronchomalacia; Returned to author children; tracheal stenosis details on associated diagnoses, intubation and 27 July 1998 ventilation characteristics, all surgical proce- Revised manuscript received dures including attempts to correct airway 22 December 1998 Accepted for publication Tracheobronchomalacia is an uncommon con- defects, reasons for non-survival, and broncho- 1 February 1999 dition that causes weakness of the tracheobron- scopic findings. 512 Burden, Shann, Butt, et al

A telephone survey was conducted (by RJB Statistical analysis was performed using and WB) to obtain information on the chronic Fisher’s exact test, the Mann-Whitney U test, health outcome of surviving children. Parents and exact confidence intervals for proportions Thorax: first published as 10.1136/thx.54.6.511 on 1 June 1999. Downloaded from were questioned about any current symptoms (StatXact, Cytel, Cambridge, Massachusetts, in their children, including the presence of USA). wheeze, cough, stridor, apnoeic or cyanotic episodes, and home oxygen dependence, as well as about the child’s exercise tolerance, Results general development, medications, school ab- TRACHEOBRONCHOMALACIA senteeism due to respiratory illness, and the Forty seven children (30 boys) with tracheo- number of lower respiratory tract infections bronchomalacia were treated in our intensive occurring each year. The outcome of the survi- care unit during the 10 year period studied. vors was classified as follows: (1) normal; (2) The diagnosis was made before one year of age functionally normal (both intellectually and in 45 of the children (96%) and 27 (57%) physically) but requiring medication or medi- weighed less than the third centile at the time cal supervision—for example, a child with of diagnosis. Thirteen children (28%) were chronic asthma on regular medication but able born prematurely, of whom 10 had associated to exercise without restriction and missing less complex congenital heart disease. Three than two weeks of school a year; (3) mild children required tracheostomy, of whom one handicap but likely to lead an independent has died and another still has the tracheostomy existence—for example, chronic asthma caus- in place. ing restriction of activity, hospital admissions, Thirty four children (72%) had at least one or loss of more than two weeks of school a year; cardiac lesion and 31 (66%) required cardiac (4) moderate handicap dependent on care—for surgery. The cardiac lesions present were left to example, Down’s syndrome; (5) or severe right shunts (ventricular septal defects in 20 handicap (vegetative, totally dependent on children, atrial septal defects in seven, truncus care).9 arteriosus in six, patent ductus arteriosus in Bronchography was performed at our hospi- five, other in one), obstructive right heart tal in a standardised manner. The patients were lesions (pulmonary atresia in six, absent lightly anaesthetised, intubated, and were pulmonary valve in four, tetralogy of Fallot in breathing spontaneously. The endotracheal four, others in three), obstructive left heart tube was maintained in a high position with the lesions (coarctation of the aorta in five, tip in the subglottic region. Opacification of the hypoplastic aortic arch in three), and other airways was obtained by bolus injections of cardiac lesions (right aortic arch in six, dextro- small volumes (1–2 ml) of contrast—initially cardia in four, double outlet right ventricle in

iodised oil (Lipiodol) and more recently three, and other in nine). Forty two children http://thorax.bmj.com/ iohexol (Omnipaque) 300 mg/ml. The contrast (89%) had non-cardiac anomalies: dysmorphic was injected through a small feeding tube syndromes (Di George syndrome in five, cleft passed through a connector in the airway lip in three, other in 21), genitourinary (single circuit close to the endotracheal tube. Contrast kidney in three, other in nine), gastrointestinal dispersal was obtained by hand ventilating the (reflux in 12, exomphalos in three, oesophageal patient; this enabled opacification of the atresia in three, other in four), respiratory (dia- trachea and first, second, and third generation phragmatic palsy in five, vocal cord palsy in bronchi. All patients had their tracheobroncho- three, tracheo-oesophageal fistula in three,

grams recorded on video and with radiographic other in ﬁve), and neurological (epilepsy in on September 27, 2021 by guest. Protected copyright. spot ﬁlms. A pressure monitor was connected four, other in three). to the airway circuit via a Y connector; the Twenty eight of the 47 children (60%) died; study was performed at atmospheric pressure the median time from diagnosis to death was

and at positive pressures of up to 20 cm H2O. two months (range 1–9 months) and 23 (82%) The bronchograms were retrospectively re- of the 28 children died before 12 months of viewed by a radiologist (MD) blinded to the age. Eighty percent of the children weighed less clinical data. Up to four aVected sites were than the third centile when they died. The identiﬁed (trachea, left main bronchus, right commonest cause of mortality was respiratory main bronchus, and peripheral bronchi). Air- failure due to airway disease (17/28). Twenty way malacia was diagnosed when narrowing four of the 28 children died in hospital without was present on expiration (mild = diameter having returned home since diagnosis. 30–50% of normal, moderate = 10–30%, and There were 19 survivors (40%). All the par- severe = 0–10%) and the diameter increased ents of the surviving children were contacted

either on inspiration or with 20 cm H2Oof and participated in the telephone survey. The positive end expiratory pressure. A record was median survival time from diagnosis was 58 kept of the end expiratory pressure required to months (range 9–116 months) and the median restore airway patency in airways aVected by age at follow up was 64 months (range 12–162 malacia. Airways were classiﬁed as stenosed months). Thirty six percent of the survivors when the narrowing was ﬁxed and unaVected were above the 50th centile for weight. Eight by positive airway pressure (mild = diameter children required subsequent admission to 60–100% of normal, moderate = 30–60%, and hospital due to respiratory illness, and a further severe = 0–30%). Lesions were termed “dif- eight and 10 children required admission to fuse” when they involved the entire length of hospital for cardiac illness or other non- the trachea or bronchus, otherwise they were cardiorespiratory reasons, respectively. Eight termed “focal”. of the 19 survivors (42%) were normal or Tracheobronchial malacia and stenosis in children in intensive care 513

Table 1 Tracheobronchomalacia: diagnosis, presentation, and outcome in survivors (n = 19) ranked by duration of ventilation Thorax: first published as 10.1136/thx.54.6.511 on 1 June 1999. Downloaded from Age Bronchogram severity score (months) Ventilation Bronchoscopic at diagnosis Main diagnosis (or presentation) (days) TLBRBPBabnormality Handicap

1 Pulmonary stenosis 2 1010None Moderate 2 Truncusarteriosis 3 1210 Normal 0 Diaphragmatic hernia 3 3002 Mild 1 Respiratory failure 3 0100 Normal 0 Oesophageal atresia 3 3000 Functionally normal 3 Absent pulmonary valve 4 1000 Mild 0 Coarctation of the aorta 6 1000TM Normal 2 Vascular ring 6 0220 Moderate 11 Coarctation of the aorta 7 1010 Mild 4 Oesophageal atresia 7 2200RBM, LBM Normal 0 Absent pulmonary valve 8 0002 Mild 152 Pulmonary atresia 10 1000 Functionally normal 3 Pulmonary atresia 11 3002 Severe 11 Absent pulmonary valve 12 1330 Moderate 0 (Respiratory failure) 14 2100 Mild 0 Laryngeal cleft 18* 2220None Mild 7 Transposition great arteries 20 1000 Normal 1 Atrioventricular septal defect 21 0100 Severe 8 Pulmonary atresia 40 0002RBM Normal

*Ventilated for aspiration rather than tracheobronchomalacia. T = trachea; LB = left main bronchus; RB = right main bronchus; PB = peripheral bronchi; TM = tracheomalacia; LBM = left main bronchomalacia; RBM = main bronchomalacia; bronchogram severity score 0 = none, 1 = mild, 2 = moderate, 3 = severe (see Methods). functionally normal, six were mildly handi- diVerence between survivors and non- capped, three were moderately handicapped, survivors in the duration of ventilation (p = and two were severely handicapped (table 1). 0.0001, see tables 1 and 2) and endotracheal The most prevalent respiratory symptom intubation (p = 0.0009). The geometric mean was cough, occurring daily in five children. duration of intubation was 0.96 days (95% CI Twelve children (63%) required regular 2.2 to 62.7) in survivors and 27.1 days (95% medication—four for asthma and five for CI 4.0 to 182.7) in non-survivors. cardiac disease. In children of school age the The severity and location of malacia airways mean annual absenteeism due to respiratory in our patients are shown in tables 1 and 2. The disease was eight days (range 0–40 days). trachea was abnormal in 35 of the 47 children Lower respiratory tract infections occurred (74%) and this was an isolated finding in eight

between 0 and 10 times a year (median 2). (17%). Twenty nine (62%) and 24 (51%) had http://thorax.bmj.com/ Three children had cyanotic episodes at home lesions aVecting the left and right main but none required home oxygen and none had bronchi, respectively. Nineteen of the 29 apnoeic episodes. There was a signiﬁcant lesions aVecting the left main bronchus and 19

Table 2 Tracheobronchomalacia: diagnosis, presentation, and cause of death in non-survivors (n = 28) ranked by cause of death (respiratory, part respiratory, non-respiratory) and duration of ventilation

Age Bronchogram severity score (months) Ventilation Bronchoscopic

at diagnosis Main diagnosis (or presentation) (days) TLBRBPBabnormality Cause of death on September 27, 2021 by guest. Protected copyright.

3 Truncusarteriosus 6 3330 Respiratory 11 Tetralogy of Fallot 17 2110TM,LBMRespiratory 8 Vascular ring 19 1020 Respiratory 0 (Respiratory failure) 23 1220TM Respiratory 11 Pulmonary atresia 24 2202 Respiratory 3 Tetralogy of Fallot 25 0132 Respiratory 1 Anomalous pulmonary venous 280230 Respiratory drainage 1 Vascular ring 29 0111None Respiratory 3 Absent pulmonary valve 35 1110 Respiratory 5 Vascular ring 37 2020TM,RBMRespiratory 2 Absent pulmonary valve 56 2020 Respiratory 0 (Respiratory failure) 60 1100 Respiratory 5 Cystic right lung 65 1020 Respiratory 4 Pulmonary atresia 68 2222LBM, PBM Respiratory 0 Pulmonary stenosis 72 3000TM Respiratory 13 Hypoplastic aortic arch 85 2220Hypoplastic LB Respiratory 0 Truncusarteriosus 852330 Respiratory 11 Atrioventricular septal defect 6 0202 Cardiorespiratory 0 Pulmonary atresia 15 0330 Cardiorespiratory 0 Interrupted aortic arch 20 1210 Cardiorespiratory 12 (Respiratory failure) 2 1000TM Neurological 3 Vascular ring 10 1100 Neurological 9 Treacher Collins syndrome 10 0200LBM Neurological 7 Transposition of great arteries 11 0200TM,LBM, RBM Multisystem 0 Coarctation aorta 19 2000None Gastrointestinal 2 Pulmonary atresia 19 3330 Gastrointestinal 1 VSD,ASD 402100 Multisystem 0 Hypoplastic aortic arch 116 3333LBM Haemorrhage

T = trachea; LB = left main bronchus; RB = right main bronchus; PB = peripheral bronchi; TM = tracheomalacia; LBM = left main bronchomalacia; RBM = right main bronchomalacia; PBM = peripheral bronchomalacia; bronchogram severity score 0 = none, 1 = mild, 2 = moderate, 3 = severe (see Methods). 514 Burden, Shann, Butt, et al

Table 3 Airway stenosis: diagnosis, presentation, and outcome (n = 15)

Bronchogram Thorax: first published as 10.1136/thx.54.6.511 on 1 June 1999. Downloaded from severity score Age (months) Main diagnosis Ventilation Bronchoscopic at diagnosis (or presentation) (days) TLBRBPBdiagnosis Handicap Cause of death

2 (Stridor) 0 2 0 0 0 TS Normal 8 (Stridor) 0 3 2 0 0 TS Normal 5 Vascular ring 4 1 0 0 0 Severe 2 (Respiratory failure) 5 2 0 0 0 TS Mild 49 Vascular ring 6 1 0 0 0 TS Normal 4 Atrioventricular septal defect 8 3 0 0 0 Moderate 2 Vascular ring 9 3 0 0 0 Normal 2 Vascular ring 10 2 0 0 0 Mild 0 Interrupted aortic arch 23 0 2 2 0 Functionally normal 0 Coarctation aorta 55 1 1 1 0 Functionally normal 15 Patent ductus 1 2 0 0 0 TS Died Respiratory 0 Vascular ring 7 3 0 0 0 TS Died Multiorgan 5 Vascular ring 13 2 0 0 0 TS Died Respiratory 3 Patent ductus 20 1 0 0 0 TS Died Cardiorespiratory 2 Interrupted aortic arch 60 0 3 0 0 Died Respiratory

T = trachea; LB = left main bronchial stenosis; RB = right main bronchial stenosis; PB = peripheral bronchi; TS = tracheal stenosis; bronchogram severity score: 0 = none, 1 = mild, 2 = moderate, 3 = severe (see Methods).

of the 24 lesions aVecting the right main bron- For the admission during which tracheo- chus occurred in association with tracheomala- bronchomalacia was diagnosed the median cia. Ten children (21%) had more distal duration of ventilation was 7 (range 2–40) days lesions. Thirty nine percent of tracheal lesions for survivors and 19 (2–116) days for non- were focal compared with 90% of lesions in the survivors (p = 0.0001). The median time in main and peripheral bronchi. Thirteen chil- intensive care was 22 (range 3–108) days for dren had evidence of airway stenosis in survivors and 31 (5–144) days for non- addition to malacia; these aVected the same survivors. The median time in hospital was 80 airway in eight children and separate airways in (range 14–270) days for survivors and 62 ﬁve. (2–250) days for non-survivors. All children who needed ventilation for malacia for longer than 14 days died if they had STENOSIS moderate or severe malacia of either main Fifteen children (eight boys) in intensive care

bronchus (15 cases), or malacia of any severity were diagnosed with airway stenosis by tra- http://thorax.bmj.com/ of both main bronchi (three additional cases); cheobronchography (table 3). In 11 cases the all children who needed ventilation for malacia diagnosis was made before the age of six for longer than 21 days died if they had mala- months, with eight children weighing less than cia of any severity of the trachea or a main the third centile. Twelve children had associ- bronchus (three additional cases). None of the ated cardiac lesions (ﬁve had vascular rings) children who survived had these ﬁndings. and 14 children had non-cardiac anomalies. These criteria were present in 21 children (all Five children were born prematurely. Five chil- of whom died) and absent in 26 (of whom dren died; the median time from diagnosis to

seven died), giving a sensitivity of 75.0% (exact death was 8.3 months (range 1.7–23) and the on September 27, 2021 by guest. Protected copyright. 95% CI 55.1 to 89.3), specificity of 100% median age at death was 12.5 months. The (95% CI 82.3 to 100), positive predictive value parents of all 10 survivors contributed to the of 100% (95% CI 83.9 to 100), negative telephone survey; six of the 10 children were predictive value of 73.1% (95% CI 52.2 to normal or functionally normal. The common- 88.4), andapvalue of <0.00005 (Fisher’s est symptom was cough which occurred on exact test). If the 11 children who died from most days in seven children. Stridor was non-respiratory causes are excluded (the crite- present every day in only one survivor. ria were present in five), the sensitivity is 94.1% Tracheobronchography demonstrated that (exact 95% CI 71.3 to 99.9), specificity 100% tracheal stenosis occurred in 13 patients and (82.3 to 100), positive predictive value 100% that it was an isolated finding in 11 cases. In (79.4 to 100), and negative predictive value four cases stenosis occurred in the main bron- 95.0% (75.1 to 99.9). chi. There were no stenoses seen more periph- The amount of positive end expiratory erally. Thirty three percent of lesions in the tra- pressure required at bronchography to com- chea and all lesions seen aVecting the main pletely expand aVected airways ranged between bronchi were focal. There was no relationship

10 and 20 cm H2O; there was no relationship between the number, site, or severity of between the amount of positive end expiratory stenotic lesions seen on the bronchogram and pressure required and either survival or the survival. duration of continuous positive airways pressure, intubation, or ventilation. No child had a SURGERY severe exacerbation of respiratory failure pre- Thirteen of the 62 patients underwent at- cipitated by the bronchogram, perhaps because tempts at deﬁnitive surgical repair of a major great care was taken to restrict the amount of airway (table 4), usually because of failure to contrast medium used, particularly when wean them from mechanical ventilation. iodised oil was being instilled. Eleven operations were performed on the Tracheobronchial malacia and stenosis in children in intensive care 515

Table 4 Main lesion and outcome in 13 children undergoing surgery

Main lesion Focal/diVuse Surgery performed Outcome Thorax: first published as 10.1136/thx.54.6.511 on 1 June 1999. Downloaded from Main bronchial stenosis Focal Resection and end to end anastomosis Alive Peripheral bronchomalacia Focal Bronchoplasty Alive Tracheal stenosis and malacia Focal Tracheopexy Alive Tracheomalacia Focal Tracheopexy Alive Tracheal stenosis Focal Resection and end to end anastomosis Alive Tracheal stenosis Focal Resection and end to end anastomosis Dead Tracheomalacia DiVuse Aortopexy Dead Tracheomalacia DiVuse Pericardial patch repair Dead Tracheobronchomalacia DiVuse Aortopexy Dead Tracheal stenosis DiVuse Tracheoplasty Dead Tracheal stenosis DiVuse Pericardial patch repair Dead Tracheal stenosis and malacia DiVuse Arteriopexy Alive Tracheal stenosis DiVuse Pericardial patch repair Alive

trachea and two on the main bronchi. Six had cheal suctioning, postural drainage, and posi- focal lesions demonstrated by tracheobronch- tive end expiratory pressure through an ography and five survived after surgery to the endotracheal tube or tracheostomy. Bron- aVected airway. Seven had diVuse disease and chodilators and steroids may improve periph- only two survived. eral obstructive airways disease and hence reduce the dynamic compression of the large 3 BRONCHOSCOPY airways. Surgery to the airway is usually Twenty five of the 62 patients were investigated reserved for patients in whom medical treat- by bronchoscopy as well as bronchography ment has failed and is generally performed near the time of the original diagnosis of either after underlying surgical causes such as con- airway stenosis or malacia (tables 1–3). There genital heart disease have been corrected. Pro- was agreement about the diagnosis in 12 of the cedures include anterior tracheopexy, tracheal 25 cases and in 10 of the 12 the abnormalities stenting, airway implants, and resection of dis- were isolated to the trachea. There was ease segments; they are reported to be more disagreement about the diagnosis in 13 of the successful when disease is focal rather than 25 cases and in 12 of these the discrepancy diVuse,3101213although there are no controlled involved lesions beyond the trachea; six of these trials of these procedures. cases had airway abnormalities in the major We reviewed all children with tracheobron- bronchi or periphery that were detected by chomalacia who had received treatment in our bronchography but not by bronchoscopy. A intensive care unit over a 10 year period. Con- fibreoptic 1.8 mm bronchoscope was available trast cinetracheobronchography was found to and was usually used in small infants. be highly predictive of death or survival. All http://thorax.bmj.com/ children who needed ventilation for malacia Discussion for longer than 14 days died if they had mod- Tracheobronchomalacia was first described in erate or severe malacia of either main 1952 in three infants1 and, since then, many bronchus, or malacia of any severity of both classifications have been proposed. Primary main bronchi (table 2); all children who malacia results from immature tracheobron- needed ventilation for malacia for longer than chial cartilages occurring in isolation or in 21 days died if they had malacia of any sever-

association with congenital conditions such as ity of the trachea or a main bronchus. None of on September 27, 2021 by guest. Protected copyright. heart disease, oesophageal atresia, and tracheo- the children who survived had these findings, oesophageal fistula. Secondary malacia results although one child with moderate malacia of from degeneration of previously normal carti- both main bronchi needed ventilation for 18 lage and is associated with extrinsic vascular days because he had a very large laryngeal compression, bronchial neoplasms, recurrent cleft. The findings were present in 21 children, bronchitis, gastro-oesophageal reflux, long all of whom died, and absent in 26 of whom term intubation, and tracheostomy.23Tracheo- seven died, six from non-respiratory causes. bronchomalacia has been classified according Although these criteria had 100% specificity to the histopathological and bronchoscopic and positive predictive value for death in our changes seen in the condition10 and also series (p<0.00005), the lower limit of the 95% according to the degree of luminal narrowing confidence interval was 82.3% for the specifi- seen during coughing.11 city and 83.9% for the positive predictive The severity of the symptoms depends on value, so that up to 16% of patients fulfilling the location, length, and severity of the abnor- these criteria might survive. The clinicians mal airway segments.10 In children, symptoms were aware of the bronchography results at the include stridor, wheeze, cough, hyperextension time that these patients died, but this is very of the neck, recurrent respiratory tract infec- unlikely to have led to a decision to discon- tions, cyanotic spells (especially when crying or tinue mechanical ventilation because the passing stool), and reflex apnoea. In adults, prognostic implications of bronchograms were chronic cough, wheeze, and sputum produc- not known then. Indeed, the main reason for tion predominate. The symptoms associated doing this study was to determine whether the with mild cases of tracheobronchomalacia gen- bronchogram provides information about the erally resolve spontaneously.11 Some patients prognosis of bronchomalacia. with severe disease respond favourably to con- We found that children with severe tracheo- servative measures including oxygen, endotra- bronchomalacia who needed prolonged venti- 516 Burden, Shann, Butt, et al

lation for malacia had a very high mortality findings on contrast cinebronchography have rate, and malacia of the main bronchi carried a diVerent implications in these patients because worse prognosis than tracheomalacia. Al- chronic respiratory failure may be due to Thorax: first published as 10.1136/thx.54.6.511 on 1 June 1999. Downloaded from though these findings provide a very useful bronchopulmonary dysplasia as well as bron- guide to the prognosis, they should not be used chomalacia. in isolation as an indication to withdraw treat- Tracheal and bronchial stenosis are also rare ment; this was a retrospective study, there were conditions.18 We reviewed 15 patients in whom only 28 deaths (six from non-respiratory cinetracheobronchography had shown airway causes), and the clinicians were aware of the stenosis without malacia while they were bronchogram results. patients in our intensive care unit. Thirteen We found that bronchoscopy was much less children had tracheal lesions and two of these sensitive than bronchography in detecting had additional bronchial lesions. The remain- lesions aVecting the main bronchi, and the ing two children had stenosis confined to the presence of malacia of the main bronchi was major bronchi. This suggests that stenotic the main factor that determined whether a lesions are commonly confined to the trachea, child died or survived. Our findings suggest in contrast to malacia where lesions are often that contrast cinetracheobronchography is widespread in the tracheobronchial tree. There more useful than bronchography for the was a high incidence of associated anomalies, diagnosis of tracheobronchomalacia, and this particularly vascular rings and prematurity. has been the experience of other Malacia and stenosis coexisted in 13 patients, investigators.81415When bronchography is per- indicating that neither represents a separate formed the endotracheal tube must be high in clinical entity in critically ill infants and the trachea and it is very important that the children. patient has active spontaneous respiration Thirteen children with malacia or stenosis of (bronchomalacia may be missed if the patient is the airways underwent attempts at surgical heavily sedated). The complication rate is very repair, mainly after failure to wean from low if very small amounts of a non-ionic mechanical ventilation. Surgical repair of focal contrast medium are used. lesions had a better outcome than repair of dif- Tracheobronchomalacia is a rare and seri- fuse lesions. Contrast cinebronchography pro- ous condition associated with a high mortality vides a record of the length of aVected in children in intensive care, and it is an segments and assists in the planning for surgi- important cause of prolonged intubation and cal repair. ventilation. The median duration of intubation Tracheobronchomalacia is rare, but children in our population was 12 days for survivors with this condition who are in intensive care and 30 days for non-survivors, and for ventila- often need prolonged respiratory support and

tion was seven days for survivors and 19 days have a high mortality rate. In contrast to bron- http://thorax.bmj.com/ for non-survivors. The median duration of stay choscopy, cinetracheobronchography provides in our intensive care unit for all patients is one detailed information about the lesions in day (reﬂecting the short length of stay in pae- tracheobronchial malacia as well as stenosis, diatric intensive care in Victoria16). Death from and it provides information that helps predict malacia is commonly the result of respiratory which children are likely to survive. Because failure and an inability to wean from ventila- many centres do not perform tracheobroncho- tion; children who die usually succumb within grams in high risk patients and other investiga- a few months of the original diagnosis of tions have a much lower sensitivity, tracheo-

malacia. bronchomalacia is probably much more on September 27, 2021 by guest. Protected copyright. Less than half the surviving children in our common in ventilator dependent children than study were normal or functionally normal at is generally realised. follow up, with only eight of the 19 survivors having both normal development and exercise We acknowledge the help of Sue Stubbings in Medical Records tolerance. However, cyanotic and apnoeic epi- at the Royal Children’s Hospital, and Bernard Higgins at the Department of Mathematics, Portsmouth University. There was sodes, which are life-threatening accompani- no ﬁnancial support for this work. ments of malacia, rarely occurred, and the overall respiratory health of the survivors was 1 Hollinger PH, Johnston KC, Parchet VN, et al. Congenital generally good (table 1). malformations of the trachea, bronchi, and lung. Ann Otol Rhinol Laryngol 1952;61:1104–59. Thirteen of the 47 patients with malacia 2 Johnson TH, Mikita JJ, Wilson RJ, et al. Acquired were born prematurely and eight of them died. tracheomalacia. Radiology 1973;109:577–80. 3 Sotomayor JL, Godinez RI, Borden S, et al. Large airway None had developed bronchopulmonary dys- collapse due to acquired tracheobronchomalacia in in- plasia; intermittent positive pressure ventila- fancy. Am J Dis Child 1986;140:367–71. 4 Joseph PM, Berdon WE, Baber DH, et al. Upper airway tion was indicated for reasons other than hya- obstruction in infants and small children. Radiology line membrane disease and the diagnosis of 1976;121:143–8. 5 Riebello T, Wartner R. Use of non-ionic contrast media for malacia was made early in their stay in tracheobronchography in neonates and young infants. Eur intensive care (after a median of seven days). J Radiol 1990;11:120–4. 6 Shepard JO, Grillo HC, Bhalla M, et al. Inspiratory- Prolonged ventilation of premature infants expiratory chest CT in evaluation of large airway disease. with hyaline membrane disease is associated Presented at the Radiological Society North America Meeting, Chicago, Illinois. Radiology 1994;193(P):181. with bronchopulmonary dysplasia and a ten- 7 Brody S, Kuhn JP, Siedel FG, et al. Airway evaluation in dency towards increased airway collapsibility children with use of ultrafast CT: pitfalls and recommen- (acquired tracheobronchomalacia).3 These dations. Radiology 1993;178:181–4. 8 MacIntyre P, Peacock C, Gordon I, et al.Useof patients may survive after many months of tracheobronchography as a diagnostic tool in ventilator- dependent infants. Crit Care Med 1998;26:755–9. intermittent positive pressure ventilation or 9 Butt W, Shann F, Tibballs J, . Long-term outcome of 17 et al continuous positive airways pressure. The children after intensive care. Crit Care Med 1990;18:961–5. Tracheobronchial malacia and stenosis in children in intensive care 517

10 Benjamin B. Tracheomalacia in infants and children. Ann 15 Bramson RT, Sherman JM, Blickman JG. Paediatric Otol Rhinol Laryngol 1984;93:438–42. bronchography performed through the ﬂexible broncho- 11 Cogbill TH, Moore FA, Accurso FJ, et al.Primary scope. Eur J Radiol 1993;16:158–61. tracheomalacia. Ann Thorac Surg 1983;35:538–41. 16 Pearson G, Shann F, Barry P, et al. Should paediatric inten- Thorax: first published as 10.1136/thx.54.6.511 on 1 June 1999. Downloaded from 12 Cohen D. Tracheopexy: aorto-tracheal suspension for sive care be centralised? Trent versus Victoria. Lancet severe tracheomalacia. Aust Paediatr J 1981;17:117–21. 1997;349:1213–7. 13 Mair EA, Parsons DS, Lally KP. Treatment of severe bron- 17 Wiseman NE, Duncan PG, Cameron CB. Management of chomalacia with expanding endobronchial stents. Arch tracheobronchomalacia with continuous positive airway Otolaryngol Head Neck Surg 1990;116:1087–90. pressure. J Pediatr Surg 1985;20:489–93. 14 Levy M, Glick B, Springer C, et al. Bronchoscopy in 18 HoVer ME, Tom LW, Wetmore RF, et al. Congenital children: experience with 110 investigations. Am J Dis Child tracheal stenosis. Arch Otolaryngol Head Neck Surg 1983;137:14–6. 1994;120:449–53. http://thorax.bmj.com/ on September 27, 2021 by guest. Protected copyright.