Abstract
Breathing pattern, gas exchange, and respiratory effort were assessed in five awake children with chronic stridor caused by laryngomalacia during spontaneous breathing (SB) and noninvasive mechanical ventilation (NIMV). During SB, the youngest children were able to maintain normal gas exchange at the expense of an increased work of breathing as assessed by calculated diaphragmatic pressure-time product (PTPdi), whereas the opposite was observed in the older children. NIMV increased tidal volume, from 8.77 ± 2.04 ml/kg during SB to 11.67 ± 2.52 ml/kg during NIMV, p = 0.04, and decreased respiratory rate, from 24.4 ± 5.6 breaths/ min during SB to 16.6 ± 0.9 breaths/min during NIMV, p = 0.04. NIMV unloaded the respiratory muscles as reflected by the significant reduction in PTPdi, from a mean value of 541.0 ± 196.6 cm H2O · s · min− 1 during SB to 214.8 ± 116.0 cm H2O · s · min− 1 during NIMV, p = 0.04. Therefore, NIMV successfully relieves the additional load imposed on the respiratory muscles. Long-term home NIMV was provided to a total of 12 children with laryngomalacia (including these five) and was associated with clinical improvement in sleep and growth.
Keywords: stridor; laryngomalacia; diaphragmatic pressure-time product; noninvasive mechanical ventilation
Laryngomalacia is an anomaly of the larynx with laxity of the pharyngeal tissues causing the epiglottis, arythenoids, and aryepiglottic folds to involute and partially obstruct breathing during inspiration. Laryngomalacia accounts for more than 75% of cases of congenital stridor and is the most common cause of symptomatic partial upper airway obstruction in infants (1, 2). The majority of infants with laryngomalacia do well with resolution of stridor during the first 2 yr of life (3). However, potentially serious complications, including airway obstruction and sudden death (4, 5), pulmonary hypertension and cor pulmonale (5, 6), failure to thrive (5, 7), and possibly intellectual impairment can develop (8). In some severe cases of laryngomalacia, even surgery (for example, endoscopic resection of the aryepiglottic folds or epiglottoplasty) can fail to relieve upper airway obstruction (5, 7). When respiratory problems persist after surgery, a tracheostomy may be required. However, tracheostomy is associated with a significant morbidity and impairs normal development and, particularly, language development (9, 10).
Noninvasive mechanical ventilation (NIMV) has been shown to reduce the work of breathing, and in particular pressure support (PS), associated with positive end-expiratory pressure (PEEP), has been recognized as an efficient treatment of upper airway obstruction associated with alveolar hypoventilation (11). Our hypothesis is that NIMV may be proposed as an effective treatment for laryngomalacia, which might be an alternative to a tracheostomy.
To confirm our hypothesis, the aim of the study was, first, to analyze the breathing pattern and the respiratory effort in children with stridor caused by severe laryngomalacia; second, to evaluate the immediate physiologic value of NIMV in unloading the respiratory muscles in these children; and third, to investigate the long-term clinical effects of intermittent NIMV.
The study was approved by our institutional board, and written informed consent was obtained from all parents. Criteria for enrollment were severe laryngomalacia with symptoms of upper airway obstruction and nocturnal hypoventilation. (Methods are detailed in the online supplement.)
The first part of the study evaluated the gas exchange, breathing pattern, and respiratory effort in five patients (Group 1) during spontaneous breathing (SB) and NIMV. The second part of the study evaluated the clinical long-term follow up of these five patients and seven additional patients (Group 2).
Two conditions were analyzed and compared; SB and NIMV (PS ventilation) delivered through a well-fitting nasal mask. Custom-made masks were used in infants (dead space < 5 ml) and commercial masks in older children. For NIMV, the initial PS level was set at 6 cm H2O and the initial positive PEEP at its lowest level, i.e., 2 cm H2O. PS and PEEP were titrated alternatively and progressively by 1 to 2 cm H2O steps, as has been used successfully by other groups (12, 13). The final pressure settings were the highest values tolerated by the patients. Therefore, target swings for esophageal pressure (Pes) and transdiaphragmatic pressure (Pdi) were the lowest value that we could obtain according to the tolerance of the patients. Other parameters were set in line with consensus guidelines (14).
Data were recorded during the last 5 min after a stable period for at least 20 min. Nasal mask was not tolerated during the SB period by the two youngest infants who desaturated immediately. For this reason, recording of flow and determination of tidal volume (Vt) and inspiratory time/duty cycle ratio (Ti/Ttot) during SB were only made in the three oldest patients.
We measured respiratory flow (this was integrated to yield Vt), airway pressure (Paw), SaO2 , respiratory rate (RR), heart rate, and end-tidal carbon dioxide (Pet CO2 ) directly from the mask. Pes and gastric pressures (Pga) were measured using catheter-mounted transducers (Gaeltec, Dunvegan, Isle of Skye, UK) (15) positioned using standard techniques (16). We measured the Pdi swings and the diaphragmatic pressure-time products (PTPdi) on the whole duty cycle as previously described (17, 18).
Long-term follow up was obtained in these five patients (Group 1) and seven additional patients (Group 2). The setting of the PS and PEEP levels in Group 2 were determined on a clinical basis, i.e., the disappearance of stridor, chest retractions, and snoring, as well as desaturations and hypercapnia during sleep (19, 20). Compliance to the treatment was systematically assessed at home at a monthly basis by means of a data logger that recorded date and time as well as the duration of the NIMV use. The effects of NIMV on nocturnal SaO2 and growth was assessed in all patients as well as the tolerance and duration of NIMV.
Data are given as mean ± SD. Comparison between SB and NIMV were made using Wilcoxon's rank test. Correlation between the different variables was made by simple regression.
The characteristics of the patients are presented in Table 1. Five infants had chest wall deformity and six had failure to thrive requiring nutritional support (Patients 1, 2, and 6 to 9). All the patients were naive to NIMV. The mean age of the patients was 32.9 ± 25.8 mo, with three patients being younger than 1 yr of age. The four youngest patients (Patients 1, 6, 7, and 8) had the most severe upper airway obstruction. Levels of PS ranged from 4 to 8 cm H2O and levels of PEEP from 4 to 10 cm H2O (Table 1).
| Patient No. | Sex | Associated Diagnosis | At Start of NIMV | Level of PS/PEEP (cm H 2 O) | Duration of NIMV (mo) | Outcome | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age (mo) | Height (cm) | Weight (kg) | ||||||||||||||
| 1 | M | None | 8 | 63 | 5.3 | 4 / 6 | 24 | Still on NIMV | ||||||||
| 2 | F | Lymphangioma | 25 | 87 | 12 | 4 / 6 | 16 | Still on NIMV | ||||||||
| 3 | F | Mental retardation | 64 | 135 | 30 | 10 / 6 | 12 | Died AH | ||||||||
| 4 | M | Recklinghausen | 77 | 113 | 27 | 6 / 8 | 6 | On LTOT | ||||||||
| 5 | M | Dysautonomia | 79 | 115 | 39 | 8 / 8 | 23 | Still on NIMV | ||||||||
| 6 | M | Prematurity | 10 | 68 | 7.1 | 6 / 6 | 30 | Well 1 yr later | ||||||||
| 7 | M | Lymphangioma | 11 | 70 | 8.0 | 4 / 6 | 14 | Well 18 mo later | ||||||||
| 8 | M | Mental retardation | 16 | 70 | 7.9 | 6 / 4 | 6 | Still on NIMV | ||||||||
| 9 | M | Laryngeal cleft | 19 | 77 | 9.2 | 4 / 6 | 6 | Well 2 yr later | ||||||||
| 10 | M | Picnodysostosis | 25 | 79 | 9.0 | 4 / 8 | 52 | Still on NIMV | ||||||||
| 11 | M | CHARGE syndrome | 25 | 85 | 11.8 | 4 / 6 | 22 | Well 30 mo later | ||||||||
| 12 | M | None | 36 | 85 | 11.5 | 8 / 7 | 2 | Still on NIMV | ||||||||
Gas exchange was severely impaired in Patient 5 and within normal range in the other patients (Figure 1). SaO2 increased from 94.9 ± 2.7% during SB to 96.9 ± 2.0% during NIMV, p = 0.07, and Pet CO2 decreased from 41.0 ± 6.5 mm Hg during SB to 34.8 ± 6.2 mm Hg during NIMV, p = 0.04 (Figure 1).
Fig. 1. Pulse oximetry (SaO2 ) and end-tidal carbon dioxide tension (Pet CO2 ) during spontaneous breathing (SB) and noninvasive mechanical ventilation (NIMV) in the five patients of Group 1. Patients 1 and 2 are represented by open circles and the three older patients by closed circles.
[More] [Minimize]Respiratory rate decreased from 24.4 ± 5.6 breaths/min during SB to 16.6 ± 0.9 breaths/min during NIMV, p = 0.04 (Figure 2). In the three oldest patients, mean Vt increased from 8.77 ± 2.04 ml/kg during SB to 11.67 ± 2.52 ml/kg during NIMV (Figure 2), and mean Ti/Ttot decreased significantly during NIMV, with a value of 0.59 ± 0.25 during SB and 0.35 ± 0.04 during NIMV. There was a negative correlation between age and SaO2 , r2 = 0.526, p < 0.0001, a positive correlation between age and Pet CO2 , r2 = 0.374, p = 0.009 (Figure 3) and a positive correlation between age and Ti/Ttot, r2 = 0.570, p = 0.04.
Fig. 2. Respiratory rate and tidal volume (Vt) during spontaneous breathing (SB) and noninvasive mechanical ventilation (NIMV) in the patients of Group 1. Patients 1 and 2 are represented by open circles and the three older patients by closed circles. Vt could not be measured in the two infants because of nontolerance of the nasal mask without NIMV.
[More] [Minimize]
Fig. 3. Correlation between the age of the patients of Group 1 and pulse oximetry (SaO2 ) and end-tidal carbon dioxide tension (Pet CO2 ).
[More] [Minimize]A tracing from Patient 3 during SB and NIMV is presented in Figure 4, which shows a decrease in Pes and Pdi swings during NIMV. In addition, Figure 5 shows the dynamic relationship between Pes and Pga during a representative cycle of SB and NIMV in Patient 3. It demonstrates, as in all patients of Group 1, the absence of an abnormal increase of Pga during expiration, i.e., during the increase of Pes during SB and/or NIMV. These results permit us to state that no patient expiratory muscle recruitment was present during SB and/or during NIMV.
Fig. 4. A representative tracing of Patient 3 during spontaneous breathing (SB) during noninvasive mechanical ventilation (NIMV) showing the decrease in esophageal pressure (Pes) and transdiaphragmatic pressure (Pdi) swings during NIMV compared with SB. Flow = airflow; Paw = airway pressure; Pgas = gastric pressure swing; I = inspiration; E = expiration.
[More] [Minimize]
Fig. 5. Schematic tracing of esophageal pressure (Pes) and gastric pressure (Pgas) swings of a representative patient (Patient 3) during spontaneous breathing (SB, loop on the left) and noninvasive mechanical ventilation (NIMV: pressure support [PS] = 10 cm H2O and positive end-expiratory pressure = 6 cm H2O, loop on the right) showing the decrease of the inspiratory muscle activity during NIMV compared with that during SB and the absence of expiratory muscle recruitment during the two situations. End-expiratory pressure values have been normalized to zero. The solid bars indicate the beginning and the end of the inspiration.
[More] [Minimize]Pdi swing decreased from 20.7 ± 4.3 cm H2O during SB to 8.9 ± 4.3 cm H2O during NIMV, p = 0.04 (Figure 6). Similar reductions were observed for PTPdi with mean PTPdi decreasing from 541.0 ± 196.6 cm H2O·s·min−1 to 214.8 ± 116.0 cm H2O·s·min−1 during NIMV, p = 0.04 (Figure 6). In addition, we observed an excellent relationship for the five patients between the decrease in Pes and Pdi swings and the disappearance of inspiratory dyspnea characterized by stridor, loud breathing, chest deformity, sweats, and a low SaO2 . There was also a negative correlation between age and indices of respiratory effort, as reflected by ΔPdi, r2 = 0.461, p = 0.03, and PTPdi, r2 = 0.730, p = 0.01 (Figure 7).
Fig. 6. Indices of respiratory effort as assessed by transdiaphragmatic pressure swings (Swing Pdi) and diaphragmatic pressure-time product (PTPdi) during wakefulness in the patients of Group 1 during spontaneous breathing (SB) and during noninvasive mechanical ventilation (NIMV).
[More] [Minimize]
Fig. 7. Correlation between the age of the patients and the indices of work of breathing as assessed by transdiaphragmatic pressure swing (Swing Pdi) and diaphragmatic pressure-time product (PTPdi) in Group 1.
[More] [Minimize]NIMV was associated with important clinical improvements in all the patients.
NIMV was started at the age of 8 mo in Patient 1, when his weight was 5.4 kg (−3.5 SD) and height 63 cm (−3 SD). He slept with his NIMV for a mean of 12 ± 2 h/d. A significant growth catch-up was observed during NIMV, with a weight gain of 3.6 kg during the first 6 mo despite the discontinuation of his nasogastric feeding 3 wk after the start of NIMV. At the age of 14 mo, he weighted 9 kg (−1 SD) and measured 73 cm (−1.5 SD). One year later, a trial of discontinuation of NIMV because of an improvement of his sleep under room air, was associated with a 1 kg weight loss in 3 wk, which was regained after recommencing NIMV. This infant also had an important chest-wall deformity with xyphoid and subcostal recession, which totally regressed 8 mo after the start of NIMV.
The tracheostomy of Patient 2, which was performed at the age of 2 mo, was closed 1 wk before starting NIMV, allowing her and her family to return home. She had been fed by a gastrostomy since birth, and oral nutrition was not possible because of a psychological blockage to eating by mouth. It is therefore of particular interest that despite similar caloric intake, she gained 1 kg in 3 mo. After 1 yr of NIMV, she began to eat by mouth and to attend nursery school without any problem. Her chest-wall deformity has also noticeably regressed.
Nocturnal SaO2 of Patient 3 before NIMV showed significant desaturations breathing room air, with a mean SaO2 of 85 ± 9.8% with 61% of the nocturnal time spent with a SaO2 below 90%. She also had excessive daytime sleepiness, falling asleep while at school. While using NIMV, nocturnal SaO2 improved greatly, with a mean SaO2 of 95 ± 2.3%, 70% of the time spent with a SaO2 > 95% and no time spent with a SaO2 < 90%. The excessive daytime sleepiness disappeared, and the school teachers noticed a dramatic improvement in her school performances. Sadly, she died 1 yr after starting NIMV from an unexplained alveolar hemorrhage. Autopsy was not performed.
Patient 4 was started on NIMV because of severe sleep disturbance and sleep apnea. These symptoms totally disappeared during NIMV, which was also associated with consistent progress in his psychoneurologic development. Because of familial problems and the discovery of glioma of the optic chiasma, NIMV was changed to long-term oxygen therapy after 6 mo.
Patient 5 had laryngomalacia associated with familial dysautonomia. NIMV resulted in a stabilization of his obesity and improvement in his daytime stamina and school performances. Compliance with NIMV was excellent, with a mean daily use of 8.2 ± 1.5 h.
NIMV was well tolerated in all the patients. The chest wall deformity, which was present in three patients (Patients 6, 7, and 11) improved noticeably after a mean of 6 mo of NIMV. NIMV was also associated with significant weight gain in four patients (Patients 6 to 9).
In this group, SaO2 was checked during a whole night in the hospital before discharge, once the patient was totally adapted to the ventilator, which occurred generally within 2 wk after the start of NIMV. Mean nocturnal SaO2 improved significantly from 91.7 ± 2.3% before to 96.2 ± 2.0% during NIMV, p = 0.03. The nocturnal nadir SaO2 also improved significantly after the initiation of NIMV, with a mean nadir SaO2 of 74.7 ± 7.5% before NIMV to a nadir of 88.0 ± 2.5% while receiving NIMV. The percentage of night time spent with a SaO2 < 90% fell from 29.5 ± 19.6% before NIMV to 0.5 ± 0.8% while receiving NIMV, p = 0.03.
Successful discontinuation of NIMV was possible in four patients after 6 to 30 mo. Patient 10, who has a picnodysostosis, is still dependent on NIMV after 52 mo.
Compliance was systematically assessed at home in the two groups of patients on a monthly basis and was excellent, with a mean daily use of 10.2 ± 2.3 h/d in the patients performing a daytime nap and a mean daily use of 7.9 ± 1.9 h per night in the other patients.
This study is the first to quantify the breathing effort in children with stridor caused by laryngomalacia and to demonstrate that NIMV can adequately unload the respiratory muscles and improve the clinical status of these children on a long-term follow up.
Different nasal masks and ventilators were used in this study. A limited number of commercial masks are available for infants. Furthermore, the volume of the mask is often too large, therefore limiting the benefit of NIMV because of the increase in dead space, especially in these young children. For this reason, we recommend custom-made masks, molded when the child sucks his or her pacifier, which favors simultaneous closure of the mouth during NIMV. The dead space of the custom-made masks did not exceed 5 ml. Concerning the ventilators, no specific comparison was made in this study. The most efficient ventilator and ventilatory mode, based on the lowest Pdi swings, as well as what was best tolerated by the patient, was adopted (Table 1).
Breathing pattern, gas exchange, and PTPdi were related to age. Normal gas exchange was preserved in the youngest children at the expense of an increased PTPdi, whereas the opposite was observed in older children. We speculate that the longer duration of the resistive load in these older children could have caused a “resetting” of the respiratory centers so that the penalty of a reduced oxygen cost of breathing was an increase in arterial CO2. Whether respiratory muscle fatigue could also be implicated in this response cannot be excluded, though the present data do not address this.
This study shows that NIMV can effectively unload the respiratory muscles in children with severe laryngomalacia. The maintenance of a continuous pressure by PEEP is the most important component of a ventilatory assistance in patients with an upper airway obstruction. This PEEP alleviates the airway obstruction, either by splinting the upper airway open, or by increasing functional residual capacity, which in turn reflexively dilates the pharynx (21, 22). The level of PEEP sufficient to relieve airway obstruction depends on the severity of the obstruction and the patient's state, i.e., sleep or wakefulness. In this study, the relatively low levels of PEEP, sufficient to adequately decrease Pes and Pdi during wakefulness, are probably insufficient during sleep. Most investigators agree that the recording of Pes represents the “gold standard” to adjust CPAP and bilevel PS titration (23, 24). However, because in the patients of Group 1, we observed an excellent relationship between the decrease in Pes and Pdi swings and the disappearance of the clinical signs of loud breathing, we decided to use a noninvasive method on the patients of Group 2 for the titration of PEEP and PS.
The tolerance of NIMV was excellent in all patients. This certainly contributed to the good compliance in these patients. The better comfort of this noninvasive technique, as compared with a tracheostomy, is likely to translate into a better quality of life, not only for the patients but also for their families.
We did not perform polysomnography in our population, but we clearly observed an improvement of nocturnal SaO2 in Group 2. In addition, all the families claimed a subjective improvement in the quality of sleep, daytime sleepiness, and quality of life of their child when NIMV was used. Abnormal respiratory effort during sleep has been shown to worsen sleep architecture (25). This makes us speculate that NIMV could improve the quality of sleep of our population by reducing the respiratory effort.
Moreover, we observed a decrease in clinical indices of malnutrition after initiation of NIMV. If the cost of breathing represents less than 5% of the total energy expenditure in normal subjects, it is well known that it can be markedly elevated in critically ill patients. This is also probably the case in our population, as reflected by the improvement of anthropometric indices during chronic NIMV, which suggests that the decrease of work of breathing during NIMV observed in a very short period (20 to 30 min) persists in a longer and nonexperimental period such as home NIMV.
In conclusion, our study documents for the first time the presence of an increase in the resistive load during wakefulness in infants and in children with severe stridor caused by laryngomalacia. This results in an increase in respiratory effort, even during wakefulness. This increase in respiratory effort was inversely related to age and resulted in abnormalities of breathing pattern and gas exchange. Although this evaluation was performed in a small group, the benefits of NIMV are of potential relevance in view of the systematic improvement of work of breathing, quality of life, and anthropometric parameters.
Supported by Breas Medical (Molndal, Sweden).
| 1. | Tucker GFLaryngeal development and congenital lesions. Ann Otol Rhino Laryngol741980142145 |
| 2. | Holinger LDEtiology of stridor in the neonate and child. Ann Oto Rhinol Laryngol891980327399 |
| 3. | McSwiney PF, Cavanagh NP, Languth POutcome in congenital stridor (laryngomalacia). Arch Dis Child521977215218 |
| 4. | Sivan Y, Ben-Ari J, Schonfeld TMLaryngomalacia: a cause for early near miss for SIDS. Int J Pediatr Otorhinolaryngol2119915964 |
| 5. | Marcus CL, Crockett DM, Ward SLEvaluation of epiglottoplasty as treatment for severe laryngomalacia. J Pediatr1171990706710 |
| 6. | Cox MA, Scheibler GL, Taylor WJReversible pulmonary hypertension in a child with respiratory obstruction and cor pulmonale. J Pediatr671965192197 |
| 7. | Roger G, Denoyelle F, Triglia JM, Garabedian ENSevere laryngomalacia: surgical indications and results in 115 patients. Laryngoscope105199511111117 |
| 8. | Phelan PD, Gillam GL, Stocks JGThe clinical and physiological manifestations of the “infantile” larynx: Natural history and relationship to mental retardation. Aust Paediatr J71971135140 |
| 9. | Dubey SP, Garap JPPediatric tracheostomy: an analysis of 40 cases. J Laryngol Otol1131999645651 |
| 10. | Wetmore RF, Marsh RR, Thompson ME, Tom LWPediatric tracheostomy: a changing procedure? Ann Otol Rhinol Laryngol1081999695699 |
| 11. | Rapoport DMMethods to stabilize the upper airway using positive pressure. Sleep191996S123S130 |
| 12. | Sanders MH, Kern NObstructive sleep apnea treated by independently adjusted inspiratory and expiratory positive airway pressures via nasal mask. Chest981990317324 |
| 13. | Reeves-Hoché MK, Hudgel DW, Meck R, Witteman R, Ross A, Zwillich CWContinuous versus bilevel positive airway pressure for obstructive sleep apnea. Am J Respir Crit Care Med1511995443449 |
| 14. | Management of pediatric patients requiring long-term ventilation. Chest 1998;113:322S–336S. |
| 15. | Stell IM, Tompkins S, Lovell AT, Goldstone JC, Moxham JAn in vivo comparison of a catheter mounted pressure transducer system with conventional balloon catheters. Eur Respir J13199911581163 |
| 16. | Baydur A, Behrakis PK, Zin WA, Jaeger MJ, Milic-Emili JA simple method for assessing the validity of the esophageal balloon technique. Am Rev Respir Dis1261982788791 |
| 17. | Field S, Sanci S, Grassino ARespiratory muscle oxygen consumption estimated by the diaphragm pressure-time index. J Appl Physiol5719844451 |
| 18. | Barnard PA, Levine SCritique on application of diaphragmatic time-tension index to spontaneously breathing humans. J Appl Physiol60198610671072 |
| 19. | Guilleminault C, Pelayo R, Clerk A, Leger D, Boclan RCHome nasal continuous positive airway pressure in infants with sleep-disordered breathing. J Pediatr1271995905912 |
| 20. | Waters KA, Everett FM, Bruderer JW, Sullivan CEObstructive sleep apnea: the use of nasal CPAP in 80 children. Am J Respir Crit Care Med1521995780785 |
| 21. | Sullivan CE, Issa FG, Berthon-Jones M, Eves LReversal of obstructive sleep apnea by continuous positive airway pressure applied through the nares. Lancet11981862865 |
| 22. | Strohl KP, Redline SNasal CPAP therapy, upper airway muscle activation, and obstructive sleep apnea. Am Rev Respir Dis1341986555558 |
| 23. | Condos R, Norman RG, Krishnasamy I, Peduzzi N, Golding RM, Rapoport DMFlow limitation as a non invasive assessment of residual upper-airway resistance during continuous positive airway pressure therapy of obstructive sleep apnea. Am J Respir Crit Care Med1501994475480 |
| 24. | Montserrat JM, Ballester E, Olivi H, Reolid A, Lloberes P, Morello A, Rodriguez-Roisin RTime-course of stepwise CPAP titration. Behavior of respiratory and neurologic variables. Am J Respir Crit Care Med152199518541859 |
| 25. | Gleeson K, Zwillich CW, White DPThe influence of increasing ventilatory effort on arousal from sleep. Am Rev Respir Dis81421990295300 |
This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org
