Increased Work of Breathing due to Tracheomalacia in Neonates

Abstract

Rationale: Dynamic collapse of the tracheal lumen (tracheomalacia) occurs frequently in premature neonates, particularly in those with common comorbidities such as bronchopulmonary dysplasia. The tracheal collapse increases the effort necessary to breathe (work of breathing [WOB]). However, quantifying the increased WOB related to tracheomalacia has previously not been possible. Therefore, it is also not currently possible to separate the impact of tracheomalacia on patient symptoms from parenchymal abnormalities.Objectives: To measure the increase in WOB due to airway motion in individual subjects with and without tracheomalacia and with different types of respiratory support.Methods: Fourteen neonatal intensive care unit subjects not using invasive mechanical ventilation were recruited. In eight, tracheomalacia was diagnosed via clinical bronchoscopy, and six did not have tracheomalacia. Self-gated three-dimensional ultrashort-echo-time magnetic resonance imaging (MRI) was performed on each subject with clinically indicated respiratory support to obtain cine images of tracheal anatomy and motion during the respiratory cycle. The component of WOB due to resistance within the trachea was then calculated via computational fluid dynamics (CFD) simulations of airflow on the basis of the subject's anatomy, motion, and respiratory airflow rates. A second CFD simulation was performed for each subject with the airway held static at its largest (i.e., most open) position to determine the increase in WOB due to airway motion and collapse.Results: The tracheal-resistive component of WOB was increased because of airway motion by an average of 337% ± 295% in subjects with tracheomalacia and 24% ± 14% in subjects without tracheomalacia (P < 0.02). In the tracheomalacia group, subjects who were treated with continuous positive airway pressure (CPAP) using a RAM cannula expended less energy for breathing compared with the subjects who were breathing room air or on a high-flow nasal cannula.Conclusions: Neonatal subjects with tracheomalacia have increased energy expenditure compared with neonates with normal airways, and CPAP may be able to attenuate the increase in respiratory work. Subjects with tracheomalacia expend more energy on the tracheal-resistive component of WOB alone than nontracheomalacia patients expend on the resistive WOB for the entire respiratory system, according to previously reported values. CFD may be able to provide an objective measure of treatment response for children with tracheomalacia.

Keywords: computational fluid dynamics; continuous positive airway pressure; tracheomalacia; ultrashort-echo-time magnetic resonance imaging; work of breathing.

Figures

Figure 1.

(A) Time course of the initial phase of each magnetic resonance radial acquisition (FID) (black line) and the smoothed respiratory waveform for each breath binned into four phases of respiration (yellow = end expiration; blue = peak inspiration; green = end inspiration; orange = peak expiration), over a representative snapshot of 4 seconds of acquisition. (B) Axial slice from respiratory-gated ultrashort-echo-time magnetic resonance imaging of the thorax showing the trachea of a tidal-breathing neonatal subject with tracheomalacia at different time points within the breathing cycle. Data from each respiratory bin were used to reconstruct an individual respiratory-gated image. (A) Note that the initial phase of the FID captures negative displacement of the diaphragm, explaining the inverted polarity of the resulting waveform. (B) The tracheal cross-section was smaller at end expiration and larger at end inspiration, with intermediate changes at peak inspiration and peak expiration. FID = free induction decay.

Figure 2.

Left and right bronchial airflow rates of an example subject with tracheomalacia. The light-gray shaded area indicates inspiration, and the white area indicates expiration during the breathing cycle.

Figure 3.

Construction of the static airway from airway surfaces throughout the respiratory cycle. Here, as in nearly all cases, the end inspiration geometry is the largest throughout most of respiration (although the geometry at peak inspiration was used above the glottis).

Figure 4.

The instantaneous TR-WOB/s due to airway motion in a realistically dynamic airway (red) and an airway held static (blue) for (A) a subject without tracheomalacia and (B) a subject with tracheomalacia. Black dashed lines represent the end inspiration. TR-WOB = tracheal-resistive component of the work of breathing.

Figure 5.

Airflow velocity distribution of a subject with tracheomalacia during peak expiration in the computational fluid dynamics simulation (A) with airway motion and (B) without airway motion. The figure shows the coronal projection of the trachea in the two simulations. Note that the model’s geometry extends into the upper airway, but this was omitted from this figure for clarity.

Figure 6.

(A) The tracheal-resistive component of the work of breathing (TR-WOB)/d of the subjects with and without tracheomalacia (TM), calculated via computational fluid dynamics simulations for static or moving airways. The increase in the TR-WOB as a (B) percentage compared with static airway for the subjects with and without TM and (C) on different respiratory support at the time of magnetic resonance imaging. (B) The average increase in the TR-WOB due to motion was 24% ± 14% (range, 10–38%) for the subjects without TM and that increase for the subjects with TM was 337% ± 295% (range, 50–848%). (C) The room-air/HFNC group demonstrated a higher increase in the TR-WOB than the RAM continuous-positive-airway-pressure (CPAP) group of the subjects with TM (P values were not calculated due to the lower number of subjects in the RAM CPAP group). The plot elements are as follows: average = cross; median = black line; interquartile range (IQR) = colored box; data within 1.5 times the IQR below 25% or above 75% = whiskers. *P values were calculated using a two-tailed paired t test. HFNC = high-flow nasal cannula.

Figure 7.

Comparison between the tracheal-resistive component of the work of breathing (TR-WOB)/d of the present study and the respiratory system–resistive WOB of previous studies. The average TR-WOB for subjects without tracheomalacia (TM) was 60.4 ± 35.7 J/d and that value for subjects with TM was 437.2 ± 454.1 J/d. The plot elements are as follows: average = cross; median = black line; interquartile range (IQR) = colored box; data within 1.5 times the IQR below 25% or above 75% = whiskers.