Dynamic Computed Tomography for Evaluation of Tracheobronchomalacia in Premature Infants with Bronchopulmonary Dysplasia

Abstract

Introduction:

Dynamic computed tomography (dCT) gives real-time physiological information and objective descriptions of airway narrowing in tracheobronchomalacia (TBM). There is a paucity of literature in the evaluation of TBM by dCT in premature infants with bronchopulmonary dysplasia (BPD). The aim of this study is to describe the findings of dCT and resultant changes in management in premature infants with TBM.

Methods:

A retrospective study of 70 infants was performed. Infants included were < 32 weeks gestation without major anomalies. TBM was defined as ≥ 50% expiratory reduction in cross-sectional area with severity defined as mild (50–75%), moderate (≥75–90%), or severe (≥90%).

Results:

Dynamic CT diagnosed malacia in 53% of infants. Tracheomalacia was identified in 49% of infants with severity as 76% mild, 18% moderate, and 6% severe. Bronchomalacia was identified in 43% of infants with varying severity (53% mild, 40% moderate, 7% severe). Resultant management changes included PEEP titration (44%), initiation of bethanechol (23%), planned tracheostomy (20%), extubation trial (13%), and inhaled ipratropium bromide (7%).

Conclusion:

Dynamic CT is a useful noninvasive diagnostic tool for airway evaluation of premature infants. Presence and severity of TBM can provide actionable information to guide more precise clinical decision making.

Introduction

The central airway is a dynamic structure that changes size and shape during the respiratory cycle. Excessive central airway collapse (ECAC) describes two pathophysiologic entities: excessive dynamic airway collapse (EDAC) and tracheobronchomalacia (TBM). EDAC is airway compromise characterized by weakness of muscular fibers in the posterior membrane of the trachea leading to luminal invagination during expiration. TBM is a dynamic narrowing of airways in the expiratory phase characterized by weakness of the airway walls due to softening of cartilaginous rings. Both ECAC entities can occur in isolation or in association with other conditions^1,2. Although the pathophysiology is different for each, the clinical features and therapeutic goals are similar.

Premature infants with bronchopulmonary dysplasia (BPD) are particularly susceptible to developing TBM because of less time for maturation and increased utilization of accessory muscles to overcome airway resistance. Although TBM is not common in the general population (1 in 2,100), estimated prevalence of TBM in neonates with BPD undergoing bronchoscopy is estimated at 10 to 46%^3,4. Hysinger et al. found an incidence of 48% and 41% of tracheomalacia and bronchomalacia respectively in infants with severe BPD⁵.

Historically direct visualization of the airways by bronchoscopy has been the gold standard for diagnosing TBM in all age groups based on airway collapse during spontaneous respiration. Most experts believe that more than 50% narrowing of the central airway lumen is abnormal^1,6. This defined threshold is limited due to the scarcity of normal ranges for central airway collapse among children with and without coexisting pulmonary disease. Bronchoscopy is not without limitations due to the need for an experienced bronchoscopist, invasive nature, decreased interrater reliability, and only suggesting semi-quantitative data^5,7–10. Thus, there is a critical need for a quantifiable, non-invasive diagnostic modality to evaluate airway anomalies including TBM in premature infants requiring prolonged mechanical ventilation. Dynamic airway evaluation and angiography using a contrast-enhanced multidetector computed tomography (CT) with two-dimensional (2D) and three-dimensional (3D) reconstructions, has been used in the assessment of thoracic anomalies including TBM in older children^11–20. These studies report excellent concordance between bronchoscopy and dynamic airway CT (dCT) for evaluation of thoracic anomalies ^13–20. Similarly dCT has been used to describe lung parenchyma changes representing significant lung disease in infants with BPD^19,21–25. However, the role of dCT for the evaluation of malacic airways, in particular, in preterm infants with BPD has not been reported.

Among preterm infants with BPD, TBM is associated with higher adverse outcomes, including higher rates of extubation failure, episodes of pneumonia and in-hospital mortality rate. TBM is associated with increased risk of gastroesophageal reflux disease, impaired airway clearance, surgeries (tracheostomy and gastrostomy tube), and increased length of mechanical ventilation and hospital stay^4,26. At discharge, infants with TBM are more likely to require prolonged mechanical ventilation, more medical therapies and less likely to receive oral feedings⁴. The inability to precisely and objectively diagnose the presence and extent of TBM limits advancements in mitigation of its adverse effects.

We present a five-year, single-center level IV neonatal intensive care unit experience of non-breath held dCT as a diagnostic modality for identification of TBM in premature infants with BPD. The primary objective of this study is to perform a retrospective review to describe the indications for assessment, findings, and resulting changes in clinical management of preterm infants with TBM diagnosed by dCT.

Methods

This was a retrospective study of preterm infants born prior to 32 weeks gestational age who underwent a diagnostic dCT identified by our electronic health record from 2017 – 2022 at the Arkansas Children’s Hospital Level IV neonatal intensive care unit. This study was approved by the Institutional Review Board at the University of Arkansas for Medical Sciences. The inclusion criteria for the study were infants less than 32 weeks of gestation without a significant genetic, congenital airway, or complex cyanotic cardiac lesion. Cardiac anomalies including a history of a patent ductus arteriosus, atrial septal defect and ventricular septal defect were not excluded. At our institution, dCT is reserved for a small subset of chronically ventilated infants with BPD and complex ventilation physiology not responding to standard ventilation management strategies. The dCT is performed with the goal of assessing the central tracheobronchial tree for TBM as well as lung parenchyma for atelectasis.

Diagnostic Technique

Dynamic CT airway protocol was performed by use of a wide-detector scanner (Aquilion One, Toshiba, Japan) obtained at 0.5-mm collimation and a minimum gantry rotation of 0.350 seconds. Most of the scans were performed without contrast. The contrast was used for patients with any underlying heart condition or suspected vascular anomalies. Each patient’s scan length was tailored to their size on the scanogram to include the field of view from thoracic inlet to beyond the segmental bronchi and staying above the diaphragm. The maximum z-axis limit was 16 cm which could be adjusted in 2 cm steps. Scans were performed without anesthesia with free breathing for non-intubated patients. For intubated patients, the endotracheal tube was adjusted as determined by the initial scanogram and was pulled just above the thoracic inlet. Our institutional protocol requires a neonatal respiratory therapist and a neonatal nurse practitioner/ neonatologist to accompany infants undergoing dynamic CT to the radiology suite. Any adjustments to the endotracheal tube at the time of imaging are performed by the respiratory therapist with assistance of the NICU nurse. The scanogram allows for precise measurement of the degree to adjust the endotracheal tube, and no complications resulted from this adjustment. A repeat scanogram is performed after the adjustment to ensure adequate position to avoid the endotracheal tube obstructing the proximal trachea evaluation. Next, PEEP (positive end expiratory pressure) was stopped or lowered to a level of 1 and after waiting for up to a minute, the scan was performed. The infant was monitored by continuous pulse oximetry to ensure adequate oxygen saturations and work of breathing was monitored by visual assessment of accessory muscle use and respiratory rate. We used 80 kVp continuous scanning, and the mA was determined based on the formula mA = [(kg × 2.0) + 5] / 0.35 ²⁰. During dynamic scanning, the maximum respiratory rate for intubated patients was set at 40 bpm (breaths per minute). For example, for a respiratory rate of 40 bpm, continuous scan was performed for 1.4s to obtain four cycles with rotation of 350 ms/rotation. The maximum total respiratory rate is set at 40 bpm, above which the scan is not performed. With a higher rate, there is excessive breathing motion that can affect the quality of the scan. Rarely for a temporary rise in respiratory rate, a small dose of benzodiazepine is given to calm the infant. The scan yields 8–10 snapshots (axial image sets) during a respiratory cycle, which were then reconstructed into 3D and cine airway imaging using a separate software (Vitrea FX, Vital Images, Minneapolis).

Defining tracheobronchomalacia

Tracheomalacia and right and left bronchomalacia were defined as > 50% reduction in cross-sectional luminal area by dCT during expiration. A 50% reduction is most commonly agreed by experts as abnormal and is consistent with the most recent clinical definition by the European Respiratory Society (ERS) Task Force ^6,10,27. The degree of severity of tracheomalacia and bronchomalacia was defined as mild (50–75% reduction), moderate (≥75–90% reduction), or severe (≥90% reduction) ^10,28.

Data collection

Patient demographics including birth gestational age and weight, sex, maternal race/ethnicity, delivery type, multiple gestation, 5-minute apgar, antenatal steroids, respiratory support at 36 weeks postmenstrual age (PMA), and respiratory support on the day of dCT were abstracted from the electronic medical record. BPD was defined by respiratory support provided at 36 weeks PMA as (grade 1) nasal cannula ≤2 L/min; (grade 2) nasal cannula >2 L/min or noninvasive positive airway pressure; or (grade 3) invasive mechanical ventilation²⁹. Gross findings and percentages of expiratory tracheal and bronchial collapse were abstracted from dCT radiological reports. Clinical indications for evaluation by dCT were collected from daily progress notes and subspecialty consultation notes. Information on clinical interventions within one week following dCT evaluation was obtained from clinical notes and flowsheets. Data were reported as descriptive measures (mean, standard deviation, median, inter-quartile range or frequency). Statistical calculations were used to help describe our population (median, interquartile range) and depict how often (N, %) there was a change in clinical management based on the results of the dCT.

Results

A total of one hundred and five preterm infants underwent a dCT for clinical concerns of needing prolonged respiratory support during the initial NICU stay. Among these, seventy infants were born at less than 32 weeks of gestation and did not have any significant congenital anomalies, thereby meeting our inclusion criteria. The median birth gestational age of the infants was 25 weeks (IQR = 24 – 28) with 41/70 (59%) of the patients having Grade 3 BPD at 36 weeks PMA with a median PEEP level of 8 (IQR = 7 – 9) (Table 1). Infants underwent dCT at a median PMA of 44 weeks (IQR = 41 – 48) and median chronological age of 4.4 months (IQR = 3.6 – 5.3). Multiple failed extubation attempts were the most common indication for obtaining dCT imaging (Table 2). On the day of dCT, 47/70 (67%) of infants required invasive mechanical ventilation with an invasive median PEEP of 8 (IQR = 7 – 10). Infants requiring invasive ventilatory support were more likely to have a greater degree of malacia than infants on non-invasive ventilation.

Table 1.

Demographics of Infants with Dynamic Computed Tomography Imaging

Gestational Age Weeks at Birtha	25 (24 – 28)
Weight in Grams at Birtha	675 (532 – 860)
SGAc	21 (30%)
Male	37 (53%)
Delivery
Cesarean Section	58 (83%)
Vaginal	12 (17%)
Singleton at Birth	61 (87%)
5-Minute APGAR ≤ 6	47 (67%)
Race
Black	32 (46%)
White	29 (41%)
Other	9 (13%)
Antenatal Steroids
Full Course	46 (66%)
Partial Course	14 (20%)
None	10 (14%)
BPDb
Grade 3	41 (59%)
Grade 2	22 (31%)
Grade 1	7 (10%)
PEEP for Invasive Ventilationa
36 Weeks PMA (N = 40)	8 (7 – 9)
Day of dynamic CT (N = 49)	8 (7 – 10)

Data is presented as N (%) with total N = 70 unless otherwise specified

^aData is presented as median (interquartile range 25^th – 75^th)

^bBPD (Bronchopulmonary Dysplasia) defined by Jensen et. al. 2019¹⁶

^cSGA (Small for Gestational Age)

Table 2.

Infant Characteristics at Dynamic CT and Malacia Findings

PMA at Dynamic CT in Weeksa	44 (41 – 48)
Chronological age at Dynamic CT in Daysa	133 (108 – 159)
Indication for Dynamic CTb
Multiple failed extubations	28 (40%)
Assess Central Airway	21 (30%)
Elevated PEEP	19 (27%)
Inability to wean positive pressure	18 (26%)
Persistent Hypercarbia/Increased WOB	9 (13%)
Desaturation/Bradycardia (BPD) Spells	4 (6%)
Respiratory Support at Dynamic CT
Endotracheal Ventilation	47 (67%)
Tracheostomy Ventilation	2 (3%)
NIPPV or NIV NAVA	3 (4%)
NC > 2 L or CPAP	10 (14%)
NC ≤ 2 L	8 (12%)
Dynamic CT findings
None	31 (44%)
Combined TBM	27 (39%)
Isolated Tracheomalacia	7 (10%)
Isolated Bronchomalacia	3 (4%)
Fixed Narrowing	2 (3%)
Dynamic CT Severity of TBM
Tracheomalacia (N = 34)
Mild	26 (76%)
Moderate	6 (18%)
Severe	2 (6%)
Bronchomalacia (N = 30)
Mild	16 (53%)
Moderate	12 (40%)
Severe	2 (7%)

Data is presented as N (%) with total N = 70 unless otherwise specified

^aData is presented as median (interquartile range 25^th – 75^th)

^b39/70 (56%) had more than 1 indication for dynamic CT

PMA (post menstrual age), PEEP (positive end expiratory pressure), NIPPV (nasal intermittent positive pressure ventilation), NIV NAVA (Non invasive neurally adjusted ventilatory assist), CPAP (continuous positive airway pressure), NC (nasal cannula)

About half of these eligible patients, that is 37 out of 70, were identified as having either tracheomalacia, bronchomalacia, or combined TBM. Tracheomalacia and bronchomalacia were seen in 33/70 and 34/70 infants, respectively. A combination of tracheomalacia and bronchomalacia was the most common finding seen in 27/70 (39%) infants. Most patients with combined tracheobronchomalacia showed mild severity, defined as a reduction in cross-sectional airway lumen between 50% to 75%. Severe malacia, defined as ≥90% reduction in the cross-sectional luminal area or near complete occlusion was identified in 6–7% of patients with tracheobronchomalacia. A common theme among all dCT imaging was changes in lung parenchyma representing significant lung disease including areas of interstitial and cystic changes, fibrosis, and areas of atelectasis similar to prior studies reporting CT findings in infants with BPD ^21–25,30. The median effective dose for the dynamic airway studies was 1.1 mSV (range 0.4–1.9 mSV) which is similar to that described previously²⁰. A dCT depicting central airway malacia with paired inspiratory and expiratory imaging is demonstrated (Figure 1).

Figure 1:

Reconstructed images demonstrating tracheomalacia (A) as well as bronchomalacia (B) during inspiration and expiration. Dynamic computography images of the mid trachea (C) and right main bronchus (D) during inspiration and expiration demonstrate more than 50% decrease in cross-sectional area consistent with TBM.

Changes in clinical management occurred in 64/70 (91%) infants who underwent a dCT for evaluation of central airway collapse (Table 3). The most common intervention following dCT imaging was adjustment in PEEP support (44%) followed by initiation of bethanechol (23%). PEEP was titrated in incremental levels of 1–2 cm H₂0 depending on the degree of severity of malacia or the absence of malacia. Our institution uses a multidisciplinary approach to PEEP titration in infants with BPD and tracheobronchomalacia. The PEEP is often titrated in increments a few times per week while monitoring for improvements in “BPD spells,” hyperexpansion, air trapping, gas exchange, and tolerance of handling cares. Two infants (3%) were identified as having fixed airway compression secondary to vascular anomalies and underwent further evaluation by computed tomography angiogram (CTA). Figure 2 depicts an infant with BPD who underwent dCT imaging to evaluate TBM but was subsequently found to have central airway compression secondary to a vascular ring. In about 20% of infants, a decision was made to proceed with tracheostomy and long-term ventilation due to diffuse and severe tracheobronchomalacia. A decision was made to proceed with an extubation attempt due to no or mild malacia in about 13% of patients. Of the infants who underwent a trial of extubation following findings on dCT imaging, 63% of infants remained extubated during the hospitalization.

Table 3.

Changes in Clinical Management following Dynamic CT

Any Change in Managementa	64 (91%)
PEEP Titrationb
PEEP Decreased	17 (24%)
PEEP Increased	14 (20%)
Bethanechol Initiated	16 (23%)
Proceed with Tracheostomy	14 (20%)
Proceed with Extubation Attempt
Without Steroids	5 (7%)
With Steroids
Anti-Inflammatory Lung Steroids	2 (3%)
Airway Steroids	2 (3%)
Atrovent	5 (7%)
Surfactant Studies (ILD panel)	4 (6%)
Transpyloric feeds	3 (4%)
Other Imaging (CTA)c	2 (3%)

Data is presented as N (%) with total N = 70

^a24/70 (34%) had more than 1 change in clinical management following dynamic CT

^bPEEP (Positive End Expiratory Pressure)

^cComputed Tomography Angiogram (CTA)

Figure 2:

Infant with chronic lung disease with dCT imaging to evaluate TBM. Dynamic CT images during inspiration and expiration demonstrates severe segmental narrowing of the mid thoracic trachea. On subsequent chest angiography evaluation, patient was found to have a vascular ring with double aortic arch.

Discussion

In this retrospective review, we present our experience of dCT in the diagnosis and management of TBM in premature infants with BPD requiring prolonged positive pressure support. Dynamic CT can be used to evaluate the exact location and severity of the airway malacia and often identifies airway pathologies leading to changes in medical management. These results suggest that dCT can be a useful tool to guide treatment of infants with BPD and chronic respiratory failure.

In infants with BPD, prematurity disrupts the natural maturation of the airways resulting in small, highly compliant structures that are prone to collapse and injury. Small changes in airway size in infants with TBM lead to decreased airflow and increased utilization of accessory muscles for exhalation, which can increase transmural airway pressure^11,12,26. The resulting positive pleural pressure and accentuated resistive intraluminal pressure drop serve to increase the collapsing transmural airway pressure resulting in airway narrowing³¹. Consequently, both dynamic and fixed central airway pathologies are common in this patient population and are associated with increased respiratory morbidity including increased air trapping, respiratory infections, and prolonged ventilatory support and hospitalizations^4,11,12,26.

There are currently no specific criteria to guide which high risk infants should be assessed for TBM. Clinical suspicion for TBM may be due to the inability to wean from invasive mechanical ventilation, the need for high PEEP, agitation episodes “BPD spells” of bradycardia or desaturations, or prolonged need for ventilatory support ^28,32. In this study, infants with BPD underwent dCT at a median PMA of 44 weeks with 67% of these infants requiring endotracheal mechanical ventilation. The most common indication for dCT in our study was failure to wean positive pressure ventilation. More than half of our infants (53%) undergoing dCT were diagnosed with tracheomalacia or bronchomalacia while more than a third (39%) of infants had combined TBM. This is slightly higher than the prevalence (10% - 46%) in previous studies, suggesting dCT may be more sensitive at diagnosing TBM than has been the case with other diagnostic tests^5,13. Similarly our prevalence is slightly above the a smaller subset of infants examined by Hysinger et al who reported tracheomalacia in 13/27 (48%) and bronchomalacia in 11/27 (41%) in children with severe BPD whom underwent flexible bronchoscopy⁵. The difference in findings could be related to underestimation of airway malacia while using flexible scope via artificial airway which can generate auto-PEEP by preventing complete exhalation which may mask TBM^33,34. Most infants we observed with TBM had mild narrowing (50–75%) while only (6–7%) of infants had severe features of collapse though the clinical significance of the varying degrees of TBM severity remain unclear. While no studies have rigorously evaluated therapeutics for TBM in patients with BPD, treatment strategies for TBM in general include positive pressure support, pharmacotherapies, and surgical intervention.

Positive pressure titration can improve respiratory mechanics in children with tracheomalacia and was the most common intervention used in our cohort³⁵. In this study PEEP optimization was done in (44%) subjects in response to results from a dCT scan. PEEP can be provided non-invasively or invasively via an endotracheal tube or tracheostomy which can bypass the collapsible segment of trachea. The optimum PEEP serves as a pneumatic stent which decreases airway resistance, reduces respiratory work, and raises lung volumes in infants with TBM. High PEEP strategies (>10 cm H₂0) are often needed to prevent airway collapse and air-trapping associated with TBM^5,36. We titrate PEEP based on lung mechanics, such as changes in compliance and resistance, as well as extrinsic PEEP settings, but PEEP can also be titrated under direct visualization with bronchoscopy or by dCT PEEP studies in infants with severe BPD who are not optimal candidates for bronchoscopy^36,37. In 20% of our infants, lung parenchymal disease as well as TBM being diffuse and severe enough that the decision was made to offer tracheostomy and long term mechanical ventilation. While in 13% of infants, the decision was made to proceed with a trial of extubation based on dCT findings.

Pharmacotherapies for the treatment of TBM are limited, but bethanechol and ipratropium bromide have been two strategies used to treat dynamic airway collapse⁵. In our study, one in four infants who were subsequently found to have TBM after undergoing a dCT were started on bethanechol. Bethanechol, a muscarinic agonist, reduces tracheal compliance in neonatal animal models and improves respiratory mechanics and symptoms in infants and children with TBM though no trials have studied efficacy in infants^38,39. Ipratropium bromide was used in 7% of our infants with TBM. Low dose Inhaled ipratropium bromide blocks type 2 muscarinic receptors, which potentiates acetylcholine activity in the neuromuscular junction and stimulates contraction of tracheal smooth muscle⁴⁰. In infants with severe BPD, the narrowing of the airway occurs in both the posterior wall and anterior cartilaginous trachea making effective treatment difficult to determine. Future investigations will need to evaluate the efficacy of these treatments in neonates with severe BPD. Currently we are investigating the use of bethanechol on the clinical effect of TBM in preterm infants with BPD⁴¹.

As therapeutic options for the management of TBM are limited, objective, standardized tests to diagnose TBM are essential to advance treatment strategies which otherwise would not be possible. Both invasive and non-invasive diagnostic tests have been used for the evaluation of TBM including rigid and flexible bronchoscopy, magnetic resonance imaging, and dynamic airway computed tomography. Pediatric flexible bronchoscopy performed under general anesthesia with the spontaneously ventilating child has historically been preferred as a safe method of evaluating the lower airways. Flexible bronchoscopy is often considered superior to rigid bronchoscopy in the evaluation of dynamic airway changes due to the ability to provide nondisruptive visualization of the distal airways^42,43. It can demonstrate structural abnormalities, assess airway dynamics and patency pressures, and quantify the extent of the involved airways which closely correlates with morbidity. Though effective for airway evaluation, a major limitation of bronchoscopy is the inability to pass the instrument distal to any point of fixed narrowing within the tracheobronchial tree because of concerns regarding potential interference with gas exchange in the more distal airways. Additionally, due to the size of bronchoscopes, it is difficult to evaluate distal small airway disease. Thus, there has been a need for a non-invasive sensitive diagnostic modality for evaluating airway abnormalities in infants who may not be candidates for bronchoscopy.

Non-invasive imaging techniques which include ultrashort echo-time magnetic resonance imaging (UTE MRI) and dCT have been developed as promising diagnostic tools to benefit infants by use of less invasive techniques. Compared to bronchoscopy, both dCT and UTE MRI offer the benefit to characterize both parenchymal lung disease and evaluate for external airway malformations in infants with BPD^37,44,45. Our dCT imaging identified two infants with external compression from a vascular ring first identified by dCT which ultimately led to CTA imaging and surgical correction.

UTE MRI has recently been validated and correlated with flexible bronchoscopy in infants with TBM and demonstrates great potential in evaluating central airways while decreasing exposure to potential radiation in infants⁴⁶. Recent studies evaluating neonatal airways by use of UTE MRI continue to be developed, though widespread use of this imaging technique may be limited by center availability in addition to the technical challenges of completing an MRI quickly in an awake mechanically ventilated infant^47,48. Dynamic CT is a noninvasive technique of increasing utility in infants that require sensitive analysis⁴⁵. By using a short, quantifiable, low radiation technique, this may be adapted to referral centers where bronchoscopy and UTE MRI are limited or unavailable. Greenburg has previously described this technique for dCT in children with low ionizing radiation, however direct correlative studies with bronchoscopy were not performed¹⁶. In older children and adults, dCT has demonstrated a high concordance with bronchoscopy though it has not been validated or extensively described in preterm infants^13–20. Montoya et al. has recently described the successful use of dCT in preterm infants with BPD though this was limited by a small subset of infants³⁷. Our report highlights the clinical investigation of infants suspected to have TBM, the findings of dCT in a larger cohort of preterm infants with BPD, and the resulting change in clinical management based on results.

Dynamic CT utilizes many techniques that help reduce the radiation dose for example it is performed using a volume CT scanner, which allows the entire length of the airway to be scanned simultaneously as opposed to a helical CT scanner, which scans one slice at a time. The volume scanner imparts 18–40% lower radiation as compared to helical scanners⁴⁹. The region of interest can be adjusted to patients’ size to only include the airways. The inherent contrast between the airways and the lungs allows a low voltage and low currents to be utilized. In addition, use of adaptive iterative dose reduction algorithms has further reduced the dose^20,50. Our institution has previously shown the median effective dose of 1.1 mSV (range 0.9–1.9 mSV) for dCT²⁰. Other studies have used dCT to assist the clinicians in determining effective PEEP in children with severe BPD and have reported the mean effective dose of 0.4 mSv/scan (range of 0.1 mSv to 0.8 mSv)³⁶. Another study which compared dCT to bronchography showed the mean effective dose for dCT to be 1.0 mSv (range 0.3–2.7 mSv) as compared to mean of 1.4 mSv (range: 0.3–3.5mSv) for bronchography. Dynamic CT dose was very similar to the routine chest CT in the age matched controls with mean of 1.0 mSv and range of 0.5–1.9 mSv⁵⁰.

This study is not without limitations. First, this is a single center retrospective review not focused on determining the clinical effectiveness of the changes in management based on dCT results. Dynamic CT gives real time physiologic data regarding the presence and severity of malacia which provides information for medical decisions, but the extent of clinical significance of these findings remains unclear. The goal of using dynamic CT at our institution is to evaluate the presence and degree of malacia. The presence of malacia was determined at the lowest level of PEEP possible, and was imaging was not obtained at the baseline PEEP for infants. Though valuable clinical information would be obtained by measuring malacia at different PEEP levels, the tradeoff would be increased radiation exposure. Future studies will be important to evaluate outcomes from change in management resulting from the use of dCT as a diagnostic tool. Second, we did not correlate our dCT findings to the historical gold standard, bronchoscopy. At our center, we do not routinely perform bronchoscopy in infants due to concerns of the invasive nature, and experience and availability of a well-defined dCT imaging protocol. Very few of our patients had a bronchoscopy within 1–2 weeks of dCT imaging which would make prolonged intervals not reliable for comparison. Further studies should investigate the correlation of bronchoscopy and dCT in a prospective manner. Finally, the degree of malacia was determined by the cross-sectional narrowing during expiration with varying degrees of severity ranging from mild (50–75% reduction), moderate (≥75–90% reduction), or severe (≥90% reduction). These definitions have not been validated against severity of clinical presentation or long-term outcomes in preterm infants with BPD. Despite such limitations, our center has a standardized approach with dCTs and decision making involves multidisciplinary discussions as part of our BPD collaborative meetings with pulmonologists, otorhinolaryngologists and neonatologists that jointly consider the infant’s clinical course and radiological findings in decision making. Our dCT images are read by a core group of radiologists minimizing inter-rater reliability. We recognize we have no consistent comparison with bronchoscopy. However, just by nature of having a consistent approach regarding indications and response to resultant findings on dCT, we are able to tailor our management based on the unique needs and objective data for each infant.

TBM is a common comorbidity in infants with BPD that is associated with increased use of medical resources and prolonged hospitalization. Currently, there are no standardized treatments or diagnostic criteria for central airway collapse in infants with BPD. Dynamic CT has potential as a non-invasive diagnostic tool for the airway evaluation of premature infants requiring protracted and prolonged positive pressure ventilation. Objective evidence of the presence and severity of TBM can provide actionable information to guide more precise clinical decision making. Further studies are needed to validate dCT imaging among premature infants and develop targeted therapies for dynamic airway pathology.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.