Delayed three-dimensional inversion recovery-prepared fast low-angle shot for airway imaging in children: More than myocardial fibrosis assessment

Abstract

Background

To investigate the ability of a delayed respiratory-navigated, electrocardiographically-gated three-dimensional inversion recovery-prepared fast low-angle shot (3D IR FLASH) sequence to evaluate the lower airways in children undergoing routine cardiovascular magnetic resonance (CMR).

Methods

This retrospective study included pediatric patients (0–18 years) who underwent clinical CMR where a delayed 3D IR FLASH sequence was performed between July 2020 and April 2021. The airway image quality and extent of lower airway visibility were graded by two blinded readers using a four-point ordinal scale (0–3). Lower airway anatomical variants and abnormalities were recorded.

Results

One hundred and eighty patients were included with a median age of 11.7 (4.6–15.3) years. Fifty-one of 180 (28%) were under general anesthesia. Overall, the median grading of airway image quality was 3 (2–3) and the extent of lower airway visibility was 3 (3–3). Interrater agreement was almost perfect (κ = 0.867 and κ = 0.956, respectively). Image quality correlated with extent of lower airway visibility (r = 0.62, p < 0.01). Delayed 3D IR FLASH was able to characterize the segmental bronchi in 137/180 (76%) and lobar bronchi in 172/180 (96%) of patients. Lower airway abnormalities were identified in 37/180 (21%) of patients and 33/129 (26%) with congenital heart disease (CHD). Identified abnormalities included tracheobronchial branching anomalies in 6/180 (3%), abnormal tracheobronchial situs in 6/180 (3%), and extrinsic vascular compression in 25/180 (14%).

Conclusion

Delayed 3D IR FLASH has excellent performance for evaluation of the lower airway anatomy and can simultaneously assess for myocardial late gadolinium enhancement. Lower airway abnormalities are not infrequently seen in children undergoing routine CMR for CHD.

Graphical abstract

1. Introduction

Both congenital and acquired large airway abnormalities frequently coexist with complex congenital cardiovascular disease [1], [2], [3], [4]. Computed tomography (CT) is considered the gold standard imaging modality for children with respiratory problems due to its ability to rapidly assess the airway and lungs [5], [6], with the disadvantage related to the potential risk of ionizing radiation in this vulnerable patient population [7]. Optimization of sequences capable of visualizing the airway would be beneficial during routine cardiovascular magnetic resonance (CMR) studies when evaluating patients with cardiovascular malformations [1], [2], [4], [8].

The respiratory-navigated, electrocardiographically-gated, three-dimensional inversion recovery-prepared fast low-angle shot (3D IR FLASH) sequence is an excellent technique for high-resolution contrast-enhanced angiography and also late gadolinium enhancement (LGE) [9], [10], [11]. When LGE imaging is performed 15–20 minutes after intravenous injection of gadolinium-based contrast medium, delayed enhancement can be seen in both areas of pathological fibrosis and also normal connective tissues, such as the respiratory adventitia. While the airways are difficult to fully characterize during the angiographic phase, they are well visualized in the delayed phase. Therefore, delayed 3D IR FLASH allows for simultaneous myocardial tissue characterization and airway imaging on routine CMR.

This study aimed to determine the quality and extent of lower airway visibility on 3D IR FLASH performed during clinical CMR. The secondary objective was to describe the lower airway variants and abnormalities observed within a cross-sectional cohort of pediatric patients with congenital heart disease (CHD).

2. Materials and methods

2.1. Patient selection

This single-center retrospective study was performed with approval from the Institutional Research Ethics Board and the need for informed consent was waived. Consecutive patients ≤18 years old who underwent a clinical CMR from July 2020 to April 2021 and had a delayed 3D IR FLASH sequence completed were included. Children were excluded if the entire lower airway was not included in the sequence, with the lower airway defined as the trachea along with the primary, secondary, and tertiary bronchi.

2.2. MRI technique

All studies were performed on a 1.5T scanner (MAGNETOM Avanto Fit, Siemens Healthineers, Erlangen, Germany). All patients underwent clinical CMR per institutional protocol. LGE imaging using 3D IR FLASH sequence was routinely performed at our institution 15–20 minutes after intravenous administration of 0.2 mmol/kg of gadobutrol (Bayer Healthcare, Leverkusen, Germany) to assess myocardial LGE [10] and the entire airway in the thorax was typically included, which did not require extension of coverage. The technique was previously described in depth [10]. Briefly, the sequence is an electrocardiographically-gated, pencil-beam respiratory-navigated, T1-weighted, gradient echo IR FLASH sequence with fat saturation, and is acquired in the coronal plane. Typical sequence parameters are field of view 250–350 mm, flip angle 18°, repetition time/echo time 330/1.4 ms, inversion time 220 ms, isotropic voxel size between 1.0 and 1.4 mm, parallel imaging factor 2, and with interpolation [10].

2.3. CMR analysis

Two independent blinded readers assessed the airway image quality and extent of lower airway visibility (L.A.I. and R.D.S.). Coronal oblique reconstructions were created along the lower airway for analysis purposes. Airway image quality was scored using a four-point ordinal scale: (0) non-diagnostic; (1) poor with motion/susceptibility artifact and poor enhancement of the lower airway walls, degrading the quality of diagnosis; (2) good with minimal motion artifact, some blurring but enhancement of the lower airway walls good for diagnostic information; and (3) excellent, no motion artifact and sharp enhancement of the lower airway walls (Fig. 1). Extent of lower airway visibility was scored using a four-point ordinal scale based on what bronchial level was able to be adequately resolved: (0) not even the trachea was depicted; (1) trachea and main or primary bronchi; (2) lobar or secondary bronchi; and (3) segmental or tertiary bronchi visualization (Fig. 2). Additionally, lower airway variants and abnormalities were recorded.

Fig. 1.

Delayed 3D IR FLASH coronal oblique reconstructions showing examples of airway image quality: (a) grade 0: non-diagnostic (14-month-old patient); (b) grade 1: poor with motion/susceptibility artifact and poor enhancement of the lower airway walls (6-year-old patient); (c) grade 2: good with minimal motion artifact (14-year-old patient); and (d) grade 3: excellent, no motion artifact (12-month-old patient). 3D IR FLASH three-dimensional inversion recovery-prepared fast low-angle shot

Fig. 2.

Delayed 3D IR FLASH coronal oblique reconstructions showing (a) the trachea and primary bronchi (arrows) (8-year-old patient), (b) secondary bronchi (arrows) (16-year-old patient), and (c) tertiary bronchi (arrows) (4-year-old patient) for purposes of extent of lower airway visibility grading. 3D IR FLASH three-dimensional inversion recovery-prepared fast low-angle shot

2.4. Statistical analysis

Demographic characteristics were summarized using descriptive statistics. Categorical variables were expressed as frequency (percentages). Continuous and ordinal variables were expressed as median (first–third quartiles). Cohen weighted kappa statistics were used to measure interrater agreement of the airway image quality and extent of lower airway visibility with <0.20 indicating slight agreement, 0.21–0.40 fair agreement, 0.41–0.60 moderate agreement, 0.61–0.80 substantial agreement, and 0.81–1.00 almost perfect agreement [12]. Spearman rank-order correlation was used for univariate associations. Mann-Whitney U test was used to compare categories. All statistical tests were two-tailed and p values <0.05 were considered statistically significant.

3. Results

3.1. Patient characteristics

During the study period, a total of 180 children had a clinical CMR that included delayed 3D IR FLASH. All studies covered the entire lower airway and so all cases were included in the study. Median patient age was 11.7 (4.6–15.3) years, ranging from 2 months to 18 years old. Fifty-one of 180 (28%) of the exams were performed with general anesthesia (GA), median age 3.4 (1.9–4.1) years. The most common indication for the study was CHD, representing 129/180 (72%) cases. The median scan time was 4:34 (3:50–5:32) minutes. Table 1 summarizes the demographic and delayed 3D IR FLASH sequence data.

Table 1.

Patient demographics and delayed 3D IR FLASH sequence data (n = 180).

Variable
Age (years)	11.7 (4.6–15.3)
Weight (kg)	38.7 (16.8–59.4)
Body surface area (m2)	1.28 (0.7–1.6)
General anesthesia	51 (28%)
Indication
Congenital heart disease*	129 (72%)
Cardiomyopathy	20 (11%)
Kawasaki disease	7 (4%)
Myocarditis	7 (4%)
Vasculopathy	6 (3%)
Other	4 (2%)
Coronary artery anomaly	3 (1.5%)
Vascular ring	3 (1.5%)
Tumor	1 (1%)
3D IR FLASH sequence
Time between injection and start of the sequence (minutes)	19 (17–22)
Acquisition time (minutes)	4:34 (3:50–5:32)
Voxel size (mm)	1.3 (1.2–1.4)

Data are presented as median (first–third quartiles) or number (%). 3D IR FLASH three-dimensional inversion-recovery prepared fast low-angle shot

^*Specific diagnoses are available as Supplemental Table 1

3.2. Quality of delayed 3D IR FLASH images for airway assessment

The median grading of airway image quality was 3 (2–3). In total, 1/180 (1%) of patients was rated as 0, 12/180 (6%) of patients were rated as 1 (all due to significant motion artifact except one who had susceptibility artifact from a vascular stent in the left pulmonary artery obscuring the distal portion of the left main bronchus), 56/180 (31%) of patients were rated as 2, and 111/180 (62%) of patients were rated as 3 (Fig. 3). Interrater agreement was almost perfect between the readers (k = 0.867 [95% confidence interval (CI) 0.817–0.917]).

Fig. 3.

Violin plots of image quality and extent of lower airway visibility grading. Solid horizontal black lines show the median and horizontal dotted lines show the 25th and 75th quartiles

The median grading of the extent of lower airway visibility was 3 (3–3). In total, 1/180 (1%) of patients was rated as 0, 7/180 (3%) of patients were rated as 1, 35/180 (19%) of patients were rated as 2, and 137/180 (76%) of patients were rated as 3 (Fig. 3). Interrater agreement was almost perfect between the readers (k = 0.956 [95% CI 0.925–0.986]). These data are summarized in Table 2. Image quality correlated with extent of lower airway visibility (r = 0.62, p < 0.01).

Table 2.

Delayed 3D IR FLASH airway image quality and extent of lower airway visibility gradings along with interobserver agreement.

	0	1	2	3	Interobserver agreement by weighted kappa [95% CI]
Image quality grading	1 (1%)	12 (6%)	56 (31%)	111 (62%)	0.867 [0.817–0.917]
Extent of lower airway visibility grading	1 (1%)	7 (3%)	35 (19%)	137 (76%)	0.956 [0.925–0.986]

3D IR FLASH three-dimensional inversion recovery-prepared fast low-angle shot, CI confidence interval

Data are presented as number (%) and κ [95% confidence interval].

For the 51/180 (28%) patients performed under GA, the median patient age was 3.4 (1.9–4.1) years at the time of CMR. In this subset of patients, the median grading of image quality was 3 (3–3) with 7/51 (14%) of patients rated as 2 and 44/51 (86%) of patients rated as 3. The median grading of the extent of lower airway visibility was 3 (3–3) with 5/51 (10%) of patients rated as 2 and 46/51 (90%) of patients rated as 3. None were rated as 0 or 1 in either category. Patients under GA had better median grading for image quality and extent of lower airway visibility compared to the patients who did not require GA (3 (3–3) vs 3 (2–3) and 3 (3–3) vs 3 (2–3) respectively, p < 0.01) (Fig. 4).

Fig. 4.

Violin plots comparing image quality and extent of lower airway visibility in the general anesthesia (GA) and non-sedated cohorts. Solid horizontal black lines show the median and horizontal dotted lines show the 25th and 75th quartiles

A total of 37/180 (21%) of patients had a lower airway finding, including 33/129 (26%) of those with CHD. Specifically, 6/180 (3%) had a tracheobronchial branching anomaly, 6/180 (3%) had abnormal tracheobronchial situs, and 25/180 (14%) had extrinsic vascular compression of either the trachea or mainstem bronchus. Twenty-one of 129 (16%) of CHD patients had extrinsic vascular compression. Representative examples are shown in Fig. 5. One patient had an additional finding of an irregular tracheal morphology due to slide tracheoplasty for tracheal stenosis in the setting of left pulmonary artery sling (Table 3). Besides CHD patients, the only other group of patients to have an airway finding were two vascular rings and two cardiomyopathy patients, which all had varying degrees of extrinsic airway compression. In 8/180 (4%), the extent of lower airway visibility was graded as 0 or 1 and considered limited to assess for variants or abnormalities beyond the trachea, and thus excluded from this analysis. The remaining 135/180 (75%) of children showed normal large airways.

Fig. 5.

Delayed 3D IR FLASH coronal- and axial-oblique reconstructions showing examples of lower airway anomalies: (a) 11-year-old patient with tracheal bronchus (arrow); (b) 3-year-old patient with left bronchopulmonary isomerism; and (c) 9-year-old patient with right aortic arch (RAA) with aberrant left subclavian artery (arrow), causing compression of the trachea (*). 3D IR FLASH three-dimensional inversion recovery-prepared fast low-angle shot, PA pulmonary artery.

Table 3.

Spectrum of lower airway variants and abnormalities (n = 180).

Variable	n (%)
Normal	135 (75%)
Tracheobronchial branching anomalies
Tracheal bronchus	5 (3%)
Tracheal diverticulum	1 (1%)
Abnormal situs
Bronchopulmonary situs inversus	1 (1%)
Right bronchopulmonary isomerism	3 (2%)
Left bronchopulmonary isomerism	2 (1%)
Extrinsic vascular compression of the trachea and/or mainstem bronchus*	25 (14%)
Artifact with limited visibility beyond trachea	8 (4%)

Data are presented as number (%).

^*Twenty-one with underlying CHD, two with vascular rings, and two with cardiomyopathy

4. Discussion

In this study, the airway of patients who underwent a delayed 3D IR FLASH sequence as part of their clinical CMR was assessed. The major findings of the study are (1) delayed 3D IR FLASH was able to characterize the tracheobronchial tree to the segmental level in 76% (137/180) and lobar level in 96% (172/180) of patients, and (2) up to 26% (33/129) of patients with CHD had a lower airway finding during routine CMR.

The principle of LGE imaging is based on the increased concentration and delayed washout of extracellular gadolinium-based contrast agent in areas of myocardial disease, typically representing areas of myocardial fibrosis, relative to normal healthy myocardium [13]. Gadolinium-based contrast agent will also accumulate in other areas of pathologic fibrosis, such as in the liver [14], along with normal connective tissues that have relatively high collagen content, such as the cardiac valves and vascular adventitia. Accordingly, contrast also accumulates in the adventitia of the tracheobronchial tree, which allows for anatomic characterization of the lower airways. However, even when airway wall enhancement is suboptimal, the airway can be well-profiled against the intermediate signal of the mediastinum during delayed post-contrast imaging [4].

Our study showed that delayed 3D IR FLASH images show excellent image quality and visualization of the trachea, main bronchi, lobar, and segmental bronchi. It performs particularly well in patients under GA as all GA patients had good or excellent image quality ratings, with 90% (46/51) providing complete airway visualization to the segmental bronchial level. The major limitation of this technique is liability to motion artifact in poorly cooperative patients because of the relatively long image acquisition time. In our experience, this sequence is therefore suboptimal when scanning infants using a feed-and-sleep technique given less consistent respiratory navigation and potential minor intermittent motion; no patients performed feed-and-sleep in this study for formal analysis. When it is performed in cooperative patients or under GA, the 3D IR FLASH sequence provides excellent images with no motion artifact, high spatial resolution, and high signal-to-noise ratio, while enabling multiplanar reconstructions in any plane (including minimum-intensity projections). In our experience, 3D modeling using these datasets is also feasible using third-party segmentation software; however, conventional 3D volume-rendered reconstructions using standard vendor-provided protocols are challenging to create. Of note, inadequate assessment of the lobar bronchi was seen in only 4% (8/180) of patients and was related to excessive motion artifact or device-related susceptibility artifact. Delayed 3D IR FLASH is therefore suitable as a primary sequence for complete lower airway assessment and can simultaneously provide information about pathological fibrosis in other areas of the body, as we have previously shown with myocardial LGE and liver fibrosis [10], [14].

This study also confirms that airway abnormalities are not infrequently found in patients with CHD, seen in up to 26% (33/129) of CHD patients in this cohort, with 16% (21/129) representing extrinsic airway compression and 9% (12/129) representing anatomic variants. This is a higher incidence than reported in studies of CHD patients using medical record coding that have shown congenital airway abnormalities in 3%–4% of patients [15] and combined congenital and acquired lower airway abnormalities in approximately 10% (460/4797) of patients [16]. However, this is expected as many of the findings in our study were variants that would not qualify as a congenital abnormality or were mild stenoses that may not have had a significant clinical impact. Conversely, our incidence is less than as seen in other imaging studies, for example, a study by Hou et al. showed tracheobronchial stenosis in 41% (31/75) [17], but was comprised of a highly selected population that included a high proportion of vascular rings (which we classified separately). Our cohort on the other hand represents consecutive patients undergoing routine CMR that had late gadolinium enhancement imaging. The exact clinical significance of lower airway abnormalities seen in our cohort is beyond the scope of this study.

At our institution, we now routinely use the delayed 3D IR FLASH for airway assessment if appropriately indicated. The 15- to 20-minute delay is a product of the gadobutrol dose we use and our institutional protocols where we often acquire other data immediately after contrast injection, such as angiography, volumetry, post-contrast T1 mapping, and flows. It may be possible in future work to refine the delay time specifically for airway assessment if these other sequences are unnecessary for the clinical indication. However, we typically perform CMR precisely for this added cardiovascular information. When the indication is strictly for anatomical assessment, such as for vascular rings, our preference is to use cardiac CT given the low-dose capabilities of modern CT scanners. As infant airway imaging with this technique is unreliable in our experience without GA, we similarly prefer CT for primary infant airway imaging to avoid GA and to preserve access to CMR for other patients. When CMR is specifically requested by the patient or ordering physician, however, we will use the delayed 3D IR FLASH sequence for airway assessment as an institutional preference. While other MRI sequences specifically designed for airway assessment are available, we were unable to perform comparative analysis due to the retrospective nature of this study. Ultimately then, whether the delayed 3D IR FLASH sequence should be incorporated into protocols specifically for airway assessment will depend on local institutional practices and resources.

5. Limitations

Limitations of this study include the single-center retrospective design, mixed clinical cohort, and lack of reference standard images for comparison. Nonetheless, we feel the sample size is sufficient to show the diagnostic capability of delayed 3D IR FLASH for lower airway assessment. The image quality and non-diagnostic rate of the sequence are comparable to previous reports [10].

6. Conclusions

Delayed respiratory-navigated electrocardiographically-gated 3D IR FLASH has excellent performance for evaluation of the entire lower airway and is suitable as a primary airway imaging modality, while also allowing for simultaneous assessment of myocardial LGE. Lower airway abnormalities are not infrequently encountered during routine CMR in patients with CHD.

Author contributions

Mike Seed: Writing—review and editing. Afsaneh Amirabadi: Writing—review and editing, Formal analysis. Christopher Z. Lam: Writing—review and editing, Writing—original draft, Visualization, Validation, Supervision, Project administration, Investigation, Formal analysis, Conceptualization. Shi-Joon Yoo: Writing—review and editing, Visualization, Validation, Methodology, Conceptualization. Laura Acosta Izquierdo: Writing—review and editing, Writing—original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Ankavipar Saprunruang: Writing—review and editing, Resources, Formal analysis, Data curation. Romina Dsouza: Writing—review and editing, Methodology, Investigation.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

Supplementary material

.