Pediatric Radiology

, Volume 37, Issue 1, pp 33–40

Utilization of low-dose multidetector CT and virtual bronchoscopy in children with suspected foreign body aspiration

Authors

    • Department of Radiology, Cerrahpasa Medical FacultyIstanbul University
  • Sebuh Kurugoglu
    • Department of Radiology, Cerrahpasa Medical FacultyIstanbul University
  • Sila Ulus
    • Department of Radiology, Cerrahpasa Medical FacultyIstanbul University
  • Harun Ozer
    • Department of Radiology, Cerrahpasa Medical FacultyIstanbul University
  • Mehmet Elicevik
    • Department of Pediatric Surgery, Cerrahpasa Medical FacultyIstanbul University
  • Fatih Kantarci
    • Department of Radiology, Cerrahpasa Medical FacultyIstanbul University
  • Ismail Mihmanli
    • Department of Radiology, Cerrahpasa Medical FacultyIstanbul University
  • Canan Akman
    • Department of Radiology, Cerrahpasa Medical FacultyIstanbul University
Original Article

DOI: 10.1007/s00247-006-0331-y

Cite this article as:
Adaletli, I., Kurugoglu, S., Ulus, S. et al. Pediatr Radiol (2007) 37: 33. doi:10.1007/s00247-006-0331-y

Abstract

Background

Foreign body aspiration is common in children, especially those under 3 years of age. Chest radiography and CT are the main imaging modalities for the evaluation of these children. Management of children with suspected foreign body aspiration (SFBA) mainly depends on radiological findings.

Objective

To investigate the potential use of low-dose multidetector CT (MDCT) and virtual bronchoscopy (VB) in the evaluation and management of SFBA in children.

Materials and methods

Included in the study were 37 children (17 girls, 20 boys; age 4 months to 10 years, mean 32 months) with SFBA. Chest radiographs were obtained prior to MDCT in all patients. MDCT was performed using a low-dose technique. VB images were obtained in the same session. Conventional bronchoscopy (CB) was performed within 24 h on patients in whom an obstructive abnormality had been found by MDCT and VB.

Results

Obstructive pathology was found in 16 (43.25%) of the 37 patients using MDCT and VB. In 13 of these patients, foreign bodies were detected and removed via CB. The foreign bodies were located in the right main bronchus (n = 5), in the bronchus intermedius (n = 6), in the medial segment of the middle lobe bronchus (n = 1), and in the left main bronchus (n = 1). In the remaining three patients, the diagnosis was false-positive for an obstructive pathology by MDCT and VB; the final diagnoses were secretions (n = 2) and schwannoma (n = 1), as demonstrated by CB. In 21 patients in whom no obstructive pathology was detected by MDCT and VB, CB was not performed. These patients were followed for 5–20 months without any recurrent obstructive symptomatology.

Conclusions

Low-dose MDCT and VB are non-invasive radiological modalities that can be used easily in the investigation of SFBA in children. MDCT and VB provide the exact location of the obstructive pathology prior to CB. If obstructive pathology is depicted with MDCT and VB, CB should be performed either for confirmation of the diagnosis or for the diagnosis of an alternative cause for the obstruction. In cases where no obstructive pathology is detected by MDCT and VB, CB may not be clinically useful.

Keywords

Foreign bodyAspirationLow-doseMultidetector CTVirtual bronchoscopyChildren

Introduction

Because children like to put objects and pieces of food in their mouth and nose, foreign body aspiration is a common problem in children, especially among those younger than 3 years, with high morbidity and mortality rates [16]. Early diagnosis is essential to prevent complications, irreversible lung damage and even death. Physical findings and clinical history give clues to establishing the diagnosis. Chest radiography is generally the first radiological modality used, although the findings are often non-specific and insufficient. The most common radiographic findings are air trapping, consolidation, atelectasis and over-aeration [2, 5]. For accurate diagnosis and management of foreign body aspiration, conventional bronchoscopy (CB) is the accepted gold standard. However, CB is an invasive procedure, requiring general anaesthesia and occasionally resulting in serious complications in children [5, 7]. The critical question is whether CB evaluation is necessary for all patients with suspected foreign body aspiration (SFBA). We describe here the use of low-dose multidetector CT (MDCT) and virtual bronchoscopy (VB) for the evaluation of SFBA in children. The findings from low-dose MDCT and VB are compared with those from CB in order to evaluate the contribution of these methods to management.

Materials and methods

From November 2004 to July 2006, 37 children (17 girls, 20 boys) with a history of acute and/or chronic SFBA who had symptoms of coughing, choking, acute dyspnoea, wheezing, and/or recurrent respiratory tract infections were included in the study. The mean age was 32 months (range 4 months to 10 years). Chest radiographs were initially obtained for each patient. MDCT examinations were performed using a 16-detector CT scanner (Cardiac Sensation, Siemens, Germany). Scanning extended from the level of the larynx to the dome of the diaphragm. Children who could hold their breath were told to take a deep breath before scanning commenced. In all of the examinations intravenous non-ionic contrast material (1.5 ml/kg) was administered via the antecubital vein using a pump. Contrast medium was utilized because the differential diagnosis prior to scanning included mediastinal lymphadenopathy due to tuberculosis. Venous cannulation was undertaken outside the Radiology Department and small children were allowed to settle or fall asleep. Anaesthesia or sedation was not necessary in any of the patients. All MDCT studies except three were free of motion artefact. However, motion artefacts did not affect image interpretation. Scanning parameters were 16 × 0.75-mm detector collimation, 80 kVp, 0.5 s gantry rotation and 30 mA. The scanning duration was 5–7 s (mean 6.81 s). The pitch was automatically adjusted to 1.

MDCT images were reconstructed with a 0.75-mm slice thickness and 0.4-mm reconstruction intervals on the workstation (Wizard, Siemens, Germany). These reconstructed axial images were processed and evaluated by two experienced paediatric radiologists. On the 3D page, the reconstructed multiplanar reformatting (MPR) images were evaluated in the axial, sagittal and coronal planes. VB was accomplished in each patient by the “fly through” technique. The final interpretation was made by consensus between the two radiologists who evaluated all of the axial, 2D, 3D and VB images in combination.

CB was performed under general anaesthesia within 24 h of an obstructive abnormality being found by MDCT (n = 16). CB was carried out using a rigid paediatric bronchoscope (Karl Storz, Tuttlingen, Germany) by experienced paediatric surgeons who were aware of the MDCT and VB findings. The tracheobronchial tree was evaluated for the presence of foreign bodies, anatomical abnormalities, inflammation, secretions, and granulation tissue. Detected foreign bodies were removed. Bronchial lavage was obtained for microbiological examination. Patients were hospitalized for one night. No complications relating to CB occurred. Patients were discharged without any symptoms and follow-up radiographs were not obtained.

Patients whose MDCT, VB and CB findings concurred were classified as true positives, while those with positive MDCT and VB findings, but negative CB findings were classified as false positives. Children with negative MDCT and VB findings were followed clinically, but CB was not performed. The mean follow-up period for these patients was 5–20 months. Institutional ethics committee approval was obtained.

Results

In 16 of 37 patients, an obstructive lesion in the bronchial lumen was detected by MDCT and VB. Chest radiographs were positive in 8 of these 16 patients. On chest radiographs, both atelectasis and ipsilateral hyperaeration (n = 2), atelectasis (n = 2), and hyperaeration (n = 4) were detected. In the remaining 29 patients no pathological radiological findings were detected on the chest radiograph.

CB demonstrated that 13 of these 16 obstructive pathologies were due to foreign bodies (82% true positives). Except in three patients (a bone fragment, a piece of grape and a plastic tube-like object), the foreign bodies that were found and removed by CB were all various types of nut fragment. The sizes of the foreign bodies varied between 3 mm and 12 mm. On MDCT and VB the foreign bodies were located in the right main bronchus (n = 5, Fig. 1), bronchus intermedius (n = 6, Fig. 2), medial segment of the middle lobe bronchus (n = 1), and left main bronchus (n = 1). The site of the foreign bodies identified on MDCT and VB fully corresponded to the CB findings. On MDCT the radiological findings, apart from the foreign body itself, were hyperaeration (n = 4), atelectasis (n = 2), and atelectasis together with hyperaeration (n = 2) in the lobe or segments distal to the obstructed bronchus. All foreign bodies were successfully removed and symptoms and signs resolved completely with no recurrence.
https://static-content.springer.com/image/art%3A10.1007%2Fs00247-006-0331-y/MediaObjects/247_2006_331_Fig1_HTML.jpg
Fig. 1

Patient 11. A 1.5-year-old girl with sudden onset of dyspnoea and coughing. Axial (a) and coronal (b) MDCT images (parenchymal window) show an obstructive lesion in the right main bronchus (arrows) and hyperaeration in the right lung. c VB image reveals almost complete obstruction of the right main bronchus (arrow). A nut fragment was removed by CB

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-006-0331-y/MediaObjects/247_2006_331_Fig2_HTML.jpg
Fig. 2

Patient 12. A 10-year-old boy with fever, cough and dyspnoea for 1 month. Axial (a) and coronal (b) MDCT (parenchymal window) images, and VB image (c), show a tube-like foreign body causing obstruction in the right bronchus intermedius (arrow). d Coronal surface-rendered image reveals a tube-like defect in the bronchial tree extending from the bronchus intermedius to the lower lobe bronchus causing partial obstruction of the middle lobe segment bronchus (arrow). A plastic tube was removed by CB

The remaining three obstructive pathologies thought to be foreign bodies on MDCT and VB were actually due to purulent secretions (n = 2, Fig. 3) and an intrabronchial soft-tissue mass (n = 1, Fig. 4) on CB (19% false positives). All three pathologies were located in the left main bronchus. Aspirated secretions were sent for microbiological examination during the procedure. The soft-tissue mass was biopsied endoscopically and shown to be a schwannoma; left pneumonectomy was performed. The clinical features of the obstructive cases and localization of the obstructive lesions are given in Table 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs00247-006-0331-y/MediaObjects/247_2006_331_Fig3_HTML.jpg
Fig. 3

Patient 15. A 15-month-old girl with acute dyspnoea. a Axial MDCT image shows complete occlusion of the left main bronchus and atelectasis of the left lung. Note the compensatory hyperaeration in the right lung. b VB image shows a foreign body causing complete obstruction of the left main bronchus. On CB no foreign body was detected. The obstruction was due to mucoid pus-like material

https://static-content.springer.com/image/art%3A10.1007%2Fs00247-006-0331-y/MediaObjects/247_2006_331_Fig4_HTML.jpg
Fig. 4

Patient 16. A 6-year-old girl with dyspnoea and fever for 3 months. Axial (parenchymal window) (a) and coronal (mediastinal window) (b) MDCT images show a soft-tissue density lesion in the left main bronchus (arrows) and non-homogeneous aeration in the left lung (a). c VB image shows a lobulated lesion causing complete obstruction of the left main bronchus (arrows). CB showed no foreign body, but a soft-tissue mass shown histologically to be a schwannoma

Table 1

The clinical features and MDCT, VB and CB findings in patients with obstruction

Patient no.

Gender

Age (months)

Symptom duration

Symptoms

Obstruction location on MDCT

Foreign body on conventional bronchoscopy

Obstruction

Dyspnoea

Coughing

Wheeze

Fever

1

M

48

3 months

+

+

Right main bronchus

+

Nut

2

F

30

15 days

+

+

Right main bronchus

+

Nut

3

F

36

1 day

+

Right main bronchus

+

Nut

4

M

21

15 days

+

Right bronchus intermedius

+

Nut

5

M

36

1 day

+

Right bronchus intermedius

+

Nut

6

M

24

15 days

+

Right middle lobe bronchus

+

Nut

7

M

12

1.5 months

+

Right bronchus intermedius

+

Nut

8

M

18

1 month

+

+

Right bronchus intermedius

+

Grape

9

M

24

1.5 months

+

+

+

Left main bronchus

+

Nut

10

F

18

1 day

+

Right bronchus intermedius

+

Bone

11

F

18

7 days

+

+

Right main bronchus

+

Nut

12

M

120

1 month

+

+

+

Right bronchus intermedius

+

Plastic tube

13

M

24

1 month

+

+

+

Right main bronchus

+

Nut

14

F

6

3 days

+

Left main bronchus

Pus

15

F

15

1 day

+

Left main bronchus

Pus

16

M

72

3 months

+

+

Left main bronchus

Schwannoma

In 21 of 37 patients (57%) there was no intrabronchial pathology indicating obstruction on MDCT or VB. In four of these patients MDCT demonstrated consolidation (n = 2), atelectasis (n = 1) and subglottic stenosis (n = 1), and all were then treated accordingly. The remaining 17 patients were diagnosed with viral bronchitis/laryngotracheitis and/or reactive airway disease according to their physical examination and additional laboratory findings. These patients were not hospitalized. Follow-up physical examinations were performed within 1 week and at the end of 1 month. None had recurrent or new symptoms on the follow-up examinations. The true-negative rate could not be determined because CB was not performed in these patients.

Discussion

The first step in the diagnosis of foreign body aspiration is careful patient history and physical examination. The most common signs and symptoms of foreign body aspiration are cough, acute dyspnoea and sudden onset of wheeze. On physical examination the most common auscultation finding is an asymmetrical decrease in breath sounds and, less frequently, the presence of rales on the side of the obstructive pathology.

Radiological methods are the next step in the diagnostic work-up. Radiological findings depend on the size and location of the foreign body and duration of aspiration. The primary radiological modality in the evaluation of SFBA is chest radiography. On radiographs, non-specific findings such as air trapping, atelectasis, segmental or lobar collapse, consolidation, and hyperaeration can be detected [1, 5]. If the foreign body is radio-opaque it can be seen on the radiograph. However, 90% of aspirated foreign bodies are food particles and are radiolucent and cannot be detected by radiography [2]. Chest radiographs can be normal in approximately 30% of children with foreign body aspiration. In studies by Svedstrom et al. [2], Haliloglu et al. [8], and Kosucu et al. [9] sensitivity and specificity of chest radiography for foreign body aspiration were 68% and 67%, 85% and 68%, and 60% and 32%, respectively. To improve the diagnostic accuracy of chest radiography, inspiratory and expiratory chest radiographs, lateral decubitus radiographs, and fluoroscopy may be used.

MDCT is an advanced imaging tool used in SFBA [914]. It can identify a foreign body in the tracheobronchial tree and the secondary parenchymal changes. The most important advantage of MDCT over single-slice CT is the increase in scanning speed. This is especially important in small children in whom anaesthesia and sedation may pose a problem and in whom foreign body aspiration is frequent. Improvements in MDCT technology have resulted in a decrease in slice thickness [15]. Currently the minimum slice thickness in MDCT devices is 0.5–0.62 mm. This provides better spatial resolution for multiplanar reformatting and 3D reconstructions. In addition, MDCT gives the option of VB, which is a non-invasive modality that enables a 3D view of the internal surface of the tracheobronchial tree [918]. VB is superior to CB in showing complex and tortuous structures and in allowing visualization of the airways distal to an obstruction [1618]. In addition to the major bronchi, the segmental and subsegmental bronchi are easily evaluated with MDCT and VB. VB images are visually pleasing and may assist the paediatric surgeon during CB, but because they are derived from the 3D MDCT data they do not provide additional diagnostic information. The limitations of VB are its inability to assess mucosal morphology, vascularity and colour; secretions and artefacts may result in false-positive findings [19]. Tracheomalacia, a dynamic pathology, cannot be depicted. A foreign body that results in incomplete obstruction may present as complete obstruction on a 3D VB image.

The main disadvantage of MDCT is radiation exposure. CT protocols are associated with a known increase in the risk of future malignancy [20, 21] and justification is essential. Measures to reduce radiation dose include reduction of tube current (mA), but this increases image noise. In our study we used 30 mA and 80 kVp, the lowest parameters currently used to provide optimal quality images without any significant loss of diagnostic data.

CB in children is performed with either a rigid or a flexible bronchoscope. Complications are rare when a flexible bronchoscope is used. As the airways in children are small, it is difficult to enter the segmental bronchi, and almost impossible to enter the subsegmental bronchi with CB. In addition, upper lobe bronchi that arise at an acute angle from the major bronchi are problematic and require a high level of expertise to evaluate. To overcome these technical problems of CB, MDCT and VB can be used to determine the presence of foreign bodies and, if present, the exact location and size.

Patients with no obstructive or mediastinal/parenchymal pathology on MDCT and VB were followed clinically. None of these patients had recurrent or new symptoms on follow-up examinations. Thus despite the SFBA, CB was not performed and the risk of complications, hospitalization and anaesthesia were eliminated. A further group of patients in whom CB was avoided were those with findings such as atelectasis, consolidation and subglottic stenosis, which explain the patients’ symptoms.

In our study, we had three patients with a false-positive result. Purulent secretions were the underlying cause of obstruction in two and a protruding mass (schwannoma) in the third. It is important to distinguish intrabronchial secretions from foreign body. To this end we evaluated the densities of all the obstructive lesions. However, we did not observe any significant density difference between foreign bodies and secretions, and this accords with the findings of previous studies [8, 9, 19]. Thus, when intrabronchial obstructive pathology is detected by CT, secretions, granuloma or soft-tissue mass should be considered in the differential diagnosis. This inability to characterize the cause of the obstruction seems to be the weak point of MDCT and VB; however, considering the clinical consequences, the cause of the obstruction is less important than its demonstration and subsequent management.

In two previous studies, the use of MDCT and VB in SFBA in children has been evaluated [8, 9]. Haliloglu et al. [8] included 23 children in their study. They had fewer patients with a true-positive result and utilized 100 mA and 120 kVp for their scans. In the study by Kosucu et al. [9], the CT examinations were redundant in two patients who had radio-opaque foreign bodies visible on chest radiographs. Due to technical limitations, the segmental and subsegmental bronchi were not investigated in either study. In our study, we were able to evaluate these territories and we had a single patient in whom the foreign body was located in the medial segment of the right middle lobe bronchus and which was removed successfully.

There are a few limitations to our study. First, not all patients underwent CB. This group of 21 patients had no foreign body detected by MDCT and VB. On follow-up, none presented with recurrent symptoms. We did not perform follow-up radiographs or ventilation scintigraphy; follow-up was by clinical examination only. McHugh [22] addressed the important question: “Is CT reliable enough, when the likelihood of foreign body aspiration is low such that bronchoscopy can be safely omitted?” Our answer to this question is “yes” because one of the findings of our study was that CB was unnecessary when MDCT/VB showed that there was no foreign body; these patients did well on clinical follow-up. Second, the paediatric surgeons who performed CB were aware of the MDCT/VB findings, which may have resulted in possible bias. Our aim in this study was not to compare MDCT/VB findings with CB, but assist in the management of children with SFBA. Thus pinpointing the exact location of the foreign body greatly facilitates the task of the surgeon and surely reduces the possible complications. All the foreign bodies except one were located in the central airways rather than subsegmental bronchi. Because of the small size of the airways in children, it should be emphasized that foreign bodies cannot usually travel beyond the central airways. The small foreign bodies that reach the subsegmental bronchi usually cause post-obstructive pneumonia and chronic and sustained symptoms.

Conclusion

Low-dose MDCT and VB are non-invasive radiological modalities that can be used easily in the investigation of SFBA in children. Identification of the exact location greatly contributes to the proper guiding of CB and so morbidity rates are reduced. Besides obstructive pathological findings, pneumonia or atelectasis, which explain the clinical condition of the patient, are also readily detected by MDCT, prompting appropriate treatment and preventing unnecessary CB. It should be noted that MDCT should be performed with low-dose settings adhering to the ALARA principle. If obstructive pathology is depicted with MDCT and VB, CB should be performed, both for confirmation of the diagnosis and evaluation of the differential diagnosis. In cases where no obstructive pathology is detected by MDCT and VB, CB may not be clinically useful.

Copyright information

© Springer-Verlag 2006