European Archives of Oto-Rhino-Laryngology

, Volume 270, Issue 3, pp 1001–1007

3D images based on MDCT in evaluation of patients with suspected foreign body aspiration

Authors

  • Busheng Tong
    • Department of Otorhinolaryngology-Head and Neck SurgeryEENT hospital of Fudan Universtiy
  • Lichun Zhang
    • Department of Otorhinolaryngology-Head and Neck SurgeryEENT hospital of Fudan Universtiy
  • Rui Fang
    • Department of Otorhinolaryngology-Head and Neck SurgeryEENT hospital of Fudan Universtiy
  • Yan Sha
    • Department of Radiologythe EENT Hospital of Fudan University
    • Department of Otorhinolaryngology-Head and Neck SurgeryEENT hospital of Fudan Universtiy
Laryngology

DOI: 10.1007/s00405-012-2279-x

Cite this article as:
Tong, B., Zhang, L., Fang, R. et al. Eur Arch Otorhinolaryngol (2013) 270: 1001. doi:10.1007/s00405-012-2279-x

Abstract

To evaluate the value of 3D images based on multi-detect computer tomography (MDCT) for the diagnosis and management of patients with suspected foreign bodies aspiration, and report our experience about diagnosis and management of the aforementioned patients using the 3D images, as well as detail our protocol for the management of those patients. Forty-four patients (37 children, 7 adults) with suspected foreign bodies aspiration who accepted pre-surgical examination and bronchoscopy in our hospital were included in this study. All the patients’ pre-surgical 3D images based on MDCT were reconstructed and analyzed. After that all the results were compared with observations in the surgeries. Among the 37 pediatrics, 34 patients were detected with FB in their tracheobronchial system by the 3D images based on MDCT, and 3 cases were detected negative. The sensitivity of 3D images is 100 %, and the specificity is 75 %.In the adult group, all the 7 cases of FB in the tracheobronchial tree were detected by 3D images and proved by bronchoscopy. 3D images based on MDCT were proved to be a valuable method for the diagnosis of the patients with suspected FB aspiration. In the diagnosis and management of those patients, we considered that both the typical FB aspiration history and the intractable symptoms of cough should be taken as the criteria and the first step for selecting those suspected patients; then 3D images based on MDCT could be taken as the second step and as the selective criteria for those patients who should accept the rigid bronchoscopy.

Keywords

Multi-detect computer tomographyVirtual bronchoscopyVirtual renderingMultiplanar reformationForeign body aspirationRigid bronchoscopy

Introduction

Aspiration of foreign body (FB) into the tracheobronchial tree is an emergency situation in the area of otolaryngology, which occurs not only in children but also in adults. Pediatrics inappropriately performed urgent FB extraction surgery would cause severe complications, such as intractable pneumonia and hemoptysis, and increase the risk of death [15]. Whether the surgery would be a success majorly depends on the pre-surgical diagnosis and the exact evaluation of the situation about the FB, e.g., location, size, and shape. Although the conventional radiography demonstrated some values for the patients, it mainly focuses on the radiopaque FB. For those radiolucent FB, chest radiograph demonstrate little values. Previous researchers demonstrated that the diagnosis for those patients with suspected FB aspiration could be induced by observing the indirect radiological findings including air trapping, atelectasis, consolidation, and bilateral over-aeration [6], even in those patients with aspirated FBs that were radiolucent. However, there are still some risks of misdiagnosis. Those patients who have tracheal stenosis and spasm could also be detected with similar findings in their chest radiography.

Many previous researches demonstrated that 3D images, including multiplanar reformation (MPR), volume render (VR) and virtual bronchoscopy based on multi-detect computer tomography (MDCT) have some value in the diagnosis and differential diagnosis of suspected FB aspiration [7, 8]. In this study, we re-evaluated the value of those aforementioned methods for these kinds of patients and report our 2 years’ experience about diagnosis and management for those patients by use of the 3D images based on MDCT, as well as detail our protocol for the diagnosis and management of suspected FB aspiration.

Materials and methods

Patients

The study was approved by the Institutional Ethics Committee, and written informed consent of the parents was obtained. From February 2010 to February 2012, 44 patients with suspected FB aspiration who accepted pre-surgical examination and rigid bronchoscopy surgery were included in this study to the Eye, Ear, Nose, and Throat (EENT) Hospital, Fudan University, Shanghai, China. Among them, 37 cases are children aged from 6 to 96 months with the medium of 28.7 months, and 7 patients are adults aged from 33 to 69 years with the medium of 50.6 years. All the children patients, 21 boys and 16 girls, had a course of disease ranged from 4 h to 8 months with the medium of 32.5 days. And 29 of the 37 children had clear history of FB aspiration. For those 7 adult patients, 3 men and 4 women had a course of disease ranged from 2 h to 1 month. All of them had clear history of FB aspiration.

Examination

Before the bronchoscopy, all the patients were examined using a Siemens Sensation 16 CT (Forchheim, Germny). Transverse CT scans were acquired in the routing thoracic mode with 2 mm collimation, 1.5 pitch, 120 kV, 100 mA, and a gantry rotation speed of 3,600/s from the level of thoracic inlet to the base of the lung. In order to acquire a clear image, all the 7 adult patients’ examinations were performed in a single breath-hold period. However, the scans for all those 37 children cases who were not able to hold their breath were conducted in spontaneous breathing. The images have a serial 1 mm thickness images with a 1-mm intersection gap, a field of the view 180 × 180 mm, and a 512 × 512 acquisition matrix. None of the children were sedated because of a possible complication of sedation causing respiratory distress (small children were allowed to settle or fall asleep). No intravenous contrast agent was used. All axial images were evaluated with standard lung (window 1,400, level −450) and soft-tissue (window 300, level 50) settings for the presence or absence of FB. After that, all images were imported into the 3D reconstructive type in the vendor workstation (Wizard; Siemens, Forchheim, Germany) after examination to reconstruct the 3D images. All MDCT studies except five were free of motion artifacts. However, motion artifacts did not affect image interpretation. [19] Three different techniques, MPR, VR and virtual bronchoscopy were used in this study for the observation of FB. For the MPR images, the reference image was created of the target structure suspected FB. In order to reconstruct the MPR image, one of the respective transverse, coronal, and sagittal sections was first chosen as the standard section, and the other two sections were adjusted. For the VR ones, the predefined algorithm for the demonstrating tracheobronchial tree was used. For the virtual bronchoscopy images, navigation through the tracheobronchial tree was performed interactively by the workstation monitor and a computer mouse until the FB was found. All those images were reconstructed and analyzed by two otorhinolaryngologists, one of which has more than 3 years experience on the radiological diagnosis of otorhinolaryngology. If any one of the images reconstructed by the three different techniques found FB, the patients would be defined positive. Only if all of the three different 3D images do not find any FB, the patient would be defined negative. Among all those patients, 32 children and 4 adults had accepted the chest fluoroscopy examination before the CT examination in our hospital.

Surgical procedure

Bronchoscopy surgeries for all of the 44 patients were performed in the operating room under general anesthesia by a rigid bronchoscope system with optical telescopes (Olympus Optical Company, Tokyo, Japan). The surgeons who practice the bronchoscopy were aware of the CT findings. During the surgery, the presence of FB, anatomic abnormalities, inflammation, purulent secretions, and granulation tissue were carefully detected. Once the FB was visualized, removal was most often accomplished with alligator jaw and cup forceps. Afterward, the tracheobronchial cavity was examined carefully again to make sure there were no fragments retained. The results of rigid bronchoscopy were compared with findings of 3D images.

Results

Children group

All the 37 children patients’ clinical data and 3D images results are summarized in Table 1.
Table 1

The clinical data and the results of 3D images bronchoscopy in 37 children cases

 

History of FB aspiration

Symptom

Signs of lung

Chest fluoroscopy

Positive

Negtive

Cough

Fever

Vomiting

Choking

Negtive

Positive

Negtive

Positive

Blank

  

29

8

37

11

7

3

1

36

12

20

5

3D images and virtual bronchoscopy in finding FBs

Positive

26

8

32

10

6

2

1

33

11

19

4

Negtive

3

0

5

1

1

1

0

3

1

1

1

In the children group, 33 of them were proved having FB, and 4 of them were proved negative in their tracheobronchial tree by rigid bronchoscopy. 34 of them were detected positive by the 3D image based on MDCT, and 3 of them were detected negative. For the 32 children patients who have accepted the fluoroscopy examination, 20 of them were diagnosed having FB, and the other 12 were diagnosed as negative. Among the 20 positive patients, 1 of them was proved negative by the rigid bronchoscopy. Among the 12 negative patients, 11 of them were proved positive by the rigid bronchoscopy.

Among the 34 positive patients, we found 17 FBs located in the right bronchus, 9 in the left bronchus, and 8 in the trachea on the 3D images based on MDCT (Fig. 1). In the subsequent rigid bronchoscopy, 31 cases showed the same results except 2 patients whose pre-surgical 3D images demonstrated FB located in the right bronchial, while actually located in the main trachea during the surgery, and the other patient which was detected in the trachea by MDCT was proved some puriform sputum in the trachea by rigid bronchoscopy. In addition, among the 34 positive cases, in 13 of them there occurred emphysema, and 2 of them presented atelectasis.
https://static-content.springer.com/image/art%3A10.1007%2Fs00405-012-2279-x/MediaObjects/405_2012_2279_Fig1_HTML.jpg
Fig. 1

A 5-year-old girl with aspiration history of melon seeds for 7 days was admitted in the hospital, whose aspired melon seed was found in the trachea below the glottis during the bronchoscopy surgery and was successfully retrieved. Coronal (a) and Sagittal (b) MPR images demonstrated a seed-like foreign body with high-density shell locating in the trachea below the glottis (arrows); virtual bronchoscopy (c) presented a foreign body locating in the trachea; and VR (d) images showed a foreign body locating in the trachea below the glottis, and the air shadow was replaced

Of the 33 proved patients, 32 cases’ FBs were removed successfully, the exception being a 6-month-old child. The nature of the removed FBs were melon seeds in 11 cases, peanut fragments in 8 cases, vegetables in 4 cases, unknown objects in 2 cases, as well as eel, pine nuts, egg shells, plastic tube, green beans, small lamp, and almonds in the other cases, respectively. For the other child whose stone was not removed, his parents requested discharge after our surgery. For the aforementioned successfully treated 32 cases, there were 3 patients who accepted the second surgery, of which, 2 cases’ FB located at the left main bronchus, and 1 cases’ FB lying in the right middle lobe bronchus. During their first bronchoscopy surgery, we could not remove the FBs due to the local swelling mucous around the FBs. After we used the antibiotics and waited for 3–6 days, we performed the second bronchoscopy and successfully removed the FBs, and the FBs were peanut fragments, plastic tube and small lamp, respectively.

Regarding the complications, for the 33 positive patients, there was only 1 case demonstrated pneumomediastinum and emphysema after the removal of the FB, who recovered after 1 week. Of the 4 negative patients, there was one 16-month-old boy who demonstrated cough, shortness of breath and vomiting after aspiration during a semi-liquid diet, there occurred postoperative extubation difficulty and he was transferred into the ICU ward. This child was eventually transferred to a pediatric hospital for treatment. After 2 weeks of therapy, he was rehabilitated and discharged with the diagnosis of asthma. Classification results with history, chest fluoroscopy and 3D images of FB aspiration in children are shown in Table 2.
Table 2

Classification results with history, chest fluoroscopy and 3D images of FB aspiration in children

 

Rigid bronchoscopy

      

+

Total

Sensitivity*

Specificity*

PPV*

NPV*

FP*

FN*

History

+

25

4

29

75.8

0

86.2

0

100

24.2

8

0

8

 

Overall correct classification rate = 67.6 %

 

Total

33

4

37

Chest fluoroscopy

+

19

1

20

63.3

50

95

8.3

50

36.7

11

1

12

 

Overall correct classification rate = 62.5 %

 

Total

30

2

32

3D images and virtual bronchoscopy

+

33

1

34

100

75

97.1

100

25

0

0

3

3

 

Overall correct classification rate = 97.3 %

 

Total

33

4

37

* Values for these items are percentages

PPV positive predictive value, NPV negative predictive value, FP false-positive, FN false-negative

Adult group

Of the 7 adult patients which all have a clear history of FB aspiration, all of them presented cough after aspiration, and 2 of them demonstrated the symptoms of sputum with bloodshot. All of the four patients who accepted the chest fluoroscopy examination were diagnosed as negative. On the 3D images based on the MDCT (Fig. 2), we found four cases located in the left bronchia, and three cases located in right bronchia, which also were proved by the bronchoscopy. All of the 7 patients’ FBs were removed successfully. The nature of the removed FBs were bone fragments in 3 cases, shrimp fragments in 2 cases, as well as dentures and a tablet in each case, respectively. None of the patients showed any complications after surgery and were discharged 2 days later.
https://static-content.springer.com/image/art%3A10.1007%2Fs00405-012-2279-x/MediaObjects/405_2012_2279_Fig2_HTML.jpg
Fig. 2

A 33-year-old woman with aspiration history of pock spareribs for 7 days was admitted, who presented cough and blood in the sputum. And there was no abnormal founding in the physical examination. A bone slice locating in the right intermedius bronchus was found and successfully retrieved during the rigid bronchoscopy surgery. MPR images (a, b) presented a high-density foreign body locating in her right intermedius bronchus, and without any complications (arrows); virtual bronchoscopy images (c) showed a slice-like foreign body locating in her right intermedius bronchus; and VR image (d) demonstrated a foreign body locating in the same location and the air shadow was replaced

Discussion

It is generally accepted that rigid bronchoscopy is often performed for definitive diagnosis and treatment for those patients suspected with FB aspiration [9, 10]. However, it is not the first choice for those patients, especially in pediatrics. Because although bronchoscopy is generally simple and safe in child patients, serious complications such as pneumothorax, pneumonia, respiratory distress, cardiac arrest, tracheal laceration, subglottic edema and death may occur, even when practiced by physicians with many years of experience[1012]. In addition, the correct identification of the exact location, form, and volume of the FB in all cases is beneficial for our performing rigid bronchoscopy, which can not only reduce the surgery time, but also decrease the occurrence rate of complications. Therefore, the pre-surgical diagnosis of FB is very important for the management of suspected FB aspiration.

The conventional diagnostic evaluation of suspected FB aspiration is based on a clear FB aspiration history plus typical clinical presentations (i.e., cough, dyspnea and so on) and abnormal physical examinations (i.e., decreased breath sounds, wheezing and stridor in one lung) [9]. Among these signs and symptoms, Jung SY et al. [13] considered that a choking event history is the most specific. And the typical history of FB aspiration should be considered as the primary selective criteria for including the suspect patients for further treatment. This may be accurate for the adult patients. In this study, all of the 7 adult cases had a clear FB aspiration history and the existence of the FB was proved by the surgery. However, for the pediatrics, the accuracy of FB aspiration history should be re-considered, because those patients who have no clear history of FB aspiration would be excluded and their therapy would be delayed. In our study, there were 8 children who did not have any history of FB aspiration but were proved as having FBs in their tracheobronchial system. In addition, among the 29 cases with clear FB aspiration history, there were 4 cases that were proved negative in the surgery. Analysis of those case histories, no matter whether they have the typical history of FB aspiration or not, we found that all of them presented repeated intractable cough. Therefore, we consider that the symptom of intractable cough, especially in the pediatrics. This should also be taken as one of the selective criteria of those patients, besides the typical FB aspiration history.

The thoracic X-ray examination, such as chest fluoroscopy examination was generally considered as the second step for diagnosis and management of those patients [6, 13, 14]. For those radiopaque FBs, there is no doubt that it can be diagnosed by this examination. However, more than 90 % of the aspirated FBs are radiolucent [15]. Some researchers considered that we can diagnose those patients by observation of the indirect characters, such as: mediastinal shifting, air trapping, atelectasis, segmental or lobar collapse, consolidation and hyperaeration [6, 13], although, these signs are not very specific. In this study, among the 7 adult patients, there were 4 cases of accepted chest fluoroscopy examinations before our surgeries, none of them discovered any abnormalities in the examination. In addition, among the 32 pediatrics who accepted chest fluoroscopy examination, there were 12 cases who were diagnosed as normal by this exam, while only one case was proved as normal by the surgery; and there were 20 cases who were diagnosed as positive, while there was still one case that was proved as normal by the surgery. Therefore, we considered that the chest X-ray examination, such as chest fluoroscopy examination, might not be the necessary step for the diagnosis and management of those patients. Because the chest X-ray examination could not only add radiation exposure, but also delay the diagnosis for those patients, this could result in severe complications, such as bronchiectasis, bronchial fistula, or even death [13, 16].

Compared with the conventional methods, the 3D images based on MDCT, including MPR, VR and virtual endoscopy, is a very sensitive and specific modality for the detecting of FBs, even for the radiolucent FBs, in the airway system [7, 8]. In addition, it can provide more accurate information about the FBs, such as location, shape, occurring complications, and so on [7, 8, 13], which are very beneficial for the planning of surgery. With all of this detailed information, it not only reduce the time for the operation, but also decrease the occurrence rate of complications. Previous researchers reported that the 3D images based on MDCT could increase the sensitivity (100 %) and specificity (81–100 %) of the diagnosis [17, 18]. A possible disadvantage of the use of VB is the false-positive examination results. The presence of purulent endobronchial secretion or protrusion of some tumor mass towards the bronchial lumen must be ruled out [17]. All the patients were allowed to achieve the CT scanning under their physical condition in this retrospective study, and we found that the sensitivity of 3D images based on MDCT for the diagnosis of FBs aspiration is 100 % and the specificity is 75 %. Therefore, we consider that the 3D images based on MDCT might be taken as the ascertaining diagnosis method for suspected FB. Nevertheless, those patients with severe, life-threatening cardio-respiratory symptoms should be rescued firstly. The principal disadvantage of this procedure is the exposure to radiation, which can be minimized by avoiding the excessive use of tomography scans and reducing the tube current (in mA) to the minimum possible [19].

Conclusion

In conclusion, 3D images based on MDCT showed significant value for the diagnosis of FB aspiration, whose overall correct classification rate (97.3 %) might reach the level of rigid bronchoscopy. Therefore, during diagnosis and management for these patients with FB aspiration, both typical FB aspiration history and the intractable symptom of repeated intractable cough should be considered the criteria and the first step for selecting those suspected patients. The 3D images based on MDCT could be taken as the second step, which would also be considered as the selective criteria for those patients who should accept rigid bronchoscopy. Also, for those who are negative on the 3D images based on MDCT, the performance of rigid bronchoscopy should not be considered temporarily. However, the role of 3D images based on MDCT cannot replace bronchoscopy and would not become daily routine.

Acknowledgments

This study was supported by the National Natural Science Foundation of China, project NSFC Grant 81028003/H1305, NSFC Grant 30973303.

Conflict of interest

None declared.

Copyright information

© Springer-Verlag Berlin Heidelberg 2012