Endoscopic technique
Before the operation, high-resolution CT data (slice thickness 1 mm,
spacing 0.6 mm) was imported into the navigation system (Archimedes VBN,
Broncus Medical, USA) for the automatic reconstruction of 3D bronchial
and vascular models. Sub-branches of the right lower lateral basal
segment and posterior basal segment (which have fewer peripheral
vessels, lower risk of bleeding, and avoid larger blood vessels) were
chosen as the target positions. After determining the relationship
between the biopsy target site and the peripheral vessels (as well as
distance from the pleura), automatic calculation of the route and
distance of the bronchoscope to the biopsy site was carried out. The
operation was performed using rigid bronchoscopy sheath ventilation
(STORZ, USA, No.4.5, outer diameter 7.3 mm, inner diameter 6.6 mm) under
general intravenous anesthesia. A flexible bronchoscope was inserted
through a rigid bronchoscopy sheath. Routine bronchoscopy (outer
diameter 4.0 mm, orifice 2.0 mm; Olympus, Japan) was performed to
exclude mucosal lesions, and bronchoalveolar lavage was performed in the
middle lobe of the right lung. Then, under the guidance of the flexible
bronchoscope, an occlusion balloon (6×30 mm, , Boston Scientific, USA)
was placed in front of the opening of the right lower lobe (Figure 2 B),
and gas was injected to inflate the balloon via the flexible
bronchoscope. When the injected volume was sufficient to cause complete
occlusion of the opening of the target bronchial lobe, the required gas
volume was recorded, and the air tightness of the balloon was tested.
The balloon was then deflated for standby use. The navigation guide line
and airway data were displayed synchronously in real time via the
endoscopic image. After the flexible bronchoscope reached the target
lesion area, the lumen secretions were aspirated. When the puncture site
was reconfirmed, the pressure of the carbon dioxide cryopexor (1.0 mm
cryoprobe, Erbe, Germany) was set at 55–65 bars, a 1.0 mm cryoprobe was
connected, and the freezing time was set to 10 s (based on the freezing
effect of the cryoprobe in a water bath before the operation); a
specially assigned person was responsible for the countdown reminder.
After the guide sheath was inserted via the working channel of the
flexible bronchoscope, the cryoprobe was inserted into the sheath and
then placed in the target position under the guidance of the flexible
bronchoscope. The distance from the front end of the cryoprobe to the
pleura was confirmed by C-arm X-ray (Figure 2C, D). As the foot pedal of
the cryopexor was depressed, a 10-s countdown was started. When the
countdown prompt tone sounded, the cryoprobe and the flexible
bronchoscope were withdrawn, and, at the same time, the assistant
quickly injected gas into the preset balloon for occlusion. The frozen
tissue was placed on a glass slide for size measurement after the
cryoprobe was thawed, and neutral formalin fixative was added for
preservation. The flexible bronchoscope was then reinserted into the
target lumen; after observation for approximately 1 min, it was
confirmed that there was no blood leakage, and the balloon was released
slowly and replaced next to the lobe segment to be occluded. Specimens
were collected four times with repeated freezing. Specimen diameter was
between 4 mm and 6 mm (Figures 2 F). When no active bleeding was
observed, the bronchoscope was withdrawn and anesthesia recovery was
started. The child had stable vital signs after the operation, and no
complications such as hemoptysis or pneumothorax occurred. One week
after the operation, the pathological results were reported. Light
microscopy of the bronchoalveolar lavage fluid revealed a large number
of degenerated red blood cells in the smear, as well as many
hemosiderin-laden cells; special staining showed the presence of iron.
The lung tissue biopsies permitted analysis of alveolar and bronchial
structures. Hemosiderin-laden cells were observed in the alveolar
cavity, and the alveolar septa showed widening as a result of
lymphohistiocytic infiltration. In addition, lymphocyte foci were
aggregated around the airway, and part of the vascular wall structure
was destroyed and thickened. Furthermore, there was crystalline
deposition around the vascular wall, and multinucleated giant cells were
observed (Figure G). Following a diagnosis of pulmonary vasculitis,
regular glucocorticoid therapy was performed, and regular outpatient
follow-ups were performed. The patient did not experience hemoptysis
again, and, after 3 months of treatment, lung imaging revealed obvious
improvement (Figure 2H).
A large number of studies have shown that the etiological diagnosis rate
achieved by TBLC in DPLD is more than 80%, with a low incidence of
serious complications[7-8]. Thus, as a minimally invasive method for
the diagnosis of DPLD, TBLC is considered a promising alternative to SLB
[9-10]. However, there is limited knowledge of the suitability of
TBLC for use in children, as the characteristics of airway development
in children differ from those in adults. In one study, Moslehi et al.
[11] performed TBLC in younger children (17 out of 28 were younger
than 4 years old) and achieved a diagnostic rate of 92.8%. Nonetheless,
TBLC can cause complications such as bleeding and pneumothorax. Based on
conventional virtual navigation, the Archimedes system combines
augmented reality and vessel mapping technology and can integrate
spatial information relating to the airway and adjacent blood vessels,
as well as distance from the pleura. Virtual bronchial trees can be
reconstructed based on thin-slice CT, with the target lung lesions being
indicated, and the biopsy pathway and procedure plan can be formulated
to accurately avoid blood vessels; distance from the pleura and biopsy
depth can also be measured. This technique has significantly improved
the accuracy of bronchoscopy for in-depth diagnosis of the terminal lung
regions and provided safer guidance for lung biopsies[5]. However, its application as a biopsy method
in children with diffuse lung disease has not yet been reported.
In this article, our results show that using VBN is a feasible way to
assist TBLC in order to conduct safe and accurate biopsy operations in
children with DPLD. The experience gained from this case is summarized
as follows. To perform the operation safely, we recommend that that TBLC
should be performed by skilled surgeons, with the children under general
anesthesia to avoid poor cooperation. For diffuse lesions, larger blood
vessels should be avoided for biopsy; in this case, two sub-segments
were chosen for lung biopsy to improve the diagnostic rate. For diffuse
pulmonary lesions, the basal sub-branch of both lungs is preferred,
especially the lumen of the lateral and posterior bronchial sub-segment,
as this is relatively straight and unobstructed. Thus, it is easy to
reach and is unlikely to become adhered to the adjacent vascular wall
during the withdrawal of the cryoprobe, which would affect the sampling
process. The freezing effect was tested repeatedly according to the
diameter of the cryoprobe and the pressure of the gas source in this
study. In terms of freezing effect and safety, the freezing time of the
1.0 mm cryoprobe was set to 10 s each time, and no intraoperative or
postoperative bleeding or pneumothorax occurred. In addition, it is
worth noting that attention should be paid to intraoperative changes in
tidal volume (i.e., lung volume) in children. As the airway is occupied
during bronchoscopy, air leakage may occur during mechanical
ventilation. This means there may be a great difference between the
actual tidal volume and the tidal volume set by mechanical ventilation.
Such differences may bring about a change in lung volume, which will
directly affect navigation accuracy and operation safety. Thus, it is
necessary to adjust the biopsy site according to ventilator parameters
during the operation and review the target site again using the C-arm.
Following this initial application of VBN-guided lung biopsy and TBLC
technology in children, it will be necessary to conduct further clinical
explorations with larger sample sizes in the future. We believe that,
with the safety advantage that comes from combining VBN and TBLC, the
application of this technique will become widespread in children,
further promoting the transformation from conventional empirical
diagnosis to a more accurate and minimally invasive diagnosis mode.
Through the combination of clinical imaging in pathology and upgrading
of the technology, it is expected that the diagnosis of DPLD in children
will reach a new level.