Contrast-Enhanced Cone-Beam CT-Guided Robotic Bronchoscopic Biopsy of a Mediastinal Lymph Node Adjacent to Major Vascular Structures.

1
Introduction
Accurate nodal staging is critical for optimal lung cancer management, especially when nodal involvement in specific stations may determine the curative intent of treatment. While endobronchial ultrasound‐guided transbronchial needle aspiration (EBUS‐TBNA) remains the standard approach, certain nodal stations are difficult or impossible to access via this technique due to their anatomical location. In recent years, robotic‐assisted bronchoscopy (RAB) and the integration of advanced imaging technologies, such as cone‐beam computed tomography (CBCT) or digital tomosynthesis (DT), have facilitated sampling of previously inaccessible locations. Despite these technologies, proceduralists may face challenging scenarios, one of which is proximity to major vessels. Although CBCT, particularly with segmentation and fluoroscopic augmentation, provides intra‐procedural three‐dimensional imaging that enhances safety and accuracy [1], it fails to clearly distinguish soft‐tissue targets from adjacent vascular structures. We present a case demonstrating the use of contrast‐enhanced CBCT to guide biopsy of a mediastinal lymph node.

2
Case Report
A 75‐year‐old female, ECOG performance status 0, was referred for further mediastinal staging following a diagnosis of primary lung adenocarcinoma. She is a lifelong non‐smoker and presented with a dry cough and unintentional weight loss of approximately 3–4 kg. Past medical history included ischaemic heart disease with prior non‐ST elevation myocardial infarction, coronary artery bypass grafting and metabolic syndrome. CT chest demonstrated a 63 mm left lower lobe mass. CT‐guided transthoracic biopsy confirmed adenocarcinoma with an exon 19 deletion EGFR mutation. A FDG‐PET scan demonstrated intense avidity in the left lower lobe primary tumour (SUV 19) along with uptake in the left hilum (station 11 L, SUV 7.5), left pulmonary ligament (station 9, SUV 11) and low left cervical (station 1 L, SUV 3.8) lymph nodes—Figure 1. Uptake in the left adrenal gland was further evaluated with MRI, demonstrating the absence of metastatic disease. A standard linear EBUS‐TBNA procedure was already performed, sampling stations 4 L, 7 and 11R, excluding malignant involvement of these nodes.
A thorough discussion at the tumour board highlighted the importance of clarifying malignant involvement of the 1 L nodal station, as curative intent concurrent chemoradiotherapy may not be an option due to the excessive toxicity associated with a large radiation field. Multidisciplinary input determined that both video‐assisted thoracoscopic surgery and bronchoscopic sampling were feasible, with the less invasive bronchoscopic approach favoured, noting her comorbidities and prior thoracic surgery.
A linear EBUS approach from the trachea was considered but not favoured, as it would have involved transversing major vessels. A transbronchial approach utilising RAB with CBCT guidance was chosen but posed the challenge of accurately segmenting the lesion, as intra‐procedural imaging without contrast does not adequately differentiate the target from surrounding vascular structures and mediastinal soft tissue.
Prior to the procedure, a sub‐segmental branch in LB1+2 was identified as the appropriate destination for launching a biopsy needle. The path to this destination was manually mapped using the Simplified Bronchial Branch Tracing technique [2], and pre‐procedural planning was performed on the Galaxy Robotic‐Assisted‐Bronchoscopy (RAB) system (Noah Medical, Sunnyvale, USA).
The procedure was performed under general anaesthesia with muscle paralysis, using an 8.5 cm endotracheal tube for airway access. Ventilation parameters were set to optimise breath control, airway patency and minimise atelectasis. The Azurion system (Philips, The Netherlands) was utilised for CBCT image acquisition. For the initial acquisition, 80 mL intravenous contrast medium (Visipaque 270, GE Healthcare) was injected through an 18G cannula in the right cubital fossa at 3 mL/s, starting 30 s prior to image acquisition, followed by 30 mL saline. Breath‐hold was initiated 8 s prior to acquisition. Using the EmboGuide software (Philips, Netherlands), the target lesion was confidently identified (Figure 2) and after comparison with pre‐procedural CT and PET scans (Figure 1), target segmentation and path planning were performed.
The robotic bronchoscope was navigated to the planned destination, 2–3 cm away from the target. DT image acquisition was performed, identifying and updating the target on the Galaxy system (Figure 2). While directing the scope towards the target with Galaxy's target mode, and correlating the augmented fluoroscopy on both the Galaxy and Azurion systems, a 21G PeriView needle (Olympus, Japan) was advanced towards the target and held at the proximal margin at a tool‐touch‐lesion position.
A second CBCT acquisition was obtained with breath hold, which confirmed tool‐touch‐lesion (Videos 1 and 2). A second DT acquisition performed during the same breath hold updated the Galaxy's target mode to the precise location at the tip of the needle. From this point on, sampling relied on Galaxy's target mode and augmented fluoroscopy. Multiple FNA samples were performed (Figure 3).
The procedure was then concluded by performing EUS‐B‐guided sampling of the station 9 lymph node, confirming malignancy at that station.
Patient recovery was uncomplicated, and a chest x‐ray did not reveal a pneumothorax or pneumomediastinum. She was discharged home the same day.
Cytological analysis of samples obtained from 1 L revealed inflammatory cells without any evidence of malignancy. Tumour board consensus was stage IIIB (cT3 N2) disease, suitable for concurrent chemoradiation with curative intent. The patient received weekly carboplatin (AUC2) and paclitaxel (45 mg/m2) with concurrent radiotherapy (60Gy/30#) with G1 fatigue and G2 oesophagitis as the principal complications. Follow‐up at 4 weeks post completion of therapy revealed a marked improvement in disease, decreasing from 58 × 38 mm to 31 × 13 mm. The patient commenced Osimertinib 80 mg daily as maintenance therapy, which she has tolerated well.

3
Discussion
Recent advancements in intra‐procedural advanced imaging, allowing segmentation, integration and augmentation, have revolutionised RAB, leading to the emergence of second‐generation ‘image‐integrated’ RAB with improved access to intra‐thoracic targets [3]. Despite this, limitations in current technology, for instance, the inability to confidently differentiate the soft‐tissue target from important vascular structures, may pose important safety and accuracy challenges in some cases. To our knowledge, this is the first reported case where intra‐procedural contrast‐enhanced CBCT imaging is used to overcome this challenge. Although we decided to combine contrast‐enhanced CBCT with DT to improve safety and accuracy in this case, it remains unclear whether contrast‐enhanced DT alone would be adequate.

Author Contributions
All authors listed above were involved in the clinical aspects of this case and contributed to the manuscript in relation to their speciality.

Funding
The authors have nothing to report.

Consent
The authors declare that written informed consent was obtained for the publication of this manuscript and accompanying images using the consent form provided by the Journal.

Conflicts of Interest
The authors declare no conflicts of interest.