Robotic-assisted Bronchoscopy Combined with Multimodal Imaging for Targeted Lung Cryobiopsies

Abstract

Robotic-assisted bronchoscopy (RAB) allows for targeted bronchoscopic biopsy in the lung. A robotic-assisted bronchoscope is navigated through the airways under direct vision after establishing a pathway to a target lesion based on mapping performed on a 3-dimensional (3D) lung and airway reconstruction obtained from a pre-procedure thin-slice computed tomography chest. RAB has maneuverability to distal airways throughout the lung, precise catheter tip articulation, and stability with the robotic arm. Adjunct imaging tools such as fluoroscopy, radial endobronchial ultrasound (r-EBUS), and cone beam computed tomography (CBCT) can be used with RAB. Studies using shape-sensing robotic-assisted bronchoscopy (ssRAB) have shown favorable diagnostic outcomes and safety profiles in both malignant and non-malignant processes for the biopsy of peripheral pulmonary lesions (PPLs). A 1.1 mm cryoprobe combined with ssRAB has been shown to be safe and effective for the diagnosis of PPLs compared to a traditional bronchoscopy with forceps biopsy. This technique can also be used for targeted lung sampling in benign processes. The aim of this article is to describe a stepwise approach to performing RAB combined with fluoroscopy, r-EBUS, and CBCT to obtain targeted transbronchial lung cryobiopsies (TBLC).

Introduction

Flexible bronchoscopy with transbronchial lung biopsy (TBBX) is a diagnostic modality used for the evaluation of abnormal chest imaging, including masses, nodules, non-resolving infiltrates, or parenchymal lung diseases¹. Diffuse parenchymal lung diseases (DPLD) can often be characterized by fibrosis and/or inflammation. While some patients can be diagnosed noninvasively with a thorough history, physical examination, relevant serologies, high-resolution computed tomography (HRCT) findings, and multi-disciplinary discussion (MDD), many patients need an invasive procedure to establish a diagnosis². Conventional transbronchial lung biopsies with forceps are limited due to small biopsy size and crush artifacts; as a result, surgical lung biopsy has been considered the gold standard, although it has significant morbidity and mortality^3,4.

Transbronchial lung cryobiopsy (TBLC) is a technique that can be used to diagnose interstitial lung disease (ILD) or diffuse parenchymal lung disease (DPLD) and might serve as an alternative to surgical lung biopsy (SLB)⁵. According to the European Respiratory Society guidelines, TBLC is recommended as a substitute for SLB in eligible patients⁶. Similarly, the American Thoracic Society guidelines offer a conditional recommendation for TBLC as an alternative to SLB in medical centers with the necessary expertise in performing and interpreting TBLC results⁷. TBLC has historically provided good accuracy in diagnosis compared to SLB but is limited by complications, including bleeding and pneumothorax⁸. A recent meta-analysis showed an overall diagnostic yield of 77% that improved to 80.7% with MDD, and reported a pneumothorax rate of 9.2% and bleeding rate of 9.9%⁹. TBLC is also used in the evaluation of PPLs¹⁰. 

The development of robotic-assisted bronchoscopy (RAB) allows for targeted sampling in the lung by navigating through the airways under direct vision with easy catheter maneuverability, precise catheter tip articulation, stability, and the ability to maintain a bronchoscopic wedge in distal airways with the catheter using a robotic arm. The Ion endoluminal system utilizes shape-sensing technology for navigation to access specific targeted areas in the lung. Studies using shape-sensing robotic-assisted bronchoscopy (ssRAB) have shown favorable diagnostic outcomes and safety profile, primarily for PPLs suspicious of malignancy^11,12,13,14. A 1.1 mm cryoprobe for TBLC combined with ssRAB has been shown to be safe and effective for the diagnosis of pulmonary nodules compared to transbronchial biopsy with forceps¹⁵. This technique can be used to obtain targeted lung biopsies larger than conventional transbronchial biopsies using forceps that are relatively free of crush artifacts. 

Radial endobronchial ultrasound (r-EBUS) and cone beam computed tomography are used in conjunction with conventional bronchoscopy, electromagnetic, or robotic navigational systems for real-time confirmation prior to sampling PPLs^{16,17,18,19,20,21,22}. R-EBUS has also been utilized during TBLC for DPLD to increase the pathologic confidence of lung specimens, decrease bleeding, and have a shorter procedure time²³. The addition of CBCT has improved the safety profile of TBLC for DPLD by confirming the probe tip is in a safe zone for biopsy, allowing objective measurement of the distance from the pleura with the ability to visualize and avoid vasculature^24,25,26. 

This protocol will describe a procedure to obtain targeted TBLC in the setting of parenchymal lung disease for patients who are able to tolerate and benefit from the procedure using the Ion endoluminal system in conjunction with fluoroscopy, r-EBUS, and CBCT in a clinical setting under general anesthesia. This multimodal approach allows for precise sampling of targeted areas of interest. 

Protocol

The protocol described in this article outlines standard clinical practice. The University of Texas Southwestern Medical Center Institutional Review Board approved the prospective data collection of patients undergoing standard-of-care bronchoscopy with ssRAB (STU-2021-0346), and individual consent is waived for inclusion in our database. Routine procedure consent is obtained from the patient prior to the procedure. Patients who have DPLD radiographically and are acceptable candidates for bronchoscopic biopsy are referred for this procedure^5,27. Patients over 18 years of age are deemed able to undergo the procedure by the referring and performing physicians. Exclusion criteria include bleeding disorders (elevated INR >1.3, thrombocytopenia <100,000/µL), hypoxia with pulse oximetry <90% on 2 L/min supplemental oxygen, pulmonary hypertension (echocardiographically measured systemic pulmonary artery pressure >50 mmHg), or severe cardiac disease. The details of the equipment used in this study are listed in the Table of Materials.

1. Pre-procedural planning

1. Upload the patient’s thin slice CT chest to the planning software. The software will automatically create a 3-dimensional reconstruction of the airways and lungs.
2. Select targets in the lungs for proposed sampling approximately 10 mm from the pleural border. NOTE: This may be an area of ground glass, infiltrate, nodularity, or fibrosis after discussion with the referring physician, radiologist, or clinical judgment. Prior reported literature for peripheral lung lesions shows an increased diagnostic yield if the targeted area is  >2 cm²⁸.
3. Plan a pathway to each target site.
   NOTE: In the event that the pathway is not feasible during the procedure, consider planning a secondary pathway.
4. Review the plan in all three CT views (axial, coronal, and sagittal) and virtual bronchoscopy (Figure 1).
5. Export the plan to the controller console. 

2. Patient preparation

1. Induce and maintain the patient under general anesthesia with a minimum size 8.0 single-lumen endotracheal tube. Use neuromuscular blockade and total intravenous anesthesia with ventilator protocols to reduce the development of atelectasis²⁹.
   NOTE: Monitoring of paralysis is performed using a train-of-four test with a peripheral nerve stimulator²⁹.
2. Tuck the patient’s arms to allow for full rotation of the c-arm during CBCT spin.
3. Position the patient and c-arm so that the targeted area for TBLC is isocentered on fluoroscopy.

3. Conventional bronchoscopy

1. Insert the diagnostic or therapeutic bronchoscope via a bronchoscope adapter into the endotracheal tube.
2. Perform an airway examination and minimize suction to reduce the development of atelectasis³⁰.
3. Remove the bronchoscope.

4. Robotic-assisted bronchoscopy

1. Docking
   1. Move the robotic bronchoscope to a position adjacent to the patient.
   2. Dock the robotic arm with the magnetic bronchoscope adapter. Insert the catheter and vision probe into the endotracheal tube.
2. Registration
   1. Position the catheter so that direct vision matches the virtual bronchoscope image at the carina.
   2. Maneuver the robotic catheter via a scroll wheel and track ball on the controller console into both mainstem airways, then bilateral upper and lower airways to collect airway data.
   3. Compare virtual versus actual bronchoscope images after registration is complete. If significant mismatching or divergence is noted, perform re-registration; otherwise, accept the registration.
3. Navigation
   1. Maneuver the catheter using the scroll wheel and track ball on the controller console through the airways to the target lesion following the planned pathway.
   2. Use the “Preview Path” feature to follow images of the airways if divergence (discrepancy between the virtual and actual airways) is noted.
4. Use fluoroscopy, r-EBUS, and CBCT to confirm location.
   1. Remove the vision probe when the catheter is within 5–10 mm of the target lesion.
   2. Advance the r-EBUS probe with probe rotation under fluoroscopy. Advance to the pleural border (Figure 2A).
   3. Retract the r-EBUS probe under fluoroscopy approximately 10 mm from the pleural border to the anticipated biopsy target site. Use the r-EBUS probe to visualize the target area and assess the surrounding parenchyma and any vasculature in the potential biopsy area. Remove the r-EBUS probe.
   4. Insert the 1.1 mm touch cryoprobe via the catheter and extend under fluoroscopy to the pre-determined target area for biopsy (Figure 2B).
   5. Perform cone beam CT spin per system-specific protocol. The ventilation may be continued or held per provider preference using an end-inspiratory breath hold with the ventilator’s adjustable pressure-limiting valve set to match the positive end-expiratory pressure (PEEP) or vital capacity maneuver.
      NOTE: The CBCT spin may be performed with the extension of the r-EBUS probe without rotation or 1.1 mm cryoprobe in the anticipated biopsy site.
   6. Interpret and compare the intra-procedure imaging to the pre-procedure CT chest and plan to ensure the catheter is at the target. If augmented fluoroscopy is available on the CBCT, segment the target for visualization with 2-D fluoroscopy during biopsy (Figure 3).
   7. Adjust the catheter based on fluoroscopy, CBCT, and r-EBUS to ensure that sampling occurs in the appropriate location.
   8. Repeat CBCT after the catheter is adjusted if necessary.
5. Tissue sampling
   1. Ensure the 1.1 mm touch cryoprobe is in the appropriate biopsy position.
   2. Press the pedal to activate the freeze cycle from 4 s to 6 s, then retract the probe in one motion while continuing to depress the pedal.
   3. Release the pedal while placing the probe tip with tissue biopsy in sodium chloride 0.9% or fixative to release the biopsy from the tip.
   4. Repeat steps 4.5.1–4.5.3 to perform TBLC.
      NOTE: The authors typically perform 1–4 biopsies at each site. During the biopsy process, the catheter may be adjusted slightly prior to each biopsy to ensure adequate tissue procurement; this may require repeating CBCT spins or using r-EBUS to confirm position in step 4.4.
   5. After the final biopsy, inject 1–2 mL of normal saline and air in a 10 mL Leuer lock syringe into the catheter to clear any blood or secretions.
   6. Insert the vision probe to view the sampling site and retract the catheter slowly. If there is evidence of bleeding via direct vision or blushing on fluoroscopy, then instill topical 1:10,000 epinephrine 1 mL, additional cold saline, or 50–100mg of tranexamic acid via Leuer lock syringe. Then, leave the catheter in place for 3–5 min for additional tamponade.
   7. Repeat step 4.5.6. If there is no evidence of bleeding, retract the catheter to the trachea.
   8. If significant bleeding is noted, then remove the robotic bronchoscope and follow protocols for the management of iatrogenic bleeding after flexible bronchoscopy³¹.
6. Additional target sites: If additional target sites are planned for biopsy, then repeat steps 4.3–4.5.
7. After the conclusion of robotic-assisted bronchoscopy, retract the catheter, undock the robotic system, and move out of position.

5. Conventional bronchoscopy 

1. Reinsert the diagnostic or therapeutic bronchoscope through the bronchoscope adapter into the airways for airway examination and suctioning.
2. If bronchoalveolar lavage is indicated, advance the bronchoscope into a subsegmental airway where TBLC was not performed to create a wedge. Instill serial aliquots of normal saline and then return via hand suction. 

6. Procedure conclusion

1. Remove the bronchoscope.
2. Perform fluoroscopy or focused ultrasound⁵ to assess for pneumothorax. If pneumothorax is identified, consider the placement of a chest tube versus conservative management with serial observation depending on size and patient’s clinical status.
3. If necessary, transfer the TBLC specimens to the container with a fixative if initially placed in 0.9% sodium chloride.
4. Reverse anesthesia, extubate, and awaken the patient.
5. Transfer the patient to the post-anesthesia care unit.
6. Perform and review post-procedure chest radiograph (Figure 4).

7. Follow-up post procedure

1. Review results from the bronchoscopy at a multi-disciplinary discussion attended by pulmonologists who are experts in interstitial lung disease, thoracic radiologists, and thoracic pathologists to determine an ILD subtype.
2. Discuss the results of the bronchoscopy and MDD conference with the patient, as well as plans for further management and follow-up. 

Representative Results

The described technique allows for targeted transbronchial lung cryobiopsies via RAB with fluoroscopy, r-EBUS, and CBCT guidance. Compared to conventional bronchoscopy with random TBLC, this technique allows for targeting specific areas of DPLD or PPLs of interest while assessing surrounding structures prior to biopsy. This technique can be used with r-EBUS and fluoroscopy only or with a combination of CBCT. While this technique had been devised for PPLs, it can be utilized in benign and diffuse parenchymal lung diseases to ensure the sampling of precise, targeted areas.

Historically, TBLC has been performed with larger cryoprobes (1.7 mm, 1.9 mm, 2.4 mm)^9,10. Evaluation of the smaller, disposable 1.1 mm cryoprobe is ongoing. A recent prospective, randomized controlled trial evaluating 1.1 mm versus 1.9 mm cryoprobes for patients undergoing bronchoscopy with TBLC in the evaluation of DPLD reported no difference in specimen quality or diagnostic rate but a smaller sample size with the 1.1 mm cryoprobe³². The FROSTBITE-2 trial, a study to compare the effectiveness of transbronchial biopsy done by a 1.1 mm cryoprobe versus standard 2.0 mm forceps for a variety of lung processes (evaluation of lung transplant allograft, DPLD or PPL) is currently enrolling.

There is limited data on DPLD using navigational bronchoscopy guidance to obtain TBLC. Kronborg et al.³³ reported a pilot study on the use of electromagnetic navigational bronchoscopy-guided TBLC in 18 patients with DPLC using a 1.7 mm or 1.9 mm touch cryoprobe. They reported that biopsies contributed to the diagnosis in 11 patients, with pneumothorax in 3 and mild-moderate hemorrhage in 7³³.

We have previously published our outcomes for patients undergoing this multi-modality technique (using ssRAB with fluoroscopy, r-EBUS, and CBCT) for targeted lung sampling in both benign and malignant diseases^21,22. To date, from the initial manuscript submission, we have had ten patients who have undergone RAB with fluoroscopy, CBCT, and R-EBUS for TBLC in DPLD, which are not included in these data sets. Among this cohort, there were no episodes of bleeding requiring intervention. One patient required admission for observation, followed by chest tube placement for an enlarging but asymptomatic pneumothorax. Eight patients had biopsy in one lobe (1 right upper lobe, 5 right lower lobe, 1 left upper lobe, 1 left lower lobe) and 2 patients in multiple lobes (right upper and right lower lobes). The mean total biopsies obtained was 7 ± 1.2. The mean largest dimension of biopsies was 0.32 mm ± 0.1 mm. Bronchoscopy results contributed to diagnosis in 70% of patients. Data collection is ongoing.

Figure 1: CT chest slices in axial, coronal, and sagittal views with airway reconstruction and selected biopsy site in right upper lobe with planned pathway.  Please click here to view a larger version of this figure.

Figure 2: Views on ssRAB console with fluoroscopy at left anterior oblique (LAO) 30 degrees and digital zoom. (A) Top with r-EBUS extended with no significant vasculature noted at the biopsy site. (B) 1.1 mm touch cryoprobe extended. Please click here to view a larger version of this figure.

Figure 3: Cone beam CT performed with r-EBUS probe visible. Biopsy area segmented for use with augmented fluoroscopy. Note significant motion artifact and atelectasis when CBCT spin performed without breath hold. Please click here to view a larger version of this figure.

Figure 4: Post-procedure chest radiograph in anteroposterior view. Bilateral interstitial densities noted with no pneumothorax or pleural effusion. Please click here to view a larger version of this figure.

Discussion

This manuscript provides a stepwise approach for performing RAB with fluoroscopy, r-EBUS, and cone beam CT to obtain targeted TBLC. 

There are several critical steps in this protocol. First, patient selection is imperative to ensure patients are both appropriate candidates (the biopsy procedure may have a direct impact on diagnosis and further care) and medically able to undergo the procedure^5,6. Pre-procedure preparation includes discussion with the referring physician and/or radiologist to determine differential diagnosis and optimal biopsy areas that may yield diagnostic material. This combination of techniques (CBCT and r-EBUS) is helpful in patients who have DPLD but may have focal areas of disease such as nodularity or ground glass opacities amenable to targeted sampling, and/or if the optimal biopsy sites are adjacent to critical structures (pleura, vasculature) or a technically challenging area of the lung to access bronchoscopically due to airway angulation or anatomy. If CBCT is unavailable, then the addition of r-EBUS can be helpful in determining the pleural border and allowing r-EBUS visualization of the planned biopsy area to identify areas of ground glass and to avoid vasculature to decrease bleeding risk. 

It is important to plan to sample multiple lesions and account for various pathways to the target areas if possible, as numerous variables such as narrowed distal airways, mucus impaction, or airway angulation can make navigation challenging. The benefits of using the ssRAB catheter, with a 3.5 mm outer diameter, is that the flexibility and stability allow access throughout the lung to areas that may have previously been inaccessible bronchoscopically. 

After navigation, adjunct imaging such as r-EBUS and CBCT serves to verify the potential biopsy site is acceptable. R-EBUS with 2-dimensional fluoroscopy is used to visualize the pleural border, parenchyma, and any vasculature structures at the potential biopsy site. CBCT serves to further ensure the biopsy site is appropriate without surrounding critical structures, approach direction for sampling, and assist with later clinical-pathologic correlation at MDD if the intra-procedure imaging is uploaded to PACS. 

When obtaining the TBLC, It is vital to maintain continued depression of the pedal while withdrawing the probe using the 1.1 mm touch cryoprobe, as early release can lead to premature defrosting and tissue loss. Modifications are based on patient characteristics, including body habitus distribution of disease, and can include adjusting areas targeted for biopsy, the number of TBLC obtained in each target site, or adjustment of freeze cycle depending on tissue size obtained. 

Given the current size of the ssRAB catheter and the described technique, this protocol is limited to the use of a 1.1 mm touch cryoprobe. The combination of a smaller cryoprobe compared to larger probes (1.7 mm, 1.9 mm, or 2.4 mm) and the reliability of ssRAB to maintain a distal wedge position has decreased the risk of bleeding and potentially obviated the need for an endobronchial blocker. This technique may be helpful in patients who otherwise may not be able to undergo a non-targeted TBLC given patient characteristics or available procedure expertise, and would otherwise have undergone standard bronchoscopy with TBBX or no procedure. Further research is needed on this technique in non-malignant parenchymal lung disease. 

This combination of techniques may currently have limited generalizability given the cost, availability, and training required with the multi-modality approach and equipment necessary: ssRAB, fluoroscopy, CBCT, and r-EBUS. However, given the increasing use and availability of these techniques to biopsy PPLs^15,34, this may allow for increased use in non-malignant diseases. Reproducibility of these techniques should be performed in centers with high volume and experience in peripheral bronchoscopy with the ability to handle complications (pneumothorax and airway bleeding), in addition to access to MDD and thoracic surgery if used for the evaluation of DPLD. Other novel techniques that may be additive in the bronchoscopic evaluation of DPLD in the future include optical coherence tomography (OCT), confocal laser endomicroscopy (CLE), rapid onset evaluation (ROSE) to assess specimen adequacy, and immunohistochemistry and genomic classifier (GC) testing^{35,36,37,38,39}. 

In summary, the multimodal approach using ssRAB with adjunct advanced imaging and 1.1 mm cryoprobe in obtaining targeted lung biopsies for those with malignant or benign pulmonary processes and abnormal chest CTs may provide a safer modality to obtain lung tissue and aid in establishing a diagnosis. Further evidence is required regarding diagnostic yield and safety for these combined techniques for both benign and malignant diseases. 

Materials

Name	Company	Catalog Number	Comments
0.9% normal saline, 1000 mL	Any make
10 mL Leuer lock syringes	Any make
20 mL slip tip syringes	Any make
Bronchoscope	Intuitive
Bronchoscope processor and video screens	Intuitive
Carbon dioxide gas tank
Cone beam computed tomography system with c-arm and controller console
Disposable valve for biopsy channel
Disposable valve for suction
ERBECRYO 2 1-pedal footswitch AP & IP X8 Equipment US	Erbe	20402-201
ERBECRYO 2 Cart	Erbe	20402-300
ERBECRYO 2 Cryosurgical unit	Erbe	10402-000
ERBECRYO 2 System	Erbe
Flexible Cryoprobe, OD 1.1 mm, L1.15 m with oversheath, OD 2.6 mm, L817 mm	Erbe	20402-401
Flexible gas hose; L 1m for Erbokryo CA/AE/ERBECRYO 2	Erbe	20410-004
Gas bottle adapter H; CO2; Pin index	Erbe	20410-011
Ion endoluminal system with robotic arm, controller console	Intuitive
Ion fully articulating catheter	Intuitive	490105
Ion instruments and accessories
Ion peripheral vision probe	Intuitive	490106
Laptop with PlanPoint planning software	Intuitive
Probe driving unit	Olympus	MAJ-1720
Radial EBUS Probe	Olympus	UM-S20-17S or UM-S20-20R-3
Radial endobronchial ultrasound system
Specimen containers with fixative per institution standards
Sterile disposable cups
Suction tubing
Topical 1:10,000 epinephrine, 10 mL
Topical tranexamic acid 1000mg, 10 mL
Universal ultrasound processor	Olympus	EU-ME2
Wire basket; 339 x 205 x 155 / 100 mm	Erbe	20180-010