Cone-beam computed tomography augmented-fluoroscopy-guided center-strike targeting during peripheral bronchoscopy: robotic vs. manual catheters

Computed tomography (CT)-to-body divergence and tool-tissue divergence (TTD) limit bronchoscopic access to peripheral pulmonary lesions (PPLs). Fixed cone-beam CT (CBCT) with augmented fluoroscopy (AF) enables intraprocedural target segmentation, live fluoroscopic overlay, and confirmation of tool position. We describe three CBCT AF-guided spatial targeting techniques as a targeting workflow and compare manual pre-curved and shape-sensing robotic catheters using center strike as the primary technical endpoint.

This retrospective single-center study included 200 consecutive patients with 219 biopsied PPLs. The first 100 patients underwent biopsy with manual pre-curved catheters (Medtronic; 106 lesions), followed by 100 patients who underwent biopsy using a shape-sensing robotic catheter (Intuitive; 113 lesions). All procedures used fixed CBCT-AF, a ventilation protocol with apneic breath-holding, needle targeting with traditional two-angle, slow-sweep, and/or center-datum techniques, and CBCT tool-position confirmation. Center strike was defined as needle position in the middle third of the lesion on axial, sagittal, and coronal CBCT images. The primary outcome was center-strike tool position; diagnostic yield and 1-year diagnostic accuracy were secondary outcomes.

Robotic catheters achieved a higher center-strike rate, the primary endpoint, than manual catheters {66% [75/113; 95% confidence interval (CI): 57-74%] vs. 42% (45/106; 95% CI: 33-52%), P<0.001}, as well as a higher tool-in-lesion (TIL) rate [100% (113/113; 95% CI: 97-100%) vs. 91% (96/106; 95% CI: 83-95%), P<0.001]. Diagnostic yield was 91% (95% CI: 84-95%) vs. 80% (95% CI: 72-87%) (P=0.02), and 1-year diagnostic accuracy was 96% (95% CI: 91-99%) vs. 86% (95% CI: 78-91%) (P=0.01). Dose-area product (DAP) was higher with robotic catheters [median 13,251 [interquartile range (IQR), 8,683-18,264] vs. 11,114 (IQR, 6,336-15,199) µGy·m², P=0.04], without significant differences in fluoroscopy time or number of CBCT spins. Two pneumothoraces occurred in the robotic cohort, one requiring chest tube placement.

A CBCT AF-guided targeting workflow combines real-time needle-vector alignment with intraprocedural confirmation of tool position before tissue sampling. Within this workflow, robotic catheters achieved higher center-strike and TIL rates and were associated with improved diagnostic performance. Reliable intraprocedural spatial confirmation may become increasingly important as bronchoscopy evolves toward precision transbronchial therapies.