Shape-Sensing Robotic Bronchoscopy Matched ENB on Yield and Pulled Ahead on Efficiency
Peripheral pulmonary lesions remain a difficult target in interventional pulmonology. Small nodules sit beyond the reach of standard bronchoscopy, and every missed biopsy prolongs the diagnostic chain for patients who may already be facing a possible lung cancer diagnosis. In this single-center retrospective comparison, shape-sensing robotic-assisted bronchoscopy and electromagnetic navigation bronchoscopy produced the same strict diagnostic yield from the index procedure, yet the robotic platform reduced the number of procedures needed to secure a malignant diagnosis.
The study examined procedures performed by a single interventional pulmonologist at a large urban hospital, creating a relatively uniform technical environment. ENB cases were collected between December 2020 and June 2023, while ssRAB cases came from September 2023 through August 2024. Every case used radial endobronchial ultrasound and two-dimensional fluoroscopy. Neither cone beam CT, digital tomosynthesis, nor rapid on-site evaluation was used. In practical terms, this matters because the comparison isolates the two navigation platforms without the added advantage of advanced imaging or on-site cytology support.
Comparable Initial Yield in a Lean Imaging Environment
A total of 359 patients were included, with 226 in the ENB group and 133 in the ssRAB group. Strict diagnostic yield from the index procedure was 64% for both modalities. That result puts the two platforms on equal footing for the first biopsy attempt, at least in this setting and under these workflow constraints.
The similarity is clinically important because it suggests that shape sensing did not deliver a clear advantage in immediate tissue confirmation when both approaches were paired with the same adjunctive tools. The sampling strategy still depended on radial ultrasound localization and fluoroscopic guidance, and neither arm benefited from cone beam CT-based confirmation of tool-in-lesion position. In a setting like this, diagnostic success remains tethered not only to the navigation system but also to lesion size, lesion conspicuity, bronchus sign, tool stability, and the inherent limits of sampling a peripheral target through the bronchial tree.
Robotic Bronchoscopy Needed Fewer Trips to Confirm Malignancy
The more striking difference emerged when the endpoint shifted from overall yield to the number of procedures required to establish a diagnosis of malignancy. Among patients undergoing ssRAB, 84.6% were diagnosed after a single procedure, compared with 60.8% in the ENB group. The difference was statistically significant, with P=0.007.
A single successful diagnostic encounter matters to patients, clinicians, and workflow alike. Fewer procedures mean fewer anesthesia exposures, fewer scheduling delays, and less time spent waiting for pathology or repeating tissue acquisition after a nondiagnostic first pass. For lesions with a high pretest probability of cancer, the ability to secure a malignant diagnosis on the first attempt can accelerate staging, surgical referral, oncology consultation, and treatment planning. The signal here is not that ssRAB transformed raw yield, but that it improved procedural efficiency in the pathway from suspicion to diagnosis.
Smaller Lesions Were Sampled With the Robotic Platform
Lesions approached with ssRAB were smaller than those targeted with ENB, with a median size of 1.7 cm versus 2.4 cm. The size difference is a crucial piece of context. Smaller peripheral lesions are harder to reach, harder to stabilize, and easier to miss with standard transbronchial sampling. A platform that can maintain orientation and distal control in more compact airways may therefore preserve diagnostic performance when lesion size would otherwise work against the operator.
Even so, the reported strict yield remained the same between groups. That combination suggests the robotic system may have helped preserve efficiency in a more challenging lesion mix rather than producing a dramatic increase in capture rate. The retrospective design does not allow separation of platform effect from case selection, temporal practice patterns, or a possible learning-curve effect across the years studied. The later ssRAB cohort may also have benefited from accumulated operator experience, even though both technologies were used by the same physician.
Final Diagnostic Accuracy Favored ssRAB Without Reaching Significance
Final diagnostic accuracy was 71.2% for ENB and 76% for ssRAB, with P=0.39. The point estimate leaned toward ssRAB, but the difference did not reach statistical significance. In a study of this size, that result argues for caution rather than overinterpretation. The numerical separation may reflect a genuine trend, a case-mix shift, or simply the limits of retrospective comparison.
The absence of advanced three-dimensional imaging also shapes how these data should be read. Cone beam CT and tomosynthesis can improve confidence that biopsy tools have reached the lesion, especially for small or poorly visible nodules. Without those technologies, both systems were operating under a more basic imaging scaffold. Under those conditions, shape sensing may offer some procedural advantages, but it did not produce a clear superiority signal in the primary yield metric.
What This Means at the Bedside
For interventional pulmonologists working without cone beam CT, the findings offer a restrained but useful message: ssRAB can perform at least as well as ENB for strict index-procedure yield and may reduce the need for repeat procedures when malignancy is ultimately confirmed. The likely mechanism is mechanical rather than biological. Shape sensing improves catheter localization and distal navigation control, which can matter most when the lesion is small and the airway path is narrow or tortuous.
The study also keeps expectations grounded. Both technologies reached the same strict diagnostic yield, and final diagnostic accuracy differed without statistical significance. Advanced imaging was not part of the workflow, so these results do not answer whether ssRAB still holds an advantage once cone beam CT or tomosynthesis is added. Nor do they establish that the robotic platform is universally superior for every peripheral lesion.
What this study does show is more specific and more actionable: in a single-operator urban practice using radial ultrasound and fluoroscopy alone, ssRAB matched ENB on index yield and reduced the number of procedures needed to diagnose malignancy. The next question is how much of that benefit persists across operators, lesion sizes, imaging strategies, and institutional settings where three-dimensional confirmation is available.