NEW ORLEANS – A novel according to an industry-sponsored, prospective study.
The system, which incorporates fluoroscopic navigation, increased the percentage of cases in which the target overlapped with the lesion, from 60% to 83%. The percentage of cases without any target overlap decreased from 32% to 5%.
“Tomosynthesis-based fluoroscopic navigation … improves the three-dimensional convergence between the virtual target and the actual target,” said, of CHI Memorial Medical Group in Chattanooga, Tenn.
Dr. Bhadra presented results with this system at the annual meeting of the American College of Chest Physicians.
He and his colleagues conducted a study of(version 7.2), which provides real-time imaging with three-dimensional fluoroscopy. The system has a “local registration” feature, which uses fluoroscopy and an algorithm to update the virtual target location during the procedure. This allows the user to reposition the catheter based on the location of the lesion.
The researchers tested the system in 50 patients from two centers (). Patients’ lesions had to be larger than 10 mm, not visible endobronchially, and not reachable by convex endobronchial ultrasound. Lesions within 10 mm of the diaphragm were excluded.
The median lesion size was 17.0 mm, 61.2% were smaller than 20 mm, 65.3% were in the upper lobe, and 53.1% had a bronchus sign present. The median distance from lesion to pleura was 5.9 mm.
Dr. Bhadra said the system performed as designed in all cases, and the protocol-defined technical success rate was 95.9% (47/49). Local registration was attempted in 49 patients and was successful in 47 patients (95.9%). In the unsuccessful cases, local registration was not completed based on the system design because the correction distance was greater than 3.0 cm.
The study’s primary endpoint was three-dimensional overlap of the virtual target and the actual lesion, as confirmed by cone-beam computed tomography. Success was defined as greater than 0% overlap after location correction. Target overlap was achieved in 59.6% (28/47) of cases before local registration and 83.0% (39/47) of cases after.
There were six cases in which local registration was successful, but these subjects weren’t evaluable because of failed procedure recording. When those subjects were excluded, target overlap was achieved in 95.1% (39/41) of cases after local registration.
The median percent overlap between the virtual target and the actual lesion was 11.4% before local registration and 32.8% after. The percentage of cases without any target overlap decreased from 31.7% (13/41) before local registration to 4.9% (2/41) after.
Focusing on the two cases without target overlap, Dr. Bhadra noted that he was able to get a biopsy that proved a malignancy in one of those patients. In the other patient, Dr. Bhadra was able to identify features of organizing pneumonia.
“Even though we did not have overlap, we must have been close enough that we were able to get malignant tissue in one [patient] and features of organizing pneumonia in a patient who’s got no history of organizing pneumonia,” Dr. Bhadra said.
He and his colleagues did not evaluate diagnostic yield in this study, but they did assess complications up to 7 days after the procedure.
The team reported one case of pneumothorax, but the patient didn’t require a chest tube. Additionally, there were two cases of bronchopulmonary hemorrhage, but the patients didn’t require any interventions.
This study was sponsored by Medtronic. Dr. Bhadra disclosed relationships with Medtronic, Boston Scientific, BodyVision, Auris Surgical Robotics, Intuitive Surgical, Veracyte, Biodesix, Merit Medical Endotek, and Johnson & Johnson.
SOURCE: Bhadra K et al. CHEST 2019. .