Initial Experience with Robotic-Assisted CT-Guided Ablation: Analysis from a Single Center's Learning Curve in the...

To report initial experience and learning curve with robotic-assisted percutaneous ablations of abdominal lesions.

Fifty-four consecutive patients undergoing robotic-assisted ablation were retrospectively analyzed in this single-center observational study. Feasibility, that is, successful robotic applicator insertion with no conversion to freehand technique or major adjustment, safety, accuracy, and procedure timings were assessed. Exploratory cumulative sum analyses were performed to assess the robotic device learning curve.

Fifty-seven lesions in 54 patients (mean age, 66.1 years [SD ± 12.6]; 64.8% [35 of 54] men); 47 (87.0%) liver, 5 (9.3%) kidney, and 2 (3.7%) adrenal lesions were treated with robotic-assisted ablation. Forty-three lesions (75.4%) were treated with microwave ablation, 10 (17.5%) with cryoablation, and 4 (7.0%) with irreversible electroporation. A total of 94 applicators were inserted with robotic assistance (40.7%, multiapplicator procedures), including 78.7% off axial-plane trajectories. Patients were divided into 2 similar-sized learning curve cohorts (initial, 21 cases and 29 applicators; middle, 18 cases and 30 applicators) and 1 later proficiency cohort (15 cases and 35 applicators) after 6-month interval without device access. Feasibility was 93.6% (88 of 94 applicators), with all nonfeasibility in the initial cohort (P ≤ .001). Five complications occurred (4 minor and 1 major [renal failure]), only 1 (self-limited pneumothorax) related to robotic guidance. Median lateral deviation was 4.8 mm (range, 0-19.7 mm). Most placements (62.8%) required no or depth-only manual adjustment. Mean device use duration was 33 minutes (SD ± 18). Performance plateaued after around 20 cases.

After an initial learning curve, proficiency in robotic-assisted ablation was retained after a 6-month interval, resulting in consistent procedure times and maintained accuracy.