National trends and perioperative outcomes of robotic resection of thymic tumours in the United States: a propensity matching comparison with open and video-assisted thoracoscopic approaches†.

Objectives: Despite the recent increased rate of adoption of robotic approaches for the resection of thymic tumours, their use is still limited to large-volume academic centres. To date, a large-scale analysis of the robotic approach has not been performed. We assessed the recent trends and outcomes of robotic thymectomies in the United States compared to those of open and video-assisted thoracoscopic surgical (VATS) approaches.

Methods: The National Cancer Database was queried for patients who underwent resection for thymic tumours (2010-2014). Predictors of using the robotic approach were estimated by logistic regression analysis. Propensity matching analysis (robotic versus open and robotic versus VATS) was done (1:1-caliper 0.05), controlling for age, gender, comorbidity index, induction treatment, tumour size and tumour extension.

Results: A total of 2558 thymectomies were performed (robotic = 300, VATS = 280, open = 1978). The use of a robotic approach increased from 6% (2010) to 14% (2014). The number of hospitals performing at least 1 robotic thymectomy increased from 22 (2010) to 52 (2014). Independent predictors influencing the choice of a robotic approach included an academic research/integrated cancer programme [odds ratio (OR) 1.66, confidence interval (CI) 1.22-2.27], later year of diagnosis (2014; OR 2.23, CI 1.31-3.80) and a patient's race (Asian) (OR 1.68, CI 1.05-2.69). A robotic approach was less likely to be utilized in midwestern hospitals (OR 0.65, CI 0.42-0.99), in larger tumours (cm) (OR 0.85, CI 0.80-0.90), with invasion of adjacent organs (OR 0.55, CI 0.37-0.82), thymic carcinoma (OR 0.62, CI 0.40-0.97) and following induction chemotherapy (OR 0.22, CI 0.08-0.61). In a propensity-matched analysis, there were no differences in the incidence of positive margins, nodal dissection, 30-day readmission rates and 30-/90-day mortality rates between the groups. However, a robotic approach was associated with fewer conversions compared to VATS, with a trend towards a shorter length of stay compared to an open approach. There were no differences in the 5-year overall survival rate between the matched groups (robotic 93% vs VATS 94%; P = 0.571; robotic 91% vs open 80%; P = 0.094).

Conclusions: Over a 4-year study period, there was a significant increase in robotic utilization for thymectomies and an increase in the number of hospitals performing the procedure. In a matched analysis, a robotic approach was comparable to a VATS or an open approach. Current trends demonstrate increased robotic utilization for small thymomas with excellent perioperative results.