BackgroundConversion to thoracotomy during minimally invasive lobectomy for lung cancer is occasionally necessary. Differences between video-assisted thoracoscopic surgery (VATS) and robotic-assisted thoracoscopic surgery (RATS) lobectomy conversion have not been described. MethodsWe queried The Society of Thoracic Surgeons General Thoracic Surgery Database from January 1, 2015 to December 31, 2018. Patients with prior thoracic operations and metastatic disease were excluded. Univariable comparisons with χ2 and Kruskal-Wallis tests and multivariable logistic regression modeling were performed. ResultsThere were 27,695 minimally invasive lobectomies from 269 centers. Conversion to thoracotomy occurred in 11.0% of VATS and 6.0% of RATS (P < .001). Conversion was associated with increased mortality (P < .001), major complications (P < .001), and intraoperative (P < .001) and postoperative (P < .001) blood transfusions. Conversion from RATS occurred emergently (P < .001) and for vascular injury (P < .001) more frequently than from VATS, but there was no difference in overall major complications or mortality. Mortality after conversion was 3.1% for RATS and 2.2% for VATS (P = .24). Clinical cancer stage II or III (P < .001), preoperative chemotherapy (P = .003), forced expiratory volume in 1 second (P = .006), body mass index (P < .001), and left-sided resection (P = .0002) independently predicted VATS conversion. For RATS clinical stage III (P = .037), left-sided resection (P = .041), and forced expiratory volume in 1 second (P = .002) predicted conversion. Lower volume centers had increased rates of conversion (P < .001) in both groups. ConclusionsConversion from minimally invasive to open lobectomy is associated with increased morbidity and mortality. Conversion occurs more frequently during VATS compared with RATS, albeit less often emergently, and with similar rates of overall mortality and major complications. Predictors, urgency, and reasons for conversion differ between RATS and VATS lobectomy and may assist in patient selection.
Read full abstract