The medical records of all patients who underwent nonintubated VATS at National Taiwan University Hospital and National Taiwan University Hospital Yun-Lin Branch, a 3,200-bed tertiary medical center, from August 2009 to July 2012, were retrospectively reviewed. The thoracic surgical team, both surgeons and anesthesiologists, selected the cases upon review of the medical records. Patients considered appropriate for nonintubated VATS met the same criteria as for intubated VATS. Patients with American Society of Anesthesiologists (ASA) scores of greater than 3, bleeding disorders, sleep apnea, or unfavorable airway or spinal anatomy were contraindicated for nonintubated VATS in our hospital. Patient consent was obtained after explaining the type of anesthesia and the surgical procedure.

Anesthetic techniques were described previously (10,11,15). Patients were pre-medicated with fentanyl 50-100 μg intravenously (IV) and were continuously monitored electrocardiographically, along with pulse oxymetry, respiratory rate, blood pressure, body temperature, and urine output. Monitoring of central venous pressure was optional. Thoracic epidural anesthesia was performed by insertion of an epidural catheter at the T5/6 thoracic interspace to achieve a sensory block between the T2 and T9 dermatomes, and was maintained by continuous infusion of 2% lidocaine. Endtidal carbon dioxide (ETCO2) was continuously monitored by insertion of a detector into one nostril. The patient was then asked to turn him or herself into the lateral decubitus position. Sedation was then started by intravenous infusion of propofol (10 mg/mL) using a target-controlled infusion method, with incremental fentanyl injection to maintain the patient in a mildly sedated, but communicable and cooperative level (Ramsay sedation score III), responding to commands only (16).

The detailed surgical setting and procedures performed in our group were described before (10,15). Thoracoscopic lobectomy, segmentectomy, or wedge resection was performed using a 3-port method, as described by McKenna (17). In brief, the patient was positioned in the full-lateral decubitus position, with slight flexion of the table at the level of the mid-chest. The thoracoscope was placed into the seventh or eighth intercostal space in the midaxillary line. A working port was placed in the sixth or seventh intercostal space in an auscultatory triangle, and an anterior 3 cm incision was placed anteriorly in the fifth intercostal space. After collapse of the lung, incomplete fissures, pulmonary vessels, and bronchi were divided with endoscopic stapling devices. The resected specimen was removed in an organ retrieval bag through the utility incision. After staging mediastinal lymph node dissection, a 28-French chest tube was placed through the lowest incision. Rib spreading, rib cutting, and retractor use were avoided in all patients, except when conversion to thoracotomy was required. Conversion from nonintubated anesthesia to intubated general anesthesia or from thoracoscopic surgery to thoracotomy was decided by the attending surgeon and anesthesiologist. Conversion from nonintubated to intubated anesthesia was performed when epidural anesthesia was ineffective, hypoxemia persisted (S_PO₂ <80%), hemodynamic status was unstable, or intraoperative bleeding requiring thoracotomy. When conversion was indicated, the surgical wounds were sealed with transparent waterproof dressings (Tegaderm Film, 3M Health Care, Neuss, Germany) after insertion of a chest tube to re-expand the lung. A single-lumen endotracheal tube was inserted under the guidance of a bronchoscope, followed by insertion of a bronchial blocker without changing the patient’s position.

From August 2009 through July 2012, nonintubated thoracoscopic lung resection was performed on 285 patients. The demographic data of the patients are shown in Table 1. The mean patient age was 59.2 years and 107 patients (37.5%) were male. Conventional VATS was performed in 237 patients (83.2%), while needlescopic VATS was performed in 48 patients. One patient received bilateral VATS for lesions in both lungs (11). Operation procedures included lobectomy in 137 patients (48.1%), followed by wedge resection and segmentectomy.

Since the introduction of thoracoscopic surgery, intubated general anesthesia with one-lung ventilation has been considered mandatory for lung resection. To avoid intubation-related or mechanical ventilation-associated complications, thoracoscopic surgery without endotracheal intubation has recently been tried in selected patients. However, the number of patients in the published reports is small (2,6-15). This is the first report of nonintubated thoracoscopic lung resection applied in large number of patients. Our results showed that nonintubated VATS is feasible and safe for lobectomy, segmentectomy, and wedge resection.