Total Intravenous Anesthesia and Endotracheal Oxygen Insufflation for Repair of Tracheoesophageal Fistula in an Adult

Abstract

Atracheoesophageal voice prosthesis facilitates speech rehabilitation after laryngectomy [1,2]. We describe the anesthetic management of a patient who required repair of a large tracheoesophageal fistula around the voice prosthesis. The requirements for access to the upper trachea excluded the use of a tracheotomy tube. Our technique of total intravenous anesthesia, spontaneous ventilation, and endotracheal O2 insufflation provided excellent operating conditions while ensuring adequate anesthesia depth, oxygenation, and patient safety. Case Report A 38-yr-old, 58-kg female had a tracheoesophageal voice prosthesis (Provox Trademark; Atos Medical, Horby, Sweden) [3] inserted 12 mo after total laryngectomy. This was complicated by a 10 times 8-mm tracheoesophageal fistula on the posterior tracheal wall producing aspiration of food and saliva. Repair of the fistula under general anesthesia was scheduled. The patient was premedicated with temazepam 20 mg orally. In the operating room, the patient was monitored with electrocardiography, noninvasive arterial pressure, and pulse oximetry. She was preoxygenated, and anesthesia was induced with fentanyl 75 micro gram and propofol 150 mg intravenously, followed by an infusion of propofol at 10 mg centered dot kg-1 centered dot h-1. The proximal trachea was sprayed with 1 mL of 10% lidocaine. A 12 Fr suction catheter was then passed into the trachea until about 2.5 cm above the carina. A gas sampling tube for capnography was attached along the insufflation catheter, ending approximately 2 cm further distally. Humidified oxygen was insufflated through the suction catheter at 5 L/min. The right radial artery was cannulated for arterial pressure monitoring and repeated blood sampling. During surgery, the patient was positioned 15 degrees head down, and surgeons were warned of the potential for aspiration and airway obstruction. Surgery lasted 100 min with excellent operating conditions Figure 1. Spontaneous breathing was maintained throughout the procedure. The total dose of propofol was 1470 mg, including four 30-mg boluses to abolish coughing. The O2 insufflation was increased to 8 and 12 L/min because of rising ETCO2Figure 2. Despite moderate respiratory acidosis, the patient remained stable hemodynamically. She was awake 9 min after the propofol infusion was stopped. The postoperative course was uneventful. There was no evidence of aspiration pneumonitis or other pulmonary complications. The patient was discharged 1 wk later.Figure 1: Photograph showing the unobstructed surgical field with insufflation catheter and expired gas sampling tube in situ.Figure 2: Graphic representation of respiratory variables during the perioperative period. Changes in arterial oxygen saturation (SaO2), ETCO2, and respiratory rate (RR) are shown. The time of arterial blood gas sampling is indicated by the arrow underneath individual results of PaO2, PaCO2, and arterial pH. The start and end of general anesthesia (GA), as well as times of changes in oxygen insufflation rate are indicated. Readings after 120 min were taken in the recovery room.Discussion Tracheoesophageal speech effected via a valve prosthesis placed through a surgical fistula between the posterior wall of the trachea and the pharyngoesophageal segment is a standard method of vocal rehabilitation after total laryngectomy. The silicone Provox Trademark prosthesis is a new tracheoesophageal valve with an outer shaft diameter of 7.5 mm. This valve is larger than other devices and allows good quality, effortless speech because of its low resistance [3]. However, when resistant infection requires prosthesis removal, the patient may be left with a large tracheoesophageal fistula necessitating surgical closure under general anesthesia. The major anesthetic challenge in this patient was to allow surgical access of the proximal trachea through the tracheotomy stoma. Surgery partially occluded the airway and restricted the choice of airway devices. A tracheotomy mask or tube would severely impair surgical access. Jet ventilation through a small catheter was acceptable, but this would have required a more complicated setup as well as muscle paralysis and introduced a greater risk of pulmonary barotrauma. Total intravenous anesthesia with spontaneous ventilation and endotracheal O2 insufflation provided controllable anesthesia depth, adequate oxygenation, and minimal risk of pulmonary barotrauma. Should apnea or airway obstruction have occurred, the catheter could have been used to guide tracheal intubation. The surgical requirement for access to the airway limited our choice of anesthetic techniques. Total intravenous anesthesia with propofol was preferable to volatile drugs because it eliminated theater pollution and difficulties in titration and delivery of volatile anesthetics. Total intravenous anesthesia with propofol allows rapid changes in anesthesia depth and facilitates rapid recovery [4]. Propofol can be used in spontaneously breathing patients, despite moderate respiratory depression [5,6]. The small, flexible catheter used for endotracheal O2 insufflation provided excellent oxygenation yet did not interfere with surgery or the patient's breathing Figure 1. Endotracheal O2 insufflation improves oxygenation by increasing the inspired O2 concentration. It provides adequate oxygenation even in apneic patients, in whom mechanisms of gas mixing include cardiogenic oscillation, lung vibration, molecular diffusion, and convection at the insufflation site [7,8]. Endotracheal O2 insufflation also enhances CO2 elimination, mainly by decreasing dead space. This, however, is inadequate in apneic humans unless O2 is insufflated at 2-3 cm down both main bronchi at rates up to 45 L/min. In spontaneously breathing patients with acute respiratory failure, however, the reduction in dead space and hypercarbia by continuous tracheobronchial O2 insufflation is proportional to the flow rate between 1 and 8 L/min [9]. Hypercarbia in our patient was decreased by a higher flow of O2 insufflation Figure 2. By fixing the gas sampling tube with its tip distal to the insufflation catheter, sampling of expired gas diluted by insufflated O2 was minimized. The sampled gas provided ETCO2 that correlated closely with PaCO2Figure 2. Airway secretions blocked the sampling tube on two occasions, and surgery was interrupted briefly until the problem was corrected. The use of endotracheal O2 insufflation could result in several complications. Tracheal mucosal desiccation was minimized by using humidified gas. Previous studies revealed no mucosal damage even after high (45 L/min) gas flows [7]. Obstruction of expiratory gas flow could cause barotrauma and gas embolism, but this was unlikely in our patient with terminal tracheotomy. We also minimized the risk of barotrauma by insufflating O2 above the carina and at flow rates below 12 L/min. The potential problem of airway soiling was minimized by positioning the patient head-down and by frequently suctioning the surgical field. We thank Professor C. A. Van Hasselt and Dr. T. A. Buckley for their useful advice and review of the manuscript.