Ninety covered tracheal stents were inserted in the TF group and 76 tracheal stents were implanted in the TS group. Tracheal fistulas were caused by malignant tumor invasion or surgical removal in 81.2% patients in the TF group. There were 47 cases of lung cancer, 2 cases of thyroid cancer and 20 cases of esophageal squamous cell carcinoma. Among lung cancer, 4 cases of lung adenocarcinoma, 1 case of lung adenosquamous carcinoma, the rest are lung squamous cell carcinoma. Tracheal stenosis was caused mainly by benign primary disease requiring tracheotomy or tracheal cannula in 82.1% patients in the TS group. There were 5 cases of lung squamous cell carcinoma, 3 cases of thyroid cancer and 4 cases of esophageal squamous cell carcinoma. The success rate of stent removal was 98.9% (89/90) in the TF group. Only one patient underwent bronchoscopic removal for retained stent pieces after failure of fluoroscopic removal. In the TS group, the success rate of stent removal was 97.4% (74 of 76 stents) and only 2 stents retained. The mean procedure time for stent implantation or removal was similar between two groups. Tumor invasion or tumor operation was the main causes for tracheal fistula, and tracheotomy or trachea cannula was the main causes for tracheal stenosis (p < 0.0001, Table 1).
Indications for stent removal and interval
Forty-seven stents implanted for tracheal fistula and 40 stents implanted for tracheal stenosis were routinely removed. Stent migration and stent intolerance were the main indications for early stent removal in both groups. Excessive granulation tissue increased as the time interval after implantation increased, with a mean interval of 142.1±25.9 days for tracheal fistula and 89.9±15.0 day for tracheal stenosis, and was the most common indications for later stent removal. Interestingly, tracheal stents were routinely removed after 107.8±9.8 days and 85.4±6.7 days, respectively, in the TF group and the TS group (Table 2). These two time points are consistent, suggesting that the stent should be removed about three months after implantation to reduce the incidence of restenosis. Six bare stents were implanted on an emergency basis in the TS group for severe tracheal stenosis. Three of these stents were removed and replaced by covered stents, with a mean duration of 8.7±3.2 days, and one stent was replaced by a tracheal T-tube.
The technical success rate of stent removal was 98.9% and 97.4%, respectively, for TF and TS groups. In the TF group, 75 stents were removed in one piece. Stent fracture and retained stent pieces were found in 14 patients and 1 stent. In the TS group, 74 tracheal stents were successfully removed, including 71 in one piece and 5 with strut fracture. Two of these stents were retained and the other three retained stent pieces were successfully removed by endoscopy (Table 3).
Recurrence of fistula or stenosis requiring stenting was the most comment complication in both groups. Two patients showed severe dyspnea in each group and underwent endotracheal intubation and mechanical ventilation. Symptoms were relieved and the endotracheal tube was removed within four hours. Massive hemoptysis occurred in two patients during stent removal in the TF group. One patient died of asphyxia caused by massive hemorrhage and bleeding ceased in the other patient after administration of pituitrin. Two patients died perioperatively in the TS group, resulting in a clinical success rate of 94.7% (72/76). One patient with tracheal stenosis after resection of esophageal carcinoma died of massive hemoptysis three days after stent removal. The other patient died of respiratory failure two days after insertion of a second stent.
Stent replacement after removal
Eight stents were replaced immediately and five stents were replaced 5 to 9 days after removal in the TF group. Replacements included Y-type tracheal stents (n=5), L-type tracheal stents (n=3), Y-shaped single-plugged stents (n=2), straight tracheal stents (n=2) and 1 combination of a large and a small Y-type tracheal stent. One patient required stenting due to proximal re-obstruction of the stent after removal of a straight tracheal stent, and the same size straight tracheal stent was implanted immediately. In TS group, 20 tracheal stents (15 straight tracheal stents and 5 Y type tracheal stents) were implanted again after removal for replacement of bare stents (n=3), or due to restenosis (n=9), migration (n=3), inadequate expansion (n=3), and intolerance (n=2).
All surviving patients were followed up. Five patients (6.0%) were lost during follow up in the TF group and four were lost (6.2%) in the TS group. Thirty-three patients were cured and 20 patients were improved in the TF group. Twenty-four patients died in the TF group, and tumor progression was the most common cause of death. One of the 24 patients died of respiratory failure due to asphyxia, and two patients died of massive hemoptysis and hematemesis. In the TS group, 29 patients were cured, 20 patients were improved and 6 patients underwent tracheotomy and tracheal T-tube implantation after stent removal (Table 4). The 0.5-, 3-, 6-year survival rates were 90.3%, 59.6%, and 36.1% for TF group, and 80.4%, 75.7%, 75.7% for TS group (Figure 3).