Short-Term Deployment of Self-Expanding Metallic Stents Facilitates Healing of Bronchial Dehiscence

Despite several decades of experience with lung transplantrelated airway complications, bronchial dehiscence remains a disastrous complication. Most cases are encountered during the early postoperative period, are difficult to treat, and are associated with high mortality (1-3). The initiating event leading to the complication is probably inadequate vascularization at the anastomotic site. Despite meticulous suturing, telescoping techniques, or covering the anastomosis with pedicles of vital tissues, dehiscence may still occur (1). Clues to the recognition of a significant airway dehiscence are shortness of breath, inability to wean from mechanical ventilation, pneumomediastinum, subcutaneous emphysema, pneumothorax, or persistent air leak. A drop in the FEV^sub 1^ could provide a clue to this complication in extubated patients. Computed tomography has a high degree of sensitivity and specificity for depicting bronchial dehiscence (4), while flexible bronchoscopy confirms the diagnosis. Tracheobronchial stents effectively restore airway patency in selected patients with large airway obstruction and palliate symptoms of fistulae in a relatively noninvasive fashion (5). Silicone stents have been used in a variety of endobronchial conditions (6, 7); however, their tendencies to migrate, obstruct lobar orifices, and interfere with normal mucociliary clearance are significant disadvantages. Self-expanding metallic stents are an alternative to silicone stents, with the advantage of less cumbersome placement via the flexible bronchoscope; they have been used in a variety of applications to maintain airway integrity and palliate respiratory symptoms (5, 8). A drawback to metallic stents in most circumstances is their propensity to cause excessive granulation tissue formation, which can lead to airway obstruction (9-11). However, this property of promoting granulation tissue may be desirable in patients with airway disruption. Based on this reasoning, we have used uncovered self-expanding metallic stents in patients with bronchial dehiscence after lung transplant with good results. Some of the results from this experience have been previously reported in abstract form (12, 13).

METHODS

We reviewed the course of all lung transplant recipients at the Cleveland Clinic Foundation with bronchial dehiscence treated using self-expanding metallic stents. Indications for treatment included the presence of Grade III and IV bronchial dehiscence as defined in Table 1 (3). All patients were treated with Ultraflex stents (Microvasive; Boston Scientific, Watertown, MA). The Ultraflex system consists of a flexible delivery catheter preloaded with an expanding metallic stent. Flexible bronchoscopy was performed on all patients to confirm the diagnosis. The length of the stent was chosen so that it completely bridged the disruption and overlapped normal mucosa at either end. The stent diameter was chosen to match the most normal-appearing portion of the recipient's native bronchus as measured on computed tomography. Stents were placed as described previously (14). Satisfactory deployment of the stent was confirmed by bronchoscopy and by chest radiography. Surveillance bronchoscopies were performed within 3 days of deployment and thereafter if complications were suspected, as dictated by symptoms, persistent air leak, or spirometric indices. During our initial experience (n = 4), we did not plan to remove the stents, yet ended up removing them piecemeal as an adjunct to management of stent-related ingrowth of scar. Subsequently, stents were electively removed in the interval between adequate healing of the dehiscence and before epithelialization of the stent.

RESULTS

From January 1995 to June 2004, 189 single and 118 double lung transplants were performed in our institution, totaling 425 at-risk bronchial anastomoses. Seven (1.6%) cases of Grade III or IV bronchial dehiscence were encountered and treated with self-expanding metallic stents. The interval between lung transplant and diagnosis of bronchial dehiscence was 29.1 ± 18.5 days. Open surgical repair was deemed too risky or failed in all seven patients.

All patients presented with respiratory distress, and three required mechanical ventilation. Bronchial dehiscence progressed despite intercostal muscle flap bronchoplasty in patient 5 and bronchial omentopexy in patient 3 before stent insertion (Figure 1). The mean improvement in FEV^sub 1^ after insertion of self-expandable metallic stent was 455 ± 233 ml (range, 290-800 ml) in four patients for whom pulmonary function test could be obtained (Table 2). Six patients reported symptomatic improvement, including three who were weaned off the mechanical ventilator. Anastomotic healing was present as early as 1 week. At follow-up bronchoscopy, the stent position was found to be satisfactory in all cases but one. Migration of the stent into the mediastinum was encountered at a repeat bronchoscopy on Day 3 in patient 5. The stent was easily removed and replaced with a stent of larger size (Figure 2). In patients 2, 3, and 4 it was not our original intention to remove the stent, yet we had to carry it out piecemeal while treating ingrowth of granulation tissue using a neodymium:yttrium aluminum garnet laser under general anesthesia. The average time to in-stent stenosis was 188 days. In patients 5 and 6 stents were removed electively after complete healing of the anastomosis and before epithelialization of the stent. Both stents were easily removed under local anesthesia using flexible bronchoscopy in the outpatient department. Two cases (patients 3 and 4) were complicated by symptomatic bronchomalacia noted after stent removal for in-stent stenosis, and both required repeat stent insertion. There were no procedurerelated deaths.

The mean follow-up to date is 410 days, with two deaths. Patient 1 died early (10 days after stent deployment) due to pulmonary embolism, and patient 4 died late (321 days after stent deployment) due to multi-organ failure (Table 2). The remaining five patients are alive and doing well with normal graft function, no evidence of obliterative broncholitis by spirometry, and patent airways during periodic surveillance bronchoscopies (routine up to 1 year at our institution).

DISCUSSION

Necrosis at the suture line frequently leads to airway dehiscence. This has been largely attributed to ischemia of the donor bronchus. The bronchial circulation is not reestablished during transplantation, and re-arterialization from the recipient circulation generally requires 2 weeks to develop (15). Other factors invoked to explain airway complications include the donor lung ischemic time and number of rejection episodes (16-19). Multiple operative techniques have been used to minimize airway dehiscence. These include bronchial artery anastomosis, shortening of the donor bronchial stump to two or less cartilaginous rings proximal to the upper lobe takeoff, reinforcing the anastomosis with a vascularized tissue pedicle such as omentum, or intercostal muscle pedicle flap (20-22). Controversy persists regarding the optimal anastomotic technique (23).

Despite advances, airway dehiscence presents a devastating problem for a few patients. Chest tube placement to achieve complete reexpansion of the lung is considered when appropriate. Placement of silicone stents has been tried (24), but is not widely accepted because the vigorous force required for stent placement can increase the defect size. At most centers, the majority of patients with high-grade bronchial dehiscence have succumbed to death, often from bacterial sepsis (8, 25), while those with partial bronchial dehiscence have been treated conservatively. Late airway strictures developed in the majority of the latter patients requiring frequent dilatations (24, 26). Bronchial dehiscence has rarely been successfully managed surgically (2), although it has been attempted with the local application of platelet-derived wound healing factor in two cases (1) and successful use of a-cyanoacrylate glue in one case (27).

From our experience it appears that temporary placement of uncovered self-expanding metallic stents in patients with life-threatening post-lung transplant bronchial dehiscence may be of value. These self-expanding wire stents mold to the shape of the bronchus and provide a constant outward radial force, maintaining airway patency. These devices are easily deployable under direct vision or fluoroscopy with the use of a flexible bronchoscope, eliminating the need for manipulation with rigid instruments. This is essential to avoid any further trauma to the already vulnerable anastomotic site. It is well documented that self-expanding metallic stents promote excessive granulation tissue formation that grows through the interstices of the stent. The evaluation of stents in both experimental animal models and human atherosclerotic coronary arteries demonstrates that arterial responses resemble wound healing: early thrombus formation with acute inflammation, a granulation phase, and ultimately smooth muscle proliferation and matrix deposition (28). Similarly, self-expanding metallic stents provide a platform for healing of the dehiscence and, in time, peribronchial soft tissue support grows in to cover the defect. In our limited sample we have observed that these stents can be easily removed by 6-8 weeks after adequate healing and before epithelialization. The mean time to stent removal in our cases was 37.5 days (range, 21-54 days). We only selected the uncovered type of the stent, to allow drainage of the mediastinal and airway secretions as well as for unobstructed ventilation of the upper lobes. There is also high likelihood that the polyurethane coating of the covered stents may be colonized with bacteria, further interfering with anastomotic healing (29). This may be why the covered version of self-expanding metallic stents in patients with bronchial dehiscence was unsuccessful in a previous report (8).

In conclusion, temporary use of uncovered self-expanding metallic stents provides a safe and noninvasive option in the management of life-threatening bronchial dehiscence after lung transplant. More than two thirds of these patients, whose prognosis was thought to be grim on making the diagnosis of bronchial dehiscence, are alive and doing well with normal graft function, with a mean follow-up period of approximately 14 months.

Conflict of Interest Statement: None of the authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

© 2005 American Thoracic Society Provided by ProQuest LLC. All Rights Reserved.