INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

During the initial decade of lung transplantation (LT), airway anastomotic dehiscence was a relatively frequent surgical complication (1), but the development of improved anastomotic techniques led to a decline in its frequency (2). Nevertheless, stenosis of the bronchial anastomosis created in LT remains a problem, affecting about 8 to 15% of bronchial anastomoses, or 7 to 17% of adults undergoing LT and at risk (5). It has been reported that children undergoing LT have a higher prevalence of obstructive airway complications than do adults (16% of anastomoses at risk, 27% of recipients) (9). A lower risk (3.8 to 12%) has been reported for adult or pediatric heart-lung transplant recipients with tracheal anastomoses (10).

Postoperative donor lung airway ischemia and impaired airway healing are felt to be major factors leading to anastomotic stenosis (13). Other associated factors include the use of mattress sutures at the anastomosis, the need for mechanical ventilation beyond the sixth postoperative day, and single-lung transplantation (SLT) versus double-lung transplantation DLT (7). Numerous interventions have been used to treat posttransplantation airway stenosis in adults. These have included forceps resection or laser ablation of granulation tissue, bronchodilatation with a rigid bronchoscope or angioplasty balloon catheter, and silicone or metal-mesh stent placement (8, 14).

Although there have been several reports of obstructive airway complications following LT or heart-lung transplantation (HLT) in adults, there are relatively few published data concerning such complications and their treatment in pediatric transplant recipients (9, 15, 17). We therefore reviewed our experience in pediatric LT and HLT to determine the frequency of airway complications, to identify clinical risk factors favoring their development, and to describe the efficacy and adverse effects of various therapeutic modalities used for their treatment.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Between July 1985 and October 1997, 57 pediatric patients underwent HLT, SLT, or DLT at Children's Hospital of Pittsburgh. There were 28 HLT, three SLT, and 26 DLT procedures done as primary transplants. Six of 57 patients (four HLT, one SLT, and one DLT) required retransplantation for obliterative bronchiolitis; each of these patients underwent DLT as their retransplantation procedure. Among the study population, there were four subjects above 20 yr of age (range: 20.5 to 28.2 yr) who had received long-term pretransplant care at the Children's Hospital of Pittsburgh.

Four of 57 patients who had primary DLT (n = 1) or HLT (n = 3), and three of six patients with double-lung retransplants, died within 72 h after the procedure. The main cause of death in these patients was either primary cardiac or donor organ failure, and none of these subjects had airway anastomotic complications. Therefore, 53 patients undergoing primary transplantation (25 HLT, three SLT, and 25 DLT), and three of the patients who underwent repeat DLT, were at risk for airway complications and were included in the present retrospective study (Table 1). All 25 surviving heart-lung transplant recipients had tracheal anastomoses, whereas the single- and double-lung transplant recipients had a total of 55 bronchial and two tracheal anastomoses.

Implantation Techniques

All HLT procedures were performed with end-to-end distal tracheal anastomosis, using a running 3-0 polypropylene suture on the membranous portion and a separate running suture (or in a few cases a running telescoping suture) on the cartilaginous portion of the airway. A strip of pericardium was interposed between the trachea and the aorta. This same tracheal anastomotic technique was used for two en bloc DLT procedures.

Most other DLT procedures involved bilateral sequential SLT (18). These recipients and the three single-lung transplant recipients had bronchial anastomoses made with a running polypropylene suture on the membranous portion and interrupted horizontal mattress or figure-of-eight telescoping sutures on the cartilaginous portion of the airway (4). In some cases, end-to-end bronchial anastomosis was performed without telescoping of either the donor or recipient airway. The donor mainstem bronchi were trimmed to two cartilaginous rings above the upper-lobe branches. Bronchial artery reimplantation was not done in any transplant. Although omental wrapping of the tracheal and bronchial anastomoses was a standard practice at the beginning of our lung transplant program, this is no longer the case. All recipients with tracheal anastomoses, and two of those with bronchial anastomoses, required cardiopulmonary bypass.

Immunosuppressive Regimen

The immunosuppressive regimen used for pediatric HLT, SLT, and DLT at Children's Hospital of Pittsburgh has been described in detail elsewhere (19). In summary, methylprednisolone 10 mg/kg was given intravenously in the operating room and at 5 to 7 mg/kg/d for 24 to 72 h thereafter. Maintenance immunosuppression included tacrolimus or cyclosprine, azathioprine, and prednisone. Episodes of acute cellular rejection were treated with a 3-d burst of methylprednisolone (10 mg/ kg/d), given intravenously.

Diagnosis of Airway Complications

Airway anastomoses were visualized bronchoscopically within 48 h after transplantation. Subsequent evaluations included history and physical examination, chest radiographs, spirometry, and lung volume measurements (for cooperative, nonventilated patients), and bronchoalveolar lavage (BAL) and transbronchial biopsy (TBB) at 2 and 6 wk, and again at 3, 6, 9, 12, 18, and 24 mo after transplantation, and then annually. Patients were also assessed with bronchoscopy, BAL, and TBB whenever they had significant respiratory symptoms or chest-radiographic changes.

Airway cross-section was evaluated visually through flexible bronchoscopy. If it was decreased by more than 30%, bronchography with Iohexol (Omnipaque, Nycomed, Inc., Princeton, NJ) was done in most cases (20). Airway caliber was estimated with a computer program that compared the diameter of a catheter of known size with the airway diameter visualized during bronchography (Model DFP-200A and Model CAS-1A; Toshiba America Medical Systems, Carrollton, TX). Bronchography was useful in confirming the semiquantitative findings of flexible bronchoscopy and also in selecting the appropriate type of intervention (e.g., appropriate size stent).

Classification of Anastomotic Airway Complications

Ideally, a bronchial or tracheal anastomosis in HLT and LT should result in healing of recipient mucosa to donor mucosa by first intention without necrosis, dehiscence, ulceration, or development of stenosis (21). We considered any deviation from the expected course of healing of an anastomosis as an airway anastomotic complication. Abnormal airway findings seen during flexible bronchoscopy were classified through a modification of the system proposed by Shennib and Massard (22) (Table 2).

Airway complications were further subdivided into major and minor complications. A complication was considered major if it was related to one or more of the following: air leak, mediastinitis, dyspnea, monophonic wheezing, more than 30% obstruction of the airway cross-section (estimated through flexible bronchoscopy and/or bronchography), and a pattern on spirometry indicating airway obstruction (23). All other anastomotic airway complications were defined as minor. Colt and coworkers, who reported their experience with airway complications in 10 LT patients (eight of whom were younger than 20 yr), considered a decrease in airway cross-section of more than 25% as an indication for treatment (15).

Types of Interventions

Only major airway complications were treated. Interventions included forceps resection, laser ablation, balloon dilatation and stent placement. Treatment strategies were tailored to the type of lesion, degree of airway compromise, and clinical status of the patient.

Obstructing, eccentric, localized granulation tissue at the anastomosis was treated by forceps resection under general anesthesia through a Storz-Hopkins rigid bronchoscope (Karl Storz Endoscopy-America, Inc., Culver City, CA).

The treatment of concentric granulation tissue was based on the estimated length of stenosis. Lesions shorter than 2 to 3 mm were ablated with a (potassium titanyl phosphate) (KTP) or (neodymium: yttrium-aluminum-garnet) (Nd:YAG) laser, using either a flexible or a rigid bronchoscope. Stenotic lesions longer than 3 to 4 mm were managed with balloon dilatation under general anesthesia, using a modification of a previously described technique for management of bronchostenosis following lung transplantation (16).

Fibrotic strictures were treated with either laser ablation (lesions shorter than 2 to 3 mm) or balloon dilatation (lesions longer than 3 to 4 mm). Fibrotic strictures not managed satisfactorily with laser ablation or balloon dilatation, and bronchomalacia at an airway anastomosis or diffuse bronchomalacia, were treated with stent placement. Dumon silicone stents (Bryan Corp., Woburn, MA) were inserted according to an established technique (24). Metal stents, specifically the Schneider Wallstent (Schneider, Inc., Minneapolis, MN), the Palmaz- Schatz stent (Johnson and Johnson Interventional Systems, Warren, NJ), or the Microvasive nitonyl self-expanding stent (Boston Scientific, Boston, MA) were inserted under fluoroscopic guidance (20, 25).

Patency of the airways after each intervention was assessed with physical examination, radiographic imaging, pulmonary function tests, and flexible bronchoscopy. These evaluations were done at the regular posttransplantation visits or, more frequently, if clinically indicated.

Data Collection

The medical records of all lung and heart-lung transplant recipients for the period from June 1985 to October 1997 were reviewed, and all inpatient and outpatient evaluations, chest radiographic abnormalities, pulmonary function test results, and bronchoscopic findings were recorded. In addition, age at transplantation, sex, pretransplantation diagnosis, and use of steroids during the month preceding transplantation were noted for each patient. The transplantation operative record was also used to gather information about the type of transplantation procedure (HLT, DLT, SLT), type of airway anastomosis (tracheal versus bronchial), use of any type of wrapping around the anastomosis, or bronchial telescoping. Other factors that were assessed included: donor Pa_O2 with a fraction of inspired oxygen of 1.0, allograft ischemic time, days of mechanical ventilation after transplantation, use of corticosteroids, and hypotension during the first month after transplantation.

Statistics

Crude morbidity rates for airway complications were calculated separately for heart-lung and lung-transplant recipients with bibronchial anastomoses. The two patients who underwent en bloc DLT with tracheal anastomosis were excluded from the analysis. Two heart-lung and one double-lung transplant recipients who underwent repeat DLT and survived beyond 72 h postoperatively were counted as separate DLT patients in the analysis of clinical factors.

We assumed that similar factors contribute to the development of both major and minor airway complications, by affecting the process of healing of anastomoses to a different degree in the former as compared with the latter type of complication. Chi-square analysis was used to compare patients with and without airway complications (major and minor) according to type of transplant, sex, pretransplantation diagnosis, donor PO₂, graft ischemic time, anastomotic wrapping, bronchial telescoping, corticosteroid therapy before and after transplantation, days of mechanical ventilation, and postoperative hypotension. Categorical variables were derived from the continuous variables, using the median split of the data for each independent factor analyzed, or from divisions that had clinical meaning. In this way, the risk of developing airway complications could be compared, for example, in patients having a graft ischemic time exceeding 300 min and a graft ischemic time of 300 min or less. In comparisons of findings with small expected frequencies, Fisher's exact test was used.

Survival rates were calculated with standard life-table methods. Comparisons of survival rates of patients with major airway complications and those of patients with minor or no complications were made with the Tarone-Ware test. For the survival analysis, patients who received a repeat lung transplant were reentered as new cases.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Characteristics of patients with major airway complications, the types and locations of lesions in these patients, interventions used to treat lesions, functional results after interventions, and long-term follow-up data are shown in Table 3. Similar information for transplant recipients with minor airway complications is presented in Table 4.

Prevalence of Airway Complications

Dehiscence did not occur in any of the 53 transplant recipients included in the present investigation. None of the three single-lung transplant recipients or the two double-lung transplant recipients with tracheal anastomoses had airway complications.

One heart-lung and seven double-lung transplant recipients had anastomotic airway complications that fulfilled the criteria for major complications. All eight of these subjects had more than 30% obstruction of their airway cross-section as estimated visually with flexible bronchoscopy. We also measured the severity of obstruction at the level of the airway anastomosis in five of the eight subjects, using bronchography (Table 3, Patients 2 through 4, 7 and 9). Five of the eight patients had dyspnea (Table 3, Patients 1 through 3, 5 and 7), whereas chest auscultation revealed monophonic wheezing in all cases. An obstructive pattern of the maximum expiratory flow-volume (MEFV) curve was noted in five of six subjects who were able to undergo spirometry (Figure 1).

View larger version (18K): [in this window] [in a new window]   Figure 1.   MEFV curves for patients with anastomotic airway complications before (solid line) and after (broken line) completion of all airway interventions. Patient 2 did not have any pulmonary function testing after placement of a Dumon stent.

Major anastomotic airway complications tended to be less frequent in HLT (one of 25 patients [4%]) than in DLT or SLT (seven of 28 patients [25%]) (p = 0.05). One of seven double-lung transplant recipients with major anastomotic airway complications developed anastomotic stenosis after repeat DLT. Overall, nine of 55 bronchial anastomoses (16%) in double-lung and single-lung transplant recipients were affected by major anastomotic complications. There was also one double-lung transplant recipient with a major nonanastomotic airway complication (Table 3, Patient 8; a small hole in the trachea).

Minor anastomotic complications were identified in three of 25 heart-lung transplant recipients (12%) and in four of 28 single-lung and double-lung transplant recipients (14%). Minor complications were seen in five of 55 bronchial anastomoses (9%) in patients with lung transplants. Airway complications (major and minor) were identified within the first posttransplant month in heart-lung transplant recipients (range: 1 to 2 mo) and single-lung or double-lung transplant recipients (range: 1 to 6 mo).

Correlation of Clinical Factors with Frequency of Anastomotic Complications

Although not statistically significant (p = 0.12), there was a trend toward more airway complications (major and minor) in single-lung or double-lung transplant recipients (39%) than in heart-lung transplant recipients (16%). Therefore, we studied the association of clinical factors with the development of airway complications separately for each group of patients.

All 25 heart-lung transplant recipients were included in the statistical analysis. There was no difference in the mean age at transplantation between heart-lung transplant recipients with airway complications (9.48 ± 4.74 yr of age [mean ± SEM]) and those without airway complications (13.2 ± 6.14 yr of age) (p = 0.34). No statistically significant correlation was found between most clinical factors and subsequent development of airway complications. Heart-lung transplant recipients who received more than 50 mg/kg methylprednisolone during the first month after transplantation were more likely to have anastomotic complications than those who received less (four of 12 [33%] versus none of 13 [0%], p = 0.04). Because only two patients in the HLT group received corticosteroids during the month preceding transplantation, use of corticosteroids before transplantation was not included in the statistical analysis.

Correlation between clinical factors and frequency of airway complications was studied in 29 subjects with single-lung or double-lung transplants and bronchial anastomoses. There was no significant difference (p = 0.12) in the mean ages of lung transplant patients with and without anastomotic airway complications (13.6 ± 3.75 and 14.9 ± 3.51 yr of age, respectively). No statistically significant correlation was identified for any other clinical factors. Neither pre- nor posttransplantation administration of corticosteroids was related to the incidence of airway complications. Lung transplant recipients whose anastomoses were not telescoped seemed to have an increased frequency of anastomotic airway complications, but the difference was not statistically significant (75% without telescoping of the airways had complications, versus only 32% with telescoping, p = 0.14). Only one lung transplant patient underwent wrapping of bronchial anastomoses, and this factor was therefore not included in the statistical analysis.

Types of Airway Complications and Interventions Used

Fibrotic stricture of the tracheal anastomosis was the most common airway complication in heart-lung transplant recipients. Patient 1 (Table 3) developed two concentric webs of granulation tissue at the level of the anastomosis, leading to a reduction of the airway cross-sectional area of approximately 80%. Four sessions of KTP laser ablation decreased the stenosis to less than 20% of the airway cross-sectional area.

Of 11 single-lung and double-lung transplant recipients with major (Table 3) or minor (Table 4) anastomotic airway complications, five developed fibrotic strictures, four developed granulation tissue, and two developed malacia at one or both airway anastomoses. Airway obstruction by granulation tissue preceded the development of a fibrotic stricture in only one of five double-lung transplant recipients (Table 3, Patient 3).

Of seven double-lung transplant recipients with major anastomotic airway complications (Table 3), two had satisfactory relief of airway obstruction with forceps resection of granulation tissue, balloon dilatation, or laser ablation only. Nevertheless, stent placement was necessary in four of the seven subjects (two Dumon stents, two Palmaz-Schatz stents, one Schneider Wallstent, and two Microvasive stents) (Figure 2). Patient 5 (Table 3) developed severe stenosis of the left bronchial anastomosis, but expired from fulminant adenoviral pneumonia before any airway therapy could be given.

View larger version (51K): [in this window] [in a new window]   Figure 2.   Bronchoscopic view of stenotic right mainstem bronchus anastomosis in Patient 7, before (A) and after (B) laser ablation (twice), balloon dilatation (twice) and placement of a Palmaz-Schatz stent.

Patient 6 (Table 3) was noted to have obstruction of the left mainstem bronchus by a flange of telescoped and invaginated donor cartilage. The cartilaginous flange was successfully ablated with the KTP laser. Severe stenosis of the right bronchus intermedius was identified in Patient 7 (Table 3) and was treated with several balloon catheter dilatation procedures and insertion of a Palmaz-Schatz stent.

A major nonanastomotic airway complication was noted in one double-lung transplant recipient (Table 3, Patient 8). He had undergone right upper lobectomy before being considered for lung transplantation. Native right pneumonectomy was especially difficult because of fibrous tissue at the site of the lobectomy. Less than 1 wk after transplantation, pneumothorax and pneumomediastinum were noted. On reexploration, a small defect was found in the lower third of the membranous portion of the trachea, and was thought to have come from a cautery injury during the dissection. The defect was repaired with an intercostal muscle patch.

Functional Results

Five of seven patients (Table 3, Patients 1, 3, 4, 6, and 7) who underwent interventions for major anastomotic airway complications had an appreciable improvement after treatment, with less dyspnea, better air exchange on auscultation, and improved airway cross-sectional area on bronchoscopy. In addition, most patients showed improvements in pulmonary function test results. Figure 1 illustrates maximum expiratory flow-volume (MEFV) curves before and after airway interventions for six patients who were able to undergo spirometry. Table 5 shows the changes in TLC and in spirometric indices for five of these six subjects after completion of all interventions. Posttreatment spirometry was not done for Patient 2. 

Before treatment, Patients 2, 4, and 7 exhibited an obstructive pattern on pulmonary function testing. Patients 3 and 9 had a biphasic MEFV curve with upward concavity and a decreased FVC, FEV₁, and FEV₁/FVC, consistent with asynchronous emptying of fast and slow lung units. One of these patients (Patient 9) had moderate bronchomalacia of the right anastomosis, and lymphocytic bronchitis/bronchiolitis documented by transbronchial biopsy. Patient 6 had a purely restrictive picture on pulmonary function testing.

After treatment, improvement in expiratory flow function in Patient 4 was due to an increase in TLC as well as a decrease in airway obstruction. In Patient 7, airway obstruction also dramatically improved after the completion of airway interventions.

Following airway treatment, the biphasic descending limb of the MEFV curve of Patient 3 changed to a straight line. In Patient 9, on the other hand, the descending limb of the MEFV curve maintained an upward concavity (Figure 1) despite treatment of right mainstem bronchomalacia with stent insertion. Therefore, in Patient 9, small-airway disease due to lymphocytic bronchitis and bronchiolitis was clinically more important than central airway obstruction resulting from bronchomalacia. Patient 6 had also alleviation of dyspnea, but this was mainly the result of an increase in TLC (Table 5).

Complications of Airway Interventions and Their Management

Some mild mucosal bleeding accompanied balloon bronchodilation. Patient 7 (Table 3) required multiple sessions of laser ablation for granulation tissue at the bronchial anastomoses. Subsequently, bronchomalacia developed at both anastomoses, most likely as a result of injury to the bronchial cartilage.

In one patient (Patient 2), a Dumon stent migrated into the trachea and was subsequently removed via a rigid bronchoscope. At the same time, the patient was in critical condition from Burkholderia cepacia pneumonia and sepsis. Stent migration may have contributed to his death. Only one of four metal stents migrated to the trachea and required replacement with a stent of the same type (Microvasive).

Patient 7 (Table 3) had Palmaz-Schatz stents in the right mainstem bronchus (Figure 2) and bronchus intermedius, as well as a Schneider Wallstent in the left mainstem bronchus. All three stents were complicated by the growth of granulation tissue infiltrating the wire mesh of the stent. The granulation tissue was treated successfully with a series of three balloon dilatations. There were no cases of fractured metal stents.

Survival of Patients with and without Major Airway Complications

No patient was lost to follow-up. Five-year survival for lung transplant recipients with major airway complications was 30%, as compared with 40% for lung transplant recipients with minor or no airway complications (p > 0.05). Five-year survival for heart-lung transplant recipients with minor or no airway complications was 65%. Only one heart-lung transplant patient had a major airway complication, and he survived for 12 mo after transplantation.

The cause of death of patients with major anastomotic complications (Table 3) was not directly related to the specific airway complication. Causes of death included bronchiolitis obliterans complicated by acute bronchopneumonia (Patients 1 and 3); multisystem organ failure associated with Burkholderia cepacia bacteremia (Patient 2); fulminant adenoviral pneumonia (Patient 5); severe acute cellular rejection (Patient 6); and diffuse alveolar damage resulting from reperfusion injury and pulmonary hemorrhage (Patient 8). As noted earlier, migration of a bronchial Dumon stent may have contributed to the death of Patient 2. The causes of death of patients with minor airway complications (Table 4) included obliterative bronchiolitis (Patients 2, 4, and 6) and disseminated Aspergillus fumigatus infection (Patient 5).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Airway complications are not uncommon after pediatric and adult HLT and LT (5). We divided our patients into those with major and minor airway complications, and our results are similar to those reported by others (10) in that the frequency of major anastomotic complications in pediatric heart-lung transplant recipients was low (4%). In contrast, our pediatric lung transplant recipients were affected by major anastomotic problems more frequently than reported in adult recipients (5). None of our subjects experienced anastomotic dehiscence. We also report that an appreciable number of heart-lung or double-lung transplant recipients had abnormal healing of anastomoses without significant functional impairment (minor airway complications). With the possible exception of high-dose corticosteroids administered during the first month after transplantation in heart-lung transplant recipients, no other clinical factor correlated with a high frequency of airway complications. Fortunately, most complications were managed safely and successfully through a combination of treatment modalities.

In a previous report of airway complications in lung transplant recipients, the majority of whom (eight of 10) were under 20 yr of age, a 25% or greater reduction in airway cross-sectional area was considered severe enough to require treatment (15). In the present investigation, a decrease of greater than 30% was accompanied by dyspnea, a monophonic pattern of the MEFV curve, or an obstructive pattern of the MEFV curve on spirometry, or by all three. The previously reported findings suggest that obstruction exceeding 30% of the cross-sectional area of the airway is appreciable mechanically, and that the patient will most likely benefit from an intervention to relieve the obstruction.

Fatal ischemic necrosis and dehiscence of tracheal anastomoses was a significant problem during the initial postoperative stages of DLT (13). As more experience was gained and the technique of bilateral sequential DLT was introduced, dehiscence of airway anastomoses became an infrequent complication (9). At Children's Hospital of Pittsburgh, we have not experienced any cases of dehiscence of the airway anastomoses since the beginning of the transplant program. Trimming of the donor's main bronchi to two cartilaginous rings above the upper lobe branches may have reduced the risk of ischemic injury and subsequent dehiscence of anastomoses (2, 6).

Ischemia of the airway anastomosis secondary to interruption of bronchial circulation is thought to be a major factor in the development of anastomotic airway complications (13). Early in the posttransplant period, the perfusion of donor airways is provided by collateralization between the pulmonary vascular bed at the capillary level and the distal bronchial circulation (26). The median time after lung transplantation to the development of airway complications in the present study was 1 mo. This suggests that early ischemia of transplanted airways could have detrimental effects on the healing of bronchial anastomoses.

HLT, on the other hand, is associated with fewer anastomotic airway complications than is en bloc DLT or bilateral sequential DLT (10), most likely because the donor airway blood supply is preserved via coronary collaterals to the region of the tracheal anastomosis (27). The reported difference in frequency of major airway complications in HLT and LT patients was also seen in the present investigation, although it did not achieve statistical significance; with only 4% of tracheal anastomoses versus 16% of bronchial anastomoses being complicated by abnormal healing. This difference was maintained even when minor complications were considered (16% of tracheal anastomoses versus 25% of bronchial anastomoses).

The prevalence of symptomatic tracheal stenosis at the level of the anastomosis for heart-lung transplant recipients in our series was similar to or lower (4%) than that reported in previous studies (3.8 to 12%) (10). Of concern was that the prevalence of symptomatic anastomotic complications in lung transplant recipients (25% of patients or 16% of bronchial anastomoses) in our patients was higher than that at adult lung transplant centers (7 to 17% of patients, or 8 to 15% of bronchial anastomoses) (5). This remarkable difference has also been reported by other pediatric transplant centers. At St. Louis Children's Hospital, airway complications occurred in 27% of lung transplant patients (or 16% of cases of anastomoses at risk) (9). Colt and coworkers in the Joint Marseille- Montreal Lung Transplant Program reported their experience with airway complications in 24 lung transplant recipients ranging in age from 7 to 55 yr (15). Despite this wide age range (children and adults), eight of 10 patients who developed bronchial stenosis were younger than 20 yr of age. We speculate that when there is an airway stenosis resulting from abnormal healing of the anastomosis, the smaller diameter of mainstem bronchi in children predisposes them to more functional problems than occur in adults.

Telescoping of the donor into the recipient bronchus (or vice versa if the recipient bronchus is larger) at the airway anastomosis may prevent direct communication of the airway with the mediastinum if donor airway necrosis occurs (4). In the present series, 75% of lung transplant patients without bronchial telescoping experienced airway complications, as opposed to 32% of those who had bronchial telescoping. Nevertheless, only four patients had end-to-end bronchial anastomoses, as compared with 25 patients on whom the telescoping technique was utilized, which makes any comparison invalid.

Despite the variety of techniques (wrapping of the anastomosis with omentum or pericardium, anastomosis of the donor's bronchial circulation with the recipient's internal thoracic arteries, telescoping technique) designed to decrease the frequency of airway complications in LT (28, 29), the lung transplant group in San Diego has suggested that simple end-to-end bronchial anstomosis may be adequate to prevent airway stenosis. On the contrary, intussusception of the donor bronchus within the recipient bronchus, and vice versa, may promote the development of late stenosis at the anastomosis (6).

High doses of corticosteroids (> 2 mg/kg/d of methylprednisolone), administered preoperatively may impair the healing of airway anastomoses (2), whereas low doses (< 15 mg/d of methylprednisolone) could promote healing by controlling rejection and reperfusion injury (7, 30). Heart-lung transplant recipients in the present study who received methylprednisolone in a dose equivalent to more than 50 mg/kg (> 1.5 mg/kg/d) during the first month postoperatively had an increased frequency of anastomotic airway complications. However, because of the small number of HLT patients with complications, the clinical significance of this finding is uncertain. None of the other factors examined correlated with the frequency of anastomotic complications.

Several different types of airway complications were noted at the level of the anastomosis in subjects in the present study. It is of interest that in most of the cases (six of eight patients), anastomotic strictures were not preceded by the growth of granulation tissue. Bronchial obstruction from a flange of invaginated cartilage was noted in one double-lung transplant recipient. This type of obstruction is a complication of telescoping of an airway anastomosis (8). It has been proposed that a modified mattress suture technique achieves the desired telescoping of the anastomosis and also avoids the potential creation of an obstructing flange of cartilage (7). In our patient, the obstruction resolved successfully after ablation with a KTP laser.

Severe diffuse stenosis of the right bronchus intermedius developed in one patient (Table 3, Patient 7). Stenosis distal to the bronchial anastomosis has been described previously in pediatric transplant recipients (11). It has been postulated that ischemia may lead to any or all of ossification, calcification, or fragmentation of bronchial cartilage, which in turn may be responsible for similar alterations in distal bronchial cartilage, leading to airway stenosis (31).

For the management of several different types of anastomotic complications, we used various interventions, including forceps resection of granulation tissue, balloon catheter dilatation, laser ablation, and stent placement. Forceps resection proved to be safe and useful for the excision of eccentric, localized granulation tissue. None of three patients who underwent debridement of granulation tissue by forceps suffered airway perforation.

Balloon catheter dilatation was effective in treating stenosis caused by granulation tissue or fibrotic stricture (Table 3, Patients 4, 6, and 7). However, the results of this intervention were usually short-lived, and repeat balloon dilatations were necessary to maintain patency of the airway. For this reason, bronchial anastomotic stenoses that were shorter than 2 to 3 mm were managed with laser ablation. The only complication noted after balloon dilatation was minimal mucosal bleeding. Potential complications discussed in the literature are partial or complete rupture of the airway, with resulting pneumothorax, pneumomediastinum, or mediastinitis (32).

The KTP or Nd:YAG laser, with light energy delivered through flexible quartz fibers via a flexible or rigid bronchoscope, was also effective in ablating cicatricial granulation or scar tissue (length less than 2 to 3 mm), with minimal bleeding (33). The main complication with laser photoresection was bronchomalacia, probably as a result of thermal injury to the bronchial cartilage. The only effective intervention for the management of bronchomalacia is stent placement, with its known technical challenges and possible complications. Therefore, multiple sessions of laser ablation should be avoided, and balloon dilatation should be used as an adjuvant for the management of recurrent development of granulation tissue or fibrotic stricture at airway anastomoses.

Although silicone stents have long been used to treat benign and malignant tracheobronchial obstruction, certain disadvantages of these devices should be noted (14). First, such stents may migrate, particularly if the area of stenosis is short and conical rather than long and consistent in internal diameter. Second, the stent may be obstructed by retained secretions or granulation tissue at either end. Third, reepithelialization of the device is not possible, and stenting therefore impairs mucociliary clearance.

The problem of stent migration has been alleviated by the use of self-expanding (Schneider Wallstent; or Cook-Gianturco, William Cook, Inc., Bloomington, IN) or expandable (Palmaz-Schatz) metal devices. Because the metallic filaments of these devices push against the airway wall, migration of the stent is not a common problem, although this complication occurred in one of our patients (Table 3, Patient 9). This expansile force, along with epithelialization of the metal mesh of the stent, make its removal difficult if not impossible.

Cryotherapy or laser therapy has been applied to manage granulation tissue, which occasionally grows within a wire-mesh stent (14). One of our patients (Table 3, Patient 7) had exuberant granulation tissue infiltrating a wire mesh stent in her bronchus intermedius. We used a series of balloon dilatations rather than laser therapy to reestablish an adequate airway. In this way we avoided possible destruction of the stent and perforation of the airway wall.

Other, more serious complications than formation of granulation tissue through the mesh of a metal stent have been reported in the literature, such as perforation of the airway wall and adjacent vascular structures (34). At the time of writing of this report, Patients 7 and 9 (Table 3) had metal stents in place for 4.5 and 2.5 yr, respectively, without any of the serious adverse events described here.

In summary, because of the technical difficulties in stent insertion and the possible short- and long-term adverse effects of stenting, stents should be used as a last resort in HLT and LT, when all other measures have failed to alleviate obstruction or when the nature of the obstructing lesion (e.g., bronchomalacia) precludes any other type of intervention. The choice of the type of stent should depend on the experience of the lung transplantation program in which the stenting is done.

With the application of the previously described therapeutic modalities, five of seven heart-lung or lung transplant recipients treated in our study for major anastomotic complications had appreciable improvement by subjective and objective criteria. Pulmonary function testing before and after treatment of airway obstruction was helpful in objectively assessing the results of each intervention. Undoubtedly, airway complications made the care of the specific patients more challenging. Nevertheless, survival of patients with airway complications was not affected significantly, and it was similar to that for patients without anastomotic problems. The same finding has been reported by adult lung transplant centers (8).

Conclusions

Airway obstruction at bronchial anastomoses is more frequent in pediatric than in adult lung transplant recipients. As with adult patients, pediatric heart-lung transplant recipients have a low frequency of airway complications. With a combination of diagnostic and treatment modalities, satisfactory lung function can be achieved in the majority of lung transplant recipients who are affected by such complications, without a dismal effect on patient survival.