INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Stenting of the central airways has become an accepted treatment in the management of a variety of benign and malignant diseases causing stenosis or obstruction of the central airways (1, 2). Stenting can be used alone, or after application of other bronchoscopic interventions for airway reopening such as Nd: YAG laser photoresection, airway dilatation, cryotherapy, electrocoagulation, or others. Major indications for airway stenting (AS) include dyspnea and (postobstructive) infections. Symptomatic and functional benefit (and, in selected cases, even increased survival) is obtained in the majority of patients (3). Nonmetallic stents such as the dedicated silicone stent designed by Dumon (2) are most often used (4, 5). Reported complications include bleeding, stent displacement, sputum retention and desiccation, and infection. To our knowledge, however, there are no prospective data on the occurrence and nature of infectious changes or complications occurring after AS, which is probably the reason why the systematic use of prophylactic antibiotherapy after stenting or other bronchoscopic interventions still is debated (6). On the other hand, because stenting may be used in order to treat postobstructive infections, stenting may have beneficial effects on already infected airways and lungs. Furthermore, even in the absence of clinical signs of infection, obstruction of the airways by tracheobronchial disorders may lead to colonization of the distal airways (which are usually sterile [7, 8]) by a variety of potentially pathogenic microorganisms (PPM) (9). Because airway stents may alter local defense mechanisms in the airways (e.g., impaired mucociliary clearance and expectoration) (12), they may in themselves lead to distal airway colonization or infection.

In order to study the effects of AS alone or in combination with other interventions including Nd:YAG laser photoresection and balloon dilatation, we prospectively analyzed quantitative bronchoscopic protected specimen brush (PSB) samples obtained distally from airway obstructing lesions before, and 3 to 4 wk after stent insertion.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Selection

Fourteen patients (age 65 ± 17 yr, range 21 to 82) with central airway obstruction referred for interventional bronchoscopic treatment including AS were included in a 4-mo period. Only patients in whom a follow-up visit in our center for 3 to 4 wk after stent placement was accepted, were included. Patient's characteristics, including demographic data, indications for treatment, clinical, radiological, or bacteriological evidence of postobstructive infection, and interventional procedure are shown in Table 1. All patients had severe symptomatic stenosis of the trachea and/or main stem bronchi. Informed consent was obtained from every patient.

Pulmonary Function Testing (PFT)

In all patients, spirometry was performed the day before, and 3 to 4 wk after stenting. PFT was performed to objectively assess the severity of the airway obstruction and to quantitate the effect of interventional treatment. At least three consecutive efforts were performed, of which two had to be within a 5% range. The best test in terms of maximal forced expiratory volume in one second (FEV₁) and forced vital capacity (FVC) was selected. All tests were performed on a Sensor Medics 2200 PFT Unit (Sensor Medics, Bilthoven, The Netherlands) in sitting position.

Recorded spirometry parameters were FEV₁, FVC, peak expiratory flow (PEF), forced expiratory flow after expiration of 50% of the vital capacity (FEF₅₀), forced inspiratory flow after inhalation of 50% of the vital capacity (FIF₅₀), and forced inspiratory vital capacity (FIVC).

Specimen Collection

Immediately prior to therapeutic bronchoscopic intervention, specimens for microbiological analysis were obtained with the patients in supine position, under total intravenous anesthesia using propofol, alfentanyl or remifentanyl, and atracurium. Patients were intubated with a rigid bronchoscope (Efer Dumon, La Ciotat, France or Storz, Tuttlingen, Germany), through which ventilation and oxygenation were assured using high-frequency jet ventilation with variable inspiratory oxygen fractions (FI_O2).

The rigid bronchoscope was positioned proximally to the airway stenosis. Immediately thereafter, a flexible fiberbronchoscope (Olympus Type 20 D; Olympus, Tokyo, Japan) was passed through the rigid instrument and, when possible, advanced through the airway stenosis. The tip of the flexible bronchoscope was immobilized immediately distal from the stenosis. A plugged telescoping catheter brush (Model 130; Mill-Rose Laboratories, Mentor, OH), was passed through the working channel and advanced 2 cm beyond the tip of the flexible bronchoscope. After ejection of the distal plug by protruding the inner cannula, the brush was advanced beyond the tip of the inner cannula, gently rotated against the airway wall or, when present, in visible secretions. It was then retracted a few centimeters into the inner cannula. The catheter was then retracted out of the flexible bronchoscope. If the flexible bronchoscope could not be passed through the airway stenosis because of a very small residual lumen or because of jeopardized ventilation and oxygenation, the tip of the bronchoscope was fixed immediately proximal to the stenosis. The catheter was then passed through the stenosis, and sampling was performed as described previously.

Three to 4 wk after stent placement, the patients were examined at the outpatient clinic. After clinical examination, PFTs were performed as before. Thereafter, fiberoptic flexible bronchoscopy was performed transorally with the patient in sitting position. After local anesthesia of the oropharynx with 10% lidocaine spray, and of the larynx with 4 ml 2% lidocaine solution, an Olympus videobronchoscope was introduced. After a short inspection of stent position and patency, and after scoring the presence or absence of secretion retention in the stent, a PSB procedure was performed. If no retained secretions were present in the stent (SR-negative), sampling was performed immediately distal from the distal end of the stent, against the airway wall or, when present, in isolated secretions. If secretions were visible in the stent (SR-positive), sampling was performed in these secretions.

The secretion retention (SR) was scored as follows: SR 0: no retention; SR +: retention. SR-positive cases were subsequently classified as follows (Figure 1): +:  25% of the circumference of the stent covered with secretions; ++: 25 to 49% of the circumference of the stent covered with secretions; +++: 50 to 100% of the circumference of the stent covered with secretions.

View larger version (147K): [in this window] [in a new window]   View larger version (142K): [in this window] [in a new window]   View larger version (135K): [in this window] [in a new window]   Figure 1.   Fiberbronchoscopic view of a stent with + SR positivity (A), ++ SR positivity (B), and +++ SR positivity (C ).

Processing of PSB Specimens

The distal portions of the catheters were wiped clean with 70% ethanol immediately after retraction from the bronchoscope, the brushes were cut with sterile scissors, and discarded into a sterile tube containing 1 ml of sterile saline, and agitated for 1 min. The tube was immediately transported to the laboratory. Volumes of 0.1 ml and 0.01 ml of the original solution were inoculated on the following agar media for quantitative culture and identification: horse blood agar supplemented with X and V factors incubated in 5% CO₂ at 37° C, MacConkey agar, mannitol salt agar, charcoal yeast agar and Sabouraud agar incubated under aerobic conditions at 37° C. If positive, identification by standard laboratory methods and counting of colony-forming units per milliliter (cfu · ml¹) were performed for each species. A cutoff value of  10² cfu · ml¹ was considered diagnostic for AC. Although cutoff values of  10³ cfu · ml¹ (for PSB samples) and 10⁴ cfu · ml¹ (for bronchoalveolar lavage samples) are considered diagnostic for "infection," a cutoff value of  10² cfu · ml¹ obtained via PSB sampling of the airways probably is suggestive for "colonization" (13). Quality assessment of the PSB samples by microscopic cell count was performed as described by Mertens and colleagues (14): on a cytospin preparation of 140 µl of the original fluid, microscopic scoring of cell counts was performed, including the number of squamous epithelial cells (< 1% was considered as uncontaminated lower airway sample) and the number of bronchial columnar cells, macrophages, neutrophils, and lymphocytes (> 10 cells/field was considered a good quality sample). Presence of white blood cells was expressed on a semiquantitative scale, one cross corresponding to 1 to 5/high-power field (HPF), two crosses to 6 to 10/HPF, three crosses to 11 to 20/HPF, and 4 crosses to > 20/HPF.

Data Analysis

Demographic and PFT data are expressed as mean ± SD. Descriptive statistics (for the demographic data) and two-tailed paired t tests were used (for the PFT data). Comparison of proportions between groups was made with Fisher exact test. Differences were considered significant at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

All interventional procedures were uneventful. PSB and bronchial washing samples could be obtained immediately distal to the stenotic lesion in every patient before, and 3 to 4 wk after the intervention.

In all but one patient there was a remarkable clinical improvement after the interventional procedure, as confirmed by the PFT results. Mean improvement in spirometry parameters is summarized in Table 2. Improvements in expiratory parameters (FEV₁, FVC, and PEF) were significant for the whole group, whereas mean inspiratory parameter (FIVC, FIF₅₀) improvement did not reach statistical significance. Colonization pattern had no influence on changes in PFT parameters (data not shown).

PSB Analysis before and after Stenting

As shown in Table 3, all PSB samples were of good quality in terms of percentage epithelial cells or cytospin preparations (< 1%) (14); in terms of number of cells per field, two PSB samples showed less than 10 cells, suggesting inadequate sampling. Because of the low (< 1%) percentage of epithelial cells, however, these samples were included for analysis. In five out of 14 patients, AC was present before stenting. In three of these, PPM were identified, including Pseudomonas aeruginosa, Streptococcus pneumoniae, and Staphylococcus aureus. Non-PPMs included Capnocytophaga sputigena and Haemophilus parainfluenzae.

Remarkably, there was no apparent relation between a history of previous postobstructive infections, and the results of PSB sampling prior to stenting. In the six patients who had experienced at least one episode of infection, three were on antibiotics while PSB was performed before stenting. Only in one patient out of six, PSB yielded significant growth of PPMs (> 10⁵ cfu · ml¹ P. aeruginosa and > 10⁵ cfu · ml¹ S. pneumoniae), and this in a patient on antibiotics.

After stenting, AC was found in 11 of 14 patients, including seven patients without prior colonization. In the colonized patients after AS who already had AC before AS, the same microorganisms were present after AS (except in two patients), as well as "new" microorganisms.

In six of these 11 cases, PPM were present, including P. aeruginosa (4), S. aureus (3), S. pneumoniae (1) and Klebsiella spp. (1). In none of these patients, however, were clinical signs of infection present. No patients were on antibiotics during the second sampling procedure, or had received antibiotics in the period between stenting and the second sampling.

Association between PSB Sampling Results and SR Scores

Positive PSB sampling results seemed associated with the presence of SR (SR positivity): PSB samples  10² cfu · ml¹ were found in 10 of 12 SR-positive cases, and in one of the two SR-negative cases.

PSB samples < 10² cfu · ml¹ were found in two SR-positive and in two SR-negative cases. Probably because of the small population size, statistical significance was not reached (Fisher exact test, p = 0.06). In the SR-positive group, there was no association between the degree of SR and the presence or absence of AC (p = 0.741).

Effects of Associated Treatments (Mechanical Dilatation, Nd:YAG Laser) on PSB Sampling Results

There was no difference in PSB sampling results after AS between the mechanical dilatation (p = 1, Fisher exact test) or Nd:YAG laser photoresection (p = 1, Fisher exact test) treated patients.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this first systematic study on the effects of AS on the microbiological status of the airways, we were able to show AC by PPM and by non-PPM in the majority (71%) of patients after stenting. This AC probably occurs more often in patients with in-stent secretion retention. AC, however, was never associated with clinical signs of infection, and no patient had to be treated with antibiotics.

Distal airways usually are sterile in healthy nonsmokers (7, 8). When mechanical airway defenses are altered, as occurs in chronic obstructive pulmonary disease (COPD), bronchiectasis, or bronchial obstruction, distal airways may become colonized by PPM and non-PPM: AC has been shown to be present in 42% of bronchogenic carcinoma patients, 83% of patients with COPD, 88% of bronchiectasis patients, and in 47% of tracheotomized patients (13).

In our series of 14 patients with obstructive lesions in the central airways, five (35%) showed AC before stenting. In contrast with Cabello's data (13), the majority of colonized airways harbored PPM, and, also contrasting, P. aeruginosa seemed more prevalent in our patients with severe obstructive lesions. However, the patients studied by Cabello differed from ours in terms of underlying airway diseases (type and severity). Nevertheless, this remains remarkable, because none of the patients had been hospitalized for more than 72 h before the first bronchoscopy. After stenting, P. aeruginosa also was the most frequently isolated PPM, occurring in 40% of colonized airways. Here also, every patient had been discharged from the hospital 1 d after the stenting procedure and was an outpatient at the time of the second bronchoscopy.

The presence of previous or current postobstructive infections before stenting, whether treated with antibiotics or not, seemed not to be of influence on the sampling results, nor to be related to the presence of AC, suggesting that it is the mechanical obstruction of the airways itself which relates to their colonization.

Stenting, although usually well tolerated, may dramatically alter the "mechanical" expectorative properties of the airways (12, 13). As a consequence, the inner stent surface frequently becomes covered with airway secretions, which may adhere and desiccate, and thus even partially obstruct the stent lumen (15).

SR occurred in 12 (85%) patients: in four only minor retention (occupying less than 25% of circumferential stent surface) was observed, in two patients between 25 and 50% of the stent surface was covered with secretions, and in six patients more than 50% of the surface was covered. In two patients there was no SR at all. The presence of retained secretions seemed to relate to the occurrence of AC, although a statistically significant correlation could not be reached (p = 0.06), probably because of too small a sample size.

The mechanism of SR and colonization in stents is unknown, but may be related to microscopical irregularities on the macroscopically smooth internal surface of the stent, in analogy with the light- and electron microscopically demonstrated irregularities on the surface of endotracheal tubes, which are large enough to harbor colonies of bacteria (16). Endotracheal tubes have been shown to act as reservoirs for bacterial growth. Comhaire and Lamy have shown that all endotracheal tubes in 25 patients requiring mechanical ventilation were colonized by the ninth day (17), and electron microscopic examination of tubes recovered from 25 intensive care unit (ICU) patients showed partial (16%) or complete (84%) covering of the tube surface by amorphous bacteria-containing matrix with bacterial aggregates projecting into the lumen of the tube (18). Binding of bacteria can be promoted by the presence of the polysaccharide glycocalyx material that surrounds the outer cell wall of most bacteria, including P. aeruginosa (19), which may explain the large presence of this colonizing microorganism in our study.

The presence of a stent in itself may also increase mucus secretion by the airways ("reflex mucus secretion"), as has been shown after tracheal intubation (20). In this situation of increased mucus secretion and mechanically impaired mucus clearance, pooling and stagnation of secretions may occur, which in turn may promote colonization by favoring the binding and growth from bacteria (21).

Diagnosis of lower respiratory tract infection and colonization requires techniques enabling avoidance of contamination of samples by oropharyngeal flora followed by quantitative culture, such as PSB sampling and (protected) bronchoalveolar lavage (22). Cutoff levels of  10² cfu · ml¹ for PSB samples are indicative for AC (13). The use of quantitative PSB sampling during the first bronchoscopy therefore virtually excludes the possibility of contamination by oropharyngeal flora of the postobstructive airway region which was sampled. However, we cannot exclude that the stents were contaminated by the therapeutic bronchoscopy itself, i.e., by translocation of bacteria that were already colonizing more proximally initially. Systematic oropharyngeal microbiological sampling theoretically could have clarified this issue. However, the percentage of qualitative agreement in microbiological flora between oropharyngeal (assessed by, e.g., oropharyngeal swabs) and airway sites is reported to be only 50% or less in the presence of airway disease (23). Furthermore, some microorganisms such as P. aeruginosa are able to adhere to tracheal cells without previous oropharyngeal colonization, particularly in patients with underlying airway disease (13). Including a control population in the study consisting of individuals in whom therapeutic bronchoscopy (e.g., laser resection or mechanical dilation) but no stenting was performed probably would represent the only unequivocal way to confirm or refute translocation of upper airway microorganisms as the mechanism responsible for the observed colonization. Although this option was considered, it was abandoned because laser and dilation therapy are not or only anecdotically associated with secretion retention or airway infection (6, 24), whereas we were primarily concerned with the impact of "putting a foreign body into an airway" on the local microbiological environment. Nevertheless, including a control population might have clarified the mechanism by which the stented airways were colonized.

In conclusion, we demonstrated that stenting is followed by AC in the majority (78%) of patients within 3 to 4 wk after the procedure. In 55% of colonized patients, PPM were identified. Furthermore, colonization seemed to be associated with the occurrence and degree of SR in the stent. Colonization however was never complicated by clinical signs of infection, but the study period may have been too short (3 to 4 wk) for infectious events to occur. In view of our findings, isolation of one or more PPM from stented airways in a patient with clinical signs of bronchial or pulmonary infection, probably does not imply a necessary causal relationship. Quantitative cultures as obtained by PSB furthermore probably cannot discriminate colonization from infection, in view of the very high concentrations of germs in "simple" AC. The mechanism by which stented airways are colonized remains to be examined in future studies.