INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Although asthma is a heterogeneous process, airway obstruction, hyperresponsiveness, and inflammation consistently characterize it. Recent studies have suggested that asthma originates in early childhood and that it is a chronic inflammatory disease of the airways. In fact, childhood asthma has become the most common disease of childhood (1). Typically, symptoms such as recurrent wheezing, airway obstruction, and/or narrowing present during the first 6 years of life. These symptoms are extremely common and account for 25% of pediatric admissions to United States hospitals, reaching 50% in the winter in children less than 5 yr of age (2, 3). Despite a 60% prevalence of wheezing in the first year of life in the United States and 32% in the first 5 yr in the United Kingdom, investigation into the pathophysiology of such airway narrowing has been quite limited (4). Of the 60% of children who initially wheeze, only 30% persist with wheezing illnesses into later life and 10% are subsequently diagnosed with asthma (4). Contributing factors to the subsequent development of asthma in these early "wheezers" are poorly understood. Unfortunately, extrapolating from adult data into this very young pediatric population will likely provide limited and potentially erroneous information.

The National Heart, Lung, and Blood Institute's (NHLBI) workshop summary on the "Effects of Growth and Development on Lung Function: Models for Study of Childhood Asthma," offered several critical recommendations concerning future directions for evaluating pediatric asthma (5). These included (1) establishing features of airway disease through the use of research bronchoscopy (6, 7), (2) identifying early markers of airway inflammation that may reflect disease activity in blood and/or urine, and (3) exploring different mechanisms of airway obstruction (5). Although some studies have evaluated points 2 (8, 9) and 3 (10, 11), few invasive (bronchoscopic) evaluations have been performed in wheezing children (3, 12).

This is the first study to examine the cellular composition and specific lipid mediators present in the airways of very young wheezing children (WC) and compare them with healthy, age-matched normal controls (NC). The results strongly support the presence of airway inflammation at a very early age. Furthermore, bronchoalveolar lavage (BAL) findings in these young infant wheezers appear to differ from patterns previously reported in older children and adults with asthma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Wheezing Children

Males/females (< 60 mo) with prolonged wheezing (> 2 mo within a 6-mo period) determined by their attending physician (not the investigator) to have a clinical indication for bronchoscopy were eligible for participation. The subjects must have been extensively evaluated and treated by their primary care physicians and pulmonologists without resolution of wheezing. Due to the persistence of respiratory symptoms despite multiple interventions and evaluations, WC underwent bronchoscopy to identify potential anatomic and/or infectious etiologies and to rule out aspiration by examining the BAL for lipid-laden macrophages, as well as the presence of a foreign body.

Children were excluded if acutely ill or receiving > 450 µg/d of inhaled steroids. Subjects receiving antibiotic therapy, oral steroids, and/or antileukotriene (LT) drugs within 1 mo of bronchoscopy/BAL were excluded from the study, although the use of -agonists and cromolyn sodium was allowed. The protocol was approved by the Colorado Institutional Human Subject's Review Committee and The Children's Hospital (TCH) Institutional Review Board and informed consent was obtained from each patient's caregivers.

Caregivers were asked to complete a "Wheezing Questionnaire" at the time of bronchoscopy. This questionnaire provided clinical information concerning the subjects' birth history, hospitalizations, acute illnesses, family history of asthma and atopy, parental smoking, exposures, and treatments. No exclusions were made based on information obtained in the questionnaire.

Normal Controls

NC were children undergoing elective surgery with general anesthesia without any history of acute (< 1 mo) or chronic respiratory symptoms.

Bronchoscopic Procedures

WC were admitted either to TCH Emergency Department or to the pediatric general medicine ward for a 23-h short stay for the bronchoscopy/BAL. All NC were evaluated with bronchoscopy/BAL in an operating room setting under general anesthesia through a minimum endotracheal tube (ETT) size of 4.5 mm diameter while under the direct supervision/care of an anesthesiologist. A minimum 4.5-mm-diameter ETT was used to allow for the use of a 3.5-mm fiberoptic bronchoscope with lavage port, while ensuring adequate airflow.

WC underwent bronchoscopy in the manner standard to the Pediatric Pulmonary section of TCH. Following placement of peripheral intravenous access, WC were placed on continuous pulse oximetry and cardiorespiratory monitoring for a total of 4 h during and after the procedure. With initiation of monitoring. WC were sedated with repeated doses of 0.05 mg/kg midazolam (not to exceed 0.2 mg/kg) and 1.0 µg/kg fentanyl (not to exceed 4.0 µg/kg). Subjects were also given an albuterol and lidocaine (titrated by weight) nebulizer treatment to anesthetize the upper airway. Subsequently, 2% viscous xylocaine (total of 20-40 mg) was applied with cotton swabs to anesthetize the nares. A 3.5-mm Pentax fiberoptic bronchoscope (Pentax Precision Instrument Corp., Orangeburg, NJ) was then passed through the nares to the posterior pharynx under direct visualization. A maximum dose of 5 mg/kg of lidocaine was used to anesthetize the airway. After anesthetizing the vocal cords, the bronchoscope was passed into the trachea. The carina and right and left main stem bronchi were sprayed with separate aliquots of 1-2% lidocaine. Following direct visualization of all primary segments of the right and left main stem bronchi, the bronchoscope was wedged into the right middle lobe bronchus. Sterile, nonbacteriostatic saline at room temperature was instilled through the bronchoscope (four aliquots of 1 ml/kg each, followed by 3 to 5 ml of air). The BAL fluid was aspirated through the bronchoscope by immediate gentle hand suction. The aliquots were pooled together into a single sample and immediately placed on ice.

Analyses of Bronchoalveolar Lavage Fluid (BALF)

Pooled BALF was placed into two aliquots. One aliquot was submitted for quantitative bacterial and viral cultures to TCH Microbiology Department. The remainder of the BALF was immediately placed on ice and taken to National Jewish Medical and Research Center for further analysis of the cellular and fluid fractions and processed as described later.

Bacterial and viral cultures. Specific culture techniques were used to identify common pathogens including Streptococcus aureus, Hemophilus influenzae, Moraxella catarrhalis, and Streptococcus species. Viral cultures were preformed to isolate common respiratory viruses (respiratory syncytial virus, influenza A and B, cytomegalovirus, parainfluenza 1, 2, and 3, adenovirus, herpes simplex virus, rhinoviruses, and enteroviruses). Culture techniques standard to TCH Microbiology Department were utilized.

Cytology. BAL fluid was centrifuged at 1650 rpm (600 × g) 4° C to separate fluid from cells. The cell pellet was resuspended in 500 µl of phosphate-buffered saline (PBS), pH 7.4. Cell counts and viability testing were performed using a hemacytometer and trypan blue exclusion staining. Results were expressed as viable cells/ml of lavage fluid.

Cytocentrifuge preparations were made from 35,000 cells using a Shandon Cytospin-2 (Shandon Southern Instruments, Sewickly, PA). Slides were stained with a Diff-Quik stain (American Scientific Products, McGaw Park, IL) and cell differentials determined counting 600 cells in a blinded manner. Individual cell number/ml were determined by multiplying the percentage specific cell type by the total cell number/ml of BALF.

Eicosanoid assays. Twenty-five thousand dpm [³H]LTE₄ (New England Nuclear, Boston, MA) was added to 2-5 ml of all samples as an index of sulfidopeptide leukotriene (LT) recovery (generally > 65%). The 2-5 ml samples were loaded onto octadecylsilyl (Sep-Pak) cartridges (Waters, Milford, MA), washed with a 10% methanol/saline solution, and the eicosanoids eluted in 5 ml of 80% methanol.

After Sep-Pak purification, samples were evaporated under negative pressure using a speed-vacuum concentrator and resuspended in 0.8 ml of immunoassay buffer (100 mM) potassium phosphate buffer (0.1% bovine serum albumin [BSA], 0.38 g/L ethylenediaminetetraacetic acid [EDTA], pH 7.4), providing a concentration of approximately 2.5-fold. Sample aliquots for prostaglandin (PG)D₂ were evaporated separately, derivatized with a methoxyamine reagent by heating at 60° C for 1 h (13), and diluted to a final concentration of approximately 1.7:1.

LTB₄, LTC₄, PGD₂, PGE₂, and 15-hydroxyeicosatetraenoic acid (HETE) were quantified by competitive enzyme immunoassays (EIA) using acetylcholinesterase-conjugated tracers as previously described (14, 15). The rabbit polyclonal antibody (Ab) for PGD₂ was a gift from Dr. R. D. Kelly (Edinburgh, UK). A rabbit polyclonal antibody for peptide LTs cross-reactive with LTC₄ (100%), LTD₄ (100%), and LTE₄ (67%) was purchased from Cayman Chemical Co (Ann Arbor, MI) and utilized for assay purposes. LTE₄ was used as the reference standard in assays and the peptidoLT products referred to as LTE₄. The rabbit polyclonal antibody used for the quantification of LTB₄ and the mouse monoclonal PGE₂ antibody were also purchased from Cayman Chemical Co. 15-HETE was quantified by enzyme immunoassay using a rabbit polyclonal Ab purchased from PerSeptive Biosystems (Framington, MA). Lavage results were evaluated as "per milliliter" of fluid as in previous studies (16, 17).

Tryptase assay. -Tryptase levels in BALF were measured by a sandwich EIA using the B₁₂ monoclonal antibody for capture and biotin-G5 monoclonal antibody for detection of -tryptase (16, 18) in the laboratory of Dr. L. B. Schwartz (Richmond, VA). The sensitivity of the assay was 0.1 ng/ml. -Tryptase data were included in the current study as an indicator of mast cell activation.

Statistical Analysis

As the majority of the data were not normally distributed, nonparametric analysis was used. Results were expressed as median with interquartile range. Wilcoxon's signed rank testing for nonparametric data were used to determine differences between the two groups. Statistical significance was defined at a level of p < 0.05. All data were analyzed using a JMP program (version 3) Statistical Software for the PC (SAS Institute, Cary, NC).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Populations

Demographic characteristics of WC and NC are shown in Table 1. A total of 24 WC (18M/6F) and 6 NC (4M/2F) were evaluated with a questionnaire, bronchoscopy, and BAL. Based on BAL findings, 4 of the original 24 WC were excluded from the study. Two male patients had lipid indices > 75% (indicative of lipid-laden macrophages in the airway consistent with aspiration). Two patients (a male and female) grew Streptococcus pneumoniae/Moraxella catarrhalis and Streptococcus pneumonia/Staphylococcus aureus, respectively, from BAL cultures. Neither of these two subjects had clinical findings suggestive of acute infection prior to or at the time of bronchoscopy despite positive BAL cultures. The remaining 20 WC had a mean age of 14.9 mo (range 4.5 to 48 mo), whereas NC had a mean age of 23.3 mo (range 14.0 to 31.5 mo).

The NC group was older than the WC group. This age difference was necessary because NC underwent bronchoscopy and lavage while intubated for elective (nonpulmonary) surgery. Intubation with a minimum 4.5-mm ETT was required to perform bronchoscopy and BAL with a 3.5-mm fiberoptic bronchoscope. Typically, only children > 12 mo can be intubated with a 4.5-mm ETT.

Surgical procedures requiring general anesthesia in the NC group included bilateral inguinal hernia repair (2), umbilical hernia repair (1), skin graft due to left-hand burn (1), anal reconstruction (1), and dental caries (1).

Wheezing Questionnaire

After exclusion of the four WC with confounding infection or aspiration, 10 of the 20 WC had a history of viral illness at symptom onset; 8 of these 10 had evidence of respiratory syncytial virus (RSV) based on viral culture and/or EIA. Fifteen of the 20 WC evaluated with bronchoscopy/BAL had a positive family history of asthma/allergies/atopy and 11 of 20 WC had reported signs of atopy defined as eczema, allergic conjunctivitis, and/or recurrent rhinitis (Table 1).

Eighteen of the 20 WC had been treated with an inhaled bronchodilator and were reported to have transient decreases in wheezing and/or respiratory symptoms postbronchodilators. A high level of clinical severity was suggested because 18 of 20 WC had received oral steroids, averaging 2.8 courses/ subject. Twelve of the 20 WC were also treated with inhaled corticosteroids (ICS) ( 450 µg/d), but had persistent wheezing and episodes of respiratory distress. Ten of the 12 receiving ICS were on < 200 µg/d. Nine WC were supported with fluticasone (8 with 176 µg/d and 1 with 440 µg/d) and three with beclomethasone diproprionate (two with 168 µg/d and one with 336 µg/d) in the form of a metered dose inhaler (MDI). All MDIs were used with an aerochamber and mask in these young WC. Additionally, caregivers reported an average of 5.3 courses of antibiotics/subject from the time of wheezing onset to preceding bronchoscopic evaluation.

Bronchoalveolar Lavage

Bronchoscopy/BAL was well tolerated in all WC and NC and no adverse effects were reported. No subject required more than a single nebulized albuterol treatment or prolonged observation postprocedure. Few desaturations to < 90% occurred, all of which responded immediately to transient increases in oxygen. BAL return was expressed as percentage return and did not differ between the two groups (46% for both groups).

Cell Counts and Differentials

Total cells × 10⁴/ml were increased in WC compared with NC (21.5 [9.1-33.8] and 7.2 [5.1-11.7]; p = 0.01). WC also had significant increases in total cells × 10⁴/ml in all the cell types compared to NC. The largest difference occurred in lymphocytes (LYMPH) (p = 0.007). Differences in eosinophil (EOS) numbers also reached significance (p = 0.04), however, these differences were the least significant compared with the other cell types. No individual cell type, expressed as a percentage of total, reached significance in WC compared with NC. The differential cell counts, as percentage of total and individual absolute total cells/ml of macrophages/monocytes (M&M), LYMPH, polymorphonuclear cells (PMN), EOS, and epithelial cells (EPI), for the two groups are summarized in Table 2. Figure 1 presents significant elevations in LYMPH, EPI, and PMN in WC compared with NC.

View larger version (13K): [in this window] [in a new window]   Figure 1.   LYMPH, EPI, and PMN expressed as cells × 104/ml in WC compared with NC. The median is shown as a single black line. Each circle, triangle, and diamond represents one individual.

No significant differences in BAL cell counts (Table 3) and mediator levels (not shown) were found in the 12 WC receiving ICS (WC-[ICS⁺]) compared to the 8 WC not receiving ICS (WC-[ICS]) at the time of bronchoscopy. Although the individual cell counts were not different in these two groups, only the percentage of epithelial cells was significantly increased in WC-(ICS) compared with WC-(ICS⁺), (p = 0.009).

Mediator Levels

BALF levels of 15-HETE (p = 0.002) and PGE₂ (p = 0.0005) were significantly greater in WC compared with the NC group (Figure 2 and Table 4) LTE₄ and LTB₄ levels were also elevated in WC compared with NC (p = 0.04 and p = 0.05, respectively). PGD₂ and -tryptase levels did not differ between WC and NC (p = 0.20 and p = 0.94, respectively).

View larger version (14K): [in this window] [in a new window]   Figure 2.   PGE2 levels expressed as pg/ml in WC compared with NC. Levels of 15-HETE are expressed as ng/ml in WC compared with NC. The median is shown as a single black line. Each circle and triangle represents one individual.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This is the first bronchoscopic study to describe the amount and type of inflammation present in the lower airways of very young children with persistent wheeze. A total of 20 WC, without complicating factors such as aspiration and/or bacterial/viral infection, were evaluated. Bronchoscopic evaluation demonstrated that these infants with recurrent wheezing had significant airway inflammation compared with NC. Total BAL cells/ml were increased more than 3-fold in WC compared with NC. This increase is higher than that reported in a similar study comparing slightly older and possibly less severely ill pediatric patients (19). Increases in overall cell numbers and not individual cell types (percentage of total cells) would imply that the inflammation in very young WC consists of nonspecific increases in many cell types occurring without specific predominance or alteration in normal airway cellular patterns.

The current findings differ from previous studies in adults and children that reported more prominent mast cell and eosinophil involvement in asthma (19). In contrast, the majority of earlier studies evaluated older children with asthma (median age over 7 yr) with less clinical severity and used a nonbronchoscopic means of airway fluid sampling (3, 12). Ennis and colleagues (12) evaluated patients undergoing elective surgery, subjects were divided into atopic and nonatopic wheezers on the basis of history and/or serum IgE levels. The clinical history of these patients at the time of their elective surgery would suggest that unlike the current population, these patients were quite stable. Additionally, these studies employed nonbronchoscopic lavage and a nonvisualized insertion of an 8-French suction catheter via an ETT. A single 20 ml normal saline lavage was then performed in patients of various sizes. Potential limitations of such a methodology include inconsistent sites of sampling and the use of a single lavage volume regardless of patient age and size.

A recent study by Marguet and colleagues used a standard bronchoscopic technique (19). The authors reported findings on a younger, more clinically severe infantile wheeze group (median age 24 mo). Unlike the current study, a majority of the infant wheezers in that study had positive bacterial cultures from their BAL. Therefore, the inflammatory findings in that study, including neutrophilia, could reflect an acute infectious process. Additional patients could have had a viral etiology for their symptoms, but BAL viral cultures were not performed. Finally, the normal control group comprised children who required fiberoptic bronchoscopy for respiratory problems, and, therefore, may not have been representative of a "normal" control group.

The current study differs from the previous studies by evaluating even younger WC (median age 14.9 mo) who had no evidence of either acute bacterial and/or viral infection based on BAL cultures. The clinical histories of these WC suggest that these children had more severe disease than has previously been reported. This WC population is, therefore, likely not representative of the more typical child who has had one or two wheezing episodes within the first 5 yr of life. Additionally, the NC group was composed of children without any history of respiratory problems. Although they were older than the WC, this was an unavoidable technical issue related to ETT size (see METHODS for details). In both groups, a standard BAL technique was pursued that regulated the amount of lavage in milliliters/kilogram/patient.

In addition to methodology, the results reported here differ in many ways from those previously reported. Although the total number of eosinophils was elevated in WC, they were present in the least significant number and percentage compared with NC (p = 0.78, comparing percentage eosinophils in WC with NC). Although it is possible that the modest elevation in eosinophils in the airways of these WC is related to the use of ICS in the majority of children, total eosinophil numbers did not differ between WC-(ICS⁺) and WC-(ICS) (Table 3). Interestingly, the percentage epithelial cells (not total epithelial cells/ml) was significantly elevated in WC-(ICS) compared with WC-(ICS⁺). The importance of such an elevation, however, is not clear. Therefore, in contrast with the older patients with an established diagnosis of asthma, eosinophils may not be as predominant as inflammatory cells in infant wheezers (3, 19). Similarly, inflammatory mediators specifically derived from mast cells and commonly associated with adult asthma, PGD₂, and -tryptase, were not found to be elevated in WC (16). Thus, despite the history of atopy in a significant number of WC (11 of 20), BAL findings were less consistent with "atopic" inflammation in terms of selective eosinophil recruitment or mast cell activation. However, it is likely that the WC in this study represent a relatively diverse group whose overall severity may be similar, but whose underlying pathology may not. Cells, such as eosinophils and mast cells, could, nonetheless, play a distinct role in subgroups of WC. Previous analysis of subgroups identified by Martinez and colleagues has suggested that < 15% of initially wheezing infants progress to chronic asthma, with those that do progress being more closely associated with familial/personal atopic or asthmatic disease (2). In our population, a subgroup (3 of 20) could be identified with > 2.0% BAL eosinophils. It is possible that this subgroup reflects the small percentage of WC who subsequently develop asthma. In contrast, the BAL findings from the majority of WC support the concept that most early/transient wheezing is not allergy driven, but more reflective of viral processes. Interestingly, perhaps due to the severity of the population studied, no WC had "noninflamed" airways comparable with NC.

Lymphocytes have also been strongly implicated in adult asthma (22, 23), especially as related to a TH₂ type phenotype (9), however, their role in persistent wheezing of childhood is less well understood. Perhaps, related to the general increase in cell numbers, the total number of lymphocytes in WC was also significantly increased compared with NC (p = 0.007). This general increase in lymphocytes requires further investigation to determine the phenotype, activation, and cytokine complement of these cells in the persistent WC population be performed. Whether different phenotypes will be identified in subgroups of wheezers is a question of considerable importance.

The epithelial cell, which forms the first barrier to external elements such as viruses and allergens and has also been linked to injury and repair processes, may play a role in both adult and childhood asthma (10, 11, 24, 25). In addition to the increased epithelial cell numbers in the BALF of WC, supporting evidence for their importance comes from the elevation of two mediators (PGE₂ and 15-HETE) potentially derived from the epithelium. The increase in PGE₂ is especially intriguing as this particular prostanoid has not been shown to be elevated in adult asthma, and, in fact, inhalation of PGE₂ appears to be protective against antigen-induced airflow limitation (26). Like PGE₂, 15-HETE has also been suggested to play a protective or "antiinflammatory" role in asthma (27, 28). Future studies evaluating epithelial chemokines, such as interleukin-8, will be helpful in delineating the role of the epithelium in viral and nonviral related WC.

As previously stated, although several studies have suggested important roles for mast cells and eosinophils in pediatric asthma (3, 19), in the current study, total numbers of both macrophages and neutrophils were increased in WC compared with NC (p = 0.02). Interestingly, a previous study in children with chronic asthma suggested a close correlation between increased BALF macrophages and heightened bronchial hyperresponsiveness (21). BAL neutrophils were also increased in the present study. Whether this is due to the increased asthma symptom severity in the population described here, or related to either undetected current or prior viral infections is not clear. LTB₄, a potent chemoattractant for neutrophils produced by both macrophages and neutrophils, and measured in high quantities in the BALF of these WC compared with NC, could be associated with the increased number of both of these cell types (29).

In summary, inflammatory cells and mediators were found at a very early age in the airways of persistent childhood wheezers. Similar to adults, this inflammation was characterized by an increased total numbers of lymphocytes and epithelial cells; however, in contrast to adults and older children with asthma, macrophage and neutrophil numbers were also elevated, whereas increases in eosinophil number were of a more limited degree. Whether these discrepancies are reflective of different pathologic processes at early stages of the development of asthma or multiple subgroups of patients, or whether any of these cells and/or mediators contribute directly to the development of chronic asthma remains unclear. Further longitudinal studies of larger numbers with prospective follow-up are necessary to identify specific inflammatory conditions that may predispose to the development of chronic asthma.