Bronchoalveolar Lavage and Tracheal Aspirate for Assessing Airway Inflammation in Children

MICHAEL D. SHIELDS and JOSEF RIEDLER

Department of Child Health, Queens University of Belfast, and Institute of Clinical Science, Belfast, Northern Ireland; and Children's Hospital, Salzburg, Austria

	WHAT DO WE KNOW?

TOP WHAT DO WE KNOW? WHAT DO WE NEED... HOW CAN WE ACHIEVE... REFERENCES

Studies using bronchoalveolar lavage (BAL) and tracheal aspirate (TA) have increased our knowledge of the underlying airway inflammation present in the childhood disorders of asthma, cystic fibrosis (CF), and neonatal chronic lung disease (CLD) (1).

A BAL sample is obtained by wedging a bronchoscope or catheter into a bronchus and isolating the distal airway. A volume of saline is instilled and the fluid aspirated back from the airway, using gentle suction. The aim of this process is to obtain aspirated fluid that will contain the microbes, cells, and noncellular constituents present in the epithelial lining fluid of the small airways and alveoli.

Flexible fiberoptic bronchoscopy (FOB) allows a visual location and wedging of the bronchoscope tip (1, 7). In blind nonbronchoscopic BAL, a catheter is passed through an endotracheal tube and blindly wedged in a distal airway. This technique has been used to obtain normal cellular and noncellular constituents and to study diffuse airway inflammation (13). In neonatal studies, a TA or bronchial washing can be obtained by inserting a catheter through an endotracheal tube for a short distance (usually not wedged) and aspirating back mucus or instilled fluid. Darlow and coworkers (17) have shown that samples obtained by dry shallow suctioning, often used routinely to maintain airway patency in neonatal intensive care, may also provide adequate samples for research purposes. The mean values of dry and lavage samples were sim-ilar with respect to myeloperoxidase and ₁-antitrypsin content.

Problems with Standardization

The procedures for obtaining, processing, and interpreting BAL sample results have not been standardized for children. Factors that may influence the nature of the recovery fluid include size of the bronchoscope, volume instilled, number of aliquots, use of first aliquot, and dwell time.

Acquistion of Sample

Size of bronchoscope or catheter. Bronchoscopes come in fixed sizes, hence, for children of different sizes, different proportions of the lower respiratory tract will be sampled for a given lavage volume. The use of a small bronchoscope (e.g., 3.6 mm outer diameter) will result in the lavage of a more peripheral part of the lung in an older and larger child than in a small child. In addition, it has been observed that the wedging of a narrow bronchoscope in a larger child is associated with a smaller recovery volume (possibly because of distal airway collapse) (18).

In blind nonbronchoscopic BAL, the catheter used has a narrow outer diameter (e.g., 8 French [8F] gauge with outer diameter of 2.6 mm), allowing wedging in a more distal airway compared with currently available fiberoptic bronchoscopes. Catheter size has been varied with the dimensions of the endotracheal tube (ETT) (6F gauge for < 3.5 ETT, 7F gauge for 3.5 ETT, 8F gauge for > 3.5 to 5 ETT, and 10F gauge for > 5.5 ETT) and therefore with the size of the child (15, 16).

When bronchoscopic wedging occurs centrally, this will potentially result in a BAL with increased neutrophil percentages but with fewer macrophages (7, 9, 12, 13).

Volume of fluid instilled. The volume of fluid instilled can be varied in an attempt to correct for patient size and has ranged from 0.5 to 3 ml/kg in different studies. Ratjen and Bruch (11) have suggested that if BAL volume is adjusted to body weight a constant fraction of epithelial lining can be obtained. They addressed this by using a weight-adjusted BAL of 1 ml of lavage fluid per kilogram of body weight for each of three lavages. They excluded the first aliquot and found that the absolute concentrations of both urea and albumin (dilutional markers) were remarkably constant throughout the age range of 3-15 yr in normal children.

Some studies have used a fixed volume, for example, 10-20 ml irrespective of patient size or weight (2, 8, 14) and others have used 10-15% of the estimated functional residual capacity. When a small fixed volume is used or if the first BAL aliquot is analyzed separately this is regarded as a "bronchial wash" sampled predominantly from the conducting airways and may have a low total cell count and percentage of macrophages but a high percentage of neutrophils (7, 9, 12, 13). Many of the conditions for which BAL is used for research predominantly involve the "airways," and this bronchial sample may be important. The blind nonbronchoscopic technique appears to sample both the proximal and distal airways, at least in intubated neonates (13). The Belfast group (2, 14) used a nonbronchoscopic technique to obtain BAL fluid with a fixed catheter size and a fixed volume lavage. The rationale behind this approach was that in smaller children the catheter would wedge more proximally and the fixed lavage volume would be proportionally larger, whereas in bigger children the catheter would wedge more peripherally, requiring a smaller volume to provide a similar sample of airway fluid. This has been substantiated by finding no age-related variations in the cellular values. In very young children the actual total volume recovered even after two or three repeated instillations may be small and it seems wise to retain the entire sample for analysis.

Dwell time. Typically 20-60% of the lavage fluid is recovered. In some studies the instilled fluid volume has been aspirated immediately (2, 7, 8, 10, 12, 14), whereas in other studies there has been a short pause with a longer dwell time (13). Increasing the dwell time will impact on studies of dilution factors when the urea dilution method is used, as this method is based on the assumption that urea will diffuse independently of vascular permeability or active transport mechanisms. Therefore, urea should be at the same concentration in the epithelial lining fluid (ELF) as in plasma and by establishing the relative dilution the exact volume of ELF can be determined. However, limitations in this method have been suggested in adult studies, where sequential aliquots (and hence a long effective dwell time) are required to perform BAL. In this situation, urea diffuses during the lavage procedure and is dependent on the dwell time of the fluid in the lung (19).

Site. In most reports, the bronchoscope has been wedged in the right middle or lingular lobe as, in adults, this has yielded high and uniform fluid recovery. The location of lavage may not be important when studying healthy children (8). However, we do not know if this also applies to supposedly diffuse lung diseases, such as asthma and neonatal CLD, in which there may be regional variations. Because it is a blind technique, nonbronchoscopic lavage cannot be standardized on location. However, it is thought that if the infant's head is turned leftward, the catheter will wedge distal to the right bronchus (13, 20).

Unknown BAL dilution factor. The recovered lavage fluid contains a small fraction of ELF (with cellular and noncellular components) that is diluted to a variable degree. The degree of dilution depends on the instilled volume, the surface area with which the lavage fluid is in contact, and the dwell time, none of which have been standardized. Dilutional markers such as urea and albumin are considered unreliable for correcting for ELF dilution (21). The urea concentration in BAL is affected by dwell time, and if epithelial permeability is altered with inflammation, there is an increase in albumin in the ELF. However, when lavage was performed with a small fixed volume and a blind nonbronchoscopic technique, using urea as a dilutional marker showed a similar degree of dilution in BAL from normal children and adults (14). In addition, Ratjen and Bruch (11) suggested that if BAL volume is adjusted to body weight a constant fraction of epithelial lining can be obtained. This is unlikely to be the case in pathological states.

It is difficult to determine whether these differences in BAL sampling techniques are of practical importance and this dilution dilemma may be a distraction. To overcome the dilutional problem, counts of each cell type can be expressed as a percentage of the total number of cells in the BAL. It is important to record whether epithelial cell counts have been included when percentages of the other cell types are calculated and whether the percentage quoted is a percentage of white blood cells or of total cell count. It appears that the small differences in percentages seen in studies of normal cellular reference values in BAL in children are accounted for by the way the first aliquot or "bronchial wash" was handled (separately or pooled) and by how the cellular percentages were expressed.

There is no standardized method of reporting acellular data. It has been suggested that a semiquantitative approach, not influenced by dilution, is to express lavage results as differential proportions of components relative to each other or as amounts per milliliter with information provided on lavage method and input and recovered volumes (21).

It is likely that both bronchoscopic and nonbronchoscopic methods yield similar cellular BAL results in normal children (22).

Safety and ethical issues. BAL sampling is safe. Transient and usually minor side effects include reversible hypoxia, and transient fever, which is usually mild (23). Despite this, it is ethically difficult to justify the sedation or anesthesia required for BAL in young children for research purposes. Nevertheless, ethically acceptable methods of obtaining BAL fluid for research in children have been found. Serial TA and BAL fluid can be obtained from suction samples recovered after routine pulmonary "toilet" in ventilated neonates and young children (13, 17). BAL samples have been obtained in normal children and children with stable asthma who happen to be undergoing elective surgery for nonpulmonary conditions (2, 9, 15, 16). The additional use of lavage specimens for research purposes from children with asthma, and young children with wheezing, CF, stridor, and chronic cough undergoing clinically indicated bronchoscopy has been reported (3, 7, 8, 12).

Processing and Analysis

Once the lavage has been collected there are no standard methods for processing samples. The first aliquot, or bronchial wash, can be analyzed separately or pooled. If the desire is to study airways rather than alveolar inflammation, then this sample is important and should be analyzed separately. However, we do not know if the differences between the bronchial wash and pooled samples noted in healthy children also occur when patients with asthma are studied. It is important that studies record the methods employed. Other processing issues that have not been standardized include mucous filtration of the sample, the use of diluted cell suspensions and cytospins, and the number of cells counted.

Total cell counts are performed on the raw lavage fluid or on fluid that has been processed by methods such as filtration through gauze and centrifugation. Filtration through gauze may result in the preferential loss of epithelial cells and macrophages. Centrifugation of samples and resuspension in buffer solutions could cause an unpredictable loss of "light" cells at low centrifugation speeds or result in the destruction of cells at high spin speeds. These methods could therefore result in an apparent increase in the other cell types. Cytospins may also cause false counts, especially with respect to lymphocytes. These procedures may not be necessary and in some studies have been replaced with the glass coverslip method (24, 25).

At least 500 cells should be counted and macrophages, lymphocytes, neutrophils, eosinophils, and mast cells enumerated. Data on mast cell counts are particularly rare. These cells require specialized fixation and staining methods and are difficult to detect with routine stains. Furthermore, several thousand cells must be counted (e.g., 3,000-5,000 cells after fixation with Carnoy's fixative and staining with toluidine blue) to obtain an accurate picture.

Because of the lack of standardization and variations described above, BAL, in its current form, must be regarded as an imprecise tool.

	WHAT DO WE NEED TO KNOW?

TOP WHAT DO WE KNOW? WHAT DO WE NEED... HOW CAN WE ACHIEVE... REFERENCES

1. Is there a minimum standard for the sampling and processing of BAL? Which parts of the process should be standardized (e.g., bronchoscope/catheter size related to sample volume, dwell time, suction pressure, and location)?

2. A detailed comparison between bronchoscopic and nonbronchoscopic methods of obtaining BAL fluid is required to determine whether results can be regarded as interchangeable.

3. Is there a suitable external marker that would safely and accurately allow calculation of the dilutional effect on sampled ELF?

4. How do BAL fluid results compare with mucosal biopsy and induced sputum as methods for assessing airway inflammation? Importantly, do the findings from BAL reflect what is happening in the tissues? Some children with persistent day-to-day asthma symptoms have no evidence of airway inflammation from BAL studies (2). Does BAL fail to detect inflammation in some patients?

5. What is the repeatability of the BAL sampling and analysis procedures?

	HOW CAN WE ACHIEVE THIS?

TOP WHAT DO WE KNOW? WHAT DO WE NEED... HOW CAN WE ACHIEVE... REFERENCES

1. A small number of factors may be responsible for the variations in cellular and acellular data between centers (e.g., how the first aliquot of "bronchial wash" is dealt with). A detailed critical appraisal of all published studies may help to determine which factor(s) is/are the most important.

2. Postmortem studies of children of different ages with normal lungs and who have died of other causes should include a radiograph taken before and shortly after instillation of BAL fluid (13). The catheter or bronchoscope size and volume of fluid could be varied. This type of study would aid comparison of techniques and would be helpful for standardizing the procedure.

3. More information on the safety of bronchoscopic mucosal biopsy needs to be published. This information will be important for research ethics committees. Comparisons between BAL and mucosal biopsy results could be undertaken.

4. Repeatability BAL studies will be difficult to justify ethically. Children attending for multiple elective surgical procedures could be studied. Stable children with no pulmonary disease in the intensive care unit could be studied serially but the effects of mechanical ventilation and prolonged intubation may be associated with changes in the airways.

	Footnotes

Correspondence and requests for reprints should be addressed to Michael Shields, M.D., Department of Child Health, Queens University Belfast, and Institute of Clinical Science, Belfast, BT 12 6BJ, Northern Ireland. E-mail: m.shields{at}qub.ac.uk

	References

TOP WHAT DO WE KNOW? WHAT DO WE NEED... HOW CAN WE ACHIEVE... REFERENCES

1. Ferguson, A. C., and F. W. M. Wong. 1989. Bronchial hyperresponsiveness in asthmatic children: correlation with macrophages and eosinophils in broncholavage fluid. Chest 96: 988-991 [Abstract/Free Full Text].

2. Stevenson, E. C., G. Turner, L. G. Heaney, B. C. Schock, R. Taylor, T. Gallagher, M. Ennis, and M. D. Shields. 1997. Bronchoalveolar lavage findings suggest two different forms of childhood asthma. Clin. Exp. Allergy 27: 1027-1035 [Medline].

3. Marguet, C., F. Jouen-Boedes, T. P. Dean, and J. O. Warner. 1999. Bronchoalveolar cell profiles in children with asthma, infantile wheeze, chronic cough, or cystic fibrosis. Am. J. Respir. Crit. Care Med. 159: 1533-1540 [Abstract/Free Full Text].

4. Schellhase, D. E., D. D. Fawcett, G. E. Schutze, S. Y. Lensing, and A. F. Tryka. 1998. Clinical utility of flexible bronchoscopy and bronchoalveolar lavage in young children with recurrent wheeze. J. Pediatr. 132: 312-318 [Medline].

5. Khan, T., J. Wagener, T. Bost, J. Martinez, F. Accurso, and D. Riches. 1995. Early pulmonary inflammation in infants with cystic fibrosis. Am. J. Respir. Crit. Care Med. 151: 1075-1082 [Abstract].

6. Ozdemir, A., M. A. Brown, and W. J. Morgan. 1997. Markers and mediators of inflammation in neonatal lung disease. Pediatr. Pulmonol. 23: 292-306 [Medline].

7. Riedler, J., J. Grigg, C. Stone, G. Tauro, and C. F. Robertson. 1995. Bronchalveolar lavage cellularity in healthy children. Am. J. Respir. Crit. Care Med. 152: 163-168 [Abstract].

8. Midulla, F., A. Villani, R. Merolla, L. Bjermer, T. Sandstrom, and R. Ronchetti. 1995. Broncholaveolar lavage studies in children without parenchymal lung disease: cellular constituents and protein levels. Pediatr. Pulmonol. 20: 112-118 [Medline].

9. Ratjen, F., M. Bredendiek, M. Brendel, J. Meltzer, and U. Costabel. 1994. Differential cytology of bronchoalveolar lavage in normal children. Eur. Respir. J. 7: 1865-1870 [Abstract].

10. Braun, J., A. Mehnert, K. Dalhoff, K. J. Wiessman, and F. Ratjen. 1997. Different BALF protein composition in normal children and adults. Respiration 64: 350-357 [Medline].

11. Ratjen, F., and J. Bruch. 1996. Adjustment of bronchoalveolar lavage volume to body weight in children. Pediatr. Pulmonol. 21: 184-188 [Medline].

12. Pohunek, P., H. Pokorna, and I. Striz. 1996. Comparison of cell profiles in separately evaluated fractions of bronchoalveolar lavage (BAL) fluid in children. Thorax 51: 615-618 [Abstract/Free Full Text].

13. Grigg, J., S. Arnon, and M. Silverman. 1992. Fractional processing of sequential bronchoalveolar lavage fluid from intubated babies. Eur. Respir. J. 5: 727-732 [Abstract].

14. Heaney, L. G., E. C. Stevenson, G. Turner, I. S. Cadden, R. Taylor, M. D. Shields, and M. Ennis. 1996. Investigating paediatric airways by non-bronchoscopic lavage: normal cellular data. Clin. Exp. Allergy 26: 799-806 [Medline].

15. Grigg, J., J. Riedler, and C. F. Robertson. 1999. soluble intercellular adhesion molecule-1 in the bronchoalveolar lavage fluid of normal children exposed to parental cigarette smoke. Eur. Respir. J. 13: 810-813 [Abstract].

16. Grigg, J., J. Riedler, and C. F. Robertson. 1999. Bronchoalveolar lavage fluid cellularity and soluble intercellular adhesion molecule-1 in children with colds. Pediatr. Pulmonol. 28: 109-116 [Medline].

17. Darlow, B. A., K. B. Sluis, T. E. Inder, and C. C. Winterbourn. 1997. Endotracheal suctioning of the neonate: comparison of two methods as a source of mucus material for research. Pediatr. Pulmonol. 23: 217-221 [Medline].

18. Zach, M. S. 1996. Bronchoalveolar lavage: don't forget the size of the bronchoscope. Pediatr. Pulmonol. 22:329-331. [And reply (F. Ratjen)]

19. Ward, C., M. Duddridge, J. Fenwick, S. Williams, P. V. Gardiner, D. J. Hendrick, and E. H. Walters. 1992. The origin of water and urea sampled at bronchoalveolar lavage in asthmatic and control subjects. Am. Rev. Respir. Dis. 146: 444-447 [Medline].

20. Limper, A. H., U. Specks, W. M. Brutinel, W. J. Martin II, and M. S. Rohrbach. 1993. Interlobar variation in the recovery of bronchoalveolar lavage fluid, cell populations and angiotensin-converting enzyme. J. Lab. Clin. Med. 121: 785-791 [Medline].

21. Haslam, P. L.. 1998. BAL standardization and measurement of acellular components. Eur. Respir. Rev. 8: 1066-1071 .

22. Grigg, J., J. Riedler, and C. Robertson. 1994. Bronchoalveolar lavage of children without pulmonary pathology using a wedged suction catheter (abstract). Am. Rev. Respir. Dis. 149: A371 .

23. Schellhase, D. E., J. R. Tamez, A. A. Menendez, M. G. Morris, G. W. Fowler, and S. Y. Lensing. 1999. High fever after flexible bronchoscopy and bronchoalveolar lavage in noncritically ill immunocompetent children. Pediatr. Pulmonol. 28: 139-144 [Medline].

24. Laviolette, M., M. Carreau, and R. Coulombe. 1988. Bronchoalveolar lavage differential on microscope glass cover. Am. Rev. Respir. Dis. 138: 451-457 [Medline].

25. Walters, E. H., and P. V. Gardiner. 1991. Bronchoalveolar lavage as a research tool. Thorax 46: 613-618 [Free Full Text].