Effect of smoking on bronchoalveolar lavage constituents.

Increasing knowledge has accumulated relating to the effect of cigarette smoking on cells and other constituents in bronchoalveolar lavage (BAL) fluid. Such knowledge is important for the correct evaluation of BAL results. When BAL records are interpreted, smoking has to be considered as a major confounding factor. In this issue of the Journal, BuRKE et al. [1] add a new aspect to the adverse effects of smoking on the alveolar microenvironment. They suggest that the apparent epithelial lining fluid (ELF) volume retrieved from smokers is 100% greater than from nonsmokers. Clearly, this finding would also affect the cell and protein concentrations in studies using the ELF volume as reference value. Comparable differences in ELF volumes among different study groups have not previously been reported. Although the mechanism of this increase remains unclear, the authors suggest that either increased production or decreased clearance of ELF could be involved, but not enhanced capillary permeability, since a parallel jncrease in albumin was not observed. Technical difficulties, however, could also account for the observed effect. One limitation lies in the use of urea to estimate the ELF volume [2). The basic disadavantage of this method is that urea passively diffuses from plasma and the interstitium into the BAL fluid during the lavage dwell-time [3]. In addition, the fluid dynamics during BAL are complex [4]. The hydrostatic pressures induced by the BAL procedure itself may cause unpredictable movements of water and low molecular weight solutes between the blood and the BAL fluid [5]. On these lines, the data by BURKE et al. [1] might be interpreted quite differently: smokers may be more susceptible to increased water and urea movement during the BAL procedure; this would give rise to an artificially increased ELF volume with a consequent artificial decrease in albumin due to the dilution of the ELF volume during the BAL procedure. In this context, it is of inter~st that another study did not show a significant difference in the concentration of albumin expressed as ng·ml· ' ELF between asymptomatic smokers and normal nonsmokers [6]. The ideal marker for the assessment of the true ELF volume is yet to be found [7]. BuRKE et al. [1] are correct when stating that their findings should be confirmed using other methods of ELF determination.