[Quality and selection of samples of bronchoalveolar lavage (BAL) in diffuse pneumopathies]

Bronchoalveolar lavage (BAL) samples used to study immune or inflammatory response in interstitial lung disease must be representative of the lower respiratory tract. Thus, the selection of suitable samples must be part of routine practice. To assess the incidence of unsuitable BAL samples used for cytology and to determine the relation between parameters related to underlying disease and the quality of samples. One hundred sixty-one patients were enrolled. Seventy-two were HIV positive and had diffuse pulmonary infiltrates, 34 had idiopathic pulmonary fibrosis (IPF), 10 had sarcoidosis, 10 had hypersensitivity pneumonitis, 19 had interstitial lung disease and collagen diseases and 2 had pulmonary eosinophilia. Fourteen individuals formed the control group. The quality study was carried out by staining the BAL samples following a modified Wright-Giemsa technique and evaluating the samples by the selection criteria described by Chamberlain and colleagues (1987). We identified unsuitable samples from 53% of the HIV positive patients, from 35% of the IPF patients and from 21% of the interstitial lung disease patients with associated connective tissue disease. In the other groups, all samples were suitable for analysis. Intolerance of BAL with decreasing percentage of fluid recovered was significantly associated with sample quality, particularly in the IPF group. The cytology results that invalidated the samples differed by group. In all groups, unsuitable specimens had low cell counts. The finding or not of evidence of associated infection in HIV-infected patients, on the other hand, did not appear to determine sample quality in and of itself, although it did in samples related to other entities. We can predict that a high rate of unsuitable BAL samples will come mainly from patients with diffuse lung disease associated to HIV infection, IPF and interstitial lung disease with associated connective tissue disease. Tolerance to the technique influences quality of the specimen obtained and, therefore, should be taken into account in interpreting the findings of cytology. The criteria applied by the various teams using BAL should be unified, and it should be determined whether the exclusion of inappropriate samples affects the final composition of study groups.