Phenotypic expression of bronchoalveolar lavage cells in lung rejection and infection

The differentiation of episodes of lung allograft rejection from infection continues to be a problem. Bronchoalveolar lavage (BAL) has recently gained some success in the diagnosis and management of interstitial lung disease. To assess the usefulness of BAL in differentiating between lung allograft rejection and infection, we examined the differences in cellular subsets of BAL and peripheral blood (PBL) samples in a controlled canine model of rejection or pneumonia. Single-lung allotransplants were allowed to undergo rejection by withdrawing immunosuppressive agents (n = 6). In another group of dogs (n = 5), pneumonia was induced by transbronchial injection of Pseudomonas aeruginosa and melted agar followed by bronchial fulguration. Cells obtained from bronchoscopic BAL and PBL samples were labeled with functionally characterized cross-reactive murine monoclonal antibodies. Transthoracic needle biopsies and transbronchial biopsies were done to assess their adequacy in examining the rejecting or infected lungs and were compared with open lung biopsies. We found the following: (1) the percentage of DT2-labeled cells was significantly higher (p less than 0.05) in BAL samples from rejecting lungs compared with infected lungs; (2) the PBL/BAL ratio of DT2-labeled cell percentages was significantly higher in pneumonia (1.7 +/- 0.3) than rejection (0.5 +/- 0.2) (p less than 0.004); (3) the percentage of E11-labeled cells in PBL samples was significantly higher (p less than 0.02) in rejection than in infection; and (4) the ratio of WIG4 to DT2 cellular subset percentages in BAL samples from rejection (26.8 +/- 9.9) was significantly lower than from infection (61.0 +/- 22.9) (p less than 0.03). Transthoracic and transbronchial biopsies did not always yield representative specimens.(ABSTRACT TRUNCATED AT 250 WORDS)