Lectin-dependent cell-mediated cytotoxicity and natural killer function in rejecting and infected lung allografts.

Differentiation between rejection and infection of lung allografts remains difficult. The effects of these two pathologic entities on the cytolytic activity of bronchoalveolar lavage (BAL) and PBL were investigated. Left lung allotransplantation was performed on 16 mongrel dogs of which 12 were available for complete studies. All animals received CA, AZA, and PRED for 2 weeks. Four grafts developed left lower lobe Gram negative pneumonia. The eight remaining recipients progressed gradually to severe rejection after acute reduction of immunosuppression. Cytolytic activity of blood and left lung BAL lymphocytes was quantitated by the natural killer (NK) and lectin-dependent cell-mediated cytotoxicity (LDCMC) assays. Two additional groups serving as controls were either given a 10-day course of immunosuppressants or had right lower lobe pneumonia induced by transbronchial inoculation of gram negative bacteria. Immunosuppressed control animals showed significant depression of PBL and BAL lymphocyte LDCMC and NK activity. Similarly, BAL lymphocytes expressed very low LDCMC in normal allografts (2.8 +/- 0.8%). Once rejection developed and progressed, LDCMC became significantly higher (15.6 +/- 2.2 and 52.7 +/- 2.8% in mild and severe rejection, respectively). There was no detectable NK activity in rejecting lung allografts. BAL lymphocytes from infected allografts, on the other hand, showed an elevation of both NK and LDCMC activity (9.1 +/- 1.1 and 14.6 +/- 1.0%, respectively). Similarly, bacterial pneumonia in control animals manifested an increase in NK and LDCMC activity in lung and blood. PBL lymphocytes of lung allograft recipients, however, had increased NK and LDCMC activity in both rejection and infection. LDCMC/NK activity ratio (LM/NK index) of lung lymphocytes was significantly higher in rejecting allografts (11.2 +/- 1.0 and 12.4 +/- 1.6 for mild and severe rejection, respectively) than in infected ones (1.2 +/- 0.3, P < 0.0001). It appears, from this study, that rejection of the lung allograft results in alterations in BAL lymphocyte phenotypes and functions that differ from those associated with bacterial infection. Such differences may be useful in distinguishing episodes of acute allograft rejection from bacterial infection.