Passive anaphylaxis in human lung fragments as a model for testing anti-allergic drugs: its variability and constraints.

Histamine release from human lung fragments, passively sensitized and challenged with antigen under standardized experimental conditions, varied between 0 and 41.2% (mean +/- SD 15.6 +/- 10.0%) in 89 experiments. Over 41 lungs, the mean coefficient of variation for release from 10 to 28 tissue replicates was 21.4% (range 7-44%). Sodium cromoglycate and chlorpromazine were both partial antagonists of histamine release producing, at best, 30-40% inhibition. The cromoglycate analogue, bufrolin, showed similar activity. There was considerable variation in the effects of these drugs both within and between experiments. Salbutamol was a more potent and more effective inhibitor of release but it, too, showed variability. Although theoretically a very appropriate model of allergic asthma, passive anaphylaxis in human lung fragments is quantitatively inconsistent and gives only a gross indication of drug efficacy.