Cytotoxic leukocyte reaction.

Blood samples from 49 subjects were submitted to cytotoxic leukocyte tests, which were read double-blind by two laboratory technicians. The results were identical in 78% of the tests and the statistical error of the method was 0.46. In a group of 32 subjects, the cytotoxic test was repeated on 3 consecutive days. Deviations in the results occurred in 18--20% of the tests, the statistical errors varying between 0.38 and 0.43. In a series of serum samples (75 subjects) with positive leukocyte reactions to food antigens, various in vitro procedures were carried out in connection with the cytotoxic test. The following observations were made: In 42 of 58 tests, a positive leukocyte reaction was transferred with the reactive serum to cells from non-reactive subjects. After heating of the reactive serum (56 degrees C, 2 h) and washing of the cells, reactivity was no longer observable in 44 out of 52 tests. Treatment with EDTA consistently blocked expected cytotoxic reactions in 36 tests. DSCG inhibited expected reactions in 45%, antihistamine in 93%, and cortisone in 20% of the tests.