Enzyme-linked immunosorbent microassay and hemagglutination compared for detection of thyroglobulin and thyroid microsomal autoantibodies.

We have evaluated for their potential use in the routine clinical laboratory enzyme-linked immunosorbent assays (ELISA) for human thyroglobulin antibodies (hTg-Ab) and microsomal antibodies (M-Ab). Results are expressed in terms of an "ELISA Index," based on comparison with a laboratory standard. The specificity of both ELISA assays is shown by dose-dependent inhibition of the hTg-Ab and M-Ab activities of the laboratory standards by the appropriate specific antigens. Similar concentrations of ovalbumin had no significant effect on the standard activity in both assays. For consecutive samples evaluated for hTg-Ab (n = 113) and M-Ab (n = 106) by ELISA and hemagglutination, rank order analysis of the results showed a highly significant correlation between the methods (r = 0.81, p = less than 0.001 for hTg-Ab; r = 0.82, p = less than 0.001 for M-Ab). However, 8/47 (17%) of samples positive in the hTg-Ab ELISA were negative by hemagglutination, and 7/69 (12%) of samples positive in the M-Ab ELISA were negative by hemagglutination. We effectively excluded the possibility of false positivity of these specimens by ELISA by blocking specimen positivity with the specific antigens in 12 of 14 specimens investigated. We conclude that ELISA techniques for human thyroid autoantibodies are sensitive and specific, easy to initiate, objective, and capable of use in large-scale screening. They are superior to standard hemagglutination techniques by having an increased detection rate for hTg-Ab and M-Ab.