Circulating antitriiodothyronine autoantibodies in two euthyroid patients: apparent lack of interference in total T3 radioimmunoassay based on second antibody or solid phase separation techniques.

Two clinically euthyroid patients were noted to have low total T3 levels as assessed by RIA using either dextran-charcoal (DC) or polyethylene glycol (PEG) for separation of bound from unbound T3, in spite of normal free T3, total and free T4 and basal and TRH-stimulated TSH concentrations. The presence of circulating substances binding T3 was suggested by high nonspecific binding in total T3 RIA system using either DC or PEG separation. The presence of anti-T3 autoantibodies was then suspected and confirmed by the presence of [125]-T3 bound to patients' gammaglobulins, precipitated with rabbit anti-human immunoglobulins. Serum T3 concentration determined by extracting T3 from patients' sera with methanol was 166 and 226 ng/dl. Similar or even lower values were unexpectedly obtained in RIA systems with solid phase or second antibody (anti-rabbit) separation and with competitive protein binding assay. To face this paradoxical finding, simulated experiments were carried out by incubating T3- and T4-free sera added with various amounts of stable T3 and T4 in the presence of goat anti-T3 or anti-T4 serum. These samples were then radioimmunoassayed. The DC separation caused a consistent underestimation of the actual T3 and T4 concentration. The second antibody separation caused a T3 and T4 overestimation for actual levels below 200 ng/dl and 10 micrograms/dl, respectively, while at the higher T3 or T4 concentrations, an overlap or, even, an underestimation of actual T3 or T4 levels were found. These data provide evidence that, with second antibody or solid phase separation methods, there could be an apparent lack of interfering effect of endogenously occurring antibodies.