Polyclonal 3,5,3'-triiodothyronine (T3) antibodies in a euthyroid woman and their effect on radioimmunoassays for T3.

Thyroid function tests in a 17-yr-old euthyroid woman with a slightly enlarged thyroid gland were normal, except for an apparently elevated serum T3 level which was over 8 times greater than the upper limit of normal, when her serum was directly analyzed for T3 by an antibody-coated tube RIA method. When T3 analysis was performed with a RIA procedure that also used serum directly and separated free and bound radiolabeled hormone by polyethylene glycol precipitation, an absurd and conflictingly low result was obtained in which the bound fraction from the patient's serum contained 22% more radiolabeled T3 than the bound fraction from the zero T3 standard. These results suggested the presence of endogenous T3 antibodies which interfered with these RIA procedures. Further studies revealed the presence of IgG-kappa and IgG-lambda polyclonal antibodies that bound T3, but not T4, which accounted for the artifactual elevation in T3. The T3 antibodies were also detected by agarose electrophoresis of serum labeled with tracer quantities of [125I]T3. Analysis for T3 by RIA on Sephadex or RIA of an ethanol extract of the patient's serum both eliminated interference from these endogenous T3 antibodies. The T3 autoantibodies occurred in this patient even though there was no previous history of thyroid medication. Thus, endogenous thyroid hormone-binding antibodies produce artifactually abnormal results with RIA methods that assay serum directly, whereas inactivation and removal of the immunoglobulins by alcohol extraction or Sephadex obviates interference from such proteins.