[Unexpected high values of TSH: the presence of high molecular weight forms (macro TSH) must be investigated].

The laboratory methods usually employed for the measurement of serum TSH present sensitivity and specificity levels, both analytical and clinical, are highly satisfactory. Additionally, the methodologies are quite robust, so that false-positive and false-negative results are rare and unexpected. In this paper we describe two individuals quoted as euthyroid clinically, with no reference to autoimmune diseases, and no reference to the use of exogenous TSH, that presented with normal to extremely high serum TSH levels, depending on the method employed for analysis. In the three tested methods, serial dilution showed that the real TSH levels were between 250 and 300 mUI/L. In both cases the increment in TSH levels were due to the presence of TSH-binding proteins, forming high molecular weight complexes ("macro TSH"), well characterized by gel filtration chromatography on Superdex S-200 column. In one of the patients the binding protein was characterized as being IgG by protein-G binding study. In the other case, protein-G binding as well as anti-IgM binding failed to characterize the protein. These two cases call attention to the importance of the clinical-laboratory correlation and suggest the need that the presence of "macro TSH" must be investigated in patients with unexpectedly high TSH values.