Use of a sensitive thyrotropin assay for monitoring treatment with levothyroxine

To determine if a sensitive assay for thyrotropin (thyroid-stimulating hormone [TSH] ) would be useful as a primary test to determine optimal doses of levothyroxine, an enzyme immunoassay for TSH that discriminated hyperthyroid from euthyroid subjects with no overlap was compared with radioimmunoassays for TSH and thyroxine (T4) and the calculated free thyroxine index (FTI) in 100 patients receiving stable doses of levothyroxine. The basal TSH level, determined with the sensitive assay, predicted the TSH response to thyrotropin-releasing hormone (TRH); all patients whose basal TSH level was less than 0.35 mU/L had absent or subnormal responses, and all patients with TSH levels of 0.35 mU/L or greater had normal responses. In patients with normal TSH levels, the T4 level and FTI were superfluous (normal) or misleadingly high; conversely, 65% of patients with low TSH levels had normal T4 levels or FTIs, or both. An assay for TSH with sufficient sensitivity to discriminate between abnormally low values and the low range of normal can be used as the primary test for monitoring levothyroxine treatment. The T4 or FTI measurements are not needed if the TSH level is normal, but they should be done in patients with low TSH levels.