Basal TSH levels compared with TRH-stimulated TSH levels to diagnose different degrees of TSH suppression: diagnostic and therapeutic impact of assay performance.

Abstract

The estimated prevalence of endogenous subclinical hyperthyroidism varies from 4% to 6% and a basal thyroid stimulating hormone (TSH) level < 0.5 mU L-1 may be associated with increased mortality in subjects over 60 years of age who are not on thyroid medication. Exogenous TSH suppression is a mainstay in the treatment of thyroid cancer. Because of recent concerns about potential adverse effects, especially of endogenous TSH suppression on bone, the cardiovascular system and cognitive functions, subclinical hyperthyroidism obtained new clinical importance. We therefore re-evaluated the diagnostic value of basal and thyrotrop in TRH-stimulated serum TSH measurements using TSH assays with different sensitivities. A total of 805 oral and nasal TRH stimulation tests were performed on 409 ambulatory subjects with low basal serum TSH concentrations of less than 0.1 mIU L-1. Basal serum TSH was measured either using a second generation assay (functional sensitivity > 0.03 mIU L-1) or two third generation assays (functional sensitivity 0.01 mIU L-1 and 0.007 mU L-1, respectively). Serum TSH concentration was determined before and 3 h after oral administration of 40 mg of TRH and before and 30 min after nasal administration of 2 mg of TRH. In the oral testing group, the basal TSH levels measured by the different TSH assays were 0.06 +/- 0.03, 0.04 +/- 0.02 and 0.03 +/- 0.02, respectively, whereas the peak TSH levels were 0.4 +/- 0.6, 0.4 +/- 0.6 and 0.3 +/- 0.5 in the patients with subclinical hyperthyroidism. In overt hyperthyroidism, the basal TSH levels were 0.06 +/- 0.02, 0.03 +/- 0.02 and 0.03 +/- 0.02, whereas the peak TSH levels were 0.19 +/- 0.3, 0.16 +/- 0.3 and 0.15 +/- 0.2, respectively. Basal TSH values could discriminate between different degrees of TSH suppression if measured with a third generation assay (P < 0.001), but not with a second generation assay. There was only a weak correlation between basal TSH and peak TSH when measured by a second generation assay (n = 126; r = 0.3; P < 0.001) in contrast to the strong correlation found using the third generation assays (n = 128; r = 0.7; P < 0.001 and n = 69; r = 0.8; P < 0.001, respectively). In view of the recent concerns about potential adverse effects in TSH suppression and based on our data, it is mandatory to select a TSH assay with a functional sensitivity of < or = 0.01 mIU L-1 for optimal titration of L-T4 suppressive therapy, especially in patients with thyroid cancer. If, however, only a second generation TSH assay is available, additional TRH testing allows a more careful titration of suppressive thyroxine therapy.