Use of a single TSH-measurement after oral thyrotrophin releasing hormone: an economical and highly sensitive thyroid screening test

Abstract

The increment of serum TSH after thyrotrophin-releasing hormone (TRH) is usually considered an extremely sensitive thyroid test, particularly useful in borderline disturbances. We evaluated 195 patients by measuring serum TSH before and 3 h after 40 mg TRH po, and by various other thyroid tests. We found a very tight correlation between the TSH\x=req-\ increment and the 3 h TSH. Thus the TSH-increment can reliably be estimated from the 3 h TSH. The Mann\x=req-\ Whitney test confirmed that the 3 h TSH identified thyroid dysfunction as well as the TSH-increment. The basal TSH can therefore be omitted and we propose the measurement of the 3 h TSH as an initial screening test. This single measurement economically excludes the many euthyroid patients sent for thyroid evaluation from further costly tests.The usual laboratory investigation of suspected hyperthyroidism starts with measurements of thyroxine and triiodothyronine together with some test of binding capacity in serum. If the thyrotoxic patient suffers from another severe illness, or if thyrotoxicosis is due to an autonomous nodule or occurs in an elderly patient, these tests are not uncommonly within the normal range or only marginally elevated (Pohl et al. 1973 ;Birkhäuser et al. 1977;Chopra et al. 1980;Köbberling et al. 1981). Further tests are then ordered, including a scintiscan and a stimulation test with thyrotrophinreleasing hormone (TRH). If the latter is normal, the patient is proved to be euthyroid and the previous tests appear wasteful in retrospect. We suggest that it is more economical to start evalua¬ tion with the TRH-test and that with the oral modification of the test (Staub et al. 1976(Staub et al. , 1979, TSH needs to be measured in one single serum specimen only. Further thyroid tests are necessary only in patients with an abnormal TRH-test. Patients and MethodsWe analyzed 195 unselected patients from our in-and out-patient department (47 men, 148 women, age 17-92 years) who had had an oral TRH-test with one 40 mg tablet TRF (Roche) after an overnight fast (Staub et al. 1976(Staub et al. , 1978(Staub et al. , 1979. Serum TSH was measured before and 3 h after TRH by radioimmunoassay (Ria-mat from Byk-Mallinckrodt, lower and upper limits of detection 0.6 and 80 mU/1, respectively, interassay variation 6.7%). The following additional tests were available: serum thyroxine (110 patients, normal range 5.0-11.5 ug/dl), serum triiodothyronine (70 patients, normal range 80-190 ng/dl), both by Amersham radioimmunoassay kits; T3-resin-uptake (110 patients, normal range 89-112% of control pool serum) by Thyopac Amersham kit; and thyroid scan (51 patients). The thyroid status of the patients was judged by one of us (H.B.), using all available clinical and laboratory information. Fig. 1 shows the very tight statistical correlation between the 3 h TSH and the TSH-increment, suggesting that the 3 h TSH provides as much diagnostic information as the TSH-increment. Results