An evaluation of six solid-phase thyrotropin (TSH) kits.

This article describes an objective evaluation of six thyrotropin (TSH) kits. One was a radioimmunoassay kit taken for comparison, three were immunoradiometric assays and one was an immunoenzymometric assay. The laboratory internal immunoluminometric assay for thyrotropin was used to measure the concentrations of thyrotropin in the kit standards using a standard curve of WHO 68/38 international reference preparation in serum from a thyrotoxic patient as matrix. The in-house assay was used to demonstrate the "sensitivity" to citrated plasma and the fact that kit standards could only measure "correctly" when used in its own kit. The study was carried out in a "blind" way, the assayist and organiser not knowing from which of the four groups under test (blood donors - serum and citrated plasma, thyroliberin-test and thyroid outpatient clinic patients) the samples came until the study had been completed. The immunometric assay kits were able to differentiate statistically between euthyroid and untreated hyperthyroid patients, although one IRMA kit (Kit F) had a large "grey zone" where both euthyroid and hyperthyroid patients overlapped. Compound precision profiles covering the range 0-5 mU/l thyrotropin were good, a mean coefficient of variation under 5% within the range 0.5-5 mU/l being demonstrated by 3 immunometric assays. The immunometric assay kit with the most cumbersome methodology showed, as was to be expected, the worst precision. The euthyroid ranges for thyrotropin were similar in 3 immunometric assay kits using the WHO 68/38 reference material as calibrator (Kits C and E, 0.25-3 mU/l) and correlated well with one kit using the 2nd IRP (NIBSC 80/558) as calibration material (Kit D 0.33-4 mU/l), although the results were around 30% higher in Kit D. The second kit (Kit A) using the 2nd IRP material as calibrator gave identical values with the kits using the WHO 68/38 reference thyrotropin-preparation for calibration purposes. In a further kit (Kit F) it was stated that both thyrotropin international reference preparations gave rise to identical serum values when used as calibrators. The thyroliberin-test may have an additional role to play in monitoring returning pituitary function in thyrotoxic patients under treatment as the immunometric assay kits were easily able to measure a thyrotropin difference of 0.5 mU/l in the range 0-1 mU/l. The conventional radioimmunoassay (Kit B) was unable to match the precision and sensitivity of the better immunometric assay kits in the range under 1 mU/1.(ABSTRACT TRUNCATED AT 400 WORDS)