Experience with the Bio-assay of the Long-Acting Thyroid Stimulator

Abstract

The assay procedure for identifying the long-acting thyroid stimulator was modified so that batches of 150 or more mice might be used. Variability of response to standard thyrotropin (at 2 and 9 hr after injection) and to the injection of serum known to contain the stimulator was studied; comparisons between responses seemed best carried out only with concomitant assays. A 2- or 9-hr peak effect in the assay could not always be identified when response to thyrotropin or to long-acting thyroid stimulator was slight, although often this did become apparent when allowance was made for responses to control solution. Logarithmic transformation of the raw cpm response, and subsequent adjustment by covariance analysis for the independent variability of the zero-hr count, was found to be statistically justifiable and of practical use. In particular, the use of transformed and adjusted data narrowed fiducial limits of responses and rendered statistically significant some responses that were otherwise not acceptable.