Biological Assay of Luteinizing Hormone (LH, ICSH) by the Ovarian Hyperemia Method of Ellis: An Evaluation1

Abstract

The results of extensive experience with biological assay of LH by the ovarian hyperemia method of Ellis, which employs radioiodinated serum albumin for estimating the hyperemic response, have been evaluated. A total of 101 assays involving approximately 5,000 animals were performed. The mean index of precision (λ) was 0.29. Doses of the reference preparation (NIH-LH-S1) which produced the most reliable slope and reduced otherwise large fluctuations in mean response to fixed doses from assay to assay were 1.0 and 4.0 μg. The mean percentage of standard error of the potency estimates of a series of 80 assays (Series I), using 4 rats/dose, was 57%, and for a series of 21 assays (Series II), using 6 rats/dose, was 44%. The interassay error was large in Series 1 but not significant in Series II, possibly because optimal doses of the reference preparation had been employed. In its responsiveness to sheep LH, the Ellis assay was 15 times more sensitive than the prostate assay of Greep and 0.4 times as sensitive as the ovarian ascorbic acid depletion assay of Parlow. The slope for LH in the presence of a large amount of FSH was found not to be flattened, contrary to a previous claim. However, evidence is presented suggesting that sheep FSH augmented the response to sheep LH.