Two-hour Adrenocorticotropic Hormone Test: Accuracy in the Evaluation of the Hypothalamic- Pituitary-Adrenocortical Axis

Abstract

Summary: The inter- and intraindividual variations of response to an iv bolus of tetracosactide (0.25 mg/1.73 m2) were studied, and a definitive diagnostic 2-hr test was established. The test was given to 179 subjects: hypopituitary, prednisone-treated ne-phrotics, patients with the autoimmune polyendocrinopathy-candidiasis syndrome (APECS), and reference subjects. It was repeated in 62 cases and the total of 84 test pairs was analyzed. The test was started either at 8–9 AM or at 10–11 AM, 2 hr after the injection of insulin for an insulin test. Thirty subjects had both the ACTH test and the insulin-ACTH test. Capillary blood samples for fluorometric determination of plasma cortisol were taken 1, 2, and 3 hr after the injection.The distributions of the basal and stimulated levels were positively skewed, and log transformation of the data gave clearly more appropriate statistics. In most subjects the cortisol level was highest 2 hr after the injection. There was no significant difference in the response to ACTH whether it was given alone or in the insulin-ACTH test. All the poststimulation cortisol levels showed a positive interindividual correlation with the basal levels, and all the increments a negative interindividual correlation with the basal levels.In repeated tests, the intraindividual variation in the responses was directly correlated with the level of the response, and this correlation could be removed by dividing the intrapair difference in the values of the parameters by the sum of the values for percentage of intrapair difference (PIPD). The SD of the distribution of PIPDs was used for an index of precision. The 2-hr cortisol level was substantially more precise than either the basal level or the increment. In analysis of correlation between PIPDs, variation of the 2-hr level showed no significant dependence on the variation of the basal level, but the variation of the 2-hr increment was highly dependent on the variations of both the basal level and the 2-hr level. Thus the 2-hr plasma cortisol level is the best single index of the response. A better definition of normal result is given by a reference area on a plot of 2-hr level versus basal level (Fig. 2.).Of 31 tests given to hypopituitary subjects with deficient cortisol responses to insulin, all but 7 gave subnormal basal and/ or 2-hr levels. In 8 the increment was also subnormal. In the nephrotic subjects, all but one of 36 tests given at the end of continuous prednisone medication gave subnormal basal and/or 2-hr levels; the increment was subnormal in 26. At the end of the following period of intermittent prednisone, 12/36 tests gave normal results and none failed to elicit a response. The increment remained subnormal in 6 despite normal basal levels.A high normal response to ACTH was constantly associated with a normal response to insulin, and was thus indicative of normal ACTH secretion. In the reference subjects, there was a highly significant correlation between the responses to insulin and ACTH. The mean ± SD of PIPDs between these responses was −9.2 ± 13.2%. Of the APECS patients, seven with incipient adrenocortical failure had a normal increment despite subnormal basal and stimulated levels. Thus the ACTH test did not differentiate this state from ACTH deficiency.Speculation: When the criteria for a normal finding in the 2-hr iv ACTH test have been carefully established, this test will serve as a definitive procedure for the diagnosis and follow-up of states associated with ACTH deficiency, as well as in primary diseases of the adrenal cortex. Like the other ACTH tests, however, it will not differentiate with certainty between primary and secondary adrenocortical failure.