Investigators will conduct a study looking at healthy subjects to establish the expected range of peak cosyntropin-stimulated cortisol levels in normal healthy subjects with two different cosyntropin administration modalities. The goal of this study is to compare intravenous (IV) and intramuscular (IM) administration, and to generate a normal range of post-cosyntropin serum cortisol measure with the new assay.
The two aims of the study are: (1) Compare the cortisol values at baseline, 30 minutes and 60 minutes after IV and IM cosyntropin stimulation testing in the same subjects. (2) Study the values obtained using the new Roche Cortisol II assay and establish a new reference range for cosyntropin-stimulated cortisol levels in healthy subjects for both cosyntropin administration modalities.
The widely-used old Elecsys® Cortisol Immunoassay from Roche Diagnostics (Cortisol I) is now being replaced worldwide by a new Elecsys® Cortisol assay from Roche Diagnostics (Cortisol II). The new Cortisol II assay is also a competitive immunoassay, but employs a monoclonal as opposed to a polyclonal capturing antibody, which was used in the old Cortisol I assay. This results in greater specificity for cortisol with less interference from other endogenous or exogenous glucocorticoids. The new Cortisol II assay was shown to have strong correlation with cortisol levels measured by liquid chromatography-tandem mass spectrometry (LC-MS/MS) which is considered the gold standard for cortisol measurement. It has been shown in earlier studies that Cortisol II assay generates cortisol values that are approximately 20-30% lower than the older Cortisol I immunoassay with proposed new cortisol cutoff for the diagnosis of adrenal insufficiency between 350-375 nmol/L (12.7-13.6 µg/dL).
This is a prospective study on healthy subjects. Investigators will perform two cosyntropin stimulation tests on participants on two separate visits, one using intravenous and another one using intramuscular cosyntropin (250 µg). Investigators will allow a minimum of one week and a maximum of 2 months between the two tests.
Blood samples for each of the two tests will be drawn at baseline, before cosyntropin administration, then at 30 minutes and 60 minutes after cosyntropin administration from indwelling catheter. A maximum of 5 ml of blood will be drawn at each time point. Serum will be separated, and samples will be stored at - 80o C after collection until analyzed.
All of the blood samples will be tested to measure cortisol levels using the new Roche Cortisol II assay. The Cortisol assay kits will be provided from Roche Diagnostics.
Overall study duration will be 18 months. There will be two study visits. Investigators will allow a minimum of one week and a maximum of 2 months between the two visits.
The study will include 40 healthy subjects between the ages of 18 and 65. Investigators will exclude subjects with known hypothalamic-pituitary adrenal axis or other endocrine disease, glucocorticoid use (chronic, current, or within 3 months), pregnancy or oral estrogen use. Women will be studied during the same menstrual cycle phase for each of the 2 tests. Participants will be recruited by word of mouth and adds placed within the medical school campus. Participants will provide written informed consent.
Study Statistics
1. Primary outcome variable.
1. Comparison between peak cosyntropin-stimulated cortisol levels measured using IV and IM cosyntropin administration in the same subjects
2. Establishing new reference range for expected peak cosyntropin-stimulated cortisol levels using the new Cortisol II assay for both cosyntropin administration modalities.
2. Secondary outcome variables.
3. Statistical plan including sample size justification and interim data analysis.
To test the hypothesis that the cortisol levels after IM administration of cosyntropin is slightly lower than IV administration, the investigator assume that 1) proportional analytical variability (constant CV) for both methods; 2) a difference \> 50% of analytical variability is clinically meaningful (i.e., an effect size of 0.5). A sample size of 30 subjects will provide 85% power at a significance level of 5% using one-sided paired t-test. When the two time point data are analyzed together, the power will be higher.
For correlation estimation, the sample size of 30 produces a two-sided 95% confidence interval with a width \< 0.153, assuming the sample correlation is \>0.900.
With the proposed 40 healthy subjects, the distance of the reference range (defined as 95 percentile of the reference population) from the upper-bound of the estimated 95% confidence interval will be \< 22%.
