Endocrine Abstracts (2007) 13 P246

A subnormal peak cortisol response to stimulation testing does not predict a subnormal cortisol production rate (CPR)

Angela Paisley1, Susannah Rowles1, Margaret Roberts1, David Brandon2 & Peter Trainer1


1Christie Hospital, Manchester, United Kingdom; 2Oregon Health and Science University, Medicine, United States.


In patients with pituitary disease stimulation tests, such as the insulin tolerance test, are performed as tests of the HPA axis and widely used as the basis for commencing hydrocortisone replacement therapy.

We conducted a pilot study to investigate the relationship between the peak cortisol response to insulin-induced hypoglycaemia and daily CPR. Isotopic CPRs were studied in 10 patients (5 male, mean age 44±13 years) with pituitary disease and a suboptimal peak cortisol response (350–500 nmol/L) to ITT (normal response >500 nmol/L). Previous studies demonstrate CPR to be 4.9–29 mg/day in healthy volunteers [Purnell et al JCEM 2004; Esteban et al JCEM 1991]. 0.594 mg deuterium-labelled cortisol was infused over 30 hours, which equates to approximately 4% of the anticipated CPR. After 6 hours, once a steady state had been achieved, blood was obtained at 30 minute intervals for 24 hours. The proportions of native and deuterated cortisol were estimated by mass spectroscopy enabling calculation of the CPR. 24 hour urinary free cortisol was also measured.

The median peak cortisol attained with hypoglycaemia was 473.5 (range 366–492) nmol/L, and the baseline 09.00 cortisol 266 (164–320) nmol/L. Median 24-hour CPR was 8.46 (4.2–29.3) mg/day and 24-hour urinary cortisol 116.5 (20.5–265.9) nmol/L. The CPR was in the above reference range in 8 of 10 patients. No correlation was found between the peak stimulated cortisol and either the isotopically calculated CPR (r=0.55, P=0.10), or 24 hour urinary free cortisol (r=−0.25, P=0.49). A positive correlation was found between weight and CPR (r=0.73, P=0.02).

In summary, this pilot study indicates that a subnormal peak stimulated cortisol response does not predict an inadequate cortisol production rate and therefore decisions to commence long-term hydrocortisone should not be based on stimulated cortisol values.

Article tools

My recent searches

No recent searches.