Background: Opioid analgesics such as morphine, fentanyl and oxycodone are being increasingly used in primary care for chronic non-malignant pain. Opioids inhibit the hypothalamic-pituitary-adrenal (HPA) axis. While there have been case reports of clinically significant adrenal insufficiency in people treated chronically with oral or transdermal opioids, the extent of the problem has not been well studied.
Methods: We have examined 19 patients (12 male, age 52.2 ± 2.7 years) attending a hospital-based Persistent Pain Service, treated with oral or transdermal opioids, mean morphine equivalent daily dose (MEDD) 120 mg (range 30 – 265 mg). Each subject had an estimate of cortisol, ACTH and DHEA-sulphate before 0900h. Those who had a 0900h cortisol of <250 nmol/L underwent a 250 μg ACTH1-24 test and an overnight metyrapone test (OMT), dose 30 mg/kg.
Results: 13/19 participants had a 0900h cortisol of <250 nmol/L, and overall the mean (± SEM) 0900h cortisol was 207 ± 22 nmol/L, range 29 – 389 nmol/L. 7/19 participants had an undetectable 0900h ACTH (<10 ng/L). The mean DHEA-S level was 2.2 ± 0.3 μmol/L (reference range 1-11 μmol/L). Of the subjects with a low basal cortisol, 6/12 failed either the 250 μg ACTH1-24 test or OMT, 1 failed both tests, and 1 has not attended for follow-up testing. There was no significant relationship between MEDD and 0900h cortisol level. However patients who failed a stimulation test had a significantly higher MEDD (183 ± 34 mg vs 90 ± 20 mg, P<0.05).
Conclusion: Over two thirds of chronic opioid users had low basal cortisol levels, of whom half failed at least one standard stimulation test of the HPA axis. These patients meet criteria which in the setting of structural pituitary disease would lead to glucocorticoid replacement. These findings have important implications for the identification and management of hormone deficiency in patients receiving opioids.