Differences in serum cortisol by immunoassay and LC-MS methodologies in patients treated with 11b-hydroxylase inhibitors for hypercortisolism — ASN Events
  1. Schedule
  2. Joint Poster Session
  3. Differences in serum cortisol by immunoassay and LC-MS methodologies in patients treated with 11b-hydroxylase inhibitors for hypercortisolism

Differences in serum cortisol by immunoassay and LC-MS methodologies in patients treated with 11b-hydroxylase inhibitors for hypercortisolism (#402)

Daniel Yagoub 1 2 , Kevin Mantik 1 , Renee Sahertian 1 , Laura K Buckton 1 , Cherie Chiang 3 4 5 , Andrea Rita Horvath 1 2
  1. NSW Health Pathology - Prince of Wales Hospital, Sydney, NSW, Australia
  2. Faculty of Medicine and Health, University of New South Wales, Sydney, NSW, Australia
  3. Department of Endocrinology, Royal Hospital for Women, Sydney, NSW, Australia
  4. Department of Medicine, University of Melbourne, Melbourne, VIC, Australia
  5. Department of Internal Medicine, Peter MacCallum Cancer Centre, Melbourne, VIC, Australia

Background

11b-hydroxylase inhibitors such as osilodrostat are often used in overt Cushing’s syndrome for patients awaiting or ineligible for surgical treatment. These treatments have shown high efficacy, but hypocortisolism is a potential complication of therapy. Close monitoring of serum cortisol is therefore recommended by the 2025 Endocrine Society guidelines. However, some commercial immunoassays may produce falsely high cortisol results due to the potential accumulation and cross-reactivity of steroid precursors such as 11-deoxycortisol.

Methods

This study compared serum cortisol results by Roche’s immunoassay and a validated reference method by liquid-chromatography mass spectrometry (LC-MS) in patients: a) with Cushing’s syndrome before and after receiving 11b-hydroxylase inhibitors (n = 19; osilodrostat 10, metyrapone 6, ketoconazole 3), and b) with stress (n = 12). In 12 cases, in addition to Roche, cortisol was measured by Abbott, Siemens and Beckman immunoassays in order to compare performance. The LC-MS method also quantified cortisol precursors (11-deoxycortisol, 17-hydroxyprogesterone), mineralocorticoids (aldosterone, 11-deoxycorticosterone) and cortisone. 

Results

Immunoassays resulted in a large positive bias compared to cortisol measurements by LC-MS in patients with hypercortisolism before treatment (mean bias +41%, range 11-87%), a trend that was not observed in patients with physiologically high cortisol due to stress.

Following commencement of 11b-hydroxylase inhibitors, these discordances improved in many cases (mean bias +22%, range 16-57%), although particular immunoassay platforms (e.g. Siemens) remained significantly affected. The observed discrepancies between methods did not correlate with increasing concentrations of cortisol precursors during treatment. The contribution of as-yet unidentified cortisol metabolites to the observed discordance of immunoassay and LC-MS results needs further investigation.

Conclusion

These findings suggest that some immunoassay platforms (Roche, Abbott) may be most suitable for monitoring of serum cortisol for patients on 11b-hydroxylase inhibitor treatment, while Siemens may be least. Knowledge of immunoassay method used by the laboratory is therefore important for safe monitoring.