Lowering the day 5 postoperative cortisol cutoff value to avoid over-diagnosis of secondary adrenal insufficiency following pituitary surgery

Daniella Bae; Emmanuel Lawal; Sirazum Choudhury; Kavita Narula; Katherine Lazarus; Tricia Tan; Deborah Papadopoulou; Niamh Martin; Karim Meeran

doi:10.1530/endoabs.109.OC7.1

OC7.1

Prev Next

Section Contents Cite

Endocrine Abstracts (2025) 109 OC7.1 | DOI: 10.1530/endoabs.109.OC7.1

SFEBES2025 Oral Communications Neuroendocrinology and Pituitary (6 abstracts)

Lowering the day 5 postoperative cortisol cutoff value to avoid over-diagnosis of secondary adrenal insufficiency following pituitary surgery

Daniella Bae , Emmanuel Lawal , Sirazum Choudhury , Kavita Narula , Katherine Lazarus , Tricia Tan , Deborah Papadopoulou , Niamh Martin & Karim Meeran

Author affiliations

Imperial College London, London, United Kingdom

Background: Identifying patients who develop secondary adrenal insufficiency (SAI), requiring longterm glucocorticoid replacement (GR) following trans-sphenoidal pituitary surgery (TSS) is crucial to prevent life-threatening Addisonian crisis. An early morning day 5 cortisol value may reliably assess hypothalamo-pituitary-adrenal (HPA) axis functioning. However, the cut-off which can reliably predict SAI is yet to be elucidated.

Method: This is a retrospective cohort study of 88 consecutive patients who underwent TSS at Charing Cross Hospital between 2019 and 2023. Receiver Operating Characteristic (ROC) curves were generated to evaluate the different predictive values of cortisol levels on postoperative day 5, using the Abbott Alinity cortisol immunoassay. We excluded 34 patients leaving 54 patients for analysis. Area Under the curve (AUC) was calculated from the ROC curve.

Results: SAI prevalence within this study was 31%. The currently used cortisol cutoff concentration of >358 nmol/l to exclude SAI demonstrated 100% sensitivity and a specificity of 53.85%. Lowering this threshold to >308 nmol/l maintained 100% sensitivity while improving specificity to 75.86%. A Day 5 cortisol <206 nmol/l predicts poor HPA function with a sensitivity of 70.59% and 100% specificity. Day 5 cortisol (AUC=0.9645) and Day 4-7 (AUC=0.8679) cortisol provided better predictive values for SAI compared to Day 2/3 samples (AUC=0.8095).

Conclusion: Day 5 serum cortisol >308 nmol/l reliably indicates an intact HPA axis. Values <206 nmol/l reliably detects SAI. Patients with Day 5 serum cortisol measurements between 206-308 nmol/l should be discharged on glucocorticoid therapy and undergo a dynamic test post-operatively to ascertain their true HPA function.