Use of day 1 early morning cortisol to predict the need for glucocorticoid replacement after pituitary surgery

Background. Assessment of adrenal reserve in patients who have undergone pituitary surgery is crucial. However, there is no clear consensus with regards to the type and timing of the test that should be used in the immediate post-operative period. Recently, there has been increased interest in measuring post-operative cortisol levels. We present our data utilising day 1 post-operative early morning cortisol as a tool to assess adrenal reserve in steroid-naive patients. Methods. A retrospective analysis of endoscopic pituitary surgery undertaken over a 2-year period. 82 patients underwent 84 surgeries in total. Patients who were already on glucocorticoids pre-operatively and patients with Cushing's disease, pituitary apoplexy and those without follow-up data were excluded, leaving a study group of 44 patients with 45 operations. A 9am day 1 post-operative cortisol value of > 400 nmol/L was taken as an indicator of adequate adrenal reserve. All the patients were reassessed at 6 weeks with a standard short synacthen test (SST) using 250 micrograms of intravenous synacthen. Results. 22 out of 45 patients had a cortisol value of > 400 nmol/L on day 1 post-operatively and were discharged without glucocorticoid supplementation. Of these, only 2 patients subsequently failed the SST when reassessed at 6-8 weeks. The remaining 23 patients had a cortisol value of < 400 nmol/L on day 1 post-operatively and were discharged on hydrocortisone 10 mg twice daily. At 6-8 weeks, nine continued to show suboptimal stimulated cortisol levels whereas the remaining fourteen patients showed adequate adrenal reserve. The 9 am cortisol value had high specificity (81.8%) and positive predictive value (90.9%) for integrity of the HPA axis. Sensitivity was 58.8% and negative predictive value was 39.1%. Conclusion. A day 1 post-operative early morning cortisol is a useful tool to predict adrenal reserve post-pituitary surgery, enabling clinicians to avoid unnecessary blanket glucocorticoid replacement.