OBJECTIVE - To determine whether a random postoperative day-3 cortisol value of 10 μg/dL or greater is predictive of adrenal sufficiency 3 to 10 weeks after transsphenoidal surgery (TSS) and during long-term clinical follow-up.
METHODS - We retrospectively reviewed the case records of patients who underwent TSS at our institution between 1991 and 2008. Inclusion criteria were as follows: random cortisol measured on the morning of postoperative day 3, adrenal dynamic testing performed 3 to 10 weeks after TSS, and clinical assessment of the hypothalamic-pituitary-adrenal (HPA) axis at least 6 months after TSS.
RESULTS - A total of 466 patients underwent TSS at our institution during the study period. Eighty-three patients met study inclusion criteria. Sensitivity of a random postoperative day-3 serum cortisol value of 10 μg/dL or greater for the prediction of adrenal sufficiency at a median follow-up of 42 days was 64.81% (95% confidence interval, 50.6%-77.32%), with an odds ratio of 3.1 (95% confidence interval, 1.08-8.58). Specificity was 62.1% (95% confidence interval, 42.3%-79.3%). At a median follow-up of 500 days, only 2 patients with a postoperative day-3 cortisol value of 10 μg/dL or greater required hydrocortisone replacement, both of whom had multiple anterior pituitary hormone deficiencies and evidence of pituitary dysfunction during the perioperative period.
CONCLUSIONS - In the appropriate clinical context, a postoperative day-3 cortisol value of 10 μg/dL or greater accurately predicts the integrity of the HPA axis. The final decision regarding corticosteroid replacement should be personalized, considering the postoperative day-3 cortisol level, the clinical context in which the measurement was obtained, and any evidence of concomitant pituitary dysfunction in the perioperative period.