To evaluate the ability of post-transsphenoidal pituitary surgery (TSS) serum cortisol levels (s-cortisol) to predict surgical remission and recurrence of Cushing's disease (CD).


One hundred and three patients with CD from a tertiary referral centre were prospectively analysed over 6·0 ± 4·8 years of follow-up. Twenty patients received perioperative glucocorticoids as routine care and had s-cortisol measured 10–12 days after TSS (Protocol I). Eighty-six patients (91 surgeries) had s-cortisol measured at 6, 12, 18, 24, 48 h, and 10–12 days after TSS, and received glucocorticoids only in case of adrenal insufficiency (Protocol II).

Main outcomes

Remission [clinical signs and symptoms of adrenal insufficiency (or hypocortisolism) plus cortisol <3 μg/dl on the 1-mg overnight test (OT) and/or normal free urinary cortisol] during follow-up. Recurrence was defined as loss of remission criteria at least 1 year after TSS.


The remission rate after first TSS was 80%; 8% had recurrence. An s-cortisol nadir ≤3·5 μg/dl within 48 h after TSS had sensitivity of 73%, specificity and positive predictive value (PPV) of 100% and negative predictive value (NPV) of 60% and an s-cortisol nadir ≤5·7 μg/dl within 10–12 days of TSS had specificity and PPV of 100% and sensitivity of 91% NPV of 78% for CD remission.


At hospital discharge, the s-cortisol nadir within 48 h after TSS was already able to predict surgical remission for some patients, and the s-cortisol nadir within 10–12 days of TSS was able to predict cohort-wide surgical remission.