Remission rate after transsphenoidal surgery in patients with pathologically confirmed Cushing's disease, the role of cortisol, ACTH assessment and immediate reoperation: a large single center experience

Abstract

Postoperative serum cortisol is used as an indicator of Cushing's disease (CD) remission following transsphenoidal surgery (TSS) and guides (controversially) the need for immediate adjuvant treatment for CD. We investigated postoperative cortisol and adrenocorticotropic hormone (ACTH) levels as predictors of remission/recurrence in CD in a large retrospective cohort of patients with pathologically confirmed CD, over 6 years at a single institution. Midnight and morning cortisol, and ACTH at 24-48 h postoperatively (>24 h after last hydrocortisone dose) were measured. Remission was defined as normal 24-h urine free cortisol, normal midnight salivary cortisol, a normal dexamethasone-corticotropin releasing hormone (CRH) test or continued need for hydrocortisone, assessed periodically. Statistical analysis was performed using PASW 18. Follow up data was available for 52 patients (38 females and 14 males), median follow up was 16.5 month (range 2-143 months), median age was 45 years (range 21-72 years), 28 tumors were microadenomas and 16 were macroadenomas, and in eight cases no tumor was observed on magnetic resonance imaging. No patient with postoperative cortisol levels >10 mcg/dl were found to be in remission. Ten of the 52 patients with cortisol >10 mcg/dl by postoperative day 1-2 underwent a second TSS within 7 days. Forty-three patients (82.7%) achieved CD remission (36 after one TSS and 7 after a second early TSS) and six patients suffered disease recurrence (mean 39.2 ± 52.4 months). An immediate second TSS induced additional hormonal deficiencies (diabetes insipidus) in three patients with no surgical complications. Persistent disease was noted in nine patients despite three patients having an immediate second TSS. Positive predictive value for remission of cortisol <2 mcg/dl and ACTH <5 pg/ml was 100%. Cortisol and ACTH levels (at all postoperative time points and at 2 months) were correlated (r = 0.37, P < 0.001). Nadir serum cortisol of ≤2 mcg/dl and ACTH <5 pg/ml predicted remission (P < 0.005), but no level predicted lack of recurrence. Immediate postoperative ACTH/cortisol did not predict length of remission. No patients with postoperative cortisol >10 mcg/dl were observed to have delayed remission; all required additional treatment. There was no significant difference in age, body mass index, tumor size and length of follow-up between postoperative cortisol groups: cortisol ≤2 mcg/dl, cortisol >5 mcg/dl and cortisol >10 mcg/dl. Immediate postoperative cortisol levels should routinely be obtained in CD patients post TSS, until better tools to identify early remission are available. Immediate repeat TSS could be beneficial in patients with cortisol >10 mcg/dl and positive CD pathology: our combined (micro- and macroadenomas) remission rate with this approach was 82.7%. ACTH measurements correlate well with cortisol. However, because no single cortisol or ACTH cutoff value excludes all recurrences, patients require long-term clinical and biochemical follow-up. Further research is needed in this area.