FRI177 Cosyntropin Stimulation Testing Is More Selective Than Postoperative Day 1 Basal Cortisol for Diagnosing Secondary Adrenal Insufficiency Following Unilateral Adrenalectomy

Disclosure: S.H. Johnson: None. C. Zhang: None. P.T. Hangge: None. T.W. Yen: None. T. Shaik: None. K. Doffek: None. J.W. Findling: None. T.B. Carroll: None. D.B. Evans: None. S. Dream: None. T.S. Wang: None. Background: Secondary adrenal insufficiency (SAI) may occur in patients after unilateral adrenalectomy for adrenal-dependent hypercortisolism (HC) and/or primary aldosteronism (PA). Perioperative protocols include routine glucocorticoid replacement (GR) or selective GR based on postoperative day 1 (POD1) basal cortisol levels or cosyntropin stimulation testing (CST). This study aimed to assess if a POD1 basal cortisol level alone was predictive of an abnormal CST and the need for GR in patients following unilateral adrenalectomy for HC and/or PA. Methods: This is a single institution retrospective review of consecutive unilateral adrenalectomies performed for HC and/or PA from 9/2014-10/2022. Patients were grouped as overt hypercortisolism (OH;n=42), mild autonomous cortisol excretion (MACE;n=70), mixed PA/HC (n=22) or PA (n=73); 22 patients with PA did not have preoperative evaluation for HC. Preoperative morning cortisol after low-dose dexamethasone suppression testing (DST), was used to define MACE (1.8-4.9 µg/dL) and OH (≥5.0). POD1 CST was performed for all patients with HC and prior to 2021, all patients with PA. SAI was defined as basal cortisol ≤5 or stimulated ≤14 (≤18 before 4/2017). Receiver operating characteristic (ROC) curves and sensitivity analyses were performed to assess the sensitivity (SN) and specificity (SP) of cutoffs within the range of observed basal cortisol levels in predicting an abnormal CST and need for GR. Results: Of 207 patients, 152 had POD1 CST (93% OH, 96% MACE, 73% PA/HC, 41% PA). Of these, 80 (53%) had SAI (67% OH, 55% MACE, 44% PA/HC, and 33% PA). Of the 10 patients with PA and SAI, 5 did not have preoperative testing for HC. Overall, 81 patients were discharged on GR. Median (interquartile range [IQR]) duration of GR was 246 (127,557), 55 (38,207), 74 (16,185), and 26 (15,36) days in patients with OH, MACE, PA/HC, and PA, respectively. At a median follow-up of 142 days (IQR 42,617), 12 patients (5 OH, 6 MACE, 1 PA/HC) remained on GR. The SN/SP of a basal cortisol ≤10 in predicting GR was 92%/77% in patients with OH and 94%/73% of patients with MACE. Compared to initiation of GR based on basal cortisol ≤10, CST identified 4 patients with SAI in the combined OH and MACE cohorts who would not have received GR, and 11 in the combined cohort who did not require GR. For patients with PA and PA/HC, a basal cortisol ≤10 had SN of 100% and SP of 85% and 67%, respectively. The optimal basal cortisol for predicting SAI in patients with PA/HC was ≤5 (SN,SP 100%). Area under the curve for all ROC curves was >0.9. Conclusion: These results demonstrate that in patients who underwent unilateral adrenalectomy for HC, PA, or mixed PA/HC, POD1 CST improved identification of patients at risk for SAI compared to basal cortisol levels alone. While basal cortisol levels alone may be considered in patients with PA or mixed HC/PA, all other patients with HC should have POD1 CST to determine need for postoperative GR. Presentation: Friday, June 16, 2023