Confirmation of Feasibility of Selective Glucocorticoid Replacement Following Unilateral Adrenalectomy for Hypercortisolism and Primary Aldosteronism

Background: Secondary adrenal insufficiency (AI) can develop following unilateral adrenalectomy (UA) foradrenal-dependent hypercortisolism (HC) and has been reported after UA for primaryaldosteronism (PA). An institutional study previously demonstrated that cosyntropin stimulationtesting on postoperative day 1 (POD1-CST) successfully identified patients who requiredglucocorticoid replacement (GR) following UA; 50% of HC patients required GR and no PApatients required GR. The aim of this study was to reevaluate the need for GR following UA forpatients with HC and PA in a larger cohort of patients. Methods We reviewed 108 patients from a prospectively maintained adrenal database who underwent UAfor HC (n=74), PA (n=22), and concurrent HC/PA (n=12) from 9/2014-10/2020. PA patientswithout preoperative evaluation for HC were excluded. Patients with 1mg dexamethasonesuppression test (DST) cortisol >1.8 (µg/dL) were defined as having mild HC, with ³5 defined asovert Cushing’s Syndrome (CS). All patients underwent our institutional POD1-CST protocoland GR was initiated for patients with basal cortisol £5 or stimulated cortisol £14 (<18 prior to4/2017). Results: Overall, 51 (47%) patients had an abnormal POD1-CST and were discharged on GR (44 HC, 1PA, and 6 HC/PA). Two (2%) patients with CS had a normal POD1-CST but developed AIrequiring GR at 8 and 12 weeks post UA. Of the 74 patients with HC, 44 (59%) had an abnormalPOD1-CST and were discharged on GR, including 19/28 (68%) with CS and 25/46 (54%) withmild HC. Preoperative DST cortisol was higher in HC patients who required GR compared topatients with a normal POD1-CST (4.1 vs 3.6; p=0.007). Median cortisol levels for HC patientswith an abnormal POD1-CST vs those with a normal test were: basal: 3.8 vs 15.6 (p=0.027); 30-minute: 10.1 vs 20.1 (p=0.403); and 60-minute 11.4 vs 22.2 (p=0.260). Of the 22 PA patients, 19(86%) had a normal POD1-CST. Median cortisol levels for PA patients with an abnormal POD1-CST vs those with a normal test were: basal: 0.4 vs 12.1; 30-minute: 8.8 vs 24.6; and 60-minute:12.2 vs 28.9. Of the 3 (14%) PA patients with an abnormal POD1-CST, 1 was dischargedwith GR and began tapering after 2 weeks; the other 2 did not require GR and did not developAI. Of the 12 patients with combined PA/HC, 6 (50%) were discharged on GR based on POD1-CST. GR was required by 30 (59%) patients for <3 months and 82% for <12 months; 7/9 whorequired GR >12 months had CS. Conclusions: Using a standard protocol for POD1-CST in patients who underwent unilateral adrenalectomyfor HC, PA, or combined PA/HC, this study demonstrated that routine GR is not required in 32%of patients with CS and 46% of patients with mild HC. POD1-CST safely identifies patients whowill require GR with no immediate concern for adrenal insufficiency. These data also suggestthat routine evaluation for AI in postoperative PA patients is not needed if cortisol excess hasbeen excluded preoperatively.