Perioperative Risks and Long-Term Outcomes of Bilateral Adrenalectomy: A Multicenter Retrospective Cohort Study...

Background and Objectives: Bilateral adrenalectomy (BADx) is a definitive treatment for diverse adrenal disorders, though data across etiologies remain limited. This multicenter study evaluated perioperative risks, long-term survival, and shifts in comorbidities in patients undergoing BADx for Cushing's syndrome (CS) and non-CS indications. Materials and Methods: We retrospectively analyzed 103 adult patients (CS, n = 52; non-CS, n = 51) from 9 tertiary centers. Primary outcomes included early (≤30 days) and late mortality, complications (adrenal crisis, Nelson's syndrome), and changes in comorbidity profiles. Results: The cohort (mean age 50.59 ± 13.82 years) had a median follow-up of 96 months. Early mortality occurred exclusively in CS patients (5.8%), while late mortality was comparable between groups. Overall 5-year survival rates were 83.3% for CS and 89.9% for non-CS (p = 0.360), with no significant differences across CS etiologies (p = 0.760). Adrenal crisis was significantly more frequent in CS (23% vs. 7%, p = 0.023). Nelson's syndrome developed in 28% of Cushing's disease patients. Postoperatively, both cohorts showed increased reliance on psychiatric and cardiovascular medications. Notably, osteoporosis rates were significantly higher than preoperative levels exclusively in the non-CS group (p = 0.006), highlighting the impact of long-term glucocorticoid replacement in this cohort. Two patients with congenital adrenal hyperplasia demonstrated remarkable long-term survival (8 and 30 years) across different surgical timings. Conclusions: BADx delivers rapid systemic remission and definitive control across a range of adrenal pathologies. However, the primary cause and the timing of surgery significantly impact the clinical course and the risk of complications. The resulting state of irreversible adrenal insufficiency requires comprehensive, lifelong surveillance and tailored hormone replacement.