Highlight
In this multi-institutional retrospective cohort of 1,954 patients with adrenal incidentaloma and mild autonomous cortisol secretion (MACS), adrenalectomy was associated with a markedly lower hazard of all-cause mortality compared with nonoperative management after propensity score matching.
Adrenalectomy was also associated with a lower risk of new-onset diabetes, whereas differences in ischemic heart disease, cerebral infarction, fracture, and hypertension were not statistically significant.
The survival finding was less robust in sensitivity analysis excluding patients with adrenocorticotropic hormone (ACTH) levels greater than 20 pg/mL, raising concern that residual confounding and phenotype heterogeneity may influence the primary result.
The study strengthens the argument that some patients with MACS derive metabolic benefit from surgery, but it does not fully resolve who should undergo adrenalectomy.
Background
Adrenal incidentalomas are increasingly encountered in routine imaging, particularly in older adults. A substantial proportion are hormonally inactive, but mild autonomous cortisol secretion is now recognized as the most common endocrine abnormality among these lesions. MACS generally refers to biochemical evidence of autonomous cortisol production without the overt physical stigmata of Cushing syndrome. Clinically, this matters because even modest chronic cortisol excess has been associated with hypertension, dysglycemia, type 2 diabetes, obesity, cardiovascular disease, osteoporosis, and excess mortality.
The challenge in practice is that MACS occupies a biologic and therapeutic gray zone. Patients often have nonspecific metabolic comorbidities that are common in the general population, making it difficult to determine whether cortisol excess is causative, contributory, or incidental. The 1-mg dexamethasone suppression test (DST) remains central to diagnosis, but the risk conveyed by post-DST cortisol values between 1.8 and 5 micrograms/dL is heterogeneous. Current management therefore relies on individualized assessment rather than a universal surgical rule.
Adrenalectomy can remove the source of autonomous cortisol secretion, but surgery carries perioperative risks and may not reverse long-standing vascular or metabolic disease. Existing studies have suggested possible improvement in blood pressure, glycemic control, and quality of life after surgery, yet robust mortality data have been limited. This new study by Swaminathan and colleagues addresses that gap by comparing clinical outcomes in patients with MACS treated with adrenalectomy versus nonoperative management across multiple institutions.
Study Design
Design and data source
This was a retrospective cohort study using a multi-institutional database spanning 2011 through 2024. The investigators identified patients with adrenal incidentalomas and MACS, defined by cortisol levels between 1.8 and 5 micrograms/dL after a 1-mg DST.
Population
A total of 1,954 patients met inclusion criteria. Of these, 296 patients, or 15.1%, underwent adrenalectomy, while 1,658 patients, or 84.9%, were managed nonoperatively. After 1:1 propensity score matching based on demographic variables and comorbidities, 291 patients remained in each group. The matched cohort was 64.9% female and 71.1% White, with a mean age of 62 plus or minus 13 years.
Exposure and comparator
The intervention was adrenalectomy. The comparator was nonoperative management, presumably consisting of surveillance and medical management of associated comorbidities, although the abstract does not provide granular detail regarding surveillance protocols, endocrine follow-up, or medical therapy intensity.
Endpoints and statistical approach
The primary outcome was mortality. Secondary metabolic and clinical outcomes included new-onset diabetes, cerebral infarction, ischemic heart disease, fractures, and hypertension. Cox proportional hazards models were used to estimate hazard ratios with 95% confidence intervals. Propensity score matching was performed to improve baseline comparability between groups.
Mean follow-up was 1,367 plus or minus 1,185 days in the adrenalectomy group and 1,315 plus or minus 1,227 days in the nonoperative group, corresponding to approximately 3.7 and 3.6 years, respectively.
Key Findings
Mortality
In the propensity-matched analysis, adrenalectomy was associated with a substantially lower hazard of mortality compared with nonoperative management, with a hazard ratio of 0.34 and a 95% confidence interval of 0.27 to 0.36. On face value, this suggests an approximately 66% relative reduction in mortality risk. That is a striking effect size for an intervention in a population with biochemical cortisol excess that is considered mild.
Such a large association is clinically attention-grabbing, but it also warrants careful scrutiny. Observational studies can overestimate benefit when treatment selection is influenced by factors not fully captured in databases, such as frailty, lesion characteristics, patient preference, access to specialty care, surgeon selection, and intensity of longitudinal follow-up. Even after propensity matching, these factors may remain unevenly distributed.
New-onset diabetes
Adrenalectomy was also associated with a lower hazard of new-onset diabetes, with a hazard ratio of 0.49 and a 95% confidence interval of 0.25 to 0.97. This finding is biologically plausible. Chronic cortisol excess promotes insulin resistance, hepatic gluconeogenesis, visceral adiposity, and beta-cell stress. Removing the cortisol source could therefore reduce progression to overt diabetes, especially in patients at an early stage of metabolic dysfunction.
From a clinical standpoint, the diabetes result may be one of the most actionable findings in the paper. Whereas mortality is vulnerable to multiple competing explanations in a retrospective design, incident diabetes is more directly connected to the known physiology of glucocorticoid excess. Even so, the confidence interval was relatively wide and approached the null, so the magnitude of benefit should be interpreted conservatively.
Cardiovascular and skeletal outcomes
No statistically significant differences were observed for several other outcomes:
Cerebral infarction: hazard ratio 0.58; 95% confidence interval 0.24 to 1.43.
Ischemic heart disease: hazard ratio 0.83; 95% confidence interval 0.54 to 1.28.
Fractures: hazard ratio 0.72; 95% confidence interval 0.37 to 1.41.
Hypertension: hazard ratio 1.23; 95% confidence interval 0.79 to 1.93.
These neutral findings are informative. If adrenalectomy truly confers broad systemic benefit in MACS, one might expect at least directional improvement in cardiovascular events and blood pressure over time. Several explanations could account for the absence of significant differences. First, follow-up may have been too short to detect divergence in vascular and skeletal outcomes. Second, these events may require larger sample sizes because of lower incidence. Third, some comorbidities may reflect cumulative damage that is only partially reversible even after cortisol normalization. Fourth, MACS itself may not be the dominant driver of these outcomes in every patient.
Sensitivity analysis
Perhaps the most important interpretive detail in the abstract is the sensitivity analysis excluding patients with ACTH greater than 20 pg/mL. In this analysis, adrenalectomy was no longer associated with a survival benefit. This materially tempers confidence in the headline mortality result.
Why does this matter? ACTH values may help refine the likelihood that cortisol secretion is truly autonomous and clinically relevant. If the survival advantage disappears after excluding patients whose hormonal profile may be less consistent with adrenal autonomy, then the original mortality association may partly reflect misclassification, biologic heterogeneity, or residual confounding. It also suggests that the subgroup most likely to benefit from surgery may not be fully captured by the broad DST-defined MACS category alone.
Clinical Interpretation
This study supports the view that MACS is not always benign and that surgery may offer meaningful benefit in selected patients. The reduction in incident diabetes is particularly compelling because it aligns with established cortisol biology and with prior smaller studies showing postoperative metabolic improvement. For endocrine surgeons and endocrinologists, the data reinforce the need to evaluate surgical candidacy not solely on tumor size or imaging appearance, but also on biochemical phenotype and comorbidity burden.
At the same time, the results should not be interpreted as a mandate for routine adrenalectomy in all patients with post-DST cortisol between 1.8 and 5 micrograms/dL. The nonrandomized design leaves room for treatment selection bias. Patients selected for surgery may have been healthier, more functionally independent, treated at higher-resource centers, or more closely monitored than those managed conservatively. Conversely, nonoperative patients may have had more competing illnesses or shorter life expectancy, which would inflate observed mortality differences independent of adrenalectomy itself.
The null findings for hypertension and cardiovascular outcomes also remind clinicians that MACS is not a uniform disease state. Some patients may have mild biochemical abnormalities without substantial end-organ impact, while others have a clinically important cortisol-driven phenotype. The practical challenge is identifying the latter group.
How This Fits With Current Guidelines and Prior Literature
Recent European Society of Endocrinology guidance emphasizes that MACS should be assessed in the context of cortisol-related comorbidities rather than as an isolated laboratory label. Surgery may be considered in patients with relevant comorbid conditions and biochemical evidence of autonomous cortisol secretion, particularly when those comorbidities are difficult to control. This study is broadly consistent with that individualized framework rather than a one-size-fits-all surgical strategy.
Earlier observational studies and systematic reviews have suggested that adrenalectomy can improve blood pressure, glucose metabolism, body weight, and quality of life in some patients with MACS. However, most prior reports were limited by small sample size, single-center design, and variable diagnostic thresholds. The present study adds scale and multicenter breadth, but still does not overcome the central limitation of observational treatment comparisons.
In clinical decision-making, the paper may modestly lower the threshold to discuss surgery when a patient with MACS has worsening glycemia or early diabetes, particularly if other features suggest meaningful cortisol autonomy. It should not, however, be read as definitive proof of a mortality benefit.
Strengths and Limitations
Strengths
The study has several notable strengths. It includes a relatively large MACS cohort assembled across multiple institutions over more than a decade. The use of propensity score matching is appropriate and improves comparability between treatment groups. The investigators also evaluated clinically important hard outcomes, including mortality and incident diabetes, rather than relying only on laboratory endpoints.
Limitations
The limitations are substantial and central to interpretation. First, the retrospective design precludes causal inference. Second, the abstract does not describe key factors that may influence both treatment choice and outcome, such as tumor size, laterality, imaging phenotype, longitudinal cortisol metrics, body mass index, smoking, frailty, medication burden, socioeconomic status, or reason for choosing surgery. Third, follow-up duration, while adequate for some metabolic outcomes, may still be insufficient for cardiovascular and fracture endpoints.
Fourth, endpoint ascertainment in database studies can be affected by coding practices and variable follow-up intensity. Fifth, the very narrow confidence interval around the mortality hazard ratio is unusual and may reflect the statistical properties of the dataset, but it also underscores the need to inspect the full paper for event counts, censoring patterns, proportional hazards assumptions, and model specification. Sixth, the sensitivity analysis meaningfully weakens confidence in the survival claim.
Finally, generalizability may be incomplete. The cohort was predominantly White and female, and the findings may not translate evenly across populations with different comorbidity profiles, healthcare access, or surgical practice patterns.
Implications for Practice
For clinicians, the study supports a more proactive discussion of surgery in selected patients with MACS, especially those with dysglycemia or high metabolic risk. In multidisciplinary adrenal boards or endocrine surgery clinics, the following questions remain practical:
Is cortisol autonomy convincingly established beyond a single DST result?
Does the patient have cortisol-related comorbidity, particularly diabetes or worsening glucose intolerance, that is likely to improve with intervention?
Is operative risk acceptably low, and is the patient fit for surgery?
Are there alternative explanations for the patient’s metabolic profile that make surgical benefit less likely?
Would continued surveillance and optimized medical management be a reasonable alternative?
In other words, the paper advances the conversation from whether MACS matters to which MACS patients matter most.
Future Research
The field still needs prospective studies with standardized endocrine phenotyping and predefined outcome assessment. Ideally, randomized or carefully emulated target-trial designs should compare adrenalectomy with structured medical management. Future work should also determine whether ACTH suppression, repeated DST results, late-night salivary cortisol, 24-hour urinary free cortisol, radiographic characteristics, or molecular markers can identify a subgroup with the highest likelihood of benefit. Patient-reported outcomes, bone health, antihypertensive medication burden, and diabetes remission metrics would also be valuable.
Equally important is defining when surgery is unlikely to help. Avoiding unnecessary adrenalectomy is as clinically meaningful as selecting the right patient for operative treatment.
Conclusion
This multi-institutional retrospective study reports that adrenalectomy for MACS, defined by post-DST cortisol of 1.8 to 5 micrograms/dL, was associated with lower all-cause mortality and a lower risk of new-onset diabetes compared with nonoperative management after propensity matching. The diabetes finding is biologically credible and clinically relevant. The mortality signal, although striking, is weakened by sensitivity analysis and remains vulnerable to residual confounding.
The most balanced takeaway is that adrenalectomy may benefit selected patients with MACS, particularly those with metabolically meaningful cortisol excess, but current evidence still supports individualized decision-making rather than routine surgery for all. This study moves the field forward, while also making clear that better phenotyping and stronger comparative evidence are urgently needed.
Funding and ClinicalTrials.gov
No funding information or ClinicalTrials.gov registration number is provided in the abstract.
References
Swaminathan N, Song Z, Abraham P, Gillis A, Fazendin J, Chen H, Lindeman B. Mortality and clinical outcomes in patients with mild autonomous cortisol secretion: Adrenalectomy versus nonoperative management. Surgery. 2026-05-07:110238. PMID: 42103509.
Fassnacht M, Dekkers OM, Else T, et al. European Society of Endocrinology clinical practice guidelines on the management of adrenal incidentalomas, in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol. 2023;189(1):G1-G46.
Di Dalmazi G, Pasquali R, Beuschlein F, Reincke M. Subclinical hypercortisolism: a state, a syndrome, or a disease? Eur J Endocrinol. 2015;173(4):M61-M71.
Bancos I, Alahdab F, Crowley RK, et al. Therapy of endocrine disease: Improvement of cardiovascular risk factors after adrenalectomy in patients with adrenal tumors and subclinical Cushing syndrome: a systematic review and meta-analysis. Eur J Endocrinol. 2016;175(6):R283-R295.
