Adrenal incidentalomas are neoplasms detected by imaging techniques originally intended for other purposes, and their prevalence has increased due to the popularity of abdominal imaging (1). Mild autonomous cortisol secretion (MACS) is the most common hormone abnormality in patients with adrenal incidentalomas (2). MACS refers to a pathological state of hypercortisolism determined by biochemical evidence without typical signs or symptoms of Cushing’s syndrome (CS). Despite the absence of conspicuous manifestations, patients with MACS have been reported to suffer from a higher prevalence of metabolic complications and an increased risk for vertebral fractures, cardiovascular events and mortality (3, 4, 5, 6). Therefore, accurately identifying patients with MACS and managing their condition can be beneficial in reversing several metabolic effects and potentially improving their health quality (7).

According to the practical guidelines of the European Society of Endocrinology (ESE) and the European Network for the Study of Adrenal Tumors (ENSAT), a 1 mg overnight dexamethasone suppression test (DST) should be performed in every patient with adrenal incidentaloma to exclude cortisol excess (1). However, there is still a lack of standardized guidelines for identifying and managing MACS, which can lead to potential misclassification and improper management of (8, 9). Additionally, in clinical practice, we have observed that many patients with adrenal incidentalomas do not undergo DST to confirm or exclude the presence of MACS.

Nowadays, steroid profile evaluation using mass spectrometry (MS) techniques is increasingly being conducted in patients with adrenal masses or in need of screening for adrenal diseases (10, 11). The advantages of MS-based steroid determination lie in its ability to simultaneously measure multiple analytes, offer high specificity, and provide similar sensitivity to the best immunoassay methods (12, 13). Compared to DST, steroid profile evaluation does not require dexamethasone administration or an additional day for blood collection. Masjkur et al. (14) found that patients with MACS had lower levels of dehydroepiandrosterone-sulfate (DHEAS) and higher levels of 11-deoxycortisol and 11-deoxycorticosterone compared to those without MACS, highlighting the significance of steroid profiling in distinguishing MACS from adrenocortical adenomas.

However, interpreting complicated results remains challenging, thus requiring professional expertise. In this scenario, steroid metabolomics – a combination of steroid profiling and computational machine learning (ML) analysis of datasets - shows promise in identifying and managing adrenal diseases using heterogeneous steroids data (15). Many articles have reported the potential of this approach in diagnosing and differentiating several adrenal disorders including adrenocortical adenomas, adrenocortical carcinomas, CS, and primary aldosteronism (PA) (16, 17, 18). However, no publications have been retrieved regarding the identification of MACS based on the combination of steroid profiling and ML techniques. Therefore, this study aims to assess the diagnostic value of a 24-steroid panel alone or in combination with clinical characteristics for detecting MACS among those detected with adrenal incidentalomas using eXtreme Gradient Boosting (XGBoost).