ELSEVIER
Contents lists available at ScienceDirect
Journal of Steroid Biochemistry and Molecular Biology
journal homepage: www.elsevier.com/locate/jsbmb
-
—
The Journal of Steroid Biochemistry & Molecular Biology
Viapy
Serum steroid profiling by LC-MS/MS in distinguishing adrenocortical carcinoma from other indeterminate adrenal masses
Check for updates
Archana Rao ª,1, Aditya Phadtea,1, Anuj Bana, Saba Samad Memona, Manjiri Karlekar ª, Anurag Ranjan Lilaa,* (D, Vijaya Sarathib, Nimmi Kansal , Rohit Barnabasa, Padma Vikram Badhe ª, Gwendolyn Fernandes e, Sameer Regef, Gagan Prakash &, Santosh Menonh, Nalini Shahª, Tushar Bandgara
a Department of Endocrinology and Metabolism, King Edward Memorial Hospital and Seth Gordhandas Sunderdas Medical College, Mumbai, India
b Department of Endocrinology, Vydehi Institute of Medical Sciences and Research Centre, Bangalore, India
” Dr Lal PathLabs, Clinical Chemistry & Biochemical Genetic, Delhi, India
d Department of Radiology, King Edward Memorial Hospital and Seth Gordhandas Sunderdas Medical College, Mumbai, India
e Department of Pathology, King Edward Memorial Hospital and Seth Gordhandas Sunderdas Medical College, Mumbai, India Department of General Surgery, King Edward Memorial Hospital and Seth Gordhandas Sunderdas Medical College, Mumbai, India
8 Department of Uro-Oncology, Tata Memorial Hospital and Advanced Centre for Treatment Research and Education in Cancer (ACTREC), Homi Bhabha National Institute (HBNI), Mumbai, India
h Department of Pathology, Tata Memorial Hospital and Advanced Centre for Treatment Research and Education in Cancer (ACTREC), Homi Bhabha National Institute (HBNI), Mumbai, India
ARTICLE INFO
Keywords: Adrenocortical Carcinoma LC-MS/MS Steroid profiling
ABSTRACT
For an adrenal incidentaloma with indeterminate imaging characteristics, urine multisteroid profiling is sug- gested for diagnosing adrenocortical carcinoma (ACC). Data on the utility of serum steroid metabolomics in this context is limited to a few studies. Here, we present data of 62 adult patients with indeterminate unilateral adrenal masses (size ≥ 3 cm and basal attenuation >10HU) where baseline serum liquid chromatography- tandem mass spectrometry (LC-MS/MS) multisteroid profiling was available. Logistic regression was used to identify the key steroid signature for differentiating ACC from other non-ACC adrenal masses. Among 62 patients (median age: 41 years, 31 males), 37 (59.6 %) had ACC. The non-ACC cohort (n = 25) comprised pheochro- mocytoma (n = 9), adrenocortical adenoma (n = 8), metastases (n = 4), schwannoma (n = 2), ganglioneuroma (n = 1), and lymphoma (n = 1). Tumour size was significantly larger in the ACC cohort (9.9 vs 7.0 cm; p < 0.001) than the non-ACC cohort. Nine of 13 steroids were significantly elevated in ACC: 11-deoxycorticosterone (DOC), 17-hydroxyprogesterone (17OHP), 11-deoxycortisol (S), cortisone (E), androstenedione (A4), dehydroepian- drosterone (DHEA), and dehydroepiandrosterone sulphate (DHEAS) in both sexes, as well as testosterone (T) in females and progesterone (P4) in males. After excluding sex-dependent steroids, univariate analysis yielded six significant steroids (17OHP, S, E, A4, DHEA, and DHEAS). A multivariate logistic regression model with back- ward elimination identified A4, S, and DHEAS as the best discriminators (AUC:0.923), with a cutoff of 0.52 yielding 83.8 % sensitivity and 96 % specificity for diagnosing ACC. Our study results suggest serum LC-MS/MS profiling of three steroids (A4, S, and DHEAS) provides a non-invasive approach to distinguish ACC from other indeterminate adrenal masses.
1. Introduction
In the differential diagnosis of a unilateral adrenal mass, specific etiologies such as myelolipoma, cyst, hemorrhage, and lipid-rich (basal attenuation <10 HU) adrenal adenoma can be accurately diagnosed
based on their characteristic computed tomography (CT) features [1-4]. Lipid-poor adrenal masses like lipid-poor adrenocortical adenoma (ACA), pheochromocytoma (PHEO), adrenocortical carcinoma (ACC), infiltrative pathologies (metastasis, lymphoma, infection, etc.), or other rare benign adrenal masses (schwannoma, ganglioneuroma, etc) have
* Corresponding author.
1 Archana Rao and Aditya Phadte contributed equally as first authors
https://doi.org/10.1016/j.jsbmb.2026.106937
overlapping CT characteristics [5]. The overall survival of patients with ACC is poor, but prognosis is better when diagnosed at an early stage; hence, its timely diagnosis is crucial [6,7]. Most ACCs are unilateral, lipid-poor, and large; only ~1 % and ~2-4 % of ACCs are < 2.0 and 2-4 cm, respectively [8,9]. Clinical and biochemical hormonal excess is apparent in ~50 % of ACC [6]. Mass spectrometry-based urinary mul- tisteroid profiling has revealed elevated steroid hormone precursors (with poor biological activity) and androgens in both endocrine-active or hormonally silent ACC tumors [10]. Urinary steroid metabolomics can distinguish between ACCs and other adrenal tumors and was also validated in a prospective cohort [8]. This non-invasive test has the advantage over adrenal biopsy, as histopathology may not unequivo- cally differentiate between ACC and ACA, and has a potential risk of biopsy needle track seeding [11]. Hence, the latest adrenal incidenta- loma guidelines recommend considering urine steroid profiling in pa- tients with indeterminate adrenal masses [12]. Guidelines and experts suggest that the measurement of adrenal steroid excess [dehydroepi- androsterone sulfate (DHEAS), 17-hydroxyprogesterone (17OHP), an- drostenedione (A4), testosterone (T), 17-ß estradiol, and 11-deoxycortisol (S)] by liquid chromatography tandem mass spec- trometry (LC-MS/MS) aids in the diagnosis and follow-up of ACC pa- tients. Nonetheless, they acknowledge that the most suitable set of steroid precursors and sex hormones is yet to be established.
Few recent studies have demonstrated that serum multisteroid profiling by LC-MS/MS can discriminate ACC from non-ACC adrenal tumors [13-16]. Notably, two of these studies, one from Germany and the other from the USA, had a sample size of ACC patients exceeding 30 [14,16]. These studies on serum steroid metabolomics have provided cutoff scores for both ruling in and ruling out ACC tumors. Advantages of serum samples over urine are ease of collection and clinical imple- mentability, as serum LC-MS/MS testing facilities are more readily available.
This study was conducted at a tertiary care referral centre to un- derstand the accuracy of serum multisteroid profiling by LC-MS/MS with application of logistic regression modeling in distinguishing ACCs from other unilateral, lipid-poor (basal attenuation >10 HU) and large (≥3 cm) indeterminate adrenal masses.
2. Materials and methods
This retrospective study was conducted at a tertiary-care centre, following approval from the Institutional Ethics Committee (EC/OA-47/ 2025). A waiver of informed consent was granted due to the retro- spective design. A review of clinical case records between January 2018 and April 2025 was carried out. Data from adult patients (aged ≥18 years) with a unilateral, lipid-poor (basal attenuation ≥10 HU) and large (≥3 cm) adrenal mass, for whom serum steroid profiling by LC-MS/MS assay was available at diagnosis, were included. Patients with an adrenal mass exhibiting distinct imaging characteristics, such as myelolipoma, cyst, hemorrhage, or a lipid-rich (basal attenuation <10 HU) adrenal adenoma, were excluded. To avoid analytical interference, patients receiving exogenous glucocorticoids within the preceding 3 months or drugs interfering with steroidogenesis pathway (mitotane, ketocona- zole, metyrapone, or mifepristone) were also excluded. Demographic, clinical, biochemical, radiological, and histological data were noted from the records. The final diagnosis of an adrenal mass was based on histopathology for ACC (modified Weiss score ≥3), ACA (modified Weiss score <3), PHEO, lymphoma, metastases, infection, ganglio- neuroma, and schwannoma [17]. ACCs were classified as low-grade or high grade based on mitotic activity, defined as ≤ 20 mitoses per 10 mm2 and > 20 mitoses per 10 mm2, respectively [18]. For a patient with a provisional diagnosis of ACA who has not undergone surgery, the final diagnosis of ACA was considered based on its radiologic stability over a minimum follow-up period of two years. Biochemical cortisol excess was defined as an overnight dexamethasone-suppressed (ONDS) serum cortisol > 50 nmol/L. Serum cortisol levels were assessed using a
solid-phase competitive chemiluminescent enzyme immunoassay (Siemens Healthcare). Plasma normetanephrines were measured with an LC-MS/MS assay, and age-specific 97.5 centile cutoffs were used to define elevation. Computed tomography (CT) imaging was performed using a 64-slice multidetector CT scanner (Brilliance 64, Philips Healthcare). The CT acquisition parameters and imaging protocol have been described in detail in our previous publication [5]. CT imaging parameters, such as the size (the largest dimension) and unenhanced phase attenuation values, were noted.
Blood samples for serum steroid profiling were collected between 8:00 and 10:00 AM. After centrifugation, serum was stored at 4℃ and shipped to a central laboratory weekly for batch analysis. In this study, serum steroid concentrations were measured using the commercially available Masschrom® Steroid LC-MS/MS (15 steroids), In-Vitro Di- agnostics (IVD)-certified, kit (Chromsystems, Gräfelfing, Germany) and Exion LC ultra high performance liquid chromatography coupled to QTRAP4500 triple quadrupole mass spectrometer (ABSciex, Framing- ham, USA). Method implementation on the LC-MS/MS system was carried out by the manufacturer’s application specialist, with all chro- matographic and MS/MS parameters set according to instrument- specific recommendations. Analyses were performed in accordance with the LC-MS/MS kit manufacturer’s instructions, using two distinct instrumental configurations and separate sets of calibrators and con- trols, constituting two analytical panels. The first panel was designed for measurement of aldosterone (A), corticosterone (B), cortisol (F), corti- sone (E), S, and 21-deoxycortisol (21-DF) while the second panel was used for quantification of A4, DHEA, DHEAS, 11-deoxycorticosterone (DOC), 17OHP, progesterone (P), and T. Both panels employed iden- tical sample preparation procedures and the same LC-MS/MS platform. Steroids were extracted from 500 uL of serum sample by solid-phase extraction using the manufacturer-provided extraction buffer and isotope-labelled internal standards. The samples were washed with buffer to eliminate unbound substances, eluted, and dried under nitro- gen at 50 ℃. The residues were then reconstituted with buffer, mixed thoroughly, and the supernatants were transferred into auto-sampler vials for LC-MS/MS analysis. For each analyte, the manufacturer sup- plied a corresponding isotopically labelled internal standard and a minimum of two multiple reaction monitoring (MRM) transitions for both the analyte and its internal standard (Supplementary Table 1). Steroid hormones were separated chromatographically and quantified using specific multiple reaction monitoring (MRM) transitions, yielding precise and reliable measurements. Optimization of mass transitions was performed using tuning mixtures (Tuning Mix® MassChrom® Steroid Panel 1, and Tuning Mix® MassChrom® Steroid Panel 2) provided by the manufacturer. Manufacture-provided 6plus1® Multilevel Serum calibrator set MassChrom® steroid panel 1, and 6plus1® Multilevel Serum calibrator set MassChrom® steroid panel 2 were used as cali- brators whereas MassCheck® Steroid Panel 1 Serum Control, Level I, II, and III, and MassCheck® Steroid Panel 2 Serum Control, Level I, II, and III (Chromsystems, Gräfelfing, Germany) were used as controls. Organic solvent for extraction, as well as mobile phases A and B for chromato- graphic separation, were supplied as part of the IVD kit. The total chromatographic run time was 23 min per sample. The lower and upper limits of quantification and intra- and inter-assay coefficients of varia- tion are summarized in Supplementary Table 2. For the prediction of ACC, the quantitative data of sex-independent steroid hormones (11 out of 13 steroid hormones, excluding testosterone and progesterone) were used.
3. Statistical analysis
Statistical analysis was performed using IBM SPSS Statistics version 28 (IBM Corp.). Categorical variables were presented as numbers and percentages, with comparisons made using either the Chi-square test or Fisher’s exact test, as appropriate. Continuous variables were reported as medians with interquartile ranges (IQRs) or as ranges. Group-wise
comparisons of continuous variables were performed using the unpaired t-test or the Mann-Whitney U test, depending on data distribution. To identify independent predictors of ACC, each steroid was individually assessed using univariate binary logistic regression, and variables with a p-value < 0.1 were retained for multivariate analysis. Multivariate model construction employed forward, backward, and stepwise selec- tion procedures to identify the combination of predictors yielding the highest area under the receiver operating characteristic (ROC) curve.
4. Results
Sixty-two patients (31 males) with a unilateral, lipid-poor, large adrenal mass (≥3 cm) and for whom serum steroid profiling by LC-MS/ MS assay was available at diagnosis, were included. The median age at diagnosis was 41 years (32-50). The median tumour diameter was 9.6 cm (3.1-19.5), and the median unenhanced attenuation was 36.2 HU (13.1-65). ACC was diagnosed in 59.6 % (37/62) patients, while the remaining non-ACC masses (n = 25) consisted of PHEO (n = 9), ACA (n = 8; 2 overt cushing syndrome, 1 mild autonomous cortisol secretion, and 5 non-functioning), metastases (n = 4; primary malignancy being lung: 2, colon: 1, and unknown: 1), schwannoma (n = 2), ganglioneur- oma (n = 1), and lymphoma (n = 1). ONDS cortisol > 50 nmol/L was seen in 20/37 (54 %) ACC and 7/25 (28 %) non-ACC masses. ENSAT stage I, II, III, and IV in 2, 14, 7, and 14 ACC patients, respectively. The final diagnosis was based on histopathology, except for three patients with ACA, who had not undergone surgery, and the diagnosis was based on the fact that there was no significant change in adrenal mass size on follow-up.
A comparison of baseline clinical, biochemical, and radiological characteristics between the ACC and non-ACC groups is presented in Table 1. The two groups were comparable in terms of age, sex distri- bution, mode of presentation, and basal attenuation of the tumor, while
the tumor size was larger (9.9 vs. 7 cm; p < 0.001) in the ACC group. Among the steroids measured by LC-MS/MS, 9/13 steroids: DOC, 17OHP, S, E, A4, DHEA, and DHEAS, T (in females), and P4 (in males) were found to be significantly (p < 0.05) elevated in the ACC group. For further analysis, 7 of the nine significantly elevated steroids were considered after excluding the two sex-dependent variables (T and P4). On univariate logistic regression analysis, 6/7 steroids (17OHP, S, E, A4, DHEA, and DHEAS) had p-values < 0.1, and hence were included in a multivariate logistic regression model. Among the selection methods evaluated, backward elimination yielded the model with the best discriminatory value, comprising three steroids: A4, S, and DHEAS, which together achieved the highest area under the ROC curve (AUC: 0.923). The ROC curve and the corresponding predictive formula are shown in Fig. 1. A cutoff value of 0.52 was 83.8 % sensitive and 96 % specific for diagnosing ACC. The link to the online calculator based on this model is available at https://endo-tools.github.io/ACC-prediction -tool/.
5. Discussion
In this study, serum steroid profiling using LC-MS/MS revealed elevated levels of steroid precursors/metabolites (DOC, 17OHP, S, E), and androgens (A4, DHEA, and DHEAS) in the ACC group vs other indeterminate (size ≥ 3 cm and basal attenuation >20 HU) adrenal masses. Applying logistic regression, a predictive model comprising three key steroids (S, A4, and DHEAS) had the highest discriminatory value (AUC: 0.923) for diagnosing ACC.
Urine steroid profiling helps differentiate ACC from other adrenal masses [8,10,19-21]. In a large prospective study, urinary steroid profiling had a sensitivity of 85 % and specificity of 83.6 % for diag- nosing ACC among indeterminate (size ≥ 4 cm and basal attenu- ation>20 HU) adrenal masses [8]. However, it is not widely available
| ACC (n = 37) | Non-ACC (n = 25) | p value | |||
|---|---|---|---|---|---|
| Age in years | 39 | (30-50) | 42 | (35.5-56) | 0.209 |
| Male/Female | 18/19 | 13/12 | 1.000 | ||
| Presentation | |||||
| Incidental/mass effect | 24/37 (64.8 %) | 21/25 (84 %)* | 0.257 | ||
| Adrenocortical hormone excess symptoms | 13/37 (35.1 %) | 4/25 (16 %) | 0.147 | ||
| ONDS Cortisol> 50 nmol/L | 20/37 (54 %) | 7/25 (28 %) | 0.067 | ||
| Radiological characteristics | |||||
| Right/Left | 17/20 | 15/10 | 0.311 | ||
| Maximum dimension (cm)@ | 9.9 | (3.8-19.5) | 7 | (3.1-15.6) | < 0.001 |
| Basal HU@ | 37.3 | (19-50) | 35.8 | (13.1-65) | 0.170 |
| Serum steroids via LC-MS/MS | |||||
| Aldosterone (nmol/L) | 0.15 | (0.06-0.37) | 0.21 | (0.08-0.32) | 0.769 |
| Androstenedione (nmol/L) | 4.99 | (3.03-11.7) | 2.01 | (1.29-2.58) | < 0.001* |
| Cortisol (nmol/L) | 364.2 | (237.9-542.1) | 269.6 | (156.9-437.3) | 0.113 |
| Cortisone (nmol/L) | 40.2 | (31.6-55.1) | 38.3 | (16.1-42.6) | 0.039* |
| Corticosterone (nmol/L) | 5.81 | (2.63-11.0) | 4.99 | (1.51-9.06) | 0.447 |
| 11-deoxycortisol (nmol/L) | 5.71 | (1.15-23.9) | 0.49 | (0.08-0.95) | < 0.001* |
| 21-deoxycortisol (nmol/L) | 0.14 | (0.07-1.05) | 0.07 | (0.07-0.17) | 0.091 |
| DHEA (nmol/L) | 8.01 | (2.97-15.1) | 2.57 | (0.84-4.61) | < 0.001* |
| DHEAS (nmol/L) | 4858 | (1593-17179) | 977 | (468-2042) | < 0.001* |
| 11-deoxycorticosterone (nmol/L) | 0.48 | (0.19-1.01) | 0.06 | (0.04-0.22) | < 0.001 |
| 17-hydroxyprogesterone (nmol/L) | 2.42 | (1.79-3.78) | 1.57 | (0.59-2.89) | 0.018* |
| Progesterone in male (nmol/L) | 1.05 | (0.51-1.83) | 0.28 | (0.19-0.47) | 0.007 |
| Testosterone in female (nmol/L) | 1.49 | (0.76-2.70) | 0.34 | (0.24-0.54) | < 0.001 |
Data expressed as n/N, median (interquartile range) unless specified.
Abbreviations: ACC, Adrenocortical carcinoma; cm, centimeter; DHEA, Dehydroepiandrosterone; DHEAS, Dehydroepiandrosterone sulfate; HU, Hounsfield units; LC- MS/MS, Liquid chromatography- tandem mass spectrometry; ONDS, overnight dexamethasone suppression test.
@ Data expressed as median (range).
* Significantly different (p < 0.1) between ACC and non-ACC groups on univariate logistic regression.
To convert from SI unit (nmol/L) to metric unit (ng/mL), Aldosterone: divide by 2.774; Androstenedione: divide by 3.491; Cortisol: divide by 2.759; Cortisone: divide by 2.774; Corticosterone: divide by 2.886; 11-Deoxycortisol: divide by 2.886; 21-Deoxycortisol: divide by 2.886; DHEA: divide by 3.467; 11-Deoxycorticosterone: divide by 3.03; 17-OHP: divide by 3.026; Progesterone: divide by 3.180; Testosterone: divide by 3.467;
To convert SI unit umol/1 to µg/L, DHEAS: divide by 2.714
| A) | Model | AUC | P(ACC) | Sensitivity(%) | Specificity(%) |
|---|---|---|---|---|---|
| 11-Deoxycortisol (S) | 0.923 | 0.52 | 83.8 | 96 | |
| Androstenedione (A4) | |||||
| DHEAS |
B) P(ACC)
1 + exp[-(-3.87895 + 0.938 × A4 +0.18877× S+0.41356 × DHEAS)]
1
C)
| At P(ACC): 0.52 | Predicted Diagnosis | |
|---|---|---|
| Actual Diagnosis | ACC | Non-ACC |
| ACC | 31 | 6 |
| Non-ACC | 1 | 24 |
D)
100
80
Sensitivity
60
40
20
AUC = 0.923
P < 0.001
0
0
20
40
60
80
100
100-Specificity
and hence is less often implemented in clinical practice [16]. More recently, the focus has shifted to serum steroid profiling via LC-MS/MS, as it may be a relatively more practically available alternative [15,16]. Only a few studies have reported the utility of serum steroid metab- olomics for diagnosing ACC [14-16]. Similar to our study, every study has observed a relative increase of steroid hormone precursors and/or androgens in the ACC group, reflecting their inefficient tumoral steroidogenesis.
The methods and results of previously published studies reporting the use of serum steroid metabolomics for diagnosing ACC, and their
comparison to the current study, are collated in Table 2 [14-16]. Serum levels of S, a cortisol precursor, have consistently been shown to be elevated in patients with ACC compared with non-ACC in all three prior studies as well as in our study [14-16]. Elevated serum levels of an- drogens and androgen precursors have been noted in our study and in the studies by Schweitzer et al. and Berke et al. [14,15]. In contrast, the study by Kai Yu et al. did not evaluate androgens; however, 17-hydrox- ypregnenolone was assessed and found to be elevated [16]. This steroid was not evaluated in our study or in the studies by Schweitzer et al. and Berke et al. Whether the addition of this steroid hormone to our panel
| Author | ACC vs non-ACC | Serum Steroids Metabolomics | Performance | ||||
|---|---|---|---|---|---|---|---|
| Steroid hormones analysed | ML model | Discriminatory steroid hormones | Cut- off | Sensitivity | Specificity | ||
| Schweitzer, | 42 ACC vs 66 ACA@ | 15 steroids: A, B, DOC, P4, E, F, S, 21-DF, | Logistic | Males: B, P4, S, E2, DHEA, A4 | NA | 80 % (12/ | 96.6 % (28/ |
| Germany | 17OHP, E2, DHT, T, DHEA, DHEAS, A4 | regression | 15) | 29) | |||
| Female: DOC, 17OHP, DHT, | NA | 77.8 % (21/ | 97.3 % (36/ | ||||
| DHEA, DHEAS, A4 | 27) | 37) | |||||
| Berke, | 19 ACC vs 558 non- | 18 steroids: A, 18-OHB, 11-DHB, B, DOC, | Logistic | A, B, DOC, 18-oxo-F, 18-OHF, | NA | 74 % (14/ | 98 % (547/ |
| Germany | ACC | P4, 18-oxo-F, 18-OF, E, F, S, 21-DF, 17OHP, DHT, T, DHEA, DHEAS, A4 | regression | F, S, 17OHP, DHEA, DHEAS, A4 | 19) | 558) | |
| Kai Yu, | 44 ACC vs 219 non- | S, 17OHP, 17-OH Preg | LGMLVQ | S, 17OHP, 17-OH Preg | 0.40 | 84.1 % (37/ | 90.9 % |
| USA | ACC | 44) | (199/219) | ||||
| (overall cohort) | |||||||
| 37 ACC vs 41 non- | S, 17OHP, 17-OH Preg | LGMLVQ | S, 17OHP, 17-OH Preg | 0.40 | 86.6 % (32/ | 78 % (32/ | |
| ACC(All ≥4 cm and | 37) | 41) | |||||
| ≥20 HU) | |||||||
| Present study | 37 ACC vs 25 non- | A, B, DOC, P4, E, F, S, 21-DF, 17OHP, T, | Logistic | S, DHEAS, A4 | 0.52 | 83.8 % (31/ | 96 % (24/ |
| ACC$(All ≥3 cm and | DHEA, DHEAS, A4 | regression | 37) | 25) | |||
| ≥10 HU) | |||||||
| 35 ACC vs 18 non- | A, B, DOC, P4, E, F, S, 21-DF, 17OHP, T, | Logistic | S, DHEAS, A4 | 0.52 | 85.7 % (30/ | 94.4 % (17/ | |
| ACC(All ≥4 cm and | DHEA, DHEAS, A4 | regression | 35) | 18) | |||
| ≥20 HU) | |||||||
Abbreviations: 11-DHB, 11-dehydrocorticosterone; 17-OH Preg, 17-hydroxypregnenolone; 17OHP, 17-hydroxyprogesterone; 18-OHB 18-hydroxycorticosterone, 18- OF, 18-hydroxycortisol; 18-oxo-F, 18-oxocortisol; 21-DF, 21-deoxycortisol; A, aldosterone; ACA, adrenocortical adenoma; ACC, adrenocortical carcinoma; A4, an- drostenedione; B, corticosterone; DHEA, dehydroepiandrosterone; DHEAS, dehydroepiandrosterone sulfate; DHT, dihydrotestosterone; DOC, 11-deoxycorticosterone; E, cortisone; E2, estradiol; F, cortisol; LC-MS/MS, liquid chromatography tandem mass spectrometry; LGMLVQ, localized generalized matrix learning vector quan- tization; ML, machine learning; NA, not available; P4, progesterone; S, 11-deoxycortisol; T, testosterone.
Overnight dexamethasone suppressed serum cortisol> 50 nmol/L: @(21/42 ACC and 24/66 ACA) and $(20/37 ACC and 7/25 non-ACC adrenal masses)
would have improved diagnostic performance remains intriguing.
The median age of our study cohort was relatively younger (41 years) compared to others, ranging from 54 to 64 years [14,16]. Only two of these previous studies have reported ACC patients with a number greater than 30, like ours. Similar to Schweitzer et al.’s study cohort, our series has included a proportion of ACC patients with ENSAT stage IV tumors, which may have severe steroid phenotypes [14]. The ~84 % sensitivity for diagnosis of ACC in our study is comparable to the pre- vious reports of 74-86.6 %. In our cohort, 6 of 37 ACC cases were misclassified as non-ACC at the cutoff probability of 0.52. These included two patients with tumors co-secreting cortisol and aldosterone (4.5 and 9.6 cm), one patient with a cortisol-secreting tumor (3.8 cm), and three patients with non-secreting tumors (10.4, 16.6, and 18 cm). Overall, 3 of 20 (15 %) low-grade ACCs and 3 of 17 (17.6 %) high-grade ACCs were misclassified, with no significant difference in misclassifi- cation rates between low- and high-grade tumors (p = 0.82). The misclassification of low-grade ACCs may reflect a biological continuum between low-grade ACC and ACA. In contrast, two of the three mis- classified high-grade ACCs were non-secretory, which may have contributed to their classification as non-ACC masses. It is plausible that a subset of ACC patients (particularly the larger lesions ≥10 cm) are truly non-functional or biochemically negative, as has been reported for PHEOs [22]. Another possibility is that a random serum sample might have missed the ACC’s biochemical signature of functioning ACCs, and a 24-hour urine sample may be more suitable in such a scenario. Further studies are needed to compare the results of serum vs. urine steroid metabolomics in this subset of ACCs.
The specificity for the diagnosis of ACC in our study (~96 %) and previous reports (90.9-98 %) is impressive. In the study by Berke et al., which reported 98 % specificity, the common cause of misdiagnosis was non-functioning ACAs, highlighting that at times, the preoperative distinction between ACC and non-functioning ACAs may be difficult, and only a histopathological diagnosis will be definitive [15]. In our cohort, only 1 of 25 non-ACC masses-a 9.8-cm right-sided PHEO-was misclassified as ACC, while all 8 ACAs were correctly classified as non-ACC masses (p = 1.00 for both low-grade ACC vs. ACA and high-grade ACC vs. ACA). Less commonly, PHEO and primary aldoste- ronism (PA) can also be misdiagnosed as ACCs. As the majority of PHEOs are secretory, the addition of plasma-free metanephrines can aid in their identification [23]. Similarly, most adrenal masses with PA are benign and smaller in size, and very rarely does an ACC exhibit autonomous aldosterone secretion [24]. The significant difference in our series compared to previous studies is that the non-ACC group was selected based on size (≥3 cm) and basal attenuation (≥10 HU). In a prospective multicenter study, ACC constituted ~40 % of the adrenal masses with a size ≥ 4 cm and an unenhanced HU ≥ 20 HU [8]. In our series, the proportion of ACCs amongst the adrenal masses (size ≥ 3 cm and an unenhanced HU >10 HU) was relatively higher (~60 %). We believe this is primarily due to a referral bias from a nearby oncology center. The nature of the non-ACC adrenal mass affects the specificity of this test, as demonstrated by the study by Yu et al., where the specificity was 90.9 % for the overall non-ACC mass cohort and 78 % (32/41) when specifically the non-ACC masses with size > 4 cm and an unenhanced attenuation ≥ 20 HU were considered [16]. Notably, specificity was maintained at ~95 % (17/18) in our cohort, with non-ACC masses with size ≥ 4 cm and an unenhanced attenuation > 20 HU were considered. The differ- ence could be due to the smaller sample size of non-ACC masses and a lesser proportion of non-functioning ACAs in our cohort. Whether the selection of the three key steroids in the Yu et al. study (S, 17OHP, 17-OHpreg) vs our study (S, A4, DHEAS) might have contributed to this difference remains an open query [16]. It is worth noting that non-secretory ACA and PHEO are associated with lower levels of an- drogens (DHEAS, DHEA, and/or A4) compared to controls; hence, a higher weightage to these androgens may contribute to the higher discriminatory power of the serum steroid panel [25,26].
Limitations of our study include the retrospective design and small
sample size. Although the number of non-ACC masses (particularly ACA) was small, the strength of our study is an exclusive focus on unilateral, lipid-poor, large adrenal masses, a subgroup in which accurate diagnosis of ACC is critical. The results of our study, like the previous ones, cannot be applied to children, adolescents, and younger adults. It is well established that androgen secretion does not discriminate ACA from ACC in childhood and adolescent adrenocortical tumors [27]. The me- dian age of our study cohort was 41 years; therefore, caution is advised when applying this in younger adults, especially in females, where coexisting ovarian hyperandrogenism may confound the results. Finally, prospective validation in a larger patient population is needed to confirm the diagnostic performance of this steroid panel.
6. Conclusion
Serum steroid profiling, comprising three key steroids (S, A4, and DHEAS), via LC-MS/MS is a non-invasive and effective approach for distinguishing ACC from other indeterminate adrenal masses (size ≥ 3 cm and basal attenuation ≥10 HU). Future research is required to validate our findings in larger cohorts of ACCs and other indetermi- nate adrenal masses with varied etiologies.
Statement of ethics
The research was performed in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of Seth GS Medical College and KEM Hospital, Mumbai (Reference number EC/OA- 47/2025). As this was a retrospective study, obtaining informed consent from participants was not applicable and was waived by the Institutional Ethics Committee.
Disclosure summary
There are no conflicts of interest related to this study. The authors have nothing to disclose.
CRediT authorship contribution statement
Archana Rao: Conceptualization, Methodology, Investigation, Data curation, Writing - original draft. Aditya Phadte: Conceptualization, Methodology, Investigation, Data curation, Writing - original draft. Anuj Ban: Methodology, Investigation, Data curation, Writing - original draft. Saba Samad Memon: Investigation, Data curation, Writing - re- view & editing. Manjiri Karlekar: Investigation, Data curation, Writing - review & editing, Anurag Ranjan Lila: Conceptualization, Method- ology, Supervision, Project administration, Writing - review & editing. Vijaya Sarathi: Investigation, Data curation, Writing - review & edit- ing. Nimmi Kansal: Methodology, Writing - review & editing. Rohit Barnabas: Methodology, Formal analysis, Writing - review & editing. Padma Vikram Badhe: Supervision, Writing - review & editing. Gwendolyn Fernandes: Resources, Writing - review & editing. Sameer Rege: Resources, Writing - review & editing. Gagan Prakash: Re- sources, Writing - review & editing. Santosh Menon: Resources, Writing - review & editing. Nalini Shah: Supervision, Writing - review & editing. Tushar Bandgar: Supervision, Writing - review & editing.
Funding
The authors received no financial support for the research, author- ship, and/or publication of this article.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jsbmb.2026.106937.
Data availability
Data will be made available on request.
References
[1] J. Calissendorff, C.C. Juhlin, A. Sundin, I. Bancos, H. Falhammar, Adrenal myelolipomas, Lancet Diabetes Endocrinol. 9 (2021) 767-775, https://doi.org/ 10.1016/S2213-8587(21)00178-9.
[2] P. Dogra, A. Sundin, C.C. Juhlin, J. Calissendorff, H. Falhammar, I. Bancos, Rare benign adrenal lesions, Eur. J. Endocrinol. 188 (2023) 407-420, https://doi.org/ 10.1093/ejendo/lvad036.
[3] Y.S. Elhassan, C.L. Ronchi, P. Wijewickrama, S.E. Baldeweg, Approach to the Patient With Adrenal Hemorrhage, J. Clin. Endocrinol. Metab. 108 (2023) 995-1006, https://doi.org/10.1210/clinem/dgac672.
[4] M. Korobkin, F.J. Brodeur, G.G. Yutzy, I.R. Francis, L.E. Quint, N.R. Dunnick, et al., Differentiation of adrenal adenomas from nonadenomas using CT attenuation values, AJR Am. J. Roentgenol. 166 (1996) 531-536, https://doi.org/10.2214/ ajr.166.3.8623622.
[5] A. Phadte, B. Krishnappa, S.S. Memon, V. Patil, A. Lila, P.V. Badhe, et al., High diagnostic accuracy of arterial phase CT in differentiating pheochromocytoma in good/poor washout adrenal masses, J. Endocr. Soc. 9 (2024) bvae199, https://doi. org/10.1210/jendso/bvae199.
[6] S. Puglisi, A. Calabrese, F. Ferraù, M.A. Violi, M. Laganà, S. Grisanti, et al., New findings on presentation and outcome of patients with adrenocortical cancer: results from a national cohort study, J. Clin. Endocrinol. Metab. 108 (2023) 2517-2525, https://doi.org/10.1210/clinem/dgad199.
[7] ] M. Daher, J. Varghese, S.K. Gruschkus, C. Jimenez, S.G. Waguespack, S. Bedrose, et al., Temporal trends in outcomes in patients with adrenocortical carcinoma: a multidisciplinary referral-center experience, J. Clin. Endocrinol. Metab. 107 (2022) 1239-1246, https://doi.org/10.1210/clinem/dgac046.
[8] I. Bancos, A.E. Taylor, V. Chortis, A.J. Sitch, C. Jenkinson, C.J. Davidge-Pitts, et al., Urine steroid metabolomics for the differential diagnosis of adrenal incidentalomas in the EURINE-ACT study: a prospective test validation study, Lancet Diabetes Endocrinol. 8 (2020) 773-781, https://doi.org/10.1016/S2213-8587(20)30218-7.
[9] C. Sturgeon, W.T. Shen, O.H. Clark, Q .- Y. Duh, E. Kebebew, Risk assessment in 457 adrenal cortical carcinomas: how much does tumor size predict the likelihood of malignancy? J. Am. Coll. Surg. 202 (2006) 423-430, https://doi.org/10.1016/j. jamcollsurg.2005.11.005.
[10] W. Arlt, M. Biehl, A.E. Taylor, S. Hahner, R. Libé, B.A. Hughes, et al., Urine steroid metabolomics as a biomarker tool for detecting malignancy in adrenal tumors, J. Clin. Endocrinol. Metab. 96 (2011) 3775-3784, https://doi.org/10.1210/ jc.2011-1565.
[11] I. Bancos, S. Tamhane, M. Shah, D.A. Delivanis, F. Alahdab, W. Arlt, et al., DIAGNOSIS OF ENDOCRINE DISEASE: the diagnostic performance of adrenal biopsy: a systematic review and meta-analysis, Eur. J. Endocrinol. 175 (2016) R65-R80, https://doi.org/10.1530/EJE-16-0297.
[12] M. Fassnacht, S. Tsagarakis, M. Terzolo, A. Tabarin, A. Sahdev, J. Newell-Price, 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. 189 (2023) G1-G42, https://doi.org/10.1093/ejendo/lvad066.
[13] D.R. Taylor, L. Ghataore, L. Couchman, R.P. Vincent, B. Whitelaw, D. Lewis, et al., A 13-steroid serum panel based on LC-MS/MS: use in detection of adrenocortical carcinoma, Clin. Chem. 63 (2017) 1836-1846, https://doi.org/10.1373/ clinchem.2017.277624.
[14] S. Schweitzer, M. Kunz, M. Kurlbaum, J. Vey, S. Kendl, T. Deutschbein, et al., Plasma steroid metabolome profiling for the diagnosis of adrenocortical carcinoma, Eur. J. Endocrinol. 180 (2019) 117-125, https://doi.org/10.1530/EJE-18-0782.
[15] K. Berke, G. Constantinescu, J. Masjkur, O. Kimpel, U. Dischinger, M. Peitzsch, et al., Plasma steroid profiling in patients with adrenal incidentaloma, J. Clin. Endocrinol. Metab. 107 (2022) e1181-e1192, https://doi.org/10.1210/clinem/ dgab751.
[16] K. Yu, S. Athimulam, J. Saini, R.J. Kaur, Q. Xue, T.J. Mckenzie, et al., Serum steroid profiling in the diagnosis of adrenocortical carcinoma: a prospective cohort study, J. Clin. Endocrinol. Metab. 110 (2025) 1177-1186, https://doi.org/ 10.1210/clinem/dgae604.
[17] S. Aubert, A. Wacrenier, X. Leroy, P. Devos, B. Carnaille, C. Proye, et al., Weiss system revisited: a clinicopathologic and immunohistochemical study of 49 adrenocortical tumors, Am. J. Surg. Pathol. 26 (2002) 1612-1619, https://doi.org/ 10.1097/00000478-200212000-00009.
[18] O. Mete, L.A. Erickson, C.C. Juhlin, R.R. de Krijger, H. Sasano, M. Volante, et al., Overview of the 2022 WHO classification of adrenal cortical tumors, Endocr. Pathol. 33 (2022) 155-196, https://doi.org/10.1007/s12022-022-09710-8.
[19] T.M.A. Kerkhofs, M.N. Kerstens, I.P. Kema, T.P. Willems, H.R. Haak, Diagnostic value of urinary steroid profiling in the evaluation of adrenal tumors, Horm. Cancer 6 (2015) 168-175, https://doi.org/10.1007/s12672-015-0224-3.
[20] J.M. Hines, I. Bancos, C. Bancos, R.D. Singh, A.V. Avula, W.F. Young, et al., High- resolution, accurate-mass (HRAM) mass spectrometry urine steroid profiling in the diagnosis of adrenal disorders, Clin. Chem. 63 (2017) 1824-1835, https://doi.org/ 10.1373/clinchem.2017.271106.
[21] N. Vogg, T. Müller, A. Floren, T. Dandekar, A. Riester, U. Dischinger, et al., Simplified urinary steroid profiling by LC-MS as diagnostic tool for malignancy in adrenocortical tumors, Clin. Chim. Acta 543 (2023) 117301, https://doi.org/ 10.1016/j.cca.2023.117301.
[22] G. Constantinescu, C. Preda, V. Constantinescu, T. Siepmann, S.R. Bornstein, J.W. M. Lenders, et al., Silent pheochromocytoma and paraganglioma: systematic review and proposed definitions for standardized terminology, Front Endocrinol. (Lausanne) 13 (2022) 1021420, https://doi.org/10.3389/fendo.2022.1021420.
[23] J.W.M. Lenders, Q .- Y. Duh, G. Eisenhofer, A .- P. Gimenez-Roqueplo, S.K.G. Grebe, M.H. Murad, et al., Pheochromocytoma and paraganglioma: an endocrine society clinical practice guideline, J. Clin. Endocrinol. Metab. 99 (2014) 1915-1942, https://doi.org/10.1210/jc.2014-1498.
[24] S.M. Patel, R.K. Lingam, T.I. Beaconsfield, T.L. Tran, B. Brown, Role of radiology in the management of primary aldosteronism, Radiographics 27 (2007) 1145-1157, https://doi.org/10.1148/rg.274065150.
[25] G. Constantinescu, K. Langton, C. Conrad, L. Amar, G. Assié, A .- P. Gimenez- Roqueplo, et al., Glucocorticoid excess in patients with pheochromocytoma compared with paraganglioma and other forms of hypertension, J. Clin. Endocrinol. Metab. 105 (2020) e3374-e3383, https://doi.org/10.1210/clinem/ dgaa423.
[26] G. Di Dalmazi, F. Fanelli, M. Mezzullo, E. Casadio, E. Rinaldi, S. Garelli, et al., Steroid Profiling by LC-MS/MS in nonsecreting and subclinical cortisol-secreting adrenocortical adenomas, J. Clin. Endocrinol. Metab. 100 (2015) 3529-3538, https://doi.org/10.1210/JC.2015-1992.
[27] N. Marti, J. Malikova, J.A. Galván, M. Aebischer, M. Janner, Z. Sumnik, et al., Androgen production in pediatric adrenocortical tumors may occur via both the classic and/or the alternative backdoor pathway, Mol. Cell Endocrinol. 452 (2017) 64-73, https://doi.org/10.1016/j.mce.2017.05.014.