High-Resolution, Accurate-Mass (HRAM) Mass Spectrometry Urine Steroid Profiling in the Diagnosis of Adrenal Disorders
Jolaine M. Hines,1 Irina Bancos,2 Cristian Bancos,3 Raman D. Singh,1 Aditya V. Avula,1 William F. Young,2 Stefan K. Grebe,2,4 and Ravinder J. Singh4*
BACKGROUND: Steroid profiling is a promising diagnostic tool for those with adrenal tumors, Cushing syndrome (CS), and disorders of steroidogenesis. Our objective was to develop a multiple-steroid assay using liquid- chromatography, high-resolution, accurate-mass mass spectrometry (HRAM LC-MS) and to validate the as- say in patients with various adrenal disorders.
METHODS: We collected 24-h urine samples from 114 controls and 71 patients with adrenal diseases. An HRAM LC-MS method was validated for quantitative analysis of 26 steroid metabolites in hydrolyzed urine samples. Differences in steroid excretion between pa- tients were analyzed based on Z-score deviation from control reference intervals.
RESULTS: Limits of quantification were 20 ng/ml. Dilu- tion linearity ranged from 80% to 120% with means of 93% to 110% for all but 2 analytes. Intraassay and inter- assay imprecision ranged from 3% to 18% for all but 1 analyte. Control women had lower excretion of androgen and glucocorticoid precursors/metabolites than men (P< 0.001), but no difference in mineralocorticoids was seen (P = 0.06). Androgens decreased with age in both sexes (P< 0.001). Compared with patients with adreno- cortical adenoma (ACA), patients with adrenocortical carcinoma (ACC) had 11 steroids with increased Z scores, especially tetrahydro-11-deoxycortisol (14 vs 0.5, P = 0.0006), pregnanetriol (7.5 vs -0.4, P = 0.001), and 5-pregnenetriol (5.4 vs -0.4, P = 0.01). Steroid profiling also demonstrated metabolite abnormalities consistent with enzymatic defects in congenital adrenal hyperplasia and differences in pituitary vs adrenal CS.
CONCLUSIONS: Our HRAM LC-MS assay successfully quantifies 26 steroids in urine. The statistically signifi- cant differences in steroid production of ACC vs ACA, adrenal vs pituitary CS, and in congenital adrenal hyper- plasia will allow for improved diagnosis of patients with these diseases.
@ 2017 American Association for Clinical Chemistry
Adrenal steroid analysis plays an important role in the diagnosis of Cushing syndrome (CS)5, disorders of ste- roidogenesis, and adrenal tumors (1-4). Over time, ste- roid assays have improved in analytical sensitivity and specificity, with the current reference standard being LC- MS/MS (5-8). However, it has been suggested that the clinical diagnostic performance of modern assays might have paradoxically worsened compared with older, less- specific immunoassays, in particular for CS (9, 10), sug- gesting that it might be beneficial to measure multiple steroids and their metabolites simultaneously to achieve optimal diagnostic accuracy (10). In addition, during the past 5 to 6 years, a GC-MS urinary 32-analyte steroid profile has shown promising results as a diagnostic tool for distinguishing adrenocortical carcinoma (ACC) from a benign adrenocortical adenoma (ACA) (11-13), fur- ther emphasizing the potential importance of measuring multiple steroids simultaneously.
Based on these observations, steroid metabolomics might be poised to make a substantial impact on endo- crine laboratory testing. Unfortunately, steroid profiles are difficult to implement in the clinical laboratory. Among the plethora of naturally occurring steroids and
1 Immunochemical Core Laboratory, Mayo Clinic, Rochester, MN; 2 Department of Medi- cine, Division of Endocrinology, Mayo Clinic, Rochester, MN; 3 Information Technology, Mayo Clinic, Rochester, MN; 4 Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN.
* Address correspondence to this author at: Department of Laboratory Medicine and Pa- thology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. Fax 202-833-4576; e-mail singh.ravinder@mayo.edu.
Received January 3, 2017; accepted July 20, 2017.
Previously published online at DOI: 10.1373/clinchem.2017.271106
@ 2017 American Association for Clinical Chemistry
5 Nonstandard abbreviations: CS, Cushing syndrome; HRAM, high-resolution, accu- rate-mass; ACC, adrenocortical carcinoma; ACA, adrenocortical adenoma; 11-OXO-
ET, 11-oxoetiocholanolone; 11B-OH-AN, 11B-hydroxyandrosterone; 11B-OH-ET, 11ß-hydroxyetiocholanolone; 16a-DHEA, 16@-hydroxydehydroepiandrosterone; 17HP, 17@-hydroxypregnanolone; 17OHPG, 17-hydroxyprogesterone; 5PT, 5- pregnenetriol; 5@THA, 5a-tetra-11-dehydrocorticosterone; 5a-THF, 5x-tetrahydro- cortisol; DHEA, dehydroepiandrosterone; Etio, etiocholanolone; PD, pregnanediol; PT, pregnanetriol; PTONE, pregnanetriolone; 5PD, 5-pregnenediol; THS, tetrahydro- deoxycortisol; THB, tetrahydrocorticosterone; THF, tetrahydrocortisol; THE, tetrahy- drocortisone; THDOC, tetrahydrodeoxycorticosterone; An, androsterone; 6B-OH-cortisol, 6ß-hydroxycortisol; IS, internal standard; MP, mobile phase; QC, quality control.
their metabolites, there are many compounds that are near isobaric, or are isomers. Many of these compounds cannot be distinguished from each other using the rela- tive low-resolution mass-filtering mass spectrometers used in clinical laboratories, even when MS/MS is used, thus necessitating complete chromatographic separation before mass spectrometry detection. This often requires gas chromatography, a methodology technically far more demanding, labor-intensive, and time-consuming than most LC-MS/MS methods. Consequently, these prom- ising assays are not yet extensively used, and have been scarcely evaluated for their clinical utility outside of the differential diagnosis of ACC vs ACA.
Liquid-chromatography, high-resolution, accurate- mass mass spectrometry (HRAM LC-MS) might be a tool that can overcome the hurdles for wider use of steroid pro- files. This methodology is rapidly gaining in popularity for quantitative clinical analysis of small endogenous molecules, anabolic drug testing, and proteins (14-25). HRAM can resolve all steroids and their metabolites except isomers, al- lowing the use of liquid chromatography, including multi- plexed liquid chromatography setups, as a front-end instead of gas chromatography.
Therefore, we decided to use HRAM LC-MS to develop a novel 26-analyte, urine-based steroid panel, to determine sex- and age-based control reference in- tervals, and to perform a limited clinical evaluation of the assay in a cohort of patients with different adrenal diseases.
Materials and Methods
SUBJECTS
This study was approved by the Mayo Clinic Institu- tional Review Board.
We obtained 24-h urine samples from 114 volun- teers (66 women, median age 47 years, range 25-83 years; 48 men, median age 42 years, range 24-83 years) to establish reference intervals. Exclusion criteria were the presence of any adrenal gland disorder, benign or malig- nant neoplasm of other endocrine glands and related structures, and secondary malignant neoplasm.
For clinical validation, we collected 24-h urine sam- ples from 71 patients with adrenal diseases: 4 patients had adrenocorticotropic hormone-dependent pituitary hy- percortisolism; 1 woman had newly diagnosed congenital adrenal hyperplasia; and 5 patients were diagnosed with ACC and 61 with ACAs (4 with cortisol-producing ACAs and 57 with nonfunctioning ACAs). The final diagnosis in all cases was based on clinical, imaging, and pathology results.
We performed all steroid measurements blinded to the clinical information.
MATERIALS AND METHODS
11-Oxoetiocholanolone (11-OXO-ET), 11ß-hydroxyandro- sterone (11B-OH-AN), 11ß-hydroxyetiocholanolone (11B-OH-ET), 16a-hydroxydehydroepiandrosterone (16a-DHEA), 17a-hydroxypregnanolone (17HP), 5-pregnenetriol (5PT), 5a-tetra-11-dehydrocorticosterone (5aTHA), 5a-tetrahydrocortisol (5a-THF), a-cortolone, B-cortol, ß-cortolone, cortisol, cortisone, dehydroepiandro- sterone (DHEA), etiocholanolone (Etio), pregnanediol (PD), pregnanetriol (PT), pregnanetriolone (PTONE), 5-pregnenediol (5PD), tetrahydrodeoxycortisol (THS), tet- rahydrocorticosterone (THB), tetrahydrocortisol (THF), tetrahydrocortisone (THE), and tetrahydrodeoxycortico- sterone (THDOC) were purchased from Steraloids. Andro- sterone (An), 6B-hydroxycortisol (6B-OH-cortisol), and dehydroepiandrosterone-2,2,3,4,4,6-d6 (DHEA-d6) were purchased from Sigma-Aldrich.
Pregnanetriol-d5 (PT-d5), tetrahydrocortisol-d5 (THF-d5), tetrahydrocortisone-d; (THE-d;), cortisone- 13 3C3, cortisol-13C3, tetrahydrocorticosterone-d; (THB-d;), and 11-deoxycortisol-13C3 were purchased from IsoSciences, while 5a-tetra-11-dehydrocorticosterone-d3 (5aTHA- d3) was purchased from Medical Isotopes and etiocholanolone-d5 (Etio-d5) and tetrahydrodeoxycortico- sterone-d3 (THDOC-d3) were purchased from C/D/N Isotopes.
Glusulase (glucuronidase and sulfatase activity) was from PerkinElmer. HPLC-grade methanol, aceto- nitrile, and ethyl acetate were purchased from Fisher Scientific.
We prepared calibration stocks from each of the pur- chased steroids in methanol (1 g/L) and combined ali- quots from all stocks to create an intermediate concen- tration calibrator, containing 10 µg/mL of each steroid in methanol. Finally, we serially diluted the intermediate calibrator to generate working calibrators containing 5000, 2500, 1250, 625, 312.5, 156.25, 78.13, 39.06, and 19.53 ng/mL, respectively, of each steroid. This large number of calibrators was chosen to ensure linear detec- tion of the large variation in concentrations of our differ- ent analytes.
Nonradioactive isotopic internal standard (IS) stock solutions were made to either 1 g/L or 5 g/L concentra- tions in 50% methanol. We combined aliquots of the stocks to create working ISs containing 400 ng/ml of each IS.
We created 3 distinct batches of quality control (QC) material to cover low, intermediate, and high ana- lyte concentrations for each analyte by spiking calibra- tion material for each analyte into charcoal-stripped urine, tailoring the concentrations of individual analytes to their expected concentration ranges; e.g., a low control might contain concentrations approximately 20 ng/ml for 1 analyte and much higher concentrations for an-
other. Across all analytes and controls, a range of 20- 3000 ng/ml was covered.
SAMPLE STORAGE AND PREPARATION
We aliquoted and froze (-70 ℃ or colder) all samples im- mediately after recording the total urine volume of each 24-h collection. Steroids and their metabolites exist in urine mainly as sulfate or glucuronide conjugates. Hy- drolysis converts these conjugates back to the uncon- jugated metabolites, simplifying mass spectrometry analysis.
For hydrolysis and extraction, we combined 150 ALL of thawed urine or QC material with 50 µL of working IS. Ideally, there should be an IS for each analyte. However, several ISs showed poor reproducibility although analyte response was linear. Therefore, we settled on a final method using 4 ISs: THB-d4 (11.1 min), DHEA-d6 (14.2 min), PT-d5 (15.1 min), and Etio-d5 (16.3 min) (see Table 1 in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol63/ issue11). We verified with dilution studies that these 4 ISs ensured linear responses for all analytes.
The samples were vortex-mixed and sat at ambient temperature for 10 min. We then added 50 pL of sodium acetate buffer (3 mol/L, pH 5.2) and 10 µL of Glusulase to the tubes, covered the tubes, and incubated them for 2 h at 50 ℃ (dry-heat block). We then removed the samples from the heat block and stopped each reaction with 50 uL of potassium carbonate.
Liquid-liquid extraction was then performed with 3 mL of ethyl acetate per sample, followed by 5 min of vortex-mixing on a multitube vortexer (speed approx- imately 1500 rpm) and 15 min of centrifugation at approximately 1500g. We then pipetted 2 mL of the organic layer of each sample into a clean glass tube and dried it at 40 ℃ under nitrogen for 30 min. These extracts were resuspended in 200 µL of 50% acetoni- trile in water, vortex-mixed for 60 s, and transferred to a 96 deep-well plate (Chrom Tech) for analysis.
HRAM LC-MS METHOD
Steroid metabolites were measured using a Thermo Uni- Cell Dionex UPLC system coupled to Thermo Q Exac- tive Plus HRAM hybrid quadrupole/orbital trap mass spectrometer with a heated electrospray ionization source (Thermo Scientific). Data were collected in full-scan mode with 70000 resolution (relative to 200 m/z). The temperature and gas settings are provided in Table 2 of the online Data Supplement.
We used reversed-phase chromatography with Zor- bax Extend-C18, Rapid Resolution HT, 2.1 × 50 mm, 1.8-um analytical columns (Agilent Technologies). Mo- bile phase (MP) A consisted of 10% acetonitrile with 0.1% formic acid. MP B consisted of 90% acetonitrile with 0.1% formic acid. The injection volume was 5 pL at
a flow rate of 300 µL/min, using MP A for 1 min, fol- lowed by a 35-min gradient to 100% MP B and column reconditioning with MP A for 4 min. Retention times of each analyte were established and confirmed by single analyte injections (Fig. 1). An on-board diversion valve was used in line with the mass spectrometer between 2 and 4 min and then again for 7-18 min of the liquid chromatography method, keeping the instrument cleaner by diverting more polar compounds at the start of the chromatography and more nonpolar compounds as the gradient increased and the column was subsequently washed.
Mass spectra were processed with Xcalibur Qual Browser (Thermo) and quantified with TraceFinder Clinical 3.3 (Thermo). The theoretical mass of each ana- lyte was calculated during method development using the chemical formula spectral simulation function in Xcali- bur’s Qual Browser and then confirmed by an injection of the respective calibration standard (Table 1). The in- use m/z for each analyte (Table 1) was found to be pro- tonated, dehydrated, or double-dehydrated (the second- most intense ion was recorded for confirmation purposes described below). Quantification was performed against a calibration curve using analyte-to-IS peak area ratios with linear regression analysis and 1/X weighting. The mass tolerance was set to ±5 ppm, and the retention time window constraint was 10 s (5PD was 30 s to afford less user manipulation because peaks tended to be low). Confirmation-ion target ratios (ratio between the quan- tification ion and a secondary ion most often defined as the second-most intense experimental ion) varied between analytes (and were set according to observed intensities). Because some analytes produced little or no confirmation ion (e.g., cortisol, which produced predominantly monoiso- topic ions), we also used isotopic pattern scores, which were flagged when they were <90; this score indicates the good- ness of fit of the isotopic distribution of the observed data to the theoretically expected pattern, with scores ranging from 0 (no match) to 100 (complete match). Isotopic pattern scoring showed poor utility at concentrations below the lower limit of quantification.
ANALYTICAL VALIDATION
Our criteria for calibration acceptance of individual cal- ibration curves was an R2 >0.995. Calibration curves ranged from 19.53 to 5000 ng/ml (52-68 nmol/L to 13211-17349 nmol/L) using linear regression analysis with 1/X weighting.
For each analyte, the limit of quantification was ar- bitrarily defined as the concentration equivalent to that of the lowest calibrator [rounded to 20 ng/ml (53-69 nmol/L)]. The signal-to-noise ratio was >3 for all ana- lytes at this concentration.
Linearity was determined by diluting urine samples with stripped urine at 2X, 4X, and 8X dilutions. Ex-
3
5.E+08
2
271.2056
317.2475
4.E+08
273.2213
331.2268
283.2420
333.2424
285.2577
351.2530
3.E+08
287.2006
361.2010
Intensity, cps
299.2369
363.2166
57
301.2526
365.2323
2.E+08
315.2319
379.2115
11
19
25
26
4
13
15
1.E+08
89
10
14
21
24
1
6
1
12
16
17 18
23
20
22
2
4
6
8
10
12
14
16
18
Time, min
pected results were defined as the neat (undiluted) result divided by the dilution factor. Recovery was determined by spiking 250, 500, and 1000 ng/ml (661-867, 1321- 1735, and 2642-3470 nmol/L) analytes into 16 different patient urine matrices. Expected values were determined by summing the spike with the neat value. Intraassay imprecision was determined by analyzing 20 QC samples within 1 assay, and interassay imprecision was calculated from 20 replicates assayed over a series of 20 batches. Any concentration with a mean that fell below the lower limit of quantification for a particular analyte was removed from the analysis, including the lower controls for 5PD, 5PT, DHEA, and PD. The imprecision was expressed as percentage CV.
CLINICAL VALIDATION
Any sample results that fell below the lower limit of quan- tification of the assay (19.53 ng/ml) were reported as <20 ng/ml (<53-69 nmol/L). Sample results that ex- ceeded the upper limit of the calibration curve (5000 ng/ml, 13211-17349 nmol/L) were diluted in stripped urine before hydrolysis and extraction until they fell into the calibration range. Subsequent results were multiplied
by the total volume recorded at the end of 24-h urine collection and divided by 1000 to give final results in micrograms per 24 h.
STATISTICS
Raw steroid results (ng/ml) were normalized to individ- ual patients’ total urine volume for final units of micro- grams per 24 h. To account for differences in sex and age, individual steroid results for each patient were trans- formed into standard scores (Z scores), defined as: Z = “0” x- u where x was the measured steroid value, and µ and o were the mean steroid value and its SD, respectively, in controls of the same sex and age groups. Data were analyzed using JMP software, ver- sion 10 (SAS). Depending on the data distributions, descriptive statistics were used to determine mean and SD, or median and interquartile ranges (IQR 25%, 75%), respectively, and intergroup steroid differences were analyzed by Student t-tests or Wilcoxon/ Kruskal-Wallis tests, respectively. P values <0.05 were considered significant.
| Table 1. Analytes listed in order of retention time with peak number corresponding to Figure 1. | ||||||
|---|---|---|---|---|---|---|
| Peak number | Retention time, min | Analyte | Full name | Molecular formula | Ion formation | m/z used for quantification |
| 1 | 2.70 | 6B-OH-Cortisol | 6ß-Hydroxycortisol | C21H30O6 | [M+H]+ | 379.212 |
| 2 | 8.44 | Cortisol | Cortisol | C21H30O5 | [M+H]+ | 363.217 |
| 3 | 8.63 | Cortisone | Cortisone | C21H28O5 | [M+H]+ | 361.201 |
| 4 | 8.67 | B-Cortol | B-Cortol | C21H36O5 | [M+H-2H2O]+ | 333.242 |
| 5 | 9.02 | a-Cortolone | @-Cortolone | C21H34O5 | [M+H-2H2O]+ | 331.2267 |
| 6 | 9.21 | 16a-DHEA | 16a-Hydroxydehydroepiandrosterone | C19H28O3 | [M+H-H2O]+ | 287.201 |
| 7 | 9.30 | B-Cortolone | B-Cortolone | C21H34O5 | [M+H-2H2O]+ | 331.227 |
| 8 | 9.41 | 5a-THF | 5a-Tetrahydrocortisol | C21H34O5 | [M+H-2H2O]+ | 331.227 |
| 9 | 9.60 | THF | Tetrahydrocortisol | C21H34O5 | [M+H-2H2O]+ | 331.227 |
| 10 | 10.30 | THE | Tetrahydrocortisone | C21H32O5 | [M+H]+ | 365.232 |
| 11 | 10.81 | PTONE | Pregnanetriolone | C21H34O4 | [M+H-2H2O]+ | 315.232 |
| 12 | 11.12 | THB | Tetrahydrocorticosterone | C21H34O4 | [M+H-2H2O]+ | 315.232 |
| 13 | 11.24 | 11B-OH-ET | 11 B-Hydroxyetiocholanolone | C19H30O3 | [M+H-2H2O]+ | 271.206 |
| 14 | 11.36 | 5PT | Pregnenetriol | C21H34O3 | [M+H-2H2O]+ | 299.237 |
| 15 | 11.40 | 11B-OH-AN | 11 ß-Hydroxyandrosterone | C19H30O3 | [M+H-2H2O]+ | 271.206 |
| 16 | 12.04 | 11-OXO-ET | 11-Oxoetiocholanolone | C19H28O3 | [M+H-H2O]+ | 287.201 |
| 17 | 13.56 | THS | Tetrahydrodeoxycortisol | C21H34O4 | [M+H]+ | 351.253 |
| 18 | 14.16 | DHEA | Dehydroepiandrosterone | C19H28O2 | [M+H-H2O]+ | 271.206 |
| 19 | 15.14 | PT | Pregnanetriol | C21H36O3 | [M+H-2H2O]+ | 301.252 |
| 20 | 15.54 | THDOC | Tetrahydrodeoxycorticosterone | C21H34O3 | [M+H-H2O]+ | 317.248 |
| 21 | 15.76 | 5PD | Pregnenediol | C21H34O2 | [M+H-2H2O]+ | 283.242 |
| 22 | 15.90 | 5a-THA | 5a-Tetra-11-dehydrocorticosterone | C21H34O3 | [M+H-H2O]+ | 317.248 |
| 23 | 16.30 | Etio | Etiocholanolone | C19H30O2 | [M+H-H2O]+ | 273.221 |
| 24 | 16.79 | An | Androsterone | C19H30O2 | [M+H-H2O]+ | 273.221 |
| 25 | 17.03 | 17HP | 17a-Hydroxypregnanolone | C21H34O3 | [M+H-2H2O]+ | 299.237 |
| 26 | 18.23 | PD | Pregnanediol | C21H36O2 | [M+H-2H2O]+ | 285.258 |
| Isomeric compounds are separated chromatographically. | ||||||
Results
ANALYTICAL VALIDATION
All calibration curves met our acceptance criteria.
The %CV intraassay imprecision ranged from 3% to 10%, and interassay imprecision ranged from 6% to 18% for all analytes with the exception of 16a-DHEA, which had interassay CVs of 16% to 25% (Table 2).
Because not all steroid conjugates were available for purchase and synthesis was prohibitively costly, we opted to use unconjugated steroids as QC material. However, we did scan for glucuronidated and sulfated analytes in a selection of samples and found no residual conjugated analytes.
Dilution linearity met our acceptance criteria with results 80% to 120% of the predicted concentration for individual points, with means of 93% to 110% for all analytes except 5PD and 5PT, which met criteria for 2X dilutions but suffered from poor linearity with overall
ranges of 63% to 93% (mean 79%) and 41% to 91% (mean 71%), respectively (Table 2).
Recovery experiments showed acceptable recoveries for all analytes except DHEA, 5PT, and 5PD (Table 2), which showed substantial loss during extraction. Loss of DHEA, 5PT, and 5PD was less pronounced when en- dogenous concentrations of DHEA, 5PT, and 5PD were high (see Table 6 in the online Data Supplement).
CLINICAL VALIDATION
The reference intervals showed differences between men and women, and among women based on pre/ postmenopausal status. The most significant differ- ences between sexes were seen in androgens (Table 3). Additionally, glucocorticoids and glucocorticoid pre- cursors were generally higher in men. Analysis based on age was also performed, and subsequent Z scores were calculated based on sex, menopausal status, and age.
| Table 2. Imprecision, dilution linearity and spiked recovery. | ||||||||
|---|---|---|---|---|---|---|---|---|
| Analyte | Intraassay mean range, ng/ml (nmol/L) | Intraassay % CV range | Interassay mean range, ng/ml (nmol/L) | Interassay % CV range | Linearity range, ng/ml (nmol/L) | Linearity % of predicted value | Linearity mean | Recovery % of predicted value |
| An | 340-1820 (1172-6271) | 5-6% | 348-1706 (1199-5878) | 9-12% | 30-2254 (103-7766) | 85-110% | 97% | 81-111% |
| Etio | 347-2962 (1196-10206) | 6-7% | 368-2703 (1268-9313) | 8-10% | 82-6697 (283-23,075) | 84-110% | 99% | 82-110% |
| DHEA | 891 (3092) | 6% | 212-821 (436-2849) | 9-17% | 26-624 (90-2165) | 100-120% | 110% | 19-91% |
| 16a-DHEA | 54-742 (178-2439) | 7-10% | 44-636 (145-2091) | 16-25% | 47-893 (155-2936) | 93-118% | 110% | 80-119% |
| 5PD | 2307 (7249) | 6% | 2188 (6875) | 16% | 25-159 (79-500) | 63-93% | 79% | 1-46% |
| 5PT | 1759 (5259) | 6% | 1610 (4813) | 7% | 32-574 (96-1716) | 41-91% | 71% | 9-110% |
| 5a-THA | 39-72 (117-215) | 8-9% | 36-86 (108-257) | 12-16% | 36-128 (108-383) | 113-119% | 115% | 80-120% |
| THB | 27-74 (77-211) | 5-10% | 40-85 (114-243) | 12-15% | 29-368 (83-1050) | 85-113% | 102% | 80-108% |
| THDOC | 32-67 (96-200) | 9% | 35-67 (105-200) | 12-15% | 20-146 (60-436) | 88-119% | 108% | 87-119% |
| PD | 665 (2076) | 5% | 129-602 (403-1880) | 11-15% | 118-3929 (368-12,268) | 88-114% | 104% | 80-107% |
| 17HP | 49-348 (147-1041) | 8-10% | 52-336 (156-1005) | 8-13% | 26-308 (78-921) | 105-120% | 109% | 80-106% |
| PT | 106-1710 (315-5082) | 6-8% | 109-1513 (324-4496) | 6-10% | 49-1809 (146-5376) | 84-107% | 95% | 82-105% |
| PTONE | 47-90 (134-257) | 5-7% | 40-84 (114-240) | 7-12% | 24-347 (68-990) | 99-115% | 108% | 81-103% |
| THS | 93-1694 (265-4833) | 3-8% | 29-1566 (83-4468) | 11-15% | 31-358 (88-1021) | 84-107% | 95% | 82-118% |
| Cortisol | 165-1972 (455-5441) | 3-6% | 150-1626 (414-4486) | 10-11% | 30-327 (83-902) | 84-115% | 104% | 82-120% |
| 6B-OH-cortisol | 152-703 (402-1858) | 6% | 126-531 (333-1403) | 16-17% | 31-1021 (82-2698) | 83-111% | 96% | 82-111% |
| Cortisone | 175-1996 (486-5538) | 3-6% | 144-1502 (400-4167) | 13-14% | 24-499 (67-1384) | 90-103% | 95% | 80-118% |
| THF | 418-2030 (1141-5539) | 3-7% | 352-1580 (960-4311) | 12-14% | 112-3613 (306-9858) | 81-110% | 93% | 86-109% |
| 5a-THF | 176-1905 (480-5198) | 5-7% | 149-1325 (407-3615) | 9-14% | 55-2249 (150-6137) | 80-110% | 94% | 81-110% |
| THE | 398-3036 (1092-8330) | 5-8% | 300-2141 (823-5874) | 12-17% | 295-9194 (809-25,226) | 84-107% | 93% | 80-115% |
| B-Cortol | 45-1077 (122-2923) | 5-10% | 37-804 (100-2182) | 8-12% | 20-608 (54-1650) | 80-113% | 102% | 81-110% |
| a-Cortolone | 494-2358 (1348-6434) | 6-9% | 406-1822 (1108-4971) | 8-10% | 109-1748 (297-4770) | 99-118% | 109% | 81-112% |
| B-Cortolone | 183-2447 (499-6677) | 6-9% | 160-1847 (437-5040) | 6-9% | 109-1748 (297-4770) | 99-118% | 109% | 84-110% |
| 11B-OH-AN | 148-1902 (483-6211) | 5-6% | 167-1741 (545-5685) | 8-9% | 73-2385 (258-7789) | 94-119% | 108% | 83-113% |
| 11B-OH-ET | 77-914 (251-2985) | 5-8% | 91-841 (297-2746) | 8-14% | 104-3521 (340-11,498) | 88-117% | 104% | 86-115% |
| 11-OXO-ET | 68-1016 (244-3340) | 7-10% | 98-950 (322-3123) | 10-18% | 61-703 (201-2311) | 85-112% | 103% | 88-117% |
| Linearity ranges include undiluted and subsequent diluted values. Recovery was performed on 16 unique urine samples spiked with 200, 500, and 1000 ng/ml of all analytes as a mixture. | ||||||||
| Table 3. Control reference interval study. | ||||||
|---|---|---|---|---|---|---|
| Analyte | Women premenopausal median, µg/24 h (IQR) | Women postmenopausal median, µg/24 h (IQR) | P valueª | Men median, µg/24 h (IQR) | P valueb | |
| Androgens and precursors | An | 370.3 (248.0, 805.0) | 1288.7 (725.5, 1796.0) | <0.001 | 2150.7 (1650.3, 3417.8) | <0.001 |
| Etio | 764.0 (528.7, 1403.0) | 1737.3 (1118.8, 2242.6) | 0.0005 | 2429.1 (1582.6, 3266.6) | <0.001 | |
| DHEA | 29.8 (21.4, 59.3) | 79.1 (51.8, 115.5) | 0.0007 | 149.4 (103.3, 235.8) | <0.001 | |
| 16a-DHEA | 216.7 (101.7, 485.7) | 552.0 (342.6, 778.4) | 0.0005 | 562.1 (420.0, 918.1) | 0.001 | |
| 5PD | 23.4 (12.0, 36.8) | 80.9 (37.5, 144.0) | <0.001 | 105.3 (72.7, 174.8) | <0.001 | |
| 5PT | 26.1 (13.9, 67.9) | 60.7 (32.5, 99.4) | 0.01 | 84.4 (43.8, 163.5) | 0.002 | |
| Mineralocorticoids and precursors | 5a-THA | 7.5 (3.8, 27.2) | 22.4 (8.8, 49.7) | 0.15 | 18.0 (5.7, 30.0) | 0.27 |
| THB | 101.0 (64.6, 185.1) | 115.2 (76.1, 155.2) | 0.79 | 128.5 (99.0, 196.4) | 0.06 | |
| THDOC | 13.3 (7.6, 23.5) | 25.0 (12.8, 48.6) | 0.0071 | 17.2 (9.8, 36.2) | 0.6 | |
| Glucocorticoid precursors | PD | 114.5 (75.2, 196.6) | 389.9 (152.4, 1326.1) | 0.0001 | 222.9 (148.3, 326.4) | 0.95 |
| 17HP | 38.4 (20.0, 69.1) | 125.9 (52.6, 328.4) | <0.001 | 185.9 (157.0, 279.7) | <0.001 | |
| PT | 256.2 (170.7, 345.9) | 544.4 (309.1, 868.8) | 0.0001 | 715.3 (514.5, 938.3) | <0.001 | |
| PTONE | 31.6 (17.9, 49.7) | 22.4 (15.3, 36.5) | 0.13 | 25.7 (16.9, 39.4) | 0.6 | |
| THS | 82.4 (55.3, 116.0) | 70.9 (48.7, 91.9) | 0.13 | 79.2 (61.3, 120.2) | 0.17 | |
| Glucocorticoids | Cortisol | 60.0 (45.5, 74.7) | 70.3 (51.3, 99.1) | 0.13 | 73.3 (59.1,95.2) | 0.07 |
| 6B-OH-cortisol | 95.6 (70.2, 137.7) | 107.7 (79.7, 131.1) | 0.6 | 109.4 (80.8,159.3) | 0.33 | |
| Cortisone | 101.5 (78.5, 125.0) | 94.2 (79.1, 132.4) | 0.93 | 122.5 (95.0, 162.2) | 0.0004 | |
| THF | 1275.5 (1156.7, 1847.3) | 1025.7 (839.2, 1379.7) | 0.012 | 1566.0 (1310.7, 2154.0) | 0.0001 | |
| 5a-THF | 529.7 (324.5, 749.3) | 379.9 (223.1, 794.4) | 0.39 | 1127.4 (753.4, 1863.8) | <0.001 | |
| THE | 2674.2 (2015.5, 3476.9) | 1851.8 (1166.6, 2985.9) | 0.06 | 3078.8 (2359.8, 4328.9) | 0.0002 | |
| B-Cortol | 242.7 (193.3, 388.2) | 229.6 (168.8, 376.3) | 0.68 | 435.5 (327.7, 598.9) | <0.001 | |
| a-Cortolone | 1319.1 (936.1, 1839.6) | 1177.2 (853.4, 1839.2) | 0.53 | 1487.9 (1065.8, 1805.9) | 0.2 | |
| B-Cortolone | 541.3 (422.6, 756.0) | 450.5 (307.4, 658.8) | 0.12 | 852.5 (530.6, 1001.5 | <0.001 | |
| 11B-OH-AN | 557.9 (440.3, 844.4) | 443.2 (330.3, 815.3) | 0.17 | 984.1 (675.1, 1267.7) | <0.001 | |
| 11B-OH-ET | 472.0 (246.2, 670.7) | 382.7 (246.7, 559.7) | 0.33 | 497.3 (308.7, 799.1) | 0.18 | |
| 11-OXO-ET | 560.4 (389.1, 712.2) | 517.6 (307.7, 748.0) | 0.66 | 771.8 (566.0, 1065.9) | 0.0001 | |
Age-, sex- and menopausal status-based Z scores for steroid metabolite analysis were calculated from urine testing of 114 controls (40 premenopausal women, 26 postmenopausal women, 48 men). IQR, interquartile range. a P values between premenopausal and postmenopausal women.
b P values between men and women.
Steroid profiling demonstrated significant differ- ences in patients with ACC when compared with patients with ACAs for 11 steroids, most notably in THS (median Z score of 14 for ACC vs 0.5 for ACA, P = 0.0006), PT (median Z score of 7.5 vs -0.4, P = 0.001), 5PT (me- dian Z score 5.4 vs -0.4, P = 0.01), and Etio (median Z score 5.4 vs -1.1, P = 0.001) (Fig. 2A; also see Table 3 in the online Data Supplement). PD, 5PD, DHEA, and 17HP also showed significant distinction from ACA with Z scores ranging from 1.1 to 3.3.
Patients with adrenocorticotropic hormone-dependent pituitary hypercortisolism (Cushing disease) had in- creased androgens (An, Etio, DHEA, 16a-DHEA, PT, and 5PD) and glucocorticoid metabolites (cortisol, 6B- OH-cortisol, THF, 5a-THF, B-cortol, 11B-OH-AN, 11B-OH-ET, cortisone, THE, a-cortolone, B-cortolone, and 11-OXO-ET), whereas patients with cortisol- producing adrenal adenomas had suppressed androgens and increased glucocorticoids, although to a smaller de- gree than Cushing disease patients (Fig. 2B; see also Ta- ble 4 in the online Data Supplement). The difference in 3 analytes met statistical significance: Etio, 11B-OH- AN, and a-cortolone.
The patient with confirmed 21-hydroxylase deficiency (see Table 5 in the online Data Supplement for clinical and genetic details) showed the expected (26) accumulation of metabolites upstream of the enzyme block, including 5PD, PD, 5PT, PT, 17HP, PTONE, DHEA, 16a-DHEA, An, Etio, 11B-OH-AN, 11B-OH-ET, and 11B-OXO (Fig. 3), with the most substantial increase being observed in PTONE (Z score 27.9). PT, DHEA, 17HP, and 11B- OH-AN were also substantially increased with Z scores of 16.8, 13.9, 12.9, and 7.1, respectively. Furthermore, metab- olites found downstream of the defect were decreased, with Z scores ranging from -0.2 to -1.5. The 24-h urine sam- ple for steroid analysis was collected before the patient re- ceived the first glucocorticoid dose.
Discussion
We have developed an HRAM LC-MS urinary 26- steroid quantitative assay and have illustrated its success- ful application in a selection of disorders affecting steroidogenesis.
Our HRAM LC-MS method demonstrated good linearity and intraassay and interassay imprecision de- spite the quantification of multiple analytes in a single injection. However, some analytes did show inconsistent recoveries, namely, DHEA, 5PT, and 5PD. We attribute the poor recovery/linearity to loss during hydrolysis and extraction because matrix-free, deconjugated standards were prone to loss as well. We improved this to some extent by the addition of 0.1% estriol and 0.1% bovine serum albumin before analysis. Interestingly, recovery was better for DHEA, 5PT, and 5PD in a sample with
exceedingly high endogenous steroids (see Table 6 in the online Data Supplement). To have precise and adequate recoveries, improvements to the procedure should be made before implementation in clinical practice and re- coveries with conjugated steroids should be performed. Another caution will be to confirm complete enzymatic hydrolysis because different conjugated steroid metabo- lites may be affected by the background matrix of the samples, including different preservatives. Understand- ing that there could be a range of responses to hydrolysis in the steroids, we optimized the hydrolysis conditions to a point by varying the volume of Glusulase (≥85000 U/mL ß-glucuronidase) added to each sample. A mini- mum of 10 µL was determined to be sufficient at achiev- ing complete hydrolysis. To verify complete hydrolysis, we scanned for masses indicating glucuronidated or sul- fated steroids in full-scan data (extended mass range) in a selection of hydrolyzed urine samples, but no ions indi- cating incomplete hydrolysis were found. An alternative would have been to forego hydrolysis and scan for, and quantify, both unconjugated and conjugated steroids by HRAM (27, 28). We chose not to use this approach because it has its own limitations, namely, (a) variable glucuronide stability in biological matrices (29), (b) the preference of some conjugates for negative ionization, which might compromise measurement of analytes pre- ferring positive ionization, such as many unconjugated steroid metabolites, and (c) the increased complexity of the generated spectra, which might make data analysis more challenging. Additionally, retention time might be affected by conjugates, which can further complicate analysis (28).
With regard to its clinical performance, establishing sex- and age-stratified control population reference inter- vals should help in the validation of our assay for the diagnosis of a variety of adrenal disorders.
Our assay confirmed the importance of steroid metabolite profiling in the ACA vs ACC noninvasive diagnosis, as previously shown using GC-MS urine steroid profiles (11, 12, 26, 30). In our sampling of ACCs, we found Etio, DHEA, 5PT, 5PD, PD, 17HP, PT, and THS to be the strongest indicators of ACC, with THS being the most critical (Fig. 2A; also see Table 3 in the online Data Supplement). Because urine collection is far less invasive, costly, or traumatic to the patient than repeated imaging or adrenal biopsy, the urinary steroid metabolite panel has great diagnos- tic potential. Indeed, given that imaging and autopsy studies put the prevalence of incidental adrenal tumors at between 1% and 9% of the population (31), with most of these tumors being benign ACA, the noninva- sive differential diagnosis of ACA vs ACC is clearly an important future application of our HRAM LC-MS urine steroid panel.
| Fig. 2. Adrenal (A) using | Z score (not to scale) | Z score (not to scale) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CS | à " | O 8 | 30 40 | 60 | 70 is | · 5 | 10 | 15 | 20 | 25 | 30 | 60 | 120 90 | 150 330 300 | |
| versus | Clinical HRAM | An | An | A | |||||||||||
| pituitary | significance LC-MS | Etio* - | B | Etio* | |||||||||||
| DHEA. | DHEA* | ||||||||||||||
| CS (B) | of steroid | 16a-DHEA | 16a-DHEA | ||||||||||||
| steroid profiling | SPT ... | SPT* | |||||||||||||
| showed | SPD | SPD* | |||||||||||||
| 3 | metabolite | 5a-THA | Sa-THA | ||||||||||||
| analytes | established | THB | THB | ||||||||||||
| with | THDOC | ||||||||||||||
| THDOC | |||||||||||||||
| profiling 11 of | PD | PD* | |||||||||||||
| statistical | in 26 | 17HP | 17HP* | ||||||||||||
| adrenal | PT | PT* | |||||||||||||
| significance | metabolites | PTONE ... | PTONE | ||||||||||||
| disease as | THS | THS* | - | ||||||||||||
| (*). See | Cortisol | Cortisol* | |||||||||||||
| Tables | states. statistically | 6BOHcortisol | 6BOHcortisol* | . .. .. | |||||||||||
| 3 and | Comparison significant | THF Sa-THF | THE Sa-THF | ACA ACC | |||||||||||
| of 4 in | B-Cortol | Adrenal | Pituitary | B-Cortol | |||||||||||
| the online | Z score (*) in | 11B-OH-An* 11-OH-Et | Cushing | Cushing | 11B-OH-An 11-OH-Et | ||||||||||
| Data | (y axis) distinguishing | Cortisone | Cortisone* | ... | |||||||||||
| THE | THE | ||||||||||||||
| Supplement | between | a-Cortolone* | syndrome | syndrome | a-Cortolone | ||||||||||
| for | ACC ACC | B-Cortolone | B-Cortolone | ||||||||||||
| details. | and from | 11-OXO-Et | 11-OXO-EL | .. ... | |||||||||||
| ACA ACA. | |||||||||||||||
Cholesterol
CYP11A1
Pregnenolone
CYP17A1
17OH-Pregnenolone
CYP17A1
Dehydroepiandrosterone
SULT2A
DHEA-S
5PD
5PT
HSD3B2
HSD3B2
HSD3B2
DHEA
16a-DHEA, An
Etio
Progesterone
CYP17A1
17OH-Progesterone
CYP17A1
Androstenedione
CYP19A1
Estrone
PD
PT, 17HP
11B-OH-AN
CYP21A2
CYP11B1
HSD17B
CYP21A2
An, 11-OXO-ET
11-Deoxycorticosterone
21-Deoxycortisol
PTONE
Testosterone
Etio, 11B-OH-ET
CYP19A1
CYP11B2
THDOC
11-Deoxycortisol
SRD5A2
THS
An
Etio
Estradiol
Corticosterone
CYP11B1
Dihydrotestosterone
CYP11B2
THB
Cortisol
THE
5a-THF
An
Z score
>25
1SOH-Corticosterone
HSD11B1
HSD11B2
Z score above reference interval
6-25
CYP11B2
6B-OH- cortisol
included in HRAM metabolite panel
Cortisone
B-cortol
2-6
Aldosterone
enzyme
= Normal function
THE
0-2
defect
= Impaired function
0 to -1
Z score below reference interval
a-cortolone, B-cortolone
-1 to-2
Data not collected
Our steroid panel also allowed us to compare adre- nocorticotropic hormone-dependent pituitary hypercor- tisolism and cortisol-producing adrenal adenomas. Dis- tinguishing adrenocorticotropic hormone-dependent hypercortisolism from primary adrenal hypercortisolism remains one of the biggest challenges in the workup of CS. The results of the panel were promising in this re- gard; despite a sample size of only 4 patients in each group, we observed substantial differences in Z scores for multiple steroids, with 3 analytes showing statisti- cally significant differences (Etio, 11B-OH-AN, and a-cortolone; Fig. 2B and also see Table 4 in the online Data Supplement).
The use of steroid profiling in congenital adrenal hyperplasia as a result of 21-hydroxylase (CYP21A2)6 deficiency (4, 32) allowed for detailed characterization of
adrenal steroidogenesis and showed the predicted re- sponse for all metabolites, both upstream and down- stream from the enzyme blockade (Fig. 3). Moreover, in this single case, several metabolites, which are rarely, if ever, measured in CYP21A2 deficiency, displayed equal or greater changes from the normal state than the stalwart measurement targets 17-hydroxyprogesterone and an- drostenedione, suggesting that the profile might be useful for early diagnosis of subtle cases (e.g., nonclassical con- genital hyperplasia).
There are some limitations to our study. Most im- portantly, the cohort of patients with various diseases, adrenal and other, will need to be extended.
With regard to the day-to-day operations, the assay run-time is relatively long. In part, this problem can be over- come by chromatography multiplexing, which is not possi- ble with the instrumentation available to us at the time, but can theoretically be achieved with compatible multiplex liq- uid chromatography front-ends. Other operational issues include maintaining QC for 26 separate analytes.
6 Human Gene: CYP21A2, cytochrome P450 family 21 subfamily A member 2, gene en- coding 21-hydroxylase enzyme.
Novel HRAM LC-MS Method for Adrenal Diseases
A limitation of steroid profiling is the large amount of data yielded. In a typical batch, including calibrators, controls, and approximately 50 patients, there are ap- proximately 2000 chromatograms to quantify and con- firm. We found the TraceFinder software to be superior at handling this amount of information compared with any of our triple quadrupole quantification programs. With our setup, retention times were reproducible, re- sulting in accurate computer-programmed peak selection and quantification with minimal user intervention. Fur- thermore, we assigned flagging rules for each analyte, including confirmatory ions (secondary ions), when available, and isotopic pattern scoring to further enhance analytical specificity.
Finally, with a panel of this size, creating an optimal reporting system may be challenging. Although itemized analyte results based on Z score could be a simple solu- tion, a clear, comprehensive interpretation is probably required. Machine-learning (11) and heat-map analysis may offer some remedy to this. However, at this stage we remain unsure about how our final report to referring physicians will be structured.
In conclusion, we have developed a novel assay to quantify 26 steroid metabolites in urine using HRAM LC-MS. We have established control reference intervals and have demonstrated statistically significant differ- ences in urinary steroid measurements in patients with ACC, adrenal and pituitary CS, and congenital adrenal hyperplasia. Although promising, we propose a larger
validation in various adrenal diseases, which might ulti- mately allow this method to become a standard diagnos- tic tool for many adrenal diseases.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intel- lectual content; and (c) final approval of the published article.
Authors’ Disclosures or Potential Conflicts of Interest: Upon man- uscript submission, all authors completed the author disclosure form. Dis- closures and/or potential conflicts of interest:
Employment or Leadership: S.K. Grebe, Mayo Clinic. Consultant or Advisory Role: None declared.
Stock Ownership: None declared.
Honoraria: None declared.
Research Funding: I. Bancos, the Robert and Elizabeth Strickland Career Development Award, the Center of Individualized Medicine at Mayo Clinic; W.F. Young, the Center of Individualized Medicine at Mayo Clinic; R.J. Singh, the Center of Individualized Medicine at Mayo Clinic.
Expert Testimony: None declared.
Patents: None declared.
Role of Sponsor: The funding organizations played a direct role in the design of study, review and interpretation of data, and final approval of manuscript.
Acknowledgments: The authors thank Ann Rivard and Robert Taylor for method consultation, the Metabolomics Core Laboratory for HRAM-LC-MS use, and the Cardiovascular and Renal Laboratories for providing control samples.
References
1. Else T, Kim AC, Sabolch A, Raymond VM, Kandathil A, Caoili EM, et al. Adrenocortical carcinoma. Endocr Rev 2014;35:282-326.
2. Nieman LK, Biller BM, Findling JW, Murad MH, Newell- Price J, Savage MO, et al. Treatment of Cushing’s syndrome: an Endocrine Society clinical practice guide- line. J Clin Endocrinol Metab 2015;100:2807-31.
3. Nieman LK. Cushing’s syndrome: update on signs, symptoms and biochemical screening. Eur J Endocrinol 2015;173:M33-8.
4. Han TS, Walker BR, Arlt W, Ross RJ. Treatment and health outcomes in adults with congenital adrenal hy- perplasia. Nat Rev Endocrinol 2014;10:115-24.
5. Nelson RE, Grebe SK, O’Kane DJ, Singh RJ. Liquid chromatography-tandem mass spectrometry assay for simultaneous measurement of estradiol and estrone in human plasma. Clin Chem 2004;50:373-84.
6. Taylor RL, Machacek D, Singh RJ. Validation of a high- throughput liquid chromatography-tandem mass spec- trometry method for urinary cortisol and cortisone. Clin Chem 2002;48:1511-9.
7. Lacey JM, Minutti CZ, Magera MJ, Tauscher AL, Casetta B, McCann M, et al. Improved specificity of newborn screening for congenital adrenal hyperplasia by second-tier steroid profiling using tandem mass spec- trometry. Clin Chem 2004;50:621-5.
8. Van Der Gugten JG, Holmes DT. Quantitation of aldoste- rone in serum or plasma using liquid chromatography- tandem mass spectrometry (LC-MS/MS). Methods Mol Biol 2016;1378:37-46.
9. Krone N, Hughes BA, Lavery GG, Stewart PM, Arlt W, Shackleton CH. Gas chromatography/mass spectrome- try (GC/MS) remains a pre-eminent discovery tool in clinical steroid investigations even in the era of fast liq- uid chromatography tandem mass spectrometry (LC/ MS/MS). J Steroid Biochem Mol Biol 2010;121:496- 504.
10. Raff H, Auchus RJ, Findling JW, Nieman LK. Urine free cortisol in the diagnosis of Cushing’s syndrome: is it worth doing and, if so, how? J Clin Endocrinol Metab 2015;100:395-7.
11. Arlt W, Biehl M, Taylor AE, Hahner S, Libe R, Hughes BA, et al. Urine steroid metabolomics as a biomarker tool for detecting malignancy in adrenal tumors. J Clin En- docrinol Metab 2011;96:3775-84.
12. Kerkhofs TM, Kerstens MN, Kema IP, Willems TP, Haak HR. Diagnostic value of urinary steroid profiling in the evaluation of adrenal tumors. Horm Cancer 2015;6: 168-75.
13. Velikanova LI, Shafigullina ZR, Lisitsin AA, Vorokho- bina NV, Grigoryan K, Kukhianidze EA, et al. Differ- ent types of urinary steroid profiling obtained by high-performance liquid chromatography and gas chromatography-mass spectrometry in patients with adrenocortical carcinoma. Horm Cancer 2016;7: 327-35.
14. Grund B, Marvin L, Rochat B. Quantitative performance of a quadrupole-orbitrap-MS in targeted LC-MS deter- minations of small molecules. J Pharm Biomed Anal 2016;124:48-56.
15. Hines J, Milosevic D, Ketha H, Taylor R, Algeciras- Schimnich A, Grebe SK, et al. Detection of IGF-1 protein variants by use of LC-MS with high-resolution accurate mass in routine clinical analysis. Clin Chem 2015;61: 990-1.
16. Bystrom CE, Sheng SJ, Clarke NJ. Narrow mass extrac- tion of time-of-flight data for quantitative analysis of proteins: determination of insulin-like growth factor-1. Anal Chem 2011;83:9005-10.
17. Ramanathan R, Jemal M, Ramagiri S, Xia YQ, Hum- preys WG, Olah T, et al. It is time for a paradigm shift in drug discovery bioanalysis: from SRM to HRMS. J Mass Spectrom 2011;46:595-601.
18. Strano-Rossi S, Castrignano E, Anzillotti L, Odoardi S, De-Giorgio F, Bermejo A, et al. Screening for exogenous androgen anabolic steroids in human hair by liquid chromatography/orbitrap-high resolution mass spec- trometry. Anal Chim Acta 2013;793:61-71.
19. Tudela E, Deventer K, Geldof L, Van Eenoo P. Urinary detection of conjugated and unconjugated anabolic steroids by dilute-and-shoot liquid chromatography- high resolution mass spectrometry. Drug Test Anal 2015;7:95-108.
20. Diao XX, Wohlfarth A, Pang SK, Scheidweiler KB, Huestis MA. High-resolution mass spectrometry for characterizing the metabolism of synthetic cannabi- noid THJ-018 and its 5-fluoro analog THJ-2201 after incubation in human hepatocytes. Clin Chem 2016; 62:157-69.
21. Gorgens C, Guddat S, Thomas A, Wachsmuth P, Orlo-
vius AK, Sigmund G, et al. Simplifying and expanding analytical capabilities for various classes of doping agents by means of direct urine injection high perfor- mance liquid chromatography high resolution/high ac- curacy mass spectrometry. J Pharm Biomed Anal 2016; 131:482-96.
22. Odoardi S, Castrignano E, Martello S, Chiarotti M, Strano-Rossi S. Determination of anabolic agents in di- etary supplements by liquid chromatography-high- resolution mass spectrometry. Food Addit Contam Part A Chem Anal Control Expo Risk Assess 2015;32: 635-47.
23. Lagojda A, Kuehne D, Krug O, Thomas A, Wigger T, Karst U, et al. Identification of selected in vitro gener- ated phase-I metabolites of the steroidal selective an- drogen receptor modulator MK-0773 for doping con- trol purposes. Eur J Mass Spectrom (Chichester) 2016; 22:49-59.
24. Raro M, Ibanez M, Gil R, Fabregat A, Tudela E, Deventer
K, et al. Untargeted metabolomics in doping control: detection of new markers of testosterone misuse by ul- trahigh performance liquid chromatography coupled to high-resolution mass spectrometry. Anal Chem 2015;87:8373-80.
25. Clarke A, Scarth J, Teale P, Pearce C, Hillyer L. The use of in vitro technologies and high-resolution/ accurate-mass LC-MS to screen for metabolites of ‘designer’ steroids in the equine. Drug Test Anal 2011;3:74-87.
26. Vitkin E, Ben-Dor A, Shmoish M, Hartmann MF, Yakhini Z, Wudy SA, et al. Peer group normalization and urine to blood context in steroid metabolomics: the case of CAH and obesity. Steroids 2014;88:83-9.
27. Jantti SE, Hartonen M, Hilvo M, Nygren H, Hyotylainen T, Ketola RA, et al. Steroid and steroid glucuronide pro- files in urine during pregnancy determined by liquid chromatography-electrospray ionization-tandem mass spectrometry. Anal Chim Acta 2013;802:56-66.
28. Ikegawa S, Hasegawa M, Okihara R, Shimidzu C, Chiba H, Iida T, et al. Simultaneous determination of twelve tetrahydrocorticosteroid glucuronides in human urine by liquid chromatography/electrospray ionization- linear ion trap mass spectrometry. Anal Chem 2009; 81:10124-35.
29. Trontel J. Quantification of glucuronide metabolites in biological matrices by LC-MS/MS, tandem mass spec- trometry-applications and principles. InTech 2012; 794.
30. Fukushima DK, Bradlow HL, Hellman L, Gallagher TF. Origin of pregnanetriol in a patient with adrenal carci- noma. J Clin Endocrinol Metab 1962;22:765-72.
31. Terzolo M, Stigliano A, Chiodini I, Loli P, Furlani L, Ar- naldi G, et al. AME position statement on adrenal inci- dentaloma. Eur J Endocrinol 2011;164:851-70.
32. Parsa AA, New MI. Steroid 21-hydroxylase deficiency in congenital adrenal hyperplasia. J Steroid Biochem Mol Biol 2017;165:2-11.