ELSEVIER
Journal of Pharmaceutical and Biomedical Analysis
journal homepage: www.journals.elsevier.com/journal-of-pharmaceutical-and-biomedical-analysis
A
JPBA Pharmaceutical and Biomedical Analysis
A quantitative, broad-panel LC-MS/MS method for the analysis of intact steroid conjugates: A novel approach to steroid profiling for biomarker research in corticoid-dependent diseases
Eleanor North ªD, Arne Gessner a,b, Max Kurlbaum c,d, Martin Fassnacht c,d, Matthias Kroiss c,e, Martin F. Fromm a,b, Nora Bartels ª a, İD
ª Institute of Experimental and Clinical Pharmacology and Toxicology, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
b FAU NeW - Research Center New Bioactive Compounds, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
· Department of Internal Medicine I, Division of Endocrinology and Diabetes, University Hospital, University of Würzburg, Würzburg, Germany
d Central Laboratory, Core Unit Clinical Mass Spectrometry, University Hospital Würzburg, Würzburg, Germany
e Department of Internal Medicine IV, University Hospital, Ludwig-Maximilians-Universität München, Munich, Germany
Check for updates
ARTICLE INFO
Keywords: Steroid conjugates Adrenocortical adenoma Adrenocortical carcinoma Method validation Mass spectrometry
ABSTRACT
In endocrine oncology changes in hormone synthesis and metabolism play a major role in disease formation and progression. Accordingly, steroid metabolism is a possible pathophysiological factor in adrenal tumours and metabolic conjugates of steroids are promising biomarker candidates for the clinically challenging differential diagnosis of highly prevalent, benign adrenocortical adenomas versus rare, aggressive adrenocortical carci- nomas. The analysis of steroidal compounds in human biological samples is most accurately achieved by mass spectrometry-based approaches. Commonly, metabolic end products of steroids, such as sulfate and glucuronide conjugates, are hydrolysed before measurements to simplify the procedure. Therefore, data on actual steroid conjugate concentrations in humans is scarce. In this work we present a compact LC-MS/MS method for the absolute quantification of 22 intact steroid conjugates in urine and plasma with an instrument run time of 11 min and a straightforward sample preparation procedure. The method was successfully validated according to established international guidelines and applied to samples of patients with adrenal tumours to demonstrate the method’s suitability regarding sample preparation, analyte selection and quantifiable concentration ranges. Furthermore, the exploratory comparison of steroid abundance in the patient samples support the previously proposed potential of steroid conjugates as differentiating biomarkers between adrenocortical carcinoma and adenoma (e.g., 17-OH-pregnenolone sulfate, median plasma concentration in carcinoma (114 ng/ml) versus adenoma (3.76 ng/ml), p <0.001). The newly developed method enables absolute quantification of multiple steroid conjugates in humans which provides the basis for profound biomarker research for the early detection of adrenocortical carcinomas and valuable data for related research questions.
1. Introduction
Steroid profiling in human urine and plasma is commonly utilised in the diagnosis, estimation of prognosis and monitoring of various human conditions e.g., endocrine diseases affecting the adrenal gland, steroid- dependent cancer or doping status [1-3]. Steroids are synthesized in the adrenal cortex or in the peripheral reproductive organs. Concerning metabolism, they are conjugated to sulfate or glucuronide moieties and mainly excreted by the kidneys, but also by the liver into bile or in saliva
and sweat [4,5]. In the past, steroid conjugates were considered only as metabolic end products since they have a relatively low activity at nu- clear steroid receptors. However, steroid sulfates can be hydrolysed by the ubiquitously expressed steroid sulfatase, in order for the free steroid to have an effect [6].
The pathophysiological role of steroid metabolism in adrenal tu- mours has been a much-discussed topic in recent literature [7] and two recent studies demonstrated the potential to utilize intact steroid con- jugates as prognostic and diagnostic biomarkers for adrenocortical
* Correspondence to: Institute of Experimental and Clinical Pharmacology and Toxicology, Friedrich-Alexander-Universität Erlangen-Nürnberg, Fahrstr. 17, Erlangen 91054, Germany.
E-mail address: nora.vogg@fau.de (N. Bartels).
https://doi.org/10.1016/j.jpba.2025.117258
carcinoma (ACC) [8,9]. Previously, Bancos et al. had notably presented the impact urine steroid profiling could have in improving the so far clinically highly demanding differential diagnosis of adrenal neoplasms [10].
For quantitative, routine high-through-put analysis of steroids the use of liquid chromatography coupled with a triple quadrupole mass spectrometer is most established [11]. Commonly, targeted urinary steroid profiling LC-MS/MS methods quantify free unconjugated ste- roids or total steroids after deconjugation [10,12-14]. Therefore, during sample preparation the conjugated moiety is removed from the molecule by enzymatic hydrolysis or less often by chemical solvolysis. Apart from the crucial fact, that all information about the originally conjugated form of the steroid is lost, concerns have been brought up lately, as to the completeness and the inter-laboratory reproducibility of enzymatic hy- drolysis using the common enzyme preparation derived from Helix pomatia containing arylsulfatase and beta-glucuronidase [9,15,16].
Profound research on the physiological and pathophysiological role of steroid conjugates could be enhanced with extensive data on absolute concentrations of metabolic conjugates of various steroids in human urine and plasma and corresponding reference values [6,17]. So far, this data is largely not available, mainly due to the above-mentioned cir- cumstances of steroid analysis and the poor commercial availability of reference standards for steroid conjugates. Furthermore, the promising biomarker potential of steroid conjugates for adrenal tumours is yet to be fully evaluated. Reliable biomarkers could refine the challenging differential diagnosis of ACC and adrenocortical adenoma (ACA) regarding time and effort for patients and medical staff, markedly improving the poor prognosis of ACC and protecting patients with ACA from unnecessary burden [13,14,18-20].
In this work we have developed a targeted, highly sensitive and quantitative LC-MS/MS method, suitable for plasma and urine samples, for a broad panel and a wide concentration range of 22 steroid conju- gates i.e., steroid sulfates, disulfates and glucuronides. The method was validated in both matrices according to renowned international guide- lines and its fit-for-purpose proof is demonstrated by the analysis of a number of real-world samples from patients with adrenal tumours.
Using this method, extensive data on steroid conjugate concentra- tions in humans can be acquired to help shed light on their possible physiological and pathophysiological properties and further evaluate their potential as biomarkers for steroid-dependent endocrine diseases.
2. Materials and methods
2.1. Instrumentation
The LC-MS/MS analysis was performed on a Shimadzu Nexera CL UHPLC coupled to a QTRAP 6500+ triple quadrupole mass spectrometer (AB SCIEX Ptc. Ltd., Darmstadt, Germany), which was equipped with the Turbo-V ion source assembled for electrospray-ionisation. Instrument control and data analysis were navigated using the SCIEX OS Software (AB SCIEX Ptc. Ltd.).
2.2. Materials
The analytical separation was achieved on an Acquity Premier CSH Phenyl-Hexyl 1,7 um 2.1 x 100 mm column, which was attached to the corresponding VanGuard FIT pre-column cartridge (Waters GmbH, Eschborn, Germany). LC-MS grade acetonitrile, methanol and ammo- nium formate were purchased from VWR International GmbH (Darm- stadt, Germany), while LC-MS grade water was purchased from Sigma- Aldrich Chemie GmbH (Taufkirchen, Germany), as were Sigmatrix urine diluent and activated charcoal. Creatinine-free synthetic urine was supplied by synthetic urine (Eberdingen-Nussdorf, Germany) and the gender pooled human plasma with K3EDTA by BioIVT (West Sussex, United Kingdom). 11-ketoetiocholanolone glucuronide (11-ketoetio-G), 11-ketoetiocholanolone sulfate (11-ketoetio-S), 16-
hydroxydehydroepiandrosterone sulfate (16OH-DHEAS), 17,20-dihy- droxyprogesterone sulfate (diOH-prog-S), 17-hydroxypregnenolone sulfate (17OH-preg-S), 17-hydroxyprogesterone sulfate (17OH-prog-S), androstenediol disulfate (andiol-diS), androstenediol sulfate (andiol-S), androsterone sulfate (an-S), dehydroepiandrosterone glucuronide (DHEA-G), dihydrotestosterone glucuronide (DHT-G), etiocholanolone sulfate (etio-S), pregnenolone glucuronide (preg-G), testosterone glucuronide (testo-G) and testosterone sulfate (testo-S) were obtained from Steraloids (Newport, RI, USA). Androsterone glucuronide (an-G), cortisol sulfate d4 (cortisol-S d4), dehydroepiandrosterone sulfate (DHEAS), estradiol sulfate (E2S), estradiol sulfate d4 (E2S d4), preg- nenolone sulfate (preg-S) and testosterone glucuronide d3 (testo-G d3) were supplied by Biomol GmbH (Hamburg, Germany). Androsterone glucuronide d4 (an-G d4), cortisol sulfate (cortisol-S), creatinine, DHEAS d6, estradiol disulfate (E2S2), estrone sulfate (E1S), pregnandiol glucuronide 13C5 (PD-G 13C5) and pregnenolone sulfate 13C2 d2 (preg-S 13C2 d2) were purchased from Sigma-Aldrich Chemie GmbH, E1S d4 from Hölzel (Cologne, Germany) and creatinine d3 from C/D/N isotopes (Quebec, Canada). All reference standards and internal standards had purity specifications of at least 95 % confirmed by a certificate of anal- ysis. Anionic compounds were provided as sodium or disodium salt. The selection of relevant analytes was based on results of our previous untargeted metabolomics study (Figure S2A and S2B) [9].
2.3. Preparation of steroid-stripped human plasma
Endogenous steroid conjugates had to be removed from the commercially obtained pooled human plasma for it to be suitable as surrogate matrix for calibration and quality control samples. Therefore, approximately 1 g of activated charcoal was added to 25 ml of plasma and placed into an overhead mixer for 20 min. Then the mixture was centrifuged for 20 min at 20.000 rpm. The supernatant was decanted, and the procedure was repeated three times.
2.4. Preparation of standard stock solutions
Stock solutions of all analytes and internal standards were prepared in 50 % (V/V) methanol/water at a concentration of 1.0 mg/ml except for cortisol-S and creatinine d3, which were prepared in water at a concentration of 1.0 mg/ml, 17OH-prog-S, which was prepared in methanol at a concentration of 1.0 mg/ml, and creatinine, which was prepared in water at a concentration of 100.0 mg/ml. Testo-G d3 was purchased as a 0.1 mg/ml solution in methanol. All stock solutions were stored in glass vials at - 80℃.
2.5. Sample preparation
Internal standards (IS) were spiked into methanol at a concentration of 20 ng/ml (cortisol-S d4, DHEAS d6, E2S d4, E1S d4, preg-S 13C2 d2), 100 ng/ml (an-G d4, PD-G 13C5, testo-G d3) or 5 µg/ml (creatinine d3). 300 ul of methanol containing the IS were added to 100 ul of sample, vortexed vigorously and then centrifuged for 10 min at 4 ℃ and 16.000 rpm. 280 ul of the supernatant was evaporated under a stream of nitrogen and the residue was reconstituted in 70 ul of eluent composed of the gradient starting conditions.
For the preparation of calibration samples and quality control (QC) samples working solutions with specific concentrations of all reference standards were prepared in water and spiked into steroid-free synthetic urine or steroid-stripped pooled human plasma. These calibration sam- ples were then prepared as described above. The quantification of creatinine is performed within the same run and sample injection and the creatinine calibration samples were prepared accordingly, but separately, in creatinine-free synthetic urine.
2.6. LC conditions
The run time of the LC method was set to 11 min to execute the gradient (Table S1 and Figure S1) at a flow rate of 0.25 ml/min. The mobile phase A consisted of 25 mM ammonium formate in water, while the mobile phase B consisted of acetonitrile. The column oven temper- ature was set to 35 ℃, the sample injection volume was 5 ul and samples were cooled in the autosampler at 4 ℃.
2.7. MS parameters
Two MRM transitions were monitored for each analyte, of which one was used for quantification and the other helped to confirm the analyte identity (qualifier ion). The precursor and fragment ions, as well as the values for declustering potential, collision energy and dwell time are listed in Table S2. Source parameters were set as follows: ion source gas 1: 40 psi; ion source gas 2: 70 psi; curtain gas: 30 psi; CAD gas: 10; source temperature: 600 ℃; spray voltage: 4500 V (5500 V for creatinine).
2.8. Method validation
The method was validated for urine and plasma matrix according to the EMA ICH guideline M10 on bioanalytical method validation and study sample analysis and the FDA guideline for bioanalytical methods [21,22].
2.8.1. Pooled real-world sample
Whenever applicable, validation steps were performed on a pooled real-world sample of five urine and plasma samples of healthy volunteers, which was aliquoted for further analysis. The samples were collected as part of the KPE37 study for analytical method development and valida- tion, which received approval by the Ethics Committee of the Friedrich- Alexander-University Erlangen-Nuremberg (application # 310_18 B) and all volunteers provided written informed consent.
2.8.2. Calibration, sensitivity and carry-over
The linearity of the analyte concentration-response relationship was assessed by the coefficient of determination (R2) of a linear regression model with a weighting of either 1/x or 1/x2, which in every case should be ≥ 0.99 (Table S3 and S4). These criteria, as well as the calculated concentration of each non-zero calibrator being 85 - 115% of the nominal concentration (80 - 120 % for the lowest calibrator), should be met in each validation run.
The lower limit of quantification (LLOQ) was defined as the lowest calibration level which had a signal/noise ratio of at least ten. Furthermore, the area of the LLOQ had to be larger than the five-fold analyte response in the blank surrogate matrix. The limit of detection (LOD) was defined as an analyte response with a signal/noise ratio larger than three.
Carry-over was tested with solvent injections directly following in- jections of the highest calibrator. The analyte response in the solvent injections should not exceed 20 % of the area of the LLOQ.
2.8.3. Accuracy, precision and reinjection reproducibility
Accuracy was determined by the ratio of calculated concentration and nominal concentration, and the acceptance criteria was defined to be ± 15% of the nominal concentration (± 20 % for the LLOQ). Preci- sion and reinjection reproducibility were evaluated with the coefficient of variation (%CV) between results of experimental or analytical repli- cates and the values were required to be ≤ 15 (≤ 20 for the LLOQ).
Intra-day accuracy and precision were verified by preparing five experimental replicates of four QC levels (LLOQ, low, medium, high) and one pooled real-world sample on the same day and analysing them within one run. One replicate of each QC level and of the real-world sample was injected five times in a row to determine reinjection reproducibility. Inter-day accuracy and precision were evaluated by
preparing and analysing five experimental replicates of four QC levels (LLOQ, low, medium, high) on three different days.
2.8.4. Recovery and matrix effects
Recovery was evaluated by comparing the analyte response of QC samples spiked with the same amount of reference standards before and after sample preparation. This was performed for three different con- centrations per analyte (low, medium, high) and three experimental replicates per concentration level.
Matrix effects were determined by comparing the analyte response of reference standards in eluent with the analyte response in urine or plasma samples, which were respectively spiked with the same amount of reference standards after sample preparation (three different con- centrations per analyte and three experimental replicates per concen- tration level; low, medium, high).
2.8.5. Selectivity
Chromatograms of all analytes in the pooled real-world sample were reviewed regarding interfering peaks. The ratio of the quantifier and qualifier signal serves to confirm the analyte identity.
2.8.6. Dilution integrity
To evaluate the possibility of diluting highly concentrated samples in order to reach the method’s calibration range, accuracy and precision were tested for 1:2 and 1:10 dilution of a QC sample. Accordingly, the highest QC level was diluted with steroid-free synthetic urine or steroid- stripped pooled human plasma (five experimental replicates) before sample preparation, and the results were compared to a triplicate analysis of the original QC sample.
2.8.7. Stability
To assess analyte stability throughout the entire analytical procedure QC samples and a pooled real-world sample were exposed to various storage conditions. The analyte concentrations in the subjected samples were analysed immediately after preparation to generate baseline values for all stability experiments. Each of the following experiments were performed for three QC concentrations per analyte (low, medium, high) and three experimental replicates per QC level and on two replicates of the pooled real-world sample.
Benchtop stability was determined by storing freshly prepared QC samples and the pooled real-world sample on the laboratory benchtop for 24 h at room temperature before analysis.
Freeze thaw stability was tested by freezing freshly prepared QC samples and the pooled real-world sample at - 80 ℃ and thawing them the next day. This procedure was repeated twice before analysis.
Since processed samples are stored in the autosampler during the entire analytical run the stability of the analytes under these conditions was tested after 24 h and 72 h storage in the autosampler at 4 ℃.
Post-preparative stability was evaluated after one and two weeks of storage at - 20 ℃.
Long-term stability of analytes was determined by storing the QC samples and the pooled real-world sample at - 20 ℃ for two weeks, one month and three months.
2.8.8. Validation of the analyte creatinine
The creatinine concentration can be used for normalization of ana- lyte concentrations in urine. Since it serves merely as a reference value, it was considered sufficient to only validate this analyte concerning its linearity, accuracy, precision, carry-over, sensitivity and selectivity. The calibration and QC samples were prepared separately in creatinine-free synthetic urine.
2.9. Samples of patients with ACC and ACA
The patients took part in the European Network for the Study of Adrenal Tumors (ENSAT) registry, which was approved by the Ethics
Committee of the Julius-Maximilians University Würzburg (# 88/11). All patients provided written informed consent. In total, 24h-urine samples and plasma samples of 20 male patients (10 ACC vs. 10 ACA) were randomly chosen for this study. Tumour status was determined based on current clinical practice guidelines for the management of adrenal incidentalomas and ACC with post-operative histopathology and/or follow-up investigations as gold standards [18,23,24]. The Mann-Whitney-U-test was applied for the statistical analysis of differ- ences in analyte concentrations between the two patient groups using GraphPad Prism 10.4.0 (GraphPad Software, San Diego, CA, USA). In cases when the analyte concentration was below the LOD, we used 20 % of the LLOQ as replacement value for statistical calculations (urine: normalized to the mean creatinine concentration of all patients). If the analyte concentration was below the LLOQ, we used the calculated concentration for statistical calculations.
3. Results and discussion
This study reports the development of a targeted, highly-sensitive LC-MS/MS based method for the quantification of a large variety of steroid conjugates in human urine and plasma. Steroid conjugates have been a much-discussed topic in recent literature regarding their bio- logical functions beyond metabolism and excretion, such as potential pathophysiological properties for adrenal tumours [7]. Published data on physiological reference ranges of various steroid conjugates in humans are rare although this knowledge is necessary to evaluate pathophysiological mechanisms and according biomarker potential of steroid conjugates in corticoid-dependent diseases [25]. The LC-MS/MS method was successfully validated according to international guidelines and its suitability was demonstrated by the application to a number of samples from patients with adrenocortical carcinoma or adenoma.
3.1. LC-MS/MS method
The analyte panel of the quantitative LC-MS/MS based method includes 22 steroid conjugates. The phenyl-hexyl modified reversed- phase chromatographic column proved to be most suitable for the fast separation of the analytes using acetonitrile and water as mobile phase, complemented with 25 mM ammonium formate as buffer additive to improve peak shapes. The modification of the pH-value of the mobile phase A by addition of formic acid was tested but led to earlier retention times, which did not favour sufficient chromatographic separation. The fragment ions used for the sulfate conjugates were 96.9 m/z and 80.0 m/z (except for estrogen derivatives, Table S2) and for the glucuronide conjugates the fragment ions 113.0 m/z and 75.0 m/z were used. Since the fragment ions were always the same, adequate chromatographic separation of isobaric analytes was critical and was achieved in all cases (Fig. 1). Etiocholanolone glucuronide was excluded from the analyte panel before validation due to insufficient separation from its isomer an-G.
Regarding the MS parameters, extensive optimisation workflows were processed and evaluated. Compromises had to be found for ion source parameters to suit all analytes, while on the other hand the in- dividual analyte MS conditions could be specifically optimized and applied for each analyte (Table S2). The detection performance for steroid conjugates was best in negative ionisation mode, while creati- nine for normalization was analysed in positive ionisation mode.
The analyte panel consists of 14 steroid sulfates, two steroid disul- fates and six steroid glucuronides comprising androgen, estrogen and progesterone precursors and metabolites and can therefore give a good impression of the overall steroid metabolome. For research on steroid metabolism LC-MS/MS methods have been published targeting five [26] and eleven [27] steroid sulfates, 15 steroid glucuronides [28] or the simultaneous quantification of six steroid conjugates and eleven un- conjugated steroids [25], leaving the method presented here to be the first one to target steroid disulfates alongside the two more common
Chromatogram of all analytes
2.0e+07
3.14
1.8e+07
4.36
6.37
4.32
6.66
Andiol-diS: RT = 3.14
3.99
Andiol-S: RT = 4.36
1.6e+07
8.34
An-S: RT = 6.37
3.67
3.87
Etio-S: RT = 6.66
5.61
An-G: RT = 4.32
1.4e+07
3.74
DHT-G: RT = 3.87
Cortisol-S: RT = 3.99
Signal intensity (cps)
3.49
1.2e+07
4.44
DHEA-G: RT = 3.74
4.77
Testo-G: RT = 3.49
3.09
DHEAS: RT = 5.63
1.0e+07
4.58
5.43
Testo-S: RT = 4.77
5.59
16OH-DHEAS: RT = 3.67
11-ketoetio-S: RT = 4.58
8.0e+06
3.41
E2S2: RT = 3.09
4.32
E2S: RT = 4.44
5.24
E1S: RT = 5.43
6.0e+06
4.77
11-ketoetio-G: RT = 3.41
Preg-G: RT = 4.77
Preg-S: RT = 8.34
4.0e+06
17OH-preg-S: RT = 5.61
DiOH-prog-S: RT = 4.32
2.0e+06
17OH-prog-S: RT = 5.24
0.0e+00
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
Retention time (min)
conjugated forms in a larger and broader analyte panel. In the research area of doping control, the focus of analytical methods has also shifted to not only identifying free steroids, but their phase-II-metabolites as well [29-33]. However, within these methods the analyte panels very much concentrate on anabolic androgenic steroids, which are not compounds of interest for our current research objective [9].
Many analytical methods for steroidal compounds focus on the quantification of unconjugated and deconjugated steroids and there are several respective, published methods for the purpose of biomarker research in adrenal tumours [13,14,34,35]. This circumstance may partly be due to the poor commercial availability of reference standards for steroid conjugates. For this work, we were able to purchase reference material for several compounds that showed significant group differ- ences of urine concentrations in patients with ACC compared to ACA in our preliminary work [9] and managed to commercially obtain eight stable isotope-labelled steroid conjugates for the use as internal stan- dards, but many steroid disulfates in particular remained unavailable. Moreover, the direct analysis of steroid metabolites has an apparent advantage concerning accuracy, reproducibility and integrity of results, since the preceding hydrolysis step of conjugate moieties during sample preparation is not necessary when targeting intact conjugated steroids. Evidence in literature has raised concerns as to the completeness, reproducibility and accuracy of such hydrolysation procedures with enzyme preparations derived from H. pomatia, due to inconsistent experimental conditions in between laboratories and steroid conver- sions caused by enzymatic side-activities [9,15]. To the best of our knowledge, we are the first to provide a targeted, quantitative, analyt- ical methodology for profound biomarker research of intact steroid conjugates that may have particular utility for the study of adrenal tu- mours building on recently published results of our untargeted metab- olomics study [9].
Although the analyte panel includes more than 20 steroid conju- gates, various steroid metabolites, as well as unconjugated steroids, are not covered by the method. Since the steroid metabolome is manifold, it will be close to impossible to include all compounds and the limitation of commercial availability of reference standards for steroid conjugates remains ever-present.
3.2. Sample preparation
The sample preparation procedure was kept straightforward, essen- tially consisting of a protein precipitation step and a subsequent con- centration step by evaporation under a stream of nitrogen.
Keeping the instrument run time short and the sample preparation as straightforward as possible was a predefined objective for this study, since future implementation of similar analytical procedures into clin- ical routine may be encouraged by reducing time and financial efforts for medical and laboratory staff. Initially, a solid phase extraction (SPE) procedure was developed using weak anion exchange sorbents to ach- ieve effective separation of steroid conjugates from the matrix (Supplementary Excel sheet). After careful evaluation and comparison with the here-reported sample preparation procedure, the benefit of higher sample purification by SPE did not overcome the disadvantages of extensive time and financial expense. While other measures to ensure long-term preservation of materials and instrumentation can be employed, such as pre-columns and regular instrument maintenance, it should be acknowledged that more laborious sample preparation e.g., by solid phase or liquid-liquid extraction, has some advantages regarding sample purification and protection of delicate instrumentation, such as mass spectrometers, from contamination.
3.3. Method validation
Regarding analytical performance, high accuracy, precision, line- arity, recovery and analyte stability throughout the procedure could be achieved during validation, which overall fulfilled the criteria of
international guidelines from the EMA and the FDA [21,22].
3.3.1. Calibration, sensitivity and carry-over
Calibration ranges and LLOQs for all analytes were defined in order to meet the acceptance criteria keeping in mind their suitability for future method application to patient samples (Tables 1 and 2). Due to the great variation of calibration ranges between the analytes, nine non- zero calibrators were prepared for the purpose of calibration, however not every calibration level was necessarily within the validated cali- bration range of all analytes and therefore not applicable.
Carry-over was observed for 13 analytes in urine and nine analytes in plasma, however none exceeded the pre-specified limit of 20 % of the signal intensity of the analytes LLOQ. The highest rate of carry-over in urine was detected for testo-S at 16.1 % of the LLOQ area and in plasma for diOH-prog-S at 4.0 %. None of the glucuronide compounds were affected by carry-over.
3.3.2. Accuracy and precision
All 22 steroid conjugates met further acceptance criteria for inter- day and intra-day accuracy and precision in urine and plasma, with respective values of less than ± 15 % of the nominal concentration and ≤ 15 %CV (± 20% and ≤ 20 for the LLOQ) except for an-S in urine and DHEAS and testo-G in plasma, where one of the QCs of the intra-day accuracy was 115.5 %, 115.7 % and 123.2 %, respectively. (Tables 1 and 2).
3.3.3. Recovery and matrix effects
The determination of analyte recovery revealed values between 91.3 % (an-S) and 111.2 % (11-ketoetio-S) in urine and 81.8 % (an-G) and 101.6 % (E2S2) in plasma, all being reproducible with %CV ≤ 15 in urine and ≤ 10 in plasma between all replicates. Testo-G showed con- tradictory recovery rates in urine and plasma with a mean value of 128.5 % ± 10.4 versus 71.8 % ± 3.8 respectively (Tables 1 and 2).
Regarding matrix effects, a minor influence was found in urine with 17 of 22 analytes being affected by less than ± 5 % and the highest ion suppression value found at 14.0 % for E2S2. Conversely, plasma caused extensive ion enhancement, most eminently observed for the glucuronide conjugates with values up to 966.8 % for testo-G and a mean value among the six glucuronides of 544.6 % ± 224.3. Steroid sulfates and disulfates with the highest values of ion enhancement were 11-ketoetio-S (1035.4 %), cortisol-S (650.0 %) and testo-S (432.3 %) while E2S2 displayed minor ion suppression (2.1 %). Between the three QC levels matrix effects fluctuated quite considerably for various ana- lytes in plasma as 13 analytes presented %CVs above 15 % (< 31.8, DHEAS), however matrix effects in urine seemed to be consistent with all respective %CVs below seven (Tables 1 and 2).
The extent of matrix effects depends on several aspects, such as the complexity of the matrix composition, coelution of analyte and matrix components, individual analyte characteristics and sample dilution. Highest matrix effects in plasma were detected for the 14 analytes eluting during the time period 3.41 - 5.24 min, which implies that matrix components causing the strong ion enhancement elute mainly during this time period. The plasma of study participants will naturally vary regarding its exact composition, but the preparation of calibration samples in equivalent matrix, such as steroid-stripped pooled human plasma, will maximise the integrity of results. In addition, the selected internal standards (stable isotope-labelled steroid conjugates) have analogous molecular characteristics and retention times compared to the analyte panel and are therefore similarly affected by matrix effects. Despite the high matrix effects, the successful method validation ac- cording to EMA/FDA guidelines demonstrates adequate analytical integrity.
3.3.4. Selectivity
The ratio of quantifier and qualifier signal intensity was specified for each analyte using the values of the calibration samples. In the
| Analyte | Calibration Range [ng/ml] | Quality control | Intra-dayª | Inter-dayb | Reinjection™ Precision [%CV] | Recoveryª Mean [%] (%CV) | Matrix effecte Mean [%] (%CV) | |||
|---|---|---|---|---|---|---|---|---|---|---|
| Level | Concentration [ng/ml] | Accuracy [%] | Precision [%CV] | Accuracy [%] | Precision [%CV] | |||||
| Andiol-diS Andiol-S | 2.00 - 5000 | LLOQ | 2.00 | 98.7 | 8.67 | 103.4 | 6.65 | 6.85 | ||
| Low | 5.00 | 92.5 | 2.18 | 99.9 | 10.1 | 1.00 | 98.3 (3.36) | 96.7 (1.41) | ||
| Medium | 200 | 100.9 | 5.83 | 98.9 | 6.85 | 8.17 | 106.9 (2.62) | 91.6 (2.17) | ||
| High | 4000 | 98.3 | 1.58 | 99.9 | 3.75 | 2.84 | 106.2 (3.65) | 100.4 (2.21) | ||
| Pool-S | 4.42 | 6.41 | ||||||||
| 1.00 - 5000 | LLOQ | 1.00 | 98.0 | 6.07 | 95.1 | 9.72 | 5.87 | |||
| Low | 2.50 | 94.2 | 5.49 | 96.9 | 6.54 | 4.74 | 96.1 (4.76) | 105.0 (1.96) | ||
| Medium | 100 | 103.9 | 4.06 | 106.5 | 4.36 | 6.15 | 104.3 (1.36) | 91.9 (4.98) | ||
| High | 2000 | 101.0 | 2.05 | 96.9 | 5.94 | 3.56 | 107.6 (5.53) | 98.4 (4.05) | ||
| Pool-S | 6.20 | 3.15 | ||||||||
| An-G An-S Cortisol-S DHEA-G DHEAS | 10.0 - 20,000 | LLOQ | 10.0 | 98.6 | 3.99 | 102.3 | 3.75 | 4.77 | ||
| Low | 40.0 | 95.0 | 6.53 | 94.3 | 8.93 | 3.07 | 106.0 (22.2) | 102.0 (5.40) | ||
| Medium | 2000 | 106.5 | 4.41 | 109.4 | 5.96 | 3.57 | 102.3 (2.56) | 97.9 (1.56) | ||
| High | 8000 | 105.9 | 2.35 | 102.3 | 5.30 | 0.92 | 109.5 (3.22) | 103.0 (0.73) | ||
| Pool-S | 2.55 | 1.95 | ||||||||
| 2.00 - 10,000 | LLOQ | 2.00 | 118.6 | 8.27 | 111.3 | 4.98 | 4.91 | |||
| Low | 5.00 | 99.4 | 4.46 | 104.0 | 7.11 | 3.32 | 77.8 (2.59) | 100.0 (5.10) | ||
| Medium | 200 | 100.5 | 3.15 | 97.7 | 5.22 | 2.95 | 91.9 (0.82) | 106.0 (0.86) | ||
| High | 4000 | 115.5 | 4.35 | 109.4 | 4.64 | 2.20 | 104.2 (1.90) | 103.8 (1.11) | ||
| Pool-S | 3.55 | 1.96 | ||||||||
| 1.00 - 2000 | LLOQ | 1.00 | 99.2 | 12.3 | 101.8 | 10.7 | 5.61 | |||
| Low | 4.00 | 100.6 | 10.7 | 104.1 | 11.1 | 8.94 | 105.0 (17.4) | 88.1 (3.70) | ||
| Medium | 200 | 98.0 | 5.05 | 100.8 | 6.97 | 6.04 | 108.6 (5.20) | 91.7 (3.34) | ||
| High | 800 | 94.6 | 5.14 | 98.7 | 3.28 | 5.24 | 102.9 (4.30) | 97.4 (4.51) | ||
| Pool-S | 8.87 | 3.00 | ||||||||
| 2.50 - 5000 | LLOQ | 2.50 | 111.8 | 7.75 | 108.4 | 12.4 | 10.5 | |||
| Low | 10.0 | 106.0 | 4.53 | 101.4 | 6.40 | 8.31 | 101.5 (11.8) | 105.8 (11.7) | ||
| Medium | 500 | 96.9 | 6.06 | 99.8 | 7.16 | 3.97 | 111.8 (3.90) | 102.0 (0.74) | ||
| High | 2000 | 99.5 | 2.95 | 98.2 | 5.21 | 1.41 | 114.1 (5.77) | 104.5 (3.48) | ||
| 2.00 - 10,000 | LLOQ | 2.00 | 99.0 | 12.7 | 99.6 | 4.21 | 5.13 | |||
| Low | 5.00 | 100.8 | 3.22 | 100.3 | 6.58 | 2.84 | 88.9 (1.25) | 89.9 (4.98) | ||
| Medium | 200 | 109.4 | 3.62 | 108.5 | 2.98 | 5.10 | 102.4 (1.14) | 93.1 (3.65) | ||
| High | 4000 | 105.2 | 4.80 | 103.0 | 5.25 | 1.24 | 103.1 (3.34) | 99.5 (1.37) | ||
| Pool-S | 4.68 | 3.70 | ||||||||
| 16OH-DHEAS DHT-G | 2.00 - 10,000 | LLOQ | 2.00 | 99.8 | 11.1 | 104.6 | 7.63 | 5.15 | ||
| Low | 5.00 | 105.5 | 9.29 | 100.5 | 7.61 | 5.88 | 98.3 (6.43) | 102.4 (2.82) | ||
| Medium | 200 | 106.2 | 4.38 | 104.7 | 4.37 | 5.90 | 103.9 (2.15) | 96.4 (3.02) | ||
| High | 4000 | 100.8 | 5.79 | 101.1 | 6.18 | 2.53 | 105.3 (4.31) | 97.5 (2.16) | ||
| Pool-S | 3.04 | 2.99 | ||||||||
| 2.50 - 2500 | LLOQ | 2.50 | 106.6 | 4.77 | 104.6 | 3.85 | 3.75 | |||
| Low | 10.0 | 97.9 | 5.84 | 100.3 | 7.88 | 2.45 | 99.5 (9.14) | 97.2 (4.06) | ||
| Medium | 500 | 96.7 | 4.69 | 101.1 | 8.37 | 4.08 | 109.6 (4.67) | 91.9 (2.21) | ||
| High | 2000 | 100.5 | 3.63 | 102.8 | 5.14 | 1.57 | 110.1 (2.78) | 98.3 (0.45) | ||
| Pool-S | 6.29 | 1.90 | ||||||||
| E2S2 E2S E1S | 1.00 - 100 | LLOQ | 1.00 | 111.9 | 3.20 | 104.5 | 9.84 | 6.30 | ||
| Low | 2.50 | 107.1 | 3.07 | 102.4 | 6.54 | 5.56 | 100.9 (1.86) | 88.4 (4.64) | ||
| Medium | 10.0 | 106.7 | 4.33 | 106.2 | 4.64 | 4.74 | 103.6 (9.40) | 84.3 (0.38) | ||
| High | 100 | 105.1 | 2.29 | 99.7 | 6.96 | 9.10 | 104.9 (3.25) | 85.2 (3.06) | ||
| 0.25 - 500 | LLOQ | 0.25 | 102.0 | 6.83 | 102.6 | 7.27 | 7.54 | |||
| Low | 1.00 | 95.5 | 9.80 | 96.3 | 11.1 | 7.00 | 95.2 (9.93) | 94.4 (8.20) | ||
| Medium | 50.0 | 96.4 | 5.13 | 99.1 | 4.33 | 5.11 | 108.3 (1.07) | 93.4 (1.88) | ||
| High | 200 | 108.9 | 2.75 | 109.8 | 3.30 | 3.06 | 113.6 (3.30) | 97.4 (1.66) | ||
| 0.25 - 500 | LLOQ | 0.25 | 94.8 | 11.8 | 94.7 | 7.63 | 12.3 | |||
| Low | 1.00 | 104.4 | 3.79 | 105.5 | 5.62 | 9.75 | 102.3 (14.3) | 86.8 (3.32) | ||
| Medium | 50.0 | 108.0 | 1.57 | 105.2 | 5.62 | 8.70 | 98.2 (3.11) | 91.0 (4.83) | ||
| High | 200 | 105.4 | 2.25 | 99.4 | 7.47 | 4.22 | 101.2 (3.03) | 95.8 (0.42) | ||
| Pool-S | 7.14 | 7.06 | ||||||||
| Etio-S | 2.00 - 10,000 | LLOQ | 2.00 | 102.5 | 5.30 | 105.8 | 7.97 | 2.05 | ||
| Low | 5.00 | 90.1 | 6.30 | 92.7 | 5.84 | 4.81 | 77.7 (2.34) | 97.7 (2.55) | ||
| Medium | 1000 | 106.7 | 4.00 | 106.8 | 3.72 | 1.71 | 102.2 (3.38) | 99.9 (0.92) | ||
| High | 4000 | 105.5 | 3.93 | 105.5 | 4.42 | 2.17 | 103.9 (3.08) | 96.8 (1.22) | ||
| Pool-S | 7.37 | 3.32 | ||||||||
(continued on next page)
| Analyte | Calibration Range [ng/ml] | Quality control | Intra-dayª | Inter-dayb | Reinjection€ Precision [%CV] | Recoveryª Mean [%] (%CV) | Matrix effecte Mean [%] (%CV) | |||
|---|---|---|---|---|---|---|---|---|---|---|
| Level | Concentration [ng/ml] | Accuracy [%] | Precision [%CV] | Accuracy [%] | Precision [%CV] | |||||
| 11-ketoetio-G | 5.00 - 10,000 | LLOQ | 5.00 | 107.2 | 9.07 | 109.6 | 7.07 | 7.72 | ||
| Low | 20.0 | 92.1 | 5.22 | 93.7 | 6.76 | 2.91 | 103.4 (19.9) | 104.4 (3.28) | ||
| Medium | 200 | 88.6 | 4.09 | 89.7 | 4.62 | 2.07 | 106.0 (3.23) | 93.9 (2.13) | ||
| High | 4000 | 101.3 | 6.91 | 102.8 | 5.10 | 4.35 | 110.1 (4.13) | 94.5 (3.19) | ||
| Pool-S | 2.04 | 1.50 | ||||||||
| 11-ketoetio-S Preg-G Preg-S | 1.00 - 2000 | LLOQ | 1.00 | 112.4 | 8.86 | 109.5 | 9.00 | 9.16 | ||
| Low | 4.00 | 101.5 | 3.57 | 100.0 | 5.02 | 3.02 | 112.1 (13.3) | 98.6 (8.18) | ||
| Medium | 200 | 107.5 | 3.84 | 108.9 | 4.55 | 5.95 | 116.1 (3.47) | 96.8 (4.42) | ||
| High | 800 | 111.3 | 2.38 | 107.7 | 4.95 | 1.11 | 105.5 (3.50) | 101.4 (2.88) | ||
| Pool-S | 6.92 | 3.62 | ||||||||
| 20.0 - 10,000 | LLOQ | 20.0 | 103.0 | 7.38 | 97.4 | 7.44 | 7.07 | |||
| Low | 200 | 101.4 | 4.47 | 106.4 | 6.05 | 2.17 | 100.3 (1.59) | 99.6 (0.46) | ||
| Medium | 1000 | 99.9 | 5.84 | 104.6 | 6.26 | 3.47 | 111.6 (3.55) | 95.4 (2.33) | ||
| High | 4000 | 92.7 | 3.30 | 94.2 | 5.25 | 2.49 | 104.9 (3.23) | 100.6 (0.12) | ||
| 0.40 - 2000 | LLOQ | 0.40 | 110.3 | 11.6 | 101.5 | 9.85 | 4.09 | |||
| Low | 1.00 | 94.3 | 7.99 | 99.0 | 8.53 | 5.37 | 81.9 (6.65) | 96.3 (3.57) | ||
| Medium | 40.0 | 98.9 | 0.97 | 96.7 | 3.23 | 1.77 | 99.6 (5.13) | 97.3 (3.03) | ||
| High | 800 | 114.1 | 0.49 | 110.8 | 6.03 | 3.21 | 99.8 (2.89) | 94.7 (5.99) | ||
| Pool-S | 7.91 | 11.5 | ||||||||
| 17OH-preg-S | 1.00 - 2500 | LLOQ | 1.00 | 102.1 | 6.27 | 102.3 | 5.94 | 8.61 | ||
| Low | 2.50 | 89.3 | 7.16 | 98.9 | 9.86 | 5.60 | 85.1 (4.19) | 104.6 (6.60) | ||
| Medium | 100 | 96.9 | 3.78 | 101.6 | 5.81 | 2.33 | 102.3 (3.38) | 101.4 (3.96) | ||
| High | 2000 | 112.1 | 1.96 | 108.2 | 5.08 | 3.60 | 109.7 (3.32) | 103.6 (3.92) | ||
| Pool-S | 2.49 | 5.42 | ||||||||
| 17OH-prog-S | 0.10 - 250 | LLOQ | 0.10 | 100.0 | 8.39 | 104.0 | 11.0 | 5.43 | ||
| Low | 0.25 | 88.8 | 7.61 | 97.4 | 7.57 | 4.49 | 78.0 (6.01) | 98.7 (3.36) | ||
| Medium | 10.0 | 92.4 | 4.98 | 92.0 | 4.34 | 10.4 | 99.3 (5.03) | 87.5 (1.23) | ||
| High | 200 | 100.7 | 5.01 | 98.2 | 2.18 | 3.79 | 106.2 (2.67) | 97.8 (3.29) | ||
| DiOH-prog-S | 1.00 - 500 | LLOQ | 1.00 | 106.7 | 3.63 | 106.1 | 6.55 | 3.09 | ||
| Low | 10.0 | 108.6 | 2.18 | 105.1 | 6.02 | 5.50 | 108.4 (2.07) | 96.3 (1.06) | ||
| Medium | 50.0 | 93.4 | 3.21 | 97.7 | 6.94 | 5.58 | 111.6 (2.67) | 97.6 (4.69) | ||
| High | 200 | 92.9 | 2.86 | 93.4 | 3.85 | 3.91 | 111.2 (5.00) | 98.6 (2.31) | ||
| Testo-G Testo-S | 20.0 - 10,000 | LLOQ | 20.0 | 109.5 | 4.51 | 111.0 | 2.98 | 5.01 | ||
| Low | 200 | 87.8 | 2.25 | 90.5 | 4.52 | 2.81 | 128.3 (2.73) | 100.3 (3.56) | ||
| Medium | 1000 | 93.3 | 1.28 | 97.3 | 5.90 | 2.29 | 139.0 (6.03) | 95.9 (1.02) | ||
| High | 4000 | 98.6 | 2.03 | 100.5 | 4.06 | 2.46 | 118.3 (4.82) | 101.2 (2.98) | ||
| Pool-S | 7.54 | 6.10 | ||||||||
| 0.40 - 2000 | LLOQ | 0.40 | 100.3 | 10.1 | 107.0 | 7.50 | 5.51 | |||
| Low | 1.00 | 96.9 | 8.76 | 100.6 | 9.33 | 5.13 | 99.9 (8.00) | 97.6 (6.08) | ||
| Medium | 40.0 | 92.2 | 4.53 | 95.6 | 7.29 | 6.81 | 104.9 (5.72) | 94.1 (3.55) | ||
| High | 800 | 109.2 | 7.71 | 108.4 | 6.42 | 3.72 | 102.4 (4.43) | 101.8 (0.59) | ||
| Pool-S | 13.7 | 4.88 | ||||||||
| Creatinine | 50.0 - 2000 µg/ml | LLOQ | 50.0 µg/ml | 102.9 | 1.67 | 97.7 | 7.44 | 2.81 | ||
| Low | 125 µg/ml | 97.8 | 2.06 | 98.9 | 5.33 | 4.37 | ||||
| Medium | 500 µg/ml | 98.6 | 7.79 | 104.0 | 7.55 | 7.15 | ||||
| High | 1250 µg/ml | 97.9 | 3.64 | 99.7 | 5.25 | 3.14 | ||||
| Pool-S | 0.99 | 2.55 | ||||||||
underlined: outside specification limits
Pool-S: pooled real-world sample
a shown as the mean and %CV of five experimental replicates, prepared on the same day and analysed within one run
b shown as the mean and %CV of five experimental replicates, prepared on three different days and analysed in three different runs
” shown as the %CV of five analytical replicates analysed within one run
d shown as the mean and %CV of three experimental replicates analysed within one run; calculated by dividing the analyte response in urine spiked with reference standards before sample preparation by the analyte response in urine spiked with the same amount of reference standards after sample preparation
e shown as the mean and %CV of three experimental replicates analysed within one run; calculated by dividing the analyte response in urine spiked with reference standards after sample preparation by the analyte response in eluent spiked with the same amount of reference standards
chromatograms of the real-world sample several mass transitions showed multiple peaks across the entire acquisition time. These were all baseline separated from the analyte peak except for 11-ketoetio-S, 17OH-prog-S, DHEA-G, E2S2, testo-G and testo-S, and in these cases all peaks remained visibly separable with retention time differences of at least 0.10 min. The isobaric compounds an-G and etiocholanolone glucuronide could not be separated in the applied chromatographic conditions. Consequently, the signal recorded for the mass transition 465.1 → 113.0 (465.1 → 75.0 // qualifier) at retention time 4.32 min represents the sum of both compounds, but only the reference standard
an-G was used for the calibration.
The analytes 17OH-prog-S, DHEA-G, diOH-prog-S, E2S2, E2S and preg-G were either not found at all in the urine or plasma real-world sample or in concentrations below the LLOQ.
While all compounds in the targeted analyte panel were baseline separated from each other, interferences with further isobaric steroid metabolites were experienced. In these cases, the analyte identity should be reviewed critically in every sample and should be excluded from the analysis in any case of doubt. Nonetheless, despite our efforts to reduce this concern by the quantifier/qualifier ion ratio and the inclusion of
| Analyte | Calibration Range [ng/ml] | Quality control | Intra-dayª | Inter-dayb | Reinjection Precision [%CV] | Recoveryª Mean [%] (%CV) | Matrix effecte Mean [%] (%CV) | |||
|---|---|---|---|---|---|---|---|---|---|---|
| Level | Concentration [ng/ml] | Accuracy [%] | Precision [%CV] | Accuracy [%] | Precision [%CV] | |||||
| Andiol-diS | 40.0 - 20,000 | LLOQ | 40.0 | 99.9 | 8.00 | 98.5 | 8.08 | 4.83 | ||
| Low | 200 | 103.4 | 6.91 | 103.0 | 6.72 | 6.14 | 98.1 (6.80) | 106.7 (2.42) | ||
| Medium | 1000 | 97.3 | 8.65 | 94.5 | 6.56 | 4.57 | 100.2 (3.32) | 103.9 (1.99) | ||
| High | 8000 | 97.6 | 5.99 | 98.5 | 6.81 | 5.70 | 95.2 (7.17) | 97.8 (2.05) | ||
| Pool-S | 6.20 | 1.74 | ||||||||
| Andiol-S | 5.00 - 500 | LLOQ | 5.00 | 111.2 | 5.59 | 106.7 | 13.8 | 4.63 | ||
| Low | 10.0 | 100.5 | 2.56 | 100.4 | 3.48 | 6.06 | 88.1 (9.15) | 507.9 (4.54) | ||
| Medium | 50.0 | 111.5 | 2.30 | 108.9 | 3.83 | 4.24 | 91.4 (9.95) | 472.2 (4.23) | ||
| High | 250 | 99.8 | 7.31 | 102.8 | 5.29 | 9.17 | 86.0 (3.53) | 425.6 (2.14) | ||
| Pool-S | 5.27 | 5.07 | ||||||||
| An-G | 1.00 - 500 | LLOQ | 1.00 | 93.1 | 11.3 | 96.3 | 7.95 | 4.04 | ||
| Low | 5.00 | 103.0 | 6.39 | 106.5 | 3.25 | 3.79 | 82.1 (5.12) | 792.7 (2.81) | ||
| Medium | 25.0 | 107.0 | 1.93 | 105.4 | 4.44 | 2.53 | 77.4 (3.84) | 778.2 (2.28) | ||
| High | 200 | 104.8 | 5.59 | 102.2 | 4.08 | 2.99 | 86.0 (5.71) | 385.2 (3.55) | ||
| Pool-S | 4.40 | 2.76 | ||||||||
| An-S | 20.0 - 20,000 | LLOQ | 20.0 | 102.4 | 5.26 | 100.8 | 13.2 | 3.11 | ||
| Low | 40.0 | 100.9 | 6.35 | 100.5 | 5.70 | 4.68 | 95.3 (1.39) | 169.3 (5.42) | ||
| Medium | 1000 | 102.4 | 3.86 | 100.5 | 9.52 | 2.77 | 95.5 (1.68) | 113.6 (5.18) | ||
| High | 8000 | 108.6 | 5.20 | 109.9 | 5.27 | 3.48 | 99.0 (4.32) | 97.3 (0.47) | ||
| Pool-S | 4.46 | 2.37 | ||||||||
| Cortisol-S | 0.10 - 100 | LLOQ | 0.10 | 97.0 | 12.6 | 97.4 | 10.5 | 11.3 | ||
| Low | 0.20 | 96.5 | 3.15 | 101.4 | 8.39 | 8.63 | 87.0 (9.56) | 842.0 (2.85) | ||
| Medium | 5.00 | 105.8 | 3.63 | 103.4 | 5.36 | 6.20 | 81.5 (2.94) | 775.6 (4.57) | ||
| High | 40.0 | 106.8 | 4.64 | 103.8 | 4.72 | 3.99 | 79.7 (3.58) | 632.3 (1.92) | ||
| Pool-S | 11.2 | 7.00 | ||||||||
| DHEA-G | 2.00 - 2000 | LLOQ | 2.00 | 95.3 | 4.62 | 100.2 | 10.9 | 5.74 | ||
| Low | 4.00 | 97.3 | 7.27 | 97.4 | 7.50 | 6.49 | 85.7 (6.25) | 509.5 (2.95) | ||
| Medium | 100 | 103.5 | 6.17 | 102.1 | 6.13 | 4.07 | 94.7 (10.2) | 439.9 (2.25) | ||
| High | 800 | 101.0 | 8.22 | 102.9 | 7.85 | 6.18 | 82.9 (6.58) | 239.0 (1.36) | ||
| DHEAS | 20.0 - 20,000 | LLOQ | 20.0 | 108.7 | 7.04 | 100.9 | 8.31 | 5.05 | ||
| Low | 40.0 | 115.7 | 3.75 | 108.5 | 7.25 | 3.41 | 88.3 (11.9) | 201.8 (8.37) | ||
| Medium | 1000 | 110.3 | 2.50 | 107.3 | 3.76 | 2.68 | 97.3 (0.94) | 129.6 (2.26) | ||
| High | 8000 | 107.0 | 5.10 | 103.2 | 4.45 | 3.26 | 97.0 (3.70) | 100.8 (1.04) | ||
| Pool-S | 5.21 | 2.73 | ||||||||
| 16OH-DHEAS DHT-G | 5.00 - 500 | LLOQ | 5.00 | 101.6 | 8.29 | 97.6 | 6.52 | 3.67 | ||
| Low | 10.0 | 104.1 | 12.3 | 101.0 | 11.3 | 8.00 | 88.5 (7.74) | 454.1 (4.79) | ||
| Medium | 50.0 | 107.5 | 8.67 | 105.4 | 5.10 | 5.37 | 85.2 (5.15) | 456.1 (2.81) | ||
| High | 250 | 99.0 | 3.25 | 98.2 | 2.10 | 6.68 | 96.5 (6.21) | 310.6 (7.44) | ||
| Pool-S | 3.81 | 1.75 | ||||||||
| 1.00 - 500 | LLOQ | 1.00 | 98.5 | 5.44 | 93.0 | 9.96 | 8.74 | |||
| Low | 5.00 | 98.1 | 8.00 | 102.8 | 10.4 | 5.88 | 87.6 (2.79) | 638.6 (3.24) | ||
| Medium | 25.0 | 101.1 | 8.04 | 102.2 | 9.46 | 5.94 | 89.2 (4.43) | 652.8 (3.34) | ||
| High | 200 | 105.7 | 6.71 | 99.9 | 8.03 | 4.80 | 83.2 (3.90) | 487.2 (1.95) | ||
| Pool-S | 12.7 | 12.1 | ||||||||
| E2S2 | 2.00 - 2000 | LLOQ | 2.00 | 89.6 | 6.77 | 91.9 | 15.1 | 7.33 | ||
| Low | 4.00 | 98.4 | 6.49 | 94.5 | 4.58 | 4.11 | 101.0 (8.21) | 94.2 (5.74) | ||
| Medium | 100 | 100.8 | 8.15 | 103.2 | 8.58 | 3.49 | 99.9 (6.38) | 101.2 (1.18) | ||
| High | 800 | 97.2 | 6.09 | 97.2 | 5.68 | 4.74 | 104.1 (13.8) | 98.3 (0.77) | ||
| E2S | 0.10 - 100 | LLOQ | 0.10 | 104.0 | 9.10 | 99.0 | 12.1 | 4.66 | ||
| Low | 0.20 | 106.0 | 3.17 | 102.3 | 3.52 | 4.54 | 87.8 (2.03) | 412.4 (1.48) | ||
| Medium | 5.00 | 104.8 | 6.13 | 103.5 | 4.71 | 3.46 | 90.5 (6.24) | 388.4 (4.46) | ||
| High | 40.0 | 103.3 | 4.27 | 101.9 | 6.58 | 2.30 | 91.6 (8.59) | 310.2 (1.72) | ||
| E1S | 0.10 - 100 | LLOQ | 0.10 | 102.0 | 7.23 | 97.8 | 3.42 | 2.48 | ||
| Low | 0.20 | 96.0 | 5.73 | 96.1 | 4.75 | 4.94 | 84.0 (6.28) | 275.6 (6.63) | ||
| Medium | 5.00 | 107.8 | 3.82 | 106.4 | 3.26 | 3.72 | 79.9 (5.42) | 273.5 (2.00) | ||
| High | 40.0 | 107.9 | 4.02 | 102.6 | 10.4 | 4.19 | 81.7 (7.86) | 232.5 (4.89) | ||
| Pool-S | 5.73 | 5.25 | ||||||||
| Etio-S | 1.00 - 500 | LLOQ | 1.00 | 104.0 | 5.41 | 103.2 | 9.87 | 7.01 | ||
| Low | 2.00 | 97.9 | 6.70 | 97.6 | 7.14 | 5.10 | 93.6 (7.99) | 154.2 (4.11) | ||
| Medium | 50.0 | 99.9 | 4.34 | 99.5 | 6.01 | 3.72 | 98.2 (8.77) | 148.7 (4.75) | ||
| High | 400 | 112.9 | 1.98 | 112.1 | 2.27 | 4.95 | 97.0 (4.88) | 111.7 (3.16) | ||
| Pool-S | 2.67 | 2.72 | ||||||||
| 11-ketoetio-G | 1.00 - 500 | LLOQ | 1.00 | 100.8 | 1.90 | 98.0 | 5.63 | 5.60 | ||
| Low | 5.00 | 102.6 | 2.85 | 104.9 | 4.15 | 4.64 | 85.9 (2.99) | 681.7 (4.26) | ||
| Medium | 25.0 | 104.1 | 4.21 | 107.2 | 5.88 | 5.33 | 89.2 (3.17) | 651.2 (4.33) | ||
| High | 200 | 101.0 | 5.80 | 99.7 | 8.14 | 2.88 | 82.5 (8.02) | 447.5 (3.88) | ||
| Pool-S | 6.13 | 6.05 | ||||||||
(continued on next page)
| Analyte | Calibration Range [ng/ml] | Quality control | Intra-dayª | Inter-dayb | Reinjection€ Precision [%CV] | Recoveryª Mean [%] (%CV) | Matrix effecte Mean [%] (%CV) | |||
|---|---|---|---|---|---|---|---|---|---|---|
| Level | Concentration [ng/ml] | Accuracy [%] | Precision [%CV] | Accuracy [%] | Precision [%CV] | |||||
| 11-ketoetio-S | 0.10 - 100 | LLOQ | 0.10 | 113.0 | 4.00 | 116.0 | 1.83 | 6.51 | ||
| Low | 0.20 | 105.5 | 10.2 | 101.4 | 11.1 | 8.12 | 97.7 (7.33) | 1247.3 (2.32) | ||
| Medium | 5.00 | 104.6 | 0.87 | 102.7 | 3.06 | 3.19 | 97.4 (5.27) | 1207.2 (4.87) | ||
| High | 40.0 | 102.9 | 8.35 | 99.1 | 7.51 | 7.80 | 97.8 (14.4) | 951.8 (9.55) | ||
| Pool-S | 4.31 | 10.1 | ||||||||
| Preg-G Preg-S | 4.00 - 2000 | LLOQ | 4.00 | 101.1 | 3.10 | 99.6 | 2.92 | 5.16 | ||
| Low | 20.0 | 99.7 | 5.30 | 102.3 | 6.72 | 7.76 | 80.3 (11.1) | 667.6 (0.88) | ||
| Medium | 100 | 103.0 | 3.52 | 103.0 | 4.94 | 2.02 | 90.5 (1.33) | 647.5 (0.69) | ||
| High | 800 | 111.3 | 3.05 | 107.5 | 6.26 | 4.04 | 92.8 (6.31) | 384.5 (14.6) | ||
| 5.00 - 5000 | LLOQ | 5.00 | 112.4 | 2.68 | 107.8 | 10.5 | 3.27 | |||
| Low | 10.0 | 111.4 | 1.46 | 106.7 | 10.0 | 2.55 | 104.3 (6.15) | 194.4 (3.41) | ||
| Medium | 250 | 106.3 | 5.10 | 107.5 | 6.41 | 4.78 | 97.7 (3.59) | 195.0 (1.13) | ||
| High | 2000 | 99.5 | 2.97 | 98.7 | 3.94 | 0.93 | 100.3 (4.06) | 129.6 (4.06) | ||
| Pool-S | 2.59 | 1.91 | ||||||||
| 17OH-preg-S | 1.00 - 200 | LLOQ | 1.00 | 106.5 | 1.68 | 106.6 | 3.17 | 4.76 | ||
| Low | 2.00 | 105.7 | 2.90 | 102.1 | 4.03 | 4.03 | 91.8 (13.8) | 214.1 (13.5) | ||
| Medium | 10.0 | 103.4 | 5.37 | 99.4 | 7.06 | 7.19 | 89.8 (6.06) | 193.7 (9.49) | ||
| High | 50.0 | 107.1 | 3.80 | 103.7 | 4.07 | 2.11 | 93.1 (5.50) | 146.2 (2.50) | ||
| Pool-S | 2.09 | 3.30 | ||||||||
| 17OH-prog-S | 0.10 - 100 | LLOQ | 0.10 | 106.0 | 4.13 | 103.2 | 6.97 | 2.42 | ||
| Low | 0.20 | 108.5 | 4.86 | 102.3 | 11.4 | 5.52 | 96.7 (5.26) | 582.0 (5.88) | ||
| Medium | 5.00 | 108.1 | 7.39 | 102.3 | 5.62 | 4.00 | 97.5 (2.86) | 525.2 (3.72) | ||
| High | 40.0 | 101.0 | 7.20 | 100.6 | 9.20 | 6.09 | 97.5 (2.86) | 390.1 (3.62) | ||
| DiOH-prog-S | 0.10 - 100 | LLOQ | 0.10 | 106.0 | 4.64 | 103.4 | 6.47 | 1.87 | ||
| Low | 0.20 | 108.0 | 3.84 | 100.4 | 8.06 | 7.05 | 91.4 (6.56) | 604.5 (4.87) | ||
| Medium | 5.00 | 101.2 | 5.42 | 99.5 | 5.43 | 5.73 | 97.9 (1.44) | 503.6 (2.43) | ||
| High | 40.0 | 107.4 | 5.46 | 106.5 | 4.31 | 2.74 | 89.2 (6.02) | 311.0 (2.21) | ||
| Testo-G | 20.0 - 2000 | LLOQ | 20.0 | 123.2 | 1.10 | 119.4 | 3.55 | 0.90 | 68.5 (9.28) | 1416.3 (4.62) |
| Medium | 100 | 96.6 | 5.18 | 95.3 | 2.28 | 1.44 | 70.9 (7.64) | 1062.5 (4.84) | ||
| High | 800 | 111.3 | 3.06 | 103.2 | 10.6 | 1.88 | 76.0 (3.76) | 721.7 (7.39) | ||
| Testo-S | 0.5 -500 | LLOQ | 0.50 | 106.4 | 10.2 | 101.0 | 15.2 | 8.02 | ||
| Low | 1.00 | 102.2 | 7.23 | 96.2 | 4.57 | 4.17 | 97.5 (3.35) | 632.2 (4.86) | ||
| Medium | 25.0 | 103.6 | 4.37 | 106.3 | 5.72 | 4.66 | 93.6 (3.84) | 607.0 (3.99) | ||
| High | 200 | 99.4 | 5.80 | 100.2 | 6.29 | 3.54 | 93.1 (5.08) | 357.7 (5.27) | ||
| Pool-S | 3.41 | 5.18 | ||||||||
underlined: outside specification limits
Pool-S: pooled real-world sample
a shown as the mean and %CV of five experimental replicates, prepared on the same day and analysed within one run
b shown as the mean and %CV of five experimental replicates, prepared on three different days and analysed in three different runs
” shown as the %CV of five analytical replicates analysed within one run
d shown as the mean and %CV of three experimental replicates analysed within one run; calculated by dividing the analyte response in plasma spiked with reference standards before sample preparation by the analyte response in plasma spiked with the same amount of reference standards after sample preparation
e shown as the mean and %CV of three experimental replicates analysed within one run; calculated by dividing the analyte response in plasma spiked with reference standards after sample preparation by the analyte response in eluent spiked with the same amount of reference standards
real-world samples during method validation, this limitation remains to be fully resolved due to the large number of possibly unknown isobaric steroidal compounds [36].
3.3.5. Dilution integrity
1:2 dilution of QC samples showed accurate and precise results for all analytes in urine and plasma within the pre-specified limits. While the precision of replicates diluted 1:10 was less than 9 %CV for all analytes, the acceptance criteria for accuracy was exceeded in the case of three analytes in urine and six analytes in plasma (Table S5).
3.3.6. Stability
The majority of QC samples in the stability experiments showed re- sults within the prespecified limits, suggesting sufficient analyte stabil- ity. Among all analytes, DHEA-G and testo-G showed the least stability across all conditions, while overall analyte stability seemed better in urine than in plasma. Detailed results can be found in the Tables S6 - S9.
3.4. Method application to samples of patients with ACC and ACA
A total of 20 urine samples and 20 plasma samples from 10 patients with ACC and 10 patients with ACA were analysed. All patients were male and the mean age was comparable among the two groups (ACC: 50.5 ± 10.9 years; ACA: 52.1 ± 6.7 years).
A summary of the analyte concentrations found in the patient sam- ples is shown in Table 3. Overall, the median concentrations were highest in the ACC group for all analytes except for DHEA-G, where the median concentration in urine was higher in patients with ACA (57.3 vs. 79.5 µg/g crea), which was also the case for DHT-G in plasma (3.0 vs. 3.3 ng/ml). The six analytes with the most significant group differences for ACC versus ACA are shown in Fig. 2. 17OH-prog-S and preg-G were not found in any of the samples, E2S2 was found in the urine and plasma of one ACC patient (17.6 ug/g crea and 11.9 ng/ml) and diOH-prog-S was found only in the urine of one ACC patient (12.8 ug/g crea). The analyte concentrations in the ACC group showed a considerably higher interindividual variability than in the ACA group. Of all 880 results (22
| Andiol-diS | Andiol-S | An-G | An-S | Cortisol-S | DHEA-G | DHEA | S DH | E2S2 | E2S | E1S | Etio-S | 11-ketoetio-G | 11-ketoetio-S | Preg-S | 17OH-preg-S | DiOH-prog-S | Testo-G | Testo-S | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Urine | [µg/g crea] | |||||||||||||||||||
| ACC (n = 10) | ||||||||||||||||||||
| Median | 4466 | 4130 | 17666 | 5028 | 175 | 57.3 | 25543 | 11714 | 60.6 | N/Aª | 6.83b | 210℃ | 2940 | 1760 | 51.9b | 140 | 713€ | N/Aª | 267 | 95.2ª |
| Min. | 128 | 24.7 | 4646 | 84.0 | 22.6 | N/A | 110 | 582 | 14.2 | N/A | N/A | 20.8 | 503 | N/A | <LLOQ | N/A | 111 | N/A | ||
| Max. | 35831 | 25673 | 344738 | 34301 | 2155 | 357 | 133693 | 46494 | 460 | 17.6 | 111 | 4776 | 14717 | 10557 | 246 | 1125 | 3111 | 12.8 | 2412 | 385 |
| ACA (n = 10) | ||||||||||||||||||||
| Median | 110 | 12.3 | 3216 | 288 | 38.0 | 79.5b | 63.1 | 265 | 29.8 | N/A | N/A | 1.05 | 61.3 | 705 | 12.9b | 0.64b | 4.27 | N/A | 202 | 2.87b |
| Min. | 51.8 | 2.86 | 2300 | 40.6 | 15.1 | N/A | 16.1 | 33.6 | 16.3 | N/A | 5.48 | 221 | N/A | N/A | N/A | 94.3 | N/A | |||
| Max. | 1521 | 144 | 5847 | 720 | 134 | 98.4 | 730 | 1046 | 112 | 2.56 | 850 | 1346 | 28.4 | 1.70 | 24.9 | 632 | 18.4 | |||
| p-value | <0.001 | <0.0001 | <0.0001 | <0.05 | <0.05 | ns | <0.001 | <0.001 | ns | <0.01 | <0.05 | <0.05 | ns | <0.001 | <0.001 | ns | <0.01 | |||
| Plasma | [ng/ml] | |||||||||||||||||||
| ACC (n = 10) | ||||||||||||||||||||
| Median | 2995 | 456 | 129 | 1350 | 0.91 | N/A | 4578 | 296 | 3.01 | N/Aª | 0.24 | 15.8 | 92.6 | 13.4 | 1.01b | 447 | 114 | N/A | N/A | 7.74 |
| Min. | 285 | 77.6 | 36.6 | 209 | 0.18 | 1504 | 65.8 | 1.48 | N/A | 0.26 | 2.33 | 3.63 | N/A | 43.5 | 12.0 | 0.96 | ||||
| Max. | 13459 | 1549 | 2049 | 4659 | 12.1 | 16395 | 834 | 12.8 | 11.9 | 0.88 | 260 | 490 | 140 | 1.92 | 1421 | 242 | 49.8 | |||
| ACA (n = 10) | ||||||||||||||||||||
| Median | 488 | 51.2 | 70.1 | 434 | 0.27 | N/A | 756 | 61.3 | 3.31 | N/A | N/A | 1.25 | 14.6 | 5.46 | 0.31 | 45.9 | 3.76 | N/A | N/Aª | 4.52 |
| Min. | 200 | 23.3 | 21.4 | 256 | 0.13 | 327 | 14.9 | 2.85 | 0.29 | 1.47 | 1.50 | 0.19 | 19.9 | 1.67 | 1.41 | |||||
| Max. | 3430 | 118 | 145 | 854 | 2.67 | 2249 | 162 | 11.8 | 3.75 | 89.1 | 14.5 | 0.90 | 191 | 25.1 | 38.9 | 15.2 | ||||
| p-value | <0.01 | <0.001 | <0.05 | <0.05 | <0.05 | <0.0001 | <0.001 | ns | <0.01 | ns | ns | ns | <0.001 | <0.001 | ns | |||||
LLOQ = lower limit of quantification; N/A = analyte not found; ns = not significant, p > 0.05; crea = creatinine; Min = lowest value found in all samples; Max .= highest value found in all samples p-value calculated with the Mann-Whitney-U-test for analyte concentrations in the ACC group versus the ACA group.
a
analyte was found only in one sample;
b analyte was not found in all samples;
analyte was not found in one sample.
analytes in 20 patients in plasma and urine) 1.3 % were below the LLOQ of few analytes, while 3.8 % were above an upper limit of quantification (ULOQ). All of the samples, in which the ULOQ of one or more analytes was exceeded were from the ACC group and were diluted with surrogate matrix and analysed again.
The method was successfully applied to a number of urine and plasma samples from patients with adrenal tumours demonstrating it is fit for purpose. The analyte panel and the calibration ranges proved to be appropriate, since all but two analytes were found in the patient samples and only few analyte concentrations exceeded the upper or lower cali- bration limits. As expected, some ACC patients had particularly high levels of steroid conjugates in both plasma and urine, but this analytical challenge could be tackled by repeating the measurements with diluted samples. Despite the small sample size, our data strengthen the previ- ously proposed biomarker potential of conjugated steroids for adrenal
tumours, as multiple steroid conjugates were significantly more abun-
dant in urine and plasma of patients with ACC compared to ACA. This was particularly evident for various sulfate conjugates and andiol-dis (Fig. 2), which is in accordance with a previously proposed patho- physiological role of sulfated steroids in adrenal tumours [7] and ob- servations made for other corticoid-dependent diseases [17]. The described method enables in-depth investigations in this area, which will be required to generate valid clinical data.
In accordance with other studies, steroid concentrations were significantly higher in patients with ACC compared to patients with ACA [13,14,34,35]. When directly comparing our results to the absolute concentrations of deconjugated/unconjugated steroids in patients with adrenal tumours in literature we found overall our values to be higher. Arlt et al. determined median concentrations of 16-hydroxydehydroe- piandrosterone in 24h-urine to be 2.14 umol/24 h in 45 ACC patients and 0.66 umol/24 h in 102 ACA patients, while our respective median values of 16OH-DHEAS in 24h-urine were 35.3 umol/24 h and 0.82 umol/24 h [13]. The median concentrations of serum 17-hydroxy- pregnenolone in the study of Taylor et al. were 0.0226 nmol/ml (ACC n = 10) and 0.0039 nmol/ml (ACA n = 16), with our median levels of 17OH-preg-S in plasma being 0.2271 nmol/ml (ACC) and 0.0091 nmol/ml (ACA) [35]. The observed differences in urine indicate a systematic discrepancy between the two analytical approaches, possibly caused by incomplete enzymatic deconjugation during sample preparation, while the variations in the blood analyses are most likely due to the different extent of conjugation of the individual steroids circulating in blood.
Some limitations should be acknowledged regarding the study of the clinical samples, namely the lack of diversity in the study population in terms of sex and ethnicity and the low patient number. Due to the low sample number, we decided to avoid factors influencing steroid levels such as sex and menopause, to enable detection of trends between ho- mogenous patient populations. As mentioned before, the analysis of the patient samples was performed with the sole purpose of proving the methods applicability and to have a first look into biomarker validation of steroid conjugates for adrenal tumours. Subsequent large-scale studies to qualify and validate biomarkers should consider the magni- tude of human diversity including all sex and age groups, ethnicities and physicalities to provide representative, clinically applicable data.
4. Conclusion
When reviewing the main attributes of the presented method, such as the broad analyte panel, wide calibration ranges, short instrument run time and an easy sample preparation procedure, it seems suitable for the collection of extensive data about steroid conjugate levels in healthy volunteers, to determine reference intervals for the respective com- pounds in human urine and plasma. Furthermore, when applied to pathological samples, the acquired data facilitates biomarker research on steroid conjugates and their validation for corticoid-dependent dis- eases. Paired analysis of adrenal tumour tissue and urine/plasma
A
16OH-DHEAS
DHEAS
B
16OH-DHEAS
DHEAS
50000
140000
1000
20000
Concentration in urine [pg/g creatinine]
Plasmaconcentration [ng/ml]
Plasmaconcentration [ng/ml]
.
26000
Concentration in urine [ug/g creatinine]
75000
600
12000
2000 1200
9000
200
4000
1200
180
2400
800
8
800
120
1600
400
400
60
800
0
0
0
0
ACC
ACA
ACC
ACA
ACC
ACA
ACC
ACA
17OH-preg-S
Preg-S
17OH-preg-S
Preg-S
3500
1200
250
1500
Concentration in urine [ug/g creatinine]
2000
Concentration in urine [ug/g creatinine]
Plasmaconcentration [ng/ml]
Plasmaconcentration [ng/ml]
600
170
900
:
500
45
15
90
300
4.5
45
210
30
3.0
30
140
15
1.5
15
70
0
ACC
ACA
0
0
ACC
ACA
ACC
ACA
ACC
ACA
Andiol-diS
Andiol-S
Andiol-diS
Andiol-S
**
40000
30000
Plasmaconcentration [ng/ml]
15000
Concentration in urine [ug/g creatinine]
21000
Concentration in urine [ug/g creatinine]
Plasmaconcentration [ng/ml]
1600
16000
9500
900
2000
2000
4000
1800
210
3600
250
150
1200-
140
2400
100
600
70-
1200
50
0
0
0
0
ACC
ACA
ACC
ACA
ACC
ACA
ACC
ACA
samples could enable further valuable insights into underlying patho- physiological processes of disease formation and progression e.g, the extent of tumour-associated steroid sulfation and secretion. By directly targeting metabolic products of steroids this methodology enables such investigations, which so far remained beyond reach of methods target- ing deconjugated and unconjugated steroids. As was indicated by the comparative analysis of urine and plasma samples from patients with ACC and ACA, profound studies in this area could result in clinically relevant findings.
Funding sources
This work was funded by the Interdisciplinary Center of Clinical Research (IZKF) at the University Hospital of the University of Erlangen- Nuremberg (junior project J115). It was also supported by the Deutsche Forschungsgemeinschaft (DFG) (project number 314061271, CRC/ Transregio 205/2). The QTRAP 6500+ was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) - 537496341. Open Access funding enabled and organized by Projekt DEAL.
CRediT authorship contribution statement
Eleanor North: Writing - original draft, Visualization, Validation,
Methodology, Investigation, Formal analysis, Data curation, Conceptu- alization. Arne Gessner: Writing - review & editing, Supervision, Re- sources, Conceptualization. Max Kurlbaum: Writing - review & editing, Resources, Conceptualization. Martin Fassnacht: Writing - review & editing, Resources, Conceptualization. Matthias Kroiss: Writing - re- view & editing, Conceptualization. Martin F. Fromm: Writing - review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Nora Bartels: Writing - review & edit- ing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization.
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.
Acknowledgements
The authors thank Daniel Auge for his support during method devel- opment. The authors thank all patients for providing urine and blood samples. The present work was performed in (partial) fulfilment of the requirements for obtaining the academic degree “Dr. rer. nat.” by Eleanor North at the Friedrich-Alexander-Universität Erlangen-Nürnberg.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jpba.2025.117258.
References
[1] U. Mareck, H. Geyer, G. Opfermann, M. Thevis, W. Schänzer, Factors influencing the steroid profile in doping control analysis, J. Mass Spectrom. 43 (2008) 877-891, https://doi.org/10.1002/jms.1457.
[2] K .- H. Storbeck, L. Schiffer, E.S. Baranowski, V. Chortis, A. Prete, L. Barnard, L. C. Gilligan, A.E. Taylor, J. Idkowiak, W. Arlt, C.H.L. Shackleton, Steroid metabolome analysis in disorders of adrenal steroid biosynthesis and metabolism, Endocr. Rev. 40 (2019) 1605-1625, https://doi.org/10.1210/er.2018-00262.
[3] M. Valko-Rokytovská, P. Očenáš, A. Salayová, Z. Kostecká, Breast cancer: targeting of steroid hormones in cancerogenesis and diagnostics, Int. J. Mol. Sci. 22 (2021) 5878, https://doi.org/10.3390/ijms22115878.
[4] L. Schiffer, L. Barnard, E.S. Baranowski, L.C. Gilligan, A.E. Taylor, W. Arlt, C.H. L. Shackleton, K .- H. Storbeck, Human steroid biosynthesis, metabolism and excretion are differentially reflected by serum and urine steroid metabolomes: A comprehensive review, J. Steroid Biochem. Mol. Biol. 194 (2019) 105439, https:// doi.org/10.1016/j.jsbmb.2019.105439.
[5] D. Thieme, P. Anielski, J. Grosse, H. Sachs, R.K. Mueller, Identification of anabolic steroids in serum, urine, sweat and hair: Comparison of metabolic patterns, Anal. Chim. Acta 483 (2003) 299-306, https://doi.org/10.1016/S0003-2670(02)01604- 5.
[6] P.A. Foster, J.W. Mueller, SULFATION PATHWAYS: insights into steroid sulfation and desulfation pathways, J. Mol. Endocrinol. 61 (2018) T271-T283, https://doi. org/10.1530/JME-18-0086.
[7] J.W. Mueller, N. Vogg, T.A. Lightning, I. Weigand, C.L. Ronchi, P.A. Foster, M. Kroiss, Steroid Sulfation in Adrenal Tumors, J. Clin. Endocrinol. Metab. (2021) dgab182, https://doi.org/10.1210/clinem/dgab182.
[8] N. Sun, T. Kunzke, S. Sbiera, S. Kircher, A. Feuchtinger, M. Aichler, S. Herterich, C. L. Ronchi, I. Weigand, N. Schlegel, J. Waldmann, M. Candida Villares Fragoso, T. G. Whitsett, A.J. Gill, M. Fassnacht, A. Walch, M. Kroiss, Prognostic relevance of steroid sulfation in adrenocortical carcinoma revealed by molecular phenotyping using high-resolution mass spectrometry imaging, Clin. Chem. 65 (2019) 1276-1286, https://doi.org/10.1373/clinchem.2019.306043.
[9] N. Vogg, E. North, A. Gessner, F. Fels, M.R. Heinrich, M. Kroiss, M. Kurlbaum, M. Fassnacht, M.F. Fromm, An untargeted metabolomics approach to evaluate enzymatically deconjugated steroids and intact steroid conjugates in urine as diagnostic biomarkers for adrenal tumors, Clin. Chem. Lab. Med. CCLM 63 (2025) 1004-1015, https://doi.org/10.1515/cclm-2024-1337.
[10] I. Bancos, A.E. Taylor, V. Chortis, A.J. Sitch, C. Jenkinson, C.J. Davidge-Pitts, K. Lang, S. Tsagarakis, M. Macech, A. Riester, T. Deutschbein, I.D. Pupovac, T. Kienitz, A. Prete, T.G. Papathomas, L.C. Gilligan, C. Bancos, G. Reimondo, M. Haissaguerre, L. Marina, M.A. Grytaas, A. Sajwani, K. Langton, H.E. Ivison, C.H. L. Shackleton, D. Erickson, M. Asia, S. Palimeri, A. Kondracka, A. Spyroglou, C. L. Ronchi, B. Simunov, D.A. Delivanis, R.P. Sutcliffe, I. Tsirou, T. Bednarczuk, M. Reincke, S. Burger-Stritt, R.A. Feelders, L. Canu, H.R. Haak, G. Eisenhofer, M. C. Dennedy, G.A. Ueland, M. Ivovic, A. Tabarin, M. Terzolo, M. Quinkler, D. Kastelan, M. Fassnacht, F. Beuschlein, U. Ambroziak, D.A. Vassiliadi, M. W. O’Reilly, W.F. Young, M. Biehl, J.J. Deeks, W. Arlt, ENSAT EURINE-ACT Investigators, 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.
[11] S.A. Wudy, G. Schuler, A. Sánchez-Guijo, M.F. Hartmann, The art of measuring steroids, J. Steroid Biochem. Mol. Biol. 179 (2018) 88-103, https://doi.org/ 10.1016/j.jsbmb.2017.09.003.
[12] M.T. Ackermans, E. Endert, LC-MS/MS in endocrinology: what is the profit of the last 5 years? Bioanalysis 6 (2014) 43-57, https://doi.org/10.4155/bio.13.300.
[13] W. Arlt, M. Biehl, A.E. Taylor, S. Hahner, R. Libé, B.A. Hughes, P. Schneider, D. J. Smith, H. Stiekema, N. Krone, E. Porfiri, G. Opocher, J. Bertherat, F. Mantero, B. Allolio, M. Terzolo, P. Nightingale, C.H.L. Shackleton, X. Bertagna, M. Fassnacht, P.M. Stewart, 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.
[14] N. Vogg, T. Müller, A. Floren, T. Dandekar, A. Riester, U. Dischinger, M. Kurlbaum, M. Kroiss, M. Fassnacht, 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.
[15] R.L. Gomes, W. Meredith, C.E. Snape, M.A. Sephton, Analysis of conjugated steroid androgens: deconjugation, derivatisation and associated issues, J. Pharm. Biomed. Anal. 49 (2009) 1133-1140, https://doi.org/10.1016/j.jpba.2009.01.027.
[16] R.L. Gomes, W. Meredith, C.E. Snape, M.A. Sephton, Conjugated steroids: analytical approaches and applications, Anal. Bioanal. Chem. 393 (2009) 453-458, https://doi.org/10.1007/s00216-008-2451-8.
[17] T.A. Lightning, T.F. Gesteira, J.W. Mueller, Steroid disulfates - Sulfation double trouble, Mol. Cell. Endocrinol. 524 (2021) 111161, https://doi.org/10.1016/j. mce.2021.111161.
[18] M. Fassnacht, S. Tsagarakis, M. Terzolo, A. Tabarin, A. Sahdev, J. Newell-Price, I. Pelsma, L. Marina, K. Lorenz, I. Bancos, W. Arlt, O.M. Dekkers, 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.
[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] M. Fassnacht, S. Puglisi, O. Kimpel, M. Terzolo, Adrenocortical carcinoma: a practical guide for clinicians, Lancet Diabetes Endocrinol. 13 (2025) 438-452, https://doi.org/10.1016/S2213-8587(24)00378-4.
[21] Bioanalytical Method Validation Guidance for Industry, (2018)
[22] ich-guideline-m10-bioanalytical-method-validation-step-5_en.pdf, (n.d.). (https:// www.ema.europa.eu/en/documents/scientific-guideline/ich-guideline-m10-bioan alytical-method-validation-step-5_en.pdf) (accessed July 1, 2025).
[23] M. Fassnacht, G. Assie, E. Baudin, G. Eisenhofer, C. de la Fouchardiere, H.R. Haak, R. de Krijger, F. Porpiglia, M. Terzolo, A. Berruti, Adrenocortical carcinomas and malignant phaeochromocytomas: ESMO-EURACAN Clinical Practice Guidelines for diagnosis, treatment and follow-up, Ann. Oncol. 31 (2020) 1476-1490, https:// doi.org/10.1016/j.annonc.2020.08.2099.
[24] M. Fassnacht, W. Arlt, I. Bancos, H. Dralle, J. Newell-Price, A. Sahdev, A. Tabarin, M. Terzolo, S. Tsagarakis, O.M. Dekkers, Management of adrenal incidentalomas: European Society of Endocrinology Clinical Practice Guideline in collaboration with the European Network for the Study of Adrenal Tumors, Eur. J. Endocrinol. 175 (2016) G1-G34, https://doi.org/10.1530/EJE-16-0467.
[25] C.E. Galuska, M.F. Hartmann, A. Sánchez-Guijo, K. Bakhaus, J. Geyer, G. Schuler, K .- P. Zimmer, S.A. Wudy, Profiling intact steroid sulfates and unconjugated steroids in biological fluids by liquid chromatography-tandem mass spectrometry (LC-MS-MS), Analyst 138 (2013) 3792-3801, https://doi.org/10.1039/ C3AN36817C.
[26] A.Y. Dury, Y. Ke, R. Gonthier, M. Isabelle, J .- N. Simard, F. Labrie, Validated LC-MS/MS simultaneous assay of five sex steroid/neurosteroid-related sulfates in human serum, J. Steroid Biochem. Mol. Biol. 149 (2015) 1-10, https://doi.org/ 10.1016/j.jsbmb.2015.01.006.
[27] A. Sánchez-Guijo, V. Oji, M.F. Hartmann, H. Traupe, S.A. Wudy, Simultaneous quantification of cholesterol sulfate, androgen sulfates, and progestagen sulfates in human serum by LC-MS/MS[S], J. Lipid Res. 56 (2015) 1843-1851, https://doi. org/10.1194/jlr.D061499.
[28] R. Wang, M.F. Hartmann, S.A. Wudy, Targeted LC-MS/MS analysis of steroid glucuronides in human urine, J. Steroid Biochem. Mol. Biol. 205 (2021) 105774, https://doi.org/10.1016/j.jsbmb.2020.105774.
[29] G. Balcells, O.J. Pozo, A. Esquivel, A. Kotronoulas, J. Joglar, J. Segura, R. Ventura, Screening for anabolic steroids in sports: Analytical strategy based on the detection of phase I and phase II intact urinary metabolites by liquid chromatography tandem mass spectrometry, J. Chromatogr. A 1389 (2015) 65-75, https://doi.org/ 10.1016/j.chroma.2015.02.022.
[30] L. De Wilde, K. Roels, P. Van Renterghem, P. Van Eenoo, K. Deventer, Steroid profiling in urine of intact glucuronidated and sulfated steroids using liquid chromatography-mass spectrometry, J. Chromatogr. A 1624 (2020) 461231, https://doi.org/10.1016/j.chroma.2020.461231.
[31] A. Esquivel, É. Alechaga, N. Monfort, R. Ventura, Direct quantitation of endogenous steroid sulfates in human urine by liquid chromatography-electrospray tandem mass spectrometry, Drug Test. Anal. 10 (2018) 1734-1743, https://doi. org/10.1002/dta.2413.
[32] G. Forsdahl, K. Zanitzer, D. Erceg, G. Gmeiner, Quantification of endogenous steroid sulfates and glucuronides in human urine after intramuscular administration of testosterone esters, Steroids 157 (2020) 108614, https://doi.org/ 10.1016/j.steroids.2020.108614.
[33] F. Ponzetto, M. Parasiliti-Caprino, I. Gesmundo, L. Marinelli, A. Nonnato, R. Nicoli, T. Kuuranne, G. Mengozzi, E. Ghigo, F. Settanni, Single-run UHPLC-MS/MS method for simultaneous quantification of endogenous steroids and their phase II metabolites in serum for anti-doping purposes, Talanta 255 (2023) 124218, https://doi.org/10.1016/j.talanta.2022.124218.
[34] J.M. Hines, I. Bancos, C. Bancos, R.D. Singh, A.V. Avula, W.F. Young, S.K. Grebe, R. J. Singh, 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.
[35] D.R. Taylor, L. Ghataore, L. Couchman, R.P. Vincent, B. Whitelaw, D. Lewis, S. Diaz-Cano, G. Galata, K .- M. Schulte, S. Aylwin, N.F. Taylor, 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.
[36] F. Ponzetto, J. Boccard, R. Nicoli, T. Kuuranne, M. Saugy, S. Rudaz, Steroidomics for highlighting novel serum biomarkers of testosterone doping, Bioanalysis 11 (2019) 1169-1185, https://doi.org/10.4155/bio-2019-0079.