Analysis of Corticosteroids and Androgens in Serum for Clinical Research

An analytically sensitive and selective clinical research method has been developed for the analysis of testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol in serum using the Xevo TQ-S micro.

The Xevo TQ-S micro enables this method to provide sufficient analytical sensitivity to analyze low levels of testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol by using only 100 μL sample volume. Excellent levels of imprecision across the calibration range have been demonstrated. Assessment of EQA samples has shown the method provides excellent agreement to the EQA LC-MS/MS mean for testosterone, androstenedione, 17-OHP, DHEAS, and cortisol. Automation of the analytical method in combination with sample tracking capabilities of the liquid handler, through the use of the Tecan File Converter and MassLynx LIMS Interface, improves laboratory workflow and reduces sample handling, which alleviates the potential for operator error.

Benefits

• Analytical selectivity of the chromatographic method provides separation of isobaric species
• LC-MS/MS enables high samplethroughput using multi-well plate automation
• Excellent agreement to the EQA Mass spectrometry mean for testosterone, androstenedione, 17-OHP, DHEAS, and cortisol

Steroid hormones encompass a large class of small molecules that play a central role in metabolic processes, such as regulation of sexual characteristics, blood pressure and inflammation. Enzymes that form part of the steroid biosynthetic pathway are pivotal in these metabolic processes and their dysfunction can be examined through measurement of steroid hormones in the pathway. Measurement of these steroids by immunoassay can be prone to analytical interference as a result of cross reactivity of reagent antibodies with structurally related steroid hormones and synthetic derivatives. Liquid chromatography – tandem mass spectrometry (LC-MS/MS) can provide analytically sensitive, accurate, and precise measurement of these steroid hormones.

Here we describe a clinical research method utilizing Oasis PRiME HLB μElution Plate technology for the extraction of testosterone, androstenedione, 17-hydroxyprogesterone (17-OHP), dehydroepiandrosterone sulfate (DHEAS), cortisol, 11-deoxycortisol and 21-deoxycortisol from serum, which has been automated using the Tecan Freedom EVO 100/4 liquid handler. Chromatographic separation was performed on an ACQUITY UPLC I-Class System using an ACQUITY UPLC HSS T3 VanGuard Pre-column and ACQUITY UPLC HSS T3 Column, followed by detection on a Xevo TQ-S micro Mass Spectrometer (Figure 1). In addition, we have examined External Quality Assessment (EQA) samples for testosterone, androstenedione, 17-OHP, DHEAS and cortisol to evaluate the bias and therefore suitability of the method for analyzing these steroids for clinical research.

Data Management

MassLynx v4.1 Software with TargetLynx Application Manager

Sample preparation

Testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol and 21-deoxycortisol certified reference solution and stable labeled internal standards for testosterone, androstenedione, 17-OHP, DHEAS, cortisol, and 11-deoxycortisol were purchased from Sigma Aldrich (Poole, UK). 21-deoxycortisol stable labeled internal standard was purchased from Isosciences (King of Prussia, PA).

Calibrators and Quality Controls (QCs) were prepared in a surrogate matrix of MSG4000 stripped human serum purchased from Golden West Biologicals (Temecula, CA). Testosterone calibrators were prepared over the range of 0.10–69 nmol/L, with QCs at 0.35 nmol/L, 3.5 nmol/L, and 49 nmol/L. Androstenedione calibrators were prepared over the range of 0.09–349 nmol/L, with QCs at 0.52, 17.5, and 245 nmol/L. 17-OHP calibrators were prepared over the range of 0.19–757 nmol/L, with QCs at 0.76 nmol/L, 38 nmol/L, and 530 nmol/L. DHEAS calibrators were prepared over the range of 0.065–43 μmol/L with QCs at 0.41 μmol/L, 2.2 μmol/L, and 30 μmol/L. Cortisol calibrators were prepared over the range of 0.69–1380 nmol/L, with QCs at 0.83 nmol/L, 69 nmol/L, and 966 nmol/L. 11-deoxycortisol and 21-deoxycortisol calibrators were prepared over the range of 0.72–144 nmol/L. QC concentrations were 1.44, 7.2, and 101 nmol/L for 11-deoxycortisol and 2.2, 7.2, and 101 nmol/L for 21-deoxycortisol.

To convert SI units to conventional mass units divide by 3.470 for testosterone (nmol/L to ng/mL), 3.494 for androstenedione (nmol/L to ng/mL), 3.028 for 17-OHP (nmol/L to ng/mL), 2.716 for DHEAS (nmol/L to ng/mL), 2.761 for cortisol (nmol/L to ng/mL), and 2.889 for 11-deoxycortisol and 21-deoxycortisol (nmol/L to ng/mL).

Sample extraction

Extraction was performed using a liquid handler. To 100 μL of sample; 25 μL of 28 nmol/L testosterone-¹³C₃ and androstenedione-13C3, 75 nmol/L 17-OHP-¹³C₃, 1.4 μmol/L DHEAS-²H₅, 137 nmol/L Cortisol-¹³C₃, 71 nmol/L 11-deoxycortisol-¹³C₃, and 21-deoxycortisol-²H₄, 200 μL methanol, and 550 μL water were added. The samples were mixed after each reagent addition. Samples were centrifuged for five minutes at 4000 g.

An aliquot of each of the pre-treated samples (600 μL) was loaded into individual wells of the Oasis PRiME HLB μElution Plate and slowly pulled through at low vacuum (100 mbar). Consecutive washes with 150 μL of 0.1% (v/v) ammonia in 35% (v/v) methanol(aq) and 150 μL 0.1% (v/v) formic acid in 35% (v/v) methanol(aq) were performed to reduce potential ionic interference. Analytes were eluted using 30 μL of 85/15 (v/v) acetonitrile/methanol, followed by addition of 70 μL water.

The chromatographic selectivity of the column is demonstrated through the baseline resolution of isobaric steroid species; 11-deoxycortisol, corticosterone, and 21-deoxycortisol; 17-OHP and 21-OHP; testosterone and epitestosterone (Figure 2). No interferences were observed at the retention time of testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol when each of these steroids and ten other structurally related compounds were individually examined (21-hydroxyprogesterone, corticosterone, DHEA, epitestosterone, dihydrotestosterone, aldosterone, cortisone, 18-hydroxycorticosterone, 17-hydroxypregnenolone, and prednisone). It has been noted, at high prednisolone concentrations (>0.5 μmol/L), >20% bias is observed at low concentrations (20 nmol/L) of cortisol. This is related to the isotopic contribution of prednisolone to the cortisol MRM trace. This interference can be detected through a change in the ion ratio of the quantifier and qualifier ions. No significant interference (≤10% bias) for testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol were observed when other endogenous compounds were examined (albumin, bilirubin, uric acid, intralipid, triglycerides, and cholesterol).

No system carryover was observed from high concentration samples into subsequent blank injections. A 1:5 dilution was successfully employed on high concentration samples, providing a mean accuracy of 102%, 98%, 99%, 96%, 94%, 107%, and 97% for testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol, respectively, with RSDs <7%.

The method would allow for precise quantification (<20% RSD) at 0.10 nmol/L for testosterone, 0.09 nmol/L for androstenedione, 0.19 nmol/L for 17-OHP, 0.65 μmol/L for DHEAS, 0.69 nmol/L for cortisol, and 0.72 nmol/L for 11-deoxycortisol and 21-deoxycortisol. Signal:noise at these concentrations in stripped serum were >10:1 on 10 separate occasions.

Total precision was determined by extracting and quantifying three replicates of tri-level QC material on two occasions per day over five separate days (n=30). Repeatability was assessed by analyzing three replicates at each QC level. Low, mid, and high concentrations were 0.35 nmol/L, 3.5 nmol/L, and 49 nmol/L for testosterone; 0.52, 17.5, and 245 nmol/L for androstenedione; 0.76 nmol/L, 38 nmol/L, and 530 nmol/L for 17-OHP; 0.41 μmol/L, 2.2 μmol/L, and 30 μmol/L for DHEAS; 0.83 nmol/L, 69 nmol/L and 966 nmol/L for cortisol; 1.44, 7.2 and 101 nmol/L for 11-deoxycortisol and 2.2, 7.2 and 101 nmol/L for 21-deoxycortisol. As seen in Table 3, total QC precision and repeatability for the steroid hormones was ≤ 7.6%.

The method was shown to be linear for testosterone (0.10–69 nmol/L), androstenedione (0.09–349 nmol/L), 17-OHP (0.19–757 nmol/L), DHEAS (0.065–43 μmol/L), cortisol (0.69–1380 nmol/L), and 11-deoxycortisol and 21-deoxycortisol (0.72–144 nmol/L) when different ratios of high and low concentration pools of the analytes were combined and analyzed. In addition, calibration lines in spiked serum were linear with coefficient of determinations (r2) >0.994 for all analyses.

Matrix effect investigations for testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol were performed using individual donor serum samples (n=6). The matrix factor calculated is shown in Table 4. Normalized matrix factor calculations, based on the analyte:internal standard response ratio demonstrated that the internal standards compensated for any ion suppression observed.

Accuracy was assessed for testosterone, androstenedione, 17-OHP, DHEAS, and cortisol through the analysis of EQA samples. The data obtained was compared to the mass spectrometry method mean for the samples and Deming regression, Altman-bland agreement and Linear regression was performed (Table 5). No statistically significant proportional or constant bias was observed for testosterone and cortisol. Statistically significant proportional bias was observed for androstenedione and 17-OHP and statistically significant constant bias was observed for DHEAS. Altman-Bland agreement for testosterone, androstenedione, 17-OHP, DHEAS, and cortisol demonstrated a mean method bias within }5.8%, demonstrating excellent agreement with the EQA mass spectrometry method mean for the steroid hormones (Figures 3A–E).

An analytically sensitive and selective clinical research method has been developed for the analysis of testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol in serum using the Xevo TQ-S micro.

The Xevo TQ-S micro enables this method to provide sufficient analytical sensitivity to analyze low levels of testosterone, androstenedione, 17-OHP, DHEAS, cortisol, 11-deoxycortisol, and 21-deoxycortisol by using only 100 μL sample volume. Excellent levels of imprecision across the calibration range have been demonstrated. Assessment of EQA samples has shown the method provides excellent agreement to the EQA LC-MS/MS mean for testosterone, androstenedione, 17-OHP, DHEAS, and cortisol. Automation of the analytical method in combination with sample tracking capabilities of the liquid handler, through the use of the Tecan File Converter and MassLynx LIMS Interface, improves laboratory workflow and reduces sample handling, which alleviates the potential for operator error.