Performance enhancement in the measurement of 5 endogenous steroids by LC-MS/MS combined with differential ion mobility spectrometry

Background: Challenges for steroid analysis by LC-MS/MS include low ionization efficiency, endogenous isobars with similar fragmentation patterns and chromatographic retention. Differential ion mobility spectrometry (DMS) provides an additional degree of separation prior to MS/MS detection, and shows promise in improving specificity of analysis. We developed a sensitive and specific method for measurement of corticosterone, 11-deoxycortisol, 11-deoxycorticosterone, 17-hydroxyprogesterone and progesterone in human serum and plasma using an ABSciex 5500 mass spectrometer equipped with a differential ion mobility interface. Methods: 250 μL aliquots of serum were spiked with deuterated internal standards and extracted with MTBE. The samples were analyzed using positivemode electrospray LC-DMS-MS/MS. Themethod was validated and compared with immunoassays and LC-MS/MS methods of reference laboratories. Results: Inter and intra assay imprecision was b10%. Limits of quantification and detection in nmol/L were 0.18, 0.09 for corticosterone and 17-hydroxyprogesterone, 0.30, 0.16 for 11-deoxycortisol, 0.12, 0.06 for progesterone and 0.06, 0.03 for 11-deoxycorticosterone. Comparison for progesterone and 17-hydroxyprogesterone with immunoassay showed slopes of 0.97 and 1.0, intercepts of 0.16 and 0.10 and coefficients of determination (r2) of 0.92 and 0.97, respectively. Progesterone by immunoassay showed positive bias in samples measuring b3.18 nmol/L. Reference intervals for progesterone and 11-deoxycorticosterone in post-menopausal women were found to be b2.88 and b0.28 nmol/L respectively. Conclusions: We developed and validated an LC-DMS-MS/MS method for analysis of five endogenous steroids suitable for routine measurements in clinical diagnostic laboratories. Specificity gained with DMS allows reducing the complexity of sample preparation, decreasing LC run times and increasing speed of the analysis.

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier's archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/authorsrights Author's personal copy Performance enhancement in themeasurement of 5 endogenous steroids by LC-MS/MS combined with differential ion mobility spectrometry Julie A. Ray a,⁎, Mark M. Kushnir a, Richard A. Yost b, Alan L. Rockwood a,c, A.WayneMeikle a,c,d a ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, UT, United States b Department of Chemistry, University of Florida, Gainesville, FL, United States c Department of Pathology, University of Utah, Salt Lake City, UT, United States d Department of Medicine, University of Utah, Salt Lake City, UT, United States a r t i c l e i n f o a b s t r a c t Article history: Received 16 April 2014 Received in revised form 25 July 2014 Accepted 28 July 2014 Available online 9 August 2014 Keywords: Liquid chromatography-tandem mass spectrometry (LC-MS/MS) Steroids Differential ion mobility spectrometry (DMS) Immunoassay (IA) Chemiluminiscent immunoassay (CIA) Congenital adrenal hyperplasia (CAH) Background: Challenges for steroid analysis by LC-MS/MS include low ionization efficiency, endogenous isobars with similar fragmentation patterns and chromatographic retention. Differential ion mobility spectrometry (DMS) provides an additional degree of separation prior to MS/MS detection, and shows promise in improving specificity of analysis. We developed a sensitive and specific method for measurement of corticosterone, 11-deoxycortisol, 11-deoxycorticosterone, 17-hydroxyprogesterone and progesterone in human serum and plasma using an ABSciex 5500 mass spectrometer equipped with a differential ion mobility interface. Methods: 250 μL aliquots of serum were spiked with deuterated internal standards and extracted with MTBE. The samples were analyzed using positivemode electrospray LC-DMS-MS/MS. Themethod was validated and compared with immunoassays and LC-MS/MS methods of reference laboratories. Results: Inter and intra assay imprecision was b10%. Limits of quantification and detection in nmol/L were 0.18, 0.09 for corticosterone and 17-hydroxyprogesterone, 0.30, 0.16 for 11-deoxycortisol, 0.12, 0.06 for progesterone and 0.06, 0.03 for 11-deoxycorticosterone. Comparison for progesterone and 17-hydroxyprogesterone with immunoassay showed slopes of 0.97 and 1.0, intercepts of 0.16 and 0.10 and coefficients of determination (r2) of 0.92 and 0.97, respectively. Progesterone by immunoassay showed positive bias in samples measuring b3.18 nmol/L. Reference intervals for progesterone and 11-deoxycorticosterone in post-menopausal women were found to be b2.88 and b0.28 nmol/L respectively. Conclusions: We developed and validated an LC-DMS-MS/MS method for analysis of five endogenous steroids suitable for routine measurements in clinical diagnostic laboratories. Specificity gained with DMS allows reducing the complexity of sample preparation, decreasing LC run times and increasing speed of the analysis. © 2014 Elsevier B.V. All rights reserved. 1. Introduction Congenital adrenal hyperplasia (CAH) refers to a group of genetic disorders of steroid biosynthesis in the adrenal glands [1,2]. They are caused by the inability of the glands to produce enzymes required for biosynthesis of several steroid hormones, such as glucocorticoids, mineralocorticoids and sex hormones. Simultaneous measurement of corticosterone, 11-deoxycortisol (11-DOC-ol), 11-deoxycorticosterone (DOC) and 17-hydroxy progesterone (17-OHP) is useful in the diagnosis of these defects and monitoring of patients. Corticosterone is produced from DOC by the 11β-hydroxylase path-way. It is further converted to 18-hydroxy corticosterone and finally to the important mineralocorticoid, aldosterone. Measurement of cortico-sterone along with 11-DOC-ol and DOC is used in the diagnosis of CYP11B1 deficiency, hyperaldosteronism and 11β-hydroxylase deficiency [3,4]. DOC is also used in the diagnosis of suspected 11β-hydroxylase deficiency, differential diagnosis of 11β-hydroxylase 1 (CYP11B1) versus 11β-hydroxylase 2 (CYP11B2) deficiency and in the diagnosis of gluco-corticoid responsive hyperaldosteronism [5,6]. Measurement of 17-OHP along with cortisol and androstenedione is useful for the diagnosis of 11- or21-hydroxylasedeficiencies [3,7]. Progesterone primarily produced in the placenta and corpus luteum, is used to monitor placental function during pregnancy and ovulation during the menstrual cycle. It is also used to evaluate patients with adrenal or testicular tumors [8-11]. The steroidsmeasured in this assay have unique biological properties, while being closely structurally related (corticosterone: 11-DOC-ol and DOC: 17-OHP being isomeric pairs). Differentiation of these pairs by LC-MS/MS alone is difficult to achieve due to similar fragmentation Clinica Chimica Acta 438 (2015) 330-336 Abbreviations: 11-DOC-ol, 11-deoxycortisol; DOC, 11-deoxycorticosterone; 17-OHP, 17-hydroxy progesterone; DMS, Differential ion mobility spectrometry; IA, Immunoassay; CIA, Chemiluminiscent immunoassay; CAH, Congenital adrenal hyperplasia; LLE, liquid- liquid extraction; MTBE, Methyl-tert-butyl ether. ⁎ Corresponding author at: ARUP laboratories 500 Chipeta Way, Salt Lake City, UT 84108-1221, United States. E-mail address: [e-mail] (J.A. Ray). http://dx.doi.org/10.1016/j.cca.2014.07.036 0009-8981/© 2014 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim UU IR Author Manuscript UU IR Author Manuscript University of Utah Institutional Repository Author Manuscript Author's personal copy patterns and chromatographic retentions. The last analyte on the list, namely progesterone is the most hydrophobic and elutes from the chro-matographic column at very high percentage of organic modifier, along with strongly retained and unknown interferences from patient samples. Failing to resolve these co-eluents in samples with low concentrations of progesterone could lead to falsely elevated results, which is especially critical in post-menopausal women [12]. High specificity and sensitivity are essential for diagnosis of the above mentioned disorders. Immunoassays (IA) which are commonly used for steroid analysis are replete with inaccuracies for lack of sensitivity at low concentrations as well as poor selectivity due to cross reactivity with antibodies andmatrix effects [13-18]. LC-tandem mass spectrome-try is increasingly the choice of methodology due to better accuracies by virtue of improved sensitivity and specificity [19-23]. In addition to LC-MS/MS we evaluated the use of differential ion mobility spectrome-try towards improving the performance of the assay. DMS (or field asymmetric ion mobility spectrometry, FAIMS) is based on gas-phase separations at atmospheric pressures and ambient temperature [24-35]. The Selexion equipped on an ABSciex 5500 mass spectrometer consists of a DMS cell comprised of 2 flat parallel plates (10 × 30 mm, separated by 1 mm). An asymmetric RF waveform (separation voltage, SV) is applied to the plates, causing the ions to oscillate, with different ionic species drifting with different velocities towards one plate or the other. A DC potential (compensation voltage, CoV) is also applied to compensate for the net drift of the analyte ions where the ions of the targeted analytes get resolved from interfering ionic species (Supple-mental Fig. 1, courtesy Richard A. Yost, University of Florida). The aim of this work was to develop a highly sensitive and robust method for quantification of the above 5 endogenous steroids in serum and plasma samples and to evaluate its performance. The novel method thatwe developed uses DMSin conjunctionwithMS/MS detec-tion as a way of enhancing specificity of analysis and reducing the complexity of sample preparation [31,36-39]. The method has been fully validated and applied to the analysis of clinical samples. 2. Materials and methods 2.1. Reagents, standards and patient samples Corticosterone, progesterone and d9-progesterone were purchased from Cerilliant (Sigma), 11-DOC-ol, DOC, 17-OHP from Sigma, d8-DOC and d8-corticosterone from CDN Isotopes and d2-11-DOC-ol and d8-17-OHP from Cambridge Isotope Laboratories, Inc. Stock standards were prepared inmethanol at concentrations of 1 g/l. Aworking calibra-tion standard of the analytes was prepared at 144.5 nmol/l for cortico-sterone and 11-DOC-ol and 151.5, 159.0, 15.1 nmol/l for 17-OHP, progesterone and DOC respectively in 1:1 methanol: water. Com-bined working internal standards were prepared at 72.2 nmol/l for d8-corticosterone and d2-11-DOC-ol and 75.1, 79.5, 18.9 nmol/l for d8-17-OHP, d9-progesterone and d8-DOC respectively in 1:1 methanol: water. Calibration standards were prepared in 0.05% BSA at nmol/l concentrations of 1.4, 2.8, 14.4, 28.9, 57.8, 86.7 for corticosterone and 11-DOC-ol, 1.5, 3.0, 15.1, 30.3, 60.6, 90.9 for 17-OHP, 1.5, 3.1, 15.9, 31.8, 63.6, 95.4 for progesterone and 0.1, 0.3, 1.5, 3.0, 6.0, 9.0 for DOC.Water, methanol,methyl t-butyl ether (MTBE), iso-propyl alcohol, acetone, ace-tonitrile and tri-fluoro acetic acid were purchased from VWR (Radnor, PA). Serum and plasma samples used in the assay were deidentified dis-card samples submitted to ARUP Laboratories for routine analysis. All studies with human serum and plasma samples were approved by IRB of the University of Utah. 2.2. Sample preparation Aliquots of 250 μl of calibrators, controls and patient serum/plasma samples were transferred into 2 ml polypropylene microcentrifuge tubes. 20 μL of combined working internal standard and 1.5 ml MTBE were added to each tube, shaken and centrifuged. The organic layer was transferred to a 96-well plate and evaporated under nitrogen at 50 °C. The residues were reconstituted using 100 μl of 1:1 methanol: water. The plate was shaken, centrifuged for 5 min at 4000 ×g and the samples were analyzed by LC-MS/MS. 2.3. LC-MS/MS Chromatographic separation was performed on a Kinetex C18, 50 × 3 mm, 2.6 μm particle HPLC column (Phenomenex) fitted with a Phenomenex Ultra security guard column (C18, 3 mm). The injection volume was 30 μl and oven temperature was set to 50 °C. The mobile phase consisted of 10 mmol/l formic acid in water and 10 mmol/l formic acid in acetonitrile and was delivered at 1 ml/min with a linear gradient 20% to 48% of organic in 3.5 min, followed by a gradient to 75% organic in 1.6 min. The column was conditioned with 98.5% organic for 1 min and then equilibrated to initial conditions. Total analysis time per sample was 6 min. The HTC PAL autosampler (LEAP Technologies,) injection sy-ringe was washed twice with methanol: water (1:1) with 10 mmol/l formic acid, and 45% acetonitrile, 45% IPA, 9.4% acetone and 0.6% TFA. Quadrupoles Q1 and Q3 were tuned to unit resolution and the MS pa-rameters optimized for maximum signal intensity for each mass transi-tion. The instrument was operated with electrospray ionization in positive mode; ion-spray voltage was 5500 V, gases 1, 2 and curtain gas were 60, 50 and 20, respectively; entrance potential (EP) of 10 V, ion source temperature of 500 °C and a separation voltage (SV) of the Selexion of 4000 V. The declustering potential (DP), collision energies (CE), exit po-tentials (CXP) and compensation voltages (CoV) of the 2 monitored MRMs for each analyte are shown in Table 1. The ratio of primary (1) to secondarymass transition (2) was used to evaluate specificity of the anal-ysis. Quantitative calibration was performed with each batch of samples; data acquisition and processing was performed with AnalystTM 1.5.2. 2.4. Assay performance characteristics Performance of the assaywas assessed based on imprecision, limit of detection (LOD), limit of quantification (LOQ), upper limit of linearity (ULOL),method comparison, extraction recovery, carryover, interference and ion suppression studies. Imprecision of the assay was determined by analyzing three replicates of human serum sample pools (low, medium and high) with concentrations of the analytes ranging between 0.1 and 181.8 nmol/l in one run per day over a period of 20 days. These were also used as the quality control samples. LOQ was determined by ana-lyzing 8 samples in triplicate over 6 days; the samples contained Table 1 Mass transitions and corresponding optimized voltages used in the method. Mass transitions Q1(Da) Q3(Da) DP(V) CE(V) CXP(V) CoV(V) Corticosterone-1 347.3 121.1 90 30 15 5 Corticosterone-2 347.3 91.1 90 65 15 5 d8-Corticosterone-1 355.2 125.1 80 41 25 5 d8-Corticosterone-2 355.2 95.1 80 71 25 5 11-DOC-ol-1 347.2 109.1 90 35 10 5.8 11-DOC-ol-2 347.2 97.1 90 31 10 5.8 d2-11-DOC-ol-1 349.3 109.1 140 33 16 5.8 d2-11-DOC-ol-2 349.3 97.1 140 32 10 5.8 DOC-1 331.2 109.1 100 33 13 4.8 DOC-2 331.2 97.1 100 29 11 4.8 d8-DOC-1 339.2 113.1 100 32 25 4.8 d8-DOC-2 339.2 100.1 100 29 25 4.8 17-OHP-1 331.2 109 150 35 13 6 17-OHP-2 331.2 97.1 150 30 11 6 d8-17-OHP-1 339.2 113.1 150 37 13 6 d8-17-OHP-2 339.2 100.1 150 30 11 6 Progesterone-1 315.1 109 90 32 13 5.8 Progesterone-2 315.1 97.1 90 30 11 5.8 d9-progesterone-1 324.2 100.2 90 28 15 5.8 d9-progesterone-2 324.2 113.1 90 31 15 5.8 J.A. Ray et al. / Clinica Chimica Acta 438 (2015) 330-336 331 UU IR Author Manuscript UU IR Author Manuscript University of Utah Institutional Repository Author Manuscript Author's personal copy progressively lower concentrations of the analytes, and were prepared by mixing serum pools containing high and low concentration of the analytes. A stripped serum which had negligible concentrations of all 5 analytes was used to dilute the high sample pool. The lowest concentration for which precision was within 15% and observed concentrations were within 20% of the expected value, was set as the lower limit of quantitation (LLOQ). Limit of detection (LOD) was deter-mined as the lowest concentration at which the peaks of the analyte were present in both mass transitions at the expected retention time and signal to noise ratio for the quantitative mass transition was ≥5. Linearity was evaluated by analyzing seven samples in triplicate over a period of six days. The highest concentration at which precision was within 10% and accuracy was within 20% of the expected values was considered to be the upper limit of linearity (ULOL) of the method. Over 400 patient sampleswere analyzed during the method evaluation. 2.5. Method comparison The method was compared with commercial IAs and LC-MS/MS methods of referral laboratories. Comparison with Radioimmunoassay (RIA) was performed for 17-OHP (n=27); with chemiluminiscent im-munoassay (CIA) for progesterone (n = 324) and with LC-MS/MS methods for 17-OHP (n = 32), 11-DOC-ol (n = 31), DOC (n = 20), corticosterone (n = 50) and progesterone (n = 20). The results for method comparison and bias estimation were evaluated using Deming regression. 2.6. Method recovery and carryover Method recoverywas evaluated by standard addition of the analytes at 2.89 nmol/l for corticosterone, 11-DOC-ol, 17-OHP and progesterone and 0.30 nmol/l for DOC in patient samples (n = 5) and analyzed in duplicate over a period of 2 days. Recovery was estimated from the difference between the observed and the expected concentrations of the analytes. Carryover potential for the method was evaluated by injecting negative controls after samples containing 3.0 to 909.0 nmol/l of the targeted analytes. 2.7. Interference and Ion suppression Twenty-three steroids and steroid metabolites were analyzed using the method to evaluate possible interferences (Supplemental Table 1). Ion suppression was evaluated by analyzing extracted serum samples with standards of the analytes infused with a syringe pump at a flow rate 0.6 ml/h. Concentration of the 5 analytes infused into HPLC effluent was 50.0 nmol/l; concentration of the targeted analytes in the serum samples was less than 0.03 nmol/l. A drop in the baseline in the MRM transitions evidenced ion suppression [40]. 2.8. Quality controls Four controls (negative, low(LQC),medium(MQC) and high (HQC)) analyzed over a period of 20 dayswere prepared in patient serum pools by spiking with standards. Acceptability of runs was based on retention times of the analytes, concentrations of the analytes beingwithin 20% of historical values and ratios of the primary to the secondary mass transi-tions N30% being interpreted as presence of interference. Negative controls were considered acceptable only when concentrations of the analytes were below LOQ of the method. 2.9. Sample stability and suitability Stability of the analytes was evaluated over a period of 30 days at three storage conditions. A sample containing 14.4, 10.6, 12.8, 4.9 and 1.1 nmol/l of corticosterone, 11-DOC-ol, 17-OHP, progesterone and DOC, respectively, was stored at room temperature (RT), 4 °C and −20 °C. The tubes were placed in a −70 °C freezer after 1, 3, 7, 14, 21 and 30 days of storage and analyzed in a single batch. Concentration of the analytes in different collection tubes was eval-uated in duplicate by using samples from 5 individuals collected in six types of collection tubes: Li-Heparin, PST, SST, Na-Heparin, Serum and K2EDTA. Concentrations in the samples were compared within same individuals. 2.10. Reference interval study for progesterone and DOC in post-menopausal women Serum samples were collected from 125 post-menopausal women (ages 55-89 y;mean 61 y).Whole bloodwas collected in SST vacutainer tubes and allowed to clot at room temperature for 30 min. Tubes were centrifuged at 3000 rpm for 10 min, aliquoted, and frozen immediately at −80 °C. The samples were analyzed for progesterone and DOC. Sta-tistical data analysis was performed using EP evaluator (v. 9.0; David G. Rhoads Associates, Inc.). 3. Results A representative chromatogram of the primaryMRM transition of the analytes in an extract from a patient sample is shown in Fig. 1; three periods are shown in separate traces, (A, 0-2.8 min; B, 2.8-3.9 min; and C, 3.9-6 min). The LLOQs (LODs) of the method were 0.18 (0.09), 0.33 (0.16), 0.18 (0.09), 0.12 (0.06), 0.06 (0.03) nmol/l for corticosterone, 11-DOC-ol, 17 OHP, progesterone and DOC, respectively. Linearity of the method was found to be 289, 116, 4545, 191, 91 nmol/l for cortico-sterone, 11-DOC-ol, 17 OHP progesterone and DOC, respectively. Values of the within-run, between-run and total imprecision are shown in Table 2. The results of method comparison are presented in Fig. 2A-F. Method recovery was found to be greater than 95% for all analytes. No carryover was detected in the blanks and solvents injected after samples containing very high concentrations of the analytes. None of the steroids and steroid metabolites evaluated for potential interference produced peaks at the retention times of the analytes of interest. There was no ion suppression at the retention times of the analytes (Supple-mental Fig. 2). The 3 controls analyzed with every batch of samples showed ≤20% imprecision and the ratios of the primary and secondary mass transi-tions for the analytes in the controls were within the expected limits. Supplemental Fig. 3 shows results for the three controls per analyte analyzed over a period of one month. The % CV for LQC, MQC and HQC were 13, 11 and 11 for corticosterone, 11, 11 and 11 for DOC, 11, 8 and 8 for 11-DOC-ol, 10, 9 and 9 for 17-OHP and 7, 10 and 8 for proges-terone evidencing ruggedness of the assay. The 5 analytes, corticosterone, 11-DOC-ol, DOC, 17-OHP and proges-terone were found to undergo 12, 21, 41, 15 and 32% degradation, respec-tively, over a period of 4 weekswhen stored at room temperature. Except DOCwhich degraded by 20% at the end of 4 weeks at 4 °C, other analytes were stable under refrigeration for a month. No degradation was observed in samples stored at −20 °C and −70 °C, and after 3 freeze thaw cycles. Specimen type evaluation showed that both serum and plasma samples were acceptable for the test. Concentrations observed in the patient samples collected in various collection tubes was found to be b10%. Reference intervals of progesterone and DOC in post-menopausal women (n = 125) were established using nonparametric method as the central 95% of the distribution and determined to be b2.88 and b0.28 nmol/l respectively (Fig. 3). 4. Discussion In evaluating the ion mobility technology towards improving the efficiency of the method, we approached the development of the assay from 2 directions: simplification of sample preparation and 332 J.A. Ray et al. / Clinica Chimica Acta 438 (2015) 330-336 UU IR Author Manuscript UU IR Author Manuscript University of Utah Institutional Repository Author Manuscript Author's personal copy enhancement of specificity. Earlier published method for 11-DOC-ol, 17-OHP, pregnenolone and 17-hydroxypregnenolone used 2 extrac-tions (SPE and LLE) in conjunction with derivatization, for enhancing sensitivity and specificity of the assay [19]. 4.1. Extraction, ionization, HPLC separation During evaluation of the method we observed an overall reduction in the signal of all the analytes by approximately 5 times by the use of DMS. However an efficient extractionmethod such a LLE which provided higher signal of the analytes as well as higher background noise of the chromatograms was a trade-off for the loss of the signal intensity due to the use of DMS. Comparison of the ionization efficiencies between ESI and APCI modes showed approximately 2-fold greater sensitivity for DOC, 11-DOC-ol, 17-OHP and progesterone using ESI while corticosterone was more efficiently ionized using APCI. Aside from the differences in ionization efficiencies among the analytes, HPLC separation also pre-sented a problem. Progesterone eluted late in the gradient along with several co-eluting peaks which posed a problem of poor specificity especially in samples with low concentration. LLE in conjunctionwith ESI boosted the signal of the analytes and the use of DMS overcame the challenges of selectivity and high background noise, resulting in an assay with simplified sample preparation. It also permitted combining all the analytes into a single assay. 4.2. Separation of isomers using DMS Among the 5 steroids, there are 2 pairs of isomers: corticosterone and 11-DOC-ol (sharing mass transitions m/z 347/96 and 347/91); and DOC and 17-OHP (sharing mass transitions m/z 331/109 and 331/97). As a way of enhancing specificity we evaluated the use of DMS for resolving peaks of the analytes from coeluting interfering substances. Optimal SV for all analytes was 4000 V; optimal CoVs were determined by ramping the DC voltage over a range of −100 to +100 V. While DOC and 17-OHP were well separated by DMS (CoVs of 4.8 V and 6.0 V) and could be resolved chromatographically, baseline separation of 11-DOC-ol and corticosterone by LC alonewas somewhat inadequate. To enhance the selectivity for these 2 analytes, we employed CoV of 5 V for corticoste-rone, and 5.75 V for 11-DOC-ol (Supplemental Fig. 4). Combination of the partial DMS separation and partial chromatographic separation allowed better resolution of the isomers than could be achieved by either tech-nique alone. 6.0e5 d9-progesterone Progesterone 1.5e4 Intensity, cps d8-17-OHP 17-OHP d8-DOC DOC 3.9 4.1 4.3 4.5 4.7 4.9 5.1 5.3 5.5 5.7 5.9 2.80 2.90 3.00 3.10 3.20 3.30 3.40 3.50 3.60 3.70 3.80 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 Time, min 5.0e4 Corcosterone d8-corcosterone d2-11-DOC-ol 11-DOC-ol A B C Fig. 1. Representative chromatogram (separated into three time periods) of patient serum sample containing 6.3, 0.3, 2.8, 34.3 and 1.2 nmol/L of corticosterone, 11-DOC-ol, 17-OHP pro-gesterone and DOC, respectively. The black and white peaks correspond to the quantitative mass transitions of the analytes and the internal standards, respectively. Mass transitions of corticosterone and 11-DOC-ol were collected during period 1 (A); DOC and 17-OHP during period 2 (B) and progesterone during period 3 (C). Table 2 Inter and intraassay imprecision. Sample Mean, nmol/L Within run, CV% Between- run/day, CV% Total Imprecision % Corticosterone Level 1 1.5 4.6 7.1 8.5 Level 2 30.0 5.0 6.9 8.5 Level 3 115.1 5.4 4.3 6.9 11-DOC-ol Level 1 1.6 5.5 4.1 6.9 Level 2 60.7 3.1 1.1 3.3 Level 3 138.7 2.4 3.5 4.3 17-OHP Level 1 1.4 3.7 3.9 5.4 Level 2 15.3 2.7 2.3 3.6 Level 3 59.3 2.6 3.2 4.1 Progesterone Level 1 1.6 3.0 1.8 3.5 Level 2 101.1 2.0 2.4 3.1 Level 3 126.4 2.5 3.2 4.1 DOC Level 1 0.1 9.3 5.0 9.9 Level 2 1.5 5.4 2.3 5.9 Level 3 15.9 2.8 2.7 3.9 CV, coefficient of variation. J.A. Ray et al. / Clinica Chimica Acta 438 (2015) 330-336 333 UU IR Author Manuscript UU IR Author Manuscript University of Utah Institutional Repository Author Manuscript Author's personal copy 4.3. Removal of unknown interferences using DMS The role of DMS in improving selectivity could be more clearly assessed by evaluating the performance of themethodwith and without the DMS. As shown in Fig. 4, for samples containing low concentrations of progesterone or DOC, the DMS effectively removed peaks of interfer-ing substances from the peaks of interest. Fig. 4B and D demonstrate the elimination of coeluting peaks with progesterone and DOC at 4.28 and 3.32 min, respectively (a 0.03 min difference in retention time of progesterone between Fig. 1 and Fig. 4 is attributed to typical variation in LC retention times between runs). In the experiments without DMS, the apparent concentrations of progesterone (due to coeluting interfer-ing peaks) were up to 3 times higher than concentrations in themethod employingDMS (Supplemental Fig. 5). This clearly establishes the utility of DMS as a tool for resolving peaks of interest from co-eluents, which becomes especially noticeable in samples containing low concentra-tions of the analyte. 4.4. Method comparison The method showed good agreement with LC-MS/MS assays of other reference laboratories with slopes of Deming regression line of 1.12, 1.08, 0.98, 1.10 and 1.03 and for 11-DOC-ol, 17-OHP, corticosterone, DOC and progesterone respectively (Fig. 2). 17-OHP measured by RIA showed good agreement with the current method (slope of regression line being 1.08). Comparison with CIA method for progesterone A B C D E F Average DOC by LC-DMS-MS/MS (nmol/L) Difference (nmol/L) 2 4 6 -4 -2 0 2 4 Average Corticosterone by LC-DMS-MS/MS (nmol/L) Difference (nmol/L) 20 40 60 80 -4 -2 0 2 4 Average 17-OHPcomparing LC-DMS-MS/MS with RIA (nmol/L) Difference (nmol/L) 20 40 60 80 -20 -10 0 10 20 Average 11-DOC-ol by LC-DMS-MS/MS (nmol/L) Difference (nmol/L) 5 10 15 20 -10 -5 0 5 10 Average Progesterone by LC-DMS-MS/MS (nmol/L) Difference (nmol/L) 20 40 60 80 100 -4 -2 0 2 4 Average 17-OHP by LC-DMS-MS/MS (nmol/L) Difference (nmol/L) 10 20 30 -4 -2 0 2 4 Fig. 2. Bland Altman plots for method comparisons. A) Comparing 11-DOC-ol measured by current LC-DMS-MS/MS methodwith reference laboratory LC-MS/MSmethod. Deming regres-sion equation: Ref LC-MS/MS = 1.12 ∗ LC-DMS-MS/MS − 0.46, n = 31, r = 0.99, Sy/x = 0.72. B) Comparing 17-OHP measured by current LC-DMS-MS/MS method with reference laboratory LC-MS/MSmethod. Deming regression equation: Ref LC-MS/MS=1.08 ∗ LC-DMS-MS/MS+ 0.26, n=32, r= 0.99, Sy/x= 0.28. C) Comparing 17-OHPmeasured by current LC-DMS-MS/MS method with reference laboratory RIA method. Deming regression equation: Ref RIA= 1.08 ∗ LC-DMS-MS/MS + 0.32, n = 27, r = 0.97, Sy/x = 2.15. D) Comparing corticosterone measured by current LC-DMS-MS/MS method with reference laboratory LC-MS/MS method. Deming regression equation: Ref LC-MS/MS Corticosterone = 0.98 ∗ LC-DMS-MS/MS + 0.55, n = 50, r = 0.99, Sy/x = 0.72. E) Comparing DOC measured by current LC-DMS-MS/MS method with reference laboratory LC-MS/MS method. Deming regression equation: Ref LC-MS/MS=1.10 ∗ LC-DMS-MS/MS− 0.022, n=20, r= 0.99, Sy/x= 0.07. F) Comparing progesteronemeasured by current LC-DMS-MS/MSmethod with reference laboratory LC-MS/MS method. Deming regression equations: Ref LC-MS/MS = 1.03 ∗ LC-DMS-MS/MS + 0.00, n = 20, r = 0.99, Sy/x = 0.31 Divide by 0.029 to convert nmol/L to ng/dL for corticosterone and 11-DOC-ol, by 0.030 for 17-OHP and DOC and by 3.18 for converting nmol/L to ng/mL for progesterone. Fig. 3. Histograms with distributions of concentrations of progesterone and 11-deoxycorticosterone in samples from post-menopausal women (n= 125). 334 J.A. Ray et al. / Clinica Chimica Acta 438 (2015) 330-336 UU IR Author Manuscript UU IR Author Manuscript University of Utah Institutional Repository Author Manuscript Author's personal copy (n = 324) showed a regression line slope of 0.97 and correlation coefficient of 0.92. However, CIA was overestimating concentration of progesterone by an average of 2.3 times in samples measuring b3.18 nmol/l (by the current LC-MS/MS method) (Supplemental Fig. 6). Considering these findings, the reported analytical sensitivity of the CIA method (0.31 nmol/l) could be misleading, especially when measuring concentrations in samples of post-menopausal women. Performance of the method validated according to CLSI guidelines (Supplemental Table 2)was found to be consistent throughout the eval-uation of the assay, suggesting sufficient robustness of the technique for use in routine analysis and the practical utility of the hyphenated tech-nique LC-DMS-MS/MS. 5. Conclusion We developed an LC-DMS-MS/MS method for analyzing corticoste-rone, 11-deoxycortisol, 11-deoxycorticosterone, 17-hydroxy progester-one and progesterone in human serum and plasma that has acceptable performance characteristics for diagnostic applications. We were able to reduce background noise, improve performance and separate the isomeric pairs of analytes by the effective combination of LC and ion mobility spectrometry. Reference intervals for progesterone and DOC in post-menopausal women were established. Use of DMS allowed us to simplify sample preparation, while achieving high specificity and sensitivity of analysis, especially in measuring low concentrations of progesterone in samples from post-menopausal women. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cca.2014.07.036. Acknowledgements We thank ARUP Institute for Clinical and Experimental Pathology® for supporting this project. We also thank Sonia La'ulu for assistance with the reference interval study, and Witold Woroniecki and Jesse Seegmiller (AB SCIEX TM) for technical assistance. References [1] Loras B, Dazord A, Roux H, Bertrand J. Study of the level of production of cortisol, cor-ticosterone, aldosterone and deoxycorticosterone in 4 cases of congenital adrenal hyperplasia due to II beta-hydroxylase deficiency. Rev Eur Etud Clin Biol 1971;16: 585-90. [2] Pang S. Congenital adrenal hyperplasia owing to 3 beta-hydroxysteroid dehydrogenase deficiency. Endocrinol Metab Clin North Am 2001;30:81-99 [vi-vii]. [3] Tonetto-Fernandes V, Lemos-Marini SH, Kuperman H, Ribeiro-Neto LM, Verreschi IT, Kater CE. Serum 21-Deoxycortisol, 17-Hydroxyprogesterone, and 11-deoxycortisol in classic congenital adrenal hyperplasia: clinical and hormonal correlations and identification of patients with 11beta-hydroxylase deficiency among a large group with alleged 21-hydroxylase deficiency. J Clin Endocrinol Metab 2006;91:2179-84. [4] Schumert Z, Rosenmann A, Landau H, Rosler A. 11-deoxycortisol in amniotic fluid: prenatal diagnosis of congenital adrenal hyperplasia due to 11 beta-hydroxylase deficiency. Clin Endocrinol (Oxf) 1980;12:257-60. [5] Doerr HG, Sippell WG, Drop SL, Bidlingmaier F, Knorr D. Evidence of 11 beta-hydroxylase deficiency in childhood adrenocortical tumors. The plasma corticosterone/11-deoxycorticosterone ratio as a possible marker for malignancy. Cancer 1987;60:1625-9. [6] Yanagibashi K, Shackleton CH, Hall PF. Conversion of 11-deoxycorticosterone and corticosterone to aldosterone by cytochrome P-450 11 beta-/18-hydroxylase from porcine adrenal. J Steroid Biochem 1988;29:665-75. [7] Santos CM, Abad LR, Cua SC, Domingo CF.Monitoring congenital adrenal hyperplasia using blood spot 17-hydroxyprogesterone assay. Southeast Asian J Trop Med Public Health 2003;34(Suppl. 3):174-8. [8] Abdelazim IA, Elezz AA, Elsherbiny M. Relation between single serum progesterone assay and viability of the first trimester pregnancy. SpringerPlus 2012;1:80. [9] Hanita O, Hanisah AH. Potential use of singlemeasurement of serum progesterone in detecting early pregnancy failure. Malays J Pathol 2012;34:41-6. [10] Wiebe JP, Zhang G, Welch I, Cadieux-Pitre HA. Progesterone metabolites regulate in-duction, growth and suppression of estrogen- and progesterone receptor-negative human breast cell tumors. Breast Cancer Res 2013;15:R38. [11] Hewedi IH, Radwan NA, Shash LS. Diagnostic value of progesterone receptor and p53 expression in uterine smooth muscle tumors. Diagn Pathol 2012;7:1. [12] Lau'lu LSAM, Roberts WL, Straseski JA. American Association for Clinical Chemistry Annual Meeting, Los Angeles, California; July 15-19 2012. [13] Lee H, Park CJ, Lee G. Measurement of progesterone in human serum by isotope dilution liquid chromatography-tandem mass spectrometry and comparison with the commercial chemiluminescence immunoassay. Anal Bioanal Chem 2010;396:1713-9. [14] Tszyrsznic W, Borowiec A, Pawlowska E, Jazwiec R, Zochowska D, Bartlomiejczyk I, et al. Two rapid ultra performance liquid chromatography/tandemmass spectrometry (UPLC/MS/MS) methods with common sample pretreatment for therapeutic drug monitoring of immunosuppressants compared to immunoassay. J Chromatogr B Analyt Technol Biomed Life Sci 2013;928:9-15. [15] Huhtaniemi IT, Tajar A, Lee DM, O'Neill TW, Finn JD, Bartfai G, et al. Comparison of serum testosterone and estradiolmeasurements in 3174 European men using platform immunoassay and mass spectrometry; relevance for the diagnostics in aging men. Eur J Endocrinol 2012;166:983-91. [16] Monaghan PJ, Owen LJ, Trainer PJ, Brabant G, Keevil BG, Darby D. Comparison of serum cortisol measurement by immunoassay and liquid chromatography-tandem mass spectrometry in patients receiving the 11beta-hydroxylase inhibitor metyrapone. Ann Clin Biochem 2011;48:441-6. [17] Moal V, Mathieu E, Reynier P,Malthiery Y, Gallois Y. Low serumtestosterone assayed by liquid chromatography-tandemmass spectrometry. Comparison with five immu-noassay techniques. Clin Chim Acta Int J Clin Chem 2007;386:12-9. [18] Ohlsson C, Nilsson ME, Tivesten A, Ryberg H, Mellstrom D, Karlsson MK, et al. Com-parisons of immunoassay and mass spectrometry measurements of serum estradiol levels and their influence on clinical association studies in men. J Clin Endocrinol Metab 2013;98:E1097-102. [19] Kushnir MM, Rockwood AL, Roberts WL, Pattison EG, Owen WE, Bunker AM, et al. Development and performance evaluation of a tandem mass spectrometry assay for 4 adrenal steroids. Clin Chem 2006;52:1559-67. [20] Zhang S, Mada SR, Sharma S, TorchM, Mattison D, Caritis S, et al. Simultaneous quan-titation of 17alpha-hydroxyprogesterone caproate, 17alpha-hydroxyprogesterone Intensity, cps 3.0 3.4 3.8 4.2 4.6 5.0 Time, min 1.3e4 4.28 3.0 3.4 3.8 4.2 4.6 5.0 3.9e4 4.28 Progesterone A) DMS OFF B) DMS ON Progesterone S/N 1.6 S/N 14.7 0.0 1.0 2.0 3.0 4.0 5.0 Time, min 2263 Intensity, cps 3.38 0.0 1.0 2.0 3.0 4.0 5.0 3.0e4 3.32 C) DMS OFF D) DMS ON S/N 3.6 S/N 5 DOC DOC Fig. 4. Patient serum sample containing 2.2 nmol/L progesterone in DMS OFF (A) and DMS ON (B) mode and 0.30 nmol/L DOC in DMS OFF (C) and DMS ON (D) mode. J.A. Ray et al. / Clinica Chimica Acta 438 (2015) 330-336 335 UU IR Author Manuscript UU IR Author Manuscript University of Utah Institutional Repository Author Manuscript Author's personal copy and progesterone in human plasma using high-performance liquid chromatography-mass spectrometry (HPLC-MS/MS). J Pharm Biomed Anal 2008;48:1174-80. [21] Wudy SA, Hartmann M, Svoboda M. Determination of 17-hydroxyprogesterone in plasma by stable isotope dilution/benchtop liquid chromatography-tandem mass spectrometry. Horm Res 2000;53:68-71. [22] Magnisali P, Chalioti MB, Livadara T,Mataragas M, Paliatsiou S,Malamitsi-Puchner A, et al. Simultaneous quantification of 17alpha-OH progesterone, 11-deoxycortisol, Delta4-androstenedione, cortisol and cortisone in newborn blood spots using liquid chromatography-tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 2011;879:1565-72. [23] McWhinney BC, Briscoe SE, Ungerer JP, Pretorius CJ. Measurement of cortisol, cortisone, prednisolone, dexamethasone and 11-deoxycortisol with ultra high performance liquid chromatography-tandem mass spectrometry: Application for plasma, plasma ultrafiltrate, urine and saliva in a routine laboratory. J Chromatogr B Analyt Technol Biomed Life Sci 2010;878:2863-9. [24] Purves RW, Guevremont R. Electrospray ionization high-field asymmetric waveform ion mobility spectrometry-mass spectrometry. Anal Chem 1999;71:2346-57. [25] Barnett DA, Ells B, Guevremont R, Purves RW, Viehland LA. Evaluation of carrier gases for use in high-field asymmetric waveform ion mobility spectrometry. J Am Soc Mass Spectrom 2000;11:1125-33. [26] Barnett DA, Purves RW, Ells B, Guevremont R. Separation of o-, m- and p-phthalic acids by high-field asymmetric waveform ion mobility spectrometry (FAIMS) using mixed carrier gases. J Mass Spectrom 2000;35:976-80. [27] McCooeye MA, Ells B, Barnett DA, Purves RW, Guevremont R. Quantitation ofmorphine and codeine in human urine using high-filed asymmetric waveformion mobility spec-trometry (FAIMS) with mass spectrometric detection. J Anal Toxicol 2001;25:81-7. [28] Guevremont R. High-field asymmetric waveform ion mobility spectrometry: a new tool for mass spectrometry. J Chromatogr A 2004;1058:3-19. [29] Kolakowski BM, Mester Z. Review of applications of high-field asymmetric wave-form ion mobility spectrometry (FAIMS) and differential mobility spectrometry (DMS). Analyst 2007;132:842-64. [30] Shvartsburg AA, Tang K, Smith RD, Holden M, Rush M, Thompson A, et al. Ultrafast differential ion mobility spectrometry at extreme electric fields coupled to mass spectrometry. Anal Chem 2009;81:8048-53. [31] Shvartsburg AA, Isaac G, Leveque N, Smith RD, Metz TO. Separation and classification of lipids using differential ion mobility spectrometry. J Am Soc Mass Spectrom 2011; 22:1146-55. [32] Purves RW. Enhancement of biological mass spectrometry by using separations based on changes in ion mobility (FAIMS and DMS). Anal Bioanal Chem 2013;405: 35-42. [33] Guddat S, ThevisM, Kapron J, Thomas A, SchanzerW. Application of FAIMS to anabolic androgenic steroids in sport drug testing. Drug Test Anal 2009;1:545-53. [34] Saba J, Bonneil E, Pomies C, Eng K, Thibault P. Enhanced sensitivity in proteomics experiments using FAIMS coupled with a hybrid linear ion trap/Orbitrap mass spec-trometer. J Proteome Res 2009;8:3355-66. [35] Jin WW, Jarvis MM, Star-Weinstock MM, Altemus M. A sensitive and selective LC-differential mobility-mass spectrometric analysis of allopregnanolone and pregnanolone in human plasma. Anal Bioanal Chem 2013;405:9497-508. [36] Varesio E, Le Blanc JC, Hopfgartner G. Real-time 2D separation by LC x differential ion mobility hyphenated to mass spectrometry. Anal Bioanal Chem 2012;402: 2555-64. [37] Blagojevic V, Chramow A, Schneider BB, Covey TR, Bohme DK. Differential mobility spectrometry of isomeric protonated dipeptides: modifier and field effects on ion mobility and stability. Anal Chem 2011;83:3470-6. [38] Shvartsburg AA, Creese AJ, Smith RD, Cooper HJ. Separation of a set of peptide se-quence isomers using differential ion mobility spectrometry. Anal Chem 2011;83: 6918-23. [39] Camara M, Gharbi N, Lenouvel A, Behr M, Guignard C, Orlewski P, et al. Detection and quantification of natural contaminants of wine by gas chromatography-differential ion mobility spectrometry (GC-DMS). J Agric Food Chem 2013;61: 1036-43. [40] Annesley TM. Ion suppression in mass spectrometry. Clin Chem 2003;49:1041-4. 336 J.A. Ray et al. / Clinica Chimica Acta 438 (2015) 330-336 UU IR Author Manuscript UU IR Author Manuscript University of Utah Institutional Repository Author Manuscript