- Email: [email protected]
Cortisol and cortisone analysis in serum and plasma by atmospheric pressure photoionization tandem mass spectrometry
Clinical Biochemistry 37 (2004) 357 – 362
Cortisol and cortisone analysis in serum and plasma by atmospheric pressure photoionization tandem mass spectrometry Mark M. Kushnir, a,* Rebecca Neilson, a William L. Roberts, b and Alan L. Rockwood a a
ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, UT 84108, USA b Department of Pathology, University of Utah, Salt Lake City, UT 84132, USA
Received 26 August 2003; received in revised form 7 January 2004; accepted 7 January 2004
Abstract Objectives: Cortisol metabolism is controlled by 11h-hydroxysteroid dehydrogenase (11h-HSD) isoenzymes, which interconvert cortisol and cortisone. Accurate measurement of the cortisol and cortisone concentrations and their ratio provide useful information about 11h-HSD activity. Methods: Cortisol and cortisone were extracted with methyl-tert-butyl ether from 100 Al of serum or plasma. The extract was evaporated, reconstituted with mobile phase, and analyzed by tandem mass spectrometry using a photoionization interface. The transitions monitored were: m/z 363 to 121 and 363 to 97 for cortisol, 361 to 163 and 361 to 105 for cortisone. Results: Within-run and between-run coefficients of variation were less than 6% and 12%; 14% and 22%; 11% and 21% for cortisol, cortisone, and their ratio, respectively. The limit of detection was 1 Ag/l for cortisol and 5 Ag/l for cortisone. Normal ranges for cortisol and cortisone concentration and for their ratio in plasma (n = 120) determined as the central 95% were 33 – 246 Ag/l for cortisol, 8 – 27 Ag/l for cortisone, and 0.081 – 0.301 for the cortisone/cortisol ratio. Conclusions: We developed a simple sensitive method for cortisol and cortisone analysis in plasma and serum that uses a small sample volume. The method is very specific, fast, does not have any known interference, and is useful for diagnosis of variety of disease and pathologic conditions. D 2004 The Canadian Society of Clinical Chemists. All rights reserved. Keywords: Cortisol; Cortisone; Mass spectrometry; 11h-hydroxysteroid dehydrogenase; Atmospheric pressure photoionization; Cushing’s disease
Introduction Cortisol plays an important role in human physiology and is a useful marker for the diagnosis on many pathologic conditions. Cortisol determinations are commonly used for diagnosis of Cushing’s disease, congenital adrenal hyperplasia, and adrenal insufficiency. Recent evidence suggests that increased cortisol secretion, altered cortisol metabolism, and increased tissue sensitivity to cortisol may be related to insulin resistance, obesity, and hypertension [1 – 5]. The above conditions are common in patients with glucose intolerance and type 2 diabetes mellitus [6– 8]. Altered metabolism of cortisol is also responsible for a
* Corresponding author. ARUP Institute for Clinical and Experimental Pathology, 500 Chipeta Way, Salt Lake City, UT 84108. Fax: +1-801-5845207. E-mail address: [email protected] (M.M. Kushnir).
condition called apparent mineralocorticoid excess (AME) syndrome [9 –12]. AME is an inherited form of hypertension, which results from insufficiency of the enzyme required for conversion of cortisol to cortisone. These and some other disorders related to cortisol metabolism can be detected based on the interconversion of cortisol and cortisone [13 –15]. Cortisol is produced and secreted by the adrenal gland, while the adrenal production of cortisone is negligible. The interconversion between cortisol and cortisone is catalyzed by 11h-hydroxysteroid dehydrogenase (11h-HSD) isoenzymes. Cortisol concentrations in tissue and body fluids are controlled by the relative activity of 11h-HSD type 1 (11hreductase), which is responsible for the reversible conversion of cortisone to cortisol, and 11h-HSD type 2 that catalyzes the irreversible conversion of cortisol to cortisone. The 11h-HSD1 is expressed in liver and adipose tissue, with the highest activity normally observed in the liver. Hepatic 11h-HSD1 conversion of cortisone to cortisol
0009-9120/$ - see front matter D 2004 The Canadian Society of Clinical Chemists. All rights reserved. doi:10.1016/j.clinbiochem.2004.01.005
358
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
is impaired in obese people, but this is compensated for by increased enzyme activity in adipose tissue. The distribution of 11h-HSD1 suggests its physiological role in maintaining adequate activation of glucocorticoid receptors, where cortisol plays an important role. The enzyme 11hHSD type 2 is in the kidney [1,6]. A balance between the type 1 and type 2 enzymatic activities maintains available cortisol, which at physiological concentrations acts as a mild anti-inflammatory agent. The ratio between cortisol and cortisone is dependent on 11h-HSD isoenzymes activity and a combined test for cortisol and cortisone is potentially useful in evaluating patients with the disorders described above. Historically, immunoassay (IA)-based methods were widely used for cortisol analysis [16]. These methods may be susceptible to interference from cortisone and other steroid metabolites. It has been suggested that chromatographic methods are more accurate for measurement of cortisol than immunoassay [17,18]. Mass spectrometry (MS)-based methods are especially attractive for cortisol and cortisone analysis. Several LC-MS/MS based methods for cortisol and cortisone have been reported [19 – 21]. Although potentially more accurate than IA, these methods are time- and labor-intensive and lack the sensitivity needed for reliable detection of cortisol and cortisone in human serum and plasma. Our objective was to develop a simple, rapid, and sensitive assay for simultaneous measurement of cortisol and cortisone in serum and plasma.
Materials and methods Standards and reagents Standards of cortisol and cortisone were purchased from Sigma Chemical Co., d4-cortisol (98% purity) was purchased from Cambridge Isotope Laboratories. Methanol, methyl-tert-butyl ether (MTBE), isopropanol, and phosphoric acid were purchased from Fisher Scientific (all HPLC grade). Stock solutions of cortisol, cortisone, and d4-cortisol were prepared in methanol at a concentration of 1 g/l. Working standards were prepared from the stock standards at a concentration of 0.5 mg/l in a mix of methanol/water (1:1) containing 10 mg/l of estriol. The internal standard (IS) d4-cortisol was prepared at a concentration of 0.5 mg/ l in a mix of methanol/water (1:1). The calibration standards were prepared at concentrations of 10, 50, 100, and 200 Ag/ l in 6% bovine serum albumin (BSA). All the standards were stored at 70jC. Assay procedures Sample preparation for the method was performed as follows. An aliquot of serum or plasma (100 Al) was pipetted into a microcentrifuge tube. Working internal standard solution (50 Al) and MTBE containing 3% phos-
phoric acid (500 Al) were added. The tube was vortexed for 5 min and centrifuged at 15,000 g for 5 min. The supernatant was transferred to a second tube; the solvent was evaporated, and reconstituted with a methanol and water mix (1:1) containing 0.1 mg/l of estriol and transferred to a labeled autosampler vial. Sample collection The plasma samples utilized in the study (n = 120) were collected after obtaining informed consent from apparently healthy adult volunteers using EDTA as an anticoagulant. To account for diurnal variation in cortisol excretion, all samples utilized in the study were collected in the morning. Seventy-one samples for routine clinical analysis of cortisol were used for the method comparison studies. The samples were stored after collection at 70jC. All studies with samples from human subjects were approved by Institutional Review Board of the University of Utah. Instrumental analysis A PE series 200 HPLC system (Perkin Elmer Analytical Instruments) was equipped with a Luna Phenyl-Hexyl column 50 2.1 mm, 5-Am particle size (Phenomenex). The mobile phase was composed of methanol and water (1:1). The mobile phase flow rate was 0.45 ml/min, the column temperature was 45jC, and the injection volume was 10 Al. The injection syringe was washed between the injections eight times with 100 Al of solution containing 20% water, 80% methanol, 0.05% trifluoroacetic acid. The injection interval was 3 min. An API 3000 (Applied Biosystems/MDS SCIEX) tandem mass spectrometer was used in the positive ion mode with atmospheric pressure photo ionization (APPI) interface. Quantitative analysis was performed in the multiple reaction monitoring (MRM) mode. The MRM transitions monitored were m/z: 363 to 121 and 363 to 97 for cortisol, 361 to 163 and 361 to 105 for cortisone, and 367 to 121 and 367 to 97 for d4-cortisol. The product ions m/z 121 and 163 were quantitative, while the product ions m/z 97 and 105 were qualitative. The calibration standard containing 50 Ag/ l of cortisol was utilized to establish the qualitative ratio of the intensities of the product ion fragments, m/z 363 to 121 and 363 to 97 for cortisol, m/z 361 to 163 and 361 to 105 for cortisone, and m/z 367 to 121 and 367 to 97 for d4-cortisol. The qualitative ratio acceptability limits for the controls and test samples were set as F40% of the values established with the calibration standard. Toluene was used as a dopant to promote ionization at a flow rate of 75 Al/min. Ionization was induced by exposure of LC column effluent to the krypton lamp. The APPI voltage was 1400 V, the orifice voltage was 35 V. The collision gas pressure was 5 mTorr; collision energy was set at 35 V for transitions m/z 363 to 121, 361 to 163, and 367 to 121; and 45 V for transitions m/ z 363 to 97, 361 to 105, and 367 to 97. Quantitative data
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
analysis was performed with Analystk software (Applied Biosystems/MDS SCIEX). Cortisol measurement by IA method were performed on an Immulite 2000 analyzer (Diagnostic Products Corporation) using manufacturer’s reagents according to their instructions.
Results Because the MS/MS detection is very selective, LC separation plays less role for the assay compared to other detection modes. A short LC column was used to separate compounds from the solvent front to eliminate possible signal suppression. The chromatogram for the monitored mass ion transitions for extracted plasma sample is presented in Fig. 1. Method sensitivity was determined by analyzing cortisone and cortisol spiked in BSA at progressively lower concentrations. Each standard was analyzed in duplicate. The limit of quantitation (LOQ) was determined as the
359
Table 1 Method imprecision Cortisol Concentration Within-run (Ag/l) CV (%)
Between-run Total CV (%) CV (%)
2a 25.5a 50a 6b 30b
12 10 10 7 8
3 6 5 5 4
13 12 11 8 9
Cortisone Concentration Within-run (Ag/l) CV (%)
Between-run Total CV (%) CV (%)
2a 5.5a 10a 2b 10b
22 13 12 9 7
5 6 6 14 5
22 14 14 16 8
Ratio cortisone/cortisol Ratio
Cortisone/cortisol Within-run concentration CV (%) (Ag/dl)
Between-run Total CV (%) CV (%)
1a 0.216a 0.200a 0.333b 0.333b
2/2 5.5/25.5 10/50 2/6 10/30
19 21 19 7 10
a b
5 5 4 11 5
19 22 19 13 11
Three replicates per run over 6 days. Two replicates per run over 6 days.
lowest concentration at which accuracy was within F20% and imprecision was within F15%. The LOQ was found to be 5 Ag/l for both cortisol and cortisone. The limit of detection (LOD) was determined as the lowest concentration at which signals were five times greater than background noise for both primary and secondary transitions. The LOD
Fig. 1. MRM chromatograms of extracted plasma sample containing 55 Ag/l of cortisol, and 18 Ag/l of cortisone. Ion transitions: cortisol m/z 363!121 and 363!97, (A); cortisone m/z 361!163 and 361!105, (B); d4-cortisol m/z 367!121 and 367!97, (C).
Fig. 2. Comparison of LC-MS/MS values with immunoassay for cortisol analysis in serum samples (n = 71).
360
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
was determined to be 1 Ag/l for cortisol and 5 Ag/l for cortisone. Method linearity was evaluated by analyzing BSA spiked samples containing cortisone and cortisol at concentrations of 50, 1600, 3200, 4800, 6400, and 8000 Ag/l. Each sample was analyzed in duplicate, and concentrations were calculated from a calibration curve established with standards containing 10, 50, 100, and 200 Ag/l of each compound. The method was found to be linear up to 8000 Ag/ l for cortisol and to 1600 Ag/l for cortisone, with inaccuracy at the highest concentration of 6% and 14%, respectively. This linearity range far exceeds the normal and pathologic concentrations of cortisol and cortisone in human serum and plasma. Within-run and between-run imprecision for the LC-MS/ MS method was determined by analysis of BSA supple-
mented with cortisol and cortisone to concentrations ranging from 25 to 450 and 20 to 100 Ag/l, respectively. The experiments were performed in triplicate over a period of 6 days. Additional imprecision data were obtained from results of controls included with samples during routine method use. A summary of the results of these imprecision studies is presented in Table 1. Method performance for cortisol analysis was evaluated by comparison with an IA method utilizing 71 human serum samples (Fig. 2). The regression equation, correlation coefficient and standard error for the comparison were y = 0.74x 0.70, r = 0.965 and Sy/x = 32.7 Ag/l respectively. Nineteen plasma samples with cortisol concentrations ranging between 0 and 650 Ag/l were analyzed by a LC-MS/MS method at another laboratory. The regression equation, correlation coefficient, and standard error for the results were y = 1.00x + 0.03 and r = 0.994 and 1.84 Ag/l. The LC-MS/MS method was used to establish reference intervals for cortisone, cortisol, and the their ratio in an adult population. The median concentration of cortisol and cortisone were 79.8 and 14.8 Ag/l for male, and 83.6 and 13.7 Ag/l for female, respectively. The median value for the ratio of cortisone to cortisol was 0.169 for male and 0.160 for female. Nonparametric reference intervals determined as central 95% were 33 –246 Ag/l for cortisol, 8 – 27 Ag/l for cortisone, and 0.081 –0.301 for the cortisone/cortisol ratio. Histograms with the distribution of cortisol, cortisone, and their ratio are presented in Fig. 3.
Discussion
Fig. 3. Distribution of concentrations of cortisol (A) and cortisone (B), and the cortisone/cortisol ratio (C) in samples utilized for the reference interval study.
Analysis of cortisol and cortisone is analytically challenging because there are many structurally similar steroids and metabolites that may be present in biological samples. Immunoassay-based methods may lack specificity. They may also utilize method-specific reference intervals, which take into account cross-reactivity with other endogenous steroids. IA-based methods are not ideal when accurate measurement of cortisol and cortisone is required. High specificity detection provided by MS/MS based methods for cortisol and cortisone can produce more accurate results and eliminate potential interference. Determination of the ratio of cortisol and cortisone requires high specificity, good accuracy, and sufficient precision for both analytes; otherwise, it may not be useful. Method precision is important for the combined test because the imprecision for measurement of the ratio of the concentrations of two analytes is potentially greater than imprecision of each individual measurement. When the ratio of concentrations needs to be determined, it is preferable to simultaneously measure both analytes because some factors affecting method performance and related to sample handling (volume of aliquot, reconstitution volume, etc.) during the analysis are canceled out as they equally affect both analytes.
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
Various ionization techniques can be used for LC-MS/ MS analysis of steroids. It was reported that photoionization (APPI) might be more efficient for analysis of steroids [22] compared to other ionization techniques. APPI interface is an evolution of atmospheric pressure ionization (APCI) in which UV light initiates the ionization process, as opposed to corona discharge as in APCI. The principle utilized for the ionization is based on irradiating the vaporized eluent stream with UV light, which initiates a cascade of gas-phase reactions resulting in ionization of the analyte molecules. Toluene has a low ionization energy (8.8 eV) and is commonly used as a dopant to improve ionization efficiency of analytes. When dopant molecules are exposed to photons emitted by the light source, they produce photo ions. In addition to a direct charge-transfer reaction from photo ions to compounds with low ionization energy, the photo ions initiate a cascade of reactions through formation of active intermediate products that lead to ionization of the analytes. A short LC column was used to separate cortisol and cortisone from potentially interfering compounds with similar characteristic MS/MS transitions and from the solvent front to eliminate possible signal suppression [24]. Because of high specificity of the MS/MS detection, the short column allowed obtaining fast analysis time without sacrifice in the method performance. Comparison of recovery between an extracted sample and cortisol and cortisone spiked in the mobile phase at the same concentrations showed that ion suppression did not occur under the LC conditions utilized. Most of the published methods for cortisol analysis utilize reverse phase octadecyl LC columns. A potential problem with this column is interference of tetrahydroprednisolone with cortisol. Tetrahydroprednisolone is a prednisolone metabolite, an isomer of cortisol, and it is present in samples of patients taking prednisone and prednisolone. A phenyl – hexyl column utilized in this method can separate these isomers; produce good peak shape and sufficient separation from the solvent front. The phenyl –hexyl bonded phase has selectivity based on k –k interactions between the stationary phase and an analyte, and proved to be useful for analysis of steroids with strong structural similarities. No compounds were found to interfere with cortisol. Prednisolone, an isomer of cortisone, has the same parent ion as cortisone, but has different product ions. This results in significantly lower sensitivity to prednisolone compared to cortisone. When prednisolone is present in a sample it elutes with a slightly shorter retention time compared to cortisone and produces elevated branching ratio for the transitions 361 to 105/361 to 163. The magnitude of the branching ratio depends on relative concentration of the analytes. Utilization of two transitions and branching ratios for each analyte allows one to qualitatively confirm the identity of the compounds and to detect interfering peaks if other isobaric compounds with similar product ions are present in a sample [23].
361
No significant differences were observed for the cortisol and cortisone concentrations and for the cortisone/cortisol ratio between the genders. Previously, cortisol concentrations in plasma for adults were reported to be 60 –230 Ag/ l by an IA-based method [16]. Lower reference intervals for urinary free cortisol were reported in adult females by LCMS/MS based methods compared with IA methods [17,20]. The results of the method comparison with the IA method suggest a negative bias of 26%. This may be due to cross-reactivity with steroid metabolites in the IA method. The difference in the cross-reactivity associated with IA may also be reflected in the reference intervals for adult cortisol values that are usually greater for IA-based methods. Measurement of the ratio of free cortisol and cortisone in urine also has clinical value and can be utilized for diagnosis of AME and other conditions. Normal interval for the ratio of urinary free cortisone and urinary free cortisol was reported to be between 1.2 and 4.2 [9,20]. The significant difference observed for the cortisone/cortisol ratio between serum and urine can be explained by inactivation of cortisol to cortisone that takes place in kidney, and lower binding of cortisone to plasma proteins that lead to its faster excretion. A potential pitfall in the analysis for cortisol and cortisone is poor stability of the compounds. The stability of cortisol and cortisone was assessed in plasma samples. Samples with a normal and an elevated concentration of each analyte were frozen and thawed three times and analyzed between each cycle. No degradation was observed after the freeze – thaw cycle for either cortisol or cortisone. In addition, samples with normal and elevated concentrations were stored at room temperature and analyzed over a week of storage. Storage of the samples at room temperature did not have an affect on the cortisol, but decreased the concentration of cortisone by 55%. We also observed an approximately 60% degradation of cortisone in calibration standard solution after 2 months of storage at 4 –8jC when it was prepared in water. No degradation of cortisol was observed under these conditions. The stability of cortisone improved when the working calibration standard solution was prepared in methanol or in mix of methanol with water (1:1) and stored at 70jC. While cortisol was stable in storage, we observed its loss during sample preparation. Sample instability is harmful for the combined test for cortisol and cortisone, and may lead to errors in the ratio of their concentrations. One possible explanation for the loss of cortisol is its binding to glass surfaces [20]. To minimize loss, Taylor et al. [20] suggested adding to the standards another steroid, estriol, which would compete with cortisol binding to active silanols on glass tube surface. Because concentration of estriol is significantly greater compared to the concentration of cortisol, this would have a stabilizing effect. To compensate for the degradation, it is also important to prepare the calibration standards in a matrix similar to the matrix of the samples, and to utilize an internal
362
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
standard with properties closely matching those of the analytes of interest.
[7] [8]
Conclusions In conclusion, we developed a sensitive and specific method for analysis of cortisol and cortisone in serum and plasma that utilizes atmospheric pressure photoionization interface and tandem mass spectrometry. The method is very specific, does not have any known interference, and has sufficient sensitivity and precision. Monitoring two ion transitions for each analyte results in high specificity of the method, which is necessary for a test targeted for determining the ratio between cotrisol and cortisone. The benefit of simultaneous measurement of both compounds in single analysis is in reduction of imprecision related to sample handling. Considering the proposed links between cortisone/cortisol ratio and a variety of diseases, the test will be very useful for diagnostic purposes. The small sample volume required for this test helps to reduce the volume of blood drawn from patients and would be especially useful for pediatric testing.
[9]
[10] [11]
[12]
[13]
[14]
[15]
[16]
Acknowledgments This work was supported by ARUP Institute for Clinical and Experimental Pathology.
[17]
[18]
References [1] Tiosano D, Eisentein I, Militianu D, Chrousos GP, Hochberg Z. 11h-hydroxysteroid dehydrogenase activity in hypothalamic obesity. J Clin Endocrinol Metab 2003;88:379 – 84. [2] Stewart PM. 11h-hydroxysteroid dehydrogenase: implications for clinical medicine. Clin Endocrinol (Oxford) 1996;44:493 – 9. [3] Bujalska IJ, Kumar S, Stewart PM. Does central obesity reflect ‘‘Cushing’s disease of the omentum’’? Lancet 1997;349:1210 – 3. [4] Stewart PM, Boulton A, Kumar S, Clark PM, Shackleton CHL. Cortisol metabolism in human obesity: impaired cortisone ! cortisol conversion in subjects with central adiposity. J Clin Endocrinol Metab 1999;84:1022 – 7. [5] Reinke M. Subclinical Cushing’s syndrome. Endocrinol Metab Clin North Am 2000;29:43 – 56. [6] Andrews RC, Herlihy O, Livingstone DEW, Andrew R, Walker BR. Abnormal cortisol metabolism and tissue sensitivity to cortisol in
[19]
[20]
[21]
[22]
[23]
[24]
patients with glucose intolerance. J Clin Endocrinol Metab 2002;87: 5587 – 93. Andrews RC, Walker BR. Glucocorticoids and insulin resistance: old hormones, new targets. Clin Sci 1999;96:513 – 23. Phillips DIW, Barker DJP, Fall CHD, et al. Elevated plasma cortisol concentrations: an explanation for the relationship between low birth weight and adult cardiovascular risk factors. J Clin Endocrinol Metab 1998;83:757 – 60. Palermo M, Delitala G, Mantero F, Stewart PM, Shackleton CHL. Congenital deficiency of 11h-hydroxysteroid dehydrogenase (apparent mineralocorticoid excess syndrome): diagnostic value of urinary free cortisol and cortisone. J Endocrinol Invest 2001;24:17 – 23. Shimojo M, Stewart PM. Apparent mineralocorticoid excess syndrome. J Endocrinol Invest 1995;51:518 – 32. Stewart PM, Wallace AM, Valentino R, Burt D, Shackleton CHL, Edwards CRW. Mineralocorticoid activity of licorice: 11h-hydroxysteroid dehydrogenase comes of age. Lancet 1987;2:821 – 4. Edwards CRW, Walker BR, Benediktsson R, Seckl JR. Congenital and acquired syndrome of apparent mineralocorticoid excess. J Steroid Biochem Mol Biol 1983;45:1 – 5. Mune T, Morita H, Suzuki T, et al. Role of local 11h-hydroxysteroid dehydrogenase type 2 expression in determining the phenotype of adrenal adenomas. J Clin Endocrinol Metab 2003;88:864 – 70. Funder JW, Pearce PT, Smith R, Smith AI. Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science 1988;242:583 – 5. White PC, Mune T, Agarwal AK. 11h-hydroxysteroid dehydrogenase and the syndrome of apparent mineralocorticoid excess. Endocr Rev 1997;18:135 – 56. Tello FL, Hernandez DM. Performance evaluation of nine hormone assays on the Immulite 2000 immunoassay system. Clin Chem Lab Med 2000;38:1039 – 42. Murphy BEP. Lack of specificity of urinary free cortisol determinations: why does it continue? J Clin Endocrinol Metab 1999;884: 2258 – 9. Murphy BEP. How much ‘‘UFC’’ is really cortisol? Clin Chem 2000;46(6):793 – 4. Vogeser M, Zachoval R, Felbinger TW, Jacob K. Increased ratio of serum cortisol to cortisone in acute-phase response. Horm Res 2002; 15:269 – 78. Taylor RL, Machacek D, Singh RJ. Validation of a high-throughput liquid chromatography-tandem mass spectrometry method for urinary cortisol and cortisone. Clin Chem 2002;48:1511 – 9. Kushnir MM, Rockwood AL, Nelson GN, Terry AH, Meikle AW. Analysis of cortisol in urine by LC-MS/MS. Clin Chem 2003;49: 965 – 7. Robb DB, Covey TR, Bruins AP. Atmospheric pressure photoionization: an ionization method for liquid chromatography-mass spectrometry. Anal Chem 2000;72:3653 – 9. Meikle AW, Findling J, Kushnir MM, Rockwood AL, Nelson GJ, Terry AH. Pseudocushings syndrome cause by interference of drug Fenofibrate with HPLC assay for cortisol. J Clin Endocrinol Metab 2003;88:3521 – 4. Annesley TM. Ion suppression in mass spectrometry. Clin Chem 2003;49:1041 – 4.
Cortisol and cortisone analysis in serum and plasma by atmospheric pressure photoionization tandem mass spectrometry Mark M. Kushnir, a,* Rebecca Neilson, a William L. Roberts, b and Alan L. Rockwood a a
ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, UT 84108, USA b Department of Pathology, University of Utah, Salt Lake City, UT 84132, USA
Received 26 August 2003; received in revised form 7 January 2004; accepted 7 January 2004
Abstract Objectives: Cortisol metabolism is controlled by 11h-hydroxysteroid dehydrogenase (11h-HSD) isoenzymes, which interconvert cortisol and cortisone. Accurate measurement of the cortisol and cortisone concentrations and their ratio provide useful information about 11h-HSD activity. Methods: Cortisol and cortisone were extracted with methyl-tert-butyl ether from 100 Al of serum or plasma. The extract was evaporated, reconstituted with mobile phase, and analyzed by tandem mass spectrometry using a photoionization interface. The transitions monitored were: m/z 363 to 121 and 363 to 97 for cortisol, 361 to 163 and 361 to 105 for cortisone. Results: Within-run and between-run coefficients of variation were less than 6% and 12%; 14% and 22%; 11% and 21% for cortisol, cortisone, and their ratio, respectively. The limit of detection was 1 Ag/l for cortisol and 5 Ag/l for cortisone. Normal ranges for cortisol and cortisone concentration and for their ratio in plasma (n = 120) determined as the central 95% were 33 – 246 Ag/l for cortisol, 8 – 27 Ag/l for cortisone, and 0.081 – 0.301 for the cortisone/cortisol ratio. Conclusions: We developed a simple sensitive method for cortisol and cortisone analysis in plasma and serum that uses a small sample volume. The method is very specific, fast, does not have any known interference, and is useful for diagnosis of variety of disease and pathologic conditions. D 2004 The Canadian Society of Clinical Chemists. All rights reserved. Keywords: Cortisol; Cortisone; Mass spectrometry; 11h-hydroxysteroid dehydrogenase; Atmospheric pressure photoionization; Cushing’s disease
Introduction Cortisol plays an important role in human physiology and is a useful marker for the diagnosis on many pathologic conditions. Cortisol determinations are commonly used for diagnosis of Cushing’s disease, congenital adrenal hyperplasia, and adrenal insufficiency. Recent evidence suggests that increased cortisol secretion, altered cortisol metabolism, and increased tissue sensitivity to cortisol may be related to insulin resistance, obesity, and hypertension [1 – 5]. The above conditions are common in patients with glucose intolerance and type 2 diabetes mellitus [6– 8]. Altered metabolism of cortisol is also responsible for a
* Corresponding author. ARUP Institute for Clinical and Experimental Pathology, 500 Chipeta Way, Salt Lake City, UT 84108. Fax: +1-801-5845207. E-mail address: [email protected] (M.M. Kushnir).
condition called apparent mineralocorticoid excess (AME) syndrome [9 –12]. AME is an inherited form of hypertension, which results from insufficiency of the enzyme required for conversion of cortisol to cortisone. These and some other disorders related to cortisol metabolism can be detected based on the interconversion of cortisol and cortisone [13 –15]. Cortisol is produced and secreted by the adrenal gland, while the adrenal production of cortisone is negligible. The interconversion between cortisol and cortisone is catalyzed by 11h-hydroxysteroid dehydrogenase (11h-HSD) isoenzymes. Cortisol concentrations in tissue and body fluids are controlled by the relative activity of 11h-HSD type 1 (11hreductase), which is responsible for the reversible conversion of cortisone to cortisol, and 11h-HSD type 2 that catalyzes the irreversible conversion of cortisol to cortisone. The 11h-HSD1 is expressed in liver and adipose tissue, with the highest activity normally observed in the liver. Hepatic 11h-HSD1 conversion of cortisone to cortisol
0009-9120/$ - see front matter D 2004 The Canadian Society of Clinical Chemists. All rights reserved. doi:10.1016/j.clinbiochem.2004.01.005
358
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
is impaired in obese people, but this is compensated for by increased enzyme activity in adipose tissue. The distribution of 11h-HSD1 suggests its physiological role in maintaining adequate activation of glucocorticoid receptors, where cortisol plays an important role. The enzyme 11hHSD type 2 is in the kidney [1,6]. A balance between the type 1 and type 2 enzymatic activities maintains available cortisol, which at physiological concentrations acts as a mild anti-inflammatory agent. The ratio between cortisol and cortisone is dependent on 11h-HSD isoenzymes activity and a combined test for cortisol and cortisone is potentially useful in evaluating patients with the disorders described above. Historically, immunoassay (IA)-based methods were widely used for cortisol analysis [16]. These methods may be susceptible to interference from cortisone and other steroid metabolites. It has been suggested that chromatographic methods are more accurate for measurement of cortisol than immunoassay [17,18]. Mass spectrometry (MS)-based methods are especially attractive for cortisol and cortisone analysis. Several LC-MS/MS based methods for cortisol and cortisone have been reported [19 – 21]. Although potentially more accurate than IA, these methods are time- and labor-intensive and lack the sensitivity needed for reliable detection of cortisol and cortisone in human serum and plasma. Our objective was to develop a simple, rapid, and sensitive assay for simultaneous measurement of cortisol and cortisone in serum and plasma.
Materials and methods Standards and reagents Standards of cortisol and cortisone were purchased from Sigma Chemical Co., d4-cortisol (98% purity) was purchased from Cambridge Isotope Laboratories. Methanol, methyl-tert-butyl ether (MTBE), isopropanol, and phosphoric acid were purchased from Fisher Scientific (all HPLC grade). Stock solutions of cortisol, cortisone, and d4-cortisol were prepared in methanol at a concentration of 1 g/l. Working standards were prepared from the stock standards at a concentration of 0.5 mg/l in a mix of methanol/water (1:1) containing 10 mg/l of estriol. The internal standard (IS) d4-cortisol was prepared at a concentration of 0.5 mg/ l in a mix of methanol/water (1:1). The calibration standards were prepared at concentrations of 10, 50, 100, and 200 Ag/ l in 6% bovine serum albumin (BSA). All the standards were stored at 70jC. Assay procedures Sample preparation for the method was performed as follows. An aliquot of serum or plasma (100 Al) was pipetted into a microcentrifuge tube. Working internal standard solution (50 Al) and MTBE containing 3% phos-
phoric acid (500 Al) were added. The tube was vortexed for 5 min and centrifuged at 15,000 g for 5 min. The supernatant was transferred to a second tube; the solvent was evaporated, and reconstituted with a methanol and water mix (1:1) containing 0.1 mg/l of estriol and transferred to a labeled autosampler vial. Sample collection The plasma samples utilized in the study (n = 120) were collected after obtaining informed consent from apparently healthy adult volunteers using EDTA as an anticoagulant. To account for diurnal variation in cortisol excretion, all samples utilized in the study were collected in the morning. Seventy-one samples for routine clinical analysis of cortisol were used for the method comparison studies. The samples were stored after collection at 70jC. All studies with samples from human subjects were approved by Institutional Review Board of the University of Utah. Instrumental analysis A PE series 200 HPLC system (Perkin Elmer Analytical Instruments) was equipped with a Luna Phenyl-Hexyl column 50 2.1 mm, 5-Am particle size (Phenomenex). The mobile phase was composed of methanol and water (1:1). The mobile phase flow rate was 0.45 ml/min, the column temperature was 45jC, and the injection volume was 10 Al. The injection syringe was washed between the injections eight times with 100 Al of solution containing 20% water, 80% methanol, 0.05% trifluoroacetic acid. The injection interval was 3 min. An API 3000 (Applied Biosystems/MDS SCIEX) tandem mass spectrometer was used in the positive ion mode with atmospheric pressure photo ionization (APPI) interface. Quantitative analysis was performed in the multiple reaction monitoring (MRM) mode. The MRM transitions monitored were m/z: 363 to 121 and 363 to 97 for cortisol, 361 to 163 and 361 to 105 for cortisone, and 367 to 121 and 367 to 97 for d4-cortisol. The product ions m/z 121 and 163 were quantitative, while the product ions m/z 97 and 105 were qualitative. The calibration standard containing 50 Ag/ l of cortisol was utilized to establish the qualitative ratio of the intensities of the product ion fragments, m/z 363 to 121 and 363 to 97 for cortisol, m/z 361 to 163 and 361 to 105 for cortisone, and m/z 367 to 121 and 367 to 97 for d4-cortisol. The qualitative ratio acceptability limits for the controls and test samples were set as F40% of the values established with the calibration standard. Toluene was used as a dopant to promote ionization at a flow rate of 75 Al/min. Ionization was induced by exposure of LC column effluent to the krypton lamp. The APPI voltage was 1400 V, the orifice voltage was 35 V. The collision gas pressure was 5 mTorr; collision energy was set at 35 V for transitions m/z 363 to 121, 361 to 163, and 367 to 121; and 45 V for transitions m/ z 363 to 97, 361 to 105, and 367 to 97. Quantitative data
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
analysis was performed with Analystk software (Applied Biosystems/MDS SCIEX). Cortisol measurement by IA method were performed on an Immulite 2000 analyzer (Diagnostic Products Corporation) using manufacturer’s reagents according to their instructions.
Results Because the MS/MS detection is very selective, LC separation plays less role for the assay compared to other detection modes. A short LC column was used to separate compounds from the solvent front to eliminate possible signal suppression. The chromatogram for the monitored mass ion transitions for extracted plasma sample is presented in Fig. 1. Method sensitivity was determined by analyzing cortisone and cortisol spiked in BSA at progressively lower concentrations. Each standard was analyzed in duplicate. The limit of quantitation (LOQ) was determined as the
359
Table 1 Method imprecision Cortisol Concentration Within-run (Ag/l) CV (%)
Between-run Total CV (%) CV (%)
2a 25.5a 50a 6b 30b
12 10 10 7 8
3 6 5 5 4
13 12 11 8 9
Cortisone Concentration Within-run (Ag/l) CV (%)
Between-run Total CV (%) CV (%)
2a 5.5a 10a 2b 10b
22 13 12 9 7
5 6 6 14 5
22 14 14 16 8
Ratio cortisone/cortisol Ratio
Cortisone/cortisol Within-run concentration CV (%) (Ag/dl)
Between-run Total CV (%) CV (%)
1a 0.216a 0.200a 0.333b 0.333b
2/2 5.5/25.5 10/50 2/6 10/30
19 21 19 7 10
a b
5 5 4 11 5
19 22 19 13 11
Three replicates per run over 6 days. Two replicates per run over 6 days.
lowest concentration at which accuracy was within F20% and imprecision was within F15%. The LOQ was found to be 5 Ag/l for both cortisol and cortisone. The limit of detection (LOD) was determined as the lowest concentration at which signals were five times greater than background noise for both primary and secondary transitions. The LOD
Fig. 1. MRM chromatograms of extracted plasma sample containing 55 Ag/l of cortisol, and 18 Ag/l of cortisone. Ion transitions: cortisol m/z 363!121 and 363!97, (A); cortisone m/z 361!163 and 361!105, (B); d4-cortisol m/z 367!121 and 367!97, (C).
Fig. 2. Comparison of LC-MS/MS values with immunoassay for cortisol analysis in serum samples (n = 71).
360
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
was determined to be 1 Ag/l for cortisol and 5 Ag/l for cortisone. Method linearity was evaluated by analyzing BSA spiked samples containing cortisone and cortisol at concentrations of 50, 1600, 3200, 4800, 6400, and 8000 Ag/l. Each sample was analyzed in duplicate, and concentrations were calculated from a calibration curve established with standards containing 10, 50, 100, and 200 Ag/l of each compound. The method was found to be linear up to 8000 Ag/ l for cortisol and to 1600 Ag/l for cortisone, with inaccuracy at the highest concentration of 6% and 14%, respectively. This linearity range far exceeds the normal and pathologic concentrations of cortisol and cortisone in human serum and plasma. Within-run and between-run imprecision for the LC-MS/ MS method was determined by analysis of BSA supple-
mented with cortisol and cortisone to concentrations ranging from 25 to 450 and 20 to 100 Ag/l, respectively. The experiments were performed in triplicate over a period of 6 days. Additional imprecision data were obtained from results of controls included with samples during routine method use. A summary of the results of these imprecision studies is presented in Table 1. Method performance for cortisol analysis was evaluated by comparison with an IA method utilizing 71 human serum samples (Fig. 2). The regression equation, correlation coefficient and standard error for the comparison were y = 0.74x 0.70, r = 0.965 and Sy/x = 32.7 Ag/l respectively. Nineteen plasma samples with cortisol concentrations ranging between 0 and 650 Ag/l were analyzed by a LC-MS/MS method at another laboratory. The regression equation, correlation coefficient, and standard error for the results were y = 1.00x + 0.03 and r = 0.994 and 1.84 Ag/l. The LC-MS/MS method was used to establish reference intervals for cortisone, cortisol, and the their ratio in an adult population. The median concentration of cortisol and cortisone were 79.8 and 14.8 Ag/l for male, and 83.6 and 13.7 Ag/l for female, respectively. The median value for the ratio of cortisone to cortisol was 0.169 for male and 0.160 for female. Nonparametric reference intervals determined as central 95% were 33 –246 Ag/l for cortisol, 8 – 27 Ag/l for cortisone, and 0.081 –0.301 for the cortisone/cortisol ratio. Histograms with the distribution of cortisol, cortisone, and their ratio are presented in Fig. 3.
Discussion
Fig. 3. Distribution of concentrations of cortisol (A) and cortisone (B), and the cortisone/cortisol ratio (C) in samples utilized for the reference interval study.
Analysis of cortisol and cortisone is analytically challenging because there are many structurally similar steroids and metabolites that may be present in biological samples. Immunoassay-based methods may lack specificity. They may also utilize method-specific reference intervals, which take into account cross-reactivity with other endogenous steroids. IA-based methods are not ideal when accurate measurement of cortisol and cortisone is required. High specificity detection provided by MS/MS based methods for cortisol and cortisone can produce more accurate results and eliminate potential interference. Determination of the ratio of cortisol and cortisone requires high specificity, good accuracy, and sufficient precision for both analytes; otherwise, it may not be useful. Method precision is important for the combined test because the imprecision for measurement of the ratio of the concentrations of two analytes is potentially greater than imprecision of each individual measurement. When the ratio of concentrations needs to be determined, it is preferable to simultaneously measure both analytes because some factors affecting method performance and related to sample handling (volume of aliquot, reconstitution volume, etc.) during the analysis are canceled out as they equally affect both analytes.
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
Various ionization techniques can be used for LC-MS/ MS analysis of steroids. It was reported that photoionization (APPI) might be more efficient for analysis of steroids [22] compared to other ionization techniques. APPI interface is an evolution of atmospheric pressure ionization (APCI) in which UV light initiates the ionization process, as opposed to corona discharge as in APCI. The principle utilized for the ionization is based on irradiating the vaporized eluent stream with UV light, which initiates a cascade of gas-phase reactions resulting in ionization of the analyte molecules. Toluene has a low ionization energy (8.8 eV) and is commonly used as a dopant to improve ionization efficiency of analytes. When dopant molecules are exposed to photons emitted by the light source, they produce photo ions. In addition to a direct charge-transfer reaction from photo ions to compounds with low ionization energy, the photo ions initiate a cascade of reactions through formation of active intermediate products that lead to ionization of the analytes. A short LC column was used to separate cortisol and cortisone from potentially interfering compounds with similar characteristic MS/MS transitions and from the solvent front to eliminate possible signal suppression [24]. Because of high specificity of the MS/MS detection, the short column allowed obtaining fast analysis time without sacrifice in the method performance. Comparison of recovery between an extracted sample and cortisol and cortisone spiked in the mobile phase at the same concentrations showed that ion suppression did not occur under the LC conditions utilized. Most of the published methods for cortisol analysis utilize reverse phase octadecyl LC columns. A potential problem with this column is interference of tetrahydroprednisolone with cortisol. Tetrahydroprednisolone is a prednisolone metabolite, an isomer of cortisol, and it is present in samples of patients taking prednisone and prednisolone. A phenyl – hexyl column utilized in this method can separate these isomers; produce good peak shape and sufficient separation from the solvent front. The phenyl –hexyl bonded phase has selectivity based on k –k interactions between the stationary phase and an analyte, and proved to be useful for analysis of steroids with strong structural similarities. No compounds were found to interfere with cortisol. Prednisolone, an isomer of cortisone, has the same parent ion as cortisone, but has different product ions. This results in significantly lower sensitivity to prednisolone compared to cortisone. When prednisolone is present in a sample it elutes with a slightly shorter retention time compared to cortisone and produces elevated branching ratio for the transitions 361 to 105/361 to 163. The magnitude of the branching ratio depends on relative concentration of the analytes. Utilization of two transitions and branching ratios for each analyte allows one to qualitatively confirm the identity of the compounds and to detect interfering peaks if other isobaric compounds with similar product ions are present in a sample [23].
361
No significant differences were observed for the cortisol and cortisone concentrations and for the cortisone/cortisol ratio between the genders. Previously, cortisol concentrations in plasma for adults were reported to be 60 –230 Ag/ l by an IA-based method [16]. Lower reference intervals for urinary free cortisol were reported in adult females by LCMS/MS based methods compared with IA methods [17,20]. The results of the method comparison with the IA method suggest a negative bias of 26%. This may be due to cross-reactivity with steroid metabolites in the IA method. The difference in the cross-reactivity associated with IA may also be reflected in the reference intervals for adult cortisol values that are usually greater for IA-based methods. Measurement of the ratio of free cortisol and cortisone in urine also has clinical value and can be utilized for diagnosis of AME and other conditions. Normal interval for the ratio of urinary free cortisone and urinary free cortisol was reported to be between 1.2 and 4.2 [9,20]. The significant difference observed for the cortisone/cortisol ratio between serum and urine can be explained by inactivation of cortisol to cortisone that takes place in kidney, and lower binding of cortisone to plasma proteins that lead to its faster excretion. A potential pitfall in the analysis for cortisol and cortisone is poor stability of the compounds. The stability of cortisol and cortisone was assessed in plasma samples. Samples with a normal and an elevated concentration of each analyte were frozen and thawed three times and analyzed between each cycle. No degradation was observed after the freeze – thaw cycle for either cortisol or cortisone. In addition, samples with normal and elevated concentrations were stored at room temperature and analyzed over a week of storage. Storage of the samples at room temperature did not have an affect on the cortisol, but decreased the concentration of cortisone by 55%. We also observed an approximately 60% degradation of cortisone in calibration standard solution after 2 months of storage at 4 –8jC when it was prepared in water. No degradation of cortisol was observed under these conditions. The stability of cortisone improved when the working calibration standard solution was prepared in methanol or in mix of methanol with water (1:1) and stored at 70jC. While cortisol was stable in storage, we observed its loss during sample preparation. Sample instability is harmful for the combined test for cortisol and cortisone, and may lead to errors in the ratio of their concentrations. One possible explanation for the loss of cortisol is its binding to glass surfaces [20]. To minimize loss, Taylor et al. [20] suggested adding to the standards another steroid, estriol, which would compete with cortisol binding to active silanols on glass tube surface. Because concentration of estriol is significantly greater compared to the concentration of cortisol, this would have a stabilizing effect. To compensate for the degradation, it is also important to prepare the calibration standards in a matrix similar to the matrix of the samples, and to utilize an internal
362
M.M. Kushnir et al. / Clinical Biochemistry 37 (2004) 357–362
standard with properties closely matching those of the analytes of interest.
[7] [8]
Conclusions In conclusion, we developed a sensitive and specific method for analysis of cortisol and cortisone in serum and plasma that utilizes atmospheric pressure photoionization interface and tandem mass spectrometry. The method is very specific, does not have any known interference, and has sufficient sensitivity and precision. Monitoring two ion transitions for each analyte results in high specificity of the method, which is necessary for a test targeted for determining the ratio between cotrisol and cortisone. The benefit of simultaneous measurement of both compounds in single analysis is in reduction of imprecision related to sample handling. Considering the proposed links between cortisone/cortisol ratio and a variety of diseases, the test will be very useful for diagnostic purposes. The small sample volume required for this test helps to reduce the volume of blood drawn from patients and would be especially useful for pediatric testing.
[9]
[10] [11]
[12]
[13]
[14]
[15]
[16]
Acknowledgments This work was supported by ARUP Institute for Clinical and Experimental Pathology.
[17]
[18]
References [1] Tiosano D, Eisentein I, Militianu D, Chrousos GP, Hochberg Z. 11h-hydroxysteroid dehydrogenase activity in hypothalamic obesity. J Clin Endocrinol Metab 2003;88:379 – 84. [2] Stewart PM. 11h-hydroxysteroid dehydrogenase: implications for clinical medicine. Clin Endocrinol (Oxford) 1996;44:493 – 9. [3] Bujalska IJ, Kumar S, Stewart PM. Does central obesity reflect ‘‘Cushing’s disease of the omentum’’? Lancet 1997;349:1210 – 3. [4] Stewart PM, Boulton A, Kumar S, Clark PM, Shackleton CHL. Cortisol metabolism in human obesity: impaired cortisone ! cortisol conversion in subjects with central adiposity. J Clin Endocrinol Metab 1999;84:1022 – 7. [5] Reinke M. Subclinical Cushing’s syndrome. Endocrinol Metab Clin North Am 2000;29:43 – 56. [6] Andrews RC, Herlihy O, Livingstone DEW, Andrew R, Walker BR. Abnormal cortisol metabolism and tissue sensitivity to cortisol in
[19]
[20]
[21]
[22]
[23]
[24]
patients with glucose intolerance. J Clin Endocrinol Metab 2002;87: 5587 – 93. Andrews RC, Walker BR. Glucocorticoids and insulin resistance: old hormones, new targets. Clin Sci 1999;96:513 – 23. Phillips DIW, Barker DJP, Fall CHD, et al. Elevated plasma cortisol concentrations: an explanation for the relationship between low birth weight and adult cardiovascular risk factors. J Clin Endocrinol Metab 1998;83:757 – 60. Palermo M, Delitala G, Mantero F, Stewart PM, Shackleton CHL. Congenital deficiency of 11h-hydroxysteroid dehydrogenase (apparent mineralocorticoid excess syndrome): diagnostic value of urinary free cortisol and cortisone. J Endocrinol Invest 2001;24:17 – 23. Shimojo M, Stewart PM. Apparent mineralocorticoid excess syndrome. J Endocrinol Invest 1995;51:518 – 32. Stewart PM, Wallace AM, Valentino R, Burt D, Shackleton CHL, Edwards CRW. Mineralocorticoid activity of licorice: 11h-hydroxysteroid dehydrogenase comes of age. Lancet 1987;2:821 – 4. Edwards CRW, Walker BR, Benediktsson R, Seckl JR. Congenital and acquired syndrome of apparent mineralocorticoid excess. J Steroid Biochem Mol Biol 1983;45:1 – 5. Mune T, Morita H, Suzuki T, et al. Role of local 11h-hydroxysteroid dehydrogenase type 2 expression in determining the phenotype of adrenal adenomas. J Clin Endocrinol Metab 2003;88:864 – 70. Funder JW, Pearce PT, Smith R, Smith AI. Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science 1988;242:583 – 5. White PC, Mune T, Agarwal AK. 11h-hydroxysteroid dehydrogenase and the syndrome of apparent mineralocorticoid excess. Endocr Rev 1997;18:135 – 56. Tello FL, Hernandez DM. Performance evaluation of nine hormone assays on the Immulite 2000 immunoassay system. Clin Chem Lab Med 2000;38:1039 – 42. Murphy BEP. Lack of specificity of urinary free cortisol determinations: why does it continue? J Clin Endocrinol Metab 1999;884: 2258 – 9. Murphy BEP. How much ‘‘UFC’’ is really cortisol? Clin Chem 2000;46(6):793 – 4. Vogeser M, Zachoval R, Felbinger TW, Jacob K. Increased ratio of serum cortisol to cortisone in acute-phase response. Horm Res 2002; 15:269 – 78. Taylor RL, Machacek D, Singh RJ. Validation of a high-throughput liquid chromatography-tandem mass spectrometry method for urinary cortisol and cortisone. Clin Chem 2002;48:1511 – 9. Kushnir MM, Rockwood AL, Nelson GN, Terry AH, Meikle AW. Analysis of cortisol in urine by LC-MS/MS. Clin Chem 2003;49: 965 – 7. Robb DB, Covey TR, Bruins AP. Atmospheric pressure photoionization: an ionization method for liquid chromatography-mass spectrometry. Anal Chem 2000;72:3653 – 9. Meikle AW, Findling J, Kushnir MM, Rockwood AL, Nelson GJ, Terry AH. Pseudocushings syndrome cause by interference of drug Fenofibrate with HPLC assay for cortisol. J Clin Endocrinol Metab 2003;88:3521 – 4. Annesley TM. Ion suppression in mass spectrometry. Clin Chem 2003;49:1041 – 4.