- Email: [email protected]
Dialyzable free cortisol after stimulation with Synacthen®
Clinical Biochemistry 35 (2002) 539 –543
Dialyzable Free Cortisol after Stimulation with Synacthen威 Michael Vogesera,*, Josef Briegelb, Reinhart Zachovalc a
Institute of Clinical Chemistry Ludwig-Maximilians-Universita¨t Munich, Klinikum Grosshadern, 81366 Munich, Germany b Clinic for Anesthesiology, Ludwig-Maximilians-Universita¨t Munich, Klinikum Grosshadern, 81366 Munich, Germany c Department of Internal Medicine II, Ludwig-Maximilians-Universita¨t Munich, Klinikum Grosshadern, 81366 Munich, Germany Received 1 May 2002Accepted 30 August 2002
Abstract Objectives: To compare the changes in free vs. total serum cortisol concentrations after acute stimulation of the adrenal cortex. Design and Methods: Paired serum samples of ten individuals taken immediately before and 1 h after stimulation with 250 g ACTH (1–24) (Synacthen威) given i.v. were analyzed. Total cortisol was quantified using liquid chromatography tandem-mass spectrometry with an online sample extraction system and tri-deuterated cortisol as the internal standard. Free cortisol was measured with the same method after equilibrium dialysis. Concentrations of the corticosteroid-binding globulin (CBG) were determined by radioimmuno assay. Results: Total cortisol increased by a mean of 106% (mean basal cortisol 312 nmol/L (SD 140 nmol/L), stimulated 686 nmol/L (SD 163 nmol/L); p ⬍ 0.001, paired t-test for differences); no significant change of CBG concentrations was found (874 nmol/L (SD 179 nmol/L) before stimulation, 869 nmol/L (SD 225 nmol/L) after stimulation). The mean increase of free cortisol was 263% (mean basal free cortisol 20.3 nmol/L (SD 13.2 nmol/L), stimulated 73.8 nmol/L (SD 26.7 nmol/L); p ⬍ 0.001) and thus substantially more pronounced compared to the increase of total cortisol (p ⬍ 0.01). The ratio of free to total serum cortisol was significantly increased after stimulation (6.1% (SD 1.7%) before stimulation, 10.6% (SD 1.9%) after stimulation; p ⬍ 0.001). Conclusions: After acute neuroendocrine stimulation of the adrenal cortex the relative increase of free bioactive cortisol concentrations is substantially more pronounced than the increase of total cortisol concentrations. © 2002 The Canadian Society of Clinical Chemists. All rights reserved. Keywords: Cortisol; ACTH; Synacthen威; Equilibrium dialysis; Liquid chromatography tandem-mass spectrometry
1. Introduction Corticosteroid-binding globulin (CBG) is the main transport protein of cortisol in the circulation [1–3]. Under normal conditions the molar concentration of CBG is about twice that of cortisol and less then 10% of serum cortisol is unbound; only this free fraction is considered to be biologically active [4,5]. Based on the binding characteristics of CBG a calculation of free serum cortisol concentrations is possible [6]. If such calculation incorporates the concentration of competing steroids (especially cortisone) and besides the concentration of CBG, that of albumin (the second most important transport protein of cortisol) very complex nonlinear equations have to be applied that can only be solved * Corresponding author. Institut fu¨r Klinische Chemie, Klinikum Grosshadern der Ludwig-Maximilians-Universita¨t, D-81366 Munich. Tel.: 0049 89/7095-3246; FAX: 0049 89/7095-3240 E-mail address: [email protected] (M. Vogeser).
by numeric methods and computer simulation [5,6]. A simplified calculation has been suggested by Coolens et al. [7] that only involves the concentrations of total cortisol and CBG and can be performed using standard software (e.g., Microsoft Excel). This approach is clinically useful for example in cases of suspected Cushing’s syndrome with increased CBG concentrations e.g., induced by exogenous estrogens. Close correlation between concentrations of free cortisol calculated according to Coolens et al. and free cortisol measured after ultrafiltration has been shown for single samples taken under baseline morning conditions [7]. Calculation of free serum cortisol using either computer simulation [6] or the simplified equation of Coolens et al. similarly predicts a nonlinear behavior of free cortisol concentrations along with increasing total cortisol and constant CBG concentrations, with an increase of the free-to-bound ratio of cortisol. The aim of the present study was to verify this mathematical prediction by measuring the serum concentrations of free cortisol after equilibrium dialysis and of
0009-9120/02/$ – see front matter © 2002 The Canadian Society of Clinical Chemists. All rights reserved. PII: S 0 0 0 9 - 9 1 2 0 ( 0 2 ) 0 0 3 5 8 - 2
540
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543
total cortisol in paired samples taken before and following stimulation with the ACTH analogon Synacthen.
2. Patients and methods 2.1. Patients Residual serum samples were used which had been submitted to the Institute of Clinical Chemistry for the determination of cortisol and/or 17␣-OH-progesterone in the context of clinically indicated Synacthen威 stimulation testing; the tests were performed to exclude heterozygote or late-onset adrenogenital syndrome or Addison’s disease. Samples were taken before and 60 min. after i.v. application of 250 g of Synacthen威 (Novartis Pharma, Nu¨ rnberg, Germany), a synthetic corticotropin preparation (ACTH 1–24) displaying all the pharmacological properties of endogenous ACTH. Ten consecutive paired samples were included in which normal serum cortisol response to Synacthen威 defined by a stimulated total cortisol concentration of at least 550 nmol/L was found. Six patients were female, four male; the median age of the patients was 24 yr (range 18 –57). The study was approved by the institutional review board. 2.2. Analytical methods Serum cortisol and cortisone were determined simultaneously as previously described [8,9] using liquid chromatography electrospray tandem mass spectrometry (LC-ESI/ MS/MS). In brief, samples and calibrators were precipitated with methanol/zinc sulfate containing tri-deuterated cortisol as the internal standard (Cambridge Isotope Laboratories, Andover, USA). After centrifugation the supernatants were applied to online solid-phase extraction with subsequent HPLC separation employing column switching (extraction column: Waters Oasis HLB (Waters, Milford, USA), washed with 5% methanol, elution in back-flush with methanol/2 mM ammonium acetate 7/3 by volume onto an analytical C18-HPLC column (Reprosil pur C18-AQ, 125 ⫻ 2 mm; 5 m, Maisch, Ammerbuch, Germany). The HPLC system was coupled to a Micromass Quattro LC tandem mass spectrometer (Micromass, Manchester, UK) run in multiple reaction monitoring mode; the following transitions generated by collision induced desintegration of the respective [M⫹H]⫹ ions were monitored: cortisol 363 ⬎ 309 m/z, cortisone 361 ⬎ 163 m/z, and d3-cortisol 366 ⬎ 312 m/z. All study samples were analyzed within two runs. Equilibrium dialysis cells were prepared according to Reinard and Jacobsen [10] from commercially available 1.5 mL polypropylene microtubes (Eppendorf, Germany; Figure. 1). Lids of these tubes were cut off, placed upside down so that their recess functioned as a reservoir for approx 240 L of a dialysis buffer (Nichols Dialysis Buffer, cat.no. 30 – 0650). A dialysis membrane (regenerated cellulose,
Fig. 1. Directly measured free serum cortisol concentrations (determined after equilibrium dialysis by liquid chromatography-tandem mass spectrometry) vs. calculated free serum cortisol concentrations [7] in 10 patients undergoing adrenal stimulation with Synathen威 (circles, baseline samples; squares, samples taken 60 min. after stimulation).
molecular weight cut-off 4.000 – 6.000 Da.; Roth GmbH, Karlsruhe, Germany) was conditioned by first soaking in a large volume of 10 mM sodium bicarbonate at 60°C for 1 h while stirring and subsequently by soaking in destilled water at the same temperature again for 1 h. After final washing in a large volume of destilled water the membrane was cut in 1.5 ⫻ 1.5 cm squares to be placed on the tube lids filled with the dialysis buffer. Tubes were then placed upside down on top of the lids, thereby fixing the dialysis membranes. Finally the tip of the tubes was cut off to fill 1 mL of serum into the dialysis cells. The cells were incubated for 16 h on a reciprocal shaker placed in a thermostated oven at 37°C (⫾ 0.1°C). After incubation the serum containing tubes were removed and the dialysate was collected from the lids to be analyzed by LC-ESI/MS/MS in the same way as serum samples. For quality control of total cortisol measurement commercially available materials were used (Lyphochek Immunoassay Plus, Biorad, Hercules, USA; level 1, target concentration 82 nmol/L, accepted range ⫾ 12%; level 2, target concentration 501 nmol/L, accepted range ⫾ 12%); additionally pooled serum was assayed in triplicate in both analytical series for total cortisol and for free cortisol after equilibrium dialysis. Serum corticosteroid-binding globulin was quantified by a coated-tube RIA (BioSource Europe, Nivelles, Belgium) in one single batch using kit control material for quality assurance. To calculate serum free cortisol concentrations from total serum cortisol and CBG concentrations the equation described by Coolens et al. [7] was applied
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543 Table 1 Total serum cortisol and free serum cortisol concentrations (nmol/L) determined by equilibrium dialysis and liquid chromatography-tandem mass spectrometry in 10 patients undergoing adrenal stimulation with Synathen威 Pat. #
1 2 3 4 5 6 7 8 9 10
Age, sex
21, 57, 21, 44, 22, 26, 18, 18, 27, 33,
female male male female female female male female male female
Total cortisol
Free cortisol
0 min
60 min
0 min
60 min
349 566 380 377 326 95 161 434 120 320
663 969 718 862 670 749 429 757 409 638
33 50 21 17 20 4 8 27 7 15
84 138 66 82 82 67 33 81 45 60
(C: CBG (mol/L); T: total cortisol (mol/L); SQR for square root): Free serum cortisol ⫽ SQR ((0.0167 ⫹ 0.0182 ⫻ (C ⫺ T))2 ⫹ 0.0122xT) ⫺0.0167 ⫹ 0.0.182 ⫻ (C ⫺T) A free cortisol index was calculated as [total cortisol]/ [CBG]. 2.3. Statistical methods The paired t-test for differences was used to test for statistically significant changes of the respective analyte concentrations and proportions before and after Synacthen威 stimulation, with significance assumed for p ⬍ 0.05.
3. Results In all subjects studied the percentage increment of free serum cortisol measured by LC-MS/MS after equilibrium dialysis was more pronounced compared to the increment of total serum cortisol (Table 1); this was shown to be statistically significant by paired t-test. Mean total cortisol was 312 nmol/L (SD 140 nmol/L) before stimulation and 686 nmol/L (SD 163 nmol/L) after stimulation, representing an increase by 106% (p ⬍ 0.001) (Table 2). Mean dialyzable free cortisol was 20.3 nmol/L (SD 13.2 nmol/L) before stimulation and 73.8 nmol/L (SD 26.7 nmol/L) after stimulation which corresponds to an increase by 263% (p ⬍ 0.001). The difference in the percentage increment significantly differed between total cortisol and measured free cortisol respectively (p ⬍ 0.01). The mean ratio of dialyzable free cortisol to total cortisol was significantly higher after stimulation than before (6.1% (SD 1.7%) before stimulation vs. 10.6% (SD 1.9%) after stimulation (p ⬍ 0.001)). CBG concentrations did not change significantly with stimulation (874 nmol/L (SD 179 nmol/L) before vs. 869 nmol/L (SD 225 nmol/L) after stimulation). Consequently, the percentage increase of the free cortisol index (total
541
Table 2 Concentrations of total serum cortisol, corticosteroid binding globulin, free serum cortisol, cortisol/CBG ratio, and percentage of free cortisol in samples of ten individuals taken before and after stimulation with 250 g Synacthen (mean and standard deviation) 0 min. Total serum cortisol (nmol/L) CBG (nmol/L) Cortisol/CBG Free serum cortisol (nmol/L) % free cortisol of total cortisol
60 min. after 250 g Synacthen
312 (SD 140) 687 (SD 163) 874 (SD 179) 869 (SD 225) 0.35 (SD 0.13) 0.80 (SD 0.15) 20.3 (SD 13.2) 73.8 (SD 26.7) 6.1 (SD 1.7) 10.6 (SD 1.9)
cortisol/CBG) was similar to the increase of total cortisol in the patients studied (mean 0.35 (SD 0.13) before stimulation and 0.80 (SD 0.15) after stimulation). Close correlation between free cortisol concentrations calculated according to Coolens et al. [7] and measured free cortisol after equilibrium dialysis was found (Figure. 1); by regression analysis a Pearson’s coefficient of correlation of 0.904 and the following equation was found: Free cortisol by equilibrium dialysis ⫽ 1.6 x calculated free cortisol ⫹ 2.2 nmol/L The cortisol/cortisone ratio increased from a mean of 5.90 (SD 2.9) before stimulation to 19.0 (SD 9.2) after stimulation. Mean total cortisone was 52.5 nmol/L (SD 14.3) before stimulation and 41.8 nmol/L (SD 13.7 nmol/L) after stimulation, mean free cortisone was 19.0 nmol/L (SD 5.2 nmol/L) before stimulation and 15.2 nmol/L (SD 5.0 nmol/L) after stimulation representing 36% of total cortisone at both sampling times. The mean ratio of free cortisol to free cortisone was 2.1 (SD 1.1) before stimulation and 8.1 (SD 3.9) after stimulation. In the two analytical series an inter-assay coefficient of variation of 2.9% for total cortisol and of 5.8% for free cortisol after equilibrium dialysis was found for the human serum pool (n ⫽ 6; total cortisol, 441 nmol/L; free cortisol, 29 nmol/L); the signal-to-noise ratio of the specific cortisol signal in the dialyzed samples was between 19:1 and 21:1. All quality control samples were found within the acceptable ranges.
4. Discussion In our study we were able to demonstrate that following adrenocortical stimulation by the ACTH analog Synacthen, free serum cortisol concentrations show a substantially more pronounced relative increase compared to total serum cortisol concentrations with a significant increase of the free-to-bound ratio of cortisol. These data suggest that measurement of bioactive free cortisol may reflect biologically more appropriately the glucocorticoid activation during neuroendocrine stimulation as compared to measurement of
542
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543
total cortisol. The nonlinear behavior of free serum cortisol concentrations in situations of acutely increased total serum cortisol and constant corticosteroid-binding globulin (CBG) concentrations is predicted by mathematical models that take into account the binding characteristics of CBG and albumin [5, 6]. These calculations, however, are most complex and require commercially not available software for numerical procedures. The calculation method described by Coolens et al. [7] is substantially simplified in that for example albumin concentrations and the concentrations of cortisone are not incorporated, however, this method is easily applicable employing standard software. We found a close correlation between free cortisol concentrations calculated in the latter way and directly measured concentrations, however, a slope of 1.6 was found with higher concentrations obtained by direct measurement; this proportional bias may arise from the simplifications made in this mathematical approach. Coolens et al. had as well compared calculated with directly measured free cortisol concentrations applying ultrafiltration and immunoassay but no relevant bias was found. This discrepancy –the apparent good agreement between calculation and measurement of free cortisol found by Coolens et al. but not in our investigation-might be due to methodological differences in the measurement of free serum cortisol. A rapid ultrafiltration method as used by Coolens et al. is - due to protein concentration effects on the filtration membrane - potentially more prone to artifacts than is equilibrium dialysis over an extended time period. A simple free cortisol index (total cortisol/CBG) as proposed previously [11,12] in contrast to the albeit simplified calculation of free serum cortisol from total serum cortisol and CBG (according to Coolens et al.) is obviously not able to display the over-proportional increase of free cortisol compared to total cortisol after neuroendocrine stimulation. The inexpensive equilibrium dialysis method applied in our study to quantify free serum cortisol proved convenient with good reproducibility (inter-assay coefficient of variation of 5.5%) and a favorable chromatographic signal-tonoise ratio in a normal serum pool; applicability of equilibrium dialysis to the assessment of free cortisol has been shown previously [13]. The simple dialysis method described here in combination with the highly specific and sensitive liquid-chromatographic tandem-mass spectrometric assay for the determination of cortisol may be considered as a candidate reference method for measurement of free cortisol. The method may be applicable in the development of direct immunoassays to quantitate free serum cortisol, by analogy to the measurement of free thyroid hormones. Such assays may be particularly useful for the monitoring of disease related hypercortisolism and neuroendocrine stimulation tests in severely ill patients [14,15] especially since critical illness is typically associated with a decline of CBG and albumin concentrations [16 –18]. We found the concentrations of CBG not to be affected by acute stimulation with ACTH; it could have been spec-
ulated that acute short term neuroendocrine stimulation could lead to cleavage and depletion of CBG as it is found in inflammatory states [16 –19]; in rats indeed a decline of CBG was found after stimulation with ACTH [20]. We report—to our knowledge—for the first time directly measured concentrations of free cortisone in men. Our results confirm that cortisone is bound to serum proteins to a lower extent compared to cortisol [6]. The ratio of free cortisol to free cortisone in consequence differs from the ratio of total cortisol to total cortisone. We have previously shown that acute stimulation with exogenous ACTH induces a systemic shift from inactive cortisone to active cortisol, suggestive of an increased activity of 11-hydroxysteroid dehydrogenase type 1 under this condition [9]; these results have been confirmed by the present study. The main results of the latter and the present investigation are in line in that both, the shift of the systemic cortisol-to-cortisone ratio and the nonlinear, over-proportional increase of free bioactive cortisol with a substantial rise in the free-to-bound ratio lead to an enhanced availability of cortisol in acute neuroendocrine stimulation in addition to the mere increase in the adrenocortical cortisol synthesis. References [1] Heyns W, Coolens JL. Physiology of corticosteroid-binding globulin in humans. Ann New York Acad Sci 1988;538:122–9. [2] Hammond GL, Smith CL, Underhill DA. Molecular studies of corticosteroid binding globulin structure, biosynthesis and function. J Steroid Biochem Mol Biol 1991;40:755– 62. [3] Rosner W. Plasma steroid-binding globulins. Endocrinol Metab Clin North Am 1991;20:697–720. [4] Brien TG. Pathophysiology of free cortisol in plasma. Ann New York Acad Sci 1988;538:130 – 6. [5] Ekins R. Measurement of free hormones in blood. Endocr Rev 1990; 11:5– 46. [6] Dunn JF, Nisula BC, Rodbard D. Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma. J Clin Endocrinol Metab 1981;53:58 – 68. [7] Coolens JL, Van Baelen H, Heyns W. Clinical use of unbound plasma cortisol as calculated from total cortisol and corticosteroid-binding globulin. J Steroid Biochem 1987;26:197–202. [8] Vogeser M, Briegel J, Jacob K. Determination of serum cortisol by isotope-dilution liquid-chromatography electrospray ionization tandem mass spectrometry with on-line extraction. Clin Chem Lab Med 2001;39:944 –7. [9] Vogeser M, Zachoval R, Jacob K. Serum cortisol/cortisone ratio after Synacthen stimulation. Clin Biochem 2001;34:421–5. [10] Reinard T, Jacobsen HJ. An inexpensive small volume equilibrium dialysis system for protein-ligand binding assays. Anal Biochem 1989;176:157– 60. [11] Bonte HA, van den Hoven R, van der Sluijs Veer G, Vermes I. The use of free cortisol index for laboratory assessment of pituitaryadrenal function. Clin Chem Lab Med 1999;37:127–32. [12] Beisenhuizen A, Thijs LG, Vermes I. Patterns of corticosteroidbinding globulin and the free cortisol index during septic shock and multitrauma. Intensive Care Med 2001;27:1584 –91. [13] Clerico A, Del Chicca MG, Ferdeghini M, Ghione S, Materazzi F. Progressively elevated levels of biologically active (free) cortisol
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543 during pregnancy by a direct radioimmunological assay of diffusible cortisol in an equilibrium dialysis system. J Endocrinol Invest 1980; 3:185–7. [14] Briegel J, Kilger E, Schelling G. Stress doses of hydrocortisone in septic shock: Beyond the hemodynemic effects. Vincent JL, editor. Yearbook of Intensive Care and Emergency Medicine. Berlin, Germany: Springer, 1999. 189 –198. [15] Annane D, Sebille V, Troche G, Raphael JC, Gajdos P, Bellissant E. A 3-level prognostic classification in septic shock based on cortisol levels and cortisol response to corticotropin. JAMA 2000;283:1038 – 45. [16] Garrel DR. Corticosteroid-binding globulin during inflammation and burn injury: nutritional modulation and clinical implication. Horm Res 1996;45:245–51.
543
[17] Bernier J, Jobin N, Emptoz-Bonneton A. Decreased corticosteroidbinding globulin in burn patients: relationship with interleukin-6, and fat in nutritional support. Crit Care Med 1998;26:452– 60. [18] Vogeser M, Felbinger TW, Kilger E, Ro¨ ll W, Fraunberger P, Jacob K. Corticosteroid-binding globulin and free cortisol in the early postoperative period after cardiac surgery. Clin Biochem 1999;32: 213– 6. [19] Pemberton PA, Stein PE, Pepys MB, Potter JM, Carrell RW. Hormone binding globulins undergo serpin conformational change in inflammation. Nature 1988;336:257– 8. [20] Armario A, Giral M, Marti O, et al. The effect of acute, and chronic ACTH adminstration on pituitary-adrenal response to acute immobilization stress. Relationship to changes in corticosteroid-binding globulin. Endocr Res 1994;20:139 – 49.
Dialyzable Free Cortisol after Stimulation with Synacthen威 Michael Vogesera,*, Josef Briegelb, Reinhart Zachovalc a
Institute of Clinical Chemistry Ludwig-Maximilians-Universita¨t Munich, Klinikum Grosshadern, 81366 Munich, Germany b Clinic for Anesthesiology, Ludwig-Maximilians-Universita¨t Munich, Klinikum Grosshadern, 81366 Munich, Germany c Department of Internal Medicine II, Ludwig-Maximilians-Universita¨t Munich, Klinikum Grosshadern, 81366 Munich, Germany Received 1 May 2002Accepted 30 August 2002
Abstract Objectives: To compare the changes in free vs. total serum cortisol concentrations after acute stimulation of the adrenal cortex. Design and Methods: Paired serum samples of ten individuals taken immediately before and 1 h after stimulation with 250 g ACTH (1–24) (Synacthen威) given i.v. were analyzed. Total cortisol was quantified using liquid chromatography tandem-mass spectrometry with an online sample extraction system and tri-deuterated cortisol as the internal standard. Free cortisol was measured with the same method after equilibrium dialysis. Concentrations of the corticosteroid-binding globulin (CBG) were determined by radioimmuno assay. Results: Total cortisol increased by a mean of 106% (mean basal cortisol 312 nmol/L (SD 140 nmol/L), stimulated 686 nmol/L (SD 163 nmol/L); p ⬍ 0.001, paired t-test for differences); no significant change of CBG concentrations was found (874 nmol/L (SD 179 nmol/L) before stimulation, 869 nmol/L (SD 225 nmol/L) after stimulation). The mean increase of free cortisol was 263% (mean basal free cortisol 20.3 nmol/L (SD 13.2 nmol/L), stimulated 73.8 nmol/L (SD 26.7 nmol/L); p ⬍ 0.001) and thus substantially more pronounced compared to the increase of total cortisol (p ⬍ 0.01). The ratio of free to total serum cortisol was significantly increased after stimulation (6.1% (SD 1.7%) before stimulation, 10.6% (SD 1.9%) after stimulation; p ⬍ 0.001). Conclusions: After acute neuroendocrine stimulation of the adrenal cortex the relative increase of free bioactive cortisol concentrations is substantially more pronounced than the increase of total cortisol concentrations. © 2002 The Canadian Society of Clinical Chemists. All rights reserved. Keywords: Cortisol; ACTH; Synacthen威; Equilibrium dialysis; Liquid chromatography tandem-mass spectrometry
1. Introduction Corticosteroid-binding globulin (CBG) is the main transport protein of cortisol in the circulation [1–3]. Under normal conditions the molar concentration of CBG is about twice that of cortisol and less then 10% of serum cortisol is unbound; only this free fraction is considered to be biologically active [4,5]. Based on the binding characteristics of CBG a calculation of free serum cortisol concentrations is possible [6]. If such calculation incorporates the concentration of competing steroids (especially cortisone) and besides the concentration of CBG, that of albumin (the second most important transport protein of cortisol) very complex nonlinear equations have to be applied that can only be solved * Corresponding author. Institut fu¨r Klinische Chemie, Klinikum Grosshadern der Ludwig-Maximilians-Universita¨t, D-81366 Munich. Tel.: 0049 89/7095-3246; FAX: 0049 89/7095-3240 E-mail address: [email protected] (M. Vogeser).
by numeric methods and computer simulation [5,6]. A simplified calculation has been suggested by Coolens et al. [7] that only involves the concentrations of total cortisol and CBG and can be performed using standard software (e.g., Microsoft Excel). This approach is clinically useful for example in cases of suspected Cushing’s syndrome with increased CBG concentrations e.g., induced by exogenous estrogens. Close correlation between concentrations of free cortisol calculated according to Coolens et al. and free cortisol measured after ultrafiltration has been shown for single samples taken under baseline morning conditions [7]. Calculation of free serum cortisol using either computer simulation [6] or the simplified equation of Coolens et al. similarly predicts a nonlinear behavior of free cortisol concentrations along with increasing total cortisol and constant CBG concentrations, with an increase of the free-to-bound ratio of cortisol. The aim of the present study was to verify this mathematical prediction by measuring the serum concentrations of free cortisol after equilibrium dialysis and of
0009-9120/02/$ – see front matter © 2002 The Canadian Society of Clinical Chemists. All rights reserved. PII: S 0 0 0 9 - 9 1 2 0 ( 0 2 ) 0 0 3 5 8 - 2
540
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543
total cortisol in paired samples taken before and following stimulation with the ACTH analogon Synacthen.
2. Patients and methods 2.1. Patients Residual serum samples were used which had been submitted to the Institute of Clinical Chemistry for the determination of cortisol and/or 17␣-OH-progesterone in the context of clinically indicated Synacthen威 stimulation testing; the tests were performed to exclude heterozygote or late-onset adrenogenital syndrome or Addison’s disease. Samples were taken before and 60 min. after i.v. application of 250 g of Synacthen威 (Novartis Pharma, Nu¨ rnberg, Germany), a synthetic corticotropin preparation (ACTH 1–24) displaying all the pharmacological properties of endogenous ACTH. Ten consecutive paired samples were included in which normal serum cortisol response to Synacthen威 defined by a stimulated total cortisol concentration of at least 550 nmol/L was found. Six patients were female, four male; the median age of the patients was 24 yr (range 18 –57). The study was approved by the institutional review board. 2.2. Analytical methods Serum cortisol and cortisone were determined simultaneously as previously described [8,9] using liquid chromatography electrospray tandem mass spectrometry (LC-ESI/ MS/MS). In brief, samples and calibrators were precipitated with methanol/zinc sulfate containing tri-deuterated cortisol as the internal standard (Cambridge Isotope Laboratories, Andover, USA). After centrifugation the supernatants were applied to online solid-phase extraction with subsequent HPLC separation employing column switching (extraction column: Waters Oasis HLB (Waters, Milford, USA), washed with 5% methanol, elution in back-flush with methanol/2 mM ammonium acetate 7/3 by volume onto an analytical C18-HPLC column (Reprosil pur C18-AQ, 125 ⫻ 2 mm; 5 m, Maisch, Ammerbuch, Germany). The HPLC system was coupled to a Micromass Quattro LC tandem mass spectrometer (Micromass, Manchester, UK) run in multiple reaction monitoring mode; the following transitions generated by collision induced desintegration of the respective [M⫹H]⫹ ions were monitored: cortisol 363 ⬎ 309 m/z, cortisone 361 ⬎ 163 m/z, and d3-cortisol 366 ⬎ 312 m/z. All study samples were analyzed within two runs. Equilibrium dialysis cells were prepared according to Reinard and Jacobsen [10] from commercially available 1.5 mL polypropylene microtubes (Eppendorf, Germany; Figure. 1). Lids of these tubes were cut off, placed upside down so that their recess functioned as a reservoir for approx 240 L of a dialysis buffer (Nichols Dialysis Buffer, cat.no. 30 – 0650). A dialysis membrane (regenerated cellulose,
Fig. 1. Directly measured free serum cortisol concentrations (determined after equilibrium dialysis by liquid chromatography-tandem mass spectrometry) vs. calculated free serum cortisol concentrations [7] in 10 patients undergoing adrenal stimulation with Synathen威 (circles, baseline samples; squares, samples taken 60 min. after stimulation).
molecular weight cut-off 4.000 – 6.000 Da.; Roth GmbH, Karlsruhe, Germany) was conditioned by first soaking in a large volume of 10 mM sodium bicarbonate at 60°C for 1 h while stirring and subsequently by soaking in destilled water at the same temperature again for 1 h. After final washing in a large volume of destilled water the membrane was cut in 1.5 ⫻ 1.5 cm squares to be placed on the tube lids filled with the dialysis buffer. Tubes were then placed upside down on top of the lids, thereby fixing the dialysis membranes. Finally the tip of the tubes was cut off to fill 1 mL of serum into the dialysis cells. The cells were incubated for 16 h on a reciprocal shaker placed in a thermostated oven at 37°C (⫾ 0.1°C). After incubation the serum containing tubes were removed and the dialysate was collected from the lids to be analyzed by LC-ESI/MS/MS in the same way as serum samples. For quality control of total cortisol measurement commercially available materials were used (Lyphochek Immunoassay Plus, Biorad, Hercules, USA; level 1, target concentration 82 nmol/L, accepted range ⫾ 12%; level 2, target concentration 501 nmol/L, accepted range ⫾ 12%); additionally pooled serum was assayed in triplicate in both analytical series for total cortisol and for free cortisol after equilibrium dialysis. Serum corticosteroid-binding globulin was quantified by a coated-tube RIA (BioSource Europe, Nivelles, Belgium) in one single batch using kit control material for quality assurance. To calculate serum free cortisol concentrations from total serum cortisol and CBG concentrations the equation described by Coolens et al. [7] was applied
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543 Table 1 Total serum cortisol and free serum cortisol concentrations (nmol/L) determined by equilibrium dialysis and liquid chromatography-tandem mass spectrometry in 10 patients undergoing adrenal stimulation with Synathen威 Pat. #
1 2 3 4 5 6 7 8 9 10
Age, sex
21, 57, 21, 44, 22, 26, 18, 18, 27, 33,
female male male female female female male female male female
Total cortisol
Free cortisol
0 min
60 min
0 min
60 min
349 566 380 377 326 95 161 434 120 320
663 969 718 862 670 749 429 757 409 638
33 50 21 17 20 4 8 27 7 15
84 138 66 82 82 67 33 81 45 60
(C: CBG (mol/L); T: total cortisol (mol/L); SQR for square root): Free serum cortisol ⫽ SQR ((0.0167 ⫹ 0.0182 ⫻ (C ⫺ T))2 ⫹ 0.0122xT) ⫺0.0167 ⫹ 0.0.182 ⫻ (C ⫺T) A free cortisol index was calculated as [total cortisol]/ [CBG]. 2.3. Statistical methods The paired t-test for differences was used to test for statistically significant changes of the respective analyte concentrations and proportions before and after Synacthen威 stimulation, with significance assumed for p ⬍ 0.05.
3. Results In all subjects studied the percentage increment of free serum cortisol measured by LC-MS/MS after equilibrium dialysis was more pronounced compared to the increment of total serum cortisol (Table 1); this was shown to be statistically significant by paired t-test. Mean total cortisol was 312 nmol/L (SD 140 nmol/L) before stimulation and 686 nmol/L (SD 163 nmol/L) after stimulation, representing an increase by 106% (p ⬍ 0.001) (Table 2). Mean dialyzable free cortisol was 20.3 nmol/L (SD 13.2 nmol/L) before stimulation and 73.8 nmol/L (SD 26.7 nmol/L) after stimulation which corresponds to an increase by 263% (p ⬍ 0.001). The difference in the percentage increment significantly differed between total cortisol and measured free cortisol respectively (p ⬍ 0.01). The mean ratio of dialyzable free cortisol to total cortisol was significantly higher after stimulation than before (6.1% (SD 1.7%) before stimulation vs. 10.6% (SD 1.9%) after stimulation (p ⬍ 0.001)). CBG concentrations did not change significantly with stimulation (874 nmol/L (SD 179 nmol/L) before vs. 869 nmol/L (SD 225 nmol/L) after stimulation). Consequently, the percentage increase of the free cortisol index (total
541
Table 2 Concentrations of total serum cortisol, corticosteroid binding globulin, free serum cortisol, cortisol/CBG ratio, and percentage of free cortisol in samples of ten individuals taken before and after stimulation with 250 g Synacthen (mean and standard deviation) 0 min. Total serum cortisol (nmol/L) CBG (nmol/L) Cortisol/CBG Free serum cortisol (nmol/L) % free cortisol of total cortisol
60 min. after 250 g Synacthen
312 (SD 140) 687 (SD 163) 874 (SD 179) 869 (SD 225) 0.35 (SD 0.13) 0.80 (SD 0.15) 20.3 (SD 13.2) 73.8 (SD 26.7) 6.1 (SD 1.7) 10.6 (SD 1.9)
cortisol/CBG) was similar to the increase of total cortisol in the patients studied (mean 0.35 (SD 0.13) before stimulation and 0.80 (SD 0.15) after stimulation). Close correlation between free cortisol concentrations calculated according to Coolens et al. [7] and measured free cortisol after equilibrium dialysis was found (Figure. 1); by regression analysis a Pearson’s coefficient of correlation of 0.904 and the following equation was found: Free cortisol by equilibrium dialysis ⫽ 1.6 x calculated free cortisol ⫹ 2.2 nmol/L The cortisol/cortisone ratio increased from a mean of 5.90 (SD 2.9) before stimulation to 19.0 (SD 9.2) after stimulation. Mean total cortisone was 52.5 nmol/L (SD 14.3) before stimulation and 41.8 nmol/L (SD 13.7 nmol/L) after stimulation, mean free cortisone was 19.0 nmol/L (SD 5.2 nmol/L) before stimulation and 15.2 nmol/L (SD 5.0 nmol/L) after stimulation representing 36% of total cortisone at both sampling times. The mean ratio of free cortisol to free cortisone was 2.1 (SD 1.1) before stimulation and 8.1 (SD 3.9) after stimulation. In the two analytical series an inter-assay coefficient of variation of 2.9% for total cortisol and of 5.8% for free cortisol after equilibrium dialysis was found for the human serum pool (n ⫽ 6; total cortisol, 441 nmol/L; free cortisol, 29 nmol/L); the signal-to-noise ratio of the specific cortisol signal in the dialyzed samples was between 19:1 and 21:1. All quality control samples were found within the acceptable ranges.
4. Discussion In our study we were able to demonstrate that following adrenocortical stimulation by the ACTH analog Synacthen, free serum cortisol concentrations show a substantially more pronounced relative increase compared to total serum cortisol concentrations with a significant increase of the free-to-bound ratio of cortisol. These data suggest that measurement of bioactive free cortisol may reflect biologically more appropriately the glucocorticoid activation during neuroendocrine stimulation as compared to measurement of
542
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543
total cortisol. The nonlinear behavior of free serum cortisol concentrations in situations of acutely increased total serum cortisol and constant corticosteroid-binding globulin (CBG) concentrations is predicted by mathematical models that take into account the binding characteristics of CBG and albumin [5, 6]. These calculations, however, are most complex and require commercially not available software for numerical procedures. The calculation method described by Coolens et al. [7] is substantially simplified in that for example albumin concentrations and the concentrations of cortisone are not incorporated, however, this method is easily applicable employing standard software. We found a close correlation between free cortisol concentrations calculated in the latter way and directly measured concentrations, however, a slope of 1.6 was found with higher concentrations obtained by direct measurement; this proportional bias may arise from the simplifications made in this mathematical approach. Coolens et al. had as well compared calculated with directly measured free cortisol concentrations applying ultrafiltration and immunoassay but no relevant bias was found. This discrepancy –the apparent good agreement between calculation and measurement of free cortisol found by Coolens et al. but not in our investigation-might be due to methodological differences in the measurement of free serum cortisol. A rapid ultrafiltration method as used by Coolens et al. is - due to protein concentration effects on the filtration membrane - potentially more prone to artifacts than is equilibrium dialysis over an extended time period. A simple free cortisol index (total cortisol/CBG) as proposed previously [11,12] in contrast to the albeit simplified calculation of free serum cortisol from total serum cortisol and CBG (according to Coolens et al.) is obviously not able to display the over-proportional increase of free cortisol compared to total cortisol after neuroendocrine stimulation. The inexpensive equilibrium dialysis method applied in our study to quantify free serum cortisol proved convenient with good reproducibility (inter-assay coefficient of variation of 5.5%) and a favorable chromatographic signal-tonoise ratio in a normal serum pool; applicability of equilibrium dialysis to the assessment of free cortisol has been shown previously [13]. The simple dialysis method described here in combination with the highly specific and sensitive liquid-chromatographic tandem-mass spectrometric assay for the determination of cortisol may be considered as a candidate reference method for measurement of free cortisol. The method may be applicable in the development of direct immunoassays to quantitate free serum cortisol, by analogy to the measurement of free thyroid hormones. Such assays may be particularly useful for the monitoring of disease related hypercortisolism and neuroendocrine stimulation tests in severely ill patients [14,15] especially since critical illness is typically associated with a decline of CBG and albumin concentrations [16 –18]. We found the concentrations of CBG not to be affected by acute stimulation with ACTH; it could have been spec-
ulated that acute short term neuroendocrine stimulation could lead to cleavage and depletion of CBG as it is found in inflammatory states [16 –19]; in rats indeed a decline of CBG was found after stimulation with ACTH [20]. We report—to our knowledge—for the first time directly measured concentrations of free cortisone in men. Our results confirm that cortisone is bound to serum proteins to a lower extent compared to cortisol [6]. The ratio of free cortisol to free cortisone in consequence differs from the ratio of total cortisol to total cortisone. We have previously shown that acute stimulation with exogenous ACTH induces a systemic shift from inactive cortisone to active cortisol, suggestive of an increased activity of 11-hydroxysteroid dehydrogenase type 1 under this condition [9]; these results have been confirmed by the present study. The main results of the latter and the present investigation are in line in that both, the shift of the systemic cortisol-to-cortisone ratio and the nonlinear, over-proportional increase of free bioactive cortisol with a substantial rise in the free-to-bound ratio lead to an enhanced availability of cortisol in acute neuroendocrine stimulation in addition to the mere increase in the adrenocortical cortisol synthesis. References [1] Heyns W, Coolens JL. Physiology of corticosteroid-binding globulin in humans. Ann New York Acad Sci 1988;538:122–9. [2] Hammond GL, Smith CL, Underhill DA. Molecular studies of corticosteroid binding globulin structure, biosynthesis and function. J Steroid Biochem Mol Biol 1991;40:755– 62. [3] Rosner W. Plasma steroid-binding globulins. Endocrinol Metab Clin North Am 1991;20:697–720. [4] Brien TG. Pathophysiology of free cortisol in plasma. Ann New York Acad Sci 1988;538:130 – 6. [5] Ekins R. Measurement of free hormones in blood. Endocr Rev 1990; 11:5– 46. [6] Dunn JF, Nisula BC, Rodbard D. Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma. J Clin Endocrinol Metab 1981;53:58 – 68. [7] Coolens JL, Van Baelen H, Heyns W. Clinical use of unbound plasma cortisol as calculated from total cortisol and corticosteroid-binding globulin. J Steroid Biochem 1987;26:197–202. [8] Vogeser M, Briegel J, Jacob K. Determination of serum cortisol by isotope-dilution liquid-chromatography electrospray ionization tandem mass spectrometry with on-line extraction. Clin Chem Lab Med 2001;39:944 –7. [9] Vogeser M, Zachoval R, Jacob K. Serum cortisol/cortisone ratio after Synacthen stimulation. Clin Biochem 2001;34:421–5. [10] Reinard T, Jacobsen HJ. An inexpensive small volume equilibrium dialysis system for protein-ligand binding assays. Anal Biochem 1989;176:157– 60. [11] Bonte HA, van den Hoven R, van der Sluijs Veer G, Vermes I. The use of free cortisol index for laboratory assessment of pituitaryadrenal function. Clin Chem Lab Med 1999;37:127–32. [12] Beisenhuizen A, Thijs LG, Vermes I. Patterns of corticosteroidbinding globulin and the free cortisol index during septic shock and multitrauma. Intensive Care Med 2001;27:1584 –91. [13] Clerico A, Del Chicca MG, Ferdeghini M, Ghione S, Materazzi F. Progressively elevated levels of biologically active (free) cortisol
M. Vogeser et al. / Clinical Biochemistry 35 (2002) 539 –543 during pregnancy by a direct radioimmunological assay of diffusible cortisol in an equilibrium dialysis system. J Endocrinol Invest 1980; 3:185–7. [14] Briegel J, Kilger E, Schelling G. Stress doses of hydrocortisone in septic shock: Beyond the hemodynemic effects. Vincent JL, editor. Yearbook of Intensive Care and Emergency Medicine. Berlin, Germany: Springer, 1999. 189 –198. [15] Annane D, Sebille V, Troche G, Raphael JC, Gajdos P, Bellissant E. A 3-level prognostic classification in septic shock based on cortisol levels and cortisol response to corticotropin. JAMA 2000;283:1038 – 45. [16] Garrel DR. Corticosteroid-binding globulin during inflammation and burn injury: nutritional modulation and clinical implication. Horm Res 1996;45:245–51.
543
[17] Bernier J, Jobin N, Emptoz-Bonneton A. Decreased corticosteroidbinding globulin in burn patients: relationship with interleukin-6, and fat in nutritional support. Crit Care Med 1998;26:452– 60. [18] Vogeser M, Felbinger TW, Kilger E, Ro¨ ll W, Fraunberger P, Jacob K. Corticosteroid-binding globulin and free cortisol in the early postoperative period after cardiac surgery. Clin Biochem 1999;32: 213– 6. [19] Pemberton PA, Stein PE, Pepys MB, Potter JM, Carrell RW. Hormone binding globulins undergo serpin conformational change in inflammation. Nature 1988;336:257– 8. [20] Armario A, Giral M, Marti O, et al. The effect of acute, and chronic ACTH adminstration on pituitary-adrenal response to acute immobilization stress. Relationship to changes in corticosteroid-binding globulin. Endocr Res 1994;20:139 – 49.