MSc Chemistry
Analytical science track
Master Thesis
Enhanced method
development for the
quantification of
endogenous steroids
by
M.J.J.
Rijks
10726136
48 EC
September 2019 – April 2020
Daily supervisor:
Dr. ir. Mariëtte Ackermans
Endocrine laboratory
Department of clinical Chemistry
Supervisor/Examiner:
Examiner:
Dr. Rob Haselberg
Prof. Dr. Govert Somsen
Endocrine laboratory, Department of
Clinical Chemistry at Amsterdam
Abstract
Nowadays in endocrine diagnostics, LC-MS/MS is the method of choice for the determination of steroids. Since the significant increase of steroid samples that should be tested from different biological matrices the need for more high-throughput methods is increasing. Here, we develop a new method that separates 9 endogenous steroids from serum samples. These steroids include isobaric compounds: 21-deoxycortisol, corticosterone and 11-deoxycortisol (mw 347.25) and 17-hydroxyprogesterone and 11-deoxycorticosterone (mw 331.25). Also, a comparison with MOREPEAKS, former PIOTR, was made in order to find out whether the developed method could even be more efficient.
Compared to the current routine method the developed method showed a reduction in analysis time of 3.5 min while maintaining a resolution of at least 1.6 between the isobaric compounds, indicating a baseline separation. Moreover with the new method also 21-deoxycortisol could be measured which was not possible with the current routine method. Compared to the current routine method the newly developed method showed good agreement based on the measured concentrations in serum. Therefore we conclude that the new method can be implemented in daily practice.
Abbreviations
11-DOCL 11-deoxycortisol
17-OHP 17-hydroxyprogesterone
21-DOCL 21-deoxycortisol
ACN Acetonitrile
AmAc Ammonium acetate
Andro Androstenedione
CCSN Corticosterone
CV Coefficient of variation
DBS Dried blood spot
DCCSN Deoxycorticosterone
GC Gas chromatography
HSS High strength silica
LC Liquid chromatography
MeOH Methanol
MRM Multi-reaction monitoring
MS/MS Tandem mass spectrometry
Table of content
1. Introduction ... ….3
2. Theorectical background 2.1 Steroid metabolism ...4
2.2 Different biological matrices………..………..……….……….5
2.3 Techniques for the quantification of steroids….………..……….……….5
2.4 Reverse phase liquid chromatography ...6
2.5 Positive Electrospray ionization………..….………..……….……….7
2.6 MOREPEAKS ……….….………..……….……….7
3. Chemicals and method 3.1 Chemical and reagents ……….8
3.2 Samples ………..9
3.3 Sample preparation ………9
3.4 UPLC-MS/MS instruments and conditions ……….9
3.5 Method development ..……….10
3.6 Statistical analysis……….….………..……….………..12
3.7 Analytical performance ……….….………..……….……….12
4. Results 4.1 Parameter tuning .………13
4.2 Optimization column length……….……….15
4.3 Optimization mobile phase and flowrate …..……….……….16
4.4 Optimization using MOREPEAKS ……….….………..……….……….…17
4.5 Method comparison ………20 5. Discussion ………..……….……….21 6. Conclusion……….……….………..21 References APPENDIX I APPENDIX II APPENDIX III
1. Introduction
Why is it so important that we continue to develop and improve healthcare executed in hospitals, despite the fact that many processes and systems used in hospitals have been endlessly tested and validated? First of all, a treatment should be as pleasant as possible for the patient: the patient’s quality of life is the first priority. Secondly, to further reduce the proceedings to be performed by health care providers and analysts. In a time where the amount of people who need medical support significantly increases so does the demand for more automated, more high-throughput and faster analysis. [1] Therefore, development and improvement of the current healthcare in hospitals is important in order to continue the provision of proper, high-quality healthcare. Making an early diagnosis of specific diseases is an example of such an improvement, as it benefits the patient’s welfare and quality of life to start the treatment as soon as possible. Moreover, it could further reduced the proceedings to be performed by health care providers and analysts.
Multiple congenital disorders like Phenylketonuria (PKU), Cystic Fibrosis (CF) and congenital Hypothyroidism (CH) are caused by problems in the steroid synthesis pathway.[2] The early diagnosis of these diseases is necessary to establish a good quality of life and personal treatment for those who suffer from one of these disorders. Moreover steroids are considered as biomarkers for several adrenal diseases.[3] Therefore, it is important to accurately measure the steroids involved. The process prior to the diagnosis of a disease is based on the preparation of biological samples. The collection of biological samples is an essential pre-analytical activity. However, there are different matrices from which steroids can be measured. These biological matrices for steroids have their own purposes. However, it has not been yet assessed whether the analysis can be accelerated and more sensitive when the determination methods of steroids from different biological matrices are combined. For this reason, the possible improvements described in the following thesis focus on developing a highly sensitive, fast and comprehensive LC-MS/MS method that quantifies 9 endogenous steroid from different biological matrices in order to reduce the workload and save time, which could be beneficial for the patient’s welfare.
Figure 1. Metabolism of 9 (bold) endogenous steroids with Cholesterol as their precursor. Also, shown are the regulatory enzymes which are involved in the synthesis.[12]
2. Theoretical background
2.1 Steroid metabolism
Research in recent years has found that endogenous steroids are synthesized in the adrenal cortex, the gonads and the placenta.[4,5] More accurately they are synthesized in the mitochondria and smooth endoplasmatic reticulum. Steroid hormones are key regulators in metabolic pathways and therefore considered as potential biomarkers in several adrenal disorders. Also, these steroids can operate as marker for doping control screening.[6,7] About 30 steroids are produced in the adrenal cortex which can be divided into three different classes: mineralocorticoids, glucocorticoids and androgens. [8] Although steroids may belong to more than one of these classes, this work focusses mostly on glucocorticoids and androgens. For example, cortisol (glucocorticoid), androstenedione and testosterone (androgens).
All the steroids hormones used here are derivatives from cholesterol. Main source of cholesterol is provided in the form of low-density lipoprotein (LDL) cholesterol. [9] Therefore, the backbone structure of the steroid hormones are chemically similar to cyclopentanophenanthrene 4-ring structure as shown in Figure 1. Due to this backbone the glucocorticoids have lipophilic characteristics. In an attempt to cover as much as possible the biosynthesis metabolism of steroids the bold hormones as shown in Figure 1 are of interest. The steroid panel used consists of; cortisone, cortisol, 21-desoxycortisol, 11-21-desoxycortisol, corticosterone, 11-desoxycorticosteron, androstenedione, testosterone and 17-hydroxyprogesterone. The key players in the synthesis of steroids are the regulatory enzymes including i.e. 11β-hydroxylase and 21α-hydroxylase.[10] All of the diseases mentioned in the introduction are correlated to fluctuating levels of these steroids due to malfunctioning of (one of) these key regulatory enzymes. Therefore, determination of the concentration of these steroids and their precursors can aid in the diagnosis of one of these steroid depending diseases.[11] Regulation of the circulating steroids is likely to be more complex than described here. However, it is beyond the scope of this thesis to go into more detail.
2.2 Different biological matrices
Steroids can be measured from different biological matrices. This work focusses on serum samples while in hospitals blood sampling collection is mostly done by venous blood sampling and therefore most samples are serum samples.[13] An alternative matrix for the determination of blood steroid levels are dried blood spot (DBS) samples. This is especially used for newborns in the first week after birth, where all newborns will undergo a heel prick and will be screened for several congenital diseases, including congenital adrenal hyperplasia (CAH). In this case DBS sampling is used instead of serum sampling as this is less invasive, the sample is more stable and transportation of the sample to the laboratory is much easier. The sensitivity of this CAH neonatal screening can be improved by a second-tier steroid profile testing using an quantitative LC-MS/MS method.
2.3 Techniques for the quantification of steroids
Accurate and fast analysis of the different steroids plays an important role in the biochemical diagnosis of steroid related diseases. Over the years several techniques have been proposed for their quantification.
In 1959, Yalow and Berson developed the first technique that was able to determine low concentrations of hormones by radioimmunoassay (RIA).[14] Using different antibodies these type of assays were applicable to a wide range of steroids and other analytes of clinical importance. [15]. However, the need for radioactive reagents limited the applications of the RIAs. After the development of a new generation immunoassays (IA) chemiluminescent and fluorescent labels as well as monoclonal antibodies were used instead[16]. These were rapidly implemented as commercial kits. Unfortunately, these assays still have a lot of challenges and limitations which have been well documented in literature [17,18,19,20]. First of all, the dynamic range of the assay can be a problem as concentrations can range over 3 orders of magnitude, depending on factors like gender, age, and the disease state. Secondly and more importantly due to the similarities in the structures of the endogenous hormones the IA may not be able to distinguish between the analyte and another structure related steroid. [21] This cross-reactivity can cause overestimation of the analyte concentrations.
The main issue of cross-reactivity can be overcome by using chromatography-mass spectrometry methods. In first instance GC-MS was used. Since the analysis and ion formation occur under high vacuum (10−5 to 10−6 Torr) GC was initially used for the separation of steroids prior to MS. In addition, the ease of coupling these orthogonal techniques with respect to the vacuum considerations resulted in the fact that GC-MS systems have been around for many years in analytical chemistry, but not without limitations regarding steroids. [22] Steroids are not naturally volatile compounds nor thermally stable. [23] These applications resulted in intensive sample preparations including several steps as purification and derivatization. First, Solid phase extraction (SPE) is performed to concentrate the steroids and removal of salts. Next, the ketones and the free hydroxyl groups need to be chemically modified in order to make the compounds more volatile. The main reason for selecting a derivatization agent is the enhanced stability and retention time of the steroids. Silylating reagents, most commonly used, are able to convert the hydroxyl groups into trimethylsilyl ethers. Also, the ketone groups were modified into silyl enol ethers. [22] The derivatives have an increased volatility in comparison with the parent steroids and are stable at very high temperatures. Moreover these methods showed long analysis times and therefore GC-MS/MS has limited throughput.[24] Therefore, with the coming of robust and affordable LC-MS systems in routine settings, readily the change was made to LC-MS/MS methods.
LC-MS/MS, as it is well known for its accuracy, particularly with regard to low concentrations. Also, due to its high analytical specificity the mass spectrometer is capable to discriminate the related structures and overcome the possibility of cross-reactivity. Despite the requirement of expertise and large setup costs, it is applicable to a wide dynamic range and it has the ability to perform high-throughput without dependence on radioactivity agents or antibodies. [25] So, since LC-MS does not require a derivatization step in combination with adaptable sample preparations liquid chromatography-tandem mass spectrometry (LC-MS/MS) has become the state-of-the art method for the analysis of steroids.
Although mass spectrometry is a very robust and sensitive way to analyze steroids it is not possible to do it without the chromatography. [26] The chromatography is essential for adequate analysis of this steroid panel since it contains some isobaric compounds. This disallows the mass spectrometer to separate the compounds with an identical nominal molecular weight. For example, 21-deoxycortisol, corticosterone and 11-deoxycortisol have the same weight. Also, 11-deoxycorticosterone and 17-hydroxyprogesterone have the same weight (m/z= 374.4 and 331.4, respectively). Baseline separation of these isobaric compounds therefore is necessary and the separation of the critical pairs therefore is a crucial subject of the method development.
2.4 Reverse phase liquid chromatography
Since, LC-MS/MS could bypass a lot of the disadvantages of the other techniques it is now the most used technique for determination of steroids in clinical use. Due to the lipophilic characteristics and the hydrophobic nature of the unconjugated steroids makes that reverse phase is more effective for the separation than normal phase.[25] Also, less organic solvent is necessary when doing reverse phase. In combination with the RPLC, silica or polymer particles are bound to the hydrocarbon backbone causing enough interaction with the hydrophobic part of the steroid for good separation. A triple quadrupole is used since it most robust and sensitive instrumentation for high throughput assays. [22] In Figure 2, a schematic overview of triple quadrupole RPLC tandem MS set-up is displayed. After the sample preparation, the steroids are injected into the analytical column. Injecting the steroids at high polarity the steroids focus at the beginning of the column. Using a gradient with increasing organic modifier, the first eluting peaks will relate to the more polar steroids.
Figure 2. A schematic overview of the LC-MS/MS technique with electrospray ionization. After the steroids have been separated based on their affinity to the stationary phase, the mobile phase enters the MS through a needle were it is evaporated under a stream of nitrogen and high temperatures. Then under vacuum, the Q1 select ions with a specific m/z ratio. These ions were fragmented in the collision cell before they enter Q3 where again only compounds with a certain m/z are send through and arrive at the detector.
2.5 Positive electrospray ionization
To introduce the chemical analytes into the MS a flowrate of 0.6 mL/min need to be forced into a vacuum of 10−6 Torr. To achieve this the effluent must be dispersed into tiny droplets. One of the ionization method to do this is electrospray ionization (ESI-MS). Previous studies have described the fragmentation patterns of steroids.[28,29] As consequence, the screening for steroids has been carried out with positive electrospray ionization due to the fragmentation behavior of the steroids, Figure 3. Negative ionization for testosterone is only possible if certain adducts are formed by introducing a derivatization step in the sample preparation. Therefore, in this thesis positive electrospray is used.
After ionization, only the compounds with a certain mass/charge ratio (precursor ion) are selected in the first mass analyzer (Q1) and send to a collision cell. Here the selected components are fragmented and collected by the second mass analyzer (Q3). Using the MRM scan mode, which indicates that Q1 and Q3 consist both of selected masses. The ions selected by the first mass analyzer are only detected if the daughter ions have the selected mass that is selected by the second mass analyzer. By putting both analyzer on fixed masses, it allows to focus on the precursor and fragment ions over longer times, increasing the sensitivity. This enables simultaneous determination of multiple steroids at very low concentrations in a single run. In this thesis the MS/MS provided increased specificity by collision-induced fragmentation of the analyte into a molecular fingerprint that could be identified by monitoring two or more fragments. [30] Measuring at least two fragments of every steroid results in a decreased risk of not detecting isobaric interferences. However, in this case some of the fragments of the isobaric compounds also have the same mass transitions products.
Figure 3. A fragmentation pathway of testosterone suggested by Nielen et al. [28] Testosterone is ionized into fragments of m/z 97 and m/z 109, which are the most abundant.
2.6 MOREPEAKS
MOREPEAKS, former PIOTR, was introduced by Pirok et al in 2016.[31] It is a virtual program developed to predict retention models, It is originally developed for the optimization of LC x LC separations. Both dimensions will be treated as one separate dimension for the optimization. Therefore, it will also be applicable for one dimension separations.
For the retention models prediction it uses an algorithm that is based upon some assumptions. First, the retention mechanisms in isocratic reverse phase follow the assumption of the linearity between the logarithm of the retention factor (k) and the volume fraction of organic modifier (ɸ), Equation 1. [32] If S, k0 and are known for a solute, for any gradient the retention factor of the analytes can be
predicted. The S and ln k0 can be obtained from some experimental input. Two different gradients with
Equation 1.
ln 𝑘 = ln 𝑘0− Sɸ
Whereas For linear gradients in RPLC, the retention time can be derived from Equation 2.[31] Equation 2. t𝑅,𝑔𝑟𝑎𝑑𝑖𝑒𝑛𝑡 = 1 𝑆𝐵ln
{
1 + 𝑆𝐵𝑘(
𝐴) [
𝑡0− 𝑡𝑖𝑛𝑖𝑡+ 𝑡𝐷 𝑘(
𝐴)
]}
+ 𝑡0+ 𝑡𝐷+ 𝑡𝑖𝑛𝑖𝑡Where A represents a fraction of organic modifier at the beginning of the gradient, B the slope of the gradient (change in organic modifier), S the change in retention factor with increasing mobile phase strength, t0 is the dead time of the column, tD the systems dwell time, tinit the time before the gradient
has started. In addition, by using a van-Deemter model to generate band broadening it allows the prediction of the chromatogram.
Theoretically, given the right input, MOREPEAKS should be able to calculate the optimal gradient with shortest run time for the separation for the steroid panel. As a part of this thesis project we gathered the necessary data and calculated and tested the solution MOREPEAKS provided. Besides, in collaboration with the CAST group this thesis provided data that could contribute to further development of the MOREPEAKS program.
3. Chemicals and methods
3.1 Chemical and reagents
Cortisone (Sigma Ceriliant art.nr. C-106), cortisol (Sigma Ceriliant art. Nr. C-130), 21-deoxycortisol (Sigma art. Nr. P-9521), corticosterone (Sigma Ceriliant art. Nr. C-117), 11-deoxycortisol (Sigma Ceriliant art. Nr. C-061), 11-deoxycorticosterone (Sigma Ceriliant art. Nr. C-105), androstenedione (Sigma Ceriliant art. Nr. A-75), testosterone (Sigma Ceriliant art. Nr. T-037), 17-hydroxyprogesteron (Sigma Ceriliant art. Nr. H-085), metanphrine (Sigma Ceriliant art. Nr. C-110). D7-cortisone (Aldrich art. Nr. 705586), D5-11-deoxycortisol (Aldrich art. Nr. 710784) were all purchased at Ceriliant/Sigma-Aldrich, Saint Louis, USA. D4-cortisol (DLM-2218), D8-corticosteron (DLM-7347), D8-17-hydroxyprogesteron (DLM-6598) were all purchased at Cambridge Isotope Laboratories, Tewksbury, USA. D4-21-deoxycortisol (Isoscience cat. Nr. 15063), D7-androstenedione (CDN isotopes, cat. Nr. D-5305) and D3-testosterone (CDN isotopes, cat. Nr. D-3793) were purchased at CDN isotopes, Point-Claire, Canada. LCMS grade Acetonitrile (art. Nr. 01207801) and LCMS grade Methanol (art. Nr. 13687801) were purchased at Biosolve, Valkenswaard, NL. Formic acid 98 -100% (art. Nr. 1.00264) was puschased at Merck, Kenilworth, USA) and , Ammonium acetate (art. Nr. 49638) at Fluka, Schwerte, Germany.
3.2 Samples
In order to test the method a system suitability sample was used. This sample contained all steroids (1 -10 nM) in water/methanol/FA (60/40/0.1) mixture.
To check matrix effects and to compare the newly developed method to the routinely used method, patient samples have been used. These samples were previously tested for clinical diagnostic purposes at the Endocrine laboratory Amsterdam UMC. All samples were anonymized and handled identically according to the sample preparation.
3.3 Sample preparation
Protein precipitation (PPT) is used for the preparation of serum sample. 25 µL of serum sample were placed into a porvair filter plate (purchased from Screening Devices , Amersfoort, NL) before adding Internal standard (IS) mix in ACN. The mix contained al deuterium labelled steroids. Subsequently, the samples were filtrated by applying pressure with a Waters Positive Pressure-96 Processor (1 minute, 10 psi), dried by nitrogen at 50 degrees Celsius and resolved again in 60 µL ACN/H2O/FA 20/80/0.1.
3.4 UPLC-MS/MS instruments and conditions
For the experiments performed in this thesis several different systems were used. IN order not to intefere with the daily routine of the lab the method development was performed on the research system of the lab. This is a Thermo Scientific Duo UHPLC (San Jose, United States) system coupled to a Thermo Scientific Quantiva TQS triple-quadrupole mass spectrometer equipped with an Ion Max NG ESI ion source. In addition, the Quantiva mass spectrometer was equipped with an additional Chemyx Fusion 100T syringe pump for direct MS infusion and a Rheodyne MX Series II modular inject valve. The Thermo Scientific system was controlled by ChromeleonTM software(v7.2.10).
Routine analyses of the steroids in the Amsterdam UMC are performed on Waters systems. The final method of the project was therefore transferred to the Waters systems to validate the new method against the current routine method. The instrument used for daily routine was a Waters I-Class/Xevo TQ-S micro IVD tandem quadrupole mass spectrometer, the back-up system consists of a Waters Acquity UPLC coupled to a Xevo TQ-S triple-quadrupole mass spectrometer (Milford, MA, United states). Both systems are equipped with an ESI ion source. Both Waters systems were controlled by MassLynxTM IVD Software (v4.1) and the data was quantified and processed using the TargetLynxTM
Application Manager.
Both Waters systems were already tuned for the quantification of the selected steroids as they are in use for the routine analysis for clinical diagnosis. MRM ion transitions, as acquisition mode, were used for the identification, quantification and validation. Positive electrospray was used with the following settings of the mass spectrometer: capillary voltage 1.00 kV, source temperature 150°C, desolvation temperature 650°C, desolvation gas flow (N2) 1200 L/hour and cone gas flow 150 L/hour. Mass
transitions and settings of cone voltages and collision energies for the steroids are mentioned in Table 1. In this table also the m/z value of metanephrine is displayed. Although metanephrine is not a steroid, it is mentioned as it is used as a T0-marker to obtain data for MOREPEAKS.
3.5 LC method development 3.5.1 Current routine LC method
The current routine analysis, which is used for the quantification of steroids, is a 10 minutes gradient and start with 20% organic modifier. The first four minutes is a linear gradient after which it continues as a step gradient till 100% organic. The injection volume is 10 µL. The aqueous phase consist of water and 0.1% FA, whereas the organic modifier is acetonitrile with 0.1% FA. For the separation a 10cm Waters acquity UPLC HSS T3 column was used at a temperature of 40 degrees Celsius. In Table 3, an overview of the routine method is provided.
Table 1. Overview of m/z values of all steroid Quantifiers and qualifiers used in this experiment. Metanephrine is added as T0-marker. Also, the cone voltage and the collision energy is displayed.
Compound Quantifier [m/z] Qualifier [m/z] Cone (V) Coll. En. Precursor ion Product ion Precursor
ion
Product ion (eV)
Metanephrine 179.90 148.00 x x 36 18 Cortisone 361.10 163.10 361.10 145.00 40 28 Cortisol 363.10 121.00 363.10 309.10 44 22 21-deoxycortisol 347.25 121.10 374.25 91.10 42 22 Corticosterone 347.25 121.10 347.25 311.10 42 22 11-deoxycortisol 347.25 97.10 347.25 109.10 46 20 11-deoxycorticosterone 331.25 97.14 331.25 109.15 42 24 Androstenedione 287.20 97.15 287.20 109.10 36 18 Testosterone 289.20 97.10 289.20 109.10 38 20 17-hydroxyprogesterone 331.25 97.15 331.20 109.10 36 22 Internal standards [m/z]
Precursor ion Product ion
D7 -Cortisone 369.25 169.15 D4 -Cortisol 367.20 121.00 D4 -21-deoxycortisol 351.40 121.60 D8 -Corticosterone 355.47 125.13 D5 -11-deoxycortisol 352.40 100.03 D7 -Androstenedione 294.20 100.15 D3 -Testosterone 292.20 97.10 D8 -17-hydroxyprogesterone 339.20 100.15
Table 3. Overview of the routine analysis. Here, analysis time, mobile phase composition, flow rate, column and temperature are mentioned
3.5.2 Optimization of column length
Another essential requisite for the final method is the chromatographic separation of the critical pairs prior to the mass spectrometric detection. Initially, for the routine analysis a Waters Acquity UPLC HSS T3 column with 100 x 2.1 mm dimensions and a particle size of 1.8 µm was used. First, in order to try to reduce the analysis time, a shorter column was used. Instead of a 10 cm column, the analysis was now performed on a Waters Acquity UPLC HSS T3 column with 50 x 2.1 mm dimensions and a particle size of 1.8 µm was used with a VanGuard Pre-column HSS T3 1.8µm 2.1 x 5 mm. The shorter column should theoretically save half of the analysis time. However, a shorter column could possibly reduce the resolution of the separation. For the development of the new method, the routine method is considered the ‘golden standard’ and taken along for comparison reasons.
3.5.3 Optimization of mobile phase
The order in which the steroids elute is partially predictable from their structure in combination with certain solvent mixtures. The difference in solvent mixtures can cause a change elution characteristics. The routine analysis for serum samples contains a mobile phase of A: H20/0.1% FA (v/v) and an organic phase of B: ACN/0.1% FA (v/v). Acetonitrile is a polar aprotic solvent which lack acidic hydrogen and therefore not a hydrogen donor. In the new developed method, the mobile phase consisted of A: H20/2mM AmAc/0.1% FA (v/v) and B: MeOH/2mM AmAc/0.1% FA (v/v). Now methanol is polar protic
solvent which have an acidic hydrogen and is able to act as a hydrogen donor. The change in organic modifier had some consequences for the development. In previous studies the use of methanol not only increased the elution of the steroids and therefore the analysis time. But also methanol is due to its acidic hydrogen able to facilitate in positive electrospray ionization. Therefore, there will be more ionization in comparison with acetonitrile.
3.5.4 Optimization of flow rate
However, changing the organic modifier from ACN to methanol also increased the viscosity of the mobile phase. Since the viscosity of methanol is higher than the viscosity of acetonitrile the system backpressure will increase accordingly. Additional requirements were made in order to reduce backpressure and compensate for the increased viscosity. By lowering the flowrate from 0.6 mL/min to 0.4 mL/min the backpressure will not exceed its limitations.
Routine analysis
Time (min) Mobile phase composition Flowrate (ml/min)
A (%) B (%) 0 80 20 0,6 4 70.4 29.6 0,6 5.5 55 45 0,6 7.8 55 45 0,6 7.85 0 100 0,6 8.85 0 100 0,6 9 80 20 0,6 10 80 20 0,6
A: water + 0.1% formic acid B: Acetonitrile + 0.1% formic acid
3.5.5 Optimization using MOREPEAKS
As input data for the program the dwell volume was calculated by dwell time times flowrate resulting in 123 µL. Consequently, the expected T0 was calculated by following equation;
T0 =
Vcol,mφ F
=
ε . 14 . 𝜋 . 𝑑𝑐2 . 𝐿 F
Where Vcol, mφ represents the column volume also calculated by the porosity (ε), squared column diameter (dc2) and the column length (L). This resulted in a T0 of 1.23 ±0.01 min at a 0.1 mL/min
flowrate. The flowrate was kept as low as possible to be more precise. Since, the experiment was carried out at a flowrate of 0.6 the expected time was 0.2 minutes. In order to check this, an experiment with a T0-marker in form of metanephrine was performed. This compound is a very polar
compound and normally measured with HILIC so it should have no retention in reverse phase chromatography under these conditions.
Next, two series were performed in order to obtain independent dataset. One series with a gradient from 0 to 100% organic modifier in 10 minutes and the other series with the same gradient in 30 minutes to achieve a factor 3 in gradient time, which is necessary for the MOREPEAKS program. For this experiment SST of steroids was used. Since the t0 is required as input, the SST of the metanephrine
was 1:1 diluted with the SST of the steroids. 3.6 Statistical analysis
For the method comparison Passing-Bablok regression analysis was performed. The Pearson correlation coefficient was calculated to assess the similarities between the methods. Also, a Bland-Altman plot was plotted to analyze the agreement between the methods. This allows for identification of possible systematic errors and possible outliers. Furthermore, a clear overview is given in which range (high or low concentrations) differences are present. Both Passing-Bablok and Bland Altman plot were obtained from calculations with MedCalc 18.5 and Microsoft Excel 2010.
3.7 Analytical Performance 3.7.1 Linearity
The linearity of the method was checked on the basis of the calibration curves which consisted of six varying concentrations covering the biological dynamic range for each steroid. Each calibration point was measured in duplicate. These curves were prepared by adding cerriliant of the steroids to steroid free plasma. Linearity was determined by weighed linear regression model (W= 1/X) with error margins of 5%.
3.7.2 Selectivity and carry-over
The selectivity was determined by only injecting blank samples which should be free of any spectral interferences at the retention times of the steroids of interest. Interference of the carry-over effect has been evaluated by injecting a blank sample right after the highest calibration standard. This has been performed in triplicate to assess whether the response of the blank samples were less than 20% of the lowest calibration concentration at the allocated retention times of the steroids.
3.7.3 Matrix effect
The assessment of the matrix effect and ion suppression was carried out according to the Matusewski experiment. [34] By using different internal standards before and after the extraction there could be determined whether the matrix effect and/or the ion suppression affected the results. Normally, deuterium labelled steroids were used as internal standards. But to determine whether there was a matrix effect and/or ion suppression also 13C labelled steroids were used. Using 13C labelled instead
has some practical advantages over the deuterium labelled ones such as having the exact same retention time as the natural occurring steroid. The 13C labelled steroids were only available for
cortisol, testosterone and androstenedione.
3.7.4 Verification of the method
Once the new method is developed the method has to be verified. Since the sample pretreatment has not changed not a full analytical validation is performed at this stage. Of crucial importance is that the newly developed method results in the same measured concentrations for patient samples. This is important because the routine steroid analysis is standardized against (inter)national standards and reference values are determined. To prove that the method can be transferred several patients’ samples were tested. The samples were selected to represent as much as possible the measurable range of the steroids, were prepared as described above and measured sequentially first with the current routine method and then with the newly developed method. Standard curve and controls are also measured. For both methods the concentrations of the steroids were calculated using Masslynx and a Passing Bablok regression was performed plotting the concentrations measured with the newly developed method against those measured using the current routine method. The result was acceptable if the slope of this Passing Bablok regression was between 0.95 and 1.05.
4. Results & Discussion
4.1 Parameter tuning
Until now the steroids were never measured on the Qauntiva system. Therefore, the mass spectrometer needed to be tuned for each and every steroid for optimal detection. Pure steroids were divided into three groups whereby the critical pairs were divided over the groups to avoid interference. Group 1: Testosterone, 21-deoxycortisol and 17 hydroxyprogesterone. Group 2: Androstenedione, cortisol and 11-deoxycortisol. Group 3: 11-deoxycorticosterone, corticosterone, cortisone.
The parameters that were tuned were: Collision energy, Collision-induced dissociation gas, Spray voltage and RF lens. The first parameter that had to be assessed was the collision energy. This parameter was different for each steroid and therefore a range between 16 – 28 KV with 2 KV steps was chosen. This range was based on the collision energy the steroids had on the Water system. As an example the results for testosterone are shown in Figure 4. For each steroid the optimum collision energy was set at the value that gave the highest peak area.
0 50000 100000 150000 200000 16 KV 18 KV 20 KV 22 KV 24 KV 26 KV 28 KV Are a Collision Energy in kV
Collision energy
Testosterone 21-deoxycortisol 17-OHP
0 50000 100000 150000 200000 0,5 1 1,5 2 2,5 3 3,5 4 Are a ClD gas
Collision-induced dissociation gas
Testosterone 21-deoxycortisol 17-OHP
Figure 4. The results of the collision energy tuning experiment. Area is plotted against the collision energy. Also, a table is provided with the optimal values for each individual steroids.
Figure 5. The results of the collision-induced dissociation gas. Area is plotted against the values of the CID gas.
The next parameter was the collision-induced dissociation gas (CID). This parameter only had preprogrammed values between 0.5 and 4.0 with steps of 0.5. Also, one value had to be chosen to cover all the steroids instead of an optimal value for each individual steroid like the collision energy. Therefore, a compromise should be made if necessary. As a result of this experiment the area were plotted against the values of CID gas, an example of which can be observed for testosterone in Figure 5. Taken in account all areas of the steroids there was easily decided that CID gas value should be 1.5.
Subsequently, spray voltage was assessed. This parameter also had preprogrammed values between 1.0 and 4.0 with steps of 0.5 kV. Again for this parameter only one value should cover all of the steroids and a compromise should be made if necessary. As a result of this experiment the area were plotted against the values of spray voltage, which can be observed in Figure 6. This result led to an optimal spray voltage of 1.00 kV. Collision energy (kV) Testosterone 22 21-deoxycortisol 18 17-hydroxyprogesterone 26 Corticosterone 26 11-deoxycorticosterone 24 Cortisone 28 Androstenedione 20 Cortisol 22 11-deoxycortisol 22
0 20000 40000 60000 80000 100000 120000 140000 1 1,5 2 2,5 3 3,5 Are a Spray voltage in kV
Spray Voltage
Testosterone 21-deoxycortisol 17-OHP
0 20000 40000 60000 80000 100000 120000 140000 55 65 85 105 115 Are a RF voltage in V
Rf Voltage
Testosterone 21-deoxycortisol 17-OHP
Figure 6. The results of the spray voltage tuning experiment. The area was plotted against the value of spray voltage in kV.
Figure 7. The results of the RF lens optimization. Area was plotted against the value of the RF in V.
Last parameter that had to be assessed was the radiofrequency value (RF). Since a wide range of values were possible initially the following RF values: 65, 85, 105 and 115 were selected. However, when discovered that a lower RF values resulted in higher areas the value of 55 was added. As can be observed in Figure X, the samples measured with a RF value of 55 resulted in the highest area. Therefore, 55 was selected for al steroids.
4.2 Optimization of column length
In Figure 8, a chromatogram is shown, as a result of a serum sample analyzed with the routine analysis. This was the starting point for the development of a new improved method. In order to decrease the analysis time the first step in the optimization process was to reduce column length by half. A SST mix of the steroids was now analyzed on a 5cm HSS T3 column with the same eluents and gradient as the routine analysis. As a result, the last eluting peak was at 5.94 minutes. This first step already led to a faster analysis time. However, the resolution of the first critical pairs, corticosterone and 11-deoxycortisol,decreased so that the baseline separation between the isobaric compounds was at risk.
In particular at the later eluting peaks of the second critical pair, 11-deoxycorticosteron and 17-hydroxyprogesteron, a lot of resolution was lost as displayed in Figure 9. Also, the resolution on the other critical pair corticosterone and 11-deoxycortisol was decreased. Note that in the SST of the steroid mix is yet without 21-deoxycortisol. Addition of this compound will increase the complexity of the first critical pair. Because of the much decrease in analysis time there was decided to continue the development with the shorter column.
Figure 8. Chromatogram as result of a SST mixture analysed by routine analysis. Eluents A: H20+0.1% FA and B: ACN+0.1% FA on a 10cm HSS T3 column. Note: 21-deoxycortisol is not measured in routine analysis.
Figure 9. Separation of the critical pairs. (A) Corticosterone and 11-deoxycortisol and the later eluting critical pair (B) 11-deoxycorticosterone and 17-hydroxyprogesterone. This result was obtained after analysing a serum SST on a 5 cm HSS T3 column.
4.3 Optimization of mobile phase and flowrate
Next, the mobile phase had to be optimized. Therefore, the mobile phase was changed from ACN to methanol and 2 mM ammonium acetate was added. The change in organic modifier did ensure for a change in retention order of several compounds. Cortisone and cortisol changed in retention order, likewise did testosterone and androstenedione. For the identification each single compound was injected separately. Although the new organic modifier was more suitable for these steroids it also increased the viscosity and therefore the systems backpressure. Therefore, the flowrate should be reduced to 0.4 mL/min in order to ensure the backpressure does not exceed its limitations. Due to the
reduce in flowrate to 0.4 mL/min, the baseline separation of the critical pairs was again at risk. To overcome this uncertainty there was decided to increase the temperature to 60 degrees Celsius. After increasing the temperature the following results were obtained, Figure 10.
Figure 10. Separation of the critical pairs analysed by the new method. (A) 21-deoxycortisol, corticosterone and 11-deoxycortisol, respectively. (B) the separation of 11-deoxycorticosterone and 17-hydroxyprogesterone.
Although the analysis time was significantly decreased, lowering the flowrate and increasing the temperature affected the separation of both critical pairs. Whereas the resolution (Rs) of the critical
pair (A) of corticosterone and 11-deoxycortisol in the routine analysis was 2.58 it costly decreased to 1.63 with the new analysis method. Since Rs > 1.5, this is still baseline separated and therefore accepted. Also, the second critical pair (B) is baseline separated and therefore a new method was developed which was able to separate both critical pairs.
Table 4. Overview of the gradient and conditions used for routine analysis and the new developed method
Method comparison
Routine analysis ‘new’ method
Time (min) Mobile phase composition Flowrate (ml/min) Time (min) Mobile phase composition Flowrate (ml/min)
A (%) B (%) A (%) B (%) 0 80 20 0,6 0 80 20 0,4 4 70.4 29.6 0,6 0 60 40 0,4 5.5 55 45 0,6 5 36 64 0,4 7.8 55 45 0,6 5.1 2 98 0,4 7.85 5 100 0,6 6 2 98 0,4 8.85 5 100 0,6 6 80 20 0,4 9 90 20 0,6 6.5 80 20 0.4 10 90 20 0,6
A: water + 0.1% formic acid B: Acetonitrile + 0.1% formic acid D: Acetonitrile
A: 2 mM ammonium acetate + 0.1% formic acid in water B: 2 mM ammonium acetate + 0.1% formic acid in methanol Waters Acquity UPLC HSS T3, 100 x 2,1 mm – 1,8 µm – 40°C Waters Acquity UPLC HSS T3, 50 x 2,1 mm – 1,8 µm – 60°C
Figure 11. A pareto optimality plot as output of the MOREPEAKS program. In this plot the resolution score is plotted against the time of the last eluting peak. The purple dot represent the gradient with the highest resolution score.
4.4 Optimization using MOREPEAKS
In order to find out if the MOREPEAKS program was able to predict an even better retention model the program required some experimental data input. In chapter 3.5 the dwell time was calculated as 0.2 minutes. In order to check this, a T0-marker in form of metanephrine was used. Injecting the system
suitability test (SST) of the metanephrine resulted in a peak at 0.26 min. A difference of 0.06 min with the calculated value was acceptable.
Next, two series were performed in order to obtain independent datasets. One series with a gradient from 0 to 100% organic modifier in 10 minutes and the other series with the same gradient in 30 minutes to achieve a factor 3 in gradient time, which is necessity for the MOREPEAKS program. The SST of the metanephrine was 1:1 diluted with the SST of the steroids.
Subsequently, the output of the program was a pareto optimality plot. In this plot the resolution score was plotted against the time of the last eluting peak, Figure 11. Since the critical pairs have to be separated the most appealing gradient was the gradient which offered the best resolution. This was a linear gradient of 10 minutes starting at 50% organic modifier till 80%. This experiment was already done with 2mM ammonium acetate in methanol, a flowrate of 0.4 mL/min and the 5cm HSS T3 column. This way a comparison could be made between the manual optimized gradient and the gradient provided by the program.
Since it was only managed to obtain a linear gradient from the MOREPEAKS program the separation was no better than the manual grafted gradient. In Figure 12, the separation of the critical pairs are shown according to the MOREPEAKS gradient. The separation of 11-deoxycorticosterone and 17-hydroxyprogesteron is clearly baseline separated. However, the other critical pair shows that 21-deoxycortisol is separated from the others but there is no separation between corticosterone and 11-deoxycortisol. Therefore, it was decided to continue the optimization process with the manual grafted gradient and leave the computer based gradient for further research. A possible reason why the program failed to separate corticosterone and 11-deoxycortisol is most likely due to the characteristics of the steroids. Under these condition they may have the same affinity to the stationary phase and there will be no change of separating them with a linear gradient. The program however does not
Figure 12. Separation of the critical pairs. (A,B) separation of the critical pairs with the manual grafted gradient. (C) separation of 21-DOCL, CCSN and 11-DOCL according to the MOREPEAKS gradient. (D) separation of 11-DCCSN and 17OHP according to the MOREPEAKS gradient.
guarantee that all peaks could be separated only provided the most likely retention model based on the experimental input.
4.5 Method comparison
In Table 4 a summary of the obtained values for the Passing Bablok regression of the measured concentrations for the patient samples are given. These were measured with the current routine method and subsequently with the newly developed method on the three LC-MS systems. The Passing- Bablock and the Bland- Altman plots can be found in the Appendix.
Table 5. Summary of the Passing- Bablock regression of each single steroid with a 95% confidence interval for each of the three systems.
Quantiva System
Steroid n Intercept Slope r (95% CI)
Testosterone 84 0.01 1.038 0.9996 (0.9994 – 0.9997) 17-OH Progesterone 43 -0.23 0.992 0.9994 (0.9990 – 0.9997) Androstenedione 84 -0.18 1.038 0.9979 (0.9967 – 0.9986) 11-Deoxycortisol 27 0.16 1.028 1.0000 (0.9999 – 1.0000) 11-Deoxycorticosterone 29 0.04 1.064 1.0000 (1.0000 – 1.0000) Corticosterone 83 0.16 0.915 0.9963 (0.9942 – 0.9976) Cortisone 84 -0.23 1.017 0.9915 (0.9869 – 0.9945) Cortisol 84 4.14 0.983 0.9942 (0.9910 – 0.9962)
Xevo TQ-S system.
Steroid n Intercept Slope r (95% CI)
Testosterone 79 0.00 0.778 0.9989 (0.9982 – 0.9993) 17-OH Progesterone 39 0.00 1.273 0.9998 (0.9997 – 0.9999) Androstenedione 79 0.03 0.808 0.9992 (0.9987 – 0.9995) 11-Deoxycortisol 24 0.15 1.299 0.9951 (0.9886 – 0.9979) 11-Deoxycorticosterone 15 0.15 0.981 0.4308 (-0.1045 – 0.7726) Corticosterone 79 0.03 1.025 0.9981 (0.9970 – 0.9988) Cortisone 79 0.00 1.017 0.9689 (0.9516 – 0.9800) Cortisol 79 -6.23 0.930 0.9974 (0.9960 – 0.9984)
Micro-I class system
Steroid n Intercept Slope r (95% CI)
Testosterone 71 -0.02 0.987 0.9998 (0.9997 – 0.9999) 17-OH Progesterone 21 -0.24 0.929 0.9958 (0.9896 – 0.9983) Androstenedione 72 0.01 0.963 0.9962 (0.9938 – 0.9976) 11-Deoxycortisol 23 -0.18 1.023 1.0000 (1.0000 – 1.0000) 11-Deoxycorticosterone 29 0.06 0.970 0.9996 (0.9991 – 0.9998) Corticosterone 72 -0.03 1.003 0.9996 (0.9994 – 0.9998) Cortisone 72 0.05 1.006 0.9939 (0.9902 – 0.9962) Cortisol 72 2.81 1.026 0.9991 (0.9985 - 0.9994)
Although the patient samples were carefully selected to cover the measurable range as can be seen for some steroids the number of measurable samples is lower. This is because some of the steroids are barely measurable in non-ill patients. This hold i.e for corticosterone, 11 –desoxycorticosterone 11-deoxycortisol and 17-hydroxyprogesterone. Moreover, the values that were detectable were often very low.
As shown in Table 4, the slopes of most of the steroids on the Quantiva system and de Micro-I class system are within the acceptable range. It can be seen that the steroids were the acceptable slope was not obtained are those were either the number is low or the concentrations measured are barely above the LloQ. Therefore, further verification is necessary for these compounds, where samples should be collected from patients, where higher values can be expected.
For the Xevo-TQS system very different results are obtained. Surprisingly there only two slopes were within the acceptable range. Since both on the Quantiva system and the Micro-I class system the current routine method and the newly developed method gave the same results the suspicion is that the fact that the Xevo system did not is due to an instrument feature and not due to one of the methods. One of the features that can explain this phenomenon could be that due to the conformation of the probe/source the ionization can be influenced by a matrix effect. Therefore, 13C labelled internal
standards were used to focus on the detection of matrix effect and ion suppression. 13C labelled
steroids were added after the sample preparation to focus on the ion suppression. After measuring the concentrations of cortisol, testosterone and androstenedione with both deuterium labelled as well as 13C labelled internal standards, it became clear that for the routine analysis it did not matter which
of the internal standards were used. The Areas of the components differed in the same way as both internal standards. However, these differences were so small that it was not likely to conclude that ion
suppression and the matrix effect caused the big difference between the results of the Quantiva and the Xevo TQ-S.
Discussion
The results of the first comparison on the Quantiva and the third on the Micro I-class were in line with what was expected. There were some exceptions, corticosterone on the Quantiva and 17-hydroxyprogesterone on the Micro I-class. However, this could be explained due to either the low numbers op replicates or the low concentrations that were measured. Since all samples were prepared in the same way it could not cause a recovery problem and deviations in the concentrations for one particular steroid.
The results measured by the Xevo were not in line with the expected results nor the results of the Quantiva and Micro I-class. Since the chromatography was the exact same as those from the other systems it indicates that it had something to do with the mass spectrometer. This could either be ion suppression or matrix effect but with a small experiment adding 13C labelled components these were
both excluded. Somehow this could have something to do with the source configuration. For future recommendations there could be done some research into the source configuration in order to explain the differences in results. In addition, the company could provide some technical support in attempt to account for the deviating results.
It is important that the concentrations should be at least the same as those from the routine analysis. Because external quality controls showed that the current routine analysis measured concentrations within the error margin set for all hospitals and research facilities measuring the steroids. In addition, there is no need for new reference values.
Also, this new method is not only an improvement of the routine analysis but could in addition be used for different biological matrices. The LC-MS/MS system can be used with adaptable sample preparations which makes it possible to measure steroids from dried blood spots (DBS) or saliva. However using the different matrices can result in lower or higher concentrations. Therefore, the calibration curve should be checked if the concentrations from different matrices are within the range of the existing calibration curve.
The new method in this thesis reduced the analysis time by 3.5 minute per sample. This was established by optimization of the chromatography. However, this was not the only way to improve the analysis time. Kamemura et al. [35] tried to improve the throughput by multiplexing sample in one injection. This has been done by derivatization with different reagents.
Conclusion
In this research project a method has been developed for the separation of 9 endogenous steroids by LC-MS/MS. The developed method provide complete baseline separation between both critical pairs (Rs >1.6) within 6.5 minutes. Since the average steroid runs can take up to 150 samples the new method
could save up to 8,5 hours each run. The quantification of steroids measured with the new method showed good correlation, for all compounds, with the routine analysis at the Amsterdam UMC.
Acknowledgement
Mariëtte Ackermans; who was my daily supervisor. I want to thank you in the first place for the
opportunity you gave me to do my master research project at the Amsterdam UMC endocrine laboratory. I want to thank you for your support and advice during this journey. Also, I am very grateful for the amount of patience you had during all my questions about unclarities of the results. Furthermore, I want thank you for the feedback you gave me on this thesis. It was great working with you.
Frans Martens; senior analyst. I want to thank you for the incredible amount of technical skill and
References
1. de Carvalho IA, Epping-Jordan JA, Pot AM, et al. Organizing integrated health-care services to meet older people’s needs. Bull World Health Organ. 2017;95(11):756-763.
doi:10.2471/BLT.16.187617
2. Rodrigues LP, Vaz Tanaka SCS, Haas VJ, Cunali VCA, De Marqui ABT. Heel prick test: Maternal-fetal conditions that may have an effect on the test results in newborns admitted to the intensive care unit. Rev Bras Ter Intensiva. 2019;31(2):186-192. doi:10.5935/0103-507X.20190030
3. Ericson-Neilsen W, Kaye AD. Steroids: Pharmacology, complications, and practice Delivery Issues. Ochsner J. 2014;14(2):203-207.
4. Schiffer L, Barnard L, Baranowski ES, et al. Human steroid biosynthesis, metabolism and excretion are differentially reflected by serum and urine steroid metabolomes: A comprehensive review. J Steroid Biochem Mol Biol. 2019;194.
doi:10.1016/j.jsbmb.2019.105439
5. Baulieu EE. Neurosteroids: A novel function of the brain. In: Psychoneuroendocrinology. Vol 23. Pergamon; 1998:963-987. doi:10.1016/S0306-4530(98)00071-7
6. Badoud F, Boccard J, Schweizer C, Pralong F, Saugy M, Baume N. Profiling of steroid
metabolites after transdermal and oral administration of testosterone by ultra-high pressure liquid chromatography coupled to quadrupole time-of-flight mass spectrometry. J Steroid
Biochem Mol Biol. 2013;138:222-235. doi:10.1016/j.jsbmb.2013.05.018
7. Teubel J, Wüst B, Schipke CG, Peters O, Parr MK. Methods in endogenous steroid profiling – A comparison of gas chromatography mass spectrometry (GC–MS) with supercritical fluid chromatography tandem mass spectrometry (SFC-MS/MS). J Chromatogr A. 2018;1554:101-116. doi:10.1016/j.chroma.2018.04.035
8. Holst JP, Soldin OP, Guo T, Soldin SJ. Steroid hormones: Relevance and measurement in the clinical laboratory. Clin Lab Med. 2004;24(1):105-118. doi:10.1016/j.cll.2004.01.004
9. Arlt W, Stewart PM. Adrenal corticosteroid biosynthesis, metabolism, and action. Endocrinol
Metab Clin North Am. 2005;34(2):293-313. doi:10.1016/j.ecl.2005.01.002
10. Cain DW, Cidlowski JA. Specificity and sensitivity of glucocorticoid signaling in health and disease. Best Pract Res Clin Endocrinol Metab. 2015;29(4):545-556.
doi:10.1016/j.beem.2015.04.007
11. Kulle AE, Welzel M, Holterhus PM, Riepe FG. Principles and clinical applications of liquid chromatography - Tandem mass spectrometry for the determination of adrenal and gonadal steroid hormones. J Endocrinol Invest. 2011;34(9):702-708. doi:10.3275/7843
12. Dahl SR, Nermoen I, Brønstad I, Husebye ES, Løvås K, Thorsby PM. Assay of steroids by liquid chromatography– tandem mass spectrometry in monitoring 21-hydroxylase deficiency.
Endocr Connect. 2018;7(12):1542-1550. doi:10.1530/EC-18-0453
13. Wallin O, Söderberg J, Van Guelpen B, Stenlund H, Grankvist K, Brulin C. Blood sample collection and patient identification demand improvement: A questionnaire study of
preanalytical practices in hospital wards and laboratories. Scand J Caring Sci. 2010;24(3):581-591. doi:10.1111/j.1471-6712.2009.00753.x
14. Yalow RS, Berson SA. Assay of plasma insulin in human subjects by immunological methods.
Nature. 1959;184(4699):1648-1649. doi:10.1038/1841648b0
15. Hill M, Lapčík O, Hampl R, Stárka L, Putz Z. Radioimmunoassay of three deoxycorticoids in human plasma following HPLC separation. Steroids. 1995;60(9):615-620. doi:10.1016/0039-128X(95)00060-4
16. Köhler G, Milstein C. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature. 1975;256(5517):495-497. doi:10.1038/256495a0
17. Hoofnagle AN, Wener MH. The fundamental flaws of immunoassays and potential solutions using tandem mass spectrometry. J Immunol Methods. 2009;347(1-2):3-11.
doi:10.1016/j.jim.2009.06.003
18. Taieb J, Mathian B, Millot F, et al. Testosterone measured by 10 immunoassays and by isotope-dilution gas chromatography-mass spectrometry in sera from 116 men, women, and children. Clin Chem. 2003;49(8):1381-1395. doi:10.1373/49.8.1381
19. Herold DA, Fitzgerald RL. Immunoassays for Testosterone in Women: Better than a Guess?
Clin Chem. 2003;49(8):1250-1251. doi:10.1373/49.8.1250
20. Taylor AE, Keevil B, Huhtaniemi IT. Mass spectrometry and immunoassay: How to measure steroid hormones today and tomorrow. Eur J Endocrinol. 2015;173(2):D1-D12.
doi:10.1530/EJE-15-0338
21. Wudy SA, Schuler G, Sánchez-Guijo A, Hartmann MF. The art of measuring steroids: Principles and practice of current hormonal steroid analysis. J Steroid Biochem Mol Biol. 2018;179:88-103. doi:10.1016/j.jsbmb.2017.09.003
22. McDonald JG, Matthew S, Auchus RJ. Steroid profiling by gas chromatography-mass spectrometry and high performance liquid chromatography-mass spectrometry for adrenal diseases. Horm Cancer. 2011;2(6):324-332. doi:10.1007/s12672-011-0099-x
23. Marcos J, Pozo OJ. Derivatization of steroids in biological samples for GC-MS and LC-MS analyses. Bioanalysis. 2015;7(19):2515-2536. doi:10.4155/bio.15.176
24. Thienpont LM, Van Nieuwenhove B, Stöckl D, Reinauer H, De Leenheer AP. Determination of reference method values by isotope dilution-gas chromatography/mass spectrometry: a five years’ experience of two European Reference Laboratories. Eur J Clin Chem Clin Biochem. 1996;34(10):853-860. http://www.ncbi.nlm.nih.gov/pubmed/8933112. Accessed April 6, 2020.
25. Wooding KM, Auchus RJ. Mass spectrometry theory and application to adrenal diseases. Mol
Cell Endocrinol. 2013;371(1-2):201-207. doi:10.1016/j.mce.2012.12.026
26. Keevil BG. Novel liquid chromatography tandem mass spectrometry (LC-MS/MS) methods for measuring steroids. Best Pract Res Clin Endocrinol Metab. 2013;27(5):663-674.
27. Kushnir MM, Rockwood AL, Roberts WL, Yue B, Bergquist J, Meikle AW. Liquid
chromatography tandem mass spectrometry for analysis of steroids in clinical laboratories.
Clin Biochem. 2011;44(1):77-88. doi:10.1016/j.clinbiochem.2010.07.008
28. Nielen MWF, Lasaroms JJP, Mulder PPJ, Van Hende J, Van Rhijn JA, Groot MJ. Multi residue screening of intact testosterone esters and boldenone undecylenate in bovine hair using liquid chromatography electrospray tandem mass spectrometry. J Chromatogr B Anal
Technol Biomed Life Sci. 2006;830(1):126-134. doi:10.1016/j.jchromb.2005.10.028
29. Musharraf SG, Ali A, Khan NT, Yousuf M, Choudhary MI, Atta-Ur-Rahman. Tandem mass spectrometry approach for the investigation of the steroidal metabolism: Structure-fragmentation relationship (SFR) in anabolic steroids and their metabolites by ESI-MS/MS analysis. Steroids. 2013;78(2):171-181. doi:10.1016/j.steroids.2012.10.017
30. Multisteroid LC–MS/MS assay for glucocorticoids and androgens and its application in Addison’s disease in: Endocrine Connections Volume 2 Issue 3 (2013).
https://ec.bioscientifica.com/view/journals/ec/2/3/125.xml. Accessed April 4, 2020.
31. Pirok BWJ, Pous-Torres S, Ortiz-Bolsico C, Vivó-Truyols G, Schoenmakers PJ. Program for the interpretive optimization of two-dimensional resolution. J Chromatogr A. 2016;1450:29-37. doi:10.1016/j.chroma.2016.04.061
32. Quarry MA, Grob RL, Snyder LR. Prediction of Precise Isocratic Retention Data from Two or More Gradient Elution Runs. Analysis of Some Associated Errors. Anal Chem. 1986;58(4):907-917. doi:10.1021/ac00295a056
33. Schoenmakers PJ, Billiet HAH, Tussen R, De Galan L. Gradient selection in reversed-phase liquid chromatography. J Chromatogr A. 1978;149(C):519-537. doi:10.1016/S0021-9673(00)81008-0
34. Matuszewski BK, Constanzer ML, Chavez-Eng CM. Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC-MS/MS. Anal Chem.
2003;75(13):3019-3030. doi:10.1021/ac020361s
35. Kamemura M, Yokota M, Ogawa S, Sugiura T, Higashi T. Sample-multiplexing by
derivatization using multiple analogous reagents for enhancing throughput in LC/ESI-MS/MS assay of steroids: Plasma 17α-hydroxyprogesterone as an example. J Chromatogr B Anal
0 10 20 30 40 50 60 70 80 0 20 40 60 80 TST_S T ST y = 0,0106 + 1,038 x n = 84 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 3,0 0 20 40 60 80 100 Mean of TST and TST_S T ST - T ST _ S Mean 0,34 -1.96 SD -0,64 +1.96 SD 1,32 0 50 100 150 200 0 50 100 150 200 Cortison_S C o rt is o n y = -0,234 + 1,017 x n = 84 -10 -5 0 5 10 15 0 50 100 150 200 250 Mean of Cortison and Cortison_S
C o rt is o n - C o rt is o n _ S Mean 1,0 -1.96 SD -5,2 +1.96 SD 7,2 0 100 200 300 400 500 600 700 800 0 200 400 600 800 Cortisol_S C o rt is o l y = 4,149 + 0,983 x n = 84 -100 -80 -60 -40 -20 0 20 40 0 200 400 600 800 1000
Mean of Cortisol and Cortisol_S
C o rt is o l - C o rt is o l_ S Mean -2,0 -1.96 SD -36,3 +1.96 SD 32,3 APPENDIX I Quantiva
Below the Passing-Bablock regression graphs are displayed. On the X-axis the concentrations measured by the serum method (old method). On the Y-axis the concentrations measured with the new developed method.
0 50 100 150 200 0 50 100 150 200 11_DOCL_S 1 1 _ D O C L y = -0,0188 + 1,028 x n = 27 -4 -2 0 2 4 6 8 10 0 50 100 150 200 250
Mean of 11_DOCL and 11_DOCL_S
1 1 _ D O C L - 1 1 _ D O C L _ S Mean 0,5 -1.96 SD -3,1 +1.96 SD 4,1 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 CCSn_S C C Sn y = 0,165 + 0,915 x n = 83 -3,0 -2,5 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 0 5 10 15 20 25 30 35 Mean of CCSn and CCSn_S C C Sn - C C Sn _ S Mean -0,39 -1.96 SD -1,68 +1.96 SD 0,90 0 5 10 15 20 0 5 10 15 20 Andro_S An d ro y = -0,188 + 1,038 x n = 84 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 0 5 10 15 20 25
Mean of Andro and Andro_S
An d ro - An d ro _ S Mean 0,02 -1.96 SD -0,83 +1.96 SD 0,87 0 20 40 60 80 100 120 140 160 180 0 50 100 150 200 17_OHP_S 1 7 _ O H P y = -0,233 + 0,992 x n = 43 -4 -2 0 2 4 6 8 10 0 50 100 150 200
Mean of 17_OHP and 17_OHP_S
1 7 _ O H P - 1 7 _ O H P_ S Mean -0,0 -1.96 SD -2,8 +1.96 SD 2,7
0 5 10 15 20 25 30 35 40 45 0 10 20 30 40 50 11_DCCSN_S 1 1 _ D C C SN y = 0,0404 + 1,064 x n = 29 0,02 0,03 0,04 0,05 0,06 0,07 0 10 20 30 40 50 Mean of 11_DCCSN and 11_DCCSN_S 1 1 _ D C C SN - 1 1 _ D C C SN _ S Mean 0,047 -1.96 SD 0,030 +1.96 SD 0,064
0 10 20 30 40 50 60 70 80 0 20 40 60 80 TST_S T ST y = 0,00191 + 0,778 x n = 79 -20 -15 -10 -5 0 5 10 0 20 40 60 80 Mean of TST and TST_S T ST - T ST _ S Mean -4,0 -1.96 SD -12,6 +1.96 SD 4,6 100 150 200 250 300 350 400 450 500 100 150 200 250 300 350 400 450 500 Cortisol_S C o rt is o l y = -6,230 + 0,930 x n = 79 -50 -45 -40 -35 -30 -25 -20 -15 -10 100 200 300 400 500
Mean of Cortisol and Cortisol_S
C o rt is o l - C o rt is o l_ S Mean -28,6 -1.96 SD -42,7 +1.96 SD -14,5 30 40 50 60 70 80 90 100 30 40 50 60 70 80 90 100 Cortison_S C o rt is o n y = 0,890 + 1,000 x n = 79 -10 -5 0 5 10 15 20 30 40 50 60 70 80 90 100 110 Mean of Cortison and Cortison_S
C o rt is o n - C o rt is o n _ S Mean 1,5 -1.96 SD -4,6 +1.96 SD 7,6 Appendix II XEVO
The Raw data can be found:
De MedCalc files and the corresponding excel files can be found in:
\\ENDO-GCLC\stagiaires\Marco\Steroiden\uitwerking vergelijking DBS-serum Xevo
Below the Passing-Bablock regression graphs are displayed. On the X-axis the concentrations measured by the serum method (old method). On the Y-axis the concentrations measured with the new developed method.
0 50 100 150 200 250 0 50 100 150 200 250 11DOC_S 1 1 D O C y = 0,150 + 1,299 x n = 24 -30 -20 -10 0 10 20 30 40 50 0 50 100 150 200 250 300 Mean of 11DOC and 11DOC_S
1 1 D O C - 1 1 D O C _ S Mean 9,0 -1.96 SD -22,3 +1.96 SD 40,3 0 2 4 6 8 10 12 14 16 0 5 10 15 Cortico_S C o rt ic o y = -0,00159 + 1,017 x n = 79 -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 0 5 10 15 20
Mean of Cortico and Cortico_S
C o rt ic o - C o rt ic o _ S Mean 0,09 -1.96 SD -0,32 +1.96 SD 0,50 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 11_Dcorti_S 1 1 _ D co rt i y = 0,152 + 0,981 x n = 15 -15 -10 -5 0 5 10 0 2 4 6 8 10
Mean of 11_Dcorti and 11_Dcorti_S
1 1 _ D c o rt i - 1 1 _ D c o rt i_ S Mean -2,4 -1.96 SD -11,2 +1.96 SD 6,5 0 5 10 15 20 25 0 5 10 15 20 25 Andro_S An d ro y = 0,0317 + 0,808 x n = 79 -5 -4 -3 -2 -1 0 1 0 5 10 15 20 25
Mean of Andro and Andro_S
An d ro - An d ro _ S Mean -0,7 -1.96 SD -2,1 +1.96 SD 0,7
0 5 10 15 20 25 30 35 40 0 10 20 30 40 17OHP_S 1 7 O H P y = 0,00818 + 1,273 x n = 39 -4 -2 0 2 4 6 8 0 10 20 30 40
Mean of 17OHP and 17OHP_S
1 7 O H P - 1 7 O H P_ S Mean 0,9 -1.96 SD -2,0 +1.96 SD 3,7
0 10 20 30 40 50 60 0 10 20 30 40 50 60 TST_S T ST y = -0,0223 + 0,987 x n = 71 -1,0 -0,8 -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0 10 20 30 40 50 60 70 Mean of TST and TST_S T ST - T ST _ S Mean -0,12 -1.96 SD -0,57 +1.96 SD 0,32 0 100 200 300 400 500 600 700 0 100 200 300 400 500 600 700 Cortisol_S C o rt is o l y = -2,816 + 1,026 x n = 72 -15 -10 -5 0 5 10 15 20 25 30 35 0 100 200 300 400 500 600 700 800 Mean of Cortisol and Cortisol_S
C o rt is o l - C o rt is o l_ S Mean 4,4 -1.96 SD -8,9 +1.96 SD 17,6 0 50 100 150 200 250 0 50 100 150 200 250 Cortison_S C o rt is o n y = 0,0489 + 1,006 x n = 72 -10 -5 0 5 10 15 0 50 100 150 200 250
Mean of Cortison and Cortison_S
C o rt is o n - C o rt is o n _ S Mean 0,6 -1.96 SD -6,0 +1.96 SD 7,2 APPENDIX III MICRO
The Raw data can be found:
De MedCalc files and the corresponding excel files can be found in:
\\ENDO-GCLC\stagiaires\Marco\Steroiden\uitwerking vergelijking DBS-serum Micro
Below the Passing-Bablock regression graphs are displayed. On the X-axis the concentrations measured by the serum method (old method). On the Y-axis the concentrations measured with the new developed method.
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 CCSN_S C C SN y = -0,0280 + 1,003 x n = 72 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 0 10 20 30 40 50 60 70 Mean of CCSN and CCSN_S C C SN - C C SN _ S Mean 0,01 -1.96 SD -0,55 +1.96 SD 0,57 0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350 11_DOCL_S 1 1 _ D O C L y = -0,181 + 1,023 x n = 23 -6 -4 -2 0 2 4 6 8 10 12 0 100 200 300 400
Mean of 11_DOCL and 11_DOCL_S
1 1 _ D O C L - 1 1 _ D O C L _ S Mean 1,0 -1.96 SD -4,9 +1.96 SD 7,0 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 11_DCCSN_S 1 1 _ D C C SN y = 0,0633 + 0,970 x n = 29 -4 -3 -2 -1 0 1 2 0 20 40 60 80 Mean of 11_DCCSN and 11_DCCSN_S 1 1 _ D C C SN - 1 1 _ D C C SN _ S Mean -0,4 -1.96 SD -2,3 +1.96 SD 1,5 0 5 10 15 20 25 0 5 10 15 20 25 Andro_S An d ro y = 0,0108 + 0,963 x n = 72 -1,0 -0,5 0,0 0,5 1,0 1,5 0 5 10 15 20 25
Mean of Andro and Andro_S
An d ro - An d ro _ S Mean -0,12 -1.96 SD -0,84 +1.96 SD 0,59
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 17_OHP_S 1 7 _ O H P y = -0,239 + 0,929 x n = 21 -3,0 -2,5 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 0 5 10 15 20 25 30 35 Mean of 17_OHP and 17_OHP_S
1 7 _ O H P - 1 7 _ O H P_ S Mean -0,47 -1.96 SD -2,34 +1.96 SD 1,41
