Multisteroid LC–MS/MS assay for glucocorticoids and androgens and its application in Addison's ...

12 Pages·2013·0.3 MB·English

PMethlieetal. Quantitationofsteroidsby 1–12 2:125
Research
LC–MS/MS
OpenAccess
Multisteroid LC–MS/MS assay for
glucocorticoids and androgens and
its application in Addison’s disease
PaalMethlie1,2,SteinarHustad1,RalfKellman2,BjørgAlma˚s2,MartinaMErichsen3,
EysteinSHusebye1,3andKristianLøva˚s1,3
Correspondence
1DepartmentofClinicalScience,UniversityofBergen,N-5021Bergen,Norwayand2TheHormoneLaboratory shouldbeaddressed
3DepartmentofMedicine,HaukelandUniversityHospital,N-5021Bergen,Norway toPMethlie
Email
[email protected]
Abstract
Objective:Liquidchromatography–tandemmassspectrometry(LC–MS/MS)offerssuperior KeyWords
analyticalspecificitycomparedwithimmunoassays,butitisnotavailableinmanyregions " liquidchromatography
ns andhospitalsduetoexpensiveinstrumentationandtedioussamplepreparation. massspectrometry
o
cti Thus,wedevelopedanautomated,high-throughputLC–MS/MSassayforsimultaneous " androgens
e
nn quantificationoftenendogenousandsyntheticsteroidstargetingdiseasesofthe " glucocorticoids
o
C hypothalamic–pituitary–adrenalaxisandgonads. " adrenalinsufficiency
e
rin Methods:Deuteratedinternalstandardswereaddedto85mlserumandprocessedby " Addison’sdisease
c
o
d liquid–liquidextraction.Cortisol,cortisone,prednisolone,prednisone,11-deoxycortisol,
n
E dexamethasone,testosterone,androstenedioneandprogesteronewereresolvedby
ultra-high-pressurechromatographyonareversed-phasecolumnin6.1minanddetectedby
triple-quadrupolemassspectrometry.Themethodwasusedtoassesssteroidprofilesin
womenwithAddison’sdisease(AD,nZ156)andblooddonors(BDs,nZ102).
Results:Precisionsrangedfrom4.5to10.1%relativestandarddeviations(RSD),accuracies
from95to108%andextractionrecoveriesfrom60to84%.Themethodwaspracticallyfree
ofmatrixeffectsandrobusttoindividualdifferencesinserumcomposition.Most
postmenopausalADwomenhadextremelylowandrostenedioneconcentrations,below
0.14nmol/l,andmediantestosteroneconcentrationsof0.15nmol/l(interquartilerange
0.00–0.41),considerablylowerthanthoseofpostmenopausalBDs(1.28nmol/l(0.96–1.64)
and0.65nmol/l(0.56–1.10)respectively).ADwomeninfertileyearshadandrostenedione
concentrationsof1.18nmol/l(0.71–1.76)andtestosteroneconcentrationsof0.44nmol/l
(0.22–0.63),approximatelyhalfofthosefoundinBDsofcorrespondingage.
Conclusion:ThisLC–MS/MSassayprovideshighlysensitiveandspecificassessmentsof
glucocorticoidsandandrogenswithlowsamplevolumesandissuitableforendocrine
laboratoriesandresearch.Itsutilityhasbeendemonstratedinalargecohortofwomen
withAD,andthedatasuggestthatwomenwithADareparticularlyandrogendeficient
aftermenopause.
EndocrineConnections
(2013)2,125–136
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 2–12 2:126
LC–MS/MS
Introduction
Glucocorticoid and androgen hormone measurements the measurement of endogenous steroids. For example,
play a decisive role in the diagnosis and management of the measurements of serum dexamethasone (DXM)
many disorders of the hypothalamic–pituitary–adrenal could provide pharmacokinetic data useful in the
(HPA) axis and gonads. Examples include Addison’s evaluation of the DXM suppression test (9, 10), and
disease (AD), Cushing’s syndrome (CS) and congenital therapeuticdrugmonitoringofprednisoneandpredniso-
adrenal hyperplasia, male hypogonadism and female lone has the potential to optimize treatment (11, 12).
hyperandrogenism.Althoughimmunoassayshavelargely Becauseofthewidespreaduseoftheseglucocorticoidsand
been used to measure steroid hormones, over the last impact on endogenous cortisol levels, information on
decade, mass spectrometry (MS) has increasingly been serum levels may also be valuable in the setting of a
adopted as it offers superior analytical specificity and routineendocrinelaboratory.
accuracy. However, MS is not commonly available to Therefore,theobjectiveofthisstudywastodevelopa
cliniciansinmanyregionsandhospitals.Thehighcostof LC–MS/MS method that quantifies endogenous and
MS instruments, lack of expertise, and cumbersome and syntheticglucocorticoidhormonesaswellasendogenous
time-consuming sample preparation procedures are androgens.Withasamplevolumeofonly85mlserum,the
factors likely to slow its adoption. The development of method provides a highly sensitive, fast, comprehensive
simple,reliable,andcost-andlabour-efficientmethodsis, and cost-effective evaluation of patients with a range of
therefore,important. disorders related to glucocorticoid and androgen hor-
Liquid chromatography coupled to tandem mass mones. Its utility was demonstrated by comparing the
spectrometry (LC–MS/MS) is a highly selective mode of steroid profiles of women with AD with those of healthy
s detection. It enables simultaneous determination of femaleblooddonors(BDs).Toourknowledge,thisreport
n
tio multiple steroids at very low concentrations in a single isthefirsttoinvestigateandrogenlevelsinalargecohort
c
ne analytical run. Although many papers have been pub- ofwomenwithADusinghighlyspecificMS.
n
o lished on LC–MS/MS assays that measure one or a few
C
ne steroidhormones,onlyfewpapershavereportedmethods Materials and methods
ri
oc that extensively exploit the capability of multiplex
d Chemicals
n determination of steroids in human serum (1, 2, 3). This
E
strategycouldsimplifythelaboratoryset-upbyeliminat- Cortisol, cortisone, prednisolone, prednisone, 11-deoxy-
ingtheneedtousenumerousseparatemethodsandmake cortisol (11DOC), testosterone, androstenedione,
instrumentutilizationmoreefficient. 17a-hydroxyprogesterone(17OHP)andprogesteronewere
In MS/MS, the specificity is increased by collision- obtainedfromSteraloids,Inc.(Newport,RI,USA),andDXM
induced fragmentation of the analyte into a molecular wasobtainedfromAlfaAesar(WardHill,MA,USA).
fingerprintthatcouldbeidentifiedbymonitoringtwoor The deuterated internal standards (ISs) cortisol-d4
more fragments. However, many previously published (cortisol-9,11,12,12-d4, 97–98% atom D), cortisone-d2
multisteroid LC–MS/MS assays monitor only one mass (4-pregnen-17a,21-diol-3,11,20-trione-1,2-d2, O98% atom
transition (collision fragment) per analyte (2, 3, 4, 5, 6) D), prednisolone-d6 (1,4-pregnadien-11b,17a,21-triol-3,
andconsequentlyhaveanincreasedriskofnotdetecting 20-dione-2,4,6,6,21,21-d6, O98% atom D), DXM-d4
isobaric interference. Another analytical concern is that (DXM-4,6a,21,21-d4, 96–98% atom D), 11DOC-d2
inter-individualdifferencesinserumcomposition,suchas (4-pregnen-17a,21-diol-3,20-dione-21,21-d2, O96% atom
lipids and binding proteins, could potentially impact D),testosterone-d3(testosterone-16,16,17-d3,O98%atom
measurements. This is often not investigated at all. Most D), androstenedione-d7 (4-androsten-3,17-dione-2,2,4,6,6,
published methods also require relatively large sample 16,16-d7,O98%atomD), 17OHP-d8(4-pregnen-17a-ol-3,
volumestoprovidesufficientsensitivity(1,2,4,7,8)and 20-dione-2,2,4,6,6,21,21,21-d8, O98% atom D) and
havelongruntimes(1,2,4,7). progesterone-d9 (progesterone-2-2-4-6-6-17a,21,21,21-d9,
Thework-upandmanagementofendocrinedisorders O98% atom D) were purchased from CDN Isotopes
commonlyinvolvetheadministrationofsyntheticsteroid (Pointe-Claire, Quebec, Canada).
hormones. The determination of synthetic glucocorti- MS-grade acetonitrile (ACN), methanol (MeOH),
coids used in dynamic endocrine testing and pharma- formic acid (O98%) and ammonium formate (O98%)
cologicaltreatmentmaythusbeavaluablesupplementto were obtained from Merck. Milli-Q water purification
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 3–12 2:127
LC–MS/MS
system (Millipore, Burlington, MA, USA) was used to hexane(80:20).Thesupernatant(600ml)wastransferred
prepare de-ionized water (O18MUcm). Dextran-coated toanewvialandwashedwith50mlammoniumformate
charcoal (product number C6241) was obtained from buffer (pH 9.0, 0.1M). Subsequently, 500ml of organic
Sigma–Aldrich. phase were transferred to a new vial and evaporated at
508Cfor25minunderN flowandfinallyreconstitutedin
2
Calibrators,ISsandqualitycontrols 50mlH O:formicacid:MeOH(49.9:0.1:50).Sampleprep-
2
aration was automated on a Hamilton Star (Hamilton
Steroid-free human serum was prepared with activated
Robotics, Inc., Reno, NV, USA) using exclusively 1.1ml
dextran-coated charcoal according to the manufacturer’s
glass vials. Mixing steps were performed by repeated
protocol.SerumwasobtainedfromhealthyBDs.
pipette aspiration/dispense cycles. Centrifugation was
All steroid hormones were separately dissolved in
not necessary because the aquatic and organic phases
MeOH at concentrations of 3mM for the analytes and
separatedwithinminutes.
500mMfortheISs.Twoseparatemixturesoftheanalytes
andtheISs,designatedstandardsubstockandISsubstock,
werepreparedinMeOH.Thestandardsubstockcontained LC–MS/MSconditions
eachsteroidhormoneataconcentration100timesthatof
An Agilent 1290 UPLC system (Santa Clara, CA, USA)
the highest working calibrator. Working calibrators were
equipped with a thermostated autosampler (48C) and a
preparedbyseriallydilutingthestandardsubstock1:4in
degasser was used for chromatographic separation. Pro-
MeOHandthenbyadding2mlofeachdilutionto198ml
cessedserum(5ml)wasseparatedoveraZorbaxRRHDC18
steroid-free serum. The final working calibrators covered
reversed-phase column (50!2.1mm, 1.8mm particle
thefollowing measuring ranges: 1.95–2000nmol/l, corti-
size, Agilent) maintained at 308C. The mobile phases
sol; 0.98–250nmol/l, cortisone; 0.98–1000nmol/l, pre-
ns were water (A) and ACN (B) with 0.1% formic acid. The
o dnisolone;0.98–1000nmol/l,prednisone;0.06–250nmol/l,
ti column was developed with stepwise linear gradient
ec DXM; 0.10–25nmol/l, 11DOC; 0.02–75.0nmol/l,
n elution according to the following timetable: 0.00min,
on testosterone; 0.12–125nmol/l, androstenedione;
C 15%B;0.25min,15%B;0.50min,25%B;1.70min,28.5%B;
e 0.24–250nmol/l,17OHP;0.24–250nmol/l,progesterone;
n 3.80min,40%B;4.20min,50%B;4.90min,70%B;5.00min,
cri and a blank control. Six calibrators were used for all the
o 95%B;5.50min,95%B;5.60min,15%B;and6.10min,
d analytes,exceptforcortisoneand11DOC(fivecalibrators)
En 15%B.Flowratewas1000ml/minandthecolumneffluent
andtestosteroneandDXM(sevencalibrators).
was delivered to the mass spectrometer in the time
TheconcentrationsofeachoftheISsintheISsubstock
window1.1–5.0min.
were 100 times that of the working IS solution. The
The LC system was coupled to an Applied Biosys-
working IS solution was prepared by diluting the IS
tems/MDS SCIEX API 5500 triple-quadrupole mass spec-
substock with H O:MeOH (1:1). The final working IS
2 trometer operating with electrospray ionization (Applied
solution contained 850nmol/l cortisol-d4, 850nmol/l
Biosystems/MDS, Foster City, CA, USA). Manual tuning
cortisone-d2, 425nmol/l prednisolone-d6, 106nmol/l
and selectionof mass transitions wereperformed using a
DXM-d4,128nmol/l11DOC-d2,51.0nmol/landrostene-
T-splittomixthe5ml/mininfusioncontaining500nmol/l
dione-d7, 31.9nmol/l testosterone-d3, 106nmol/l
ofthecompoundwiththe1000ml/minmobilephaseflow
17OHP-d8and106nmol/lprogesterone-d9.
(35% B). Using this set-up, additional Q1 and Q3 scans
Quality controls (QCs) were prepared in steroid-free
(90–500 m/z) were performed by varying the ion source
serum at three levels (low, medium and high) by adding
parameters (voltage, temperature and gas flows) to
eachsteroidhormonetoafinalconcentrationcorrespond-
investigatethestabilityofthesubstancesunderconserva-
ingto0.30,5.0and50%ofthehighestworkingstandard.
tive and extreme conditions, as well as the formation of
All standards, calibrators and QCs were stored in Nunc
adducts. Ion source parameters and mass transitions are
cryovials (Nalge Nunc International, Roskilde, Denmark)
reportedinTable1.Entrancepotentialwas10V,collision
atK808C.
gas was set to medium and resolution was set to unit for
Q1 and Q3. To increase scan time for each analyte, the
Samplepreparationprocedure
multiple reaction monitoring (MRM) was divided into
To85mlofserum,workingcalibratorsorQCs,10mlofanIS seven periods based on the time of elution. All hardware
were added, mixed and equilibrated for 1h. The steroid was managed using the Analyst Software (version 1.5.1;
hormoneswerethenextractedwith825mlethylacetate: AppliedBiosystems/MDS).
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 4–12 2:128
LC–MS/MS
Table1 MRMtransitionsandcompound-dependentparameters.
Quantifier Qualifier
Dwell Dwell
MRMtransition time DP CE CXP MRMtransition time DP CE CXP
Period Q1 Q3 (ms) (V) (V) (V) Q1 Q3 (ms) (V) (V) (V)
1
Cortisol 407.2 297.0 35 K50 K42 K23 407.2 282.0 10 K50 K49 K23
Cortisol-d4 411.4 301.1 35 K50 K45 K23 411.4 286.1 10 K50 K48 K23
Prednisolone 405.2 295.0 35 K50 K42 K23 405.2 280.0 10 K50 K48 K23
Prednisolone-d6 411.3 333.2 35 K50 K23 K23 411.3 284.1 10 K50 K48 K23
Prednisone 403.3 299.1 35 K50 K25 K23 403.3 285.1 10 K50 K40 K23
Cortisone 405.3 301.1 35 K50 K26 K23 405.3 311.1 10 K50 K41 K23
Cortisone-d2 407.3 303.1 35 K50 K28 K23 407.3 313.0 10 K50 K41 K23
2
DXM 437.4 361.1 80 K80 K24 K24 437.4 307.1 80 K80 K42 K15
DXM-d4 441.3 309.0 80 K80 K42 K20 441.3 363.2 80 K80 K24 K24
3
11DOC 347.2 96.9 80 50 30 12 347.1 108.9 80 50 31 12
s
n
o
ti 11DOC-d2 349.3 108.9 80 50 33 12 349.3 96.9 80 50 30 12
c
e
n 4
n
o
C Testosterone 289.2 97.0 80 85 30 12 289.2 109.0 80 85 33 12
e
rin Testosterone-d3 292.2 109.0 80 91 35 12 292.2 97.0 80 91 33 12
c
o
d 5
n
E
Androstenedione 287.2 96.9 60 105 30 16 287.2 108.9 60 105 33 15
Androstenedione-d7 294.1 99.7 60 116 29 16 294.1 113.0 60 116 35 15
6
17OHP 331.3 96.9 60 88 30 10 331.3 108.9 60 88 34 14
17OHP-d8 339.2 100.0 60 106 33 16 339.2 113.1 60 106 41 12
7
Progesterone 315.3 96.9 35 95 30 12 315.3 108.9 35 95 28 12
Progesterone-d9 324.2 100.2 35 110 37 12 324.2 113.0 35 110 27 12
DP, declustering potential; CE, collision energy; CXP, collision exit potential; 11DOC, 11-deoxycortisol; DXM, dexamethasone;
17OHP,17a-hydroxyprogesterone.
QuantificationandQC calibrator) of their nominal concentration. Two sets of
QCs (low, medium and high) were run in each batch
Quantificationwasbasedonpeakarearatiosoftheanalyte
(96 samples). The measured concentrations of at least
to the corresponding IS. Standard curves were computed
one QC at each level and at least 75% of all the QCs
usingalinearleast-squaresregressionanalysiswith1/xor
1/x2weightingtoassignprioritytothelowerrangeofthe were required to be within G15% their nominal levels.
calibrationcurves. Interference was suspected if the peak area ratio of
To accept analytical runs, the back-calculated quantifier to qualifier transitions deviated more
concentrations of at least 80% of the working calibra- than G40% from the corresponding ratio of the third
tors had to be within G15% (20% for the lowest highest calibrator.
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 5–12 2:129
LC–MS/MS
Methodvalidation supplementary data given at the end of this article) were
prepared separately in water at concentrations of
Within-dayandbetween-dayprecisionswereinvestigated
2000nmol/l. These included all drugs containing steroid
atthreelevelsbyrunningtheQCsinreplicates(nZ6)on
hormonesthatarecommerciallyavailableinNorway.We
12 different days. The coefficients of variance (CV) were
also reviewed data from random samples received at the
calculated using one-way ANOVA. Accuracy (%) was
Hormone Laboratory (nZ926), from rheumatic patients
computed as (measured concentration/nominal concen-
onprednisolonetherapy(nZ50)andfrompatientstaking
tration)!100, using the means of all the measured
DXM (nZ25) for interfering peaks and deviations of the
concentrations. The method was also compared with
quantifier:qualifierpeakarearatio.
immunoassays using the Passing–Bablok regression and
Stability was investigated by computing the mean of
Bland–Altmanplots.
triplet analysis of QCs at three concentration levels after
The limits of detection (LoDs) and lower limits of
storageunderdifferentconditions.Stabilitywasexamined
quantification (LLoQs) were determined by adding the
afterstorageatambienttemperaturefor1,2,4,8,24and
analytesatprogressivelylowerlevelstosteroid-freeserum
96h,inarefrigerator(48C)for1,2,5,10,20and30days,
and analysing samples in replicates (nZ6). The LoD was
and in a freezer (K208C) for 1, 2 and 3 months. Freeze–
defined as the concentration with a signal-to-noise ratio
thawstabilitywasdeterminedafterfivecyclesoffreezing
above3andtheLLoQasthelowestconcentrationwitha toK808C.Thestabilityoftheprocessedsamples(nZ96)
CVbelow20%andaccuracywithinG20%.
in the autosampler was studied by repeatedly analysing
Recovery, matrix effects (MEs) and linearity were
thesamebatchon4consecutivedays.
investigated in charcoal-stripped serum obtained from
differentindividuals,adoptingtheprinciplessuggestedby
s Matuszewskietal.(13).Analyteswereaddedtosteroid-free Application:steroidprofilesofwomenwithADandBDs
n
tio serum obtained from six individuals; aliquots of each The method was applied to samples obtained from the
c
ne serum were spiked to six levels corresponding to the Norwegian registry of organ-specific autoimmune
n
o workingstandards.Duplicateserawerespikedbothbefore
C diseases. The study population has been described
ne extraction(pre-extractspike)andafterevaporation(post- previously (15), and sera obtained from women with AD
ri
oc extractspike)byadding20mlofthestandardsubstockper (nZ156) and female BDs (nZ102) were analysed. The
d
En 980mlserum.Moreover,analyteswereaddedtoa‘blank’ Regional Committee for Medical Ethics of Western Nor-
reconstitutionsolvent(reference).Referencesampleswere way and the National Data Inspectorate approved the
analysedimmediatelybefore,inthemiddleandafterthe study. All of the patients gave written informed consent
pre/post-spike series. Extraction recoveries (ERs), ion- after complete explanation of the purpose and nature of
source-related MEs and overall process efficiency (PE) alltheproceduresused.MostoftheADpatientswereon
were calculated from the mean of duplicate samples cortisoneacetatereplacementtherapy,butfivewereusing
according to the formulas given in Supplementary prednisolone.Tenpatientshadprematureovarianfailure
Table 1, see section on supplementary data given at the and seven were on DHEA replacement therapy. Blood
endofthisarticle.Evaporationoftheorganicphaseduring samples were drawn between 0800 and 1600h. For
samplepreparationwascloselymonitoredbygravimetric statistical analyses, the subjects were stratified into three
measurementsofsimultaneouslyruncontrolsamples,and groups:young(!40years),middleaged(40–60years)and
ERs,MEsandPEwereadjustedaccordingly. old (O60 years). The groups were compared by the
Linearity was evaluated based on residual plots Kruskal–Wallis test using the Wilcoxon test as a post hoc
and correlation coefficients. Relative MEs (14), that is, analysis, and Bonferroni-adjusted P values are reported.
the impact of inter-individual differences in serum P!0.05wasconsideredstatisticallysignificant.
composition, were assessed by computing the slope
constant based on each of the six pre-extraction spiked
sera from the recovery experiment. The slope constants Results
were derived from weighted (1/x or 1/x2) least-squares
Precision,accuracy,lowerlimitofdetection
linear regression of nominal concentration vs analyte:IS
andquantification
peakarearatio.
All analytes and ISs and other potentially interfering Data from the precision and accuracy experiments are
steroidhormones(SupplementaryTable2,seesectionon reported in Table 2. Total CV were %10.1% for all the
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 6–12 2:130
LC–MS/MS
Table2 Precisionandaccuracy.
Andro- Pro-
Cortisol Cortisone Prednisolone Prednisone DXM 11DOC Testosterone stenedione 17OHP gesterone
QC1
Mean 5.80 3.01 2.96 3.05 0.747 0.305 0.223 0.372 0.757 0.736
Within-dayCV(%) 5.8 6.8 7.7 7.2 4.3 8.6 4.6 9.3 7.1 7.0
TotalCV(%) 7.4 6.9 10.1 9.8 5.1 9.7 5.2 9.3 9.8 8.4
Accuracy(%) 97 100 99 102 100 102 99 99 101 98
QC2
Mean 101 50.4 51.3 53.8 12.5 5.00 3.71 6.10 12.2 12.5
Within-dayCV(%) 3.0 3.9 4.3 3.8 3.7 4.7 5.1 3.9 4.5 6.1
TotalCV(%) 4.5 4.7 5.3 8.7 4.8 5.7 5.6 5.0 7.7 6.9
Accuracy(%) 101 101 103 108 100 100 99 98 98 100
QC3
Mean 1004 – 517 535 128 – 36.4 61.6 121 118
Within-dayCV(%) 4.7 – 5.0 4.7 4.0 – 4.7 4.1 4.6 4.5
TotalCV(%) 5.7 – 5.5 8.8 5.2 – 4.9 5.0 8.0 6.8
Accuracy(%) 100 – 103 107 102 – 97 99 96 95
s
n
o QCswereruninreplicatesof6on12differentdays.TotalCVincludeswithin-dayandbetween-dayvariability.Allconcentrationsarereportedasnmol/l.QC3
ti levelsofcortisoneand11DOCareabovetheupperlimitofquantificationandthusomitted.CV,coefficientofcorrelation;11DOC,11-deoxycortisol;DXM,
c
e dexamethasone;17OHP,17a-hydroxyprogesterone.
n
n
o
C
e compounds,andaccuracieswereintherangeof95–108%. to intra-individual differences in the serum. On running
n
cri The LoDs and LLoQs are reported in Table 3, and routinesamples,itwasobservedthatthepeakareasofthe
o
nd chromatograms of the lowest calibrators are shown in ISs were consistent over a batch of 96 samples with CV
E
Fig. 1. Comparison of the LC–MS/MS method with typicallyintherangeof10–20%.
immunoassaysavailableinourlaboratoryisshowninFig.2.
Selectivity
Recovery,linearityandMEs
Potential analytical interference of 48 synthetic and
Recoveryandlinearitywereinvestigatedinseraobtained naturally occurring steroid hormones and metabolites
fromsixindividuals(SupplementaryTable1).TheERsand (SupplementaryTable2)wasinvestigated.Alltheanalytes
PE were high and consistent across low-to-high concen- were chromatographically separated, except prednisone
tration levels. The ER was 60% for progesterone, and it and prednisolone. Individual injection of these two
ranged from 71 to 85% for the other analytes. There was compounds at 2000nmol/l showed that the selected
no significant ion depression or enhancement. The fragment ions of these glucocorticoids did not interfere
method was linear for all the compounds, except for in the negative ionization mode. In the systematic
cortisone and 11DOC. For these two compounds, the evaluationofchromatogramsfrom1000routineanalyses,
highestcalibratorsof250and25nmol/lrespectivelywere suspicion of interference was very rare. The quantifier:
slightly underestimated, but had acceptable back-calcu- qualifierpeakarearatiowaswithinG40%oftheexpected
latedaccuraciesabove90%. valueforalltheanalytesofallthesamples.
Thevarianceoftheslopeconstantsbetweendifferent
individualsprovidesausefulindexofrelativeMEs.TheCV
Stability
of the slope constants were %3% for all the analytes
(Supplementary Table 3, see section on supplementary Steroidhormonesintheserumwerestablewhenstoredat
data given at the end of this article). According to 48Coralowertemperature.Afterstorageforupto25days
Matuszewski(14),thisindicatesthatthemethodisrobust at 48C or over 6 months at K208C, all the compounds
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 7–12 2:131
LC–MS/MS
Table3 Limitsofdetection(LoDs)andlowerlimitsofquantification(LLoQs).
Andro- Pro-
Cortisol Cortisone Prednisolone Prednisone DXM 11DOC Testosterone stenedione 17OHP gesterone
LoD(nmol/l) !0.1 !0.1 !0.2 !0.2 !0.03 !0.03 !0.01 !0.02 !0.06 !0.06
LoQ(nmol/l) 1.95 1.58 0.49 0.97 0.061 0.098 0.018 0.122 0.244 0.122
CV(%) 11.4 6.7 12.3 18.3 5.4 17.3 18.8 16.9 14.2 16.9
Accuracy(%) 94.6 117.9 86.0 103.6 112.0 103.5 116.2 103.8 95.9 113.4
LoDisdefinedasasignal-to-noiseratioabove3.LoQisdefinedasthelowestlevelthatcouldbemeasuredwithaccuracywithinG20%ofthenominallevels
andaCVbelow20%(nZ6).AttheLLoQ,thesignal-to-noiseratiowasabove10forallthecompoundsinallthereplicates.DXM,dexamethasone;11DOC,
11-deoxycortisol;17OHP,17a-hydroxyprogesterone.
werestillmeasuredwithin G10%oftheirnominallevels middle-agedgroup(P!0.00001),whichagainwerelower
withnoapparenttrends.Thecompoundswererobusttoat than those of the young group (PZ0.006). For BDs,
least five freeze–thaw cycles. Cortisone concentrations androstenedione concentrations were not significantly
were within G10% of the nominal levels after 48h of different in the middle-aged group and the old group
storage at ambient temperature; however, for samples (PZ0.160).Theyounggrouphadhigherlevelsrelativeto
containinglevelsof50nmol/,concentrationsdeclinedto the middle-aged group (PZ0.0002) and the old group
86% after 96h. Similarly, at this time point, the (PZ0.00007).
concentrations of samples with prednisone at an initial On comparing women with AD with BDs, it was
level of 50nmol/l had declined to 84%. Reconstituted found that testosterone levels were significantly lower in
ons sampleswerestableforatleast4daysat48C. all the age-stratified groups (all P%0.00020). Similarly,
cti androstenedione concentrations were lowest in AD
e
nn patients (all P%0.00005). The levels of 17OHP in the old
Co Application:steroidprofilesofwomenwithADandBDs groups were lower in AD patients than in the BDs
e
rin The steroid profiles of patients with AD and BDs are (P!0.00001). There were no systematic differences in
c
o serum cortisol and cortisone levels between AD patients
d reportedinTable4.Themethodprovidedinformationon
n
E and BDs, as expected since the timing of specimen
the cortisol and cortisone levels of patients on oral
collectionwasnotstandardized.
hydrocortisone replacement therapy and successfully
identifiedthefivepatientstakingprednisolone(predniso-
lone317nmol/l,range220–419;prednisone47.8nmol/l,
range17.2–86).NineADpatientshadlowlevelsof11DOC Discussion
(median0.20;range0.14–0.32nmol/l).
WedevelopedamultiplexedLC–MS/MSassaythattargets
Testosterone and androstenedione concentrations
the work-up of adrenal and gonadal disorders. As
were above the LLoQs in all the BDs and most of the
demonstrated by the steroid profiles of women with AD,
womenwithadrenalinsufficiency.ThetwoADgroupsof
itallowsthequantificationofverylowlevelsofandrogens,
women aged above 60 years and women with premature
as well as measurements of endogenous and synthetic
ovarianfailurestoodoutasinthemandrogenlevelswere
glucocorticoids. The method is automated, requires only
often not even detectable. Generally, patients with AD 85ml serum and has a chromatographic run time of
exhibited decreasing serum testosterone concentrations
6.1min, which makes it suitable for use in routine
with age, and levels were significantly lower in the old endocrinelaboratoriesaswellasinresearch.
group vs the young group (PZ0.0195). This was in The panel of analytes includes the most commonly
contrast to the testosterone levels in BDs, where no systemicallyusedsyntheticglucocorticoids.Thiscouldbe
differencebetweentheoldandyounggroupswasobserved particularly useful in a clinical setting as it enables a
(PZ1). The middle-aged BDs, however, had lower levels comprehensive assessment of the patient. For example,
thantheyounggroup(PZ0.003). the overnight low-dose DXM suppression test is com-
Serum androstenedione concentrations also declined monly used in the work-up of CS, but its diagnostic
with age in women with AD. Androstenedione concen- specificity of only 80% is a major drawback (16). The
trations in the old group were lower than those in the determination of serum DXM could help in identifying
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 8–12 2:132
LC–MS/MS
100% 100% 100% 100% 100%
Cortisol Cortisone Prednisolone Prednisone 11DOC
90% m/z 407/297 90% m/z 405/301 90% m/z 405/295 90% m/z 403/299 90% m/z 347/97
Period I Period I Period I Period I Period 2
80% 80% 80% 80% 80%
70% 70% 70% 70% 70%
60% 60% 60% 60% 60%
50% 50% 50% 50% 50%
40% 40% 40% 40% 40%
30% 30% 30% 30% 30%
20% 20% 20% 20% 20%
10% 10% 10% 10% 10%
0% 0% 0% 0% 0%
1.2 1.4 1.6 1.2 1.4 1.6 1.2 1.4 1.6 1.2 1.4 1.6 1.8 2.0 2.2
100% DXM 100% Testosterone 100% 100% 17OHP 100% Progesterone
Androstenedione
m/z 437/361 m/z 289/97 m/z 331/97 m/z 315/97
90% 90% 90% m/z 287/97 90% 90%
Period 3 Period 4 Period 6 Period 7
Period 5
80% 80% 80% 80% 80%
70% 70% 70% 70% 70%
60% 60% 60% 60% 60%
50% 50% 50% 50% 50%
40% 40% 40% 40% 40%
30% 30% 30% 30% 30%
s
n
o 20% 20% 20% 20% 20%
ti
c
e 10% 10% 10% 10% 10%
n
n
o 0% 0% 0% 0% 0%
C 2.4 2.6 2.8 3.0 3.2 3.4 3.5 3.6 3.7 3.8 3.9 4.0 4.5 5.0
e
n
ri
c
o Figure1
d
En MSchromatogramsofthequantifiermasstransitionofthelowestworking 0.06nmol/l,dexamethasone(DXM);0.02nmol/l,testosterone;0.12nmol/l,
calibratorforeachanalyte.Thenominalconcentrationswereasfollows: androstenedione;0.24nmol/l,17a-OH-progesterone(17OHP);and
1.95nmol/l,cortisol;0.98nmol/l,cortisone;0.98nmol/l,prednisolone; 0.24nmol/l;progesterone.m/z,mass-to-chargeratiooftheionized
0.98nmol/l,prednisone;0.10nmol/l,11-deoxycortisol(11DOC); compoundandfragment.
patients with false-positive tests due to reduced gastro- Our LC–MS/MS assay allows the quantification of
intestinal DXM absorption or abnormal DXM metab- serum testosterone in the picomolar range found in
olism.ThisstrategyisindeedsuggestedbytheEndocrine women and children, which traditional immunoassays
Society (17), but rarely implemented. Moreover, the cannot measure reliably (21). We achieved a LLoQ for
detection of synthetic glucocorticoid hormones could testosteroneoflessthanone-fifthofthatreportedbyGuo
revealiatrogeniccausesofCSaswellaspreventunnecess- etal.(2)andCeglareketal.(3),andfunctionalsensitivity
ary investigations in patients with iatrogenic cortisol wassimilartothatofmethodsemployinglabour-intensive
suppression, both circumstances commonly encountered derivatization procedures to enhance the ionization
in the routine laboratory. Moreover, prednisolone and of androgens (5, 8). The LLoQs in the range of
prednisone are commonly used as immunosuppressants 0.02–1.97nmol/l are, for most of the analytes, lower
andinthetreatmentofcancer.Itisaclinicalobservation than those of previously published LC–MS/MS methods,
that these drugs show considerable inter-individual althoughourmethodrequireslessserumvolume(1,2,3,
variation in dose required for therapeutic effect. Analo- 4,5,7,8).Combinedwithsimultaneousdeterminationof
gously,somepatientsaremoreresistanttoadverseeffects. multiplehormones,lowserumvolumerequirementisan
Therapeutic monitoring of prednisolone and prednisone advantage when the specimen is scarce, such as samples
levels may ultimately optimize clinical response and from children and population-based biobanks. The low
reduceadverseeffects,suchasosteoporosisandcardiovas- sample volume used also facilitates high-throughput
culardisease(18,19,20). analysis by allowing unattended automation of the
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 9–12 2:133
LC–MS/MS
Cortisol Androstenedione Testosterone (males) Testosterone (females)
800 RSl o=p 0e. 9=3 1 .04 (0.94, 1.4) 25 RSl o=p 0e. 7=8 12. 96 (1.72, 2.26) 50 RSl o=p 0e. 9=5 05. 78 (0.75, 0.81) 3.0 RSl o=p 0e. 5=6 04.97 (0.74, 1.3)
Intercept = 28.65 (–6.88, 64.25) Intercept = –0.35 (–1.32, 0.48) Intercept = 0.43 (0.33, 0.56) Intercept = 0.24 (–0.13, 0.45)
mulite 2000 (nmol/l) 640000 n = 57 mulite 2000 (nmol/l) 211050 n = 136 mulite 2000 (nmol/l) 432000 n = 265 mulite 2000 (nmol/l) 22..50 n = 86
m m m m 1.5
DPC i 200 DPC i 5 DPC i 10 DPC i 1.0
0
0 200 400 600 800 2 4 6 8 10 0 10 20 30 40 0.5 1.0 1.5 2.0 2.5 3.0 3.5
LC–MS/MS (nmol/l) LC–MS/MS (nmol/l) LC–MS/MS (nmol/l) LC–MS/MS (nmol/l)
100 100
1500
300
% difference DPC immulite2000 vs LC–MS/MS1500000 % difference DPC immulite2000 vs LC–MS/MS 210000 % difference DPC immulite2000 vs LC–MS/MS –55000 % difference DPC immulite2000 vs LC–MS/MS –55000
0
0 –100 –100
0 200 400 600 800 2 4 6 8 10 0 10 20 30 40 0.5 1.0 1.5 2.0 2.5 3.0 3.5
LC–MS/MS (nmol/l) LC–MS/MS (nmol/l) LC–MS/MS (nmol/l) LC–MS/MS (nmol/l)
11DOC 17OHP Progesterone
80 RSl o=p 0e. 6=6 36. 47 (2.59, 4.31) 150 35 RSl o=p 0e. 9=0 06. 95 (0.85, 1.04)
nections RIA, Biosource (nmol/l) 642000 Inn =te r7c9ept = 2.11 (1.48, 2.74) RIA, Coat-A-Count (nmol/l) 10500 RSInn l t=o=e p r60ce7.e 8=p8 t23 =. 6 02. 6(23. 3(01.,3 27.,9 19.)02) DPC immulite 2000 (nmol/l) 22311500505 Inn t=e r4c4ept = 0.69 (0.6, 0.86)
on 0 0 0
C 0 5 10 15 0 20 40 60 80 100 120 140 0 5 10 15 20 25 30
e LC–MS/MS (nmol/l) LC–MS/MS (nmol/l) LC–MS/MS (nmol/l)
n
ri 1000
c
Endo % difference RIA,Biosource vs LC–MS/MS 846200000000 % difference RIACoat-A-Count vs LC–MS/MS 264000000 % difference DPC immulite2000 vs LC–MS/MS432100000000
0
0 0
0 5 10 15 0.2 0.5 1.02.0 5.010.020.0 50.0 0 5 10 15 20 25 30
LC–MS/MS (nmol/l) Log LC–MS/MS (nmol/l) LC–MS/MS (nmol/l)
Figure2
ComparisonoftheLC–MS/MSmethodwithimmunoassaysavailableatthe Arrowsincortisolplotsdenoteapatientputonprednisolonetherapyin
HormoneLaboratory.Statisticalanalyseswerecarriedoutusingthe whomtheimmunoassayreportedfalselyelevatedcortisollevels
Passing–Bablokregression(upper)andBland–Altmanplots(lower). (220vs12nmol/l).
extractionprocedureinastandard96-wellformatusinga contributedtothe longlifetimeofthe analyticalcolumn,
liquid-handling robot. Within an 8-h workday, 288 whichtypicallywithstoodthousandsofinjections.
samplescanbeprepared,andmorethan200samplesper AlthoughMSisahighlyspecificanalyticaldetector,it
daycanbeanalysed. is necessary to chromatographically separate the steroid
When hormones with different chemical properties analytes from isobaric isomers and other interfering
are assayed simultaneously, there is a balance between compounds.WefoundthattheZorbaxRRHDC18column
obtaining high recoveries for all the compounds and performed well for both glucocorticoids and androgens.
eliminating substances that cause ion suppression or A reversed-phase column with phenyl-bonded stationary
impactmethodruggedness.Overall,extractionwithethyl phase may better separate cortisol, cortisone, predniso-
acetate:hexaneyieldedhighrecoverieswithoutsignificant lone and prednisone (22), but is incapable of resolving
ion suppression or enhancement. The clean extracts androgens from interfering compounds. Because the
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access
Research PMethlieetal. Quantitationofsteroidsby 10–12 2:134
LC–MS/MS
above60yearsZ(n13) Median(IQR) 66(62–68) 256(175–303) 42.1(40.6–51.3) NA NA NA 0.32(0.26–0.45) 0.65(0.56–1.10) 1.28(0.96–1.64) 0.31(0.25–0.63) NA ementtherapy; gipcnrlruoecsdtsohn-cieinosotrnetnirecefgeoariaedtnnisvdcfeeo.primroeneddinziassptoieolconinfiecmawoddadseu,cwttsihtwheoitcuhot-feoslirugmntiiiocfincaacnoidtf
Age !LoQ 0 0 13 13 13 0 0 0 3 13 eplac instrTuhmeeinnctsreahsainsgisnecnrseiatsiveidtyaowfathreenleatsesstagnednecroatniocenrnofMfoSr
r
Blooddonors ZAge40–59years(n41) !LoQMedian(IQR) 48(45–55) 0211(153–278) 041.5(35.6–49.4) 41NA 41NA 41NA 10.29(0.22–0.48) 00.56(0.43–0.67) 01.64(1.18–2.21) 70.68(0.35–1.86) 230(0–3.52) failure;DHEA,DHEA irfppsraeeaoncagbgtkmaeedra.eircndeMtaesisvo)ninstaaheitdtriooofduernsrlidneaongnfrcaotetithsswhe,eoesrirunamltsaipopyaiescaosirroftionttcrfhuaoenlsfapsqriei,ntucitaioiefinnnrctfsiietfitryhe(eienr.aecst.epoaic(cqn2oou3lmula,ilnsoi2ifieol4axen)rr-.
n Although monitoring two mass transitions reduces scan
Agebelow40yearsZ(n48) !LoQMedian(IQR) 28(25–32) 0290(218–407) 044.8(37.3–53) 48NA 48NA 48NA 00.30(0.20–0.61) 00.71(0.56–0.86) 02.47(1.87–2.81) 60.68(0.44–1.38) 230.249(0–5.82) OF,prematureovariaetected. trawoientmechaoeelbmyreasmlminised.uevOnletthdiusiuestredsaars(osn2esina5diy)msmiotmpirevotoaihrtdntyoavitdniaossnterdds(i2mtaL,wspL3omor,oQ4amv,,ne5atmdh,sas6eitsn)ot.rstratyrnca(os2timet6gi)opyfnaohrsre,adsswtebwhreoiiectinhhd
Pd
ZHEA(n7) Median(IQR) 51(46–54) 133(83.6–289) 40.6(25.2–52.5) NA NA NA NA 0.56(0.47–0.95) 1.74(0.87–1.85) 0(0–0.78) 0(0–0.78) uantification;monewasnot psptoreotretoIeninidtntiseasirlna-liynntdhidmeilvipspiedairdcuutsam.lmdAmeinfafaesiyusroervnetaomcrpeyesicnwintaistns.heaFrtlouohrmgeueacxemoaommofupptnohletse,iotatifhonbneailncEyodRteuinolidgsf
EndocrineConnections ofilesoffemaleswithAddison’sdiseaseandinfemaleblooddonors. PatientswithAddison’sdisease below40yearsAgeabove60yearsZZZZ(n25)Age40–59years(n71)(n43)POF(n10)D !!!!Median(IQR)LoQMedian(IQR)LoQMedian(IQR)LoQMedian(IQR)LoQ 35(30–38)48(44–52)68(64–76)56(53–62) 326(116–540)2273(102–452)0290(149–524)0404(163–554)1 58.5(27.5–67.5)245.1(24.9–59.6)151(34.7–65.7)053.5(41.6–61.1)1 NA690(0–0)410(0–0)90(0–0)7 NA690(0–0)410(0–0)90(0–0)7 NA71NA43NA10NA7 0(0–0)660(0–0)420(0–0)10NA7 0.44(0.22–0.63)130.26(0.12–0.45)160.15(0–0.41)50.03(0–0.27)0 1.18(0.71–1.76)190.63(0–10.2)380(0–0)60(0–0.16)1 0.65(0.37–1.59)260.41(0–1.21)400(0–0)80(0–0)4 0(0–2.72)430(0–1.96)43NA10NA5 !asmedian(IRQ,interquartilerange)nmol/l.LoQ,numberofsamplesbelowthelimitofqa11DOC,11-deoxycortisol;17OHP,17-hydroxyprogesterone,NA,notapplicablebecausehor cNverpaTvaciiswsmfiidAnndetyqountaahrnvDeriledenedlamurrtdaeeeivints.hviilistdvtirehovslvTHIohlmTeetpeaauiSr(itiuiohsdnbfoedt1neatdfihroomhdsitrertee,utueoooinrbwieie2ucaedeaantLastuiycllc)e,olnnymeoeylnC.olrsrogdsiac,tbTddeoman,kstslfoiuwywh,uhisisffinirfpmfnssstpcoipfecieeeBeoeretteradtoontrlrechpdDeswrhailhiereincnossdnacetanpestaopluingerapnsacpes.hvatcrrittntsrssdoTaitnrteeeediheeteishlietgtrsdshhgeolacfaMasitdteeuonteisoeenpltenyn,eseodhnflEniioaitnmictdfdsnadnsoaetolsthsoIoteys,steortseSinuerrrailsootwir.esooearitbsenuiintszsthosnnhnss.ldhmuomaooueeotnAeidettleulrcrgafptietuniaeachcodshalncrtrstumlotooaehnnoayoieawdhococstpmmtfiotbdrehsiofeoeaenlusdttalrntaqppfyelceaehomfiyltcmrtsuetooosnieihgescenhsncccrasfeioemefiatiooeebotnestesntiitferdhafnnfisi-icrhteensdeotepwioeecrcrwneofilnddnosinoeyniuotdcntdhcfi.rbottodfnormbaoneteetoutottwomotwyghfnn,hresiiemgpnooetttaixonpiecuehinsrannmvsaafarrnlautervaetecwdllttopocidrepsethhiobaipahiyltncuoefsyteyner-iffeutimhgsfhea.nniawrw(IllagtrpiaTf2SaslototiegAteyaeiinoerr7hhestnvnhtcrtlteDrshn)ddoyeeeeeett----..
teroidpr Age !LoQ 0 0 25 25 25 22 2 ne2 5 16 expressedethasone; tlhaeirsggthoelsytceosrphoeoncreitfiocanfMdwSo.amTnodenrdoisfwfteeirtnehendatiidoartneeenbaecltowinnecseuennffitpcrraieetminocenynsouipsnianuga-
S o em
Table4 Age(years) Cortisol Cortisone Prednisolone Prednisone DXM 11DOC Testosterone Androstenedi 17OHP Progesterone AlllevelsarDXM,dexa yssauelbarjaesn)ctdsanpidnosttoomldyeon(uoOnp6gau0s(!ayle4as0tras)tyuegsa,rrosw)u,epsms.tirdTadhtilefie-eamdgietddhdel(e4s-0atug–d6ed0y
http://www.endocrineconnections.org (cid:1)2013Theauthors ThisworkislicensedunderaCreativeCommons
DOI:10.1530/EC-13-0023 PublishedbyBioscientificaLtd Attribution3.0UnportedLicense.
Downloaded from Bioscientifica.com at 02/12/2023 12:07:42AM
via free access