Tandem Mass Spectrometry - CiteSeerX

Journalof AnalyticalToxicology,Vol, 22, July/August1998

FastScreeningfor Drugsof Abuseby Solid-Phase ExtractionCombinedwith Flow-Injection IonsprayTandem Mass Spectrometry Wolfgang Weinmann* Institute of Forensic Medicine, Klinikum der Albert-Ludwigs-Universit~t Freiburg i. Br., Albertstrasse 9, D-79104 Freibur& Germany

Michal Svoboda Perkin-Elmer Applied Biosystems, PauI-Ehflich-Str. 17, D-63225 Langen, Germany

I Ahst,act I A fast analytical approach for the simultaneous quantitative screening for illicit drugs in serum and urine without tedious chromatographic separation stepswas developed by combining solid-phase extraction (SPE)followed by flow-injection analysis (FIA) with ionspray-ionization and tandem massspectrometry (MS-MS) detection using a PESciex API 300 triple-quadrupole MS. MS-MS analysiswas performed by sequentially isolating the precursor ions of the analytes and their deuterated standards with subsequent fragmentation and monitoring of one fragment ion for each substance.A multiple-reaction monitoring experiment was set up for morphine (MO), codeine (COD), amphetamine (AMP), benzoylecgonine (BZE), and their deuterated analogues. For method evaluation, serum samples spiked with 2-1000 ng of each drug and deuterated standards were extracted by mixed-mode SPE, redissolved in CH3CN-NH4OAc-buffer, and directly injected by flow injection into the ionspray source. The specificity of this new method was demonstrated by testing compounds with similar chemical structure for interferences from the analytes of interest (e.g., dihydromorphone, morphine glucuronide, and 6-monoacetylmorphine with MO; dihydrocodeine and hydrocodone with COD; cocaine [COC] and ecgonine methylester with BZE;methamphetamine with AMP). The possibility of interferences of such compounds with the FIA-ionspray-MS-MS screening method is discussed. Spiked serum samples and serum and urine samplesfrom drug addicts and victims of drug abuse were ana/yzed with F/A-MS-MS and, after derivafization, with gas chromatography-mass spectrometry (GC-MS). Comparable quantitative resultswere obtained with both methods; no interferences with metabolites or other compounds were found. The FIA-ionspray-MS-MSmethod is a fast, quantitative, sensitive, and highly specific alternative method to drug-screening by immunoassays, high-performance liquid chromatography, and 9Aulhor to whom correspondence should be addressed: Dr. rer. nat. Wolfgang Weir~mann, Inslilule of Forensic Medicine, Klinikum der AlberbLudwigs-Universit~t, Albenstr 9, D-79104 Freiburg, Germany, e-maJJ: weinmann@sun'~ 1.u kl.uni-treiburg.de.

GC-MS. It can be used for the simultaneous detection of different drugs and metabolites such as opiates, COC, AMP derivatives, and many other drugs.

Introduction /onspray and etectrospraymass spectrometry (MS) (1,2) have been used in many different fields for the analysis of organic compounds such as drugs (3,4), explosives(5), pesticides (6), and natural products (7), as well as for the characterization of biomacromolecules such as proteins and oligonucleotides (8--10). For mixture analysis, electrospray ionization is usually coupled with a separation technique such as high-performance liquid chromatography (HPLC) or capillary zone electrophoresis prior to MS analysis (11,12). Combinations of liquid chromatography, electrospray-ionization, and tandem mass spectrometry (LC-MS-MS) have recently been used for the identification and quantitation of polar drugs and their metabolites (e.g., LSD, opiates, and cocaine [COC]) in body fluids and hair samples (4,13). In contrast to many gas chromatography-mass spectrometry (GC-MS) procedures using electron impact or chemical ionization, no derivatizationof polar compounds is required before MS analysis when using ionspray- or electrospray-ionization MS. This technique is amenable to polar drugs dissolved in salt-free, volatile solvents such as acetonitrile, methanol, water, acetic acid, or ammonium acetate solutions. The high efficiency of the ionization process, resulting from the formation of primarily protonated molecular ions ([M+H]+) from organic compounds in positive ion mode or deprotonated molecular ions ([M-HI-) in negative ion mode, leads to a highly sensitive MS method (14). In most forensic laboratories, qualitative and quantitative

Reproduction(photocopying)of editorialcontentof thisjournalisprohibitedwithoutpublisher'spermission.

319

Journal of Analytical Toxicology, Vol, 22, July/August 1998

determinations of drugs of abuse (e.g., opiates, COC, amphetamine [AMP] derivatives, cannabinoids, and LSD) from serum, urine, and hair samples are performed by GC-MS analysis (15-20). However,several sample preparation steps, such as extraction and derivatization of involatile or thermally labile compounds, are necessary before analysiswith GC-MS. In some forensic applications, MS-MS has been used in combination with GC to enhance specificity of a GC-MS experiment by adding a further MS dimension (21-23). For quantitative analyses of illicit drugs in urine, serum, and hair samples, which are usually performed by selected ion monitoring with GC-MS, one could think about the exchange of one analytical dimension against another, that is, the separation in time (chromatography) against the separation in molecular weight (MS) (24). If, for target analytes to be quantitated in a mixture, deuterated standards are available, an MS-MS experiment using multiple-reaction monitoring (MRM) to switch from one compound to another with short dwell times can replace the GC separation step of a GC-MS experiment performed in selected ion monitoring (SIM) mode. A prerequisite for this approach is the homogeneity of the sample. Matrix effectswhich may interfere with the ionization process have to be taken into consideration, hence internal standards have to be employed. Interferences between substances with structural similarity have to be investigated. In combination with solid-phase extraction (SPE) for sample cleanup and preconcentration, and with the use of deuterated internal standards, we developedan ionspray-MS-MSmethod for the quantitative screening for morphine (MO), codeine (COD), benzoylecgonine (BZE),and AMP in blood serum and urine.

Experimental Reagents Morphine-d3, codeine-d3, benzoylecgonine-d3, D,L-amphetamine-d5 (deuterium labeled on the side chain), and all other drugs were purchased from Promochem/Radian (Wesel, Germany) or Sigma (Deisenhofen,Germany). All solvents were of high purity (p.a. or HPLC grade). Phosphate buffer consisted of 0.1M NaH2PO4adjusted to pH 6 by addition of 1M NaOH. Solid-phase extraction Mixed-modeSPE has been successfullyapplied to the analysis of serum samples with trace amounts of opiates, AMP,and COC metabolites (25). The extraction of spiked serum samples and real-case samples was performed by a manual SPE method validated in-house (20) or by an automated SPE device (RapidTrace| Zymark GmbH, Idstein, Germany) (26). For all extractions, mixed-mode SPE cartridges (Chromabond Drug, 3 mL/200 mg, Macherey-Nagel, Dtiren, Germany) were used. For method evaluation, 1-mL serum samples spiked with deuterated standards (25 ng in 50 IJL methanol) were diluted with 1 mL phosphate buffer (pH 6) before SPE. Aliquots of real-case samples (serum or urine, 0.1-2 mL) were spiked with deuterated standard mixture (200 ng each) and diluted with phosphate buffer. After applying the samples on the precondi-

320

tioned cartridges (methanol and phosphate buffer), extraction steps were as follows:washing with 2 mL deionizedwater, 1 mL 0.1M acetic acid, and 2 mL methanol; drying; and eluting with 1.5 mL of dichloromethane/2-propanol/25% NH4OH (80:20:2, v/v). The eluate of spiked serum samples and real-case samples was divided in two equal portions (one for flow-injection analysis-MS-MS [FIA-MS-MS] analysis and the other one for GC-MS analysis) and evaporated to dryness. The recoveries of this SPE method were determined by GC-MS analysis in previous work and were 92% for AMP,86% for BZE, 88% for COD, and 87% for MO (25). FIA-ionspray-MS--MSanalysis An AP1300 triple-quadrupole MSwas used with an ionspray source (PE Sciex, Perkin-Elmer Sciex Instruments, PerkinElmer Applied Biosystems, Langen, Germany) and a series 200 HPLC pump with autosampler (Perkin-Elmer). FIA was performed at a flow-rate of 50 laL/min.The SPE eluate was redissolved in 100 IJL acetonitrile/10mM ammonium acetate (1:1, v/v); aliquots of 5 or 10 IJL were injected for each analysis. The orifice voltage was 20 V, and the fragmentation energy was set to 35 V. GC-MS analysis GC-MS analysis was performed after a two-step derivatization. For trifluoroacetylation, the evaporated residue was incubated with 100 IJL trifluoroacetic anhydride (60~ for 30 rain), and for esterification of BZE, 100 IJL 2,2,3,3,3-pentafluoropropanol was added to the reaction mixture and incubated again (60~ for 30 rain). For quantitation, an MD800/250 GC-MS instrument (CE Instruments) was used in SIM mode (20). Opiates (MO, COD,dihydrocodeine [DHC],AMP,methamphetamine, and 3,4-methylenedioxyamphetamine derivatives [methylenedioxyarnphetamine, methylenedioxymethamphetamine, and methylenedioxyethylamphetamine]) and ecgonine methylester (EME)were registeredas trifluoroacetylated derivatives; benzoylecgonine was detected as its pentafluorpropionylester (19). Serum and urine samples were analyzed for drugs of abuse semi-quantitatively by HPLC (BioRadRemedi| Munich, Germany) or by radioimmunoassay (DPC Biermann, Bad Nauheim, Germany).

Results A combination of SPE for sample cleanup and injection of the extracts directly into an ionspray-MS-MS was developed for the simultaneous screening for MO,COD, BZE,AMP,and some re]ated drugs and metabolites. For SPE, a mixed-mode cartridge (RP-C8 and cation exchanger) that showed good recovery (> 86%) and reproducibility for basic drugs in previous work (25) was used. The FIA-ionspray-MS-MSmethod used for analyzing the extracts is based on the high efficiencyof the ionspray-ionization process for polar substances and the extremely high specificity of MS-MS, which uses the fragmentation of precursor ions to fragment ions within a triple-quadrupole MS (Figure 1). The

Journal of Analytical Toxicology, Vol. 22, July/August 1998

separation of compounds with different molecular weights occurs in the first quadrupole (molecular weight selectivity) by the selection of a precursor ion (Q1). The identification is performed by the isolation of a fragment ion (Q3) after collision-

ally induced dissociation (CID) of the precursor ion in the second quadrupole (Q2) (structure selectivity). For quantitation, internal stable-isotope standards are used. The resolving power of the quadrupole MS was set to 1 atomic mass unit,

Flow-injection-ionspray-MS-MS Curtain Plate

Colllslon quadrupole Sample loop

Orfflce Skimmer

Autosampler

.,

~

=

5-p.Lsample

I +SkVl " "

t

I)-[-.

IQI

I I--':

ilC-II

,I

Q1

i I| fragment ion) used for the MRM experiment are emphasized by arrows.

321

Journal of Analytical Toxicology, Vol. 22, July/August1998

which is sufficient in most cases. However, interferences of structurally similar compounds may occur when two compounds have the same molecular weight or when fragment ions form in the ion source during the ionization process. In the case of compounds having the same molecular weight (and therefore the same precursor ion) as the target analyte, such as hydromorphone and MO or hydrocodone and COD, MS-MS

Ionspray-Mass

spectra of these related compounds were acquired, and the MRM experiment was set up using analyte-specific fragment ions (Figures 2 and 3 and Table I). In the case of fragmentation, certain conditions could lead to a precursor ion with the same molecular weight as a target compound as a result of the loss of a functional group (e.g., loss of an acetylgroup of 6-monoacetylmorphine [6-MAM] yields MO). Under standard

Spectra

MS-MS-Spectra

D

MH +

124.2

of MH +

93.0

141.1

1800@

Z" i --

12000,

i

93.1

_c 4000

6000,

I

Amphetlmine-d s 8000

/

~

oo

8.0

123.9

18o

1~o

m/z, amu

1;o

0'0

8'0

i

1;o

" i

120

E

1.25e5 1.2n,5.

Will + 3o0.1

1 2e6-

IO

Codeine

6.0eS-

I

_r

89

6004'

214.9 3.0e4'

30e5149.1 l

7'0

11o

m/z, amu

,92., J ~24.; ...... JL,I . _ .28:.9

215.0 243.1

l;o

%" 140

7'0

l~o

m/z,

MH +

1.596-

,

i

210

280

ainu

303. 'I

303.0

Codeine.d3

' • "1.096. P

.=.

--= 5.09! 149.1 ]

'~

1.4o9

n

7.0o4-

215,0

,L = 80

180

1 696-

.

32o

240

m/z, amu

G

MH §

.

'7'0

.

.

,;0

.

..... .9 ,=L

2,f,B

9 i

210

280

m/z, amu

286.0

285.9

1.2e68,005

r

4 0oS

Morphine " ~ 8.094

o

99.0 '

70

149.1

4.0e4

/165,o

140

201,0 ~211,1

9

l

210

.I

7'0

280

m/z, amu

H 1.5115'

cl~

300.0

9.005.

'i

140

m/z, amu

,._,JJ.JJ J

1;o

.,

2,0

288.1 ,L . 280

tlrl/z, ainu

,o~o.

MH+ 280,0

289.0

Mo~hine-d 3

, • 1.096,

J

i

= 5.o~5.

149.1 99.0 I 165,0 a,...L . .i . . . . . L I . ~ ! ..... k .[ . . . . 70

140

m/z, amu

8,0e4' e..

--

~._ . . . . . , 210

....

h.... 280

201.0 210.9

4.0e4'

. . . . . . 7'0

,JI,Ji..J ' 140

m/z, ainu

' 210

..

2:,i.2 280

Figure 2. (continued) Full-scan ionspray mass spectra and MS-MS spectra of the protonated molecular ions of amphetamine, benzoylecgonine, morphine, and codeine. The transitions (molecular ion ~ fragment ion) used for the MRM experiment are emphasized by arrows.

322


direct infusion (Figure 2). These mass spectra were used to create an MRM method for all eight substances by monitoring eight transitions (MH§ ~ F§ with dwell times of 150 ms each (Table I). Different degrees of fragmentation were obtained for amphetamine-ds, which is deuterated at the propyl side chain, and non-deuterated amphetamine. The deuteration of the side chain greatly influences the fragmentation process. This was observed recently in the fragmentation of perfluoropropionylated amphetamine-ds by GC-MS (25). Thus the fragment m/z 93 of amphetamine-ds is more intense than the fragment m/z

conditions, the occurrence of fragmentation in the ionspray source is very low. However, it can be promoted by increasing the orifice voltage. To check if orifice-induced fragmentation took place in the ionspray source under the conditions used for the experiments, full-scan mass spectra of related compounds were acquired and are shown in detail (Figure 3). Setup of an MRM experiment Full-scan mass spectra of standard substances and MS-MS spectra of the protonated molecular ions (MH*) of MO, COD, AMP, BZE, and their deuterated analogues were acquired by

C Morphine-6-Glucuronide

B Hydrocodone

A Hydromorphone

y 242

. I::

.

.

I~

.

,

~

300

-,-

.8

.Q

.Q

_o

_o|

187

:. . . . . . . .

242

[.

=

.l

g,

.

. . . . .

,

9

2O0

J,I, d L L L ~ , . =

300

- - -

,

400

1:1 m

300

186 286

/ 1 ~ 227 213 24t . .

=

.

.............

.

2~ m/z, amu

m/z, amu

.L,.

m/z, amu

F Diacetylmorphine

E 6-Monoacetylmorphine

D Morphine-3-Glucuronide

[;

3o0

,i

|

i

200

3O(]

" 9 400

328 --

J ~

,8 r 3 "~ m

. . . . . 100

,

9 3;0

200

i| , 400 370

100

211

1~o

,,

J

, i

2~ 3o0 m/z, amu

, l

~o

ill

2tl '-

m/z, ainu

z

~

. l

.

1~

2~8

328

l

2~o ' m/z, amu

40

Figure 3. Full-scan ionspray mass spectra and MS-MS spectra of opiates (3A-3G), which could interfere with the quantitation of morphine and codeine; methamphetamine (3H), which could interfere with the quantitation of amphetamine; and cocaine and ecgonine methylester (31 and 3J), which could interfere with the quantitation of benzoylecgonine.

323


91 of the non-deuterated AMP (Figure 4). In contrast, similar degrees of fragmentation were obtained for the deuterated and non-deuterated analogues of MO and for BZE and COD. This is also reflected in a slope of approximately 1 in the linear range (Figure 5). For calibration, serum samples spiked with drug mixtures of 0, 10, 25, 50, 100, 250, 500, and 1000 ng/mL and 100 ng/mL of deuterated standards each were analyzed. Linearity of calibration was obtained in this concentration range with high correlation factors (P-_>0.995). For the determination of the limit of

quantitation (LOQ) and the limit of detection (LOD),analyses of spiked serum samples in the low-concentration range were performed as follows: 10 ng of each deuterated standard was added to each serum sample; the samples were spikedwith 0, 2, 4, 8, 20, or 40 ng/mL of the analytes and then extracted by $PE (Figure 5). After evaporating the eluate, the extract was redissolved in 100 pL acetonitrile/ammonium acetate buffer, and 10 pL was directly injected by flow-injection into the ion source. Signal-to-noise ratios (S/N) at low drug concentration are shown in Table II. Based on these signal-to-noise ratios, LODs

G Dihydrocodeine

H Methamphetamine 3(

|

157 II "D r

11g

18

300

iI

il m

/:

_o

gl

lgg

100

200

loo

300

m/z, amu

2~o

~o

m/z, amu

I Cocaine

J Ecgoninemethylester 3O4

157

242 I

I= m "0

e, i1 "o t-,

c

5o

,~o

2~o

JD

JD tl

182

82

9- -

182 82 150

J .L

3O4

i

1~

2~

m/z, amu

~

5'o

1. m/z,

2~o

amu

Figure 3. (continued) Full-scan ionspray massspectraand MS-MS spectraof opiates (3A-3G), which could interfere with the quantitation of morphine and codeine; methamphetamine (3H), which could interfere with the quantitation of amphetamine; and cocaine and ecgonine methylester (31 and 3J),which could interfere with the quantitation of benzoylecgonine.

324


(at S/N = 3) were between 1 and 4 ng/mL, and LOQs (at S/N = 10) were between 2 and 12 ng/mL. For the determination of the reproducibility of the FIA-ionspray-MS-MS method, one serum extract spiked with 8 ng of each drug and 10 ng of their deuterated analogues was analyzed seven times by SPE and FIA-ionspray-MS--MS.The resuits of the analyses are shown in Table III.

Table I. Mass-to-Charge Ratios Used for the Multiple-Reaction-Monitoring (MRM) Experiment* Amphetamine Benzoylecgonine Codeine Non-deuterated drug Deuterated standard

136 ---* 91 141 ~ 93

290 --> 168 293 ~ 171

300 ~ 215 303 ~ 215

Blank I

x

I ~ .o.d,: .,) 2o,

MO-d$:2lIl>201

12(

t t I I

1201

I

Testing substanceswith structural similarity for interferences

?,

ii

160'

9r

286 ~ 201 289 ~ 201

Sample (4 ng/mL)

A 161

Morphine

I

90~

8s

~0~

I 3(

304

~

leo

/

120,

m

~ AMP-dIl: 141 > e3

18(

l&0,

$0'

/

12(

[ I

0!

;I

~o. ~o..~

AMP: 1.10) I)1

11I(

,

Il(

,I~ ~

1~1> el 3( "x. 1

|

Time min)

Figure 4. Selected ion current of the MS-MS transitions of amphetamine and morphine in blank (A) and spiked sample (B, 4 nglml_ of each drug). Deuterated amphetamine and morphine (10 nglml.) were used as internal standards.

!

Io T |

I Ampl~amlrm

f

/

8/

/

9Benzoyloogonlne

ill

ec o~--+

r -I L

,'.o~g

/

~,Morpl~tne ~

/

1

,+ !

= 2.24OSx+ o . ~

j * "

. /

II 1 I

~ y

y . 1.2e74x + 0.01152

m/

... - "

/

/

"~

...~

,...4

;,5" ; "

y.I.o2x-o.olo7 y . O . M O x + O.OO7S

+/_.: o.5

.

I

.

I

.

I

.

l

.

I

.

I

I

+ ~.s 2 :,..5 3 3.8 4 Ratlo of non-cleutemted to deuterater drug In tmrurn (10 ng deutecmter drugs mdder

I

I

4.5

+

FigureS. Calibration curves obtained by analyzingserumstandardsspiked with amphetamine, benzoylecgonine, morphine, and codeine at low concentrationsbetween 0 and 40 nglmL.

The following substances were tested for interferences with this MRMmethod for MO, COD, AMP, and BZE. Hydromorphone and hydrocodone have the same molecular weights as MO and COD, respectively.All other compounds were analyzed because of their structural similarity with the target analytes. ]onspray-MS and MS-MS spectra of the tested substances are shown in Figure 3. Figures 3A and 3B show full-scan ionspray mass spectra (upper spectra) of hydromorphone and hydrocodone and the MS-MS spectra (lower spectra) of their MH§ ions. Fragment ions of both substances differ from the fragment ions of MO and COD by two masses (e.g., 199 instead of 201, 213 instead of 215; see Figure 2 for comparison). Therefore, no interference with the quantitation of MO (transition 286 -4 201) and COD (transition 300 -4 215) can occur when monitoring these fragment ions. In Figures 3C-3G, full-scan ionspray-MS and MS-MS spectra of other related opiates are shown. From the full-scan mass spectra, which are acquired with the same orifice voltage as used for the MRM method, the following conclusions can be drawn: morphine-6-glucuronide (M6G) (Figure 3C) shows only minor fragmentation to MO (m/z 286) in the ion source and does not severely interfere with the MRM method for MO; morphine-3-glucuronide (M3G), 6-MAM, and heroin (Figures 3D-3F) do not interfere with the transition of MO (286 -4 201); and DHC (Figure 3G) does not interfere with any other opiate tested here. Figure 3H shows a full-scan ionspray mass spectrum of methamphetamine and the MS-MS spectra of its quasi-molecular ion (m/z 150). Demethylation to AMP in the ion source produces a fragment ion of m/z 135, which is not protonated, thus no interference with the transition of AMP (136 -4 91) can occur (Figure 2). MS and MS-MS spectra of COC and EME are shown in Figures 3I and 3J. No interference of either compound were found for the quantitation of BE (transition 290 -4 168). It can be concluded that the tested substances

325

Journalof AnalyticalToxicology,Vol. 22, July/August1998

show no major interference with the MRM method for MO, COD, AMP, and BZE. Theoretically, the orifice-induced deglucuronidation of M6G to MO in the ion source could produce a minor interference with the MO transition. This is not of practical importance because the intensity of this fragmentation is extremely ]ow (< 5%) and the extraction efficiency of the SPE method used is lower than 12% for the MO glucuronides. M6G Table II. Signal-to-Noise Ratios (S/N) at Low Drug Concentrations Drug concentration S/N AMP* 2 ng/mL 4 ng/mL 8 ng/mL

S/N BZE

S/N COD

S/N MO

8.6 15 32

11 18 41

2.0 3.4 7.6

4.4 6.7 14

* Abbreviations: AMP, amphetamine; BZE, benzoylecgonine; COD, codeine; MO, morphine.

Table III. Reproducibility of the Method Tested by Replicate Analysis of a Serum Sample Spiked with 8 ng of Each Drug and 10 ng of Each Deuterated Analogue Drug

Concentrationfound, mean(n = 7)

SD* (cx)

RSD(%)

AMP BZE COD MO

7.64 ng/mL 8.28 ng/mL 8.02 ng/mL 7.84 ng/mL

0.34 0.33 0.33 0.18

4.3 3.5 4.2 2.3

* Abbreviations: SD, standard deviation; RSD, relative standard deviation; AMP, amphetamine; BZE, benzoylecgonine; COD, codeine; MO, morphine.

Table V. Quantitation of Drugs in Serum and Urine Samples of Drug Addicts: Comparison of FIA-IonsprayMS-MS and GC-MS Sampleand volume Serum 1 1 mL

MO t COD DHC

Urine 1 1 mL

25 ng/mL

27 ng/mLAMP 25 ng/mLBZE 24 ng/mLCOD 23 ng/mLMO


50 ng/mL

44 ng/mLAMP 47 ng/mL BZE 53 ng/mLCOD 47 ng/mLMO






100 ng/mL

200 ng/mL

* Abbreviations: AMP, amphetamine; BZE, benzoylecgonine; COD, codeine; MO, morphine.

326

FIA-MS-MS GC-MS HPLC* (ng/mL) (ng/mL) (ng/mL) 11 38 +

8 11 +

MO DHC M3G or M6G

177 + +

151 + n.a.

160 20000 n.a.

Serum2* 1 mL

MO COD M3G or M6G

4145 98 §

3831 117 n.a.

3600 80 n.a.

Urine 2' 0.2 mL

MO COD MAM M3G or M6G

19174 160 + +

18650 190 + n.a.

15800 190 + n.a.

Serum3* I mL

MO COD

Urine 3* I mL

MO 6-MAM

Serum4* 0.2 mL

MO COD 8ZE EME COC

Urine 4* 0.1 mL

MO COD MAM M3G/M6G BZE COC

Urine 5 0.1 mL

AMP

Table IV. Quantitation of Drugs in Spiked Serum Samples: Comparison of FIA-ionspray-MS-MS and GC-MS

Drugconcentration Drugconcentration Drugconcentration (AMP,* BZE,COD, and MO) determined determinedby spikedto serumsample by FIA-MS-MS GC-MS

Drug

+

820 36

740 46

54 +

76 +

-

1324 1100 1120 + +

1123 1310 942 1140 n.a.

1400 1800 1650 n.a. 800

57370 235314 + + 111830 + 65110

500 RIA n.a.

37500 18000w 2 3 6 9 0 0 116000 + 5000 n.a. n.a. 103780 + 110500 85000 68000

+

* All HPLC results are semi-quantitative BioRad Remedi| Abbreviations: MO, morphine; -, not detected, negative; COD, codeine; DHC, dihydrocodeine; +, positive; M3G, morphine-3-glucuronide; M6G, morphine-6glucuronide; n.a., not determined; MAM, monoacetylmorphine; RIA, radioimmunoassay; 6-MAM, 6-monoacety[morphine; BZE, benzoylecgonine; COC, cocaine; AMP, amphetamine. * Samples 2-4 were from victims of drug abuse. Sample was not diluted; calibration was out of range.

Table Vl. MS-MS Transitions Used for the Qualitative Analysis for Other Drugs in Case Samples (Precursor Ion ~ Fragment Ion) Compound

Transition

6-Monoacetylmorphine Morphine-3-glucuronide/Morphine-6-glucuronide Dihydrocodeine Hydromorphone Hydrocodone Cocaine Ecgoninemethylester Metamphetamine

328 ~ 268 462 ~ 286 302 ~ 199 286 -~ 199 300 -~ 213 304 -~ 182 200 ~ 182 150 ~ 91

Journal of Analytical Toxicology,Vol, 22, July/August1998

and M3G (Figures 3C and 3D) cannot be differentiated by the transition (462 ~ 286), which is a major MS-MS transition of both MO glucuronides, because of identical molecular weight and CID by deglucuronidation. In contrast, 6-MAMand heroin can be identifiedby their specifictransitions (328 ~ 268 and 370 268, respectively). With these preconditions, four solid-phase-extracted spiked serum samples (with 25, 50, 100, and 200 ng/mL certified nondeuterated drug standards AMP, BE, COD, and MO, respectively) as well as real-case serum and urine samples from drug addicts and victims of drug overdosagewere analyzed with SPE and FIA-ionspray-MS-MS and, after derivatization, with GC-MS. An aliquot of sample (2 mL of spiked serum or a smaller aliquot of a real-case sample) was spiked with 200-ng deuterated standards and extracted with SPE. Extracts were separated into two equal portions before evaporation. One portion was analyzedby FIA-MS-MS;the other portion was derivatized by trifluoroacetylation/pentafluoropropionylation and analyzed by GC-MS. Results of both methods are compared in Tables IV and V. Screening analyses for other drugs in realcase sampleswere performed by immunoassays and HPLC. For the qualitative analysis for other drugs in the real-case samples, characteristic transitions for methamphetamine, COC, EME, DHC, M3G/M6G, 6-MAM, hydrocodone, and hydromorphone were added to the MRMmethod (TableVI). Good correlations of the results of both MS methods were found in all samples.

Discussion A new method for the screening for drugs of abuse in serum and urine samples has been developed by the combination of SPE and direct flow-injectionof the underivatized extracts into an ionspray-(triple-quadrupoie)-MS-MS instrument without any further chromatographic separation before MS-MS analysis. First applications of this new method to real-case samples, which also contained other related drugs and metabolites, and the comparison with GC-MS results show that this method can be used for quantitative drug screening. Quantitative results obtained by FIA-ionspray-MS-MS are comparable with results obtained with GC-MS methods commonly used for drugs-of-abuse screening. The MS-MS transitions used for the target analytes were shown to be highly specific when testing other related compounds and metabolites for interferences by formation of precursor ions and fragment ions. Possible variations of this method could be single-reaction monitoring (using longer dwell-times) and/or employing modified ion sources such as Turbo-Ionspray (PE Sciex) for improving detection and quantitation limits if necessary. To enhance specificity, two or more transitions per analyte and deuterated standard could be monitored. This of course would lead to increased scan times which could then be compensated for by injection of larger sample volumes. Further applications of this method to other drugs of abuse and sample material (e.g., hair extracts) are possible, although related substances and matrices have to be tested for interferences whenever a new MS-MS transition is added to this method.

Benefits of the FIA-ionspraymethod are that the derivatization of polar compounds is not necessary and short analysis times of approximately 3 rain can be realized. The analysis yields high specificitybecause of MS-MS filtering. High sensitivity by ionspray-ionization of polar compounds is obtained with LODs of 1-4 ng/mL (at S/N = 3) for the quantitated compounds. No influences by matrix effects of serum and urine samples on the quantitation of drugs were detected so far. No consumables such as GC or HPLC columns or injector liners are needed. Because both SPE and ionspray-MS-MS are automatable, this is a promising approach for the time-efficient yet reliable analysis of large numbers of samples. This method is suggested to be used as a quantitative screening technique for drugs of abuse as an alternative method to immunoassays or HPLCDAD.Although this work is the beginning of the evaluation of a very fast method with high specificity which could be comparable or at least complementary to laborious GC-MS methods, further validation of this method has to be performed. For forensic cases, the verification of results must be performed by a second, specific method.

References 1. I.B. Fenn, M. Mann, C.K. Meng, S.F. Wong, and C.M. Whitehouse. Electrospray ionization for mass spectrometry of large biomolecules. Science 246" 64-71 (1989). 2. P. Kebarle and T. Liang. From ions in solution to ions in gas phase~the mechanism of electrospray mass spectrometry. Anal. Chem. 65:A972-A986 (1993). 3. B. Kaye, M.W. Clark, N.J. Cussans, P.V.Macrae, and D.A. Stopher. The sensitive delermination of abanoquil in blood by high-performance liquid chromatography/atmospheric pressure ionizatior~ massspectrometry. BioL Mass Spectrorn. 21: 585-589 tl 992). 4. R Picotte, G, Sanchez, B. Shushan, and T. Sakuma. HPLC/ionspray/MSMS analysis of common and illicit drugs in case samples. Proceedings of the 43rd Conference on Mass Spectrometry and Allied ~opics (ASMS), American Society for Mass Spectrometry, Santa Fe, NM, 1995, p 517. 5. B. Casetta and F. Garofolo. Characterization of explosives by liquid chromatography mass spectrometry and liquid chromatography tandem massspectrometry using electrospray ionization and parent ion scanning techniques. Org. Mass Spectrom. 29:517-525 (1994). 6. D. Volmer and K. Levsen. Mass spectrometric analysis of nitrogencontaining and phosphorus-containing pesticides by liquid chromatography mass spectrometry. J. Am. Soc. Mass Spectrom. 5: 655-675 (1994). 7. G. Rule, L,G. McLaughlin, and J. Henion. A quest for oleandrin in decayed human tissue. Anal. Chem. 65:A857-A863 (1993). 8. M. Svoboda, W. Meister, and W. Vetter. A method for counting disulfide bridges in small proteins by reduction with mercaptoethanol and electrospray mass spectrometry. J. Mass Spectrom. 30:1562-1566 (1995). 9. M. Przybylski and M.O. Glocker. Electrospray mass spectrometry of biomacromolecular complexes with noncovalent interactions-mew ana[yticar perspectives for supramorecuFar chemistry and molecular recognition processes. Angew. Chem. Int. Ed. Engl. 35:806-826 (1996). 10. S.C. Beu, M.W. Senko, J.R Quinn, F.M. Wampler, and F.W. McLafferry. Fourier-tranform electrospray instrumentation for tandem high-resolution massspectrometry of large molecules. J. Am. Soc. Mass 5pectrom. 7:557-565 (1993).

327

Journal of Analytical Toxicology,Vol. 22, July/August1998 11. C.E. Parker, J.R. Perkins, K.B. Tomer, Y. Shida, K. O'Hara, and M. Kono. Application of nanoscale packed capillary liquid chromatography (75 iJm id) and capillary zone electrophoresis/electrospray ionization mass spectrometry to the analysis of macrolide antibiotics. J. Am. Soc. Mass Spectrom. 3:563-574 (1992). 12. W. Weinmann, C.E. Parker, K. Baumeister, C. Maier, K.B. Tomer, and M. Przybylski. Capillary electrophoresis combined with 2s2Cf plasma desorption and electrospray mass spectrometry for the structural characterization of hydrophobic polypeptides using organic solvents. Electrophoresis 15:228-233 (1994). 13. J. Cai and J. Henion. Elucidation of LSD in vitro metabolism by liquid chromatography and capillary electrophoresis coupled with tandem mass spectrometry. J. Anal. Toxicol. 20:27-37 (1996). 14. F.W. McLafferty. Tandem Mass Spectrometry. Wiley, New York, NY, 1983. 15. K. Pfleger, H.H. Maurer, and A. Weber. Mass Spectral and GC Data of Drugs, Poisons, Pesticides, Pollutants and their Metabolites. Verlag Chemie, Weinheim, Germany, 1992, pp 5-32. 16. R.L Foltz. Recent applications of mass spectrometry in forensic toxicology. Int. ]. Mass Spectrom. Ion Processes| 18/119:237-263 (1992). 17. J. Fehn and G. Megges. Detection of 06-monoacetylmorphine in urine samples by GC/MS as evidence for heroin use. J. Anal. Toxicol. 9:134-138 (1985). 18. M.R. Moeller. Drug detection in hair by chromatographic procedures. ]. Chromatogr. 580:125-134 (1992). 19. M.R. Moeller, P. Fey, and R. Wennig. Simultaneous determination of drugs of abuse (opiates, cocaine and amphetamine) in human hair by GC/MS and its application to a methadone treatment program. Forensic Sci. Int. 63:185-206 (1993).

328

20. W. Weinmann and M. Bohnert. Lethal monointoxication by overdosage of MDEA. Forensic Sci. Int. 91: 91-101 (1998). 21. C.C. Nelson, M.D. Fraser, J.K. Wilfahrt, and R.L. Foltz. Gas chromatography/tandem mass spectrometry measurement of delta 9tetrahydrocannabinol, naltrexone, and their active metabolites in plasma. Ther. DrugMonit. 15:557-562 (1993). 22. C.C. Nelson and R.L. Foltz. Determination of lysergic acid diethylamide (LSD), iso-LSD, and N-demethyl-LSD in body fluids by gas chromatography/tandem mass spectrometry. Anal. Chem. 64: 1578-1585 (1992). 23. U. Hofmann, M.F. Fromm, S. Johnson, and G. Mikus. Simultaneous determination of d i hydrocodeine and d ihyd romorphine in serum by gas chromatography tandem massspectrometry. J. Chromatogr. Biomed. Appl. 663:59-65 (1995). 24. Mass Spectrometry~Mass Spectrometry: Techniques and Applications of Tandem Mass Spectrometry. K.L. Busch, G.L. Glish, and S.A. McLuckey, Eds. VerTagChemie, Weinheim, Germany, 1988, p153. 25. W. Weinmann and C. Bohn. Quantitation of drugs of abuse by SPE and GC/MS using microliter-amounts of serum. Proceedings of the 44th Conference on Mass Spectrometry and Allied Topics (ASMS), American Society for Mass Spectrometry, Santa Fe, NM, 1996, p 977. 26. S. Vogt, M. Renz, R. Rickli, Th. Briellmann, and W. Weinmann. Automatische Festphasenextraktion zum Nachweis yon Drogen und Psychopharmaka aus Blutserum und Vollbiut. Toxichem. + Krimtech. 64:96-101 (1997). Manuscript received October 7, 1997; revision received January 16, 1998.