• To our knowledge, the presented workflow is the first Toxtyper / TargetScreener approach featuring complete automation from sample preparation to evaluation of data. The extraction time of about 14 min fits into the runtime of the QTOF-Screening (20 min) and only slightly exceeds the Toxtyper runtime (11 min).

• LC-MSn screening of fortified urine using µSPE led to similar or better results than the routine sample preparation. This results could be confirmed in a batch of urine samples from real cases. All µSPE-LC-MSn screening results were in good agreement with the initial routine analysis.

• LODs of the QTOF are probably lower than the ones found for the Toxtyper and will be determined separately.

• Evaluation of ME showed a maximum ion suppression of 50% which was considered acceptable for a screening approach.

• Implementation of a cleavage step would overcome the extraction issues seen for some of the glucuronides and therefore might lower the LOD of the respective parent substance but would also prolong the overall extraction time.

• Direct injection of the µSPE eluate limits the choice of solvents that could be used for extraction. Optimization of the protocol, including incorporation of an evaporation unit, might increase the overall performance of the DAU cartridge.

o Routine LC-MSn screening (Toxtyper®, TT) o LC-QTOF-MS screening (TargetScreener, TS) o Elute UHPLC equipped with a PAL RTC o Three types of smartSPETM cartridges (ITSP Solutions):

UCT C18 endcapped cartridges 10 mg UCT DAU cartridges 10 mg UCT C18 endcapped cartridges 30 mg

The Final Step Towards Automated LC-MS Screening - Implementation of an Online µSPE for Urine Screening

Michaela Schmidt1,2, Marina Schumacher3, Birgit Schneider3, Laura M. Huppertz2, Jürgen Kempf2

1Faculty Medical and Life Sciences, Furtwangen University, Schwenningen, Germany 2Institute of Forensic Medicine, Forensic Toxicology, Medical Center - University of Freiburg, Germany 3Bruker Daltonik GmbH, Bremen

Institute of Forensic Medicine

Forensic Toxicology

Systematic toxicological analysis (STA) using comprehensive screening approaches is a major part of everyday work in forensic toxicology, usually involving different analytical techniques.

The great benefit of immunoassays (IA) in routine screening of body fluids is the high degree of automation regarding sample preparation and reporting of results. In recent years, LC-MS has become a key technique in STA, but in contrast to IA an appropriate sample preparation is crucial for screening of biological samples. Offline liquid-liquid extraction (LLE), solid phase extraction (SPE) or protein precipitation (PP) are often laborious but mandatory steps and their integration into the analytical workflow is the missing piece towards a fully automated LC-MS screening analysis.

The aim of this project is to implement an online µSPE into an existing LC-MS method to achieve a fully automated LC-MS screening of urine samples.

LC Eluents

TT A: H2O, 0.1% HCOOH, 2 mM NH4+COO-, 1% acetonitrile TT B: Acetonitrile, 2 mM NH4

+COO-, 0.1% HCOOH, 1% H2O

TS A: H2O, 0.2% buffer concentrate, 1% eluent B TS B: Methanol, 0.2% buffer concentrate

µSPE Steps

1. Wash: 0.2% NH4acetate in MeOH

2. Condition: 20% NH4 acetate in H2O

3. Sample load: 200 µl

4. Desalt: 20% NH4 acetate in H2O

5. Elution: 0.2% NH4 acetate (V = 75 µl)

TT LM B 989 mL Acetonitrile 10 mL NH4 formate (0.2 M) 1 mL HCOOH

TT LM A 979 mL H2O 10 mL Acetonitrile 10 mL NH4 formate (0.2 M) 1 mL HCOOH

TS LM B

1 L MeOH 2 mL Buffer

Bruker

Elu

te

UH

PLC

- - - -

LC Tool 1 979 mL H2O dest. 10 mL Acetonitrile 10 mL Ammonium formate (200 mM) 1 mL HCOOH

Fast Wash 2 300 mL Acetonitrile 300 mL Methanol 300 mL Isopropanol 100 mL HCOOH (0.1%)

Fast Wash 1 0.1 % HCOOH

LC Tool 2 300 mL Acetonitrile 300 mL Methanol 300 mL Isopropanol 100 mL HCOOH (0.1%)

Stop

TS LM A

1 L H2O 2 mL Buffer 10 mL DE

Waste

Back

PAL RTC

B R U K E R

B R U K E R

PAL RTC µSPE System and Elute UHPLC

ESI (alternating polarity) UltraScan: 70 - 800 Da Scan Speed: 32,500 Da/s AutoMSn mode (n = 3)

Bruker amaZon Speed Ion Trap

ESI positive mode Scan range: 30 - 1000 Da Full Scan TOF MS / bbCID

Bruker impact II QTOF

Automated evaluation of HR data using a database containing over 1700 drugs, drugs of abuse and metabolites

TargetScreener HR QTOF-Screening (TS)

Toxtyper Screening (TT)

Fully automated data evaluation and reporting of results

LC Columns

TT: Acclaim® RSLC C18 2.2 µm 120A 100x2.1 mm

TS: Intensity Solo 1.8 C18-2 100x2.1 mm

Starting Equipment

Method Development

General Proof of Concept Comparison of µSPE Cartridges Reproducibility Recovery, Matrix Effects, LOD Real Samples

C18-10 C18-30

DAU

3x 75/

01

SARSTEDT

uri

ne

Proba

n

d 3

75/

01

SARSTED

T

urin

e

Proba

nd

2

75/0

1

SARSTED

T

urin

e

Proba

nd

6

75/

01

SARSTE

DT

urin

e

Proba

nd

5

75/0

1

SARSTED

T

urin

e

Proba

nd

1

75/0

1

SARSTEDT

urin

e

Proba

nd

4

1x

75/0

1

SARSTEDT

urin

e

Proba

nd

3

75/0

1

SARSTEDT

urin

e

Proba

nd

2

75/0

1

SARSTEDT

urin

e

Proba

nd

6

75/0

1

SARSTEDT

urin

e

Proba

nd

5

75/0

1

SARSTEDT

urin

e

Proba

nd

1

75/0

1

SARSTEDT

urin

e

Proba

nd

4

n = 50

75/0

1

SARSTEDT

urin

e

Proba

nd

2

75/0

1

SARSTEDT

urin

e

Proba

nd

5

75/0

1

SARSTEDT

urin

e

Proba

nd

3 7

5/0

1

SARSTEDT

urin

e

Proba

nd

6

75/0

1

SARSTEDT

urin

e

Proba

nd

6

75/0

1

SARSTEDT

PO

OLED

UR

IN

E

N = 6

75/0

1

SARSTEDT

urin

e

Proba

nd

1

75/0

1

SARSTEDT

urin

e

Proba

nd

2

75/0

1

SARSTEDT

urin

e

Proba

nd

5

75/0

1

SARSTEDT

urin

e

Proba

nd

1 7

5/0

1

SARSTEDT

urin

e

Proba

nd

2

75/0

1

SARSTEDT

urin

e

Proba

nd

5

75/0

1

SARSTEDT

urin

e

Proba

nd

3 7

5/0

1

SARSTEDT

urin

e

GS 2

3-19

75/0

1

SARSTEDT

urin

e

Proba

nd

6

75/0

1

SARSTEDT

PO

OLED

UR

IN

E

N = 6

75/0

1

SARSTEDT

urin

e

Proba

nd

2

75/0

1

SARSTEDT

urin

e

Proba

nd

5

75/0

1

SARSTEDT

urin

e

GS 1

25-18 7

5/0

1

SARSTEDT

GS 4

2-19

UR

IN

E

n = 137

Cocain

e

Am

pheta

min

e

Low, medium, high concentration

Comparison of µSPE and PP using urine from routine cases (post / ante mortem)

75/0

1

SARSTEDT

PO

OLED

UR

IN

E

N = 6

n = 6

Low, med, high concentration

Cocain

e

Am

pheta

min

e

10x

75/

01

SARSTEDT

uri

ne

Proba

n

d 3

75/

01

SARSTED

T

urin

e

Proba

nd

2

75/0

1

SARSTED

T

urin

e

Proba

nd

6

75/

01

SARSTE

DT

urin

e

Proba

nd

5

75/0

1

SARSTED

T

urin

e

Proba

nd

1

75/0

1

SARSTEDT

urin

e

Proba

nd

4

75/0

1

SARSTEDT

PO

OLED

UR

IN

E

N = 6

n = 6

25 and 100 ng/ml 27.5, 275 and 500 ng/ml for ME and RE

Cocain

e

Am

pheta

min

e

75/0

1

SARSTEDT

urin

e

Proba

nd

2

75/0

1

SARSTEDT

urin

e

Proba

nd

6

75/0

1

SARSTEDT

urin

e

Proba

nd

5

75/0

1

SARSTEDT

urin

e

Proba

nd

1

75/0

1

SARSTEDT

urin

e

Proba

nd

4

75/0

1

SARSTEDT

urin

e

Proba

nd

3

n = 6

75/0

1

SARSTEDT

urin

e

GS 0

07-18

Comparison of µSPE and routine PP

2x (ME, RE) 3x (LOD)

Cocain

e

Am

pheta

min

e

Introduction Methods

General Proof of Concept

In comparison to routine PP with acetonitrile, the identification rate of the LC-MSn screening could be improved from 74% to 84% at low concentration levels and from 90% to 96% at high concentration levels, when using µSPE.

Due to the higher sensitivity of the QTOF-MS system all spiked compounds could be detected even at low concentrations.

Comparison of µSPE Cartridges

A set of 18 compounds of different compound classes covering the retention time and mass range of the method was chosen for further testing of different µSPE cartridges. For six analytes, higher S/N ratios could be observed using the DAU cartridges. For all other analytes, no preferable cartridge could be determined.

C18-30 cartridges led to higher S/N ratios for EME and norbuprenorphine but to low absolute peak areas, probably due to higher amounts of sorbent. Higher eluent volumes might enhance the elution but would also dilute the extract injected to the LC-MS system. Therefore, the C18-30 cartridge was excluded from further method development.

n = 18

Reproducibility First tests using two different PAL tools for sample and solvent handling - to circumvent carry over at all costs - led to poor results concerning reproducibility. So, a new 250 µl LC-MS Tool was tested for all liquid handling including the injection step. Injection reproducibility of different volumes (1, 2, 5 and 10 µl) ranged from 1.5 to 7%. Optimizing the cleaning procedures after the different extraction steps led to no detectable carryover caused by the µSPE system.

Reproducibility of the complete extraction process using a single tool was tested by tenfold preparation of pooled urine fortified with a chosen set of compounds. Morphine-glucuronide was the only outlier in this test with RSDLow of 23.3 and 19.8% and RSDHigh of 75.7 and 38.3%.

RSD of ten fold extraction

C18-10 DAU

Low 5.6 - 10.9% 10.0 - 14.9%

High 4.3 - 11.2% 1.8 - 6.8%

Matrix Effects (ME) and Recovery (RE) As ion source and ion transfer of both instruments are identical, ME and RE were only evaluated by QTOF-MS using a protocol adapted from Matuszewski et al.. For the C18-10 cartridge, maximum ion suppression in six tested urine matrices was around 50%. DAU cartridges showed comparable matrix effects in a pooled urine matrix. While ion suppression will have negative effects on the LOD, ion enhancement is not an issue in screening approaches.

For morphine-glucuronide and EME the overall recovery was very poor. The C18-10 material seems to have problems properly retaining these early eluting compounds. Average RE and ME of tested analytes are depicted below.

Limits of Detection (LOD)

LOD for the µSPE-LC-MSn screening approach were determined in pooled blank urine (n = 10) fortified with compounds most often found in routine during the last year, in decreasing concentrations down to 25 ng/ml. The lowest concentration automatically identified (n = 3) was set as LOD.

Met

ham

ph

.

MD

A

Am

ph

.

Su

pp

ress

ion

RE cLow RE cMed RE cHigh Loss

ME cLow ME cMed ME cHigh

MD

MA

BE

EME

THC

-CO

OH

Met

had

on

e

EDD

P

Mo

rph

ine

6-M

AM

Mo

rph

.-G

luc

No

rco

dei

ne

Fen

tan

yl

No

rfen

tan

yl

Alp

razo

lam

α-O

H-A

lpra

z.

Bro

maz

epam

3-O

H-B

rom

az.

No

rdaz

epam

Oxa

zep

am

Zolp

idem

85

1

05

1

14

56

7

0

79

En

han

cem

en

t

16

8

17

3

16

4

89

1

23

1

76

77

9

2

11

8

79

7

7

80

13

6

13

2

13

2

10

3

92

8

4

66

9

2

10

8

70

9

5

10

4

10

7

80

6

8

72

9

2

95

76

6

9

66

59

7

6

78

95

8

1

75

81

8

0

79

84

1

09

1

18

89

1

15

1

27

10

8

13

2

14

6

94

1

28

1

30

20

0

16

7

13

3

15

7 x

3

83

x3

3

6 x

3

Average ME and RE (n = 6) of 22 Selected Analytes at Three Concentrations

Re

cove

ry

Mat

rix

Effe

cts

0%

25%

50%

75%

100%

0%

50%

100%

150%

200%

250%

Real Samples

The µSPE-LC-MSn screening results of 28 urine and 22 post-mortem urine samples from real cases were in good agreement with the findings from routine analysis. Using µSPE, 90% (C18-10) and 88% (DAU) of the substances could be identified in accordance to routine analysis. Routine LC-MSn screening could identify 80% of the compounds. The sum of all different analytes identified by µSPE and PP and confirmed by further routine analysis corresponds to 100%.

Unfortunately, neither of the cartridges could extract ethyl glucuronide (EtG) and ethyl sulfate (EtS) which were therefore excluded from this evaluation.

In one post-mortem case, the antiparkinson drug pramipexole could be detected in both µSPE runs but not in the initial routine screening. The routine Toxtyper only identified pramipexole in the corresponding cardiac blood and vitreous humor.

In a second case, amphetamine could be found in both µSPE runs but not in the routine Toxtyper. The amphetamine finding was confirmed by the more sensitive LC-QTOF-MS.

Screening Results in Real Samples (n = 50)

Conclusion

≈

≥ ID Rate

ID Rate

-

-

-

-

µS

PE

To

ol

?

-

-

-

-

LC

P T

oo

l

100 µ

L

-

-

-

-

LC

P T

oo

l

250 µ

L

10 µL 5 µL 2 µL 1 µL

Analytical Results

6 6 4 5 6 8 5

23

5 4 10 9

5 8

11 15

9 8 5

8 10 15 13

7 11

5

13 17

5 6 5 4 8

4 4

12 11 10 6

9 8 7 13

10 10 6 5 5 5

22 7 7

4 6 6

9

4

27

4 4

10 10

7

10 10

17

11 11

6

12 12

16

13

8

12

6

13

19

6 10

6 6

9

8 5

13 11

9

5

9 11 8

15

10 9

5 7 5 6

28

6 6

4

6 4

6

4

24

5 3

12

7

5

10

11

19

11 12

6

13 12

16

11

10

12

6

15

21

6

8

5 5

9

9

4

12

11

9

5

9 8

9

16

10 10

5 6

5 6

25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

Total # of analytes in the sample

7 7 7 6

5

9

5 5 5

27

12

12

12 11

5

19

11 12

6

13 15

17

13 13

10

6

18

22

6

10

6 6 6 5

8

10 13

17

28

9 9

5

10

7 7

11 10 10

13 16

0

25

50

75

100

3-O

H-B

rom

aze

pam

4-F

orm

ylam

ino

anti

pyr

ine

4-M

eth

ylam

ino

anti

pyr

ine

7-A

min

ocl

on

aze

pam

7-A

min

ofl

un

itra

zep

am

9-H

ydro

xyri

sper

ido

ne

Ace

tam

ino

ph

en

Am

bro

xol

Am

isu

lpri

de

Am

itri

pty

line

Am

itri

pty

lino

xid

Am

lod

ipin

e

a-O

H-M

idaz

ola

m

Ari

pip

razo

le

Be

nzo

yle

cgo

nin

e

Bis

op

rolo

l

Bro

maz

epam

Bu

pre

no

rph

ine

Bu

pro

pio

n

Car

bam

azep

ine

Clo

naz

ep

am

Co

cain

e

Co

de

ine

Co

de

ine-

6-G

luc.

De

smet

hyl

mir

taza

pin

e

De

xtro

me

tho

rph

an

Dih

ydro

cod

ein

e

Dip

hen

hyd

ram

ine

Do

xep

in

Ecgo

nin

emet

hyl

este

r

Flu

nit

raze

pam

Flu

oxe

tin

e

Flu

pir

tin

e

Gab

apen

tin

Hyd

roco

do

ne

Hyd

roxy

zin

e

Ke

tam

ine

Lam

otr

igin

e

Lora

zep

am LSD

MD

A

MD

EA

MD

MA

Me

lper

on

e

Me

thad

on

e

Me

top

rolo

l

Mid

azo

lam

Mo

rph

ine

Nal

oxo

ne

Nic

oti

ne

No

rbu

pre

no

rph

ine

No

rdaz

ep

am

No

rdo

xep

in

No

rfe

nta

nyl

No

rmo

rph

ine

No

rtili

din

e

No

rtri

pty

line

Op

ipra

mo

l

Oxy

cod

on

e

Par

oxe

tin

e

Pip

amp

ero

ne

Pre

gab

alin

e

Pro

maz

ine

Rit

alin

ic a

cid

Ro

curo

niu

m

Sert

ralin

e

Tem

azep

am

THC

-CO

OH

-Glu

c

Traz

od

on

e

U-4

77

00

LODs of µSPE-LC-MSn and Routine LC-MSn Analysis (70 Selected Compounds)

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

* n.d. at conc. ≤ 100 ng/ml with µSPE

* n.d. at conc. ≤ 100 ng/ml with PP

ng/ml

25, 50, 75 and 100 ng/ml for LOD

These LODs are suitable for STA in emergency and post mortem toxicology, but not sufficient for trace analysis as required for sobriety testing or analysis of DFC cases.

n = 12

n = 22