1. Introduction
Alternatives to animal testing and the development of methods that help
reducing the number of animals in preclinical studies are much sought
after. In capillary micro sampling, a low volume of blood (for example 30
µL) is collected in a capillary, centrifuged and an exact volume of plasma
(for example 10 µL) transferred into another capillary. For analysis, the
plasma is washed or blown out into a sample tube and diluted to a
volume, that can be handled reliably and reproducibly. In contrast to
DBS techniques, plasma is processed and analyzed, thus avoiding
investigation of critical parameters like effect of hematrocit and offering
standard procedures for addition of stabilizers or internal standards
during analysis. Even though necessary, dilution should be minimized as
it directly affects the achievable limits of quantification. In this poster we
describe an automated SPE procedure using small, single use
cartridges in a modified CTC liquid handling auto sampler. Results from
this approach are compared to results obtained from identical samples
processed manually by protein precipitation.
Evaluation of an automated SPE method for samples
derived from capillary microsampling
Paul-Gerhard Lassahn1, Petra Lombardi1, Winfried Wagner-Redeker1, Marie-Claude
Roubaudi-Fraschini2, Reto Bolliger3, Guenter Boehm3
1 Swiss BioAnalytics AG, Basel, Switzerland
2 Debiopharm International S.A., Lausanne, Switzerland
3 CTC Analytics AG, Zwingen, Switzerland
Experimental
General: Samples were processed by dilution of spiked plasma with water containing the internal standard
followed by online sample-preparation described below.
Working-solutions and sample preparation: The analyte (MW 581) was dissolved in DMSO and diluted in
ACN/DMSO (50:50, v/v) to obtain working-solutions at concentrations twenty times higher than those of the
final CAL and QC samples. Aliquots of the working-solutions were spiked into blank rat plasma and diluted with
water containing the internal standard in a ratio of 1:10 (v/v). The samples were vortexed, centrifuged and
transferred to the ITSP sample-preparation plate.
Online-sample-preparation: ITSP (ITSP Solutions Inc.) cartridges were conditioned with 80µL MeOH and 80µL
0.2% aqueous formic acid before loading 10µL sample. After drying the cartridge with 80µL air, the ITSP
cartridges were washed twice with 80µl 0.2% aqueous formic acid and dried again. The compounds were
eluted using 80µL EtOH and 10µL of the eluate was injected onto the analytical column. Sample-preparation of
each plasma sample was performed directly before the injection (see Figure 1 for the workflow and the results
of the optimization process).
Chromatography: 10 µL of the extracted samples were injected onto a Thermo Hypersil Gold (2.1 x 100 mm, 3
µm) analytical column. Eluent A consisted of 0.5% aqueous formic acid, eluent B of 0.5% formic acid in ACN
and LC flow rate was 0.5 mL/min. The eluent composition was changed linearly from 70% to 5% A in 2.25
minutes. Eluent composition was kept at 5% B for one minute and then went back ballistically to starting
conditions.
Calibration range: 50.0 - 25’000 ng/mL (high-range, HR) and 5.00 - 2’500 ng/mL (low-range, LR).
Instrumentation and MS-method: TSQ Quantum Ultra (Thermo Scientific), Rheos 2200 quarternary pump (Flux
Instruments), HTX PAL autosampler (CTC Analytics).
Data-Processing: Raw data were acquired with Xcalibur 2.0 software and processed with LCQuan 2.5 (Thermo
Scientific).
Validation parameters: Intra- and inter-assay accuracy and precision, linearity, and carry-over.
Table 1. MS-parameter (SRM-assay).
www.swissbioanalytics.com
Results
Blank rat plasma was spiked with the analyte and diluted with water
containing the internal standard in a ratio of 1:10 (v/v). CALs and QCs
were prepared to cover concentrations of 50.0 - 25’000 ng/mL in the
high-calibration range and 5.00 - 2’500 ng/mL in low calibration range.
The diluted plasma samples were transferred to the ITSP sample-
preparation plate and automatically extracted directly before analysis.
Sample preparation and LC-MS/MS measurements were performed in
parallel using the “Prep Ahead Function” of the CTC autosampler
(Figure 4). Chromatograms of the lowest calibration standard Cal1 are
presented in Figure 2 and compared with the chromatograms of the
same calibration standard prepared by protein-precipitation.
Three independent precision/accuracy runs were conducted for both
calibration ranges and the intra- and inter-run variabilities are
presented in Table 2. An example of a typical calibration curve for
each range is presented in Figure 3. Both assays gave very similar
results and the intra- and inter-run performance is equivalent to the
data obtained using a standard protein-precipitation method.
The carry-over in the analytical assay does not meet the acceptance
criteria of the EMA guideline but can be significantly reduced by using
Analyte Parent ion mass
[Da]
Product ion mass
[Da]
Collision energy
[V]
Analyte 582.3 255.1 35
Internal Standard 588.3 487.2 30
ESI positive, spray voltage 3.5 kV, capillary temperature 350 °C, collision gas pressure
(Ar) 1.5 mTorr, scan-time 100 msec
Figure 6. Linearity of the loading-volume and capacity of cartridges.
Loading
Eluting
Area Analyte Area IS Area RatioArea Analyte Area IS Area Ratio
5µL 10µL Injection volume 40µL
Conclusions
•The automated workflow described above gave equivalent
results (LLOQ, precision, accuracy) compared to samples
processed manually by protein precipitation. The overall
time/sample is slightly higher for the automated protocol.
However, since only minimal intervention is required, process
safety is superior and human resources are freed up for
other tasks.
•The automated protocol may be further optimized in several
aspects. As shown in Figure 7, a volume of 20 µL is sufficient
for complete elution of the analyte from the cartridge. Figure
6 indicates that an loading volume of 2-5 µL diluted plasma is
sufficient allowing the use of low-volume samples. Both
aspects are important for further optimizing this protocol for
analysis of samples originating from studies using
microsampling techniques.
•Methods for the automated optimization of the SPE
conditions exist and will be applied in future studies.
Reference: Nilsson LB, Ahnoff M, Jonsson O; Bioanalysis
(2013) 5(6), 731–738
Low-Range
Y = 0.000781592+0.00128746*X+1.0801e-008*X^2
R^2 = 0.9990 W: 1/X
0 1000 2000
ng/mL
1
2
3
AreaRatio
High-Range
Y = 0.00191642+0.000195293*X-1.32222e-009*X^2
R^2 = 0.9993 W: 1/X
0 10000 20000
ng/mL
1
2
3
4
Figure 3. Calibration-curves (low-range assay left, high-range assay right).
Figure 1. Workflow and results of the optimization process.
0.E+00
5.E+06
1.E+07
2.E+07
2.E+07 Wash 1
Wash 2
Elution 1
Elution 2
0.E+00
1.E+07
2.E+07
4.E+07
5.E+07
6.E+07
7.E+07
8.E+07
1.E+08
1.E+08
1.E+08
Elution 1
Elution 2
Elution pre-testing:
(a) 10mM NH4Fo pH9
(b) 10mM NH4Fo pH9/MeOH (50:50)
(c) 10mM NH4Fo pH9/ACN (50:50)
(d) ACN
(e) MeOH
(f) ACN/MeOH (50:50)
(g) ACN/MeOH (75:25)
(h) ACN/MeOH (75:25)
(i) EtOH
selected
0%
20%
40%
60%
80%
100%
0.1% HFo 0.2% HFo 0.5% HFo
*
* *
*
Elution 1
Elution 2
Wash 1
Wash 2
Elution 2
Elution 1
0.1% HFo 0.2% HFo 0.5% HFo
(e) (f) (i)
*
(a) (b) (c) (d) (e) (f) (g) (h) (i)
Washing variable Washing variable
Figure 2. Comparison of protein-precipitation (PP) and ITSP at LLOQ.
J:Projekte...Cal1_LR_Precipitation_02 8/23/2013 1:40:16 PM 9 3 13020 CAL1 1 1
RT: 0.0 - 4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Time (min)
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
RT: 2.6
RT: 2.7
NL: 8.83E3
TIC F: + c ESI SRM
ms2
582.300@cid35.00
[255.050-255.150]
MS
Cal1_LR_Precipitatio
n_02
NL: 8.32E5
TIC F: + c ESI SRM
ms2
588.300@cid30.00
[487.150-487.250]
MS
Cal1_LR_Precipitatio
n_02
RT: 0.0 - 4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (min)
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
RT: 2.5
RT: 2.5
NL: 2.57E4
TIC F: + c ESI
SRM ms2
582.300@cid35.0
0
[255.050-
255.150] MS
04112013_05
NL: 4.25E6
TIC F: + c ESI
SRM ms2
588.300@cid30.0
0
[487.150-
487.250] MS
04112013_05
J:Projekte...Cal1_HR_Precipitation 7/30/2013 12:10:33 PM 2 3 13020 CAL1 1 1
RT: 0.0 - 4.0 SM: 7G
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (min)
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
RT: 2.5
RT: 2.6
NL: 1.82E5
TIC F: + c ESI SRM
ms2
582.300@cid35.00
[255.050-255.150]
MS
Cal1_HR_Precipitati
on
NL: 1.20E7
TIC F: + c ESI SRM
ms2
588.300@cid30.00
[487.150-487.250]
MS
Cal1_HR_Precipitati
on
RT: 0.0 - 4.0 SM: 7G
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Time (min)
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
0
10
20
30
40
50
60
70
80
90
100
RelativeAbundance
RT: 2.4
RT: 2.5
NL: 7.44E4
TIC F: + c ESI
SRM ms2
582.300@cid35.0
0
[255.050-
255.150] MS
24102013_5
NL: 8.40E6
TIC F: + c ESI
SRM ms2
588.300@cid30.0
0
[487.150-
487.250] MS
24102013_5
Low-Range
Analyte
Internal
Standard
PP ITSP
High-Range
PP ITSP
0.0 100.0 200.0 300.0 400.0 500.0 600.0
Zentrifugieren
Verdünnen
Transfer ITSP
Fällung IS
MS
0.0 200.0 400.0 600.0 800.0 1000.0 1200.0 1400.0
PP
PP
ITSP
PP
ITSP
Gradient
5min
7min
10min
Run-Time [min]
ITSP
# of Samples: 100
centrifuge
dilute
Transfer ITSP-Plate
Protein precipitation
MS-measurement
#ofSamples
Run-Time [min]
100
200
400
Sample-Preparation:
Figure 5. Comparison of protein-precipitation (PP) and ITSP (#
Samples: variable; gradient-time: variable; ITSP: sample-preparation
7min in parallel; PP: pre-injection-time 1.0min).
PP
ITSP
PP
ITSP
PP
ITSP
at LLOQ (LR) at LLOQ (HR) > LLOQ (LR) > LLOQ (HR)
PrecIntra 9.7-10.5% 5.0-6.0% 1.5-10.5% 1.4-8.3%
AccIntra 93.2-107.4% 95.7-113.0% 101.0-105.2% 95.5-113.5%
PrecInter 11.1% 8.9% 4.4-9.5% 2.5-8.1%
AccInter 100.6% 105.1% 101.0-105.2% 104.8-108.6%
Table 2. Intra- and inter-run precision and accuracy (3 runs, [%]).
Calibration-Range Carry-over*
[% of CAL8]
Carry-over*
[% of Cal1]
Low-Range [1st Blank after Cal8] 0.6 ** 229.7 **
Low-Range [2nd Blank after Cal8] 0.2 ** 80.3 **
High-Range [1st Blank after Cal8] 0.7 278.6
High-Range [2nd Blank after Cal8] 0.3 108.5
* Standard-Wash-solution; Wash 1: H2O/EtOH (50:50, v/v); Wash 2
ACN/MeOH/i-Prop (1:1:1, v/v/v)
** Carry-Over can be significantly reduced using 0.3% Tween in
Wash-solution 1 (below 30% of Cal1 for the 2nd Blank after Cal8)
Table 3. Carry-over.
Figure 4. Prep Ahead Function of the CTC autosampler.
Area Analyte
20µL10µL 80µL
Area Analyte
40µL
Figure 7. Volume of the elution solvent EtOH (area of analyte
normalized to concentration).
0.3% Tween as an additive in the aqueous wash-solution
(see Table 3).
The influence of the amount of (diluted) plasma loaded onto
the cartridge and the volume of the extraction solvent is
presented in Figures 6 and 7. Elution from the ITSP
cartridge is complete with as little as 20 µL of extraction
solvent (Figure 7) resulting in a factor 4 higher
concentration of the analyte in the eluate compared to the
standard-elution method. The capacity of the cartridges has
been tested between 2 to 40 µL diluted plasma (Figure 6).
The area ratio of analyte to internal standard is constant
across this range and demonstrate excellent performance
of the method across a wide range of sample volumes.
The sample-preparation time and the overall run-time of
the automated and manual method are compared in Figure
5. Since only minimal manual intervention is required for the
automated approach, this workflow offers significant
advantages by high sample-workload and longer LC-
MS/MS run-times.