Talanta 176 (2018) 108–115

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

HPLC-APCI-MS/MS method development and validation for determination MARK

of tocotrienols in human breast adipose tissue
Ewa Bartosińskaa, Julia Jacynaa, Agnieszka Borsuk-De Moora, Michał Kaliszanb,
⁎
Wiesław Janusz Kruszewskic,d, Zbigniew Jankowskib, Danuta Siluka,
a
Department of Biopharmaceutics and Pharmacodynamics, Medical University of Gdańsk, Al. Gen. J. Hallera 107, 80-416 Gdańsk, Poland
b
Department of Forensic Medicine, Medical University of Gdańsk, Dębowa 23, 80-204 Gdańsk, Poland
c
Department of Oncological Surgery, Gdynia Centre of Oncology, Maritime Hospital in Gdynia, Powstania Styczniowego 1, 81-519 Gdynia, Poland
d
Division of Propedeutics of Oncology, Medical University of Gdańsk, Powstania Styczniowego 9b, 81-519 Gdynia, Poland

A R T I C L E I N F O A BS T RAC T

Keywords: For the last decade, significant attention has been paid to the potential role of tocotrienols in prevention and
Tocotrienols therapy of breast cancer. Therefore, the aim of this study was to develop and validate analytical method for
Vitamin E quantitative determination of tocotrienols (α-, β-, γ- and δ-tocotrienol) in human breast adipose tissue with the
Adipose tissue use of high performance liquid chromatography coupled with APCI-MS/MS detection. Separation of target
Triple quadrupole
compounds was achieved within 10 min with the use of naphthylethyl Cosmosil 2.5π-NAP column with
LC-MS
Method validation
methanol/water mixture (90:10, v/v) under isocratic elution. Adipose tissue samples were obtained from breast
cancer patients and women deceased as a result of accidents. Sample preparation procedure was optimized with
the application of the Plackett-Burman design and included tissue homogenization with the use of isopropanol/
ethanol/aqueous 0.1% FA mixture (13:3:8, v/v), centrifugation and solid phase extraction (SPE). The method
was validated in terms of linearity, precision, accuracy, stability (bench top, autosampler, postpreparative,
freeze and thaw stability), matrix effect (ME), recovery (RE) and process efficiency (PE). As for all four
tocotrienols ME was negligible (< 15%), precision and accuracy tests were performed with the use of
tocotrienols’ standard solutions within the ranges of 10.0–400.0 ng/g for all four tocotrienols. As the validation
requirements were met, the validated method was applied for quantitative analysis of tocotrienols in breast
cancer patients.

1. Introduction
Tocochromanols, namely tocopherols (T) and tocotrienols (T3),
belong to a group of eight naturally occurring substances widely known
as vitamin E family members [1–3] (Fig. A.1, Supplementary materi-
als). For a few decades, they were reported to present potential activity
against cancer diseases, especially breast cancer, in numerous in vitro
and in vivo tests [4–7].
Interestingly, the most promising anticancer tocochromanol prop-
erties are attributed to tocotrienols, the unsaturated homologues of
tocopherols [2,5]. Apart from being well known potent antioxidants,
they were also suggested to exhibit anti-inflammatory [8], neuropro-
tective [9], antidiabetic [10] and cholesterol-lowering properties, along
with protective activity against cardiovascular diseases [11,12]. Recent
findings have also proved tocotrienols' antiangiogenic [5,13] and
antiproliferative [14,15] activity against human breast cancer cells,
which may be further investigated in terms of potential anticancer
properties. Since very little is known about the actual quantity of
tocotrienols in adipose tissue, development and validation of a
sensitive method for their determination is necessary.
Isolation of tocopherols and tocotrienols from the adipose tissue and
their subsequent reliable quantification is challenging and in the litera-
ture, only a few examples can be found. Shim et al. described a method for
quantitative analysis of α- and γ-tocopherols after saponification of breast
adipose tissue samples with further extraction of target compounds using
ether/hexane mixture [16]. In another example, Gleize et al. implemented
Bligh and Dyer extraction and saponification in order to determine α- and
γ-tocopherol with other lipid antioxidants in subcutaneous adipose tissue
with photodiode array detection [17]. To our best knowledge, only one
paper [18] has described simultaneous quantification of tocopherols and
tocotrienols in human breast adipose tissue. Nesaretnam et al. [18]
described extraction procedure which included homogenization of 500 mg
of adipose tissue in hexane/ethanol/0.9% aqueous NaCl (4:1:1, v/v/v)
mixture.

⁎
Corresponding author.
E-mail address: dsiluk@gumed.edu.pl (D. Siluk).

http://dx.doi.org/10.1016/j.talanta.2017.08.004
Received 17 May 2017; Received in revised form 27 July 2017; Accepted 1 August 2017
Available online 02 August 2017
0039-9140/ © 2017 Elsevier B.V. All rights reserved.
E. Bartosińska et al.
The structural similarity of tocol derivatives significantly affects
their chromatographic separation. Satisfactory separation of all toco-
chromanols is easily achieved with the use of normal-phase high
performance liquid chromatography [19,20]. However, poor analytical
reproducibility and low stability of silica phases directed researchers
towards reversed-phase HPLC [21]. Furthermore, as tocols possess
natural fluorescence, a considerable part of papers concerning their
quantification describes the application of HPLC-FLD methods for this
purpose [22–25]. The use of fluorescence detection for the determina-
tion of tocopherols and tocotrienols provides high sensitivity; however,
achieving lower limits of quantification is always a desirable result
[26,27]. In contrast to fluorescence detection and at the expense of
problematic ionization of target compounds, mass spectrometry offers
advantageous sensitivity as well as selectivity and was therefore
implemented in experiments described below.
So far, very few methods for quantitative determination of toco-
trienols with the use of reversed-phase liquid chromatography techni-
ques coupled with mass spectrometry (LC-MS) have been described. In
general, atmospheric pressure chemical ionization (APCI) is a suitable
technique for the analysis of low-molecular-weight compounds of low-
polarity. Therefore, it is usually preferred to perform the analysis of
tocochromanols in both positive [19,20,28–30] and negative [31,32]
polarities. However, only in a few cases validation procedures were
performed for tocotrienols' quantification in human plasma [19] and
plant food [31]. Nevertheless, Liang et al. developed an ESI-LC-MS/MS
method for γ-tocotrienol quantification in rat plasma with lower limit
of quantification (LLOQ) equal to 10 ng/mL [33]. In the majority of
papers, negative ESI-LC/MS methods were applied for the determina-
tion of tocols in biological and plant matrices [26,34–36]. A detailed
description concerning tocochromanol analysis with the use of chro-
matographic techniques coupled with mass spectrometry was discussed
in our previous review article [21].
The objective of the study was to develop the analytical method
for quantitative determination of α-, β-, γ- and δ-tocotrienol in
adipose tissue with the use of LC-MS/MS. The method proposed by
our group allows using a limited amount of adipose tissue sample
(100 mg) followed by its rapid homogenization and extraction with
the use of more environmental friendly solvents, such as isopropanol
and ethanol instead of hexane. Moreover, low quantification limits of
target compounds (10 ng/g), as well as short time of the chromato-
graphic separation (10 min) with the application of naphthylethyl
column stationary phase were achieved in this work. The established
method was validated and applied to determine tocotrienols’ con-
centrations in breast adipose tissue from patients enrolled in a
clinical study.
2. Experimental
2.1. Chemicals
Analytical standards of tocotrienols (T3) (α-, β-, γ- and δ-),
methanol (LC-MS grade) and buthylated hydroxytoluene (2,6-di-tert-
butyl-4-methylphenol, BHT) were obtained from Sigma Chemical Co.
(Sigma Aldrich, St. Louis, MO, USA). Ethanol (HPLC grade) and 2-
propanol (LC-MS grade) were obtained from Baker (Avantor
Performance Materials B.V., Deventer, The Netherlands). Formic acid
(97%) was delivered by Alfa Aesar (A. Johnson Matthy Company,
Karlsruhe, Germany). Deionized water was produced with the use of a
D-100TUIM water system (Labopol-Polwater, Kraków, Poland).
2.2. Instrumentation and HPLC-MS/MS conditions
Quantitative determination of T3 was performed with the use of
Agilent 1200 HPLC system (Agilent Technologies, Santa Clara, CA,
USA) coupled with a triple quadrupole mass analyzer (Agilent 6430
Triple Quadrupole Mass Spectrometer, Santa Clara, CA, USA). The
109
Talanta 176 (2018) 108–115
Table 1
MRM ion transitions of tocotrienols (T3) chosen for the study.
Compound MRM transitiona Fragmentor voltage [V] Collision energy [V]
α-T3 425/165 130 19
425/205 130 11
β-T3, γ-T3 411/151 135 27
411/191 135 15
δ-T3 397/137 125 30
397/177 125 15
a
Upper and lower panel refer to the quantitative and qualitative MRM transition ions,
respectively.
HPLC system consisted of membrane degasser (G1322A), binary pump
(G1312B) and thermostated autosampler (G1329B).
The chromatographic analyses were performed with the use of
naphthylethyl column (Cosmosil 2.5π-NAP, 2.5 µm, 3.0 × 100 mm,
Nacalai Tesque, Inc., Kyoto, Japan) and methanol/water mixture
(90:10, v/v) under isocratic elution. The total analysis time, sufficient
for the separation of all tocochromanols was 10 min. Column tem-
perature, flow rate and injection volume were set at 20 °C, 0.5 mL/min
and 4 μL, respectively. The autosampler was thermostated at 4 °C.
APCI source was used under the following conditions: gas temperature:
320 °C, vaporizer temperature: 375 °C, gas flow: 4 L/min, nebulizer
pressure: 30 psi, corona current: 4 μA, capillary voltage: 3500 V. The
fragmentor voltage parameters and multiple reaction monitoring
(MRM) transitions selected with the use of the Mass Hunter
Optimizer software (Agilent Technologies, Santa Clara, CA, USA) are
summarized in Table 1. Fragmentation pathways of T3 are presented in
Fig. 1.
The analyses were carried out with the use of Agilent Qualitative
Mass Hunter Workstation (Agilent Technologies, Santa Clara, CA,
USA) composed of LC/MS Data Acquisition for 6400 Series Triple
Quadrupole version B.07.01, Qualitative Analysis version B.06.00 and
Quantitative Analysis for QQQ version B.07.00.
2.3. HPLC-FLD conditions used for experimental design step
In order to optimize sample preparation procedure (tissue homo-
genization and extraction, solid phase extraction (SPE)) Agilent HPLC
1200 system (Agilent Technologies, Santa Clara, CA, USA), composed
of membrane degasser (G1322A), quaternary pump (G1311A), ther-
mostated autosampler (G1329B) and fluorescent detector (G1321B)
were used. The excitation and emission wavelengths were set at 298
and 330 nm, respectively.
2.4. Sample collection and storage conditions
Breast adipose tissue samples were obtained from two groups of
subjects. Women who underwent a surgical intervention due to the
presence of malignant breast cancer were enrolled in the first group
and signed an informed consent before entering the study (5 partici-
pants). Adipose tissue was obtained from the peripheral area dissected
around the breast lump during surgery. The second group of subjects
consisted of women deceased in accidents (5 women). In this case
breast adipose tissue was dissected up to 24 h after the time of death.
The project was approved by the Bioethical Committee of the Medical
University of Gdańsk (No. NKBBN 464/2013).
Tissue fragments obtained from oncological surgery and forensic
medicine departments were sealed in cryogenic tubes and frozen at −
20 °C. Within a week, samples were transferred to − 80 °C refrigerator
for a long-time storage and kept frozen until analysis.
2.5. Sample preparation
Each adipose tissue fragment was thawed on ice for 30 min and
E. Bartosińska et al.
Fig. 1. Product ion spectra of the quantitative transitions for tocotrienols: m/z 425– α-tocotrienol (A), m/z 411– β-tocotrienol (B) and γ-tocotrienol (C), m/z 397– δ-tocotrienol (D).
then washed with deionized water three times in order to clear out the
blood residues. Subsequently, each sample was dissected into 100 mg
( ± 5 mg) fragments, sealed in the individual eppendorf tubes and
refrozen at − 80 °C.
On the day of analysis breast adipose tissue fragments were thawed
on ice for 30 min. Then, 400 μL of 0.1% formic acid solution, 150 μL of
ethanol and 650 μL of isopropanol (each organic solvent with 0.04%
BHT addition) were successively added into the tubes containing
adipose tissue. The resulting suspension was vortex–mixed for 15 s
at 3000 rpm. Next, samples were homogenized with the use of Bullet
Blender homogenizer (Next Advance, Inc., Averill Park, NY, USA) for
3 min at maximum operating speed (10). Homogenates obtained were
centrifuged at 1180 × g for 15 min. The 950 μL of supernatant was
collected to empty eppendorf tubes immediately after centrifugation.
Then, in order to purify the obtained tissue isolate, SPE technique
was applied (Bond Elut Plexa, non-polar polymer, 30 mg, 1 mL, Agilent
Technologies Inc.). The following procedure was carried out: cartridge
110
Talanta 176 (2018) 108–115
conditioning with 1 mL of methanol, washing with 1 mL of 0.1% formic
acid solution, and after sample application - bed cleaning with 1 mL of
20% methanol solution in water. Compounds were eluted with 1 mL of
isopropanol. The extracts collected were then concentrated in the
vacuum rotary evaporator (miVac Duo Concentrator, Genevac, SP
Scientific, Suffolk, UK) at 38 °C within 50 min. Afterwards, the dry
residue was dissolved in 100 μL of 0.04% BHT in ethanol, vortex–
mixed for 15 s at 3000 rpm and analyzed.
2.6. Optimization of adipose tissue extraction steps
2.6.1. Development of the adipose tissue homogenization/extraction
method
In the preliminary study, the following solvent mixtures suitable for
adipose tissue homogenization/extraction were selected: heptane,
isopropanol/water mixture (4:1), isopropanol/heptane mixture (4:1)
and isopropanol. Afterwards, specific extraction steps of the adipose
E. Bartosińska et al.
tissue were tested with the use of Plackett-Burman experimental design
(8 experiments, 3 dummy factors), according to the methodology
described below [37]. The tested factors included: the type of the
aqueous phase in the extracting mixture, organic phase volume used in
the extracting mixture, centrifugation temperature and the presence of
stainless steel beads enhancing tissue homogenization. The results
were determined as peak heights recounted per 100 mg of the tissue.
Factors and levels chosen for homogenization and extraction with
Plackett-Burman design are presented in Supplementary materials
(Table A.1A) while the detailed experimental plan is given in Table
A.2. In the last step, the number of homogenization cycles sufficient to
disintegrate the tissue sample was evaluated.
Standard error of the effects was obtained from a priori negligible
effects (effects of dummy variables). To recognize the significant effects,
t-test statistics and half-probability plots were used. The significance
level was set at 0.05.
2.6.2. Optimization of SPE procedure
In order to optimize the SPE procedure with the use of Bond Elut
Plexa (30 mg, 1 mL) SPE cartridges (Agilent Technologies), the 8-
factor 12-experiment Plackett-Burman screening design [37–39] was
performed with experiments carried out in a randomized order,
following methodology described in Section 2.6.1. The investigated
factors were as follows: aqueous phase type used for cartridge
conditioning, organic phase type in the extracting mixture, organic
solvent content in the extracting mixture, cleaning solution type, its
volume and composition, as well as eluting solvent volume and type.
Peak heights of three T3 (α-, γ- and δ-T3) were chosen as the
responses. The experimental plan was summarized in Supplementary
materials (Tables A.1B and A.3).
2.7. Preparation of stock solutions, working standard solutions for
calibration curves and for QC samples
Each T3 stock solution was prepared as follows: 1 mL of ethanol
with 0.04% BHT was added to 100 mg of every T3 in original amber
glass vial. The resulting solution was vortex– mixed and used for
further dilutions. Working solutions I of α-, β-, γ- and δ-T3 were
prepared by diluting 5 μL of stock solution in ethanol with 0.04% BHT
in the 10 mL volumetric flasks, yielding 50 μg/mL solutions.
Appropriate volumes of these working solutions I were used in order
to prepare 0.050, 0.150, 0.250, 0.500, 1.000 and 2.000 µg/mL of
working solutions II (L1-L6) used for the preparation of calibration
curves and 0.100, 0.750 and 1.500 µg/mL for the quality control
solutions (LQC – lower quality control, MQC – medium-quality control
and HQC – high quality control). The 20 μL of working solutions II was
added to 80 μL of ethanol with 0.04% BHT and used for validation
studies in order to obtain calibration curve points (0.010, 0.030, 0.050,
0.100, 0.200 and 0.400 µg/g) and quality control samples (0.020, 0.075
and 0.300 µg/g). All procedures were carried out under a gentle stream
of nitrogen and in a dim light.
Stock solutions of T3 were flushed with nitrogen and stored at −
80 °C. Working solutions I and II were filled with nitrogen and stored
at − 20 °C.
2.8. Method validation
The method was validated in terms of linearity, precision, accuracy,
stability (bench top, autosampler, postpreparative, freeze and thaw
stability), matrix effect (ME), recovery (RE) and process efficiency (PE)
following the FDA guidance [40].
In order to ensure the reliability of the validation studies, breast
adipose tissue samples were homogenized and pooled to obtain a
unified sample matrix. Forty eight breast adipose tissue samples
(~ 100 mg each) from 8 sources were homogenized following the
procedure described in Section 2.5. After centrifugation, 950 μL of
111
Talanta 176 (2018) 108–115
supernatant from each sample was collected into one falcon tube,
vortex–mixed and divided into 48 eppendorf tubes for validation
purpose.
2.8.1. Evaluation of matrix effect, recovery and process efficiency
In order to determine ME, RE and PE, three sets of samples were
prepared: set A consisted of standard sample solutions of the T3 at LQC
(6 samples) and HQC (6 samples) levels; set B included breast adipose
tissue homogenates spiked at LQC (4 samples) and HQC (4 samples)
levels after extraction; set C was composed of breast adipose tissue
homogenates spiked at LQC (4 samples) and at HQC (4 samples)
before extraction. The results were calculated with the use of the
following equations [41]:
⎛B ⎞
a) ME = ⎜ − 1⎟ × 100%
⎝A ⎠
C
b) RE = B × 100%
C
c) PE = A
× 100%
As vitamin E constituents are present in the adipose tissue, the
mean peak area values of T3 present in the unspiked homogenate were
subtracted in the case of B and C sample sets to calculate ME and PE.
The ME should not exceed 15% [42], RE and PE values are thought to
ensure reproducible and reliable results. In addition, obtaining this
biological matrix is problematic due to ethical reasons; therefore a
limited number of biological samples was used in this part of validation
study.
Additionally, ME was assessed with the use of autosampler injec-
tion program by simultaneous injection (sandwich injection) of pure
standard solutions (a) with sample matrix (b) or with 0.04% BHT in
ethanol (c). The QC samples (3 LQC, 3 MQC and 3 HQC) and matrix
sample used for this method were prepared similarly as described in ,
with the exception that after SPE they were finally dissolved in 50 μL of
0.04% BHT in ethanol. The injection program used for ME evaluation
was as follows: 2 μL of 0.04% ethanolic BHT (c) or sample matrix (b)
was collected from a vial. Then, the needle was washed in ethanol for
three times and another 2 μL of pure standards (a) were collected from
a chromatographic vial. The resulting 4 μL solution was mixed and
injected onto the chromatographic column. Peak areas were used for
the ME determination following methodology described above.
2.8.2. Linearity, LOD, LOQ
The calibration range and linearity of T3 were defined within 10
and 400 ng/g. Limit of detection (LOD) was assessed at signal–to-noise
ratio (S/N) equal to 3 based on peak height. Limit of quantification
(LOQ) was established for α-, β-, γ- and δ-T3 in relation to S/N equal to
10 based on peak height. The peak height instead of peak area for LOD
and LOQ calculations was chosen in this case, as a more restrictive
parameter.
2.8.3. Precision and accuracy
As ME studied with two different approaches was found negligible,
precision and accuracy studies were performed with the use of standard
T3 solutions. Intra-day precision was evaluated using 6 replicates of T3
standard samples at four concentration levels: LLOQ (10 ng/g), LQC
(20 ng/g), MQC (75 ng/g) and HQC (300 ng/g). The CV should not
exceed 15% for LQC, MQC and HQC, however in the case of LLOQ
level, the CV values should not exceed 20% [40,42]. Inter-day precision
was evaluated with the use of 18 T3 standard samples at three quality
control levels: LQC, MQC and HQC with CV not greater than 15%.
Accuracy was determined by calculating the measured concentration/
nominal concentration ratio. The intra- and inter-day accuracy should
not exceed the range of 85–115% and for LLOQ intra-day precision
might fall within the range from 80% to 120% [40,42].
E. Bartosińska et al.
2.8.4. Stability of tocotrienol standard samples and breast adipose
tissue extracts
Bench-top stability (4 h), the stability of QC standard solutions and
adipose tissue extract samples after 7- and 12-h stored at 4 °C
(autosampler conditions), as well as freeze-thaw (− 80 °C) stability
assays, were performed. Bench-top stability was determined after 4-h
storage of the breast adipose tissue homogenates spiked at MQC level
before extraction (3 samples) at room temperature. The 4-h period
refers to the maximum time required for the overall sample prepara-
tion procedure. The freeze and thaw stability assays were performed
along three consecutive days, including three cycles. Twelve replicates
of LQC and HQC (24 samples totally) were prepared as described above
(See: Section 2.7), using freshly prepared breast adipose tissue homo-
genate. The 3 LQC and 3 HQC samples were analyzed immediately
after preparation and the remaining 18 samples were kept frozen for
24 h at − 80 °C. Then, all the samples were thawed at room tempera-
ture, and analogous control samples were subjected for the analysis.
Two more cycles were performed after consecutive 12 and 24 h. The
concentration difference from “0” time should not exceed 15%.
3. Results and discussion
The current study reports a novel and validated method for
quantitative analysis of T3 in human breast adipose tissue. To our
best knowledge, so far only two clinical trials have been performed in
women with malignant and benign breast lumps, in order to assess the
tocochromanol content in adipose tissue [16,18]. Nesaretnam et al.
described a method for quantification of four tocopherols and three out
of four T3 in benign and malignant breast cancer patients using NP-
HPLC-FLD technique, however with no validation step [18]. In the
paper presented by Shim et al. quantitative analysis of lipid antiox-
idants along with only α- and γ-tocopherols was carried out [16].
However, no conclusive outcome regarding therapeutic activity of
tocopherols and T3 was established.
3.1. Development of the breast adipose tissue sample preparation
procedure
3.1.1. Optimization of adipose tissue extraction steps
As it was mentioned above, the preparation of breast adipose tissue
for analytical purposes remains a long-term process. Moreover, very
limited amount of the biological material is available. Design of
experiments approach implemented for the sample pretreatment
optimization allowed using very small amounts of the biological
material.
Although in the case of solid tissue extraction, it is usually
performed by application of two immiscible phases, such as water
and hexane. This strategy, however, might not be appropriate for
adipose tissue analysis. In this case, the use of organic solvents alone in
extraction procedure enables disjunction of potential tocotrienol
associations with matrix components [27], however hinders selective
extraction of these compounds. Nesaretnam et al. successfully applied
hexane/ethanol/0.9% aqueous NaCl mixture (4:1:1) for the extraction
of tocopherols and T3 from breast adipose tissue [10]. However,
extracting mixture composed of non-polar solvents was not applicable
in our study with MS detection, as the exhaustive extraction of lipid
compounds with hexane or heptane would increase matrix effects. In
our preliminary study, three constituents of extracting mixture were
selected from heptane, hexane, acetone, isopropanol, ethanol and
water [data not shown]. Surprisingly, the peak areas obtained for T3
in the case of hexane and other non-polar extracts were decreased in
comparison to the results obtained with the application of more polar
extracting mixtures, such as ethanol/water and isopropanol/water.
Saponification of adipose tissue samples would be a time-consuming
and deteriorating sample pretreatment method for target compounds
[27]. What is more, it would affect their ionization due to the use of
112
Talanta 176 (2018) 108–115
strong electrolytes during saponification process. For these reasons, the
final extraction method proposed by our group consisted of 0.1%
formic acid solution, ethanol and isopropanol.
3.1.2. Plackett-Burman screening design applied for optimization of
extraction process
Results of the Plackett-Burman design used for optimization of T3
isolation from the adipose tissue proved that among studied factors,
only organic phase content in the extracting mixture was significant
(Supplementary materials, Table A.4). As it was assumed, higher
organic content improved extraction yields and did not affect solid
phase extraction. Although it was not recognized as an important factor
in Plackett-Burman screening design, nut based on our prior knowl-
edge, we included 0.1% formic acid solution in the extracting mixture.
In reference to the literature, the application of formic acid into the
water phase was implemented in order to enhance the stability of
resulting homogenate and to facilitate degradation of the cell mem-
branes [43]. Similarly, it was also noticed that the decreased centrifu-
gation temperature negatively affected isolation of T3 from homoge-
nates, which was observed in the obtained chromatograms as peak
tailing. Although tocol derivatives are temperature sensitive, it was
decided to centrifuge adipose tissue extracts at 18 °C collecting super-
natant immediately in order to maintain extraction equilibrium within
all of the samples.
Analyzing the results from the Plackett-Burman design, it was
decided not to use stainless steel beads in the homogenization step.
Additionally, during optimization of T3 isolation from adipose tissue it
appeared that less homogenization cycles (1 cycle × 3 min) resulted in
higher extraction recoveries in comparison to repeated homogenization
(3 cycles × 3 min) [data not shown]. The explanation for this might be
related to the concurrent extraction of interfering compounds during
longer sample decomposition. Therefore just one homogenization cycle
in the bullet blender was used throughout further studies with
beneficial decrease of sample preparation time.
3.1.3. Experimental design results for SPE optimization
In the course of T3 isolation from the adipose matrix, a supple-
mentary purification step had to be implemented in the sample
preparation procedure as it was found that single homogenization
and target compounds extraction to polar solvent is not sufficient
before LC analysis. Thus, purification process was intensified by the
addition of solid phase extraction step to the procedure. The organic-
aqueous phase ratio was taken into account as it was crucial to observe
the effectiveness of SPE bed retention of tocols, e.g. high organic
content leads to decreased tocol recovery. In the case of homogeniza-
tion and extraction optimization, it was necessary to choose the
isopropanol/ethanol-water ratio effective for sample extraction.
Therefore, this factor was analyzed twice in two different experimental
plans. The screening design applied for SPE optimization revealed one
significant factor affecting this procedure (Supplementary materials,
Table A.5), namely organic solvent amount in the cleansing solution.
Taking this factor into account, from the tested range of concentrations
(10–80%), 20% methanol in water as cleaning solution was chosen for
the final SPE procedure.
3.2. Validation study
In case of tocochromanols’ analysis, the choice of an appropriate IS
is still a challenging task. In the proposed study three tocopheryl esters,
including succinate (TS), acetate (TA) and nicotinate (TN), were first
tested as IS. Unfortunately, TS revealed poor ionization which hindered
its application in our study. The TA and TN significantly extended the
time of analytical run and were susceptible for degradation during
sample preparation procedure, which can be due to the acidic
environment in the extracting mixture. Another compound tested as
the IS was D6-α-tocopherol. However, it proved to be unstable during
E. Bartosińska et al. Talanta 176 (2018) 108–115

Table 2
Validation results obtained for tocotrienols.

Analyte LOD LOQ Nominal concentration Intraday repeatibility Interday repeatibility Accuracy [%] ME [%] RE [%] PE [%] Precision of
[ng/g] [ng/g] [ng/g] [CV%] (n = 6) [CV%] (n = 18) A B matrix spiked
samples [CV%]

α-T3 3 10 LLOQ 10 3.6 – 115.1 – – – – –

LQC 20 4.0 3.9 103.4 10.9 4.0 90.4 74.4 7.1
MQC 75 3.4 2.7 88.9 10.8 13.7 75.6 66.5 4.6
HQC 300 7.4 5.7 96.4 6.9 6.5 76.6 78.3 3.7
β-T3 2 8 LLOQ 10 7.7 – 91.4 – – – – –
LQC 20 8.6 12.0 104.0 10.7 4.2 96.7 61.0 6.1
MQC 75 7.6 9.0 97.7 14.8 9.2 90.3 82.9 6.1
HQC 300 10.3 7.5 98.9 3.2 2.1 91.1 89.2 5.3
γ-T3 3 10 LLOQ 10 9.2 – 94.0 – – – – –
LQC 20 4.6 7.5 99.9 0.4 − 0.5 89.7 61.8 9.5
MQC 75 2.1 5.1 92.2 15.0 12.0 80.3 76.9 6.4
HQC 300 5.9 5.9 98.2 10.4 12.8 73.4 76.3 4.4
δ-T3 0.2 0.7 LLOQ 10 8.1 – 93.3 – – – – –
LQC 20 7.8 11.2 98.6 3.0 2.1 75.4 65.4 7.1
MQC 75 5.7 7.9 100.0 10.0 7.5 71.6 74.9 4.6
HQC 300 5.9 7.4 100.3 9.4 9.7 77.0 83.4 3.7

A – ME calculated with the use of the reference method, B – ME calculated on the basis of the sample injection program (sandwich injection).

sample storage at autosampler conditions. Therefore, it was decided to
analyze samples basing solely on external standards (tocotrienols’
analytical standards) for calibration curves construction instead of IS
methodology.
3.2.1. Matrix effect, recovery and process efficiency results
Matrix effect, recovery and process efficiency results are summar-
ized in Table 2. The ME values were found negligible and therefore
linearity, precision and accuracy were further investigated with the use
of T3 standard solutions. Moreover, both ME evaluation methods
usually led to similar findings (Table 2). The CV values obtained for the
alternative method were satisfactory and ranged within 2.7–12.1% for
the standard samples and 1.8–12.0% for the matrix samples. Recovery
ranged 71.6–96.7% and provided repeatable results when analyzing all
T3 compounds under study. Process efficiency was satisfactory and
ranged within 61.8–89.2% for all the T3 tested. The representative
chromatograms of T3 found in QC samples and breast adipose tissue
homogenate were presented in Fig. 2A and B.
The proposed method for ME evaluation with the use of injection
program requires only one adipose tissue sample for successful
analysis, which might represent an alternative approach of method
validation in terms of compounds extracted from poorly accessible
biological matrices.
For all four T3, ME values were acceptable, with only one value
attributed to γ-T3, which slightly exceeded 15% of concentration
difference between the standard solutions and spiked biological
samples at MQC level. This phenomenon might be associated with
the use of the APCI mode which is generally less prone to matrix effects
due to different ionization mechanisms compared to ESI [44,45]. The
non-polar nature of these compounds is reflected in their ionization
providing an explanation for both, problematic efficiencies [1,34,46],
as well as reproducibility [47] of this process (CV ≈ 10%) also observed
in our study. For this reason, the investigation of the main factors
affecting ionization of tocopherols and T3 is necessary.
3.2.2. Linearity, LOD, LOQ, precision, accuracy and stability
Responses for all four T3 were linear within the calibration ranges
described above (10–400 ng/g). The correlation coefficients obtained
from linear regression analysis were as follows: 0.9994–0.9997 for α-
T3; 0.9972–0.9999 for β-T3; 0.9979–0.9998 for γ-T3 and 0.9992–
0.9997 for δ-T3. The LOD and LOQ values obtained for T3 are shown
in Table 2.
All four T3 were characterized by satisfactory intra- and inter-day
precision and accuracy during the validation study. The results were
presented in Table A.6 (Supplementary materials). Although the
precision studies were performed with the use of T3 standard solutions,
it is worth mentioning that homogenate samples spiked before extrac-
tion were also characterized by good precision (Table 2).
Vitamin E compounds are thermally labile, susceptible for oxida-
tion and light-sensitive, which makes sample preparation and storage
conditions a laborious task. In our study T3 standard samples and
breast adipose tissue samples met the validation criteria in reference to
their preparation process, storage and analysis (Table A.6). The α- and
γ-T3 in breast adipose tissue extracts proved their stability during three
freeze and thaw cycles. The δ-T3 spiked at LQC level exceeded the
acceptance criterion of accuracy after the first freeze- and thaw cycle in
reference to “0” time (> 15%). However, β- and δ-T3 demonstrated
decreasing stability after the second freeze- and thaw cycles at HQC
level. For this reason, careful handling of breast adipose tissue samples,
as well as further investigation of this phenomenon are necessary. The
probable issue of freeze and thaw stability test might be related to the
histological characteristics of adipose tissue. It is demanding and
problematic to prepare identical biological samples daily and receive
reliable results, as adipose tissue often contains also glandular and
subcutaneous tissues. For this reason, the samples for the validation
tests, including freeze and thaw stability tests, were pooled. We
presume that the resulting homogenate is less stable than the tissue
fragments kept frozen at − 80 °C.
3.3. Application of the validated method
The validated method was successfully applied for quantitative
determination of T3 in human breast adipose tissue (Fig. 2C). The
estimated T3 concentrations in breast adipose tissue were expressed as
ng/g per adipose tissue (Table 3). However, one of the tested samples
was below the calibration ranges for all T3, which was likely due to the
presence of breast glandular tissue in addition to adipose tissue in the
analyzed sample. In the case of another sample, β-T3 concentration
was found above the upper limit of quantification (ULOQ).
Taking into account that only one method for the determination of
these compounds in the adipose tissue has been provided so far,
limited number of comparative analyses in terms of the type of
extraction, quantification limits and T3 quantities in the biological
material, are currently available. As very limited amount of T3 is found
in plant food in comparison to tocopherols [46], dietary supplementa-
tion of this compound affects its concentration in human organism,
especially in adipose tissue. According to the findings by Nesaretnam
et al. [18], T3 content in human breast adipose tissue significantly

113
E. Bartosińska et al. Talanta 176 (2018) 108–115

Fig. 2. Chromatographic separation of tocotrienols: (A) – quality control (QC) samples of tocotrienol standard solutions; (B) – extract of breast adipose tissue pooled homogenate;
(C) – an exemplary chromatogram of the extract obtained from patients’ breast adipose tissue sample.

Table 3
concentration level of 10 ng/g. Application of extract purification steps
Estimated T3 concentrations in breast adipose tissue obtained from breast cancer
patients and women deceased in accidents (n = 9) recounted as ng/g of breast adipose
including SPE allowed for significant improvement of extraction
tissue. performance. Environmental friendly and rapid sample preparation
procedure, along with fast chromatographic method (10 min analytical
Compound Median 25% quantile 75% quantile Mean SD run) proved to be sufficient for the separation of target compounds.
[ng/g] [ng/g] [ng/g] [ng/g]
Taking into account therapeutic potential of T3 and related com-
α-T3 69.8 46.0 80.2 79.2 52.2 pounds, as well as their potential significance as anticancer agents, the
β-T3 214.3 153.1 264.2 235.4 115.6 present study may support further breast cancer research.
γ-T3 104.0 91.8 134.5 138.5 113.1
δ-T3 18.5 15.2 23.7 23.2 16.4
Acknowledgements

Authors would like to thank Magdalena Walentynowicz for her

exceeds the amount reported by our group, e.g. mean α-T3 concentra-
tions obtained for benign and malignant breast cancer patients were
equal to 11.4 and 7.2 μg/g, respectively, whereas in our study the mean
concentration was equal to 79.2 ng/g. However, taking into account
mean values, the proportion of given tocotrienols in the adipose tissue
is slightly different: according to Nesaretnam et al., [α-T3] > [γ-T3] >
[δ-T3], whereas our findings suggest that β- and γ-T3 occur in excess in
relation to α-T3 and δ-T3, which is hardly detectable in some
individuals. The reason for this might be related to different dietary
habits and pharmacokinetics-driven inter-individual variability.
input in the experimental part of the study. We would also like to thank
Łukasz Kubik for his help in preparation of graphics for this manu-
script and Emilia Daghir-Wojtkowiak for her helpful comments.
This study was supported by the Polish National Science Centre,
Grant no. DEC-2013/11/N/NZ5/00164.
Appendix A. Supplementary material
Supplementary data associated with this article can be found in the
online version at doi:10.1016/j.talanta.2017.08.004.

4. Conclusions References

The developed and validated method proved to be fast and selective [1] H. Ahsan, A. Ahad, W.A. Siddiqui, A review of characterization of tocotrienols from
for successful quantification of T3 in human breast adipose tissue at the plant oils and foods, J. Chem. Biol. 8 (2015) 45–59. http://dx.doi.org/10.1007/

114
E. Bartosińska et al.
s12154-014-0127-8.
[2] K. Nesaretnam, W.Y. Wong, M.B. Wahid, Tocotrienols and cancer: beyond
antioxidant activity, Eur. J. Lipid Sci. Technol. 109 (2007) 445–452. http://
dx.doi.org/10.1002/ejlt.200600212.
[3] B.B. Aggarwal, C. Sundaram, S. Prasad, R. Kannappan, Tocotrienols, the vitamin E
of the 21st century: its potential against cancer and other chronic diseases,
Biochem Pharmacol. 80 (2010) 1613–1631. http://dx.doi.org/10.1016/
j.bcp.2010.07.043.
[4] M.T. Ling, S.U. Luk, F. Al-Ejeh, K.K. Khanna, Tocotrienol as a potential anticancer
agent, Carcinogenesis 33 (2012) 233–239. http://dx.doi.org/10.1093/carcin/
bgr261.
[5] K. Nesaretnam, P. Meganathan, S.D. Veerasenan, K.R. Selvaduray, Tocotrienols
and breast cancer: the evidence to date, Genes Nutr. 1 (2012) 3–9. http://
dx.doi.org/10.1007/s12263-011-0224-z.
[6] A.K. Smolarek, N. Suh, Chemopreventive activity of vitamin E in breast cancer: a
focus on γ- and δ-tocopherol, Nutrients 3 (2011) 962–986. http://dx.doi.org/
10.3390/nu3110962.
[7] S. Wada, Y. Satomi, M. Murakoshi, N. Noguchi, T. Yoshikawa, H. Nishino, Tumor
suppressive effects of tocotrienol in vivo and in vitro, Cancer Lett. 229 (2005)
181–191. http://dx.doi.org/10.1016/j.canlet.2005.06.036.
[8] A.S. Nazrun, M. Norazlina, M. Norliza, S.I. Nirwana, The anti-inflammatory role of
vitamin E in prevention of osteoporosis, Adv. Pharmacol. Sci. 2012 (2012). http://
dx.doi.org/10.1155/2012/142702.
[9] S. Khanna, S. Roy, N.L. Parinandi, M. Maurer, C.K. Sen, Characterization of the
potent neuroprotective properties of the natural vitamin E α-tocotrienol, J.
Neurochem. 98 (2006) 1474–1486. http://dx.doi.org/10.1111/j.1471-
4159.2006.04000.x.
[10] F. Fang, Z. Kang, C. Wong, Vitamin E tocotrienols improve insulin sensitivity
through activating peroxisome proliferator-activated receptors, Mol. Nutr. Food
Res. 54 (2010) 345–352. http://dx.doi.org/10.1002/mnfr.200900119.
[11] K. Prasad, Tocotrienols and cardiovascular health, Curr. Pharm. Des. 17 (2011)
2147–2154. http://dx.doi.org/10.2174/138161211796957418.
[12] H. Ahsan, A. Ahad, J. Iqbal, W.A. Siddiqui, Pharmacological potential of toco-
trienols: a review, Nutr. Metab. 11 (2014) 52. http://dx.doi.org/10.1186/1743-
7075-11-52.
[13] Y. Li, W.-G. Sun, H.-K. Liu, G.-Y. Qi, Q. Wang, X.-R. Sun, B.-Q. Chen, J.-R. Liu, γ-
Tocotrienol inhibits angiogenesis of human umbilical vein endothelia cell induced
by cancer cell, J. Nutr. Biochem. 12 (2011) 1127–1136. http://dx.doi.org/10.1016/
j.jnutbio.2010.09.012.
[14] V.B. Wali, P.W. Sylvester, Synergistic antiproliferative effects of gamma-tocotrienol
and statin treatment on mammary tumor cells, Lipids 42 (2007) 1113–1123.
http://dx.doi.org/10.1007/s11745-007-3102-0.
[15] P.W. Sylvester, N.M. Ayoub, Tocotrienols target PI3K/Akt signaling in anti-breast
cancer therapy, Anticancer Agents Med. Chem. 13 (2013) 1039–1047. http://
dx.doi.org/10.2174/18715206113139990116.
[16] E. Shim, K.-J. Yeum, G. Tang, S.H. Ahn, J. Hwang, Y.C. Lee-Kim, Retinoids,
carotenoids, and tocopherols in breast adipose tissue and serum of benign breast
disease and breast cancer patients, Nutr. Cancer 64 (2012) 956–963. http://
dx.doi.org/10.1080/01635581.2012.717678.
[17] B. Gleize, M. Steib, M. André, E. Reboul, Simple and fast HPLC method for
simultaneous determination of retinol, tocopherols, coenzyme Q10 and carotenoids
in complex samples, Food Chem. 134 (2012) 2560–2564. http://dx.doi.org/
10.1016/j.foodchem.2012.04.043.
[18] K. Nesaretnam, P.A. Gomez, K.R. Selvaduray, G.A. Razak, Tocotrienol levels in
adipose tissue of benign and malignant breast lumps in patients in Malaysia, Asia
Pac. J. Clin. Nutr. 16 (2007) 498–504.
[19] K. Nagy, M.-C. Courtet-Compondu, B. Holst, M. Kussmann, Comprehensive
analysis of vitamin E constituents in human plasma by liquid chromatography-
mass spectrometry, Anal. Chem. 79 (2007) 7087–7096. http://dx.doi.org/
10.1021/ac0708689.
[20] W.M. Stöggl, C.W. Huck, H. Scherz, M. Popp, G.K. Bonn, Analysis of vitamin E in
food and phytopharmaceutical preparations by HPLC and HPLC-APCI-MS-MS,
Chromatographia 54 (2001) 179–185. http://dx.doi.org/10.1007/BF02492241.
[21] E. Bartosińska, M. Buszewska-Forajta, D. Siluk, GC-MS and LC-MS approaches for
determination of tocopherols and tocotrienols in biological and food matrices, J.
Pharm. Biomed. Anal. 127 (2016) 156–169. http://dx.doi.org/10.1016/
j.jpba.2016.02.051.
[22] D. Siluk, R.V. Oliveira, M. Esther-Rodriguez-Rosas, S. Ling, A. Bos, L. Ferrucci,
I.W. Wainer, A validated liquid chromatography method for the simultaneous
determination of vitamins A and E in human plasma, J. Pharm. Biomed. Anal. 44
(2007) 1001–1007. http://dx.doi.org/10.1016/j.jpba.2007.03.033.
[23] N. Grebenstein, J. Frank, Rapid baseline-separation of all eight tocopherols and
tocotrienols by reversed-phase liquid-chromatography with a solid-core penta-
fluorophenyl column and their sensitive quantification in plasma and liver, J.
Chromatogr. A. 1243 (2012) 39–46. http://dx.doi.org/10.1016/j.chro-
ma.2012.04.042.
[24] H. Che, D.M. Tan, P. Meganathan, Y. Gan, G.A. Razak, J. Fu, Validation of a HPLC/
FLD method for quantification of tocotrienols in human plasma, Int. J. Anal. Chem.
2015 (2015) 5–12. http://dx.doi.org/10.1155/2015/357609.
[25] M.N. Irakli, V.F. Samanidou, I.N. Papadoyannis, Development and validation of an
HPLC method for the simultaneous determination of tocopherols, tocotrienols and
carotenoids in cereals after solid-phase extraction, J. Sep. Sci. 34 (2011)
1375–1382. http://dx.doi.org/10.1002/jssc.201100077.
[26] M.M. Delgado-Zamarreño, M. Bustamante-Rangel, S. Sierra-Manzano,
115
Talanta 176 (2018) 108–115
M. Verdugo-Jara, R. Carabias-Martínez, Simultaneous extraction of tocotrienols
and tocopherols from cereals using pressurized liquid extraction prior to LC
determination, J. Sep. Sci. 32 (2009) 1430–1436. http://dx.doi.org/10.1002/
jssc.200800707.
[27] F.J. Rupérez, D. Martín, E. Herrera, C. Barbas, Chromatographic analysis of alpha-
tocopherol and related compounds in various matrices, J. Chromatogr. A. 935
(2001) 45–69. http://dx.doi.org/10.1016/S0021-9673(01)01101-3.
[28] C. Nimalaratne, C. Sun, J. Wu, J.M. Curtis, A. Schieber, Quantification of selected
fat soluble vitamins and carotenoids in infant formula and dietary supplements
using fast liquid chromatography coupled with tandem mass spectrometry, Food
Res. Int. 66 (2014) 69–77. http://dx.doi.org/10.1016/j.foodres.2014.08.034.
[29] A. Gentili, F. Caretti, Evaluation of a method based on liquid chromatography-
diode array detector-tandem mass spectrometry for a rapid and comprehensive
characterization of the fat-soluble vitamin and carotenoid profile of selected plant
foods, J. Chromatogr. A 1218 (2011) 684–697. http://dx.doi.org/10.1016/j.chro-
ma.2010.12.001.
[30] W. Zarrouk, A. Carrasco-Pancorbo, M. Zarrouk, A. Segura-Carretero, A. Fernández-
Gutiérrez, Multi-component analysis (sterols, tocopherols and triterpenic dialco-
hols) of the unsaponifiable fraction of vegetable oils by liquid chromatography-
atmospheric pressure chemical ionization-ion trap mass spectrometry, Talanta 80
(2009) 924–934. http://dx.doi.org/10.1016/j.talanta.2009.08.022.
[31] P. Viñas, M. Bravo-Bravo, I. López-García, M. Pastor-Belda, M. Hernández-
Córdoba, Pressurized liquid extraction and dispersive liquid-liquid microextraction
for determination of tocopherols and tocotrienols in plant foods by liquid
chromatography with fluorescence and atmospheric pressure chemical ionization-
mass spectrometry detection, Talanta 119 (2014) 98–104. http://dx.doi.org/
10.1016/j.talanta.2013.10.053.
[32] A.M. Drotleff, C. Bohnsack, I. Schneider, A. Hahn, W. Ternes, Human oral
bioavailability and pharmacokinetics of tocotrienols from tocotrienol-rich (toco-
pherol-low) barley oil and palm oil formulations, J. Funct. Foods 7 (2014)
150–160. http://dx.doi.org/10.1016/j.jff.2014.01.001.
[33] S. Liang, X. Bian, J. Ma, M. Arogunjo, A.A. Deorukhkar, S. Krishnan, H. Xie,
Development and validation of a sensitive LC/MS/MS method for the determina-
tion of γ-tocotrienol in rat plasma: application to pharmacokinetic studies, Biomed.
Chromatogr. 27 (2013) 58–66. http://dx.doi.org/10.1002/bmc.2748.
[34] M. Bustamante-Rangel, M.M. Delgado-Zamarreño, A. Sánchez-Pérez, R. Carabias-
Martínez, Determination of tocopherols and tocotrienols in cereals by pressurized
liquid extraction-liquid chromatography-mass spectrometry, Anal. Chim. Acta 587
(2007) 216–221. http://dx.doi.org/10.1016/j.aca.2007.01.049.
[35] K. Hejtmánková, J. Lachman, A. Hejtmánková, V. Pivec, D. Janovská, Tocols of
selected spring wheat (Triticum aestivum L.), einkorn wheat (Triticum monococ-
cum L.) and wild emmer (Triticum dicoccum Schuebl [Schrank]) varieties, Food
Chem. 123 (2010) 1267–1274. http://dx.doi.org/10.1016/j.food-
chem.2010.05.064.
[36] J.U. Ziegler, R.M. Schweiggert, R. Carle, A method for the simultaneous extraction
and quantitation of lipophilic antioxidants in Triticum sp. by HPLC-DAD/FLD-
MSn, J. Food Compos. Anal. 39 (2015) 94–102. http://dx.doi.org/10.1016/
j.jfca.2014.11.011.
[37] B. Dejaegher, Y. Vander Heyden, Experimental designs and their recent advances in
set-up, data interpretation, and analytical applications, J. Pharm. Biomed. Anal. 56
(2011) 141–158. http://dx.doi.org/10.1016/j.jpba.2011.04.023.
[38] R.L. Plackett, J.P. Burman, The design of optimum multifactorial experiments,
Biometrika 33 (1946) 305–325. http://dx.doi.org/10.2307/2332195.
[39] Y. Vander Heyden, A. Nijhuis, J. Smeyers-Verbeke, B.G.M. Vandeginste,
D.L. Massart, Guidance for robustness/ruggedness tests in method validation, J.
Pharm. Biomed. Anal. 24 (2001) 723–753. http://dx.doi.org/10.1016/S0731-
7085(00)00529-X.
[40] FDA, Food and Drug Administration, Guidance for Industry: Bioanalytical method
validation, 2001. 〈http://www.labcompliance.de/documents/FDA/FDA-Others/
Laboratory/f-507-bioanalytical-4252fnl.pdf〉.
[41] B.K. Matuszewski, M.L. Constanzer, C.M. Chavez-Eng, Strategies for the assess-
ment of matrix effect in quantitative bioanalytical methods based on HPLC-MS/
MS, Anal. Chem. 75 (2003) 3019–3030. http://dx.doi.org/10.1021/ac020361s.
[42] EMA, Guideline on bioanalytical method validation, EMEA, Comm. Med. Prod.
Hum. Use, vol. 44 pp. 1–23. 〈http://dx.doi.orgEMeA/CHMP/EWP/192217/2009〉.
[43] P. Li, M.G. Bartlett, A review of sample preparation methods for quantitation of
small-molecule analytes in brain tissue by liquid chromatography tandem mass
spectrometry (LC-MS/MS), Anal. Methods 6 (2014) 6183–6207. http://dx.doi.org/
10.1039/C4AY00915K.
[44] H. Mei, Y. Hsieh, C. Nardo, X. Xu, S. Wang, K. Ng, W.A. Korfmacher, Investigation
of matrix effects in bioanalytical high-performance liquid chromatography/tandem
mass spectrometric assays: application to drug discovery, Rapid Commun. Mass
Spectrom. 17 (2003) 97–103. http://dx.doi.org/10.1002/rcm.876.
[45] S. Souverain, S. Rudaz, J.L. Veuthey, Matrix effect in LC-ESI-MS and LC-APCI-MS
with off-line and on-line extraction procedures, J. Chromatogr. A 1058 (2004)
61–66. http://dx.doi.org/10.1016/j.chroma.2004.08.118.
[46] R.K. Saini, Y.-S. Keum, Tocopherols and tocotrienols in plants and their products: a
review on methods of extraction, chromatographic separation, and detection, Food
Res. Int. 82 (2016) 59–70. http://dx.doi.org/10.1016/j.foodres.2016.01.025.
[47] S.A. Lanina, P. Toledo, S. Sampels, A. Kamal-Eldin, J.A. Jastrebova, Comparison of
reversed-phase liquid chromatography-mass spectrometry with electrospray and
atmospheric pressure chemical ionization for analysis of dietary tocopherols, J.
Chromatogr. A 1157 (2007) 159–170. http://dx.doi.org/10.1016/j.chro-
ma.2007.04.058.