Journal of Chromatography B, 1008 (2016) 242249

Contents lists available at ScienceDirect

Journal of Chromatography B
journal homepage: www.elsevier.com/locate/chromb

Quantitative analysis of pasireotide (SOM230), a cyclic peptide, in

monkey plasma using liquid chromatography in combination with
tandem mass spectrometry
Yunlin Fu, Wenkui Li , Jimmy Flarakos, Francis L.S. Tse
Department of Drug Metabolism and Pharmacokinetics, Novartis Institutes for BioMedical Research, One Health Plaza, East Hanover, NJ 07936, USA

a r t i c l e i n f o a b s t r a c t

Article history: A novel liquid chromatographic method with tandem mass spectrometric detection (LCMS/MS) for
Received 24 March 2015 the determination of Pasireotide (SOM230) was developed and validated with a dynamic range of
Received in revised form 7 November 2015 0.5250 ng/ml using 50 l of monkey plasma. SOM230 and the internal standard, [M + 6]SOM230, were
Accepted 15 November 2015
extracted from monkey plasma via Elution SPE. The acidied sample matrix was loaded onto the pre-
Available online 1 December 2015
conditioned Waters SPE plate for further processing. The analyte was eluted from the SPE plate using
freshly prepared elution solvent, which was followed by dilution and LCMS/MS analysis. By eliminating
Keywords:
a step of evaporation of the SPE eluent, instead, injecting the eluent after a simple dilution, compound
Pasireotide
SOM230
loss due to non-specic binding to the 96-well materials was prevented. Furthermore, freshly prepared
LCMS/MS elution solution was found a key to optimal extraction recovery of the analyte from monkey plasma. The
Quantication optimal chromatographic separation was achieved on an Atlantis dC18 (50 2.1 mm, 5 m particle size)
Monkey plasma column using gradient elution with a total analysis cycle time approximately 4 min per injection. The
SPE mobile phases were water containing 0.5% acetic acid and 0.05% triuoroacetic acid (TFA) (mobile phase
Elution A) and acetonitrile containing 0.5% acetic acid and 0.05% TFA (mobile phase B). The incorporation of TFA
(0.05%, v/v) and acetic acid (0.5%, v/v) into the mobile phases was accompanied by the improved chro-
matography and minimized carryover due to the HPLC column. The current method was validated for
specicity, sensitivity, matrix effect, recovery, linearity, accuracy and precision, dilution integrity, batch
size and stability. The accuracy and precision for the LLOQs (0.5 ng/ml) were within 5.6% bias and 7.8%
CV, respectively. From the intra-day and inter-day evaluations, the precision of the other QC samples (1.5,
7.5, 75 and 190 ng/ml) ranged from 2.7 to 4.9% CV and the accuracy (% bias) from 1.3 to 7.3%, respectively.
Additional assessment of incurred sample reanalysis (ISR) was conducted to demonstrate the ruggedness
and robustness of the assay method. The validated method was successfully implemented to support a
toxicity study in monkeys administered with 5 and 30 mg of SOM230 in a single intramuscular injection
of a long acting release (LAR) formulation.
2015 Elsevier B.V. All rights reserved.

1. Introduction Pasireotide, also called SOM230 [(3S,6S,9S,12R,15S,18S,20R)-

Cushings disease is caused by a pituitary adenoma that secretes
excess corticotrophin, which triggers excessive secretion of corti-
sol in the adrenal glands [1]. The disease has been attributed as the
cause for 70% of the cases of Cushings syndrome, which is charac-
terized by a range of signs and symptoms, including central obesity,
diabetes mellitus, dyslipidemia, hypertension, muscle weakness,
hirsutism, depression and osteoporosis [2]. If left untreated, Cush-
ings disease can lead to severe complications and mortality, often
as a result of cardiovascular and cerebrovascular events [24].
Corresponding author. Fax: +1 973 781 7579.
E-mail address: wenkui.li@novartis.com (W. Li).
9-(4-aminobutyl)-3-benzyl-12-(1H-indol-3-ylmethyl)-
2,5,8,11,14,17-hexaoxo-15-phenyl-6-[(4-phenylmethoxyphenyl)
methyl]-1,4,7,10,13,16-hexazabicyclo[16.3.0]henicosan-20-yl
N-(2-aminoethyl) carbamate, Fig. 1 insert], is a subcutaneously
administered somatostatin analog. The compound binds to four
of the ve human somatostatin receptors (SSTRs) [5,6] that are
expressed by many tumor cells, including cortioutroph adenomas
and, therefore, are an important target for medical treatment
for pituitary and neuroendocrine tumors. In contrast to other
commercially available somatostatin analogues, such as octreotide
and lanreotide, which bind preferentially to SSTR subtype 2 (SSTR-
2), pasireotide binds to SSTR-1, -2, -3 and -5. While it binds to
SSTR-2 at a lower afnity (1.0 0.1 nM) as compared to octreotide
(0.4 nM) or lanreotide (0.5 nM), its binding afnity to SSTR-1

http://dx.doi.org/10.1016/j.jchromb.2015.11.025
1570-0232/ 2015 Elsevier B.V. All rights reserved.
 Y. Fu et al. / J. Chromatogr. B 1008 (2016) 242249
(9.3 0.1 nM), SSTR-3 (1.5 0.3 nM) and SSTR-5 (0.16 0.01 nM)
were, respectively, 30-, 5- and 39-fold, or 19-, 9- and 106-fold
higher than those of octreotide or lanreotide [57]. This broader
binding prole may translate into a higher efcacy with respect
to suppression of hormone production and cell growth in certain
tumors [8].
In 2012, pasireotide (trade name Signifor ) was approved in the
EU and the USA for the treatment of adult patients with Cushings
disease for whom surgery is not an option or for whom surgery has
failed [5,6]. As the rst pituitary-directed therapy for the treatment
of Cushings disease, pasireotide suppresses corticotropin secretion
from the pituitary adenoma, leading to decreased cortisol secretion
from adrenal glands. The drug signicantly decreases cortisol lev-
els and improves clinical signs and symptoms and health-related
quality-of-life [9].
To date, radioimmunoassay (RIA) has been used for quantita-
tive analysis of SOM230 in various biological samples in support of
toxicokinetic or pharmacokinetic assessment of the drug [10,11].
In order to facilitate the analysis of the compound at higher tox-
icokinetic and/or pharmacokinetic concentrations, an LCMS/MS
method has been developed and validated for the determination of
SOM230 in monkey plasma.
2. Materials and methods
2.1. Chemicals and reagents
Pasireotide (SOM230), C58 H66 N10 O9 , purity of 78.2%, was
synthesized at Novartis Pharma AG (Basel, Switzerland). The
Fig. 1. Q1 scan MS spectrum (top panel) and Q3 product ion MS spectrum (bottom panel) of SOM230 under positive electrospray ionization.
243
internal standard, [M + 6]SOM230, 13 C6 C52 H66 N10 O9 (mass
purity > 99%) was synthesized in-house. HPLC grade solvents,
acetonitrile, methanol, DMSO and isopropanol, were purchased
from SigmaAldrich (Louis, MO, USA), and reagents, formic acid,
TFA, ammonium hydroxide, and acetic acid, were purchased
from Fisher (Fair Lawn, NJ, USA). Deionized water was produced
in-house via an ELGA PureLab Ultra Water System (Lowell, MA,
USA). Monkey plasma with K2 -EDTA as the anticoagulant was
obtained from Bioreclamation (Westbury, NY, USA).
2.2. Chromatographic conditions
A HPLC system consisting of a multi-channel degasser, one
autosampler, two HPLC pumps, a controller and a column oven
from Shimadzu Corp. (Columbia, MD, USA), and an Waters Atlantis
dC18 (50 2.1 mm, 5 m particle size) column (Milford, MA, USA)
was used for the chromatographic separation of SOM230 and the
internal standard. The mobile phases were water (containing 0.5%
acetic acid and 0.05% TFA, mobile phase A) and acetonitrile (con-
taining 0.5% acetic acid and 0.05% TFA, mobile phase B). Optimal
chromatographic resolution of SOM230 and the internal standard
from the matrix components was achieved by running 10% mobile
phase B for 0.2 min and then from 10% to 90% B in 2.3 min, held at
90% B for the next 0.5 min, followed by 10% B for the next 1 min
before the subsequent injection cycle. The total analysis cycle time
was about 4 min. The ow rate was 0.6 ml/min and the entire col-
umn efuent was directed to the mass spectrometer interface from
1 to 2.5 min without splitting.
244 Y. Fu et al. / J. Chromatogr. B 1008 (2016) 242249
Fig. 2. A comparison of carryover observed using 0.1% formic acid (2A) vs. 0.05% TFA and 0.5% acetic acid (2B) as the mobile phase modier(s) for the LCMS/MS analysis of
blank monkey plasma extracts injected immediately after the ULOQ (250 ng/ml) calibration standard.
2.3. Mass spectrometric conditions
A Sciex Qtrap6500 mass spectrometer (Concord, ON, CA) with
a TurboIonSpray (TIS) interface operated in the positive ioniza-
tion mode was used for the multiple reaction monitoring (MRM)
LCMS/MS analysis. The mass spectrometric conditions were opti-
mized for SOM230 and [M + 6]SOM230, respectively, by infusing
a 10 g/ml of the standard solution in 50% acetonitrile (contain-
ing 0.1% formic acid) at 10 l/min using a built-in infusion pump
directly connected to Tee, where the compounds were mixed
with the mobile phase at a ow rate of 0.6 ml/min before entering
the mass spectrometer. The optimized instrument conditions were
as follows: TIS source temperature, 600 C; TIS voltage, 5500 V; cur-
tain gas, 40 units; nebulizing (GS1), 70 units; TIS (GS2) gas, 80
units; CAD, medium; collision energy (CE), 23 eV for both SOM230
and [M + 6]SOM230. The precursor product ion transitions of m/z
524 472 and m/z 527 475 were used for multiple reaction
monitoring (MRM) of SOM230 and [M + 6]SOM230, respectively,
with a dwell time of 100 ms for each mass transition. The mass
spectrometer was operated at unit mass resolution (peak width at
half-height of 0.7 Da) for both the rst quadrupole (Q1) and the
third quadrupole (Q3).
2.4. Standards and quality control (QC) samples
Two primary stock solutions of SOM230 were each prepared in
DMSO at concentration of 1 mg/ml in 5-ml amber vials. The stock
solutions were stored at 28 C. For validation and sample analy-
sis purposes, the stock solutions from the two separate weighings
must have less than a 5% difference in the LCMS/MS responses.
An intermediate containing 10,000 ng/ml of SOM230 was prepared
by appropriate dilution of the stock solution with blank monkey
plasma. This intermediate was serially diluted with blank monkey
plasma to prepare the standards and QCs at the desired concentra-
tions. The concentrations of SOM230 calibration standard samples
were 0.5, 1, 2, 4, 20, 100, 200 and 250 ng/ml. QC samples were
prepared at concentrations of 0.5 (LLOQ), 1.5 (LQC), 7.5 (MQC), 75
(MQC) and 190 (HQC) ng/ml. Dilution QCs were prepared at a con-
centration of 2500 ng/ml. All QC samples were transferred into 2-ml
polypropylene vials and stored at 60 C. The internal standard
working solution containing 10 ng/ml of [M + 6]SOM230 was pre-
pared from the ISTD stock solution (1 mg/ml) with 50% acetonitrile
in water (v/v). A
2.5. Sample preparation
A 50 l volume of blank monkey plasma or freshly prepared
calibration standards, QC samples, or unknown samples was added
to a 1-ml 96-well plate. A 25 l aliquot of the internal standard
working solution (10 ng/ml in 50% acetonitrile in water, v/v) was
added to each well except for the matrix blank, to which a 25 l
aliquot of 50% acetonitrile in water (v/v) was added. After adding
100 l of 2% formic acid in water (v/v) to each well, the plate was
vortex-mixed briey and centrifuged. The sample mixtures in the
above plate were loaded onto the corresponding wells of a Waters
Oasis WCX (weak cation exchange) Elution plate (96-well format,
30 mg, Milford, MA, USA), which were pre-equilibrated with 200 l
of methanol, followed by 200 l of water. The plate was washed
with 200 l of 5% ammonium hydroxide in water (v/v), followed
by 200 l of methanol. Finally, the Elution plate was eluted with
100 l of 5% formic acid in methanol (v/v). The eluent was diluted
with 50 l of water. After brief vortex-mixing and centrifugation
at 3000 g for 5 min, a 10 l of the diluted sample extract was
injected onto the LCMS/MS system.
 Y. Fu et al. / J. Chromatogr. B 1008 (2016) 242249
Fig. 3. Representative LCMS/MS chromatograms of blank monkey plasma (double blank, A), blank monkey plasma spiked with internal standard only (zero sample, B) and
blank monkey plasma spiked with SOM230 at the LLOQ (0.5 ng/ml) level (C).
2.6. Data analysis
LCMS/MS peak integration for SOM230 and the internal stan-
dard was performed using the Sciex Analyst, version 1.6 software
(Foster City, CA, USA). Data processing was conducted using Wat-
son LIMS, version 7.4.2 (Thermo Fisher Scientic, Philadelphia, PA,
USA). The calibration curves (analyte peak area/IS peak area vs. ana-
lyte concentration) were constructed using the least squares linear
regression t (y = ax + b), and a weighting factor of 1/x2 was applied
to the regression. Acceptance criteria were established to be >0.99
for the calibration curve coefcient of correlation (r2 ), and within
15% bias of the nominal concentration (accuracy) and 15% CV
(precision) of the mean measured values for QC samples except
the LLOQ samples in the validation assays. For LLOQ samples, the
assay accuracy and precision should be within 20% bias of the
nominal values and 20% CV of the mean measured values.
).
3. Results and discussion
3.1. Method development
The positive ion electrospray full scan MS spectrum of SOM230
is shown in Fig. 1 A. The most abundant ion was the doubly charged
245
protonated molecule at m/z 524. The product-ion MS spectrum of
SOM230 is shown in Fig. 1B, to which the proposed product ion
was inserted. The most abundant product ion was observed at m/z
472, which is the doubly charged molecule following the loss of
the benzyl alcohol moiety. It is noteworthy that collision energy
(CE) was critical for the abundance of this product ion with opti-
mal product ion response observed at a CE value of 23 eV. A little
lower CE setting (e.g., 21 eV) resulted in almost no fragmentation
of the doubly charged parent ions. However, a slight increase in CE
setting (e.g., 25 eV) led to over-fragmentation as evidenced by the
appearance of many poorly resolved fragment ions with low abun-
dance. The initial LCMS/MS experiments were carried out with
mobile phases only containing 0.1% formic acid. It was noticed that
after multiple consecutive injections, the peak shapes started dete-
riorating and the retention time began shifting for both SOM230
and the internal standard. Furthermore, although multiple injector
washes were incorporated in the assay method by using two wash
solvents (wash solvent 1: acetonitrile/isopropanol/water/acetic
acid/formic acid at 60/30/10/0.1/0.1, v/v/v/v/v, and wash solvent
2: acetonitrile/isopropanol/water, 20/20/60, v/v/v), signicant car-
ryover (peak area > 20% of the LLOQ for SOM230 and >5% for the
internal standard at the working concentration) was observed
(Fig. 2A). This carryover was conrmed to be due to column rather
than the autosampler. In order to improve the performance of the
246 Y. Fu et al. / J. Chromatogr. B 1008 (2016) 242249

Table 1
Summary of assay precision and accuracy for SOM230 in monkey plasma QC samples.

SOM230 nominal concentration (ng/ml) 0.500 1.50 7.50 75.0 190

Calculated concentrations (ng/ml)

Day- 0.501 1.69 7.78 81.9 204
1 0.540 1.57 7.70 80.9 191
0.489 1.63 7.92 83.0 210
0.482 1.61 8.00 79.9 201
0.543 1.68 7.75 78.3 208
0.589 1.49 8.28 77.0 198
Intrarun mean 0.524 1.61 7.91 80.2 202
Intrarun% CV 7.8 4.6 2.7 2.8 3.4
Intrarun% bias 4.8 7.3 5.5 6.9 6.3
n 6 6 6 6 6

Day- 0.502 1.60 7.91 78.9 182

2 0.472 1.51 8.10 78.5 186
0.478 1.52 8.07 79.0 198
0.499 1.39 8.29 80.1 186
0.450 1.54 8.13 83.3 191
0.429 1.45 7.46 83.3 196
Intrarun mean 0.472 1.50 7.99 80.5 190
Intrarun% CV 6.0 4.9 3.6 2.8 3.3
Intrarun% bias -5.6 0.0 6.5 7.3 0.0
n 6 6 6 6 6

Day- 0.536 1.48 7.80 76.1 182

3 0.529 1.56 7.40 80.1 192
0.470 1.50 7.12 83.1 189
0.474 1.51 7.46 76.6 194
0.489 1.47 7.34 80.5 182
0.503 1.36 7.50 79.9 192
Intrarun mean 0.500 1.48 7.44 79.4 189
Intrarun% CV 5.5 4.5 3.0 3.3 2.8
Intrarun% bias 0.0 1.3 0.8 5.9 0.5
n 6 6 6 6 6

Overall mean 0.499 1.53 7.78 80.0 193

Inter-run% CV 7.6 5.9 4.4 2.9 4.4
Inter-run% bias 0.2 2.0 3.7 6.7 1.6
n 18 18 18 18 18

Table 2
Summary of stability assessment results for SOM230 in monkey plasma QC samples (n = 4).

Storage period Nominal concentration (ng/ml) Calculated concentrations (ng/ml) Mean bias (%)

24 h on laboratory bench top 1.50 1.48, 1.50, 1.59, 1.54 2.0

190 186, 183, 191, 176 3.2

3 freeze/thaw cycles (stored at 60 C) 1.50 1.45, 1.48, 1.41, 1.57 1.3

190 183, 189, 183, 178 3.7

3 days in the reconstituted monkey plasma sample extracts stored in 1.50 1.53, 1.62, 1.66, 1.72 8.7
autosampler (at 5 C) 7.50 7.72, 7.53, 7.72, 7.41 1.3
75.0 78.5, 76.7, 77.4, 78.7 3.7
190 203, 205, 208, 201 7.4

55 days 1.50 1.41, 1.48, 1.48, 1.47 2.7

(at 60 C) 7.50 7.57, 7.11, 7.59, 7.45 0.9
75.0 75.0, 69.3, 75.7, 75.2 1.6
190 176, 187, 195, 183 2.6

chromatography and eliminate the carryover, an ion-pair reagent,
TFA, was added at a concentration of 0.05% (v/v) to both mobile
phase A and mobile phase B. In the meanwhile, to minimize the
MS signal suppression anticipated by the addition of TFA to the
mobile phases, acetic acid (v/v) at a nal concentration of 0.5% was
incorporated into both mobile phases [12]. As a result, symmet-
ric peak shape and reproducible retention time were obtained for
the analyte and the internal standard. No carryover was observed
Fig. 2B).
Peptides often show greater non-specic binding or container
surface adsorption compared to many other small molecules [13].
The non-specic binding is particularly noticeable for the cali-
bration standard working solutions prepared with solvent at low
percentage of organic, and is often evidenced by the unexpected
low extraction recovery of the analyte, non-linearity of the calibra-
tion curves and/or highly variable QC sample results. In the current
experiment, the calibration standards and quality control samples
were prepared from the intermediate by spiking the appropriate
volume of the stock solution directly into blank monkey plasma.
This was followed by serial-dilution of the intermediate using blank
monkey plasma for the preparation of calibration standards and
QCs. Solid phase extraction (SPE) via Waters Oasis WCX plate was
selected for sample extraction to separate both the analyte and
internal standard from bulk of matrix components. A common WCX
SPE protocol was implemented in the sample preparation process,
i.e., (1) acidied plasma sample matrix was uploaded onto the plate
which was preconditioned using methanol and then water, (2) the
plate was washed using 5% ammonium hydroxide in water (v/v)
 Y. Fu et al. / J. Chromatogr. B 1008 (2016) 242249 247

1000 5 mg dose
30 mg dose

Plasma SOM230 concentraon (ng/ml) 100

10

1
0 500 1000 1500 2000 2500
Time post dose (h)

Fig. 4. Monkey plasma SOM230 concentrations vs. time proles after a single intramuscular injection of 5 and 30 mg of SOM230 in long acting release (LAR) formulation.

followed by methanol, and (3) analyte and the internal standard 3.2. Method validation
was eluted using 5% formic acid in methanol. Interestingly, a sig-
nicant decrease in the LCMS/MS response of the analyte was The full validation of the present method was performed accord-
observed when the SPE eluent was evaporated and the resulting ing to the FDA Guidance for Industry on Bioanalytical Method
residue reconstituted prior to LCMS/MS analysis. At both the QC Validation [14] and internal standard operation procedures. The
low (1.5 ng/ml) and high (190 ng/ml) levels each evaluated in three dynamic range validated was from 0.5 to 250 ng/ml based on 50 l
replicates, the measured LCMS/MS peak areas of SOM230 in the sample volume.
reconstituted sample extracts (in 150 l of 75% aqueous methanol,
v/v) after solvent evaporation were about 50% less than those mea- 3.2.1. Specicity and selectivity.
sured in the corresponding eluents diluted with water (nal volume Under the current LCMS/MS conditions, SOM230 was
of 150 l at 75% methanol) (Figure not shown). Apparently, the separated from interferences in blank matrix. LCMS/MS chro-
loss of the analyte in the former was due to non-specic binding matograms of two lots of blank monkey plasma were found to
of the analyte and the internal standard to the wells of the 96-well contain no co-eluting endogenous peak with the analyte or internal
plate and, however, the reconstitution solvent (75% methanol in standard. Representative chromatograms of blank monkey plasma
water, v/v) was not strong enough to reverse the binding when sam- samples without (blank) or with internal standard (zero sample)
ple extract became dried. While the use of higher organic solvent are shown in Fig. 3 and B, respectively. The small peak of SOM230
(e.g., 100% methanol) may help reverse the non-specic binding to in zero sample has been conrmed to be the impurity in the
some extent, such a high organic solvent was found detrimental to
the chromatography. This was evidenced by severe front peak tail-
ing for both the analyte and internal standard (details not shown). Table 3
Taking this into consideration, the evaporation step of the resulting Summary of incurred sample reanalysis (ISR) results for 21 randomly selected mon-
key plasma samples.
SPE eluent was omitted. Instead, the eluent was directly subjected
to LCMS/MS analysis after a very simple dilution using water with Sample index Initial conc. (ng/ml) ISR conc. (ng/ml) %Bias
a nal water content of 25%. 1 83.8 81.9 2.3
2 10.2 9.47 7.4
On the other hand, it was noticed that the overall extraction
3 10.2 10.4 1.9
recovery was signicantly decreased upon using aged elution solu- 4 56.8 60.8 6.8
tion in the nal step of sample preparation. In an experiment, 5 97.0 103 6.0
freshly prepared elution solution was compared to those that had 6 258 294 13.0
been stored at room temperature for up to 6 days after prepara- 7 33.6 35.6 5.8
8 124 118 5.0
tion. The measured LCMS/MS peak areas of the anlayte in the 9 408 446 8.9
eluent using the freshly prepared elution solvent was, respectively, 10 612 575 6.2
130, 166, 210 and 300% better that those measured using the elu- 11 424 431 1.6
tion solvent that had been stored for 1, 2, 3 and 6 days. Although 12 74.0 77.8 5.0
13 880 883 0.3
the cause of the difference remains unknown, it is clear that the
14 339 350 3.2
use of freshly prepared elution solution is necessary for optimal 15 197 183 7.4
extraction recovery. 16 137 123 10.8
17 26.2 27.7 5.6
18 140 143 2.1
19 36.8 39.0 5.8
20 9.40 9.73 3.5
21 44.1 44.0 0.2
248 Y. Fu et al. / J. Chromatogr. B 1008 (2016) 242249
[M + 6]SOM230 reference material (Fig. 3B). Injection of SOM230
at the highest concentration (250 ng/ml) did not show signi-
cant interference (<2% of the internal standard response) at the
[M + 6]SOM230 channel (gure not shown).
3.2.2. Sensitivity
The current assay has a LLOQ of 0.5 ng/ml of SOM230 in
monkey plasma based on a 50 l plasma volume. Reliable preci-
sion (CV% 7.8%) and accuracy (bias within 5.6%) results were
obtained by analyzing six replicate LLOQ samples along with the
respective calibration standards and regular QCs in three separate
validation runs (Table 1A representative LCMS/MS chromatogram
of the LLOQ sample (0.5 ng/ml) is shown in Fig. 3C.
3.2.3. Matrix effect and recovery
The matrix effect was estimated by spiking the analyte neat
solutions into blank monkey plasma sample extracts with SOM230
at 1.5 (low), 75 (mid) and 190 (high) ng/ml (n = 3) concentration
levels. The matrix factors were calculated by comparing the mean
peak areas of SOM230 in these post-spiked sample extracts with
those in the corresponding neat solutions. The overall matrix factor
is 1.071.18 for SOM230 across the low, mid and high concentra-
tion range, indicating that the current sample preparation method
is suitable for the intended studies. The recovery was estimated
by comparing the mean peak areas of the analyte in the extracted
QC samples at 1.5 (low), 75 (mid) and 190 (high) ng/ml (n = 3) with
those obtained from extracted blank monkey plasma samples post-
spiked with the corresponding neat solutions. The overall recovery
is 52.057.2% for SOM230 (details not shown).
3.2.4. Linearity
Over a calibration standard concentration range of
0.5250 ng/ml evaluated in the validation, the peak area ratio
of the analyte to the internal standard was tted to a weighted
linear equation (weighted 1/x2 ). The correlation coefcients (r2 )
obtained for all individual calibration curves were better than
0.990. Back-calculated analyte concentrations and statistics for
calibration standards (table not shown) indicated that the curve t
was appropriate, and the method was accurate (bias within 20%
for LLOQs and 15% for all other QCs) and precise (CV 20% for the
LLOQs and 15% for all other QCs) over the course of validation.
3.2.5. Precision and accuracy
Three main validation batches were used to assess the precision
(CV%) and accuracy (Bias%) of the method. Each validation batch
was processed on a separate day and had two sets of calibration
standards and six replicates of QCs at all concentration levels. The
QC samples and other test samples were interspersed between the
two sets of calibration curves. The accuracy and precision of the
method were determined by analyzing six replicates of QC sam-
ples at 1.5 (low), 7.5 (mid), 75 (mid) and 190 (high) ng/ml in each
of the three validation runs. The accuracy of the method was deter-
mined by calculating bias (%) and the precision by calculating CV(%).
Table 1 summarizes the accuracy and precision on each of the three
validation runs. The obtained accuracy ranged from 1.3% to 7.3%
(bias) and precision from 2.7% to 4.9% (CV).
3.2.6. Dilution integrity.
The dilution integrity was evaluated by 20-fold dilution of the
dilution QCs at a concentration of 2500 ng/ml with blank monkey
plasma prior to extraction and assayed in six replicates along with
calibration curve and regular QCs in a validation run. The obtained
precision and accuracy of the dilution QCs was 3.5% (CV) and 2.8%
(bias) (Table 1), respectively, demonstrating that samples with con-
centration higher than the upper limit of quantication (ULOQ) can
be analyzed to obtain acceptable concentration data after dilution
with blank plasma.
3.2.7. Stability
The bench-top stability of SOM230 in monkey plasma was eval-
uated at ambient temperature (22 C) over 24 h using QC samples
at 1.5 (low) and 190 (high) ng/ml. The measured concentrations of
SOM230 in these QC samples were compared to the nominal values.
The observed bias (%) ranged from 3.2 to 2.0% (Table 2), indicating
that SOM230 was stable for at least 24 h in monkey plasma when
stored at ambient temperature. Freezethaw stability of QC sam-
ples at 1.5 (low) and 190 (high) ng/ml, experiencing three cycles
of freezethaw, were analyzed along with calibration standards
and regular QC samplers. The bias (%) ranged from 3.7% to 1.3%
(Table 2).
During the validation, one of the validation batches was stored
in the HPLC autosampler for over 3 days and then re-analyzed
and quantied. The bias (%) values of these processed QC samples
ranged from 1.3 to 8.7% bias (Table 2), demonstrating that extracted
samples could be analyzed after standing in the HPLC autosam-
pler (5 C) for at least 3 days with acceptable assay precision and
accuracy.
The long term stability of SOM230 in monkey plasma was
evaluated by analyzing frozen QC samples in four replicates at con-
centrations of 1.5, 7.5, 75 and 190 ng/ml along with calibration
standards and regular QCs freshly prepared from stock solution.
The measured SOM230 concentrations in the above frozen QC sam-
ples were compared with nominal values with the mean bias (%) of
2.7 to 0.9% after at least 55 days of storage at 60 C (Table 2).
3.2.8. Batch size
The batch size was dened as a group of calibration standards,
matrix blanks, zero samples, reagent blanks, quality controls, and
study samples that were processed through the procedures as
described in Section 2, analyzed through sequential LCMS/MS
injections, and calculated based on the calibration curve. As a part
of validation, a total of 192 samples including calibration standards,
quality controls, blanks, zero samples and unknown samples were
analyzed. The precision and accuracy were, respectively, 2.34.3%
(CV) and 0.92.7% (bias) (details not shown), indicating that the
batch size could go up to 192 injections for study sample analysis
with acceptable assay accuracy and precision.
3.2.9. Incurred sample reanalysis
As part of method robustness evaluation, incurred sample
reanalysis (ISR) was conducted for one set of study samples col-
lected from a monkey study. For ISR, BLLOQ samples or samples
with initial measured analyte concentrations less than three times
of LLOQ (0.5 ng/ml) were not selected and 2/3 of the differences (%)
between the initial values and the mean of the initial and repeat
values must be within 20%. As shown in Table 3, all 21 ISR results
met the above acceptance criteria with the difference (%) within
10% for 19 out of 21 ISR results and within 15% for the remain-
der. This evaluation further demonstrated the robustness of the
current assay method.
4. Application
The present LCMS/MS method was implemented in measur-
ing SOM230 concentrations in the plasma samples collected from
monkeys in a toxicity study with a single intramuscular injection of
a long acting release (LAR) formulation. Illustrated in Fig. 4 are the
mean (n = 6 or 10) SOM230 plasma concentration vs. time proles
up to 13 weeks after doses of 5 and 30 mg of the compound. The
present method allowed for the determination of SOM230 in mon-
key plasma up to 27 days after a 5 mg dose and up to 10 weeks (70
 Y. Fu et al. / J. Chromatogr. B 1008 (2016) 242249
days) after a 30 mg dose. Overall, an increase in SOM230 intramus-
cular dose was accompanied with an approximately proportional
increase in the measured monkey plasma exposure.
5. Conclusions
In conclusion, an LCMS/MS method was developed and vali-
dated for quantitative analysis of SOM230 in monkey plasma with
a dynamic range of 0.5250 ng/ml based on 50 l sample volume.
The addition of ion-pairing reagent, TFA (0.05%, v/v), to the mobile
phases signicantly improved the chromatograpy and minimized
carryover on the silica based reversed-phase column used in this
study. Solid-phase extraction with Elution plate was found useful
in extracting the compound from the matrix with an overall extrac-
tion recovery more than 50%. By eliminating a step of evaporation
of the SPE eluent, compound loss due to non-specic binding of the
96-well materials was alleviated. Furthermore, freshly prepared
elution solution was found a key to optimal recovery of the ana-
lyte from monkey plasma. SOM230 was demonstrated to be stable
under various conditions of sample storage. The bioanalytical batch
size was validated to be up to 192 injections per run sequence. The
assessment of incurred sample reanalysis (ISR) further conrmed
that the assay method is rugged and robust.
References
[1] B.M. Biller, A.B. Grossman, P.M. Stewart, S. Melmed, X. Bertagna, J. Bertherat,
M. Buchfelder, A. Colao, A.R. Hermus, L.J. Hoand, A. Klibanski, A. Lacroix, J.R.
Lindsay, J. Newell-Price, L.K. Nieman, S. Petersenn, N. Sonino, G.K. Stalla, B.
Swearingen, M.L. Vance, J.A. Wass, M. Boscaro, Treatment of
adrenocorticotropin-dependent Cushings syndrome: a consensus statement,
J. Clin. Endocrinol. Metabol. 93 (2008) 24542462.
[2] R.A. Feelders, S.J. Pulgar, A. Kempel, A.M. Pereira, The burden of Cushings
disease: clinical and health-related quality of life aspects, Eur. J. Endocrinol.
167 (2012) 311326.
[3] J. Lindholm, S. Juul, J.O. Jrgensen, J. Astrup, P. Bjerre, U. Feldt-Rasmussen, C.
Hagen, J. Jrgensen, M. Kosteljanetz, L. Kristensen, P. Laurberg, K. Schmidt, J.
Weeke, Incidence and late prognosis of cushings syndrome: a
population-based study, J. Clin. Endocrinol. Metabol. 86 (2001) 117123.
249
[4] Y. Reznik, J. Bertherat, F. Borson-Chazot, T. Brue, P. Chanson, C. Cortet-Rudelli,
B. Delemer, A. Tabarin, S. Bisot-Locard, B. Vergs, Management of
hyperglycaemia in Cushings disease: experts proposals on the use of
pasireotide, Diabetes Metabol. 39 (2013) 3441.
[5] European Medicines Agency. Signifor solution for injection: summary of
product characteristics. 2012. http://www.ema.europa.eu/docs/en GB/
document library/EPAR Product Information/human/002052/
WC500128056pdf.
[6] Novartis Pharmaceuticals Corporation. Signifor: full prescribing information.
2012. http://www.pharma.us.novartis.com/product/pi/pdf/signifor.pdf.
[7] C. Bruns, I. Lewis, U. Briner, G. Meno-Tetang, G. Weckbecker, SOM230: a novel
somatostatin peptidomimetic with broad somatotropin release inhibiting
factor (SRIF) receptor binding and a unique antisecretory prole, Eur. J.
Endocrinol. 146 (2002) 707716.
[8] R.A. Feelders, W.W. de Herder, S.J. Neggers, A.J. van der Lely, L.J. Hoand,
Pasireotide: a multi-somatostatin receptor ligand with potential efcacy for
treatment of pituitary and neuroendocrine tumors, Drugs Today 49 (2013)
89103.
[9] K. McKeage, Pasireotide: a review of its use in Cushings disease, Drugs 73
(2013) 563574.
[10] T.H. Lin, K. Hu, J. Flarakos, M. Sharr-McMahon, J.B. Mangold, H. He, Y. Wang,
Assessment of the absorption: metabolism and excretion of [14 C]pasireotide
in healthy volunteers using accelerator mass spectrometry, Cancer
Chemother. Pharmacol. 72 (2013) 181188.
[11] B. Mariniello, I. Finco, P. Sartorato, A. Patalano, M. Iacobone, V. Guzzardo, A.
Fassina, F. Mantero, Somatostatin receptor expression in adrenocortical
tumors and effect of a new somatostatin analog SOM230 on hormone
secretion in vitro and in ex vivo adrenal cells, J. Endocrinol. Invest. 34 (2011)
e131e138.
[12] W. Li, S. Luo, S. Rebello, J. Flarakos, F.L. Tse, A semi-automated LCMS/MS
method for the determination of LCI699, a steroid 11-hydroxylase inhibitor,
in human plasma, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 960
(2014) 182193.
[13] E.E. Chambers, M.E. Lame, J. Bardsley, S. Hannam, E. Collins, K.J. Fountain, E.
Thomas. Development of a quantitative SPE-LC/MS/MS assay for teriparatide
in human plasma. www.waters.com/webassets/cms.
[14] U.S. Department of Health and Human Services Food and Drug
Administration, Center for Drug Evaluation and Research (CDER) and Center
for Veterinary Medicine (CVM), Guidance for Industry Bioanalytical Method
Validation, May 2001.