Purohit 2019

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

JID: CHROMA

ARTICLE IN PRESS [m5G;October 14, 2019;22:9]


Journal of Chromatography A xxx (xxxx) xxx

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Simple and reliable extraction and a validated high performance liquid


chromatographic assay for quantification of amoxicillin from plasma
Trusha J. Purohit, Zimei Wu, Sara M. Hanning∗
School of Pharmacy, The University of Auckland, 85 Park Road, Grafton, Auckland 1010, New Zealand

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

Article history: This study presents the development of an efficient extraction protocol for amoxicillin from plasma with
Received 2 July 2019 improved solubility and stability using pH control. Solubility and stability of amoxicillin in commonly
Revised 6 October 2019
used extraction solvents were determined using a newly developed stability-indicating high-performance
Accepted 8 October 2019
liquid chromatography (HPLC) method. Following this, protein precipitation (PP) mediated sample purifi-
Available online xxx
cation protocol was developed and validated along with the HPLC method for the extracted amoxicillin
Keywords: from rabbit plasma. The protocol was applied in a pharmacokinetic study in rabbits. A five-fold increase in
Amoxicillin solubility and two-fold increase in stability of amoxicillin was found by addition of acetate buffer (0.1 M,
Stability and solubility pH 5.0) in acetonitrile. PP mediated extraction protocol containing acetate buffer-acetonitrile (1:18 v/v)
Protein precipitation resulted in an extraction recovery of >80% for all the samples. The HPLC assay following extraction
Quantitative analysis was found linear (R2 >0.9999) over the range of 0.2–20 μg/mL with a lower limit of quantification of
Plasma samples
0.2 μg/mL. The accuracy of the quality control samples was found between 97–115% and the relative stan-
Rabbit pharmacokinetics
dard deviation (RSD) was found to be below 6% for all samples. The samples were stable in the mobile
phase (pH 5.0) for 72 h post-extraction. Amoxicillin-spiked plasma samples were found stable for up to
three freeze-and-thaw cycles but, nearly 50% samples had degraded following storage for two months
at −20 °C. Pharmacokinetic analysis indicated a half-life of amoxicillin of nearly 1 h following intravenous
injection in rabbits, which is similar to that in humans. Thus, a simple and repeatable, extraction protocol
was developed using pH control for quantification of amoxicillin from plasma based on its physicochem-
ical properties.
© 2019 Elsevier B.V. All rights reserved.

1. Introduction removes the proteins by lowering the solubility of proteins in the


plasma due to change in dielectric constant or isoelectric pH of the
Extraction and quantitative analysis of a drug from biological proteins. Organic solvent mediated protein precipitation technique
samples such as plasma plays an important role in pharmaceuti- has been reported to be the most common and efficient amongst
cal research. Plasma contains a variety of endogenous substances all different methods [4]. However, development of a specific pu-
including a large number of proteins, which need to be removed rification method with high extraction efficiency is challenging es-
before chromatographic drug analysis [1,2]. Incomplete removal of pecially for chemically unstable molecules with low solubility, such
proteins from plasma samples leads to compromised column per- as amoxicillin.
formance and low sensitivity of an analytical method [1,3]. It is a Amoxicillin is a broad spectrum β -lactam antibiotic used in the
complex process involving exposure of a drug to a rigorous sepa- treatment of a variety of infections. It is an amphoteric molecule
ration process and a variety of solvents such as acids and organic having a log P value of 0.87 and pKa values of 3.2 and 11.7 [6].
solvents [4]. It has a “U-shaped” pH-dependent solubility profile in aqueous
Different purification methods such as extraction (liquid-liquid media with lowest solubility between pH 4–6 [7]. In contrast, a
or solid-phase) and protein precipitation (PP) have been developed pH-degradation profile with optimal stability between pH 4–6 has
for the analysis of biological samples [5] with PP being the most been reported [2,7]. As it degrades at extreme pH, acids cannot be
common and simplest amongst all. PP can be carried out by addi- used to extract amoxicillin from biological samples [2]. Moreover,
tion of either organic solvents, acids, salts or metal ions. It mainly amoxicillin readily undergoes methanolysis in presence of alcohols
such as methanol [8,9] and is practically insoluble in acetonitrile,
two of the most commonly used and efficient organic solvents for

Corresponding author. protein precipitation [10].
E-mail address: s.hanning@auckland.ac.nz (S.M. Hanning).

https://doi.org/10.1016/j.chroma.2019.460611
0021-9673/© 2019 Elsevier B.V. All rights reserved.

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611
JID: CHROMA
ARTICLE IN PRESS [m5G;October 14, 2019;22:9]

2 T.J. Purohit, Z. Wu and S.M. Hanning / Journal of Chromatography A xxx (xxxx) xxx

There are numerous high performance liquid chromatography This chromatographic method was used in solubility and stability
(HPLC) methods available for direct and indirect quantification studies as well as plasma sample analysis following extraction.
of amoxicillin in biological matrix [2,11–15]. Previously described
methods have involved solid phase extraction (SPE) [16], liquid- 2.3. Solubility and stability analysis of amoxicillin
liquid extraction (LLE) [11,17] and PP [18–22]. SPE and LLE shows
minimum matrix effect but they are often complex and time con- The solubility of AMT was determined in commonly used ex-
suming [23] and the extraction efficiency depends upon various traction solvents, with water as control. Acetonitrile was mixed
parameters such as type of extraction cartridges, solubility and par- with 0.1% formic acid (pH 2.5), acetate buffer (0.1 M, pH 5) and
titioning in different solvents [1,23]. In contrast, PP mediated ex- phosphate buffer (0.1 M, pH 7.4) to understand the effect of pH on
traction is simple and quick, which makes it the most common solubility and stability of amoxicillin. Water, acetonitrile, and ace-
method of extraction. Furthermore, PP with an organic solvent is tonitrile mixed with either water, methanol, acetate buffer (0.1 M,
considered the most effective extraction method for quantitative pH 5.0), or phosphate buffer (0.1 M, pH 7.4) at a ratio of 95:5 v/v
analysis of a drug with a low log P value from plasma samples were investigated. Rabbit plasma was also added to the binary sys-
[24]. Previously, quantification of amoxicillin from human plasma tems (90:5:5). Excess amount of AMT (>12 mg/mL) was added to
by external standard method has been reported using methanol as each of the above solvents and the samples were maintained on a
an extraction solvent [25]. However, our preliminary trials for ex- shaking water-bath at 25 °C until equilibrium was achieved (24 h).
traction of amoxicillin by standard addition resulted in varying and The solubility in different solvents was determined from the super-
low extraction recovery, possibly due to poor stability. This neces- natants using a stability-indicating HPLC method described above.
sitated pH-control by addition of buffer to maintain solubility and Systems showing solubility of amoxicillin above 120 μg/mL were
stability of amoxicillin during the extraction process. This led to further selected for stability analysis, where the supernatant was
development of a PP mediated protocol for extraction of amoxi- split and stored, protected from light, at either 4 °C or ambient
cillin from plasma samples by understanding its solubility and sta- temperature (22 °C). Stability samples (10μL) were taken at pre-
bility in the extraction solvent(s) prior to development of an ex- determined time intervals for up to 21 days or until 80% degra-
traction protocol. dation of the drug was observed, whichever came first. The degra-
In this paper, the solubility and stability of amoxicillin in com- dation kinetics were further studied by application of a first order
monly used solvent systems during extraction was first studied us- kinetic model. Degradation rate constant (observed) (kobs ), half-life
ing a stability-indicating HPLC method. An optimal extraction pro- (t1/2 ) and the time at which the drug retained 90% of its origi-
tocol for amoxicillin from plasma was developed in the selected nal concentration (t90 ) were determined using the following equa-
solvent system by leveraging its stability and solubility. The HPLC tions:
method for quantification of amoxicillin from plasma was validated 0.693
and applied in the analysis of a preliminary pharmacokinetic study t1/2 = (1)
kobs
following intravenous (IV) drug administration to rabbits.
and
0.105
2. Materials and methods t90 = (2)
kobs
2.1. Materials 2.4. Extraction protocol

Amoxicillin trihydrate (AMT) – European Pharmacopoeia stan- To develop an optimised extraction protocol, initially various
dard was purchased from Sigma Aldrich, New Zealand. Formic acid, extraction solvent systems were screened for efficient and repeat-
sodium acetate, sodium/potassium dihydrogen phosphate, and dis- able extraction of amoxicillin from plasma by PP. Solvent systems
odium hydrogen phosphate were of reagent grade and purchased containing 100, 80, 70, 60, 50% acetonitrile in methanol, acidified
from Merck and ThermoFisher Scientific, New Zealand. Acetoni- methanol or acetonitrile with either 0.1% formic acid [21, 22] or
trile and methanol used in the extraction process and liquid chro- 10–20% perchloric acid [26, 27] were evaluated for extraction ef-
matography were HPLC grade. The Milli-Q water for HPLC was ficiency. Following the solubility and stability studies, extraction
obtained from Millipak (Millipore, 0.22 μm). Plasma was obtained protocols were developed using the acetate buffer (0.1 M, pH 5.0)
from healthy rabbits (The University of Auckland Animal Ethics - acetonitrile binary system (1:18) for extraction of amoxicillin
Committee reference number 001,663) and for the pharmacoki- from plasma samples. The protocol was evaluated by applying one-
netic study, New Zealand White rabbits weighing >3 kg and 3– step and two-step extraction processes, where the residue was ex-
4 months old were used (The University of Auckland Animal Ethics tracted with another 900 μLL acetonitrile containing acetate buffer
Committee reference number 002,012). using the same process. The extraction recoveries were determined
using the equation below:
2.2. Chromatographic conditions Peak area of the analyte extracted in plasma
recovery (% ) =
Peak area of analyte in solution
An Agilent 1260 HPLC machine (Agilent Corporation, Germany)
× 100 (3)
was used for sample analysis. The samples were analysed using
a Luna omega PS C18 100 Å, (250 × 4.6 mm, 5 μm) column from The samples were prepared as per the following method.
Phenomenex (Auckland, New Zealand). The mobile phase com- Briefly, 50 μL of plasma was mixed with and 900 μLL acetonitrile
prised 0.5 mM phosphate buffer (pH 5.0), methanol and acetoni- and 50 μLL of 0.1 M acetate buffer (pH 5.0). The mixture was vor-
trile (93:5:2 v/v) with the flow rate set at 1 mL/min, while the col- tex mixed for 5 min at 4 °C at 20 0 0 RPM and centrifuged at 13,400
umn temperature was maintained at 30 °C. The injection volume RPM at 4 °C. The supernatant was collected in another Eppendorf
was 20 μLL. Detection was carried out at 228 nm using a photo tube and evaporated at 40 °C with a gentle nitrogen stream. The
diode array (PDA) detector. The stability-indicating nature of the residue was then reconstituted using 50 μLL of mobile phase and
assay was validated with PDA using purity index >99.9% under var- 20 μLL was injected for HPLC analysis.
ious forced degradation conditions (acidic, alkaline, UV and oxida- The extraction recoveries of spiked quality control (QC) samples
tive stress) with the above mentioned chromatographic conditions. prepared at three different concentration levels (0.5 μg/mL, 4 μg/mL

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611
JID: CHROMA
ARTICLE IN PRESS [m5G;October 14, 2019;22:9]

T.J. Purohit, Z. Wu and S.M. Hanning / Journal of Chromatography A xxx (xxxx) xxx 3

and 16 μg/mL) were evaluated by injecting triplicate samples of 2.5.6. Post-extraction and sample storage stability
each concentration. The extraction recoveries were calculated us- The stability of amoxicillin kept in auto sampler (without tem-
ing Eq. (3) above. perature control) was determined by reinjecting amoxicillin sam-
ples prepared at two concentration levels, 0.5 and 16 μg/mL, in
triplicate over a period of 72 h. Freeze-and-thaw stability of spiked
2.5. Method validation
plasma samples at two concentrations (0.5 and 16 μg/mL) were
studied in triplicate for stability of amoxicillin over three freeze-
Following extraction the method for plasma sample analysis
and-thaw cycles (−20 °C to ambient temperature). Separate sets of
was validated according to the guidelines for validation of bioana-
spiked plasma samples with the same concentrations were pre-
lytical method [28].
pared to determine long-term freezing (−20 °C) stability for up to
two months. The stabilities were determined provided the accuracy
2.5.1. Preparation of calibration and quality control (QC) standards in at each concentration level remained within ±15%.
plasma
Different spiking solutions ranging from 2–200 μg/mL were pre- 2.6. Pharmacokinetic study
pared from a 1 mg/mL stock solution of AMT in water. Blank rabbit
plasma was spiked with the appropriate volume of respective spik- The analytical study to determine blood concentration of amox-
ing solution to prepare eight calibration standards (0.2, 0.3, 0.4, 0.8, icillin in rabbit plasma was approved by The University of Auckland
2, 6, 12 and 20 μg/mL) for determination of linearity. Three differ- Animal Ethics Committee (reference number 002,012). The study
ent QC standards: 0.5 μg/mL, 4 μg/mL and 16 μg/mL, were spiked was carried out on New Zealand white rabbits (n = 3). All the rab-
similarly as standards to determine accuracy and precision of the bits were acclimatized for five days prior to the start of the study
method. and were allowed free access to food and water. The rabbits re-
ceived an IV bolus of amoxicillin sodium solution (53 mg/mL) at
a total dose equivalent to 100 mg amoxicillin. Blood samples were
2.5.2. Sensitivity
collected at 0, 0.25, 0.5, 1, 2, 4, 6 and 8 h then centrifuged at 7800
The sensitivity of the method was determined in nine samples
RPM for 10 min at 4 °C. Plasma was separated and stored at −20 °C
by measuring the analyte response from blank samples and the
until analysis. The samples were extracted and analysed as per the
lowest non-zero concentration. To meet the criteria the analyte re-
method described under Section 2.5.
sponse from the lowest non-zero concentration should be ≥5 the
response from blank calibrator.
2.7. Data analysis

2.5.3. Linearity Pharmacokinetic parameters were calculated using PKsolver 2.0,


Linearity shows that the sample solutions are in a concentration an add-in program for pharmacokinetic and pharmacodynamic
range where analyte absorption is linearly proportional to concen- data analysis in Microsoft Excel [29]. The pharmacokinetic param-
tration. This was performed by analysing freshly prepared spiked eters: half-life (t1/2 ), area under the drug-concentration – time
plasma samples (n = 3) ranging from 0.2–20 μg/mL for three con- curve (AUC), area under the moment curve (AUMC), volume of dis-
secutive days. A linearity standard curve was obtained by plotting tribution (Vd ) and clearance (Cl) were determined using a non-
the peak area against concentration. compartmental pharmacokinetic model and all the results were re-
ported as mean ± standard deviation (SD). Statistical evaluation
was carried out using IBM SPSS statistics software (Version 25, IBM
2.5.4. Selectivity and specificity
Corporation). An independent t-test was used to evaluate the re-
Blank plasma samples (n = 6) were prepared at different time
coveries of one-step and two-step extraction protocol. Differences
points were compared and evaluated for interference of an en-
were considered significant if p < 0.05.
dogenous compound during analysis to establish the selectivity of
the method. The specificity of the method was evaluated based
3. Results and discussion
on the selectivity of the method and ability to detect the ana-
lyte of interest (±20% LLOQ) from the sample matrix. The chro-
3.1. Challenges in the extraction of amoxicillin during preliminary
matograms of blank samples and spiked samples (0.2 μg/mL; n = 3)
studies
were analysed and compared for determination of specificity of the
method.
Extraction and PP with the most commonly used organic sol-
vents, acetonitrile and methanol, either alone or in combination
2.5.5. Accuracy and precision did not give acceptable extraction recovery of amoxicillin. Extrac-
The accuracy (%A) of the method is the degree of closeness of tion with methanol resulted in extremely high recoveries rang-
the measured concentration of the analyte to the true concentra- ing from 140–180% and further resulted in compromised column
tion of the analyte in QC samples. The accuracy was determined by performance. The cause for this high extraction recovery may be
analysing nine QC samples at three different concentration levels, related to evaporation of methanol due to low boiling point or
0.5, 4 and 16 μg/mL, using equation mentioned below: methanolysis of amoxicillin, although this is yet to be fully ex-
plored. It was observed that acetonitrile was a better precipitating
%A = [ Measured concentration/Nominal concentration] × 100 agent as compared to methanol, which is similar to the findings by
(4) Polson et al. [4]. Therefore, an acetonitrile mediated PP approach
was employed for extraction trials.
Precision is the degree of closeness between the results. Intra- One and two-step extraction with acetonitrile resulted in low
day and inter-day precision was determined by applying multiple and varying extraction recoveries ranging from 10–40%. This could
injections of QC samples in a single day and for three consecutive be due to limited solubility of amoxicillin in acetonitrile, which is
days, respectively, and calculating the relative standard deviations in agreement with the findings by Bhattacharya [10] that amox-
(RSD). The accuracy and the precision of the developed method icillin is practically insoluble in acetonitrile. Lindegardh et al. re-
was determined provided the% RSD values remain below 15. ported an improvement in extraction recovery using a plasma to

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611
JID: CHROMA
ARTICLE IN PRESS [m5G;October 14, 2019;22:9]

4 T.J. Purohit, Z. Wu and S.M. Hanning / Journal of Chromatography A xxx (xxxx) xxx

3.2.2. Stability of amoxicillin


Solvent systems containing water, acetonitrile with acetate
buffer (0.1 M, pH 5.0) with or without plasma and acetonitrile with
plasma qualified for stability analysis. Amoxicillin was found more
stable at 4 °C than 22 °C in all solvent systems. Following storage
at 4 °C for 24 h, 87% and 78% amoxicillin was found as such in the
solvent systems containing acetonitrile and plasma with or with-
out acetate buffer, respectively. A near two-fold increase in stability
was observed due to addition of acetate buffer to the binary sys-
tem. Fig. 2 shows the degradation of amoxicillin in solvent systems
containing acetate buffer with or without plasma.
The results of linear regression analysis for first-order degrada-
tion kinetic model for both the systems at 4 °C (Table 1) shows the
effect of acetate buffer (pH 5.0) on drug stability. Amoxicillin was
found stable for about 2 days in the ternary solvent system con-
taining acetate buffer. This also justified the sample stability for
the maximum analysis time, 12 h a day.
Thus, it was confirmed that, addition of acetate buffer improved
both, solubility and stability of amoxicillin. Therefore, the protocol
for extraction of amoxicillin from plasma was developed using ac-
etate buffer (0.1 M, pH 5.0) with acetonitrile.
Fig. 1. Solubility of amoxicillin in different solvent systems (n = 3) at 25 °C: Water
(), Acetonitrile (ACN) (█), ACN+ Methanol+ Water (), ACN+ Water (), ACN+
Acetate buffer, 0.1 M, pH 5.0 (♦), ACN+ Phosphate buffer, 0.1 M, pH 7.4 (), ACN+
3.3. Development and selection of extraction protocol
Formic acid, pH 2.5 (), ACN+ Plasma (), ACN+ Acetate buffer, 0.1 M, pH 5.0+
Plasma (◦). A plasma to acetonitrile ratio under 1:2 has been reported to be
sufficient for protein removal from plasma [4]. Addition of buffer
may dissolve some of the plasma proteins; therefore, a ratio of
acetonitrile ratio greater than 1:12 [16], however, our trials with organic to inorganic solvents of 1:9 was selected to avoid accu-
plasma to acetonitrile ratios 1:12 and 1:15 contradicted this. We mulation of any endogenous proteins and to improve the column
found that increasing the quantity of acetonitrile did not improve life and performance. The extraction protocol was developed with
the extraction recoveries, due to the limited solubility of amox- the solvent system containing acetonitrile, acetate buffer (0.1 M, pH
icillin in acetonitrile. This was later confirmed by the solubility 5.0) and plasma in a ratio of 90:5:5, respectively.
and stability analysis (Section 3.2.1). pH change mediated PP was Both the one-step and two-step extraction protocols were found
also carried out by adding smaller proportion of either 0.1% formic repeatable, with extraction recoveries of amoxicillin from plasma
acid or 10% perchloric acid. It was observed that PP was better samples ranging between 80 and 88%. At the concentration range
with formic acid than with perchloric acid. Upon drying, the super- of 1–20 μg/mL, one-step and two-step extraction protocols resulted
natant collected from the perchloric acid treated samples resulted in extraction recoveries of 81.0–83.6% and 82.9–87.0%, respectively.
in dense, cloudy precipitates, indicating a higher amount of dis- When comparing extraction recoveries, no significant improvement
solved proteins. This rendered the samples inappropriate for anal- in extraction recovery was achieved using the two-step protocol
ysis. On the other hand, formic acid samples resulted in low recov- compared with the one-step extraction protocol (p = 0.834). Fur-
eries, possibly due to pH dependent stability of amoxicillin. Thus, thermore, the two-step process was more time consuming and
variable extraction recoveries and non-repeatability in the results complex compared with the one-step extraction protocol. There-
with different sample purification approaches provided the basis to fore, a one-step extraction was selected for sample preparation and
carry out preliminary analysis of solubility and stability of amoxi- method validation.
cillin in commonly used solvent systems in order to develop a re-
peatable and reliable extraction protocol.
3.3.1. Extraction recovery
The extraction recovery was found between 82 and 89% for
all samples. The extraction recovery of QC samples at 0.5 μg/mL,
4 μg/mL and 16 μg/mL were found to be 84.6%, 83.5% and 82.6%,
3.2. Solubility and stability analysis
respectively. The results of extraction recovery have been sum-
marised in Table 2.
3.2.1. Solubility of amoxicillin
The solubility equilibrium was established within 24 h in all
solvent systems (Fig. 1). There was no further increase in drug 3.4. Method validation
concentration in the supernatants found at 48 h. The solubility of
amoxicillin in water was found to be 2.7 mg/mL at ambient tem- The developed method was found to be simple, rapid, sensitive
perature which was close to the reported values [6]. It has been and reproducible with amoxicillin eluting at 6.2 ± 0.2 min. The re-
reported that amoxicillin is practically insoluble in acetonitrile [10]. sults of various validation parameters are discussed below.
This was supported by the low solubility of amoxicillin (below
70 μg/mL) in acetonitrile alone or in binary systems along with ei-
ther methanol, water or formic acid. Solubility in phosphate buffer 3.4.1. Sensitivity
(0.1 M, pH 7.4) was found slightly above 100 μg/mL. Addition of ac- The sensitivity of the method was determined with a concen-
etate buffer (0.1 M, pH 5.0) resulted in a 16 and 5 fold increase in tration as low as 0.2 μg/mL giving a response that was five times
solubility of amoxicillin in the acetonitrile alone and along with higher than the blank sample. Therefore, this was considered the
plasma, respectively. lower limit of quantification (LLOQ) for this method.

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611
JID: CHROMA
ARTICLE IN PRESS [m5G;October 14, 2019;22:9]

T.J. Purohit, Z. Wu and S.M. Hanning / Journal of Chromatography A xxx (xxxx) xxx 5

Table 1
Degradation kinetics of the solvent systems with or without acetate buffer (0.1 M, pH 5.0). All the samples
were kept at 4 °C.

Solvent system R2 kobs (10− 2 d−1 ) t1/2 (day) t90 (day)

Acetonitrile: plasma (without buffer) (95:5 v/v) 0.9785 8.412 8.2 1.2
Acetonitrile: plasma: acetate buffer (90:5:5 v/v) 0.9864 4.925 14.1 2.1

Table 2
Accuracy, precision and extraction recovery data for amoxicillin at concentrations of 0.5, 4 and 16 μg/mL. All the values are expressed as a mean
percent (%). RSD – Relative standard deviation.

Concentration (μg/mL) Intra-day (n = 3) Inter-day (n = 9) Extraction recovery (n = 3)

Accuracy (% Nominal) Precision (% RSD) Accuracy (% Nominal) Precision (% RSD) Mean (%) RSD (%)

0.5 113.6 1.4 108.3 5.3 84.6 2.3


4 99.9 2.8 97.9 2.6 83.5 3.2
16 99.7 1.5 98.7 2.1 82.6 2.2

Fig. 2. Degradation graphs for amoxicillin in two different solvent systems (n = 3), acetonitrile and plasma (95:5 v/v) and acetonitrile, plasma and acetate buffer (90:5:5 v/v):
at 4 °C (a), 22 °C (b) and first-order degradation graphs at 4 °C (c) and 22 °C (d).

3.4.2. Linearity concentration, indicating the specificity and sensitivity to amoxi-


The linear regression equation (n = 3) for amoxicillin was linear cillin of the developed method.
in the concentration range of 0.2 to 20 μg/mL in rabbit plasma. The
R2 value was 0.9999 ± ± 0.0 0 02 for all linearity curves.
3.4.4. Accuracy and precision
3.4.3. Selectivity and specificity The method was found accurate and precise at three different
The developed method was found selective for amoxicillin in concentration levels. The accuracy was found between the ranges
the presence of other endogenous compounds. Analysis of multiple of 97–115%. The percentage bias was found to be below 6, which
blank samples showed that there was no interference from the en- is below 15% for both inter-day and intra-day samples. This indi-
dogenous compounds in the sample matrix. Fig. 3 shows compar- cates that the method is repeatable and reproducible. The results
ison of the chromatograms of blank plasma and lowest standard of accuracy and precision have been summarised in Table 2.

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611
JID: CHROMA
ARTICLE IN PRESS [m5G;October 14, 2019;22:9]

6 T.J. Purohit, Z. Wu and S.M. Hanning / Journal of Chromatography A xxx (xxxx) xxx

Fig. 3. Overlaid chromatograms of blank sample and a spiked sample at LLOQ concentration (0.2 μg/mL). Inset highlights the difference between blank and spiked sample at
LLOQ concentration (0.2 μg/mL).

Fig. 4. Chromatograms showing amoxicillin peak eluting at 6.2 ± 0.2 min; spiked sample, 12 μg/mL (a) and a pharmacokinetic study sample (b).

3.4.5. Post-extraction and sample storage stability cillin was found unstable upon long-term freezing at −20 °C. At
Stability of amoxicillin was determined on two aspects, post- the end of two months, the plasma samples spiked with amoxi-
extraction and freeze-and-thaw to mimic the conditions during cillin showed degradation of amoxicillin, with 58.3 and 48.9% re-
sample analysis, handling and storage. Results of the stability study maining in the samples spiked with 0.5 and 16 μg/mL amoxicillin,
have been reported in Table 3. Amoxicillin was found stable for respectively. Amoxicillin has been reported stable for one month
up to three days post-extraction when kept in the auto sampler. when kept at −80 °C [30]. However, it is advisable to analyse clin-
Amoxicillin was found stable at both the concentrations, 0.5 and ical study samples with in short period of time from collection, to
16 μg/mL for up to 72 h. Only 2–3% of amoxicillin was degraded quantify amoxicillin as accurately as possible.
from the initial concentration at both concentration levels. This in-
dicated post-extraction stability of amoxicillin in all the samples 3.5. Pharmacokinetic study
until analysis of the final sample in the sequence, despite the tem-
perature not being controlled in the auto sampler. The method was successfully applied in the determination of
Amoxicillin was found stable for three freeze-and-thaw cycles concentration of amoxicillin in rabbit plasma following IV amoxi-
(−20 °C to ambient temperature). At the end of three freeze-and- cillin administration. Fig. 4 shows chromatograms of the samples
thaw cycles only 2–5% amoxicillin was degraded from the sam- prepared from spiked blank plasma and pharmacokinetic study
ples containing 0.5 and 16 μg/mL amoxicillin. In contrast, amoxi- sample.

Table 3
Auto sampler, freeze-and-thaw and long term storage (- 20 °C) stability of amoxicillin. ∗ RSD (relative standard deviation).

Concentration (μg/mL) Auto sampler, 72 h (n = 3) Three cycles, freeze-and-thaw (n = 3) 2 months, storage at −20 °C (n = 3)

Drug remaining (%) RSD (%) Drug remaining (%) RSD (%) Drug remaining (%) RSD (%)

0.5 99.5 6.0 98.4 6.2 58.3 6.3


16 98.3 8.0 94.9 3.7 48.9 2.1

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611
JID: CHROMA
ARTICLE IN PRESS [m5G;October 14, 2019;22:9]

T.J. Purohit, Z. Wu and S.M. Hanning / Journal of Chromatography A xxx (xxxx) xxx 7

Table 4
Pharmacokinetic parameters following IV administration of amoxicillin at a dose of 53 mg/mL amoxicillin
sodium solution equivalent to 100 mg amoxicillin. Data are expressed as Mean ± SD for n = 3 rabbits.

Dosing t1/2 (h) AUC0-∞ (μg/mL∗ h) AUMC (μg/mL∗ h2 ) Vd (L) Cl (L/h)

IV amoxicillin 0.62 ± 0.04 20.16 ± 4.69 12.94 ± 5.13 4.56 ± 0.79 5.14 ± 1.15

dependent solubility and stability, which is frequently overlooked


for many drugs.

Acknowledgements

The authors declare no conflict of interest. This research was


supported by a School of Pharmacy Performance Based Research
Fund (2017) grant for Sara Hanning (principal investigator) and
Zimei Wu, The University of Auckland, New Zealand.

References

[1] R.D. McDowall, E. Doyle, G.S. Murkitt, V.S. Picot, Sample preparation for the
HPLC analysis of drugs in biological fluids, J. Pharm. Biomed. Anal. 7 (1989)
1087–1096, doi:10.1016/0731- 7085(89)80047- 0.
[2] G. Delis, G. Batzias, G. Kounenis, M. Koutsoviti-Papadopoulou, Application and
validation of a LC/fluorescence method for the determination of amoxicillin in
sheep serum and tissue cage fluid, J. Pharm. Biomed. Anal. 49 (2009) 375–380,
doi:10.1016/j.jpba.2008.10.019.
[3] T.M. Alshammari, A.A. Al-Hassan, T.B. Hadda, M. Aljofan, Comparison of differ-
Fig. 5. Drug release profiles in rabbits (n = 3) following IV amoxicillin administra- ent serum sample extraction methods and their suitability for mass spectrom-
tion at a dose of 53 mg/mL amoxicillin sodium solution equivalent to 100 mg amox- etry analysis, Saudi Pharm. J. 23 (2015) 689–697, doi:10.1016/j.jsps.2015.01.023.
icillin. [4] C. Polson, P. Sarkar, B. Incledon, V. Raguvaran, R. Grant, Optimization of protein
precipitation based upon effectiveness of protein removal and ionization effect
in liquid chromatography–tandem mass spectrometry, J. Chromatogr. B. 785
(2003) 263–275, doi:10.1016/S1570- 0232(02)00914- 5.
[5] S. Soltani, A. Jouyban, Biological sample preparation: attempts on productivity
The drug release study following IV administration of amox- increasing in bioanalysis, Bioanalysis 6 (2014) 1691–1710, doi:10.4155/bio.14.
icillin was carried out and the release profile (n = 3, each time 118.
point) is shown in Fig. 5 and the pharmacokinetic parameters are [6] National center for biotechnology information, PubChem compound database
; CID=33613, https://pubchem.ncbi.nlm.nih.gov/compound/33613, (Accessed 3
summarised in Table 4. July 2018).
Table 4 shows that, following IV administration, amoxicillin was [7] A. Tsuji, E. Nakashima, S. Hamano, T. Yamana, Physicochemical properties of
cleared rapidly with a t1/2 of 0.6 h (35–40 min). This is close to amphoteric β -Lactam antibiotics I: stability, solubility, and dissolution behav-
ior of amino penicillins as a function of pH, J. Pharm. Sci. 67 (1978) 1059–1066,
the reported t1/2 values of 1 h in humans [31]. However, the values doi:10.10 02/jps.260 0670810.
of Vd and Cl were found noticeably lower than those in human [8] M. Frańska, Reaction of ampicillin and amoxicillin with alcohols, Ars. Separa-
treated with 500 mg IV amoxicillin, 20.2 L and 13.3 L/h, respectively toria Acta 9/10 (2012) 25–35.
[9] P.G. Navarro, I.H. Blázquez, B.Q. Osso, P.J. Martıńez de las Parras,
[31].
M.a.I.M.n Puentedura, A.A.M. Garcıá, Penicillin degradation catalysed
by Zn(II) ions in methanol, Int. J. Biol. Macromol. 33 (2003) 159–166,
doi:10.1016/S0141-8130(03)0 0 081-3.
4. Conclusion [10] P.K. Bhattacharyya, W.M. Cort, Amoxicillin, in: K. Florey (Ed.), Analytical Pro-
files of Drug Substances, Academic Press, 1978, pp. 19–41.
[11] Q. Pei, G.-.P. Yang, Z.-.J. Li, X.-.D. Peng, J.-.H. Fan, Z.-.Q. Liu, Simultaneous anal-
Quantitative analysis of a drug from plasma samples usually in-
ysis of amoxicillin and sulbactam in human plasma by HPLC-DAD for assess-
volves protein removal and maximum extraction of the drug using ment of bioequivalence, J. Chromatogr. B. 879 (2011) 20 0 0–20 04, doi:10.1016/
a suitable extraction protocol with the consideration of drug stabil- j.jchromb.2011.05.021.
ity. In this research, the solubility and stability analysis of amoxi- [12] M.-.C. Verdier, O. Tribut, P. Tattevin, Y. Le Tulzo, C. Michelet, D. Bentué-Ferrer,
Simultaneous determination of 12 beta-lactam antibiotics in human plasma
cillin was carried out in commonly used solvents by a stability- by high-performance liquid chromatography with UV detection: application to
indicating HPLC assay. Addition of a small amount of acetate buffer therapeutic drug monitoring, Antimicrob. Agents Chemother. 55 (2011) 4873–
(0.1 M, pH 5.0) improved the solubility as well as stability of amox- 4879, doi:10.1128/AAC.00533-11.
[13] J. Pullen, L.M. Stolk, C. Neef, L.J. Zimmermann, Microanalysis of amoxicillin, flu-
icillin. The synergy of organic solvent and pH-control was further cloxacillin, and rifampicin in neonatal plasma, Biomed. Chromatogr. 21 (2007)
utilised in the development of a suitable extraction protocol for 1259–1265, doi:10.1002/bmc.881.
extraction of amoxicillin from plasma. A simple, rapid, one-step [14] X. Du, C. Li, H.K. Sun, C.H. Nightingale, D.P. Nicolau, A sensitive assay of amox-
icillin in mouse serum and broncho-alveolar lavage fluid by liquid–liquid ex-
and easy to follow sample purification procedure by PP was de- traction and reversed-phase HPLC, J. Pharm. Biomed. Anal. 39 (2005) 648–652,
veloped for extraction of amoxicillin from plasma before subjecting doi:10.1016/j.jpba.2005.04.021.
the samples to HPLC for quantitative analysis. The HPLC method for [15] W. Khayata, D.A. Zakri, Two simple spectrophotometric methods for the simul-
taneous determination of amoxicillin trihydrates and flucloxacillin sodium, Res.
quantification of amoxicillin was found to be simple and repeatable
J. Pharm. Technol. 10 (2017) 1327–1332, doi:10.5958/0974-360X.2017.00235.9.
with the stability-indicating nature. Amoxicillin was found stable [16] N. Lindegardh, T. Singtoroj, A. Annerberg, N.J. White, N.P. Day, Development
for up to 3 days post-extraction in the auto sampler without tem- and validation of a solid-phase extraction-liquid chromatographic method for
determination of amoxicillin in plasma, Ther. Drug. Monit. 27 (2005) 503–508,
perature control. However, the stability was compromised when
doi:10.1097/01.ftd.0 0 0 0158082.38330.85.
the samples were kept at −20 °C for two months with nearly 50% [17] M. Carlier, V. Stove, J.A. Roberts, E. Van de Velde, J.J. De Waele, A.G. Ver-
of amoxicillin was found degraded. The method was successfully straete, Quantification of seven beta-lactam antibiotics and two beta-lactamase
applied for the quantitative analysis of amoxicillin from a pharma- inhibitors in human plasma using a validated UPLC–MS/MS method, Int. J. An-
timicrob. Agents 40 (2012) 416–422, doi:10.1016/j.ijantimicag.2012.06.022.
cokinetic study in rabbits. This research highlighted the importance [18] G. Hoizey, D. Lamiable, C. Frances, T. Trenque, M. Kaltenbach, J. Denis, H. Mil-
of understanding the physicochemical properties, particularly pH lart, Simultaneous determination of amoxicillin and clavulanic acid in human

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611
JID: CHROMA
ARTICLE IN PRESS [m5G;October 14, 2019;22:9]

8 T.J. Purohit, Z. Wu and S.M. Hanning / Journal of Chromatography A xxx (xxxx) xxx

plasma by HPLC with UV detection, J. Pharm. Biomed. Anal. 30 (2002) 661– [25] Z. Yuan, H.Q. Russlie, D.M. Canafax, Sensitive assay for measuring amoxi-
666, doi:10.1016/s0731- 7085(02)00289- 3. cillin in human plasma and middle-ear fluid using solid-phase extraction
[19] E. Guncum, T. Bakirel, C. Anlas, H. Ekici, N. Isiklan, Novel amoxicillin nanopar- and reversed-phase high-performance liquid–chromatography, J. Chromatogr.
ticles formulated as sustained release delivery system for poultry use, J. Vet. B. Biomed. Appl. 674 (1995) 93–99, doi:10.1016/0378- 4347(95)00302- 1.
Pharmacol. Ther. 41 (2018) 588–598, doi:10.1111/jvp.12505. [26] H.J. Mascher, C. Kikuta, Determination of amoxicillin in human serum
[20] M. Carlier, V. Stove, J.J. De Waele, A.G. Verstraete, Ultrafast quantification of and plasma by high-performance liquid chromatography and on-line post-
beta-lactam antibiotics in human plasma using UPLC–MS/MS, J. Chromatogr. B. column derivatisation, J. Chromatogr. A. 812 (1998) 221–226, doi:10.1016/
978–979 (2015) 89–94, doi:10.1016/j.jchromb.2014.11.034. s0 021-9673(98)0 0391-4.
[21] A. Abdulla, S. Bahmany, R.A. Wijma, B.C.H. van der Nagel, B.C.P. Koch, Simul- [27] B. Charles, S. Chulavatnatol, Simple analysis of amoxycillin in plasma by high
taneous determination of nine beta-lactam antibiotics in human plasma by an performance liquid chromatography with internal standardization and ultra-
ultrafast hydrophilic-interaction chromatography-tandem mass spectrometry, J. violet detection, Biomed. Chromatogr. 7 (1993) 204–207, doi:10.1002/bmc.
Chromatogr. B. 1060 (2017) 138–143, doi:10.1016/j.jchromb.2017.06.014. 1130070407.
[22] S. Lefeuvre, J. Bois-Maublanc, L. Hocqueloux, L. Bret, T. Francia, C. Eleout-Da [28] Bioanalytical method validation: guidance for industry, https://www.fda.gov/
Violante, E.M. Billaud, F. Barbier, L. Got, A simple ultra-high-performance liquid downloads/Drugs/Guidances/ucm070107.pdf , (Accessed 18 December 2018).
chromatography-high resolution mass spectrometry assay for the simultaneous [29] Y. Zhang, M. Huo, J. Zhou, S. Xie, PKSolver: an add-in program for pharmacoki-
quantification of 15 antibiotics in plasma, J. Chromatogr. B. 1065-1066 (2017) netic and pharmacodynamic data analysis in microsoft excel, Comput. Methods
50–58, doi:10.1016/j.jchromb.2017.09.014. Programs Biomed. 99 (2010) 306–314, doi:10.1016/j.cmpb.2010.01.007.
[23] M. Szultka, R. Krzeminski, J. Szeliga, M. Jackowski, B. Buszewski, A new [30] K.M. Matar, Simple and rapid lc method for the determination of
approach for antibiotic drugs determination in human plasma by liquid amoxicillin in plasma, Chromatographia 64 (2006) 255–260, doi:10.1365/
chromatography-mass spectrometry, J. Chromatogr. A. 1272 (2013) 41–49, s10337- 006- 0021- 9.
doi:10.1016/j.chroma.2012.11.056. [31] A. Arancibia, J. Guttmann, G. González, C. González, Absorption and dispo-
[24] W. Wang, J. Liu, Y. Han, W. Huang, Q. Wang, The most convenient and general sition kinetics of amoxicillin in normal human subjects, Antimicrob. Agents
approach for plasma sample clean-up: multifunction adsorption and supported Chemother. 17 (1980) 199–202, doi:10.1128/aac.17.2.199.
liquid extraction, Bioanalysis 4 (2012) 223–225, doi:10.4155/bio.11.332.

Please cite this article as: T.J. Purohit, Z. Wu and S.M. Hanning, Simple and reliable extraction and a validated high performance liquid
chromatographic assay for quantification of amoxicillin from plasma, Journal of Chromatography A, https://doi.org/10.1016/j.chroma.2019.
460611

You might also like