LC-MS/MS characterization of forced degradation products of ambrisentan: Development and validation of a stability-indicating RP-HPLC method

NJC

View Article Online

View Journal | View Issue

PAPER

LC-MS/MS characterization of forced degradation

products of ambrisentan: development and

validation of a stability-indicating RP-HPLC method

Cite this: New J. Chem., 2014,

38, 3050

Nageswara Rao Ramisetti* and Ramakrishna Kuntamukkala

The current study reports the characterization of degradation products of ambrisentan by liquid

chromatography-tandem mass spectrometry, and development and validation of a stability-indicating

reversed phase high performance liquid chromatographic method for determination of ambrisentan in

the presence of its process related impurities in bulk drugs. The drug was subjected to various stress

conditions such as hydrolysis, oxidation, photo- and thermal degradations to investigate the stability-

indicating ability of the method. Significant degradation was observed during acidic and oxidative

stresses. Ambrisentan was well resolved from its process related impurities and degradation products

formed under stress conditions. The chromatographic separation was accomplished on an Agilent XDB

C₁₈column (150 Â 4.6 mm; 5 mm) with the mobile phase consisting of 10 mM NH₄OAc (pH = 5.2) and

CH₃CN in a gradient elution mode at a flow rate of 1.0 ml min^À1. The eluents were monitored by a

photodiode array detector at 215 nm and quantitation limits were obtained in the range of 0.07–0.25 mg

ml^À1for ambrisentan and all its process related impurities. The developed liquid chromatographic

method was validated with respect to accuracy, precision, linearity, robustness and limits of detection

and quantitation. The degradation products were characterized by comparing their collision induced dis-

sociation mass spectral data with that of ambrisentan and the most possible degradation and fragmenta-

tion pathways were proposed.

Received (in Montpellier, France)

16th January 2014,

Accepted 12th March 2014

DOI: 10.1039/c4nj00075g

www.rsc.org/njc

determination of AMB in rat plasma.⁵Normal and reversed

phase HPLC were also used for separation of AMB enantiomers

Introduction

Ambrisentan (AMB), (+)-(2S)-2-[(4,6-dimethylpyrimidin-2-yl)- in the absence and presence of impurities.^6,7Quite recently,

oxy]-3-methoxy-3,3-diphenylpropanoic acid, is an orally active Satheeshkumar and Naveenkumar have reported a stability-

nonsulfonamide class endothelin receptor antagonist selective indicating RP-HPLC method for determination of AMB assay,⁸

for type-A endothelin. It is used in the treatment of pulmonary but it failed to establish the stability-indicating ability of the

arterial hypertension (PAH), a rare life-threatening disease method in the presence of all possible process related impu-

characterized by its progressive increase in pulmonary vascular rities. Narayana et al. have also reported a RP-HPLC method for

resistance due to elevations in pulmonary arterial pressure determination of AMB assay in the presence of a limited

causing right ventricular failure and leading finally to number of process related impurities using a mobile phase

death.^1–4It was approved worldwide under brand names comprised of non-volatile phosphate buﬀer which is incompa-

Letairis (US), Volibris (EU), and Pulmonext (India) in various tible for LC-ESI-MS characterization.⁹However, these methods

strengths ranging from 2.5–10 mg. There is a strong demand did not address the characterization of stress DPs by mass

for the characterization of potential impurities (process related spectrometry. Thus, there is a great potential for development

impurities and degradation products (DPs)) in the quality of analytical methods to monitor the levels of all possible

control of AMB active substances, not only to ensure safety of impurities in the bulk drugs of AMB during process develop-

its formulations but also regulatory compliance.

ment and characterization of DPs to ensure its safety in

A thorough literature survey has revealed that only a few LC formulations. In the present study, four process related impu-

methods are available for analysis of AMB. LC-MS was used for rities: methyl 3,3-diphenyloxirane-2-carboxylate (Imp-2), methyl

2-hydroxy-3-methoxy-3,3-diphenylpropanoate (Imp-3), (S)-2-

hydroxy-3-methoxy-3,3-diphenylpropanoic acid (Imp-4), and

HPLC Group, Analytical Chemistry Division, IICT, Tarnaka, Hyderabad-500007,

4,6-dimethyl-2-(methylsulfonyl)pyrimidine (Imp-5), were prepared

from benzophenone (Imp-1) as described in Scheme 1,^10,11

India. E-mail: rnrao55@yahoo.com, rnrao@iict.res.in; Fax: +91 40 27160387;

Tel: +91 40 27193193

3050 | New J. Chem., 2014, 38, 3050--3061

This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

View Article Online

Paper

NJC

CBM-20A communications bus module was used for method

development, validation and stress degradation studies (all

from Shimadzu, Kyoto, Japan.). The chromatographic system

was controlled by LC Solution data acquisition software. The

chromatographic column used was XDB C₁₈(150 Â 4.6 mm,

5 mm) from Agilent. The mobile phase components are (A)

10 mM NH₄OAc (pH adjusted to 5.2 by using 10% AcOH

solution), and (B) CH₃CN. The separation was accomplished

in a gradient elution program (time (min)/% B: 0.01/27, 12/27,

17/55, 33/55, 34/27, 45/27) at a flow rate of 1.0 ml min^À1and at a

column temperature of 25 1C. The chromatographic eluents

were monitored at a detection wavelength of 215 nm using a

photodiode array (PDA) detector. The sample injection volume

was 20 ml.

Mass spectrometry. An Agilent 1200 series HPLC instrument

coupled to a quadrupole time-of flight (Q-TOF) mass spectro-

meter (Q-TOF LC/MS 6510 series classic G6510A, Agilent Tech-

nologies, USA) equipped with an ESI source was used for

identification and characterization of degradation products of

AMB. The data acquisition and processing were under the

control of Mass Hunter workstation software. In order to allow

the entry of only 40% of the chromatographic eluent a splitter

was placed before the ESI source. The typical operating source

conditions for MS scan in positive ESI mode were optimized as

follows: the fragmentor voltage was 80 V; the capillary voltage

was 3000–3500 V; the skimmer voltage 60 V. Nitrogen gas was

used for nebulization (45 psi) and drying (300 1C, 9 L h^À1). For

collision-induced dissociation (CID) experiments, keeping MS¹

static, the precursor ion of interest was selected using the

quadrupole analyzer and the product ions were analyzed using

a time-of-flight (TOF) analyzer. Ultra-high purity nitrogen was

used as collision gas at 18 Torr. All the spectra were recorded

under identical experimental conditions, and are averages of

20–30 scans.

Scheme 1 A schematic representation of the synthesis of ambrisentan.

a validated stability-indicating RP-HPLC method for the determi-

nation of AMB and its all five process related substances was

developed, and five of its DPs formed under acidic and oxidative

stress were characterized by LC-ESI-MS/MS and accurate mass

measurements.

Experimental

Chemicals

High purity H₂O was obtained by using a Millipore Milli-Q

water purification system (Millipore synergy, France). HPLC

grade CH₃CN was purchased from Sigma Aldrich Chemicals

Pvt. Ltd, Bangalore, India. Benzophenone (minimum assay

99%), and AR grade NH₄OAc, AcOH, HCl, and NaOH were

purchased from SD Fine Chemicals Pvt. Ltd, Mumbai, India.

H₂O₂(27% w/w) from Acros Organics was used for oxidative

degradation. A mixture of CH₃CN and H₂O (50 : 50 v/v) was used

as diluent. AMB (potency 99.8%) was received as a gift sample

from Matrix Laboratories, Hyderabad, India. Related impurities

of AMB were synthesized in our laboratory. ¹H NMR (CDCl₃,

BRUKER AVANCE 300) and HPLC area percentage purity data of

synthesized impurities are given below.

Preparation of analytical solutions

Stock solutions of AMB (2.0 mg ml^À1) and all process related

impurities (0.5 mg ml^À1each) were prepared separately by

dissolving the appropriate amounts in the minimum amount of

CH₃CN and diluted to volume with diluent. A stock solution of

impurity mixture (0.05 mg ml^À1each) was also prepared by

mixing impurity stock solutions and made to volume with diluent.

Working solutions were prepared by adequately mixing the stock

solutions for method development and validation studies.

Imp-2. d 3.52 (s, 3H), 3.99 (s, 1H), 7.28–7.46 (m, 10H); HPLC

purity 99.2%.

Imp-3. d 2.96 (d, J = 8.7, 1H), 3.15 (s, 3H), 3.62 (s, 3H), 5.17

(d, J = 8.7, 1H), 7.29–7.44 (m, 10H); HPLC purity 99.8%.

Imp-4. d 3.17 (s, 3H), 3.41 (br, 1H), 5.08 (s, 1H), 7.29–7.50

(m, 10H); HPLC purity 99.7%.

Imp-5. d 2.51 (s, 6H), 3.55 (s, 3H), 7.00 (s, 1H); HPLC

purity 98.8%.

Specificity and forced degradation

Specificity is the ability of the method to measure the analyte

(AMB) response unequivocally in the presence of its possible

impurities. The specificity of the developed LC method for AMB

was determined in the presence of its process related impu-

rities (Imp-1 to Imp-5 at 0.15%) and DPs. Forced degradation

studies can help to identify the likely DPs, also they in turn can

Instrumentation

High performance liquid chromatography. A prominence help to establish the degradation pathways and the intrinsic

series HPLC system equipped with a quaternary UFLC LC-20AD stability of the molecule.¹²AMB (1 mg ml^À1) was subjected to

pump, a DGU-20A₅degasser, a SIL-20AC auto sampler, a CTO- stress conditions such as acidic (0.5 N HCl, 60 1C, 3 h), basic

20AC column oven, an SPD-M20A diode array detector, and a (0.5 N NaOH, 60 1C, 8 h) and neutral (H₂O, 60 1C, 8 h)

This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

New J. Chem., 2014, 38, 3050--3061 | 3051

View Article Online

NJC

Paper

hydrolyses, and oxidation (10% H₂O₂, 24 h) in solution state. concentration levels ranging from LOQ to 250% to the specification

AMB was also subjected to thermolytic (60 1C, 10 days) and level of impurities (i.e., LOQ, 0.45, 0.9, 1.35, 1.8, and 2.25 mg ml^À1

)

photolytic (UV light 254 nm, 10 days) stress in solid state. spiked into AMB drug substance (600 mg ml^À1). The calibration

Diﬀerent stress conditions were followed to achieve significant curves were drawn by plotting the peak areas of impurities against

degradation. Acid and base hydrolyzed samples were neutra- their corresponding concentrations. Similarly, assay method linear-

lized, and all the degradation samples were diluted five times ity was established by injecting AMB at five diﬀerent concentration

for assay determination. AMB assays were performed by com- levels ranging from 50 to 150% (i.e., 50, 75, 100, 125, and 150%) to

parison with standard and the mass balances (%assay + AMB concentration 100 mg ml^À1. The correlation coeﬃcients (r²),

%impurities + %DPs) were calculated for stressed samples. slopes and Y-intercepts of impurities and AMB were determined

The degradation samples were injected into an LC-PDA system from their respective calibration plots.

to check the purity and homogeneity of the AMB peak. LC-MS

was used for the characterization of the DPs.

Robustness. The robustness study was carried out to evaluate

the influence of small variations in the optimized chromato-

graphic conditions. The factors chosen for this study were tem-

perature of the column (Æ2 1C), mobile phase pH (Æ0.2) and flow

Method validation

The validation of the HPLC method was carried out for the rate (Æ0.1 ml min^À1). System suitability parameters, and changes

determination of related substances (i.e., Imp-1 to Imp-5) and in assay of AMB and recoveries of impurities were checked. In all

assay of AMB as per ICH guidelines to demonstrate that the the above deliberately altered experimental conditions, the com-

method is appropriate for its intended use.¹³

ponents of the mobile phases were held constant.

System suitability. The system suitability tests were conducted

throughout the validation studies by injecting 600 mg ml^À1of AMB

solution containing 0.9 mg ml^À1(0.15%) of all related substances

and 100 mg ml^À1of AMB solution for related substance and assay

methods, respectively.

Results and discussion

Method development and optimization of chromatographic

conditions

Precision. The system precisions were checked by analyzing

six replicates of standard solutions for both assay (AMB Preliminary experiments were carried out to develop a chroma-

100 mg ml^À1) and related substance (AMB 600 mg ml^À1spiked tographic system not only capable of eluting and resolving AMB

with 0.15% of each impurity) methods individually. The method from its process related impurities and stress DPs but also

precisions for assay and related substances were evaluated by compatible with LC-MS characterization. In the initial stages of

injecting six individual test preparations of AMB (100 mg ml^À1), method development, C₁₈columns (Phenomenex Luna and

and AMB (600 mg ml^À1) spiked with 0.15% of each impurity, Agilent Extend (250 Â 4.6 mm, 5 mm)) were tried. A desirable

respectively. The intermediate precision were evaluated with resolution (>2) between AMB and impurities was achieved with

same concentration solutions used for methods precision pre- several mobile phase compositions in isocratic elution modes

pared separately on a diﬀerent day by diﬀerent analysts using with NH₄OAc/HCOONH₄buﬀers and CH₃CN/CH₃OH. But AMB

diﬀerent instruments located within the same laboratory. Preci- peak asymmetry was found to be high (A_s> 1.5) in CH₃OH when

sion at LOQ levels was also determined by injecting six indivi- compared to CH₃CN. At lower organic mobile phase content,

dual preparations of mixtures of all impurities spiked into AMB higher retention times of non-polar impurities (Imp-1, Imp-2,

(600 mg ml^À1) at their LOQ level. The %RSDs of the areas of each and Imp-3), and DP-6 were observed; whereas co-elution of

impurity and AMB were calculated for precision studies.

polar impurities (Imp-4, and Imp-5) and DP-3 was observed at

Limits of detection (LOD) and quantitation (LOQ). The LOD higher organic content. A very good separation between all

and LOQ values for AMB and related substances (Imp-1 to Imp-5) impurities and AMB was accomplished with 20 mM NH₄OAc

were determined at signal-to-noise ratios of 3 : 1 and 10: 1, (pH = 4.2) and CH₃CN in 50 : 50 (v/v) ratio on the Luna-C18

respectively, by injecting a series of dilute solutions with known column, but with lack of desirable resolution between polar

concentrations.

components Imp-4 and DP-3. Replacement of the C₁₈column

Accuracy. Accuracy of the related substance method was eval- with a C₈column also showed no improved results. To decrease

uated by spiking known amounts of the impurities into the test HPLC method run time and good resolutions between all

sample, analyzing the same and calculating the percent recovered. impurities, DPs, and AMB the elution mode was shifted to

For related substances, the recovery studies were performed in gradient. In the gradient elution program, the method was

triplicate at four concentration levels (LOQ, 50, 100 and 150%) to initiated with lesser organic content to achieve resolution

specification level (0.15%) of impurities (i.e., LOQ, 0.45, 0.9, and between polar analytes; and later increased to achieve

1.35 mg ml^À1) with respect to AMB drug substance concentration decreased retention times of non-polar analytes. Under these

600 mg ml^À1. The accuracy of the AMB assay was evaluated in conditions (20 mM NH₄OAc–HCOONH₄, CH₃CN), large base

triplicate at the three concentration levels 50, 100 and 150% (i.e., 50, line drift to the negative side (>250 mAU) was observed with

100 and 150 mg ml^À1) to AMB concentration 100 mg ml^À1, and the different pH (4.0–6.5) buffer combinations. Employing a 15 cm

recovery was calculated for each concentration.

column instead of a 25 cm column was found useful in decreas-

Linearity. Linearity of the related substance method was estab- ing the negative base line drift without disturbing the peak

lished by analyzing series of dilute solutions at six diﬀerent parameters to a significant extent by decreasing the variation

3052 | New J. Chem., 2014, 38, 3050--3061

This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

View Article Online

Paper

NJC

attempts, optimized chromatographic conditions were identified

as described in the experimental section. The chromatographic

separation of AMB (600 mg ml^À1) and impurities (6 mg ml^À1

)

under optimized conditions is shown in Fig. 1.

Method validation

System suitability data indicating that the system was suitable

for use as the tailing factor for all the analytes was less than 1.5

and the resolution between any of the two adjacently eluting

analytes was greater than 2.5. It also confirms the good selectiv-

ity of the method. LOD and LOQ results of AMB and its related

substances were obtained in the ranges of 0.025–0.08 mg ml^À1

and 0.07–0.25 mg ml^À1, respectively, indicating the higher sensi-

tivity of the method. The %RSD values of peak areas obtained

were below 3.7 and 0.8 for related substances and AMB, respec-

tively, indicating good precision of the method. The percentage

recoveries of related substances ranged from 93.74 to 105.74 and

Fig. 1 Chromatogram of AMB (600 mg ml^À1) spiked with 1% of all process

impurities under optimized chromatographic conditions.

in mobile phase composition from initial to later conditions in

the gradient program. The detector wavelength was optimized at

215 nm instead of 225 and 262 nm as mentioned in earlier

reports,^8,9as sensitivity of all the analytes was found to be higher

and with visibility of DP-2 and DP-3. Finally, after several

Table 1 Validation data

Parameter

Imp-4

Imp-5

AMB

Imp-3

Imp-1

Imp-2

System suitability^a

t_R(min)

RRT

RRF

R_s

3.48

0.330

0.84

—

1.43

3252

5.11

0.484

0.26

5.91

0.47

1.20

4490

10.55

1.000

1.00

11.94

2.03

1.08

4834

23.83

2.259

1.12

29.02

5.84

1.12

95 243

24.69

2.340

1.02

2.53

6.09

1.06

72 124

25.95

2.460

0.98

3.42

6.45

1.21

78 311

k^l

A_s

N

Linearity

Range (mg ml^À1

)

0.09–2.25

0.9999

46 179

À56.43

0.03

0.25–2.25

0.9987

13 467

341.8

0.08

50–150

1

66 199

À91 467

0.033

0.1

0.07–2.25

0.9984

57 828

2010

0.025

0.07

0.1–2.25

0.9985

52 904

1009

0.033

0.1

0.1–2.25

0.9992

54 277

942.6

0.033

0.1

r²

Slope

Intercept

Detection limit (mg ml^À1

)

Quantitation limit (mg ml^À1

)

0.09

0.25

RRF

0.84

0.26

1.00

1.12

1.02

0.98

Precision (%RSD)^a

System

Method

Intermediate

LOQ

0.804

1.041

3.612

2.497

0.418

2.133

0.766

2.233

0.358

0.483

0.797

—

1.588

0.705

2.076

1.572

1.514

0.445

0.997

1.363

1.239

0.987

1.691

1.986

Accuracy at 50% level^b

Amount added (mg ml^À1

)

0.45

0.461

102.51

0.45

0.442

98.21

50

49.59

99.19

0.45

0.456

101.32

0.45

0.457

101.59

0.45

0.445

98.86

Amount recovered (mg ml^À1

%Recovery

)

Accuracy at 100% level^b

Amount added (mg ml^À1

)

0.90

0.893

99.24

0.90

0.903

100.35

100

100.23

0.90

0.904

100.42

0.90

0.878

97.59

0.90

0.887

98.6

Amount recovered (mg ml^À1

%Recovery

Accuracy at 150% level^b

Amount added (mg ml^À1

)

1.35

1.327

99.27

1.35

1.327

98.31

150

151.06

100.71

1.35

1.33

98.51

1.35

1.379

102.14

1.35

1.325

98.12

Amount recovered (mg ml^À1

%Recovery

Accuracy at LOQ level^b

Amount added (mg ml^À1

)

0.09

0.092

102.09

0.25

0.24

95.84

—

0.07

0.066

93.74

0.1

0.106

105.74

0.1

0.095

95.26

Amount recovered (mg ml^À1

%Recovery

t_R, retention time; RRT, relative retention time; R_s, resolution; k^l, retention factor; A_s, tailing factor; N, number of theoretical plates; r², correlation

a

b

coeﬃcient; RRF, relative response factor. Average of six determinations. Average of three determinations.

This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

New J. Chem., 2014, 38, 3050--3061 | 3053

View Article Online

NJC

Paper

AMB from 99.19 to 100.71 determining the accuracy of the Significant degradation of AMB drug substance was observed in

method. Linearity results of AMB and related substances showed acidic hydrolysis, and oxidation (Fig. 2). Peak purity test results

the existence of an excellent correlation (r²> 0.998) between the obtained from the PDA detector confirmed that the AMB peak was

peak area and the concentrations. In all deliberately altered pure and homogeneous in all the analyzed stress samples. The

chromatographic conditions (flow rate, pH and column tem- mass balance of stressed samples was found in the range of

perature), all analytes were adequately resolved and elution order 99.3–99.8%. The forced degradation results are summarized in

remains unchanged. The resolutions between any two adjacent Table 2. Peak purity test results obtained from the PDA detector

analytes were obtained greater than 2.0 and tailing factors of all confirmed that the AMB peak was homogeneous and pure in all

analytes were obtained less than 1.6. The variability in the the analyzed stress samples. Insignificant change in assay of AMB

estimation of AMB assay and related substances was within Æ2 in the presence of related substances and peak purity results of

and Æ9%, respectively, indicating the robustness of the method. stress samples confirm the specificity and stability-indicating

The results of system suitability, LOD, LOQ, precision, accuracy ability of the developed method.

and linearity are summarized in Table 1.

Characterization of AMB and its DPs by LC-MS/MS

Specificity and degradation behavior of AMB

AMB and all the DPs (DP-1 to DP-6) were well separated by LC and

The assay of AMB for three determinations was found to be their retention times (t_R) are given in Table 3. AMB, DP-1, DP-3,

99.62% with %RSD 0.16, while in the presence of impurities DP-5, and DP-6 exhibited abundant protonated molecular ions

(0.15% w/w) it was 99.69% with %RSD 0.39. No considerable ([M + H]⁺) in positive ionization mode whereas DP-2 showed

degradation of AMB drug substance was observed under neutral abundant molecular ion ([M À H]^À) in negative ionization mode.

and basic hydrolytic, photolytic and thermolytic stress conditions. Collision induced dissociation (CID) spectra of the molecular ions

Fig. 2 Chromatograms of AMB: (a) acidic stress and (b) oxidative stress samples.

Table 2 Summary of forced degradation of AMB

Assay of

Mass balance

ambrisentan (assay + total impurities)

Stress conditions

Time

3 h

(% w/w)

83.9

(% w/w)

99.8

Remarks

Acid hydrolysis (0.5 N HCl, 60 1C)

Five degradation products (DP-2, DP-3, DP-4, DP-5,

and DP-6) were observed

Base hydrolysis (0.5 N NaOH, 60 1C) 8 h

99.7

99.2

98.2

99.7

99.2

99.7

99.9

99.3

No degradation was observed

One degradation product (DP-1) was observed

No degradation was observed

Water hydrolysis (60 1C)

Oxidation (10% H₂O₂)

Thermal (60 1C)

8 h

24 h

10 days 99.9

10 days 99.3

Photo (UV light)

3054 | New J. Chem., 2014, 38, 3050--3061

This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2014

View Article Online

Paper

NJC

of AMB and DPs were recorded to obtain structural information. ion of AMB showed abundant product ions at m/z 347 and 303

For CID experiments, the AMB was directly infused into the ESI which might be due to 2-carboxy-2-((4,6-dimethylpyrimidin-2-

source and on line LC-ESI-MS/MS experiments were performed to yl)oxy)-1,1-diphenylethan-1-ylium ion (loss of CH₃OH from m/z

characterize all the DPs formed under various stress conditions. 379) and protonated 2-((2,2-diphenylvinyl)oxy)-4,6-dimethylpyr-

The fragmentation patterns were obtained based on MS/MS experi- imidine (loss of CH₃OH followed by CO₂from m/z 379), respec-

ments and accurate mass measurements from high resolution tively. These two product ions confirm the presence of the methoxy

mass spectral (HRMS) data. The ESI-MS/MS spectra of AMB and and carboxyl groups in AMB molecule. In addition, the spectrum

all the DPs (except DP-4, no MS/MS achieved) are shown in Fig. 3, also showed low abundant product ions at m/z 329 and 125. The

and the proposed fragmentation patterns of their molecular ions ion at m/z 329 ((2-((4,6-dimethylpyrimidin-2-yl)oxy)-3,3-diphenyl-

are shown in Schemes 2 to 7. The elemental composition of allylidyne)oxonium) might be formed due to the loss of CH₃OH

molecular ions of AMB and all the DPs and their fragment ions followed by loss of a H₂O molecule from m/z 379, and the ion of

obtained from HRMS data are summarized in Table 3.

m/z 125 might be due to the formation of protonated 4,6-

dimethylpyrimidin-2-ol from m/z 379, confirming the presence of

dimethyl and oxy-group substituted pyrimidine moiety in AMB

MS/MS of AMB (m/z 379)

To elucidate the degradation behavior of AMB, the ESI-MS molecule. The elemental compositions of all these proposed

spectrum of its abundant protonated molecular ion ([M + H]⁺) fragment ions (Scheme 2) have been confirmed by accurate mass

of m/z 379 was examined. The ESI-MS/MS spectrum of [M + H]⁺measurements and are given in Table 3.

Table 3 Elemental compositions of molecular and fragment ions in MS/MS spectra of AMB (m/z 379), DP-1 (m/z 395), DP-2 (m/z 271), DP-3 (m/z 125),

DP-5 (m/z 183), and DP-6 (m/z 303)

Analyte (t_Rin min)

Proposed formula

₂₂H₂₃N₂O₄

C₂₁H₁₉N₂O₃

C₂₁H₁₇N₂O₂

C₂₀H₁₉N₂O

C₆H₉N₂O

Observed mass (Da)

Calculated mass (Da)

Error (ppm)

Proposed neutral loss

Ambrisentan m/z 379 (10.6)

C

379.1652

347.1386

329.1287

303.1488

125.0709

379.1657

347.1395

329.1290

303.1497

125.0714

À1.32

À2.59

À0.91

À2.97

À4.00

CH₃OH

CH₃OH, H₂O

CO₂, CH₃OH

C₁₆H₁₄O₃

DP-1 m/z 395 (2.1)

C₂₂H₂₃N₂O₅

C₂₂H₂₁N₂O₄

C₂₁H₁₉N₂O₄

C₂₀H₁₉N₂O₂

C₁₆H₁₂NO₃

C₁₆H₁₅O₃

C₁₄H₁₁

C₆H₉N₂O₂

C₈H₉O₂

C₇H₅O

395.1600

377.1491

363.1328

319.1462

266.0801

255.1024

179.0854

141.0659

137.0602

105.0336

80.0501

395.1606

377.1501

363.1344

319.1446

266.0817

255.1021

179.0860

141.0664

137.0594

105.0340

80.0500

À1.52

À2.65

À4.41

5.01

H₂O

CH₃OH

CO₂, CH₃OH

C₆H₁₁NO₂

C₆H₈N₂O₂

C₈H₁₂N₂O₅

C₁₆H₁₄O₃

C₁₄H₁₄N₂O₃

C₁₅H₁₈N₂O₄

CH₃O₂

À6.01

1.18

À3.35

À3.54

5.84

À3.81

C₅H₆N

1.25

C₂H₇N₂

59.0615

59.0609

10.16

C₂₀H₁₆O₅

DP-2 m/z 271 (3.5)

DP-3 m/z 125 (3.9)

C₁₆H₁₅O₄

271.0980

197.0969

195.0812

119.0493

72.9920

271.0976

197.0972

197.0815

119.0502

72.9931

1.62

À1.57

À1.70

À7.55

À14.83

C₁₄H₁₃

C₁₄H₁₁

C₈H₇O

C₂HO₃

O

CO₂, CH₂O

CO₂, CH₃OH

C₆H₆

C₁₄H₁₄

O

C₆H₉N₂O

C₆H₇N₂

C₅H₈N

125.0709

107.0605

82.0660

80.0500

67.0301

61.0524

60.0453

43.0286

42.0340

33.0344

125.0714

107.0609

82.0656

80.0500

67.0296

61.0527

60.0449

43.0296

42.0343

33.0340

À4.00

À3.74

4.87

0

7.46

H₂O

CHNO

C₅H₆N

CH₃NO

C₃H₆O

C₅H₄

C₄H₃N

C₅H₆O

C₄H₅NO

C₅H₄O

C₃H₃N₂

CH₅N₂O

C₂H₆NO

CH₃N₂

C₂H₄N

CH₅O

À4.91

6.66

À23.24

À7.14

12.11

DP-5 m/z 183 (24.7)

DP-6 m/z 303 (29.5)

C₁₃H₁₁

C₇H₅O

C₆H₇O

C₆H₅

O

183.0802

105.0336

95.0493

77.0388

183.0804

105.0335

95.0491

77.0386

1.09

À0.95

À2.10

À2.60

C₆H₆

C₇H₄

C₇H₆O

C₂₀H₁₉N₂O

C₁₄H₁₁

C₆H₉N₂O

303.1492

179.0854

125.0709

303.1497

179.0860

125.0714

À1.65

À3.35

À4.00

C₆H₈N₂O

C₁₄H₁₀

New J. Chem., 2014, 38, 3050--3061 | 3055

View Article Online

NJC

Paper

Fig. 3 ESI-MS/MS spectra of molecular ions of (a) AMB, (b) DP-1, (c) DP-2, (d) DP-3, (e) DP-5, and (f) DP-6.

clearly suggests the presence of 2-hydroxy-4,6-dimethylpyrimidine

1-oxide (i.e., oxidation of pyrimidine moiety in AMB molecule

MS/MS of degradation products

DP-1 (m/z 395). The ESI-MS/MS spectrum of the ion m/z 395 during oxidation). The spectrum also showed other abundant

corresponding to the [M + H]⁺of DP-1 formed in oxidative product ions at m/z 363 (m/z 347 + 16 Da, i.e., loss of CH₃OH from

stress is shown in Fig. 3. The ion of m/z 395 possibly formed m/z 395) and m/z 377 (loss of H₂O from m/z 395) which are

due to the addition of an oxygen atom to AMB during oxidation characteristic peaks of N-oxide product. In addition, the spectrum

as it possesses 16 Da more in its molecular mass compared to also showed low abundant product ions at m/z 319 (loss of CH₃OH

AMB. It is confirmed by MS/MS studies of m/z 395 ion and followed by CO₂from m/z 379), m/z 255 (loss of pyrimidine moiety

comparison with MS/MS study of AMB. The ESI-MS/MS spec- from m/z 395), m/z 179 (loss of CH₃OH, CO₂and pyrimidine groups

trum of m/z 395 showed abundant product ions at m/z 141 (base from m/z 395), m/z 105 (due to the formation of benzylidyneoxo-

peak) and 363. The formation of the ion at m/z 141 (increase of nium ion), m/z 80 (due to the formation of 2,4-dimethylazet-3-ylium

m/z value by 16 Da might be due to addition of oxygen to m/z 125) ion) and m/z 59 (due to the formation of 3-hydroxyoxiran-2-ylium

3056 | New J. Chem., 2014, 38, 3050--3061

View Article Online

Paper

NJC

Scheme 2 Proposed fragmentation mechanisms of [M + H]⁺ion of AMB.

Scheme 3 Proposed fragmentation mechanisms of [M + H]⁺ion of DP-1.

New J. Chem., 2014, 38, 3050--3061 | 3057

View Article Online

NJC

Paper

Scheme 4 Proposed fragmentation mechanisms of [M À H]^Àion of DP-2.

Scheme 5 Proposed fragmentation mechanisms of [M + H]⁺ion of DP-3.

ion). The formation of product ions at m/z 377, 363, 319, 255, 179,

DP-2 (m/z 271). The ESI-MS spectrum of the degradation

and 141 clearly indicates the presence of an oxygen atom on the product DP-2 showed an abundant molecular ion at m/z 271

pyrimidine moiety in DP-1. The elemental compositions of DP-1 ([M À H]^À) under negative ionization mode. The ESI-MS/MS

and all its fragment ions (Scheme 3) have been confirmed by spectrum of m/z 271 (Fig. 3) showed abundant product ion at

accurate mass measurements and are given in Table 3. Therefore, m/z 73 (base peak) which might be due to 2-oxoacetate formed

DP-1 was identified as 2-(1-carboxy-2-methoxy-2,2-diphenylethoxy)- by the loss of (methoxymethylene)dibenzene. In addition, the

4,6-dimethylpyrimidine 1-oxide.

spectrum also include low abundant product ions at m/z 197,

3058 | New J. Chem., 2014, 38, 3050--3061

View Article Online

Paper

NJC

–C–OH group, respectively, in the molecular structure of m/z

125. The elemental compositions of all these ions have been

confirmed by accurate mass measurements and are given in

Table 3. From the above fragmentation pattern (Scheme 5),

DP-3 was identified as 4,6-dimethylpyrimidin-2-ol.

DP-4 (m/z 319). The ESI-MS spectrum of the minor degrada-

tion product DP-4 showed a low intensity molecular ion peak at

m/z 319. The fragmentation in CID experiments was not suc-

cessful because of low abundance molecular ion peak. Hence a

possible structure of DP-4 could not be proposed.

Scheme 6 Proposed fragmentation mechanisms of [M + H]⁺ion of DP-5.

DP-5 (m/z 183). The ESI-MS/MS spectrum [M + H]⁺ion of the

minor degradation product DP-5 (m/z 183) formed in acidic

stress is shown in Fig. 3, and it showed two abundant product

195, and 119. The ion at m/z 197 may correspond to 2,2- ions at m/z 105 (base peak) and 95. The product ion at m/z 105

diphenylethanol anion formed by the loss of CO₂followed by show a decrease in mass by 78 Da compared to m/z 183, and

CH₂O from [M À H]^Àion, the ion at m/z 195 may correspond to may be formed by the loss of C₆H₆from m/z 183 resulting in a

2,2-diphenylethenol anion formed by the loss of CO₂followed benzylidyneoxonium ion. The ion at m/z 95 may be formed due

by CH₃OH from [M À H]^Àion, and the ion at 119 may to formation of protonated phenol. In addition, the spectrum

correspond to 2-phenylethenol anion formed by the loss of also showed a low abundant product ion at m/z 77 which may

C₈H₈O₃from [M À H]^Àion or C₆H₆from m/z 197. All these be due phenyl ion by loss of C₆H₅CHO (À106 Da) from m/z 183.

product ions confirm the presence of two phenyl and carboxylic The formation of ions at m/z 105 and 77 confirms the presence

acid groups in the molecular structure of DP-2. Moreover, its of two phenyl rings with a carbonyl substitution in the mole-

retention time and molecular masses were matched with Imp- cular structure of DP-5, i.e., the ion at m/z 183 (DP-5) may be

4. Peak purity test and CID experimental studies of Imp-4 due to protonated benzophenone which is one of the process

confirm it as one of the degradation products DP-2. The related impurities (Imp-1) of ambrisentan. The elemental com-

elemental compositions of molecular ion ([M À H]^À) of DP-2 positions of molecular ion ([M + H]⁺) of DP-5 and all its

and all its fragmentation ions have been confirmed by accurate fragmentation ions (Scheme 6) have been confirmed by accu-

mass measurements and are shown in Table 3. From the above rate mass measurements and are shown in Table 3. The

discussion on the fragmentation pattern (Scheme 4), DP-2 was formation of DP-5 (benzophenone) during stress is also

identified as 2-hydroxy-3-methoxy-3,3-diphenylpropanoic acid. checked by comparing the fragmentation pattern obtained for

DP-3 (m/z 125). The ESI-MS/MS spectrum of the ion m/z 125 Imp-1. The obtained similar patterns for both Imp-1 and DP-5

corresponding to the [M + H]⁺of DP-3 formed in acidic stress is confirm DP-5 as benzophenone.

shown in Fig. 3, and it showed abundant product ions at m/z 82,

DP-6 (m/z 303). The ESI-MS/MS spectrum of the ion m/z 303

61, 60, and 42. The product ion at m/z 82 (2,4-dimethylazet-1- corresponding to the [M + H]⁺of major degradation product

ium ion) might be resulted by the loss of cyanic acid from m/z DP-6 formed in acidic stress is shown in Fig. 3. The MS/MS

125. The ions at m/z 61 (isouronium ion), m/z 60 (1-hydroxy- spectrum of m/z 303 showed only two abundant product ions at

N-methylenemethanaminium ion), and m/z 42 (etheniminium m/z 179 and 125. Unlike the ambrisentan (m/z 379) spectrum,

ion) might result due to loss of C₅H₄, C₄H₃N, and C₄H₅NO from the product ion spectrum of DP-6 (m/z 303) showed no product

m/z 125, respectively. In addition, the spectrum also showed ions corresponding to the loss of CH₃OH (À32 Da) and CO₂

low abundance product ions at m/z 107 (loss of H₂O from m/z (À44 Da), clearly indicating the absence of methoxy and car-

125), m/z 80 (loss of formimidic acid from m/z 125), boxylic acid groups in its molecular structure. Exact loss of

m/z 67 (((1H-azirin-2-yl)methylidyne)ammonium), m/z 43 (pro- mass by 76 Da (32 + 44) clearly matches with the m/z value of

tonated cyanamide), and m/z 33 (methyloxonium ion). The DP-6. This may result of the CQC double bond between

formation of ions at m/z 107, 61, and 33 indicates the presence diphenyl and pyrimidine moieties in its molecular structure.

of pyrimidine moiety, urea type functional group linkage and It is confirmed by its product ion spectrum. The base peak at

Scheme 7 Proposed fragmentation mechanisms of [M + H]⁺ion of DP-6.

New J. Chem., 2014, 38, 3050--3061 | 3059

View Article Online

NJC

Paper

Scheme 8 Mechanism of degradation of AMB in acidic and oxidative stress conditions.

m/z 125 might be due to the formation of 4,6-dimethylpyrimidin- Conclusions

2-ol by the loss of a biphenyl substituted ethylene moiety (i.e.,

The degradation behavior of ambrisentan was studied

under various stress conditions as per International Confer-

ence on Harmonization (ICH) prescribed guidelines. In

total, six degradation products were formed and five of them

were characterized by liquid chromatography and tandem

mass spectrometry and accurate mass measurements. A sim-

ple gradient reversed phase high performance liquid chroma-

tographic method has been developed and validated for the

determination of a stability-indicating assay of ambrisentan

and its five related substances in bulk drugs. The developed

method has been found to be selective, precise, linear, accu-

rate, and sensitive, and is applicable for detecting process

related impurities and other possible degradation products

which may be present at trace levels in bulk drugs. Thus, the

method can be used for process development as well as quality

assurance of ambrisentan in bulk drugs.

ethene-1,1-diyldibenzene, À178 Da) attached to the pyrimidine

ring. The product ion at m/z 179 might be formed due to the

formation of 2,2-diphenylethen-1-ylium ion by the loss of a

pyrimidine moiety (À124 Da) from m/z 303. From these frag-

mentation studies (Scheme 7), DP-6 was identified as 2-((2,2-

diphenylvinyl)oxy)-4,6-dimethylpyrimidine. The elemental com-

positions DP-6 and its fragment ions have been confirmed by

accurate mass measurements and are given in Table 3.

Mechanism of degradation

The mechanism of formation of degradation products (DP_1–3

and DP_5–6) under both acidic and oxidative stress conditions

has been proposed. DP-1 in oxidative stress could be formed by

the addition of one oxygen atom from H₂O₂to one of the

nitrogens of the pyrimidine moiety of ambrisentan. DP-2 and

DP-3 might be due to the hydrolytic cleavage of C–O bond

present between pyrimidine and diphenyl moieties in ambri-

sentan in an acidic medium. The formation of DP-5 is relatively

complex and it could be due to the hydrolysis of methoxy group

followed by the loss of the pyrimidine moiety, whereas DP-6

Acknowledgements

could be formed by the loss of methoxy and CO₂groups present The authors thank Director, IICT, Hyderabad, India for encourage-

in ambrisentan. A pictorial representation of mechanism of ment and permission to communicate results for publication. RK

formation of DPs is shown in Scheme 8.

thanks CSIR, New Delhi, for a Research Fellowship.

3060 | New J. Chem., 2014, 38, 3050--3061

View Article Online

Paper

NJC

ˆ

7 M. Dousa and P. Gibala, J. Sep. Sci., 2012, 35, 798.

8 N. Satheeshkumar and G. Naveenkumar, J. Chromatogr. Sci.,

2013, DOI: 10.1093/chromsci/bmt138.

References

1 J. D. Croxtall and S. J. Keam, Drugs, 2008, 68, 2195.

2 H. W. Farber and J. Loscalzo, N. Engl. J. Med., 2004, 351, 1655.

3 J. R. Klinger, R. J. Oudiz, R. Spence, D. Despain and C. Dufton,

Am. J. Cardiol., 2011, 108, 302.

9 M. B. V. Narayana, K. B. Chandrasekhar and B. M. Rao,

J. Chromatogr. Sci., 2013, DOI: 10.1093/chromsci/bmt121.

10 H. Riechers, H. P. Albrecht, W. Amberg, E. Baumann,

`

4 N. Galie, D. Badesch, R. Oudiz, G. Simonneau, M. D. McGoon,

¨

H. Bernard, H. J. Bohm, D. Klinge, A. Kling, S. Mu¨ller,

A. M. Keogh, A. E. Frost, D. Zwicke, R. Naeije, S. Shapiro,

H. Olschewski and L. J. Rubin, J. Am. Coll. Cardiol., 2005, 46, 529.

5 R. Nirogi, V. Kandikere, P. Komarneni, R. Aleti, N. S. P. Padala

and I. Kalaikadhiban, Biomed. Chromatogr., 2012, 26, 1150.

6 R. K. Kaja, K. V. SurendhraNath, K. Babuji, P. V. V.

Satyanarayana and K. S. Kumar, Anal. Chem.: Indian J.,

2009, 8, 155.

M. Raschack, L. Unger, N. Walker and W. Wernet, J. Med.

Chem., 1996, 39, 2123.

11 L. A. Sorbera and J. Castaner, Drugs Future, 2005, 30, 765.

12 ICH. 2003. Q1A(R2): Stability testing of new drug substances

and products.

13 ICH. 2003. Q2(R1): Validation of analytical procedures: test

and methodology.

New J. Chem., 2014, 38, 3050--3061 | 3061

Article Doi

LC-MS/MS characterization of forced degradation products of ambrisentan: Development and validation of a stability-indicating RP-HPLC method

DOI: 10.1039/c4nj00075g

Source and publish data:

Authors:

Article abstract of DOI:10.1039/c4nj00075g

Full text of DOI:10.1039/c4nj00075g

Products guided by the article

R&D Labs maybe for 178306-47-3

Relevant to this article

Hot Product