Determination of Ethinyl Estradiol and Levonorgestrel in Human Plasma with Prednisone as Internal Standard Using Ultra-performance Liquid Chromatography-Tandem Mass Spectrometry.

OBJECTIVE: Ethinyl estradiol and levonorgestrel as a combination of oral contraceptive drugs have very low dosage levels; hence, a highly sensitive and selective method of using ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) is needed.
MATERIALS AND METHODS: This method was developed using prednisone as an internal standard, thus the purpose of this research was to get the optimum condition. The analytical method had been fully validated according to the European Medicines Agency guidelines, 2011. A reverse-phase chromatography separation was performed on ACQUITY UPLC ethylene bridged hybrid C18 column (1.7 μm, 2.1 × 50mm), eluted at a 0.3 mL/min flow rate under a gradient of mobile phase of 0.1% formic acid in water and acetonitrile within 5 min. Sample preparation used protein precipitation followed by liquid-liquid extraction. Quantification analysis was performed by a triple quadrupole mass spectrometry with electrospray ionization in positive-ion mode. The multiple reaction monitoring was set at m/z: 530.16 → 171.08 for ethinyl estradiol derivatized by dansyl chloride; m/z: 313.16 → 245.10 for levonorgestrel; and m/z: 359.10 → 147.04 for prednisone.
RESULTS: The validated method was accurate, precise, and sensitive with a lower limit of quantification at 5 and 100 pg/mL for ethinyl estradiol and levonorgestrel, respectively.

Introduction

One method of contraception is the method of low-dose combined oral contraceptives (COCs), a combination of low-dose COCs containing synthetic estrogens such as ethinyl estradiol and synthetic progestogens, such as levonorgestrel or Norethisterone.[1] Examples of contraceptive drugs, including COCs, are a combination of oral contraceptives of ethinyl estradiol and levonorgestrel. This combination of drugs works synergistically by suppressing gonadotropin and inhibiting ovulation.[2]

Levonorgestrel does not undergo first-pass metabolism; hence, it is absorbed rapidly after oral administration (bioavailability, ±100%). In contrast, ethinyl estradiol is absorbed rapidly in the gastrointestinal tract but undergoes first-pass metabolism in the intestinal mucosa and liver (bioavailability, ±38%–48%). Levonorgestrel in serum is bound to sex hormone binding globulin (SHBG) and ethinyl estradiol is 97% bound with plasma albumin and induces SHBG synthesis.[3] The pharmacological and pharmacokinetics differences between ethinyl estradiol and levonorgestrel as a combination of oral contraceptives may have different effects on each individual, so it is needed for monitoring of drug therapy to ensure pharmacological response.[4]

The monitoring of drug therapy is based on the pharmacological and pharmacokinetic knowledge of a drug obtained by determining the concentration of the drug and its metabolite by means of bioanalysis method. Bioanalysis is performed on biological samples, one of which is blood plasma obtained by taking blood from a syringe through vein (venipuncture).[5] In addition, ethinyl estradiol and levonorgestrel are drugs that are mandatory for bioequivalence testing because they require a definite therapeutic response; hence, bioanalysis must be performed.[6] The combined doses of oral contraceptives of ethinyl estradiol and levonorgestrel are very low, so a highly sensitive and selective analysis method using ultra-performance liquid chromatography–tandem mass spectrometry (UPLC-MS/MS) is needed.[7]

The validated method for analyzing ethinyl estradiol and levonorgestrel in human plasma by UPLC-MS/MS has been carried out by several researchers using internal standard (IS)-labeled radioisotopes. However, because the price is very expensive, its use is often replaced by compound isomers, such as analog hormones for contraceptive compounds. This study aims to obtain optimum conditions and validated methods for the analysis of ethinyl estradiol and levonorgestrel in human plasma, simultaneously with prednisone as IS, using UPLC-MS/MS. Prednisone as an IS is used because it is easy to obtain and its price is cheaper. Validation parameters of analysis methods tested refer to the European Medicines Agency (EMEA), 2011, which are selectivity, linearity, lower limit of quantification (LLOQ), accuracy and precision, recovery, carry over, dilution integrity, matrix effect (ME), and stability.[8]

Materials and Methods

Chemical and reagents

Ethinyl estradiol and levonorgestrel (Beijing Keyifeng Biotech Development, Beijing, China) and prednisone as IS (Tianjin Tianyao Pharmaceuticals, Tianjin, China) were used. The mobile phase was formic acid and acetonitrile, which were obtained from Merck (Darmstadt, Germany). Human plasma with citrate anticoagulant was obtained from the Indonesian Red Cross (Jakarta, Indonesia). Reagents such as methanol, sodium chloride, ethyl acetate, n-hexane, and sodium bicarbonate were obtained from Merck, and dansyl chloride was obtained from Sigma Aldrich (Darmstadt, Germany).

Ultra-performance liquid chromatography–tandem mass spectrometry conditions

The experiment was performed on an ACQUITY UPLC system (Waters, Milford, Massachusetts) and a Xevo triple quadrupole (TQD) mass spectrometer (Waters, Manchester, United Kingdom) equipped with a positive electrospray ionization (ESI+). All data were acquired in centroid mode by the MassLynx NT4.1 software and analyzed by QuanLynx program (Waters). The analyte was separated on ACQUITY UPLC ethylene bridged hybrid C18 column (1.7 μm, 2.1 × 50mm, Waters). The mobile phase was composed of 0.1% formic acid solution and acetonitrile, flow rate was of 0.3mL/min, autosampler temperature was at 8°C, and injection volume was 10 µL. The gradient elution used within 5 min is shown in Table 1.

Table 1

The first gradient elution profile

Min to-	Mobile phase A (%)	Mobile phase B (%)
0.00	35	65
1.00	15	85
2.25	10	90
2.75	5	95
3.00	5	95
3.05	35	65
5.00	35	65

The mass spectrometric detector parameters were optimized and set as follows: nitrogen desolvation temperature as 498°C with a flow rate of 694L/h, dwell time of 0.044s, and capillary voltage of 3.50kV. The cone voltage was 56 V for ethinyl estradiol; 32 V for levonorgestrel; and 24 V for prednisone as IS, and collision energy was 38 V for ethinyl estradiol; 18 V for levonorgestrel; and 18 V for prednisone as IS. The detector was performed in positive-ion mode obtained by positive mode of ESI+ technique, and quantification was acquired with multiple reaction monitoring (MRM) with ion transition at 530.16 → 171.08 for ethinyl estradiol derivatized by dansyl chloride; m/z: 313.16 → 245.10 for levonorgestrel; and m/z: 359.10 → 147.04 for prednisone as IS, respectively.

Preparation of stock and working standard solution

Stock solution of ethinyl estradiol and levonorgestrel was prepared at 1.0mg/mL in methanol and then serially diluted with methanol to obtain the working solution of ethinyl estradiol with a concentration of 10ng/mL and levonorgestrel with a concentration of 200ng/mL. The stock solution of prednisone as IS was prepared at 1.0mg/mL in methanol and then serially diluted with methanol to obtain the working solution of prednisone with a concentration of 100 µg/mL. All solutions were stored at 4°C and brought to room temperature before use.

Calibration standards and quality control (QC) samples were used to estimate precision and accuracy of the method and were prepared from two separate sets of solutions in human plasma. Calibration standard samples of ethinyl estradiol (5, 10, 20, 50, 200, and 500 pg/mL) and levonorgestrel (100, 200, 400, 1,000, 4,000, and 10,000 pg/mL) were obtained by spiking 500 μL of the appropriate working solutions to 5,000 μL human plasma (calibration curve in plasma). The QC samples of ethinyl estradiol were prepared separately in the same procedure at a concentration of 15, 200, and 400 pg/mL for low, medium, and high QC (QCL, QCM, and QCH). The QC samples of levonorgestrel were prepared separately in the same procedure at a concentration of 300, 4,000, and 8,000 pg/mL for QCL, QCM, and QCH, respectively. All aliquot plasma solutions were stored at freezer –20°C and brought to room temperature before use.

Preparation of reagent solution

Methanol–water solution (80:20) was prepared by mixing methanol and water with a ratio of 80:20 to the amount required, and then in a homogeneous vortex. Sodium chloride (concentration of 100mM) was prepared by dissolving as much as 146.275mg sodium chloride in 25mL aquadest, and then in a homogenous vortex. Ethyl acetate–hexane solution (10:90) was prepared by mixing ethyl acetate and n-hexane with a ratio of 10:90 to the amount required, and then in a homogeneous vortex. All solutions were stored at room temperature (25°C). Sodium bicarbonate (concentration of 100mM; pH, ±11) was prepared by dissolving 210mg of sodium bicarbonate in 25mL aquadest, and then in a homogenous vortex. The pH of the solution was measured by using a pH meter device to obtain pH of ±11 and stored in refrigerator (4°C). Dansyl chloride (concentration of 1,000 µg/mL) was prepared by dissolving dansyl chloride (10mg) in acetonitrile (10mL), then in a homogenous vortex, and stored in freezer (–20°C).

Preparation of sample in human plasma

A 500 μL aliquot plasma was added of 50 μL methanol–water (80:20), 50 μL IS solution (100 µg/mL), and 50 μL sodium chloride (100mM), and vortex-mixed for 1min. Then, 2.5mL of ethyl acetate–hexane (10:90) was added in centrifugation tube. The mixture was vortex-mixed for 2min and centrifuged at 3,000rpm for 10min at 4°C. The supernatant phase was transferred into evaporation tube and evaporated to dryness under nitrogen at 45°C for 10min. The residue was reconstituted with 50 μL of natrium bicarbonate (100mM; pH, ±11) and 50 μL of dansyl chloride (1,000 μg/mL) in acetonitrile. The mixture was vortex-mixed for 1min and was incubated in water bath at 60°C for 10min. Methanol as much as 100 μL was added and vortex-mixed for 1min. The mixture was transferred into sample cup and centrifuged at 14,000rpm for 5 min. The supernatant phase was transferred into vial, and then 10 μL aliquot of the solution was injected into the UPLC-MS/MS system for analysis.

Method validation

Validation of the quantitative UPLC-MS/MS method was assessed including selectivity, linearity, LLOQ, accuracy and precision, recovery, carry over, dilution integrity, ME, and stability of the analytes in biological matrix, according to the EMEA guidelines on bioanalytical method validation by the Committee for Medicinal Products for Human Use.[8]

Selectivity

The selectivity of the method was evaluated by analyzing six blank plasma and spiked plasma at the LLOQ. The peak areas of the endogenous interference co-eluted with the analytes should be less than 20% of the peak area of the LLOQ standard and less than 5% of the peak area of the IS.[8]

Linearity

Calibration standards were prepared and analyzed by plotting the peak area ratios of the analyte to the IS versus the nominal concentration using a linearity weighted regression method in triplicate. Calibration curve was calculated with weighted linear curve fit equation, 1/x.[9] Calibration curves were considered acceptable when the correlation coefficient (r) was greater than 0.98 for biological matrix[10] and the bias of the calculated concentrations was within ±15% of the nominal concentrations, except the LLOQ with an allowed deviation of ±20%.[8]

Lower limit of quantification

LLOQ was established by analyzing blank plasma samples spiked with half or one-fourth of the lowest concentration of ethinyl estradiol and levonorgestrel in the sample. The analyte response should be identifiable, discrete, and reproducible with acceptable precision and accuracy (less than 20% for each criteria).[8]

Accuracy and precision

Accuracy and precision were evaluated by assessing five replicates of the QC samples at four concentration levels (LLOQ, low, medium, and high) on three consecutive validation days. Intra- and inter-day precisions were required, % coefficient of variance (%CV) was not to exceed 15%, and accuracy (%diff) should be within ±15% except the LLOQ with an allowed deviation of ±20%.[8]

Recovery

Recovery values (%) were calculated at three QC levels (QCL, QCM, and QCH) by comparing the peak areas of the regularly processed QC samples with those of spiked post-extraction samples. The %CV of the recovery values should be less than 15%.[2]

Carry over

Carry over was assessed by injecting blank samples after calibration standard at the upper limit of quantification. The measured peak area should not be greater than 20% of the peak area of the analyte at LLOQ and 5% of the peak area of the IS, respectively.[8]

Dilution integrity

The standard work solution of ethinyl estradiol and levonorgestrel was diluted in plasma until the concentration was above upper limit of quantification (ULOQ) and twice the concentration of QCH. Then, it was diluted to half the concentration and a quarter by using blank plasma. The test is performed in five replicates. Dilution should not affect accuracy and precision with the requirements of %diff and %CV not more than ±15%.[8]

Matrix effect

Blank plasma from six lots was extracted and then spiked with analyte at a concentration of QCL and QCH to evaluate the ME of the analyte. The peak area in spiked plasma post-extraction samples was then compared with those of standard solutions containing the analyte at equivalent concentrations. The %CV of the ME should not be more than ±15%.[8] The standardized matrix factor values with IS should be obtained in the acceptance range of 0.80–1.20.[11]

Stability

Stock solution stability of ethinyl estradiol, levonorgestrel, and prednisone was evaluated in short term at 0, 6, and 24h at room temperature (25°C) and long term at 1, 15, and 30 days at storage temperature (–20°C). The test was performed in two replicates and the %diff value should not be more than 10%.[1213] Sample stability was tested by analyzing the QCL and QCH after short-term storage (kept at room temperature for 0, 6, and 24h) and long-term storage (at freezer –20°C for days 1, 10, and 15). It was also tested by analyzing the QCL and QCH after three freeze–thaw cycles and autosampler stability (kept at autosampler temperature for 0 and 24h). The test was performed in three replicates and the %diff and %CV values should not be more than 15%.[8]

Results and Discussion

The UPLC-MS/MS is currently considered as the best choice for supporting bioanalytical studies due to its high specificity, sensitivity, and rapidity; however, only few reports are available on the determination of ethinyl estradiol and levonorgestrel by UPLC-MS/MS that was using analog compounds such as prednisone as IS.[14] This study described the development and validation of UPLC-MS/MS method for the quantitative analysis of ethinyl estradiol and levonorgestrel in human plasma.

Method development

Selection of IS

The usage of internal standard in LC-MS/MS should be a stable isotope labeled (SIL) compound or similar in physicochemical properties to the analyte.[14] Prednisone was chosen as IS because it has similar characteristics, including classes of steroid hormones. It was also chosen because it is easy to obtain and its price is cheaper.

Optimization of mass condition

To optimize ESI condition for ethinyl estradiol, levonorgestrel, and prednisone as IS, the MS parameters were tuned in positive-ionization mode. This positive ionization was related to their basic properties. The spectra showed a high-intensity signal at m/z: 530.16, 313.16, and 359.10 for ethinyl estradiol derivatized by dansyl chloride, levonorgestrel, and prednisone, respectively, as protonated molecular ions [M + H]+. The product ion of mass spectra for ethinyl estradiol, levonorgestrel, and prednisone, resulting from the fragmentation process, was observed at m/z: 171.08, 245.10, and 147.04, respectively [Figure 1A–C]. Following the optimization of mass spectrometry conditions, the quantification was acquired with MRM with ion transition at 530.16 → 171.08 for ethinyl estradiol derivatized by dansyl chloride, m/z: 313.16 → 245.10 for levonorgestrel, and m/z: 359.10 → 147.04 for prednisone as IS, respectively.

Figure 1

Fragmentation spectrum of (a) ethinyl estradiol as Internal Standard; (b) levonorgestrel as Internal Standard; (c) prednisone as Internal Standard

Ethinyl estradiol is derivatized by phenolic hydroxyl reaction in the secondary amine part of dansyl chloride through a nucleophilic aromatic substitution.[15] Dansyl chloride reacts with phenolic hydroxyl groups, primary amines, and secondary amines.[16] Ethinyl estradiol has low ionization capability and fragmentation of the core structure, so the process of derivatization by dansyl chloride produces a secondary nitrogen group that causes the molecule to be easily ionized.[17] The reaction of ethinyl estradiol derivatized by dansyl chloride is shown in Figure 2.

Figure 2

The reaction of ethinyl estradiol derivatizated by dansil chloride

Optimization of mobile phase combination

It tested three types of mobile phase combinations, such as 0.1% formic acid in water with acetonitrile, 0.2% formic acid in water with acetonitrile, and 0.1% formic acid in water with 0.1% formic acid in acetonitrile. This mobile phase was tested using isocratic methods with aqueous phase and acetonitrile as an organic phase (30:70). On the basis of the results, a combination of 0.1% formic acid in water with acetonitrile was chosen because it produces best chromatogram with the largest area.

Optimization of mobile phase composition

It tested three types of mobile phase compositions between formic acid (0.1%) in water (A) and acetonitrile (B), such as 20:80, 30:70, and 40:60. On the basis of area, the ethinyl estradiol and levonorgestrel produced at 0.1% formic acid in water with acetonitrile (30:70) was greater than that of the other mobile phase compositions.

Optimization of flow rate

It tested the variation of the flow rate, which was 0.2, 0.3, and 0.5mL/min. A 0.3mL/min flow rate was chosen because it produces the best chromatogram with large area and the retention time is not too fast or too long. Increasing the flow rate will result in a sharper peak and faster retention time, so that the analysis time is also faster, but the resulting area is smaller. Increased flow rate also causes greater column pressure, which can accelerate column damage.

Optimization of mobile phase gradient elution

In the development of this method, isocratic elution has been able to produce large areas and a good chromatogram, but the resulting peak still has tailing or fronting. Therefore, the next optimization method was carried out using gradient elution. The gradient elution profile is shown in Tables 1 and 2. The resulting area in the second elution profile was greater than the first elution profile, but the retention time on the first elution profile was faster than the second elution profile. This indicated the separation or elution process in the first elution profile (gradual increase of the organic phase) was less suitable for the ethinyl estradiol, levonorgestrel, and prednisone than the second elution profile (the increase in the organic phase directly). Hence, the second gradient elution profile was chosen.

Table 2

The second gradient elution profile

Min to-	Mobile phase A (%)	Mobile phase B (%)
0.00	75	25
1.50	5	95
3.00	5	95
3.05	75	25
5.00	75	25

System suitability test

After obtaining optimum conditions of analysis, the system suitability was tested to ensure the system works well to produce accurate data. On the basis of the test results, the %CV of the area produced by ethinyl estradiol, levonorgestrel, and prednisone was 2.10%, 3.70%, and 1.12%, respectively, whereas the %CV of the retention time produced by ethinyl estradiol, levonorgestrel, and prednisone was 0.00%, 0.19%, and 0.00%, respectively. From the results obtained, it can be concluded that the system is running well because it meets the requirements of %CV not more than 6%.[18]

Optimization of sample preparation

Development of this method was tested based on the different methods of extraction and mixing method. Sample preparation methods were tested with protein precipitation followed by twice of liquid–liquid extraction and protein precipitation followed by once of liquid–liquid extraction. Sample preparation methods were also tested by mixing with vortex for 2min and rotospin for 10min. On the basis of the test results, protein precipitation and once of liquid–liquid extraction was selected as well as mixing with vortex for 2min because it produces the largest area at concentrations of LLOQ and ULOQ.

The representative chromatograms resulting from the UPLC-MS/MS analysis of 500 μL plasma from blank plasma sample and spiked LLOQ of ethinyl estradiol, levonorgestrel, and prednisone are given in Figure 3A and B. No significantly interfering peaks because of the endogenous components or reagents were observed for ethinyl estradiol, levonorgestrel, and IS.

Figure 3

Representative UPLC-MS/MS chromatograms of ethinyl estradiol, levonorgestrel, and prednisone in (a) human blank plasma spiked with analyte at LLOQ; (b) human plasma spiked with analyte at LLOQ

The calibration curve over the concentration range of 5–500 pg/mL for ethinyl estradiol and 100–10,000 pg/mL for levonorgestrel was linear and acceptable. The calibration equation obtained was correlation coefficient (r) >0.98 and %diff of the calculated concentrations was acceptable. Data of inter-day calibration curve of ethinyl estradiol and levonorgestrel are shown in Table 3.

Table 3a

Data of inter-day calibration curve of ethinyl estradiol

Replica inter-day	R	Slope (b)	Intercept (a)
1	0.9958	0.5066	−0.0428
2	0.9984	0.3883	0.0078
3	0.9989	0.6911	−0.0110
Mean	0.9977	0.5287	−0.0153

The LLOQ was 5 pg/mL for ethinyl estradiol and 100 pg/mL for levonorgestrel, %CV and %diff of back calculated concentrations of LLOQ were 2.17% and <20% for ethinyl estradiol and 5.49% and <20% for levonorgestrel, respectively [Table 4]. On the basis of the EMEA (2011) requirements, LLOQ concentrations were achieved already.[8] Another study conducted by Dara et al.,[19] in 2014, and Huang et al.,[20] in 2016, developed the analysis of ethinyl estradiol in human plasma using liquid–liquid extraction and solid phase extraction (SPE) and produced LLOQ concentration of 5 pg/mL. Another study conducted by Praveen Kumar et al.,[21] in 2014, and Moser et al.,[22] in 2011, developed the analysis of levonorgestrel in human plasma using SPE and produced LLOQ concentrations of 100 and 50 pg/mL, respectively. Compared with this study, LLOQ concentration obtained in this study has an advantage because of its preparation methods used, that is, protein precipitation and liquid–liquid extraction, which provide cost efficiency over SPE method.

Table 4a

The accuracy and precision from LLOQ of ethinyl estradiol

LLOQ conc. (pg/mL)	Measured conc. (pg/mL)	Accuracy (%diff)	Precision (%CV)
5.00	4.51	−9.75	2.17
	4.76	−4.71
	4.73	−5.32
	4.62	−7.54
	4.62	−7.61

The intra- and inter-day accuracy and precision are shown in Table 5. The data show that the accuracy and precision values are within the acceptable criteria.

Table 5a

The intra- and inter-day accuracy and precision of ethinyl estradiol

Nominal conc. (pg/mL)	Intra-day	Inter-day
	Mean accuracy (%diff)	Precision (%CV)	Mean accuracy (%diff)	Precision (%CV)
5.00	−3.81 to 9,01	5.73	−15.05 to 9.01	6.27
15.00	−14.79 to 14.69	12.45	−14.80 to 14.69	9.21
200.00	−12.59 to 5.69	7.64	−14.72 to 11.63	7.85
400.00	−14.56 to 1.68	7.76	−14.84 to 4.74	7.06

The mean extraction recoveries of ethinyl estradiol were 68.03%, 84.74%, and 77.76% (n = 3) at the concentrations of QCL, QCM, and QCH, with %CV values of 2.56%, 5.86%, and 7.35%, respectively. The mean extraction recoveries of levonorgestrel were 88.83%, 87.81%, and 89.71% (n = 3) at the concentrations of QCL, QCM, and QCH, with %CV values of 8.92%, 5.42%, and 2.96%, respectively, whereas for the IS, it was 83.14% at a concentration of 1,000 µg/mL with %CV value of 3.09%.

The measured peak area of the blank sample injected after calibration standard at the ULOQ (500 pg/mL) was 16.414% of the peak area of the analyte at LLOQ for ethinyl estradiol; ULOQ (10,000 pg/mL) was 6.580% of the peak area of levonorgestrel and 1.251% of the peak area of the IS, respectively.

The dilution integrity testing results were acceptable because the dilution still fulfilled accuracy and precision requirements with %diff and %CV not more than 15%, which was diluted in human blank plasma until the concentration of QCH and a half of QCH.

The mean MEs of ethinyl estradiol were 100.39% and 102.60% at the concentration of QCL and QCH, with %CV of 1.66% and 9.37%, respectively. The mean MEs of levonorgestrel were 102.02% and 108.00% at the concentration of QCL and QCH, with %CV of 3.78% and 4.55%, respectively, whereas for the IS, it was 105.25% with %CV of 4.99%. These data indicate that the ME (ion suppression or enhancement) from human plasma was negligible under the current conditions.

Storage of stock solutions of ethinyl estradiol, levonorgestrel, and prednisone in methanol at room temperature for 24h and in refrigerator (–4°C) for 1 month did not alter the analyte of ethinyl estradiol and levonorgestrel; however, prednisone as IS was stable until 15 days. The stability test results of ethinyl estradiol and levonorgestrel in human plasma are given in Table 6. The data indicate that ethinyl estradiol and levonorgestrel are stable enough during sample preparation and storage conditions.

Table 6

The stability test results of ethinyl estradiol and levonorgestrel in human plasma

Stability experiments	Stable to-
	Ethinyl estradiol	Levonorgestrel
Short-term storage (24h, 25 °C)	24 h	24 h
Long-term storage (freezer −20 °C)	10 days	15 days
Autosampler (24h)	24 h	24 h
Three freeze-thaw cycles	3 cycles	3 cycles

Conclusion

In conclusion, the UPLC-ESI-MS/MS method for the quantitative analysis of ethinyl estradiol and levonorgestrel in human plasma was successfully developed and validated. This method provides very rapid, sensitive, and specific measurements of ethinyl estradiol and levonorgestrel concentrations. The LLOQ obtained in this study was 5 pg/mL for ethinyl estradiol and 100 pg/mL for levonorgestrel with sample preparation of protein precipitation and liquid–liquid extraction, and faster analysis time of 5 min. Prednisone as IS is cheaper and easy to be found, can be used for ethinyl estradiol and levonorgestrel analysis, and is able to control MEs.

Ethical approval and consent

We hereby declare that all experiments have been examined and approved by the Ethics Committee of the Faculty of Medicine, Universitas Indonesia (no.: 0034/UN2.F1/ETIK/2018, protocol no.: 18-01-0037), and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. A written informed consent was obtained from the subjects before inclusion into the study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Table 3b

Data of inter-day calibration curve of levonorgestrel

Replica inter-day	R	Slope (b)	Intercept (a)
1	0.9984	0.0023	−0.000012
2	0.9992	0.0015	0.000100
3	0.9974	0.0021	0.000043
Mean	0.9983	0.0020	0.000044

Table 4b

The accuracy and precision from LLOQ of levonorgestrel

LLOQ conc. (pg/mL)	Measured conc. (pg/mL)	Accuracy (%diff)	Precision (%CV)
100.00	92.98	−7.02	5.49
	90.71	−9.29
	104.29	4.29
	93.68	−6.32
	96.10	−3.90

Table 5b

The intra- and inter-day accuracy and precision of levonorgestrel

Nominal conc. (pg/mL)	Intra-day	Inter-day
	Mean accuracy (%diff)	Precision (%CV)	Mean accuracy (%diff)	Precision (%CV)
100.00	−13.16 to −0.02	5.33	−13.16 to 13.22	8.80
300.00	12.70 to 13.47	12.13	−12.70 to 14.60	8.91
4000.00	−7.00 to 14.05	8.33	−13.60 to 14.05	8.31
8000.00	8.71 to 14.53	2.24	−14.75 to 14.53	10.82