International Journal of Pharmaceutics 360 (2008) 77–82
`
`Contents lists available at ScienceDirect
`
`International Journal of Pharmaceutics
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / i j p h a r m
`
`Stability indicating validated HPLC method for quantification of levothyroxine
`with eight degradation peaks in the presence of excipients
`R.B. Shah a, A. Bryant a, J. Collier a,b, M.J. Habib b, M.A. Khan a,∗
`
`a Division of Product Quality Research, Office of Testing and Research, Office of Pharmaceutical Sciences, Center for Drug Evaluation and Research,
`Food and Drug Administration, United States
`b Department of Pharmaceutical Sciences, School of Pharmacy, Howard University, United States
`
`a r t i c l e
`
`i n f o
`
`a b s t r a c t
`
`Article history:
`Received 12 December 2007
`Received in revised form 20 March 2008
`Accepted 12 April 2008
`Available online 20 April 2008
`
`Keywords:
`Levothyroxine
`Impurities
`Degradation products
`Validation
`Excipients
`
`A simple, sensitive, accurate, and robust stability indicating analytical method is presented for identi-
`fication, separation, and quantitation of l-thyroxine and eight degradation impurities with an internal
`standard. The method was used in the presence of commonly used formulation excipients such as
`butylated hydroxyanisole, povidone, crospovidone, croscarmellose sodium, mannitol, sucrose, acacia,
`lactose monohydrate, confectionary sugar, microcrystalline cellulose, sodium laurel sulfate, magnesium
`stearate, talc, and silicon dioxide. The two active thyroid hormones: 3,3(cid:3),5,5(cid:3)-tetra-iodo-l-thyronine (l-
`thyroxine-T4) and 3,3(cid:3),5-tri-iodo-l-thyronine (T3) and degradation products including di-iodothyronine
`(T2), thyronine (T0), tyrosine (Tyr), di-iodotyrosine (DIT), mono-iodotyrosine (MIT), 3,3(cid:3),5,5(cid:3)-tetra-
`iodothyroacetic acid (T4AA) and 3,3(cid:3),5-tri-iodothyroacetic acid (T3AA) were assayed by the current
`method. The separation of l-thyroxine and eight metabolites along with theophylline (internal stan-
`dard) was achieved using a C18 column (25 ◦C) with a mobile phase of trifluoroacetic acid (0.1%, v/v,
`pH 3)–acetonitrile in gradient elution at 0.8 ml/min at 223 nm. The sample diluent was 0.01 M methanolic
`NaOH. Method was validated according to FDA, USP, and ICH guidelines for inter-day accuracy, preci-
`sion, and robustness after checking performance with system suitability. Tyr (4.97 min), theophylline
`(9.09 min), MIT (9.55 min), DIT (11.37 min), T0 (11.63 min), T2 (14.47 min), T3 (16.29 min), T4 (17.60 min),
`T3AA (22.71 min), and T4AA (24.83 min) separated in a single chromatographic run. Linear relationship
`(r2 > 0.99) was observed between the peak area ratio and the concentrations for all of the compounds
`within the range of 2–20 ␮g/ml. The total time for analysis, equilibration and recovery was 40 min. The
`method was shown to separate well from commonly employed formulation excipients. Accuracy ranged
`from 95 to 105% for T4 and 90 to 110% for all other compounds. Precision was <2% for all the compounds.
`The method was found to be robust with minor changes in injection volume, flow rate, column tem-
`perature, and gradient ratio. Validation results indicated that the method shows satisfactory linearity,
`precision, accuracy, and ruggedness and also stress degradation studies indicated that the method can be
`used as stability indicating method for l-thyroxine in the presence of excipients.
`Published by Elsevier B.V.
`
`1. Introduction
`
`Levothyroxine sodium pentahydrate, the sodium salt of the
`levo-isomer of thyroxine is an active physiological substance
`secreted by thyroid gland. With three ionizable moieties: carboxyl
`group (pKa = 2.4), phenolic group (pKa = 6.87) and amino group
`(pKa = 9.96), its aqueous solubility reduces from pH 1 to 3 and
`increases above pH of 7 (Patel et al., 2003). Another thyroid hor-
`
`∗ Corresponding author at: 10903 New Hampshire Avenue, Life Sciences Building
`64, Silver Spring, MD 20993-0002, United States. Tel.: +1 301 796 0132;
`fax: +1 301 796 9816.
`E-mail address: Mansoor.khan@fda.hhs.gov (M.A. Khan).
`
`0378-5173/$ – see front matter. Published by Elsevier B.V.
`doi:10.1016/j.ijpharm.2008.04.018
`
`mone, 3,3(cid:3),5-tri-iodo-l-thyronine (T3), is also pharmacologically
`active. The precursors or metabolites include di-iodothyronine (T2),
`the parent compound of the iodinated series of thyroid-active
`hormones, thyronine (T0), tyrosine (Tyr), di-iodotyrosine (DIT),
`mono-iodotyrosine (MIT) (Gika et al., 2005), as well as 3,3(cid:3),5-tri-
`iodo-l-thyroacetic acid, and 3,3(cid:3),5,5(cid:3)-tetra-iodo-l-thyroacetic acid
`have no pharmacological activity.
`Stability is considered one of the most important requirements
`of pharmaceutical product quality. Only stable preparations would
`promise precise delivery of the drug to the patients. Expiration dat-
`ing on any drug product is based upon scientific studies at normal
`and or stressed conditions of certain batches and strengths of prod-
`ucts that are developed in multiple strengths. Levothyroxine is one
`such example where products are available in multiple strengths.
`
`Mylan Ex 1016, Page 1
`
`

`
`78
`
`R.B. Shah et al. / International Journal of Pharmaceutics 360 (2008) 77–82
`
`Previous studies have shown that different dosage forms of levothy-
`roxine are susceptible to degradation under the influence of various
`environmental stress factors such as humidity and temperature.
`Won (1992) reported that levothyroxine degrades with high tem-
`perature and extremes of pH. Levothyroxine has been a subject
`of FDA Advisory Committee meetings where the clinical conse-
`quences of marketing product with approved specification limits
`of 90–110% has been reported as a problem. There were numerous
`recalls of levothyroxine due to stability issues (FDA, 2006). Further,
`lacks of potency and stability assurances has brought in concerns
`from physicians regarding their therapeutic substitutions and are
`believed not to deliver right doses to the patients (Thyroid, 2004).
`In order to understand the degradation mechanisms of levothy-
`roxine systematically, there is a need for a reliable and simple
`validated stability indicating method (ICH Q1A (R2), 2003). The
`stability indicating method should not only identify the degra-
`dation products of levothyroxine but also quantitate them. There
`are numerous reported methods to assay levothyroxine (Takahashi
`et al., 2002; Smith et al., 1981; Rapaka et al., 1981; Garnick et
`al., 1984; Richheimer and Amer, 1983). However, these methods
`either require derivatization of levothyroxine and liothyroxine for
`separation on HPLC (Takahashi et al., 2002; Smith et al., 1981)
`or lengthy and more tedious extraction–evaporation procedures
`before injecting into HPLC (Rapaka et al., 1981). There are meth-
`ods reported to assay content uniformity (Garnick et al., 1984)
`which cannot be used as stability indicating method. Richheimer
`and Amer (1983) reported a stability indicating assay method for
`levothyroxine. However, it is limited by the number of impurities.
`Quantification of impurities was not proposed in this method. Thin
`layer chromatography (TLC) has traditionally been used to identify
`degradation kinetics of levothyroxine (Won, 1992). However, it is
`not very accurate method to quantify the related compounds.
`Thus none of the previously reported methods satisfied the
`criteria of stability indicating methods. Very recently, a novel
`HPLC-based assay to quantify the impurities of levothyroxine was
`reported in the literature (Gika et al., 2005). The method included
`quantification of levothyroxine and six of its degradation products.
`However, in an attempt to reproduce the method in our laboratory,
`it was found to be erroneous in the order of mobile phase gradi-
`ent. The other limitation of the method was that the major acidic
`impurities of levothyroxine, namely, tri-iodo thyroacetic acid and
`tetra-iodo thyroacetic acid were not a part of impurity profile. The
`purpose of the current work was to modify this assay method to
`include these two impurities and also demonstrate that the assay
`was stability indicating as per FDA and ICH guidelines. Stress con-
`ditions used were high temperatures, acid and base hydrolysis,
`oxidation, and photolysis (Bakshi and Singh, 2002). Also some of the
`commonly used formulation excipients were mixed with levothy-
`roxine and eight impurities, and the chromatography was evaluated
`with a good resolution of all the peaks.
`
`2. Materials and methods
`
`l-Thyroxine sodium (l-T4) was obtained from KVPharmaceutical
`(St. Louis, MO). 3,3(cid:3),5-Tri-iodo-l-thyronine (l-T3) 3,5-di-iodo-l-
`thyronine (l-T2), 3,5-di-iodo-l-tyrosine (l-DIT), 3-iodo-l-tyrosine
`(l-MIT), l-thyronine (l-T0), l-tyrosine (l-Tyr), 3,3(cid:3),5-tri-iodo-l-
`thyroacetic acid, and 3,3(cid:3),5,5(cid:3)-tetra-iodo-l-thyroacetic acid, buty-
`lated hydroxyanisole, mannitol, sucrose, acacia, sodium laurel sul-
`fate, magnesium stearate, Inertsil 5 ␮m column, 250 mm× 4.6 mm,
`and security guard cartridge were purchased from Sigma (St.
`Louis, MO). Theophylline reagents, methanol, 0.01 M NaOH, 0.1%
`trifluoroacetic acid (TFA), Acetonitrile, and Fisherbrand low adhe-
`sion specialty tips (21-381-83) were purchased from Fisher Sci
`
`(Suwanee, GA). Povidone (BASF, Florham Park, NJ), crospovidone
`(ISP technologies Inc., Wayne, NJ), lactose monohydrate (Kerry
`BioScience, Chicago, IL), confectionary sugar (Domino’s sugar, Balti-
`more, MD), talc (Spectrum Chemicals, Gardena, CA), silicon dioxide
`(Aerosil, Evonik Degussa, Orange, CA) croscarmellose sodium, and
`microcrystalline cellulose (FMC Biopolymer, Philadelphia, PA),were
`used as received. For all studies, distilled and deionized water was
`used.
`
`2.1. Preparation of calibration standards
`
`In all cases the sample diluent used for preparing the standards
`and samples was the 0.01 M methanolic sodium hydroxide solution,
`which was prepared as described in USP monograph (USP, 2007).
`Two stock solutions (I and II) of each of nine components (T4, T3,
`T2, T0, MIT, DIT, T3AA, T4AA, and Tyr) prepared at 1000 ␮g/ml were
`prepared by dissolving them individuallly in the sample diluent.
`From the stock solution I, a working mix I was prepared by mix-
`ing 10 ml of each of these components and making the volume to
`100 ml. In a similar way working mix II was prepared from stock II.
`This was used on 3 different days but final dilutions were made on
`each day of validation. Working mix I was used for the calibration
`standards, and working mix II was used for quality control samples.
`Six different standard solutions were prepared from the working I
`to yield all the nine components in a concentration range from 2
`to 20 ␮g/ml. An internal standard, theophylline, was also added to
`all the above diluted calibration ranges. The standards were then
`transferred to an automatic injector for HPLC analysis.
`
`2.2. Preparation of quality control (QC) standards
`
`Three quality control standards were prepared from the working
`mix II to yield concentrations of 8, 10, and 12 ␮g/ml with 10 ␮g/ml
`as target concentration (100%). These were then transferred to an
`automatic injector for HPLC analysis.
`
`2.3. Preparation of resolution mixture and system suitability
`standard
`
`A combination solution containing all nine components at
`10 ␮g/ml each and theophylline was prepared from stock solution
`I and was used as system suitability standard.
`
`2.4. Chromatography
`
`HP 1100 HPLC equipment from Agilent (Wilmington, DE) con-
`sisted of quaternary pump, an automatic injector, a diode array
`wavelength detector, and a column oven. Various columns and
`mobile phases were tested. Finally, the method was validated
`with a reversed phase Inertsil ODS 2 column (250 mm× 4.6 mm,
`5 ␮m, 150 A) with a Inertsil ODS Security Guard cartridge
`(4.0 mm× 3.0 mm, 10 ␮m). It provided baseline separation with
`gradient conditions with 0.1% TFA (A) and acetonitrile (B) from 92
`to 8% A in 25 min, at 8% A from 25 to 30 min, from 8 to 92% A from
`30 to 35 min run time of 40 min for all the nine components and
`IS in a single chromatographic run. The flow rate was 0.8 ml/min,
`column temperature was 25 ◦C and the injection volume was 50 ␮l.
`The UV detection wavelength was set at 215, 223, 228, 232, and 240.
`However, all the calculations were performed at 223 nm.
`
`2.5. Validation
`
`Validation was carried out according to ICH and FDA guidelines
`for chromatographic methods (Bakshi and Singh, 2002). Speci-
`ficity, selectivity, linearity, accuracy, precision, and robustness were
`
`Mylan Ex 1016, Page 2
`
`

`
`R.B. Shah et al. / International Journal of Pharmaceutics 360 (2008) 77–82
`
`79
`
`established for the method. System suitability and resolution was
`performed utilizing related compound C and ranitidine HCl as the
`standards.
`
`2.6. Stress degradation studies
`
`Stress conditions applied for degradation of levothyroxine pow-
`der include refluxing it (1 mg) at room temperature under acidic
`(0.1N HCl, 24 h) and alkaline (0.1N NaOH, 24 h) conditions, oxida-
`tion (3% hydrogen peroxide, 24 h), and photolysis (exposure to UV-A
`and UV-B rays). Also the degradation was carried out at 40 ◦C for a
`period of 14 h. All these samples were appropriately diluted with
`sample diluent and injected into the HPLC.
`
`2.7. Excipient analysis
`
`Some of the commonly used formulation excipients were
`selected based on the commercial product inserts. They included
`butylated hydroxyanisole, povidone, crospovidone, croscarmellose
`sodium, mannitol, sucrose, acacia, lactose monohydrate, confec-
`tionary sugar, microcrystalline cellulose, sodium laurel sulfate,
`magnesium stearate, talc, and silicon dioxide. Initially, all the excip-
`ients were diluted in the sample diluent, filtered, and analyzed
`on HPLC for detection and evaluating their retention times. Fol-
`lowing that, l-thyroxine along with internal standard and all the
`eight degradation compounds were mixed with the excipient in
`1:1 ratios and were analyzed on HPLC as described earlier. The
`detector was set at multiple wavelengths of 215, 223, 228, 232,
`and 240 so as to ensure non-interference of excipients with either
`the active pharmaceutical ingredient (API), internal standard or
`impurities.
`
`3. Results and discussion
`
`3.1. Analytical method development
`
`Levothyroxine products have a history of stability failures which
`could result into sub-potency products with patients receiving less
`than optimal dose. The problem is aggravated while switching
`from one product to another although they are listed as ther-
`apeutic equivalents. To understand mechanism of levothyroxine
`degradation, an assay which will not only identify but quan-
`tify the impurities is essential. A reported method (Gika et al.,
`2005) was corrected and used with modifications to add two
`additional degradation products of levothyroxine, T3AA and T4AA.
`These two are considered to be significant degradation prod-
`ucts for levothyroxine. Five different wavelengths were used to
`observe the chromatograms, but only 223 nm was used for calcu-
`lation purpose. Although the peak areas were highest at 215 nm,
`the baseline showed very high negative drift. The wavelength
`of 223 also showed comparatively higher peak area for all the
`impurities as well as levothyroxine compared to 228, 232, and
`240 nm. Therefore, that wavelength was used. However, the detec-
`tion at all wavelengths was continued considering that stability
`samples might show some impurities which might be detected
`at one wavelength as opposed to other. Fig. 1 depicts the chro-
`matogram obtained with the current method. The peaks of all
`the impurities and levothyroxine were well resolved. This method
`was further validated as given in Section 2.7. The order of elu-
`tion was Tyr (4.97 min), theophylline (9.09 min), MIT (9.55 min),
`DIT (11.37 min), T0 (11.63 min), T2 (14.47 min), T3 (16.29 min), T4
`(17.60 min), T3AA (22.71 min), and T4AA (24.83 min). The current
`method can be used to assay levothyroxine and its major degrada-
`tion products.
`
`Fig. 1. Typical chromatogram of the iodothyronines and iodotyrosines separation
`using the HPLCs. All analytes are at 10 ␮g/ml. Peaks: Tyr (4.97 min), theophylline
`(9.09 min), MIT (9.55 min), DIT (11.37 min), T0 (11.63 min), T2 (14.47 min), T3
`(16.29 min), T4 (17.60 min), T3AA (22.71 min), and T4AA (24.83 min).
`
`3.2. Analytical method validation
`
`Specificity was established by determining that levothyroxine,
`internal std and degradation products have no co-eluting peaks
`in preparative solvents, mobile phase (blanks), or related matrices
`(Fig. 1).
`Selectivity was tested by running solutions containing the eight
`impurities and one internal standard in the same quantities and
`conditions as the samples to show that there was no peak at the
`retention times corresponding to the API.
`The detection limit (LOD) was 1 ␮g/ml which was evaluated by
`measuring the baseline noise and by calculating the analyte con-
`centration that gave S/N = 3, while the limit of quantification (LOQ)
`was 2 ␮g/ml which was established for the analyte concentration
`that gave S/N = 10.
`Linearity was established across the analytical calibration range.
`At least five non-zero calibration standards and a zero calibration
`standard and or blanks were utilized for each calibration curve.
`Table 1 shows the calibration curves of levothyroxine and all eight
`degradation compounds on 3 different days showing a linear corre-
`lation with R2 > 0.99 for all the components. Range was established
`by demonstrating a suitable level of accuracy, precision, and linear-
`ity.
`
`Accuracy and precision of the analytical method was established
`across its analytical range (Table 2). The accuracy was measured at
`each quality control (QC) standard level (n = 3) over the analyti-
`cal range as defined by the 80% of target concentration (8 ␮g/ml),
`l00% of target concentration (10 ␮g/ml), and 120% of target concen-
`tration (12 ␮g/ml), against the calibration curve. The levels were
`selected based on FDA and ICH guidelines (FDA, 1994; ICH Q2 (R1),
`1995). Nominal values are no greater than 15% at the LLOQ and 10%
`at the low, intermediate and high QC levels.
`
`Table 1
`Linearity and sensitivity data
`
`Analyte
`
`Calibration range (␮g/ml)
`
`Tyr
`MIT
`DIT
`T0
`T2
`T3
`T4
`T3AA
`T4AA
`
`2–20
`2–20
`2–20
`2–20
`2–20
`2–20
`2–20
`2–20
`2–20
`
`Slope
`0.051 ± 0.004
`0.124 ± 0.008
`0.192 ± 0.014
`0.149 ± 0.009
`0.183 ± 0.013
`0.187 ± 0.012
`0.157 ± 0.011
`0.208 ± 0.014
`0.192 ± 0.014
`
`Intercept
`0.016 ± 0.008
`0.015 ± 0.002
`0.009 ± 0.012
`0.002 ± 0.003
`0.008 ± 0.009
`0.010 ± 0.003
`0.014 ± 0.006
`0.007 ± 0.003
`0.012 ± 0.010
`
`R2
`
`0.9992
`0.9999
`1.0000
`0.9998
`0.9999
`0.9998
`0.9999
`1.0000
`1.0000
`
`Mylan Ex 1016, Page 3
`
`

`
`80
`
`R.B. Shah et al. / International Journal of Pharmaceutics 360 (2008) 77–82
`
`Table 2
`Accuracy and precision data
`
`Inter-day (n = 9)
`
`80
`
`100.09 ± 3.04
`102.18 ± 1.34
`100.19 ± 2.85
`98.19 ± 3.18
`99.46 ± 3.32
`99.26 ± 3.28
`100.25 ± 2.96
`97.30 ± 2.91
`103.61 ± 3.02
`
`Intra-day (n = 3)
`
`80
`
`96.38 ± 0.24
`100.52 ± 0.13
`96.40 ± 0.07
`94.08 ± 0.05
`95.04 ± 0.03
`94.94 ± 0.34
`96.31 ± 0.00
`93.55 ± 0.04
`99.59 ± 0.03
`
`Accuracy (%)
`Tyr
`MIT
`DIT
`T0
`T2
`T3
`T4
`T3AA
`T4AA
`
`Precision (%)
`Tyr
`MIT
`DIT
`T0
`T2
`T3
`T4
`T3AA
`T4AA
`
`100
`
`120
`
`100.72 ± 0.56
`100.57 ± 0.12
`99.82 ± 0.11
`97.36 ± 0.09
`98.39 ± 0.05
`98.20 ± 0.28
`99.91 ± 0.10
`96.86 ± 0.20
`103.22 ± 0.15
`
`101.62 ± 1.33
`100.57 ± 0.02
`100.53 ± 0.07
`98.25 ± 0.10
`99.21 ± 0.10
`100.01 ± 0.41
`100.71 ± 0.08
`97.58 ± 0.06
`103.90 ± 0.05
`
`Intra-day (n = 6)
`
`0.94
`0.12
`0.08
`0.09
`0.08
`0.29
`0.10
`0.16
`0.12
`
`100
`
`120
`
`100.63 ± 1.47
`100.62 ± 0.72
`100.04 ± 0.45
`98.19 ± 1.35
`99.43 ± 0.98
`99.31 ± 1.27
`100.38 ± 0.60
`97.29 ± 1.07
`103.55 ± 0.51
`
`102.92 ± 2.21
`101.28 ± 0.54
`101.00 ± 0.46
`99.28 ± 0.98
`100.47 ± 0.95
`100.59 ± 0.60
`101.41 ± 0.56
`98.29 ± 0.68
`104.52 ± 0.53
`
`Inter-day (n = 18)
`
`1.48
`0.70
`0.43
`1.33
`0.95
`1.22
`0.58
`1.07
`0.47
`
`The system suitability standard contains levothyroxine, inter-
`nal standard, and all eight degradation products. Table 3 depicts
`the system suitability and resolution factors. The specifications are
`also given in Table 3. It was observed that all the parameters passed
`the USP specifications. An important system suitability parameter
`is resolution, a measure of how well two peaks are separated. For a
`reliable quantification, well-separated peaks are essential. This is a
`
`very useful parameter if potential interference peak may be of con-
`cern. The degradation impurities of levothyroxine were selected to
`measure the resolution parameter. It is desirable to have resolu-
`tion of >1 between the two peaks. With the current method, we
`obtained satisfactory resolution of >1 in all cases. The tailing fac-
`tor was also considered as the accuracy of quantification decreases
`with increase in peak tailing because of the difficulties encountered
`
`Table 3
`System suitability parameters
`
`RT %R.S.D.
`
`Peak area %R.S.D.
`
`USP tailing
`1.00 ± 0.00
`1.01 ± 0.01
`1.00 ± 0.00
`1.07 ± 0.03
`1.10 ± 0.01
`1.07 ± 0.01
`1.23 ± 0.00
`1.17 ± 0.00
`1.17 ± 0.00
`1.22 ± 0.00
`1.16 ± 0.00
`1.15 ± 0.00
`1.32 ± 0.00
`1.25 ± 0.00
`1.25 ± 0.01
`1.35 ± 0.01
`1.29 ± 0.00
`1.31 ± 0.01
`1.38 ± 0.01
`1.34 ± 0.00
`1.38 ± 0.01
`1.30 ± 0.01
`1.24 ± 0.00
`1.26 ± 0.00
`1.29 ± 0.01
`1.23 ± 0.08
`1.26 ± 0.00
`
`Theoretical plates (X05)
`3.03 ± 0.04
`2.54 ± 0.05
`3.07 ± 0.05
`1.64 ± 0.02
`1.79 ± 0.02
`1.34 ± 0.05
`2.54 ± 0.01
`2.61 ± 0.02
`2.55 ± 0.02
`2.78 ± 0.02
`2.84 ± 0.03
`2.78 ± 0.00
`3.73 ± 0.03
`3.72 ± 0.06
`3.56 ± 0.02
`4.26 ± 0.07
`4.16 ± 0.04
`3.93 ± 0.03
`4.37 ± 0.06
`4.18 ± 0.07
`3.90 ± 0.03
`5.76 ± 0.09
`5.64 ± 0.06
`5.40 ± 0.05
`6.38 ± 0.09
`6.18 ± 0.08
`5.93 ± 0.02
`
`Resolution
`
`Selectivity
`
`1.82 ± 0.04
`1.05 ± 0.00
`1.03 ± 0.00
`1.20 ± 0.00
`1.19 ± 0.00
`1.20 ± 0.00
`1.02 ± 0.00
`1.02 ± 0.00
`1.02 ± 0.00
`1.25 ± 0.00
`1.24 ± 0.00
`1.25 ± 0.00
`1.13 ± 0.00
`1.12 ± 0.00
`1.13 ± 0.00
`1.08 ± 0.00
`1.08 ± 0.00
`1.08 ± 0.00
`1.30 ± 0.00
`1.30 ± 0.00
`1.29 ± 0.00
`1.09 ± 0.00
`1.09 ± 0.00
`1.09 ± 0.00
`
`1.03 ± 0.00
`2.49 ± 0.05
`1.55 ± 0.02
`10.3 ± 0.05
`10.1 ± 0.07
`10.3 ± 0.04
`1.54 ± 0.02
`1.48 ± 0.01
`1.53 ± 0.01
`15.9 ± 0.08
`15.6 ± 0.14
`15.6 ± 0.06
`9.52 ± 0.06
`9.30 ± 0.10
`9.18 ± 0.04
`6.43 ± 0.04
`6.21 ± 0.07
`6.14 ± 0.03
`23.6 ± 0.29
`22.8 ± 0.10
`22.2 ± 0.14
`8.97 ± 0.06
`8.69 ± 0.05
`8.58 ± 0.03
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Day-1
`Day-2
`Day-3
`
`Tyr
`
`MIT
`
`DIT
`
`T0
`
`T2
`
`T3
`
`T4
`
`T3AA
`
`T4AA
`
`Spec
`
`0.05
`0.05
`0.08
`
`0.15
`0.23
`0.13
`
`0.09
`0.17
`0.05
`
`0.21
`0.16
`0.05
`
`0.26
`0.09
`0.03
`
`0.03
`0.04
`0.04
`
`0.02
`0.02
`0.04
`
`0.05
`0.07
`0.04
`
`0.04
`0.06
`0.04
`
`1.38
`1.97
`1.02
`
`0.24
`0.07
`0.18
`
`0.23
`0.09
`0.13
`
`0.24
`0.09
`0.11
`
`0.11
`0.14
`0.10
`
`0.16
`0.15
`0.27
`
`0.08
`0.12
`0.09
`
`0.09
`0.06
`0.12
`
`0.05
`0.10
`0.10
`
`<2
`
`<2
`
`<2
`
`>1
`
`>1
`
`>1
`
`Mylan Ex 1016, Page 4
`
`

`
`R.B. Shah et al. / International Journal of Pharmaceutics 360 (2008) 77–82
`
`81
`
`Table 4
`Robustness with flow rate variation (nominal was 0.8 ml/min)
`
`Flow rate (ml/min)
`
`RT %R.S.D.
`
`Peak area %R.S.D.
`
`0.7
`0.9
`
`0.7
`0.9
`
`0.7
`0.9
`
`0.7
`0.9
`
`0.7
`0.9
`
`0.7
`0.9
`
`0.7
`0.9
`
`0.7
`0.9
`
`0.7
`0.9
`
`Tyr
`
`MIT
`
`DIT
`
`T0
`
`T2
`
`T3
`
`T4
`
`T3AA
`
`T4AA
`
`Spec
`
`0.03
`0.05
`
`0.04
`0.06
`
`0.02
`0.07
`
`0.01
`0.07
`
`0.02
`0.04
`
`0.01
`0.03
`
`0.01
`0.03
`
`0.01
`0.02
`
`0.01
`0.02
`
`<2
`
`0.83
`1.51
`
`0.19
`0.16
`
`0.07
`0.19
`
`0.04
`0.12
`
`0.05
`0.09
`
`0.03
`0.05
`
`0.11
`0.20
`
`0.06
`0.05
`
`0.10
`0.08
`
`<2
`
`in calculating the area under the peak. A desirable USP tailing factor
`is <2 which is consistent with the factor obtained with the current
`method (Table 3).
`Robustness was established by analyzing the system suitability
`standard (n = 6) at 20 and 30 ◦C (nominal = 25 ◦C), at flow rates of 90
`and 110% of the nominal flow (i.e., 1 ml/min.) and injector volumes
`at 50 and 150% of system suitability standard injection volume. Also
`since gradient method was used, a slight variation in gradient was
`also evaluated which included 93% (A) −7% (B) and 91% (A) −9%
`(B). Table 4 shows the results obtained for robustness of the sys-
`tem. A low CV (%) indicates that the system was robust and could
`be used without any problem if a minor change is to occur. The
`determination of robustness or ruggedness is especially important
`for gradient elution systems which might be impacted significantly
`by minor variations due to gradient ratio, temperature or other fac-
`tors. However the current method was found to be robust for such
`minor changes as demonstrated in Table 4 for flow rate variation.
`The specifications were met for all of the peaks under minor flow
`rate condition. Similar data was obtained for minor variations in
`injection volume, gradient, as well as column temperature (data
`not shown).
`
`3.3. Stress degradation studies
`
`Stress studies were carried out following an ICH guideline which
`establishes the requirements of stability indicating methods. A vari-
`ety of conditions, such as pH, light, oxidation, dry heat, etc. were
`applied and separation of drug from the degradation products was
`observed in the chromatograms. Similar studies are carried out in
`the literature for many drugs (Bakshi et al., 2004; Ojha et al., 2003).
`However all the literature methods fall short in meeting the current
`regulatory requirements for levothyroxine assay methods (Garnick
`et al., 1984; Graham et al., 1974; Rapaka et al., 1981). Therefore, the
`current work comprised of performing forced degradation studies
`to establish suitability of the method as stability indicating. Another
`ICH guideline on stability of testing of new drug substances and
`products (ICH Q1A (R2), 2003) advocates the use of stability testing
`assay methods for highly susceptible drugs such as levothyroxine.
`
`USP tailing
`0.94 ± 0.02
`0.95 ± 0.03
`1.06 ± 0.01
`1.00 ± 0.00
`1.10 ± 0.00
`1.15 ± 0.01
`1.21 ± 0.01
`1.14 ± 0.00
`1.27 ± 0.00
`1.23 ± 0.00
`1.31 ± 0.00
`1.28 ± 0.01
`1.40 ± 0.00
`1.35 ± 0.01
`1.29 ± 0.01
`1.24 ± 0.01
`1.27 ± 0.01
`1.23 ± 0.01
`
`Theoretical plates (X05)
`3.58 ± 0.02
`2.18 ± 0.06
`1.87 ± 0.00
`1.46 ± 0.01
`2.97 ± 0.01
`2.44 ± 0.02
`2.73 ± 0.02
`2.71 ± 0.03
`3.63 ± 0.05
`3.61 ± 0.04
`3.99 ± 0.05
`4.07 ± 0.05
`3.89 ± 0.04
`4.07 ± 0.06
`5.27 ± 0.03
`5.67 ± 0.07
`5.74 ± 0.05
`6.21 ± 0.06
`
`Resolution
`
`Selectivity
`
`1.03 ± 0.00
`1.04 ± 0.00
`1.02 ± 0.00
`1.22 ± 0.00
`1.18 ± 0.00
`1.03 ± 0.00
`1.23 ± 0.00
`1.27 ± 0.00
`1.12 ± 0.00
`1.13 ± 0.00
`1.08 ± 0.00
`1.08 ± 0.00
`1.29 ± 0.00
`1.30 ± 0.00
`1.09 ± 0.00
`1.10 ± 0.00
`
`1.81 ± 0.01
`1.76 ± 0.01
`1.31 ± 0.00
`10.69 ± 0.07
`9.80 ± 0.02
`1.70 ± 0.01
`14.55 ± 0.12
`16.56 ± 0.10
`8.66 ± 0.07
`9.79 ± 0.08
`5.88 ± 0.03
`6.48 ± 0.06
`21.61 ± 0.10
`23.06 ± 0.19
`8.31 ± 0.03
`8.96 ± 0.05
`
`<2
`
`>1
`
`>1
`
`>1
`
`In the current study, stress decomposition studies at temperatures
`in 40 ◦C increments above the accelerated temperature, extremes
`of pH and under oxidative and photolytic conditions were carried
`out on the drug substance, levothyroxine. The suitability of the
`proposed analytical method as a stability indicating method was
`supported by these stress degradation studies. Fig. 2 represents
`the stress degradation of the drug substance and the drug product
`in acidic, alkaline, oxidative, and photolytic (UVA and UVB) con-
`ditions, respectively. It was observed that the degraded products
`eluted far from the drug peak in case of oxidized sample where the
`degradation peaks were well separated from l-thyroxine peak. The
`main degradation compound formed by oxidation was observed at
`a retention time of 22.9 min which corresponds to T3AA. Thus it was
`even possible to identify the degradation compounds of l-thyroxine
`under the stress condition. There was no degradation observed with
`acidic, alkaline or with UV exposure under the conditions specified.
`Basic pH condition was found to enhance l-thyroxine pentahydrate
`stability in one of the studies (Patel et al., 2003).
`
`Fig. 2. Stres-degradation samples of levothyroxine with acid, alkali, oxidation, UV-
`A, and UV-b conditions (expanded view). The degradation compounds were seen
`under oxidation condition which were well separated from l-thyroxine peak. The
`primary degradation peak with retention time of 22.9 min corresponded to T3AA
`impurity. There was no degradation observed under acidic, alkaline, or photolysis
`conditions.
`
`Mylan Ex 1016, Page 5
`
`

`
`82
`
`R.B. Shah et al. / International Journal of Pharmaceutics 360 (2008) 77–82
`
`iodo-l-tyrosine (l-MIT), l-thyronine (l-T0), and l-tyrosine (l-Tyr).
`The method employed is under gradient condition in 40 min of
`total run. The method has been validated and it has been shown
`that it is reliable, linear, and precise both in upper and lower
`concentration range as well as robust with minor variations in
`chromatographic parameters. Therefore, it can be applied for quan-
`tification of the active compound and all of the eight degradation
`compounds. Excipient analysis study indicated that the method
`was found to separate l-thyroxine from the commonly used for-
`mulation excipient, butylated hydroxyanisole. The other excipients
`did not interfere with the analysis as they did not elute in the chro-
`matographic run.
`
`Acknowledgements
`
`Authors wish to express their sincere thanks to Everett Jefferson
`and Anthony Ciavarella from DPQR for their invaluable help.
`Disclaimer. The opinions expressed in this work are only of
`authors, and do not necessarily reflect the policy and statements
`of the FDA.
`
`References
`
`Bakshi, M., Singh, S., 2002. Development of validated stability-indicating assay
`methods—critical review. J. Pharm. Biomed. Anal. 28, 1011–1040.
`Bakshi, M., Ojha, T., Singh, S., 2004. Validated specific HPLC methods for deter-
`mination of prazosin, terazosin and doxazosin in the presence of degradation
`products formed under ICH-recommended stress conditions. J. Pharm. Biomed.
`Anal. 34, 19–26.
`FDA CDER (1994). Reviewer Guidance, Validation of chromatographic methods. FDA.
`FDA, 2006. Regulatory History and Current Issues. Dept. Health Human Ser.
`Garnick, R.L., Burt, G.F., Long, D.A., Bastian, J.W., Aldred, J.P., 1984. High-performance
`liquid chromatographic assay for sodium levothyroxine in tablet formulations:
`content uniformity applications. J. Pharm. Sci. 73, 75–77.
`Gika, H.G., Samanidou, V.F., Papadoyannis, I.N., 2005. Development of a validated
`HPLC method for the determination of iodotyrosines and iodothyronines in phar-
`maceuticals and biological samples using solid phase extraction. J. Chromatogr.
`B: Anal. Technol. Biomed. Life Sci. 814, 163–172.
`Graham, J.H., Banes, D., Biesemeyer, M.E., Nadkarni, A., 1974. Analysis of sodium
`levothyroxine or sodium liothyronine in tablets. J. Pharm. Sci. 63, 763–766.
`ICH Q2 (R1), 1995. Validation of analytical procedures: text and methodology. In:
`Proceedings of the International Conference on Harmonization.
`ICH Q1A (R2), 2003. Stability testing of new drug substances and products. In: Pro-
`ceedings of the International Conference on Harmonization.
`Ojha, T., Bakshi, M., Chakraborti, A.K., Singh, S., 2003. The ICH guidance in prac-
`tice: stress decomposition studies on three piperazinyl quinazoline adrenergic
`receptor-blocking agents and comparison of their degradation behaviour. J.
`Pharm. Biomed. Anal. 31, 775–783.
`Patel, H., Stalcup, A., Dansereau, R., Sakr, A., 2003. The effect of excipients on the
`stability of levothyroxine sodium pentahydrate tablets. Int. J. Pharm. 264, 35–43.
`Rapaka, R.S., Knight, P.W., Prasad, V.K., 1981. Reversed-phase high-performance liq-
`uid chromatographic analysis of liothyronine sodium and levothyroxine sodium
`in tablet formulations: preliminary studies on dissolution and content unifor-
`mity. J. Pharm. Sci. 70, 131–134.
`Richheimer, S.L., Amer, T.M., 1983. Stability-indicating assay, dissolution, and content
`uniformity of sodium levothyroxine in tablets. J. Pharm. Sci. 72, 1349–1351.
`Smith, D.J., Biesemeyer, M., Yaciw, C., 1981. The separation and determination of
`liothyronine and levothyroxine in tablets by reversed-phase high performance
`liquid chromatography. J. Chromatogr. Sci. 19, 72–78.
`Takahashi, M., Nagashima, M., Shigeoka, S., Kamimura, H., Kamata, K., 2002. Determi-
`nation of thyroid hormones in pharmaceutical preparations, after derivatization
`with 9-anthroylnitrile, by high-performance liquid chromatography with fluo-
`rescence detection. J. Chromatogr. A 958, 299–303.
`Thyroid Foundation of America, 2004. Thyroid Patient Advocacy Group Warns Thy-
`roid Patients of Potential Problems with Thyroid Drug Resulting From FDA
`Approval of Generic Versions. All Thyroid Org.: Thyroid Issues in the News.
`USP monograph, 2007 Levothyroxine Sodium. United States Pharmacopeia.
`Won, C.M., 1992. Kinetics of degradation of levothyroxine in aqueous solution and
`in solid state. Pharm. Res. 9, 131–137.
`
`Fig. 3. Expanded view of chromatograph of l-thyroxine degradation compounds
`(T3AA and T4AA) with formulation excipient, BHA. The retention times are T3AA
`(22.71 min), BHA (23.36 min) and T4AA (24.83 min).
`
`3.4. Excipient analysis
`
`The current method was used to assay various excipients which
`are commonly used in l-thyroxine formulations. The information
`was obtained from package insert or label of currently marketed
`lev