Evaluation of Effect of Excipients, Temperature and Humidity on the Stability of Levothyroxine Sodium

Levothyroxine Sodium is one of the most critical drugs for a significant segment of the population. The development of HPLC method of analysis for levothyroxine sodium revealed that there were significant stability problems associated with this drug. On August 141\ 1997, FDA announced that there was evidence which shows significant stability and potency problems associated with orally administered levothyroxine sodium products. This lack of stability and potency has a potential to cause serious health consequence to the public. The literature, on determining the stable formulation of levothyroxine sodium is very little. Much of the published work on levothyroxine sodium describes the innumerable bioavailability and potency problems with this drug. However, little in the literature explored the stability of levothyroxine sodium in presence of various excipients. The portion of work in this thesis represents a series of investigations we have performed to evaluate the stability of levothyroxine sodium in presence of some of the most commonly used excipients at different temperatures and humidities. These studies indicated that levothyroxine sodium is unstable in presence of carbohydrates such as dextrose, lactose and starch. These excipients have an aldehyde group, which may be reacting with the free amino group of levothyroxine sodium leading to a Schiff-base reaction along with oxidation reaction. In furthe~, series of investigations, the kinetics of degradation reactions were evaluated. First order and bi-phasic first order models were evaluated for our studies. Nonlinear regression was performed using Sigma Plot and the results suggest a biphasic-first order degradation pathway in most of the cases. The shelf life for levothyroxine sodium was

consequence to the public. The literature, on determining the stable formulation of levothyroxine sodium is very little. Much of the published work on levothyroxine sodium describes the innumerable bioavailability and potency problems with this drug. However, little in the literature explored the stability of levothyroxine sodium in presence of various excipients. The portion of work in this thesis represents a series of investigations we have performed to evaluate the stability of levothyroxine sodium in presence of some of the most commonly used excipients at different temperatures and humidities. These studies indicated that levothyroxine sodium is unstable in presence of carbohydrates such as dextrose, lactose and starch. These excipients have an aldehyde group, which may be reacting with the free amino group of levothyroxine sodium leading to a Schiff-base reaction along with oxidation reaction.
In furthe~, series of investigations, the kinetics of degradation reactions were evaluated.
First order and bi-phasic first order models were evaluated for our studies. Nonlinear regression was performed using Sigma Plot and the results suggest a biphasic-first order degradation pathway in most of the cases. The shelf life for levothyroxine sodium was determined using the k-values obtained from the best fit models. The lower t9o values indicated that levothyroxine sodium is highly unstable in presence of moisture and higher temperatures and so the tablets should be formulated and stored at or below room temperature with 0% humidity.    Among the different types of drug delivery systems, solid dosage forms are by far the most common due to their dose precision, low cost, easy and inexpensive packaging and shipping. In addition product identification for these dosage forms is simple when done using embossed or monogrammed punch faces; lend themselves to certain special-release profiles, such as enteric or delayed released products; having the best combined properties of chemical, mechanical and microbiological stability of all oral forms and are the lightest and most compact of all oral dosage forms. The advantages of solid dosage forms are reflected by widely accepted practice of delivering drugs in this manner during initial clinical trials [ 1,2]. Solid state stability of a drug is one of the most intrinsic and vital parts in the formulation of drug substance. One of the primary concerns during product development is producing a drug substance that has consistent stability properties in the bulk form as well as when formulated [3]. Stability testing provides evidence of how the quality of the drug substance or drug product varies over time under the influence of a variety of environmental factors such as temperature, humidity and light. A systematic approach should be adopted in the testing and evaluation of stability information which should cover, as necessary, physical, chemical, biological, and microbiological quality characteristics, including unique properties of the dosage form (for example, dissolution rate for oral solid dose forms). The design of the stability study is to establish, based on testing a minimum of three batches of the drug product, shelf life and label storage instructions applicable to all future batches of the dosage form manufactured and packed under similar circumstances. The degree of variability of individual batches establishes the confidence that a future production batch will remain within specification until the expiration date [ 4]. Stability testing of the drug substances and drug products is required to support the defined expiry period for the following categories of drug regulatory submissions: Investigational New Drug Applications (IND's), New Drug Applications (NDA's) for both the New Molecular Entities (NME's) and non-NME's, Abbreviated New Drug Applications (ANDA's), Supplements and annual reports, Biologics License Application (BLA's) and product license applications (PLA's) [5].

LIST OF FIGURES
The term 'pharmaceutical stability' encompasses a range of parameters. The most common interpretation is the chemical stability of the drug substance in a dosage form.
However, the performance of a drug when given as a tablet, capsule, syrup or injection is not only dependent upon the content of the drug substance, but also on its pharmaceutical properties (dissolution, disintegration, hardness, friability, content uniformity etc.) [ 1].
All of these parameters must, therefore, be a part of the stability program. In 1994 International Conference on Harmonization (ICH) guidelines for stability testing were published and are as follows [ 4].
Information on the stability of Active Pharmaceutical Ingredient (API) under defined storage conditions is an integral part of the systematic approach to stability evaluation.
Stability information from accelerated and long-term testing should be provided on at least three production batches. Long term stability should cover a minimum of 12 month's duration on at least three production batches at the time of submission. The testing should cover especially those features susceptible to change during storage and likely to influence quality, safety and/or efficacy. Stability information should cover as I necessary the physical, chemical, biological, and microbiological test characteristics. The length of studies and storage conditions should be sufficient to cover storage, shipment, and subsequent use. The 6-month accelerated testing should then be carried out at a temperature at least 15°C above long term storage temperature (25°C ± 2°C, 6D% ± 5% RH. The general guidelines for storage conditions and testing periods for bulk drugs and drug product is given in Table 1.

Drug Product
Long Term Testing 25°C ± 2°C, 60% ± 5% RH 12 months Accelerated Testing 40°C ± 2°C, 75% ± 5% RH 6 months Intermediate Testing 30°C ± 2°C, 60% ± 5% RH 6 months If 'significant change' occurs due to accelerated testing, additional testing at an intermediate condition e.g., 30°C ± 2°C, 60% ± 5% RH should be conducted. 'Significant change' at the accelerated condition is defined as [1]: a) A 5 percent potency loss from the initial assay value of the batch; b) Any specific degradant exceeding its specification limit; c) The product exceeding its pH limits; d) Dissolution exceeding the specification limits for 12 capsules or tablets; e) Failure to meet specifications for appearance and physical properties e.g., color, phase separation, resuspendibility, delivery per actuation, caking, hardness etc.

Modes of Drug Degradation
Drug degradation occurs by four main processes [6]: Trace metal ion cata ys1s ue to Fe , Fe , Cu , Co , etc.
The decomposition of drugs are most often classified as either hydrolysis or oxidation.
Most drugs contain more than one functional group, and hence may be subjected simultaneously to oxidation as well as hydrolysis. Other reactions such as epimerization, isomerization and photolysis may also affect the stability of drugs.
Hydrolysis may be caused by reaction of water with amides or esters. Water may also react with the ions of salts of weak acids and weak bases. Molecular hydrolysis reactions proceed much more slowly then the ionic hydrolysis (protolysis) [7]. In solid dosage formulations, free moisture is contributed by various additives or excipients, as well as the drug. In tablets a small percentage, typically 2% (w/w) of moisture, is required to facilitate good compression. This free water has the ability to act as a vector for chemical reactions between the drug and the excipients [6]. Some of the most common examples are: the hydrolysis of aspirin above pH 10 and hydrolysis of chloramphenicol which is pH independent is catalyzed by general acids and bases [7].
Oxidation reactions involve the removal of electrons or loss of hydrogen (dehydrogenation) from the molecules. When the reaction involves molecular oxygen, it is called auto-oxidation. Oxidation reactions frequently involve free radicals of atoms or molecules containing one or more unpaired electrons, or free hydroxy and molecular oxygen (0-0). Oxidation may be catalyzed by the presence of trace amounts of heavy metals and organic peroxides [7]. The oxidation of unsaturated fats and oils proceeds in the presence of atmospheric oxygen, light and trace amounts of catalysts according to free radical chain reactions. The hydroperoxide (R'-CHOOH-CH=CH-R") formed in the reaction may further decompose, and the reaction continues until the free radicals formed in the reactions are destroyed by inhibitors or by side reactions which will break the chain. Another classic example is the oxidation of ascorbic acid to dehydroascorbic acid in the presence of copper ions and oxygen. When the solution of ascorbic acid is freed from copper ions, the oxidation of ascorbic acid in alkaline medium will cease [7].
Oxidation and, to some extent, photolysis may be catalyzed by light. The energy of light is inversely related to wavelength (ultravisible > visible > infrared) and is independent of temperature. When the molecules are exposed to electromagnetic radiation (EMR), they absorb light at characteristic wavelengths, which causes an increase in the energy state of Photodegradation is dependent on both the intensity and the wavelength of light and is usually mediated by free radicals to produce dark colored substituents [6]. Some of the simple strategies for improving drug stability are given in Table 2.
51 Table 2. Simple strategies for improving drug stability (

Drug-Excipient Compatibility Testing
What emerges from a drug discovery program is an Active Pharmaceutical Ingredient (API) or drug substance. The API is the basis for producing the therapeutic activity expected of the drug and becomes a drug product after formulation with various excipients to produce a dosage form. Hastening the drug development process and optimization of dosage form stability are two major goals of any drug development program. The successful formulation of a stable and effective solid dosage form depends on the careful selection of the excipients used to facilitate administration, promote consistent release, enhance bioavailability of the drug and protect it from degradation.
Hence, excipients are the integral components of almost all pharmaceutical dosage forms.
In mixtures of solids, incompatibilities or chemical interactions can occur by following mechanisms [ 6]: degradation by nucleation via the gaseous phase contracting surface due to nucleation with coverage by the breakdown products degradation mediated by surface moisture or eutectic films oxidation photolysis.
Degradation in the solid state may be affected by several factors, such as the proportion of the drug to excipient(s), method of mixing, hygroscopicity of the powder mixture, hygroscopicity of the substance involved, temperature, humidity, particle size distrib.ution, particle packing, porosity of the powder bed, etc. Hence, in pharmaceutical preformu.lation drug-excipient compatibility studies are obligatory. Interestingly, with the 8 • importance of drug-excipient compatibility studies, no general method is available for these studies.
Ahlneck et al. [8] studied the three commonly used methods for dn:g-excipient screening: the suspension technique, storage of powder mixtures and compacts at specified temperatures and stored at specified relative humidities; and evaluated the variables influencing drug degradation. They concluded that the suspension technique is a fastscreening method for detecting chemical stability problems but gives limited information on the stability of a drug in solid dosage forms. The solid state techniques, i.e. the powder mixtures and compacts, gave a better picture of the stability profile of a solid dosage form composition. The solid state procedure took into account a large number of variables such as powder mixing, particle size, surface area, moisture adsorption, etc.
Monkhouse et al. [9] suggested that one should eliminate drug-excipient compatibility testing and instead select excipients on the basis of the physical and chemical characteristics of the drug substance and the literature data for the excipients. They recommended that the final composition should be selected on the basis of the accelerated stability testing of one or more target formulations at high temperature and humidity.
In 1999, Serajuddin et al. [10] reported a method that may be used successfully to identify the relative influence of different excipients on the stability of the drug. They proposed a model which involved storing drug-excipient blends with 20% added water in a closed glass vials at 50°C and analyzing at one and three weeks for chemical and physical stability. The amount of the drug substance in the blend was determined on the basis of the expected drug-to-excipient ratio in the final formulation. The effect of several 9 key factors such as the chemical nature of the excipient, drug-to-excipient ratio, moisture, micro environmental pH of the drug-excipient mixture, temperature, and light on the ..
dosage form stability could be identified by using this model. They suggested that selection of the dosage form composition by using this model at the outset of the drug development program would lead to a reduction of surprise problems during long term stability testing of drug products.

Quality and Functionality of The Excipients
Excipients are better known as promoters of degradation rather than as stabilizers of drug substances. Different functional groups or residues present in the excipients may have the propensity to interact with labile active ingredients or drugs, causing loss of molecular integrity or degradation. The quality of the final product depends not only on the active principles and production processes, but also on the performance of the excipients.
Excipients have undergone an evolution from an 'inert' and cheap vehicle to an essential constituent of the formulation that enhances the stability and the bioavailability of the drug substance in the drug product and improves its manufacturability on a production scale. The studies of interactions between a drug and the excipients shows that complexation, hydrogen bonding, ion-dipole, dipole-dipole and van der waals attractions can modify the physicochemical, pharmacological or pharmacokinetical behavior of the final product especially in the solid dosage forms [11,12,13].
If excipients have to act as stabilizers they must obviate or alternate the factors that cause molecular transformation of drug substances. These factors may be environmental components such as water vapor and sunlight. Other factors include stress during the • processing of the dosage form such as size reduction, compaction or sterilization, or interactions between adjacent molecules of the drug or functional groups on the same molecule [11].

Moisture Related Degradation
Water is one of the major factors responsible for causing degradation in pharmaceutical formulations. It may be associated with the drug or the excipients, may be incorporated during the processing of dosage forms or may be acquired from the environment during packaging or storage. Because of its ubiquitous nature and its ability to exist as a vapour, water is virtually impossible to avoid and difficult to control. The molecular mass of water is low, and so small amounts may be significant in terms of molecular reactivity.
Water is also capable of diffusing, to some extent, through packaging materials, pack seals or through compacted solid dosage forms [ 11].
Excipients with a greater affinity for moisture might aid in mitigating moisture sensitivity. Thus, formulation with a substance having a greater affinity for water as compared to the drug may help in sequestering moisture in the product. Perrier et al. [ 14] used nitrogen sorption isotherms to predict the effect of common excipients on the stability of nitrazepam. They determined that excipients with higher absorption energies caused less degradation, meaning, if the excipient has a higher binding energy for water as compared to the drug, the excipient may act as a desiccant and stabilize the drug.
Along with moisture, residues of lower alcohols (methanol, ethanol, isopropanol) might be present in the final formulation, as a result of synthesis, isolation of the drug or the process used for manufacturing the dosage form. Nimry et al. [15] and Tobyn et al. [16] showed that materials such as amorphous silica and microcrystalline cellulose may act as 'scavangers' of volatile residues and help in stabilizing the formulation.

Degradation by Oxidation
Loss of drug quality due to oxidation is usually secondary to hydrolytic breakdown.
These reactions are often complex and caused by factors that are difficult to separate and clarify. Oxidation can be catalyzed by exposure to air or light, the presence of trace residues (metal ions), by other components in the formulation or the combination of all the above mentioned factors.
Formulation additives have been effective in the stabilizaion of vitamin preparations. The antioxidants tocopherol, butylated hydroxy anisole, butylated hydroxy toluene and propyl gallate have all be used to stabilize vitamins A and D3 [17,18]. Reyes [19] showed that magnesium, calcium and aluminium stearates helped in the stabilization of ascorbic acid .

Photodegradation
Exposure to light may precipitate a plethora of degradation reactions, such as polymerization, isomerization, addition reactions in unsaturated systems, substitution reactions and photo-oxidation [20]. Thoma and Klimek introduced the concept of spectral overlay. This approach involves formulating with an excipient whose UV absorption spectrum overlaps (or substantially overlaps) that of the compound requiring stabilization. The excipient would thus compete with the active compound for photons from a radiation source and hence the impact of damaging radiation would be attenuated. They showed that the photolabile calcium antagonist nifedipine can be stabilized by the natural food colorant, curcumin, or by riboflavin [21]. Sanderson et al. [22] showed that the stability of a 13-lactam BRL42715B can also be enhanced by addition of a 'blocker' such as titanium dioxide and addition of soft paraffin with a UV spectrum that provided a partial spectral cover.
The spectral overlay approach is an elegant way to stabilize a drug. However the list of potentially useful materials which are free from pharmaceutical activity and are non-toxic is very limited.

Other Modes of Degradation
Some degradation reactions like isomerization, dimerization, polymerization and other forms of molecular rearrangements do not involve species other than the active ingredient. These of reactions are common in drugs of large molecular mass or those of biological origin. Hence it might seem that molecules with an intrinsic 'self-destructing' capability would be most difficult to stabilize.
Cyclodextrins are unique compounds, due to their umque, molecular complexation capability has been shown to improve the stability of compounds such as clofibrate and isosorbide which have tendency to . sublime [23]. Cyclodextrins have also been shown to stabilize labile materials such as PGE1 and PGF2 by forming molecular encapsulation [24].
The use of excipients to stabilize an unstable ingredient is an attractive concept. A product can be developed that will retain its quahty while the drug and other formulation ingredients are in close association. However, there are few examples, as listed in Table   3, in which excipients may destabilize or decrease the efficacy of the drug.

Methods for Studying Interactions between Drug and Excipients in Pharmaceutical Formulations
Presently, infra-red (IR)-spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, differential scanning calorimetry (DSC) and X-ray diffraction are the most commonly used methods to investigate interactions between drugs and the excipients in pharmaceutical formulations. IR spectroscopy is sensitive to crystalline form changes.
Since it is inherently based on molecular vibrations, IR has an advantage of being sensitive to functional group changes in low or non-crystalline materials. IR spectroscopy helps in looking at interactions between the drug and the excipient by following important characteristics such as the appearance of new IR absorption band(s), the broadening of band(s), and alteration in intensity. These comparisons are performed by comparing the IR spectra of the drug alone, the excipient alone, the complex and the simple physical mixture (prepared in same stoichiometry as a complex from both the drug and the excipient(s) [13]. For example, IR was used to determine which excipients participate in the salt to free base conversion of delavirdine. The spectra of delavirdine mesylate indicated that free base was formed in all· the mixtures except the one where croscarmellose sodium was absent. The spectra revealed that croscarmellose sodium in the tablet matrix was necessary to prevent conversion of the delavirdine salt to a free base [27]. An electronic database is available for comparison and identification of absorption and transmission spectra for excipients and the drugs which can be used to determine functional groups or degradation products present in the mixtures. This can be done by extrapolating the spectra obtained with that present in the database.

15
Solid state NMR can be used to determine inter and intra-molecular interactions and crystal packing in the solids at a molecular level; and are, therefore, useful for distinguishing between closely related solid forms of the same molecular entity. Rohrs et al. (28] used solid state NMR to characterize and quantify the forms of delavirdine in the tablet form. Three crystalline forms of delavirdine mesylate have been characterized, and the distinct NMR features observed among these and other forms were used for diagnostic purposes. The identification of the drug form(s) were used to verify that no major interferences occurred from tablet excipients. It was seen that there were no interferences from the tablet excipients with the drug forms in the tablet matrix.
DSC is another important tool which is used to study changes in the melting point of compounds and hence determine polymorphic changes in the mixtures. Jain et al. [29] used DSC to evaluate the amount of moisture pickup in a drug-moisture-lactose interaction. They found that stable solid lactose exists in a-monhydrous, a-anhydrous, and ~-anhydrous forms. DSC studies revealed that the a-anhydrous lactose which is present in small quatities in ~-lactose is responsible for moisture uptake and hence makes the formulation unstable.
Crystalline materials in powder form exhibit highly characteristic X-ray diffraction patterns in which the positions and relative intensities of peaks are well-defined and reproducible. The diffraction pattern of a crystalline powder is characteristic for the crystal lattices of that particular polymorph. By measuring the rate of disappearance (or appearance) of a peak unique to a reactant (or product), the kinetics of a reaction or transformation can be determined. This method has been used to follow the desolvation reaction of several crystalline hydrates. Since crystals serve as unique micro-reaction vessels, X-ray crystallography can often be used to determine the exact position of atoms as they re-orient themselves during a reaction. This can provide valuable information in predicting the path followed by atoms throughout the course of a reaction. Because of the wealth of information offered by X-ray analysis, it can be expected to play an important role in solid state stability studies [30].

INTRODUCTION
In 1891 Between 1987 and 1994, FDA received 58 adverse drug experience reports associated with the potency of orally administered levothyroxine sodium products. Out of the 58 reports, 47 suggested that the products were sub-potent, while 9 suggested super-potency.
Some of the problems reported were the result o( switching products from different manufacturers. However, other adverse events occurred when a patient received a prescription refill of a product from the same manufacturer on which they had previously been stablized, indicating a lack of consistency, potency and bioavailability between different lots of tablets from the same manufacturer. From 1991 to 1997, there had been no less than 10 firms initiating recalls of levothyroxine sodium tablets, involving 150 lots and more than 100 million tablets. In all but one case, the recalls were initiated because the tablets were found to be either sub-potent or potency could not be assured through the expiration date. The remaining recalls were initiated on the products that were found to be superpotent. It appears that the stability problem for levothyroxine sodium is quite complex and with some products, at least, stability is batch dependent [35].
After evaluation of the above mentioned problems, on August 14

viii) Stability:
Levothyroxine is sensitive to irradiation, hydrolysis, oxidation and heat.

ix)
Solubility: The study of drug-decomposition kinetics, the development of a stable dosage forms, and the establishment of expiration dates for drug products should evaluate the parameters that are likely to affect the quality, safety, and/or efficacy of the drug product. The analytical procedure should be fully validated and the assays should be stabilityindicating.
During pre-formulation stage, one of the objectives of the stability studies is to establish excipient-compatibility information to help the formulator design a stable and efficacious formulation. Excipients can affect the stability of drugs by: ( 1) acting as surface catalysts; (2) altering the pH of the moisture layer; and (3)  This study helps in developing analytical, physical, and chemical data that will facilitate formulation studies and manufacturing. These studies also allows the formulator to: (1) determine reactivities of the drug substances; (2) establish whether or not special handling and storage procedures are required to protect the drug substance; (3) ensure that the potency is sufficient and the level of significant degradation products may be significant throughout the life of the supplies; (4) develop supporting data for subsequent stability studies of the formulated drug; and ( 5) delineate any interaction between selected formulation excipients and the drug substance.
Once the initial screening of the excipients is done, the next step is to perform the stability studies on the formulated dosage form(s) under several storage conditions of temperature and humidity. In case of tablets, stability studies should also include evaluation of characteristics such as: appearance, friability, strength, hardness, color, disintegration and dissolution.

I) STABILITY INDICATING ASSAY FOR LEVOTHYROXINE SODIUM:
The selected HPLC method was selected to provide high sensitivity for levothyroxine sodium along with its reference standard liothyronine sodium at very low concentrations, as well as to be highly reliable and simple to use within the constraint of these studies.

STANDARD PROCEDURE:
If the HPLC system has not been used for a week or more, the system should be purged and primed as described in the instruction manual.

PREPARATION OF MOBILE PHASE:
Prepare a mixture of water and acetonitrile (70 : 30) containing 1 ml of a-phosphoric acid in each 1000 ml of the mixture. While making the mobile phase water and a-phosphoric acid should be mixed prior to the addition of acetonitrile. Mix the mobile phase for 5 minutes with magnetic stirrer. Filter the mobile phase using vacuum pump. Degas the mobile phase in ultra-sonifier for 5 minutes.

PREPARATION OF HPLC SYSTEM:
Connect the L-10 column to the system and make sure that the connections are tight. Turn on the system and set the following parameters: The column thermostat temperature should be set to 30°C.
The absorbance of the spectrophotometer should be set to 225 nm.
The flow rate should be set to 1 ml/minute.

RECOMMENDATIONS:
Confirming to the standard procedure described above, together with the following recommendations allowed the assay of levothyroxine sodium in a consistent and reproducible manner.
Prepare fresh standards everyday.
Wait for the HPLC system to stabilize.
The mobile phase should be properly degassed and the pump should be free of air bubbles. If there are erratic changes in the pump pressures, it might be due to entrapment of air in the pump heads. The pumps should be primed to remove the entrapped air. If the problem still persists, inlet and outlet check valves of the pump should be checked.
Always mount the magnetic stirrer in the mobile phase.
Any problem with the assay may be due to pumps, column, standard, sample, technique and should be isolated by checking each in turn.

A)VALIDATION OF ASSAY:
Investigations into the validity of the method have been undertaken. Using suitable standards, all containing liothyronine, the slope, measuring range and linearity of the calibration curve were determined for levothyroxine sodium. Additional studies were also completed to determine the validity of the typical analytical parameters used in the assay such as precision, accuracy, linearity, selectivity etc. In a further series of investigations, the storage of levothroxine sodium in .01 M methanolic NaOH at room temperature over a period of 2 days was also investigated. This was performed to make sure that levothyroxine sodium does not degrade during the analysis, which takes 35 hours.
PRECISION: The precision of an analytical method is the closeness of the test results when the procedure is applied repeatedly to multiple aliquots from a single homogenous sample. Thus, it is a measure of the degree of reproducibility of the analytical method under normal operating circumstances. Precision is determined by repeatedly assaying multiple samples removed from the homogenous sample.
ACCURACY: The accuracy of an analytical method is the closeness of test results obtained by the method to the true value. It is a measure of the exactness of the analytical 29 method. Accuracy is determined by usmg the calibration curve to determine the concentration of a sample (with known concentration). The accuracy of our assay method, expressed as percent recovery was found to be greater than 95% throughout the linear range used.

LINEARITY:
The linearity of an analytical method is its ability to elicit test results that . are directly, or by a well defined mathematical transformation, proportional to the concentration of analyte in samples within a given range. Linearity expressed in terms of the variance around the slope of the regression line.

ACCURACY SELECTIVITY I INTERFERENCES:
Placebo analysis was performed to check for interference or junk peaks due to excipients. No interfering peaks were observed with all nine excipients.

REPRODUCIBILITY:
Reproducibility IS limited by factors such as temperature, fluctuations and noise. The reproducibility was found to be independent of concentration and temperature, as the temperature of the column was controlled by a thermostat.

LIMIT OF DETECTION (LOD):
Determination of the signal-to-noise ratio IS performed by comparing measured signals from samples with known low concentrations of analyte with those of blank samples and establishing the minimum concentration at which the analyte can be readily detected. A signal-to-noise ratio between 3 or 2: 1 is generally considered acceptable for estimating the detection limit. In our assay method LOD was found to be 0.4µg/ml, with a signal-to-noise ratio of 3: 1.

LIMIT OF QUANTITATION (LOQ):
Determination of the signal-to-noise ratio is performed by comparing measured signals from samples with known low concentrations of analyte with those of blank samples and establishing the minimum concentration at 30 which the analyte can be readily quantified. A typical signal-to-noise ratio is 10: 1. In our assay method LOQ came out to be 0.6µg/ml, with a signal-to-noise ratio of 10: 1.
No significant differences were found in the concentrations of the samples stored for 2 days at room temperature. Placebo analysis were performed for all excipients to be used to make sure that the excipient peaks do not interfere with the levothyroxine peaks. No additional peak were seen which confirmed that the excipients did not interfere with levothyroxine peaks. Tables 5 and 6 shows the mean values of the data generated out of five days of validation studies for ten separate and distinct samples for series of levothyroxine sodium and liothyronine sodium concentrations ranging from 0.2 to 10 µg/ml. Figures 2 and 3 illustrate the complete as well as linear protion of the calibration curve for these samples. The correlation coefficient was found to be 0.9999 and 0.9997 for levothyroxine sodium and liothyronine sodium respectively.    Three samples each containing 5 µg/ml of levothyroxine sodium obtained from Sigma Chemical Company and Biochemie, Inc. were prepared using the standard method. The samples were analyzed by HPLC and the results compared using a two sided t-test with 95% confidence interval as shown in Table 7 and Table 8. It was seen that there was no · significant difference between the levothyroxine sodium obtained from Sigma Chemical Company and that obtained from Biochemie, Inc.

II) STABILITY STUDIES FOR LEVOTHYROXINE SODIUM
Evaluation of a stable solid dosage formulation often begins with drug-excipient compatibility studies. The present study was designed to investigate the stability of levothyroxine as a pure drug as well as in the presence of different excipients at different temperatures and humidities. From results of these investigations, excipients as well as optimal formulation and processing conditions were identified that would yield a more stable and reliable dosage form of levothyroxine. Of the many excipients available in the market, selection for this study was based on the physical and chemical properties. Nine excipients were selected for inclusion in the study. The studies were divided into four sections:

1)
Use of saturated salt solutions to achieve specified relative humidities.

4)
Stability studies for levothyroxine sodium tablets.

22.0
. .  Recovery Procedure: To a 4 ml HPLC vial containing 0.0009gm of a drug, 4ml of .01 M methanolic NaOH was added. The vials were shaken and kept in sonifier bath for 5 minutes, such that the drug dissolves in 4 ml of 0.01 M methanolic NaOH, to give an approximate concentration of 225 µg/ml. One ml of this solution was withdrawn and diluted in a 25 ml volumetric flask with 0.01 M methanolic NaOH, to get a concentration of 9 µg/ml. The samples were then analyzed by the HPLC method discussed earlier.

RH
Note: The concentration of 9 µg /ml was selected to ensure that the sample concentration falls within the validated range of our HPLC method. Also, if the drug degrades over a period of time it would still be within the detection limits of the HPLC method.

RESULTS AND DISCUSSION:
The data presented in Table 9 shows the amount of drug remaining over a period of 1 O weeks. Figure 7 shows a graphical presentation of the data. As can be seen from the plot, the drug appeared to be relatively stable at the different conditions evaluated except at 50°C. This is further corroborated by the fact that none of the kinetic models designed to 40 measure degradation fitted to the stability profiles of pure drug at the different conditions tested. It is believed that any trend in the degradation behavior of pure drug would have been accentuated and revealed by these models evaluated. Absence of any such trend within the limits of experimental variability coupled with the high values of percent drug left support this conclusion. Temperature induced degradation appeared to be significant at 50°c, as compared to the lower temperatures (see 25°C, 40°C and 50°C curves in Won, studied the kinetics of degradation of levothyroxine sodium in aqueous solution and in the solid state. The author concluded that levothyroxine sodium followed simple first-order degradation by the process of deiodination in aqueous solution. In contrast to solution degradation, levothyroxine sodium did not deiodinate in the solid state. Instead, the isolated degradation products   As stated earlier, the drug exhibited stability at the different conditions monitored in this study. Studies at rather adverse conditions might have provided more insight into the degradation pathways. However, these are outside the scope of this study considering the FDA stipulated guidelines. While the possibility of multiple complex degradation pathways may be indicated, what becomes important in context of predicting a model are those mechanisms that contribute significantly to that degradation. First order and biphasic first order models were evaluated to find the best fit for degradation kinetics of levothyroxine sodium under various conditions. Statistical analysis of the data was performed using a non-linear curve fitting procedure from Sigma Plot for Windows (SPSS Inc., 1997). The raw data generated following the experimental procedures described earlier was evaluated using nonlinear regression. The goodness of fit of the experimental data for first-order and biphasic first-order model was evaluated using residual plots, adjusted r 2 values, normality testing etc. The equations used for first order and biphasic first order are shown below:

First-Order Equation
C =Co* e-kt

Biphasic First-Order Equation
Attempts to fit a model to understand the nature of degradation in this regime were unrewarding considering the stable behavior of pure drug. Inherent variability in the data arising as a result of the complex sample preparation (refer to page 40) and storage conditions involved further complicated the prediction of model. While no particular trend in the degradation profile of pure drug is seen at various conditions tested, it can be stated that the drug's stability seems unaffected at these conditions for short period of testing. An understanding of the stability behavior of pure levothyroxine sodium at the FDA stipulated conditions will form a basis for later studies which are expected to involve more complex multiple interactions.

III) DRUG-EXCIPIENT COMPATIBILITY TESTING
Drug-Excipient compatibility studies were conducted to determine the formulation ingredients and conditions that might provide a stable formulation of levothyroxine sodium.
Experimental Design: For these studies, nine commonly used excipients were selected.
These excipients were dextrose, dicalcium phosphate dihydrate, calcium sulfate, mannitol, lactose anhydrous, lactose monohydrate, starch 1500, talc and ferric-oxide. The maximum strength of levothyroxine sodium tablets available in the market is 300 µg.
According to FDA c-GMP guidelines, the sampling from the mixture should not be more than 3 times the maximum dose concentration available. Thus, it was decided to work at three times 900 µg. The levothyroxine sodium-excipient ratio was kept at 1: 10.
Levothyroxine sodium and the excipients were mixed usmg a Crescent Wtg-L-Bug mixer. 0.0099gm of the mixture was carefully weighed out in separate 4 ml HPLC vials and recovery analysis was performed using the method described earlier. The vials were pla. ced in the dessicators under the different conditions previously described. At defined time intervals, three samples for each drug-excipient mixture were analyzed. The studies were conducted for a period of 20 weeks.

RESULTS AND DISCUSSION:
The data presented in Tables   Dextrose 120.00 80.00                        Considering that a Schiff-base reaction and an oxidation reaction may be the predominant pathways [43,46], first-order and biphasic first-order models were evaluated to find the best fit for degradation kinetics of levothyroxine sodium in the presence of different excipients under various conditions. Significant degradation involving any single pathway is best explained by simple first order kinetics as seen with dextrose, where Schiff-Base reaction is seemingly predominant. On the other hand, degradation following two different mechanisms (Schiff-base reaction plus oxidation) that are regulated by different rate kinetics seems to be best explained by bi-phasic first order models. The majority of the drug-excipient mixtures that were evaluated exhibited dual mechanism as reflected in the number of successfully fitted biphasic first order models. A non-linear regression analysis similar to the one used for the pure drug data was performed on the experimental data obtained from drug-excipient compatibility studies. The complete output of this statistical analysis along with the model parameters and constraints are shown in detail in Appendix A. Residual plots are generated from this statistical output using suitable graphing procedure (Microsoft® Excel 97) and are shown in Appendix B.
After selecting the model that fits the data following the above analysis, expiration dates (t90) for levothyroxine sodium in the presence of different excipients were calculated. For all conditions, the k-values for first order reactions and ai, a2, k1 and k2 for biphasic first order reactions were calculated from the models and used to calculate t90 values. The summary of the various results obtained is presented in Table 19.
It was noticed that levothyroxine degraded to a higher extent at 50°C in presence of all the excipients. These findings further support the fact that 50°C-60°C is the threshold free aldehyde group of the carbohydrates and the highly reactive ammo group of levothyroxine sodium. The general Schiff-base reaction is shown in Figure 17.  The availability of the free aldehyde group varies among the carbohydrates. For example, dextrose which is composed of only one ring structure, can easily break open to give a free aldehyde group, and so by this mechanisn the drug would degrade very quickly in presence of dextrose. Lactose, which is composed of a two-ring structure, would be more stable with the drug as compared to dextrose but less stable as compared to starch.
Levothyroxine sodium was found to be much more stable in presence of mannitol. This could be attributed to the absence of the ring structure and the free aldehyde group in mannitol.
The literature also indicates that levothyroxine has a tendency to undergo oxidation reaction [43]. Accordingly, in the presence of various components capable of oxidizing the pure drug, the degradation can be expected to be rather severe. It was seen that the mixture of levothyroxine with carbohydrates had turned brown, and this may be attributed to the oxidation of the drug by the mechanism shown in Figure 18.

58
The overall results showed that the drug was most stable at 25°C and 25°C I 60% RH.
One of the most interesting findings observed was in the case of talc and starch, where levothyroxine sodium was much more stable at 25°C I 60% RH and 40°C I 75% RH in . presence of these excipients ( Figure 14 and 15) than at the same temperatures with added humidity. This behavior may be due to a higher moisture uptake ability of starch and talc as compared to other excipients [ 4 7]. A higher affinity of these excipients towards moisture might have sequestered the free moisture, thereby reducing the availability of water for reaction with pure drug.
Overall for this group of excipie. nts the lowest t-90 values of levothyroxine sodium were observed in the presence of dextrose, dicalcium phosphate dihydrate and calcium sulfate at 25°C, 25°C I 60% RH and 40°C I 75% RH (Table 19) with stability well below usable levels.
Interestingly, the stability of the mixtures of levothyroxine sodium and lactose anhydrous were much lower as compared to that of levothyroxine sodium and lactose monohydrate as seen in Table 19 ( Figures 23 and 24). HPLC analysis demonstrated that at higher humidity/temperature, the mixture containing lactose anhydrous exhibited relatively greater degradation of the drug than that containing lactose monohydrate. This type of degradation was reported by Jain et al. [29], who studied the stability of a proprietary hydrophobic drug in the presence of hydrous and anhydrous lactose. The authors concluded that, lactose anhydrous becomes hydrated on exposure to high humidity/temperature condition and that the transition state of the lactose, not its stable state may be responsible for its greater interaction and subsequent degradation of the drug. They also concluded that in certain cases, lactose anhydrous may absorb a 59 significant amount of moisture, which can affect its inherent properties and may directly come in contact with the drug. For a moisture sensitive drug like levothyroxine sodium, this behavior may drastically affect its stability. Therefore, the general belief that lactose anhydrous, which has less than 0.5% moisture, should provide greater stability as compared to lactose hydrous needs to be properly evaluated.  EXPERIMENTAL PROCEDURE: Based on the results obtained from drug-excipient compatibility studies, the more stable excipients, mannitol(diluent), starch 1500(disintegrant and binder) and talc(lubricant) were selected for use in formulating the tablets. Due to the low concentration of the levothyroxine sodium in the tablets drug-to-excipient ratios of the excipients slightly different from those studied in the compatibility studies were used to obtain tablets of consistent weight and hardness. The total weight of the tablets was fixed at 115mg. Table   20 indicates the selected formulation.

Tabletting Procedure
Ingredients for the batch of 30,000 tablets, each weighing 115mg and containing 345 µg of levothyroxine sodium were weighed as shown in Table 20.
All the excipients as well as the drug were passed through a #40 sieve to get uniform particle size distribution. 282 gm of mannitol, 15 gm of starch 1500 and 3 gm of talc were weighed into a jar and mixed for 5 minutes using Turbula® mixer. Levothyroxine sodium was added to the remaining ingredients using geometric dilution while mixing.
The overall optimum mixing time was estimated to be 20 minutes. This was determined by addition of a red dye to the excipient mixture and observation of the homogeneity of the dye dispersed in the mixture at various time intervals. Further to confirm the homogeneity of the mixture, a 1 OOmg sample of the powder was taken from 6 different points in the container and compared with standard L-4 (6mcg/ml). Results of homogeneity testing are shown in Table 21. The amount of drug recovered expressed as mean (% CV) of six random samples was found to be 98.6% (7.00). Accordingly, the homogeneity of the drug mixture passes the USP specified limits (85% -115% ). An ANOV A was applied to the data and the p-value of .07 obtained indicated no significant difference between these samples. The mixture was considered acceptable and tablets were compressed using a Stokes® single punch tabletting machine. The following tests were performed to check the general properties of the tablets:

1)
Weight Variation: Twenty tablets were selected at random and weighed. The weight of these tablets are s_ hown in Table 22.
USP limits: For tablet weights below 130mg, percentage difference accepted is ±10% which allows weight limits for these tablets of 105mg to 125mg.
Average weight of 20 tablets= 116.05mg Allowable deviation is 10% Number of tablets in sample exceeding limits was 0 tablets Result: Therefore the batch passes the weight variation test  Table 23.
USP Limits: The concentration of the drug should lie in the range of 85% to 115% and the %CV should be less than or equal to 6.0%.

Results:
The tablets passed the USP content uniformity test.

3)
Tablet Hardness: Hardness ·of 10 tablets was determined using the Erweka® tablet hardness tester and the data is shown in  Percentage weight loss of 10 tablets = 0.69%.
USP Limits: A maximum weight loss of not more than 1 % of the weight of tablets being tested is considered acceptable for most products.
Results: The tablets passed USP friability test.

5)
Disintegration Test: The USP (Vanderkamp) disintegration tester was used for these studies. 1000 ml of distilled water was placed in a glass beaker and was kept in the water bath maintained at the temperature 3 7 ± 2°C. Six tablets were selected and the disintegration time noted. All tablets disintegrated within 5 minutes.
USP Limits: All six tablets should disintegrate within 30 minutes.

Results:
The tablets passed the USP disintegration test. 6) Dissolution Test: USP apparatus 2 was used. The apparatus was set to 100 rpm.
The temperature of water bath was set to 37 ±_i°C.  Table 25 shows the data obtained. Figure 28 shows the dissolution profile of levothyroxine sodium from its tablets over the 100 minutes of run time.
Results: USP limits state that not less than 55% (Q) of the labeled amount of Levothyroxine sodium is dissolved in 80 minutes. Hence the tablets passed the dissolution test.  where the potency could not be assured through the expiration date. The many problems and issues associated with levothyroxine sodium formulations have been discussed in the earlier section (Chapter 1 ).
In view of the above facts, studies were performed to evaluate the stability of pure drug and excipient compatibility. Based on these results an optimum formulation components were selected. Further tablet testing was performed on the selected formulation and the tablets produced were found to conform to the USP standards. This portion of work deals with the stability testing of these tablets.
Experimental Design: The previously formulated and manufactured levothyroxine sodium tablets were stored in stability chambers under same conditions as excipients. The stability of the tablets was evaluated over a period of 10 weeks. Tablets were sampled at predetermined time points and drug content was determined after extraction using the previously discussed HPLC method. At the end of 1 O weeks, dissolution studies were also conducted on these tablets, that had been stored at 25°C I 60%RH and at 40°C I 75%RH to determine if dissolution was affected by moisture.
72 RESULTS AND DISCUSSION: From the results obtained from the drug-excipient compatibility studies, it was decided to use mannitol (diluent), starch 1500 (binder and disintegrant) and talc (lubricant) as the excipients in the formulation of the levothyroxine sodium tablets. Immediately after manufacture, the tablets were characterized by good mechanical strength and a shorter disintegration time than the limits specified in the USP.
Further dissolution studies indicated that 80% of levothyroxine sodium was released within 30 minutes. The data for stability studies after 10 weeks is shown in table 26 and Figure 29.
From the data (Table 26) it can be seen that the drug degraded to less than 90% in less than a week, at all conditions studied. The stability of tablets was found to be much worse than the stability of levothyroxine sodium determined in presence of individual excipients (refer to Figure 29 versus Figures 11,14 and 15). It appears that the rate of degradation of levothyroxine sodium contained in the tablets was the synergistic effect of the individual excipients used. This is reflected in the shape of the degradation curves for the tablets when compared to those of individual excipients. A general trend showing an initial faster degradation rate followed by a relatively sfower phase is apparent for all the conditions studied.
The results of the dissolution studies of the tablets stored for 10 weeks are presented in Basing on these findings it could be stated that the physical stability of the formulation did not change at the end of 1 O weeks. Dissolution testing of tablets is considered as a quality-indicating tool in ensuring batch to batch uniformity [2]. For our purposes, it is therefore believed that any minor changes in the physical stability of the formulation would have been accentuated and reflected in the dissolution behavior of tablets.        (Table 19) with stability well below usable levels.
The results of the compatibility studies led to the selection of mannitol, starch 1500 and talc for use in formulating levothyroxine sodium tablets. The formulated tablets were found to conform to USP standards. The data on the stability of tablets revealed that the drug degraded to less than 90% in less than a week at all the conditions studied. These results indicate that in the presence of multiple excipients, the stability of the drug appeared to be much less. No particular trend was observed in the degradation profile as a function of temperature and humidity on the final formulation. It appears that the rapid degradation of the drug in tablets may be due to the synergistic effect of the individual excipients used or the processing variables such as mixing techniques, compaction pressures etc.
from the overall results obtained it can be concluded that levothyroxine sodium has very little stability at 50°C. It is also recommended that dextrose, lactose hydrous, lactose anhydrous, calcium sulfate and dicalcium phosphate dihydrate should not be used as excipients along with levothyroxine due to the very low stability in presence of these excipients.

VI) RECOMMENDATIONS FOR FUTURE WORK
The future work on the stability studies of levothyroxine sodium should focus on the following: • Increase the drug concentration and study the degradation pattern using IR, NMR or DSC for better understanding of the degradation pathways.
• Investigate the stability of the drug in presence of different formulation environments like adjusting micro-environmental pH or addition of anti-oxidants 85 or coating the drug particles with inert polymers and thereby reducing the contact of the drug with excipients.
• Stability of the drug in presence of different mini-formulations (multiple excipients) consisting of different compositions and ratios.
• Evaluate the effect of processing conditions and equipments on the stability of the drug. indicating the goodness of the fit followed by providing the validity of the assumptions.