Citation
Drug Delivery Utilizing Fat and Wax Matrices

Material Information

Title:
Drug Delivery Utilizing Fat and Wax Matrices
Creator:
Vaghefi, Farid
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Materials Science and Engineering
Committee Chair:
Batich, Christopher D
Committee Members:
Singh, Rajiv K
Brennan, Anthony B
Goldberg, Eugene P
Khan, Saeed R
Graduation Date:
5/5/2012

Subjects

Subjects / Keywords:
drug-delivery
mathematical-modeling
Genre:
Unknown ( sobekcm )

Notes

General Note:
In this body of work, the critical parameters that can be utilized to manipulate release profile from controlled release fat and wax matrices are evaluated. Active Pharmaceutical Ingredients (API) of various solubility were encapsulated in wax matrices by spray congealing or granulation techniques. Dissolution experiments identified the parameters with the most significant impact on the release profile. These included the raw material composition, API/drug particle size distribution, drug loading, cooling rate of the wax matrix (affecting polymorph structure), and the microsphere size distribution.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
5/31/2014

Downloads

This item has the following downloads:


Full Text

PAGE 1

1 DRUG DELIVERY UTILIZING FAT AND WAX MATRICES By FARID VAGHEFI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPH Y UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Farid Vaghefi

PAGE 3

3 T o the people I love

PAGE 4

4 ACKNOWLEDGMENTS I thank my p rofessor Dr. Christopher Batich for his support and guidance. Also, my thanks to Dr. Brennan for introducing me to percolation, which turned out to be a critical aspect in comprehending the release profiles in the studied systems. I also would like to acknowledge and thank my colleagues and management at Verion Inc. Special thanks to the management for providing me access to and permitting me to utilize my work at the firm for my dissertation.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FIGURES .......................................................................................................... 9 LIST OF ABBREVIATIONS ........................................................................................... 11 ABSTRACT ................................................................................................................... 12 CHAPTER 1 BACKGROUND .......................................................................................................... 16 2 THEORETICAL APPROACH ..................................................................................... 22 Material Properties .................................................................................................. 22 Saponification Value ......................................................................................... 23 Iodine Value ..................................................................................................... 23 Melting Characteristics ..................................................................................... 23 Polymorphism ......................................................................................................... 23 Micro Encapsulation Processes .............................................................................. 24 Emulsion encapsulation ................................................................................... 24 Advantage .................................................................................................. 25 Disadvantages ........................................................................................... 25 Spray Congealing ............................................................................................. 25 Advantages ................................................................................................ 26 Disadvantage ............................................................................................. 26 Granulation ....................................................................................................... 26 Advantage .................................................................................................. 27 Disadvantage ............................................................................................. 27 Conventional Methods for Controlling the Release Profile ............................... 27 Microsphere Particle Size ................................................................................. 27 Drug (API) Particle Size .................................................................................... 27 Drug Loading .................................................................................................... 27 Coating ............................................................................................................. 28 Physical Modification of Structure (Annealing) ................................................. 28 Discussion and Co nclusions ................................................................................... 28 3 EXPERIMENTAL DESIGN ......................................................................................... 30 Manufacturing Process Optimization ...................................................................... 30 Model Active Pharmaceutical Ingredients (APIs) .................................................... 32 Propranolol ....................................................................................................... 33

PAGE 6

6 Diltiazem ........................................................................................................... 34 Nifedipine ......................................................................................................... 34 Theophylline ..................................................................................................... 35 Materials and Methods ............................................................................................ 37 4 E XPERIMENTAL RESULTS ...................................................................................... 38 Manufacturing Process Optimization ...................................................................... 38 Characterization Results ......................................................................................... 42 Differential Scanning Calorimetry (DSC) .......................................................... 42 Optical Microscopy ........................................................................................... 46 X Ray Diffraction .............................................................................................. 48 Scanning Electron Microscopy (SEM) .............................................................. 49 Dissolution Analysis ......................................................................................... 51 Particle Size Analysis ....................................................................................... 52 Confocal Microscopy ........................................................................................ 53 In Vivo Rat Experiments ................................................................................... 56 Data Analysis ............................................................................................. 58 Results ....................................................................................................... 58 Discussion and Conclusions ................................................................................... 63 Manufacturing Process Optimization ................................................................ 63 Confocal Microscopy and Image Analysis ........................................................ 63 In vivo Rat Experiments .................................................................................... 64 5 M ATHEMATICAL MODELING ................................................................................... 65 Modeling Release Kinetics ...................................................................................... 65 Fickian Diffusion ............................................................................................... 65 Monolithic Systems, DiffusionControlled Release Profile, and Percolation ..... 67 Discussion and Conclusions ................................................................................... 75 6 C ONCLUSIONS ......................................................................................................... 77 Future Work ............................................................................................................ 78 APPENDIX A: STANDARD PROTOCOLS AND PROCEDURES ................................. 80 Encapsulation Process ........................................................................................... 80 Sample Spray Congealing Procedure .............................................................. 81 Sample Granulation Procedure ........................................................................ 82 Sieving .................................................................................................................... 84 Dissolution .............................................................................................................. 85 Propranolol H Cl HPLC Procedure .......................................................................... 87 DSC Procedure ....................................................................................................... 92 Confocal Microscopy ............................................................................................... 94 X R ay Diffraction ..................................................................................................... 95 In Vivo Rat Experimental Procedures (Vaghefi et al; 2005) .................................... 96

PAGE 7

7 LIST OF REFERENCES ............................................................................................... 97 BIOGRAPHICAL SKETCH .......................................................................................... 101

PAGE 8

8 LIST OF TABLES Table page 3 1 Model APIs and their characteristics .................................................................. 33 3 2 Partition coefficient and solubility of model APIs ................................................ 36 4 1 Rat In Vivo Experimental Design (Vaghefi et al; 2005) ....................................... 57 4 2 Results from IV route of administration (Vaghefi et al; 2005) ............................. 59 4 3 In vivo Rat experiment 1 plasma results of Nifedipine SR formulations (Vaghefi et al; 2005) ........................................................................................... 59 4 4 Results of the in vivo Rat Experiment 2 (n per group doubled), validated observed results in Experiment 1 (Vaghefi et al; 2005) ...................................... 60 A 1 Sieving Configuration ......................................................................................... 84

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Biopharmaceutics Classification System ............................................................ 17 1 2 Chemical structure of fats. A. Stearic Acid (a fatty acid). B Tristearin (triglyceride) ....................................................................................................... 18 1 3 Name and chemical structure of common saturated fatty acids ......................... 19 3 1 Design of Experiment for Manufacturing Process Optimization .......................... 30 3 2 Experimental Design Schematic for Manufact uring Process Optimization ......... 31 3 3 Chemical Structure of Propranolol ...................................................................... 34 3 4 Chemical Structure of Diltiazem ......................................................................... 34 3 5 Chemical Structure of Nifedipine ........................................................................ 35 3 6 Chemical Structure of Theophylline .................................................................... 36 4 1 Release profile from granulated vs. spray congealed Microspheres .................. 38 4 2 3D Surface Plot of effect of Cooling Method (rate) and API Loading on the % API released in 24 hours. ................................................................................... 40 4 3 3D Surface Plot of effect of API Loading and Mixing Technique on the % API released in 24 hours. .......................................................................................... 40 4 4 3D Surface Plot of effect of A PI Mixing Technique and Cooling Method on the % API released in 24 hours. ......................................................................... 41 4 5 Spray congealed product ................................................................................... 43 4 6 Spray congealed product ................................................................................... 44 4 7 The unannealed/granulated product batch ......................................................... 45 4 8 The annealed/granulated product batch ............................................................. 46 4 9 Optical light microscopy of Propranolol MSs; 50x, bar = 500 microns ................ 47 4 10 Polarized light microscopy of Propranolol HCl microspheres, 50x ...................... 48 4 11 Powder X Ray Diffraction of Spray Congealed Sterotex NF ............................... 49

PAGE 10

10 4 12 SEM micrographs of Propranolol HCl microspheres at 100x and 1000x magnification (lacks surface porosity) ................................................................. 50 4 13 SEM micrographs of cross section of Propranolol HCl microspheres at 100x and 1000x magnification ..................................................................................... 50 4 14 SEM micrographs of granulated Propranolol HCl microspheres at 50x and 500x magnification .............................................................................................. 51 4 15 Effect of MS size on drug release profile ............................................................ 53 4 16 Principles of Confocal Microscopy (Courtesy of Wikipedia) ................................ 54 4 17 Confocal Microscopy of Propranolol loaded microsphere ................................... 55 4 18 Confocal Microscopy of Nifedipine loaded microsphere (a. unmilled API; b. milled API) .......................................................................................................... 55 4 19 Nifedipine Mean Plasma Concentrations in Rat; x and o are Example 1 and 2, respectively (Vaghefi et al; 2005) ................................................................... 61 4 20 Nifedipine Plasma AUC in Rat ............................................................................ 62 5 1 Illustration of uniform drug distribution within the shell matrix (Monolithic System) .............................................................................................................. 68 5 2 Propranolol Release Profiles at 10, 20, and 30 % loading .................................. 70 5 3 Plot of time versus % released of Propranolol HCl compared with Higuchi model .................................................................................................................. 71 5 4 Plot of square root of time versus % released of Propranolol HCl compared with Higuchi model ............................................................................................. 72 5 5 Release profile of Propranolol HCl microsphere in acidic environment (pH 2) ... 72 5 6 Sample size distribution chart for jet milled Propr anolol HCl showing that 95% of the particles are below 10 microns with a volume averaged mean diameter of 1.96 microns. ................................................................................... 75 A 1 Encapsulation Process Spray Congealing ....................................................... 80 A 2 Kitchenaid Mixer ................................................................................................. 82 A 3 Dissolution Apparatus (Model 2100C) set as Apparatus II (paddles) ................. 85 A 4 Powder X Ray Diffraction data conditions .......................................................... 95

PAGE 11

11 LIST OF ABBREVIATION S A I nitial amount of drug in the matrix per unit volume C C oncentration (g/cm3) CS Solubility of the drug in the permeating fluid D D iffusion coefficient ( cm2/s ) P orosity F F lux T ransfer of mass per unit area and time across a defined plane (g/(cm2s)) M A mount of the drug released after time t Q A mount of the drug released after time t per unit of exposed area S0 S urface area of matrix exposed to the fluid T ortuosity Actual distance travelled divided by linear distance between point a and b X D istance (cm)

PAGE 12

12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillm ent of the Requirements for the Degree of Doctor of Philosophy DRUG DELIVERY UTILIZING FAT AND WAXES By FARID VAGHEFI M ay 2012 Chair: Christopher Batich Major: Material Science and Engineering Summary : In this work, Active Pharmaceutical Ingredients (APIs) of varying solubility were taken through formulation development and manufacturing process optimization to achieve the desired controlled release profile. Fat and wax matrices were used as the encapsulating shell material utilizing a moisturefree manufacturing process suited for APIs with poor solubility and moisturesensitivity Oral route of administration was the targeted route of delivery. The resulting API loaded microspheres exhibited controlled release profiles in vitro and in vivo. This approach enhanced the oral bioavailability of the poorly soluble model API (Nifedipine). Objectives: The main objectives of this effort w ere to evaluate the feasibility of utilizing fats and waxes as drug delivery matrices to improve oral bioavailabilit y and to identify the critical manufacturing parameters that enhance reproducibility of the API release profile. The specific aims of the effort were as follows: 1. Screen parameters (drug loading, drug particle size, microsphere size, cooling rate of molten fat/wax matrix, polymorph structure, etc.) and identify the critical parameters and techniques with the highest impact on the

PAGE 13

13 polymorphic crystalline structure of the encapsulating matrix (fats and waxes) and its API release profile. 2. Develop a process and manufacture an oral dosage form with modified release profile, utilizing fats and/or waxes. 3. Generate in vitro release profiles and analyze the mechanism of drug release based on mathematical modeling and compare to established theoretical models. 4. Determi ne the ability of fat and wax matrices to improve bioavailability (in vivo) of a model API. Materials and Methods: Hydrogenated cottonseed oil (Sterotex NF) was used as the main shell material. A full factorial experimental design was applied to screen f or drug loading (10, 20, 30% wt./wt.), mixing technique (impellor, homogenizer, microfluidization), and encapsulation technique (spray congealing vs. granulation) and their impact on the release profile utilizing response surface plots Release Profiles w ere generated using Apparatus I and II Distek 2100C dissolution apparatus. Nifedipine was chosen as the model API to be evaluated in vivo. Its poor aqueous solubility (19 g/mL) and low oral bioavailability (4556%) combined with its light and moisture sensitivity, made it the most challenging candidate and the choice API for this effort. In vivo analysis was conducted in rats utilizing both Instant Release (IR) and Intra Venous (IV) controls. After favorable results from the initial rat study, a second study was conducted increasing the number of rats per test group. Results: T he critical parameters affecting the release profile of Active Pharmaceutical Ingredients (API) from fat and wax matrices were evaluated. Porosity

PAGE 14

14 of the microsphere matrix was found to be one of the most important factors affecting release profile. The drug loading directly affected the ultimate porosity of the matrix. The higher the drug loading, the larger the resulting porosity after the dissolution of the drug into the sur rounding environment. Percolation theory [Stauffer et al; Sahimi] proved helpful in explaining the observed release profiles from these monolithic systems (insoluble matrix with random and uniform distribution of drug within). The resulting monolithic sy stems in our approach are diffusioncontrolled, but require a continuous network of voids (porosity) to facilitate complete release of the encapsulated drug. At 10 % loading, this continuous network was not formed. A phenomenon known as Percolation T hres hold signifies the loading at which the drug particles form a tortuous, but connected path for release. The literature [Zaller] report this to occur at about 15% drug loading (wt. of drug/total wt.) which is in line with results from this body of work R elease profile reproducibility diminished with drug loading >35%. Reducing the drug particle size to below 10 microns enhanced reproducibility. Granulation significantly reduced the release rate as compared to spray congealing This may be due to potent ial introduction of air bubbles within the matrix during atomization step of the spray congealing process. The faster cooling rate produced microspheres with the less stable polymorphs With time, the Alpha and/or Beta Prime polymorphs transitioned to the more stable Beta Polymorph. Differential Scanning Calorimetry (DSC) confirmed this transition when the lower melting temperature endotherm that had resulted with spray congealing, was eliminated over time. A nnealing was identified as a technique for accelerating the aging of the wax to the more stable polymorph (artificial accelerated aging) [Emas; Riiner]. Results demonstrated the elimination of the Alpha

PAGE 15

15 endotherm (Tm = 48.72C) by annealing. The Alpha endotherm was not observed in the DSC thermogr am of the granulated microparticulates. The rat plasma AUC (Area Under the Curve for plasma concentration vs. time) was improved from the Instant Release (IR) formulations 238 + 95 to as high as 2093 + 526 (ng*hr/mL) with lower variability (IR= + 39.9%; Ex 2= + 25%). This is about a ninefold increase in oral bioavailability. Conclusions : APIs of various solubility ( Biopharmaceutics Classification System : Class I and II) were encapsulated in fat and wax matrices by spray congealing and granulation tec hniques. Dissolution experiments identified the parameters with the most significant impact on the release profile to be the drug loading that in turn affected the porosity of the matrix, encapsulation technique, API/drug particle size distribution, cooli ng rate of the matrices (affecting polymorph structure), and the microsphere size distribution. Therefore, it is feasible to produce modified release matrices from fats and waxes by controlling the critical parameters. In vivo experiments in rats demons trated sustained release as well as about a ninefold increase in oral bioavailability of the model API (Nifedipine) over its IR formulation. In vivo rat study was repeated and confirmed the enhanced bioavailability observed in the initial study. Theref ore, this approach was able to enhance the oral bioavailability of Nifedipine in rats.

PAGE 16

16 CHAPTER 1 BACKGROUND Pharmaceutical industry is in constant pursuit of new APIs to address therapeutic needs. Many of these APIs possess properties that make it di fficult or impossible to develop them into products. The industry looks to formulation development to compensate for the molecules shortcomings. Preformulation studies focus on identifying the API properties/characteristics and its suitability for form ulation/product development The BioPharmaceutics Classification System categorizes APIs into four Classes: Class I H igh permeability, high solubility Class II H igh permeability, low solubility Class III L ow permeability, high solubility Class IV L ow permeability, low solubility The API models used in this work belong to Class I and II (see Figure 1 1 below) These model APIs are detailed under Experimental Design in Chapter 3.

PAGE 17

17 Propranolol NifedipineI IIDiltiazem TheophyllineIII IVHigh Low Solubility High Low Permeability Figure 11. Bio pharmaceutics Classification System Fats a nd waxes are naturally occurring molecules that belong to a subset of the lipid family This includes fatty acids, monog l ycerides, diglycerides, and triglycerides (mono, di, and tri esters of fatty acids, respectively). Figure 12 depicts the chemical st ructure of such molecules.

PAGE 18

18 O O A) Stearic Acid (a fatty acid) -----------------------------O O O O O O B) Tristearin (triglyceride) Figure 12. Chemical structure of fats. A ) Stearic Acid (a fatty acid) B ) Tristearin ( triglyceride) The hydrophobicity of the fatty acids increases with the length of the hydrocarbon ch ain length. Figure 13 outlines the name and chemical structure of some common saturated fatty acids.

PAGE 19

19 Common Name Chemical Structure # of Carbons Caprylic acid CH3(CH2)6COOH 8 Capric acid CH3(CH2)8COOH 10 Lauric acid CH3(CH2)10COOH 12 Myristic acid CH3(CH2)12COOH 14 Palmitic acid CH3(CH2)14COOH 16 Stearic acid CH3(CH2)16COOH 18 Common Saturated Fatty Acids Figure 13. Name and chemical structure of common saturated fatty acids Fats and waxes have been considered for controlled release applications of Active Pharmaceutical Ingredients for many decades [Giannola; Miyagawa; Prasad; Reza; Schwartz (1968a&b); Varshosaz]. They have been utilized as tablet lubricants and coating ingredients, lotion and cream ingredients for topical applications, and sustained release applications for oral delivery of APIs [Wade et al.]. The hydrophobic nature of fats and waxes could be advantageous for protecting APIs that may be susceptible to moisture (hydrolysis) and/or oxidation [Phuapradit et al.]. However, fats and waxes possess a polymorphic crystal structure [deMan; Ensikat; Heertje (1987); Herrera; Hui; Rivarola] that makes it difficult to control the behavior of such systems. The polymorphs of fats and waxes have been identified as Alpha Beta prime, and Beta which are mentioned here i n order from least to most stable, respectively [Rivarola et al.]. Alpha has been observed to be the stable form at high temperatures near the melting point of fats and waxes [Ensikat et al.]. The rate of cooling has a tremendous effect on the

PAGE 20

20 resulting morphology of the crystalline phase [Emas; Herrera; Litwinenko; Phuapradit; Rivarola; Heertje (1988)] The faster the cooling rate, the more prevalent the Alpha form. The reason may be that the Alpha form is the stable form at higher temperatures and due to the rapid cooling, the crystal structure is frozen in this polymorph without the opportunity to r earrange to the more favorable (stable) polymorph. However, with time, this Alpha form transforms to Beta Prime and eventually to the Beta form, which is believed to be the most stable form. The Alpha polymorph has a hexagonal, the Beta Prime, an orthorh ombic, and Beta a triclinic unit cell crystal structure [Ensikat et al.]. Beta polymorph is the most stable form and hence the desired polymorph to ensure product performance over its intended shelf life. The polymorphs exhibit differences in melting poi nts that are easily detected using Differential Scanning Calorimetry (DSC; Elem et al.). Emas et al., conducted studies on the possibility of annealing the product to provide the molecular mobility to allow the rearrangement of the crystals into the more stable polymorph. Their finding suggests that artificial aging of the crystal structure is feasible and can significantly reduce the amount of time it would normally take at room temperature for the crystal transformations to occur. However, even though the energy state of the annealed samples were close to the reference bulk material, they were not able to reproduce the exact polymorph of a bulk material, which had been stored for 5 years. This would be important since any changes in the polymorphic str ucture of the material could affect its properties (solubility, hydrophobicity, melting temperature, etc.) and the release profile of the API. Other factors such as characterization (e.g. Differential Scanning Calorimetry, X ray diffraction, etc.) and processing techniques (such as spray congealing, granulation, etc.)

PAGE 21

21 and their part in accomplishing the specific aims are cov ered in more detail in the Chapter 2 ( Theoretical Approach). The work outlined here, identifies the critical parameters with the most significant impact on the quality of the product.

PAGE 22

22 CHAPTER 2 THEORETICAL APPROACH In this effort three aspects of control are proposed that may enhance the feasibility of utilizing fats and waxes for pharmaceutical applications. The proposed controls include scrutinizing materials properties, evaluating processing techniques, and modifying processing parameters to improve reproducibility of results (drug release profile). It should be mentioned that even though this body of work is focused on fat and wax as encapsulating matrices, the covered processing and characterization techniques are applicable to other encapsulation applications. Material Properties It is very important to identify a qualified vendor that can provide the high quality material tha t is required to generate reproducible results. For this reason, it is good practice to utilize the vendors suggested by APhAs (American Pharmaceutical Association) Handbook of Pharmaceutical Excipients [Wade et al.]. Most fats and waxes are provided fr om the manufacturer with a certificate of analysis (C of A) that provides the specifications of the product that has been verified through their quality control. It is recommended to assess and determine the critical material specifications for achieving the desired release profile and using them as an acceptance criterion for raw material. This may result in a narrower set of specifications than even the vendors specifications. Properties such as the crystalline polymorph composition can be manipulat ed by processing (annealing).

PAGE 23

23 Saponification V alue Saponification value (Hui Vol. II: 55) is the weight of potassium hydroxide (in milligrams) needed to saponify one gram of fat. The higher the molecular weight, the lower the saponification value. Iodine V alue Iodine value (Hui Vol. II: 55) is a test that determines the degree of unsaturation of fats and oils. It represents the number of grams of Iodine absorbed by 100 grams of fat. Melting C haracteristics Melting Characteristics (Hui Vol II: 41) is probably the most important aspect of the fat/wax material as it pertains to its performance as a modified release formulation matrix. However, this information is not supplied by the vendor. This characterization will have to be conducted to qualify a material for use. Each vendor may process their material differently. A control sample will have to be generated that can act as the reference for future materials being purchased. Differential Scanning Calorimetry (DSC, described in more detail in sect ion 5.1) is a technique that can provide a fingerprint of the melting characteristics of the material. DSC can provide the melting temperature (Tm) as well as the glass transition temperature (Tg) of the material. Polymorphism Fat s and waxes can exist in many polymorph forms such as Alpha Beta and Beta Alpha is considered to be the stable polymorph at higher temperatures near the Tm of the fat/wax. Beta is considered to be the most stable form with the lowest energy state of all the polymorphs. The Beta crystals range in size from 5 25 microns and may be observed under the optical microscope [Hui Vol. II: P41]. However, visualizing the Beta Prime fat crystal, which are generally submicron, may

PAGE 24

24 require utilization of other techni ques such as Scanning Electron Microscopy (SEM) [ Eldem et al (1991); Heertje (1997)]. The crystal structure of the naturally aged bulk material may serve as a reference as to its most stable crystal form [Emas et al.]. However, the crystal structure of t he bulk material may not be important if processing takes the fat/wax to a molten state, which erases all traces of crystallization and provides the opportunity to engineer the desired crystal structure by controlling the processing parameters. One of the most important of these parameters in forming the desired polymorph is the cooling rate of the fat/wax from its molten state. This is covered in more detail in the processing sections (section 4.2). Micro E ncapsulation P rocesses There are many microenca psulation techniques that have been utilized in the industry. For this effort, emulsion encapsulati on, spray congealing (Appendix A ), and the granulation processes have been chosen as the encapsulation methods. The advantages and disadvantages of these processes (along with some others that were considered) are discussed below Emulsion encapsulation In this case, the emulsion is an oil in water emulsion (O/W) [Varshosaz]. This technique utilizes water as the outer continuous phase, which is initially maintained above the melting temperature of the microsphereforming matrix material (fat/oil/wax). The fat/oil/wax is heated to above its melting point to form a liquid. API is added to this melt and mixed to form a homogenous suspension/solution. This su spension/solution is then emulsified into the outer continuous phase and the system cooled down to below the solidification temperature of the fat/oil/wax. The advantages and disadvantages of this process are outlined below:

PAGE 25

25 Advantage : In the case where the API is not hydrophilic and has a higher affinity for the oil phase, this may be an acceptable approach. Disadvantages: Not Suitable for moisturesensitive APIs. If the API to be encapsulated is hydrophilic, it may result in significant loss of the dr ug to the outer continuous water phase. The drug at the surface of the oil droplets during emulsification stage is constantly migrating to the continuous phase (water). Therefore, it could be difficult to achieve high microsphere loading levels. The micr ospheres can retain water (residual), which can affect the long term stability of the drug and the matrix shell material. The residual moisture can also have a plasticizing effect on the shell material, lowering its melting temperature (Tm), which could be deleterious for the annealing stage. The emulsion may not be stable resulting in rapid phase separation, which may require utilization of surfactants. This also further enhances the loss of hydrophilic drugs to the outer continuous water phase. This procedure may result in a large size distribution of the microspheres. This technique would not be feasible if the melting temperature of the fat/wax is near the boiling point of water. Since the disadvantages outweighed the advantages, this process was aba ndoned. Spray C ongealing In spray congealing [Emas; Ghebre Sellassie; Phuapradit] as illustrated in Appendix A the microsphere matrix material is heated to above its melting temperature,

PAGE 26

26 at which point the API is added. This mixture is then properly agi tated to form a homogeneous suspension. This homogeneous suspension is in turn sprayed into a chill zone as fine droplets to form solid microspheres. Advantages: This is a high volume technique suitable for industrial production. The encapsulating fat/ wax shell material can be heated to above its melting temperature without the complication of boiling water as in the previously mentioned emulsion technique. Disadvantage: The cooling rate of this procedure is very rapid, resulting in solidification fro m the molten state in less than one second. This rapid solidification produces the less stable polymorph ( Alpha ), which given time will transform to the more stable crystal forms, namely Beta prime and Beta This can change the release profile significantly over the shelf life of the product. Granulation In granulation technique [Mathiowitz, Prasad] the matrix material is heated to above its melting temperature in a jacketed vessel with a mixing blade that rotates around the vessel in close proximity to its inner wall. The desired amount of API is added to the melt and agitated to form a homogeneous mixture. The vessel is then cooled at a predetermined cooling rate based on the recirculating baths program. This results in cooling at the surface of t he vessels inner wall, which in turn results in solidification (crystallization) of the homogeneous mixture. The rotating blades result in the fracture of the crystalline layers into smaller particulates. This process repeats itself until the entire ves sel content is solidified and broken into particulates. The size of the

PAGE 27

27 particles can be controlled by the level of crystallinity of the shell material as well as the amount of shear (intensity and time of mixing) introduced into the solidification proces s. Advantage : The cooling rate of the solidification can be precisely controlled. With a slow cooling program, more stable polymorph can be generated in the solidification process. Disadvantage: The resulting multi particulates are irregularly shaped (larger surface area than spheres) and have a wide size distribution. Conventional Methods for C ontrolling t he Release P rofile This section is not the focus of the research effort since the basic concepts of their effect are well understood. However, to provide some review of the factors contributing to the final API release profile from the formulated microspheres, they are briefly discussed here. Microsphere Particle Size For a certain mass or volume of material, the smaller the particle size of the micr ospheres, the larg er the surface area and hence the faster the release profile. Drug (API) Particle S ize The drug particle size can be utilized to manipulate the release profile. This can in turn determine the porosity of the matrix, and tortuosity of the result ing pathways for drug dissolution into the surrounding medium Drug Loading For microspheres where the drug is uniformly distributed within an insoluble matrix, the release of the drug depends on the solubility of the drug, the porosity of the m atrix, and the tortuosity of the pathway through which the solubilized drug would have

PAGE 28

28 to travel. The higher the drug loading, the higher the number of channels and pathways that are generated upon dissolution of the drug from the matrix. Coating Coatin g the microspheres can provide another control m echanism for manipulating the release profile. This could modify the release of the drug by either acting as a permeable membrane or a secondary barrier to diffusion An example of the latter is enteric coating of microspheres to prevent the drug release in the gastric region (low pH region) until it is exposed to the higher pH region of the intestine. It can also provide a permeability barrier to further control the release rate of the API. Physical Modification of Structure (Annealing) The most promising approach suggested by the literature for transforming the crystal structure into the more stable polymorph, is annealing (artificial accelerated aging) [Emas; Riiner]. This is done by heating the sample to above its glass transition temperature, but below its melting point. This allows the extra molecular mobility that is required for the transformation of the crystal polymorphs into the more stable forms (lowest energy state). Discussion and Conclusions Literature and experience have proven the need for tight control on the shell materials specifications. Polymorphism, if not controlled, can affect the quality and reproducibility of the product over time. The release characteristics of the product c an change over time if the manufacturing process results in the less stable polymorphs, which will transition to the more stable form over time.

PAGE 29

29 The emulsion microencapsulation was abandoned since it was not suitable for moisturesensitive APIs. Also, f or shell materials with melting temperatures close to the boiling temperature of water, this is not a feasible process.

PAGE 30

30 CHA PTER 3 EXPERIMENTAL DESIGN Manufacturing Process Optimization In order to identify the critical parameters affecting manufactur ing process, the polymorph str ucture of the MS shell material, and ultimately the release profile of the formulation, a full factorial design was applied Figure 31. Design of Experiment for Manufacturing Process Optimization As shown in Figure 31, t he following parameters were evaluated: a) Effect of mix ing technique The microfluidization, homogenization, and stirrer mixing were examined. b) Effect of solidification technique Spray congealing vs. granulation was examined. c) Effect of loading Loadings of 10% and 20 % were examined. d) Reproducibility The manufactured batches were duplicated to eval uate reproducibility of results Sterotex NF (CAS Number: 68334009) a hydrogenated vegetable oil of the cottonseed was used as the shell m aterial Sterotex NF conforms to United States Pharmacopeia/National Formulary (USP/NF) monograph for Hydrogenated Vegetable Oil, Type I. Non hydrogenated c ottonseed oils fatty acid profile generally consists of 70% unsaturated fatty acids including 18% (13% 44%) monounsaturated (oleic), and

PAGE 31

31 52% (33.1% 60.1%) polyunsaturated (linoleic & linolenic) [Daniel D. R.]. However, hydrogenated c ottonseed oil predominantly contains stearic acid. Stearic acid is the saturated fatty acid with an 18carbon chain and has the IUPAC name octadecanoic acid. It is a waxy solid with the chemical formula of CH3(CH2)16COOH. The larger the number of carbon atoms in the structural back bone, the less water soluble the fatty acid becomes. The following is a schematic of the experimental design: Pretreatment Solidification Analysis Figure 32 Experimental Design Schematic for Manufacturing Process Optimization Propranolol HCl was used as the model API for all batches prepared in the manufa cturing process optimization effort. In all cases micronized Propranolol HCl was used (mean diameter of about 2 microns). Sample procedures for spray Stirrer Mixing (10%&20 % Load) Homogenization (10%&20 % Load) Microfluidization (10%&20% Load) Granulation Granulation Granulation Spray Congealing Spray Congealing Spray Congealing (1) Dis solution Apparatus I (50 rpm, phosphate buffered media, pH 6.8, 37 C) (2) Differential Scanning Calorimetry (DSC)

PAGE 32

32 congealing and granulation are provided in Appendix A. The loadings of 10%, 20% and 30% refer to wt. of API divided by total wt. of API plus the shell material Model Active Pharmaceutical Ingredients (APIs) Model APIs chosen for sustained release formulation development plus some of the ir characteristics of interest are listed in Table 3 1 These API s have been extensively utilized for controlled release applications in the literature.

PAGE 33

33 Table 31 Model APIs and their characteristics Chemical Structure Half Life (Hours) BCS Bioavailability (%) Propranolol 4-5 1 26 Diltiazem 3-4.5 1 40 Nifedipine 2 2 45-56% Theophylline 5-8 2 100 H N O O N O O O O O O H N H N S O N O O O N N N N O O Propranolol Propranolol (brand names Inderal, Inderal LA) is a Beta blocker ( C16H21NO2; MW 259.34 g/mol) This is a water soluble API, which is used to treat hypertension and has been modeled for controlled drug delivery [Mathiowitz; Varshosaz].

PAGE 34

34 O O N Figure 33 Chemical Structure of Propranolol Diltiazem Thi s is a non dihydropyri dine (nonDHP) member of the class of drugs known as calcium channel blockers ( C22H26N2O4S, MW 414.52 g/mol) This is a water soluble API for hypertension (Ca channel blocker; Leeuwenka mp et al. (1994) ) N S O N O O O Figure 34 Chemical Structure of Diltiazem Nifedipine Nifedipine (brand names Adalat, Nifediac, Cordipin, Nifedical, and Procardia) is a dihydropyridine calcium channel blocker ( C17H18N2O6; MW 346.34 g/mol) It is poorly

PAGE 35

35 solu ble in water (19 micrograms per mL) as well as moisture and light sensitive, combined with a short half life (2 hours ; Frishman et al. (1989) ) All processing for this drug had to be done under red light. This was the most challenging API in our list of model APIs and the one most in need of sustained release (shortest half life, 2 hours), improvement in aqueous solubility, and enhancement in oral bioavailability. Therefore, this API was chosen as the model API for formulations developed for in vivo evaluation. N O O O O N O O Figure 3 5 Chemical Structure of Nifedipine Theophylline This is a methylxanthine drug used in therapy for respiratory diseases such as asthma ( C7H8N4O2; MW 180.164 g /mol) It is a poorly water soluble and light sensitive drug with a half life of 5 8 hours.

PAGE 36

36 N N N N O O Figure 36 Chemical Structure of Theophylline Table 3 2 below, summarizes the solubility of these compounds. As can be seen, most have better solubility in oil than in aqueous phase. Table 32 Partition coefficient and solubility of model APIs Solubility (mg/ml) Log P P Oil Solubility (mg/ml) Propranolol63.5 1.05 11.22 712.48 Diltiazem79.6 2.01 102.33 8145.41 Nifedipine0.019 3.57 3715.35 70.59 Theophylline5.5 -0.02 0.95 5.25

PAGE 37

37 Materials and Methods The following is a summary of the steps required to characterize the microspheres and generate the release profiles that will be presented in the results chapter. The procedures mentioned below are detailed in Appendix A. 1. Microencapsulate the API a. Spray Congealing b. Granulation 2. Sieve the microspheres generated is step 1a or 1b above. 3. Characterize the sieved microspheres as well as the bulk sample a. Microscopy i. SEM select formulations sent to contractor ii. Optical iii. Confocal select formulations b. DSC analysis c. Dis solution d. HPLC analysis of samples from dissolution e. Generate the cumulative release profile from diss olution/HPLC data

PAGE 38

38 CHAPTER 4 EXPERIMENTAL RESULTS Manufacturing Process Optimization Solidification technique (spray congealing vs. granulating) and API loading (resulting porosity) were found to have the most significant impact on the release pro file. Effect of solidification technique The granulation batches showed slower release than spray congealing. This difference was significant as observed in Figure 41 below: 0.00 20.00 40.00 60.00 80.00 100.00 120.00 0 5 10 15 20 25 30 % Released Time (Hrs) Granulated vs Spray Congealed (Loading at 20%) I Gr 20 I SC 20 H Gr 20 H SC 20 M Gr 20 M SC 20 Figure 41. Release profile from granulated vs. spray congealed Microspheres Six microsphere batches are represented in Figure 41. The loading for the six batches represented in this graph is fixed at 20%. The variation between the three spray congealed lots are not significant as is the case between the three granulated lots. It is therefore concluded that the encapsulation technique can significantly affect

PAGE 39

39 the release profile. This conclusion is further confirmed when the 3D response surface plots are examined (Figures 42, 4 3, and 4 4). Figures 4 2 and 4 4 demonst rate the significant impact of the encapsulation technique. These differences were evaluated using DSC thermal analysis that is detailed in the next section. In summary t he DSC results showed a lower melting point endotherm (at ~48.7 C) for all un anneal ed spray congealed product s that were not observed in the un annealed granulated samples. Annealing of the spray congealed product eliminated the lower endotherm. It is interesting to note that annealing the spray congealed microspheres, did not lower the release profile to the extent to resemble the slower release from the granulated microparticulates This may suggest to other reasons for the faster release observed with spray congealing. The SEM cross section of the spray congealed Propranolol HCl microsphere shows void due to entrapped air bubbles (probably due to air atomization). This added porosi ty is expected to be the reason for the faster release profile resulting from the spray congealing technique that cannot be remedied with annealing. API Loading (wt. API divided by total wt.; %wt. by wt.) Loading affected the release profile as expected. The higher the loading, the faster the drug was released (Figures 4 2 and 4 3 ). This was expected since the higher loading results in a more porous mat rix, enhancing drug dissolution and diffusion rate.

PAGE 40

40 0 2 50 100 30 1 20 10 % Released (24 hrs) Cooling Loading 3D Surface Plot of Results for Encapsulation DOECooling: 1-Granulation; 2-Spray Congealed Figure 42. 3D Surface Plot of effect of Cooling Method (rate) and API Loading on the % API released in 24 hours. Figure 43. 3D Surface Plot of effect of API Loading and Mixing Technique on t he % API released in 24 hours.

PAGE 41

41 0 25 50 75 2 100 3 2 1 1 % Released (24 hrs) Cooling Mixing 3D Surface Plot of Results for Encapsulation DOEMixing: 1-Impellor; 2-Homogenizer; 3-Microfluidizer Cooling: 1-Granulation; 2-Spray Congealed Figure 44. 3D Surface Plot of effect of API Mixing Technique and Cooling Method on the % API released in 24 hours. These surface plots can identify the experimental conditions that will lead to the desired effects. These plots can be generated for any measured response to guide in choosing the correct experimental conditions. Reproducibility Each one of the microencapsulation batch conditions generated by the factorial experimental design, was repeated to determine r eproducibility between the two batches F irst and second set of batches were examined and showed good correlation (0.950.99) The only variability in release profiles observed between repeat batches was between the two batches manufactured by microfluidization mixing and granulation technique at 20% loading. There was no significant difference in the cooling rates of the two batches to account for this difference. All other product preparations, regardless of pretreatment and/or final solidification, exhibited good reproducibility. For

PAGE 42

42 batches that were microfluidized and spray congealed, reproducibility was better than the homogenized or batches with stirrer mixing Characterization Results Differential Scanning Calorimetry (DSC) DSC can provide a th ermal signature for each material. This procedure requires 1 10 milligrams of material and measures the heat absorption (endotherm) or release (exotherm) by the material over a prespecified temperature range. This can provide information on glass transi tion and melting temperature as well as any phase transformations that may occur. Many waxes have been evaluated by DSC [Eldem; Herrera; Nassu; Rivarola]. An interpretation of these results is as follows: Spray congealing results in a much faster cooli ng of the matrix (shock/immediate) than granulation. The thermal analysis (DSCs) suggests that this difference results in a mix of all three polymorphic crystal forms for unannealed spray congealed samples.

PAGE 43

43 Figure 4 5 Spray congealed product (20% spray congealed, unannealed), before annealing, exhibiting the lower melting point endotherm. Thermogram above is the result of instantaneous matrix solidification with spray congealing The less stable endotherm (Tm = 48.72C) observed with spray congealing is nonexistent in the granulated samples. This thermogram is representative of those gen erated from Unannealed spray congealed samples.

PAGE 44

44 Figure 4 6 Spray congealed product (20% spray congealed, annealed), after annealing, the low melting point endotherm was eliminated. A slight melt temperature shift in the primary melt endotherm can be noted compared to that seen in Figure 4 5 (previous thermogram). This thermogram is represent ative of those generated from all the spray congealed and annealed samples.

PAGE 45

45 Figure 4 7 The unannealed/granulated product batch (20% granulated, unannealed) lacked the low melting point endotherm. This thermogram is representative of those generated from all the unannealed granulated samples.

PAGE 46

46 Figure 48 The annealed/granulated product batch (20% granulated, annealed) lacked the low melting point endoth erm. This thermogram is representative of others generated from the granulated samples that are annealed. Optical Microscopy This is probably the simplest approach to identifying the Beta polymorph (long range order; triclinic) since its characteristic long needles of 2550 microns distinguish it from the short needles of Beta Prime (1 micron) and the 5 micron platelets of the Alpha polymorph. Also, with the use of polarized light it can assist in distinguishing the crystalline reg ions from the amorphous ones (the crystalline region will be illuminated). Probably one of the most pertinent procedures for our application is the hot stage optical microscopy [Rivarola, et al.]. This allows the temperature of the slide and hence

PAGE 47

47 the sample to be controlled while it is being observed and recorded under an optical microscope. Rivarola et al. utilized the hot stage microscopy for determining the melting temperature of various Hydrogenated Vegetable Oils (HVOs). Figure 4 9 Optical lig ht microscopy of Propranolol MSs ; 50x, bar = 500 microns Other techniques such as polarized light microscopy can be utilized if one constituent is crystalline and the other is not. However, if both drug and matrix are crystalline, it would not be possi ble to distinguish them apart. Figure 4 10 shows the microsphere ill uminating under polarized light

PAGE 48

48 Figure 4 10. Polarized light microscopy of Propranolol HCl microspheres 50x However, polarized light microscopy could not be utilized to determine A PI distribution within the microspheres since both the drug and the shell material were of crystalline form and could not be distinguished from one another. The technique also lacks the resolution to observe 1 micron particles and the ability to establi sh a focal point within the microsphere (feasible with Confocal Microscopy). X Ray Diffraction molecular and crystal structure of solids. An X Ray beam is directed at the sample. The sample can be in a powder or single crystal form. The powder form is a qualitative anal ytical technique and may not provide the precise information that the analysis of a single crystal can provide [Bernstein, page 112]. The identifying d spacing values for the Alpha (hexagonal symmetry), Beta Prime (orthorhombic), and Beta (triclinic) poly morphs have been reported by Ensikat et. al. as 4.13, 3.72, and 4.58 Angstroms, respectively. This technique can also determine the degree of crystallinity for the sample (crystalline vs. amorphous).

PAGE 49

49 The main driver of the Alpha to Beta polymorph trans ition was determined to be the cooling rate. This is in agreement with the literature on polymorphism of these materials. It is important to note that having the most stable matrix results in the most stable end product. X Ray diffraction of spray congealed hydrogenated vegetable oil (Sterotex NF, cottonseed) showed 52% crystallinity with the Alpha polymorph constituting >60% of the composition. Upon annealing the crystallinity increased to 62%. Data conditions for this powder X Ray Diffraction are lis ted in Appendix A. Figure 411. Powder X Ray Diffraction of Spray Congealed Sterotex NF (hydrogenated cottonseed oil) Scanning Electron Microscopy (SEM) This technique provides high resolution topography images of its subjects. It bombards the sample with an electron beam of varying intensity that excites the sample into emitting electrons that are captured by the detector. If the sample is not conductive, localized over charging (heating) can be destructive to the sample. For this reason, a thin lay er of conductive material (Gold/Palladium) can be plasmacoated onto nonconductive samples. Heertje et al. (1997) have provided a procedure for visualizing the crystal structure of fats.

PAGE 50

50 Analysis of SEM micrographs of spray congealed microspheres ( Figu re 4 12) prior to dissolution showed no signs of visible porosity Figure 4 12. SEM micrographs of Propranolol HCl microspheres at 100x and 1000x magnification (lacks surface porosity) Cross section of the spray congealed microspheres showed porosi ty (Figure 4 13) that is suspected to be the result of air bubbles introduced into the microspheres during the air atomization process. Figure 4 13. SEM micrographs of cross section of Propranolol HCl microspheres at 100x and 1000x magnification

PAGE 51

51 Figure 414 shows the SEM micrographs of granulated microparticulates. These microparticulates are formed by shearing the solidifying drug suspension (in wax) as it cools slowly over the course of hours. See Appendix A for the procedure. Figure 41 4. SEM micrographs of granulated Pr opranolol HCl microspheres at 50x and 5 00x magnification SEM is a characterization technique for observing topography and cannot distinguish between encapsulating matrix and drug particles. Also, it cannot provide the d istribution of API within the matrix. This was accomplished by utilizing Confocal microscopy as outline later in this chapter. Dissolution Analysis A dissolu tion apparatus is an USP (United States Pharmacopeia) approved equipment for analyzing the release of API from a matrix/capsule/tablet into a desired medium with controlled temperature and agitation. The agitation is provided by rotating basket (apparatus I) or paddles (apparatus II). For marketed products the approved procedures are listed in the US P. All release profiles presented in this body of work, were generated using the Distek dissolution apparatus. The procedure is outlined in Appendix A. Dissolution analysis was conducted for Propranolol HCl microspheres at

PAGE 52

52 both neutral pH (~7) and gastr ic pH (~2) showing no significant difference in release profile behavior (square root of time release type, see Figures 54 and 5 5). Particle Size Analysis Size distribution of the resulting microspheres can be determined by examining with optical or Scanning Electron Microscopy. The microspheres were sieved (see Appendix A for the procedure) to enable release profile evaluation for the desired size range. The microspheres from the 43% loading of Diltiazem were sieved into different size fractions ( between 20, 40, and 60 mesh sieves ; see Appendix A for corresponding size range in micrometers ) and tested for dissolution according to USP procedure Test 5 (Apparatus I, 900 mL 0.05 M phosphate buffer, pH 7.2, 37C, 50 rpm ; see Appendix A for detailed dissolut ion procedure). The drug release rate based on the size of the microspheres followed the expected trends (faster release with smaller microspheres due to larger cumulative surface area per unit mass, see Figure 4 15 below)

PAGE 53

53 Effect of microspheres particle size (mesh) on release profile of Diltiazem HCl, Drug loading: 43%, Shell material: Dritex C, Phosphate buffer7.2, Basket 50 rpm 0 10 20 30 40 50 60 70 80 90 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Hours Percent pf release -20 +40 (1) -20 +40 (2) -40 +60 (1) -40 +60 (2) -60 +80 (1) -60 +80 (2) Figure 4 15. Effect of MS size on drug release profile Confocal Microscopy This technique provides the unique ability of looking within an object with better resolution than an optical microscope. Confocal microscopy is a technique that has generally been utilized for biologi cal applications (fixed or living cells and tissue). Confocal microscopy can focus on a shallow depth of field and eliminate out of focus glare. These advantages over optical microscopy enable visualizing planes within the sample (in our case microspheres). Nikon provides a web page that outlines the basic concepts of Laser Scanning Confocal Microscopy (LSCM) for further review [see reference under Nikon Corp.]. This technique allowed determination of the drug distribution within the microspheres. Figur e 4 16 illustrates the c oncept of Confocal Microscopy.

PAGE 54

54 Figure 4 16. Principles of Confocal Microscopy (Courtesy of Wikipedia) A Confocal Microscopy (fluorescent emission) procedure was developed (Appendix A), which allowed observation of the drug distribution within the microspheres. For biological samples, fluorescent probes need to be attached to conduct Confocal microscopy. However, for our application, Nifedipine and Propranolol HCl, autofluoresce and did not require further modification. Figure 4 17 shows Confocal microscopy of a 35% loaded Propranolol HCl microsphere. The image shows a uniform distribution of Propranolol drug particles within the microsphere

PAGE 55

55 Figure 4 17. Confocal Microscopy of Propranolol loaded microsphere a b Figure 4 18. Confocal Microscopy of Nifedipine loaded microsphere (a. unmilled API; b. milled API) Figure 4 18 shows Confocal microscopy results from microspheres that were prepared with unmilled (a) and milled (b) API. T he resulting images suggest that this technique is feasible for determining the API particle distribution wit hin the

PAGE 56

56 microspheres. Image analysis of these Confocal micrographs with MatLab software package enabled assessment of the particle size distribution of the encapsulated API to assess potential aggregation of the particles during the encapsulation process. The image analysis also enabled determining the porosity of the matrix that would result upon dissolution of the API from the matrix. As previously discussed, porosity is one of the most important factors affecting the release profile of the API from these monolithic systems. This technique has shown promise for predicting porosity of the microspheres and deserves further investigation. In Vivo Rat Experiments Several Nifedipine formulations were prepared for in vivo rat experiments (conducted at cont ract research organization under appropriate guidelines) to assess potential enhancement in oral bioavailability. The first in vivo rat experiment suggested that some of the formulations are capable of enhancing the oral bioavailability of Nifedipine to a lmost 100%. The in vivo rat experiment was repeated, doubling n per group (from 3 to 6), to confirm this observation. Details of this effort are outlined below. The studies evaluated the pharmacokinetic (PK) profiles of sustained release (SR) oral formul ations of Nifedipine in male rats in comparison to an immediate release (IR) oral solution and an intravenous dose. M ale SpragueDawley rats were divided into 7 groups of 3 rats/group. Each group was dosed with one of 5 sustained release (SR) formulations the immediate release (IR) formulation or the intravenous (IV) formulation a s described in the following Table:

PAGE 57

57 Table 41. Rat In Vivo Experimental Design (Vaghefi et al; 2005) The Nifedipine load for each m icrosphere formulation was deter mined by HPLC and the appropriate weight of microspheres to deliver a dose of 3.6 mg/kg to a 250 g rat was placed in an opaque gelatin capsule. For dosing, the contents of the capsule were emptied into the hub of a gavage needle. A syringe filled with 3 mL of wat er was attached to the gavage needle and the microspheres were flushed into the stomach. The IR (0.5 mg Nifedipine / mL in a 30% sodium benzoate solution) and intravenous (0.05 mg Nifedipine / mL in a 30% sodium benzoate solution) dose solutions are prepared using the same lot of Nifedipine used to make the sustained release formulations. Nifedipine concentrations in the dose solutions were verified by UV absorbance. The IR dose solution (1.8 mL ) was administered via oral gavage. For intravenous administration, each rat was restrained and 0.5 mL of the dose solution was slowly infused over 4 minutes by hand via the lateral tail vein.

PAGE 58

58 N ecessary precautions were taken during the preparation, storage and dosing of the formulations to protect the microspheres from n atural and artificial light (Nifedipine is light sensitive). Data Analysis For each pharmacokinetic profile, the highest observable concentration is assumed to be the maximum concentration ( Cmax). The time that Cmax is reached is denoted Tmax. Area under t he plasma concentrationtime curve (AUC) is calculated from zero to the last quantifiable plasma concentration, AUC(tf). Results Individual concentrations of Nifedipine in plasma for the 7 formulations evaluated are reported in Tables 5 3 All analytical data met the acceptance criteria for the assay. Plasma concentrations that were below the quantifiable limit of 5 ng/ mL were reported as zero. Following intravenous administration of Nifedipine, maximum plasma concentrations of Nifedipine are observed at the end of the 4minute infusion. Plasma Nifedipine concentrations then declined rapidly and were not detectable 1.5 hours post infusion. The terminal half life was estimated at ~ 15 minutes which is in close agreement with that reported in the literature f ollowing a 6 mg/kg intravenous dose to rats. The data are summarized in the following Table 5 2.

PAGE 59

59 Table 42. Results from IV route of administration (Vaghefi et al; 2005) Following oral administration of Nifedipine in solution (IR formulation), Nifedipine is rapidly absorbed and Cmax is observed at the first sampling time post dose (15 minutes). Plasma Nifedipine concentrations steadily declined out to 2 hours and then showed some fluctuation over the next 8 hours (Table 5 3). Once Tmax occurred, plasma Nifedipine concentrations are not maintained beyond this time. Table 43. In vivo Rat experiment 1 plasma results of Nifedipine SR formulations (Vaghefi et al ; 2005) Formulation Cmax (ng/mL) Tmax (hr) tfa (hr) AUC(tf)(ng*hr/mL) Durationb (hr) Control Example A81-120 10 10 508-680 6 Control Example B188-469 10 10 1068-1767 2-4 Control Example C260-406 4 10 1352-1745 2-4Example 1374-864 1-4 10 1980-2736 1-4Example 2471-611 2-6 10 2390-2917c3-5 IR (Instant Release)90-149 0.25 10 324-378 0.25 a. Time of last quantifiable plasma concentration b. Time plasma concentrations are greater than half Cmax c. Normalized to a targeted dose of 3.6 mg/kg Compared to the IR formulation, there is a slower rate of absorption of Nifedipine from all of the mi crosphere formulations with Cmax generally being achieved 24 hours post dose. In addition, the duration that plasma N ifedipine con centration exceeds

PAGE 60

60 greater than half Cmax generally ranged from 2 to 4 hours. There are, however, some marked differences in the extent of absorption of the Nifedipine load between the different microsphere formulations t hat resulted in some formulations having greater Cmax and associated AUC values than others. Although the intravenous dose is 36times less than the oral doses, it would appear, assuming linear pharmacokinetics, t hat the bioavailability of Nife dipine (based upon AUC ratios) from the two microsphere formulations would be about 100%. The in vivo rat experiment was repeated with higher number of rats per group to validate initial results The second experiment confirmed results of the first experiment as evi dent in Table 54 below: Table 44. Results of the in vivo Rat Experiment 2 (n per group doubled), validated observed results in Experiment 1 (Vaghefi et al; 2005) Parameter Mean SD (Range) [N=5 or 6] Formulation Cmax (ng/mL) Tmax (hr) tf a (hr) AUC(tf) (ng hr/mL) C ontrol Example A 25.0 6.3 (17.4 34.2) 3.6 0.9 (2 4) N/A (8 12) 129 46 (84.1 195) Control Example B 170 75 (34.0 248) 3.2 1.3 (1 4) N/A (6 12) 612 285 (141 853) Control Example C 119 21 (99.8 148) 2.7 1.0 (2 4) N/A (8 12) 536 79 (432 628) Example 1 674 281 (402 1030) 1.4 0.5 (1 2) N/A (10 12) 1847 542 (1356 2676) Example 2 449 125 (216 588) 2. 7 2.0 (1 6) N/A (8 12) 2093 526 (1288 2693) Control Example D 37.0 8.1 (26.6 49.6) 4.0 1.3 (2 6) N/A (6 12) 188 75 (132 320) Control Example E 38.3 7.3 (28.3 46.3) 4.8 2.7 (2 8) N/A (8 12) 209 29 (167 248) Control Example F 56.1 7.8 (44.1 66.9) 2.8 2.6 (1 8) N/A (8 12) 280 33 (235 323) Control Example G 83.8 41.9 (31.0 124) 3.2 1.8 (2 6) N/A (6 12) 314 145 (99.2 468) IR 77.9 19.0 (55.7 97.6) 0.25 N/A (8 10) b 238 95 (103 374) N/A: Not applicable a: Time of last quantifiable plasma concentration b: Last collection time point for IR administration was 10 hours post dose

PAGE 61

61 As can be seen from Table 5 4 t he IR Tmax of 0.25 hours was increased to as high as 8 hours ( most 4 6 hours). It should be mentioned that Nifedipine half life in rat is much shorter than human (15 min ute s vs. 2 hours respectively). Therefore, the Tmax observed in this experiment should be longer in human. The Nifedipine Cmax for the IR fo rmulation of 77.9 + 19 ng/mL was increased to as high as 674 + 281 ng/mL ( over 8 fold) for one of the formulations These are promising results and as can be seen in Figure 514 below, with mean plasma levels much higher than the Instant Release or even t he IV dose could achieve. Figure 4 19. Nifedipine Mean Plasma Concentrations in Rat ; x and o are Example 1 and 2, respectively (Vaghefi et al ; 2005)

PAGE 62

62 The rat plasma AUC was improved from the Instant Release (IR) formulations 238 + 95 to as high as 209 3 + 526 (ng*hr/mL) with lower variability (IR= + 39.9%; Ex 2= + 25% ; See Figure 515) Figure 4 20. Nifedipine Plasma AUC in Rat It is therefore concluded that it is feasible to enhance bioavailability of the poorly soluble API, Nifedipine in a Sus tained Release (SR) formulation by utilizing this approach. The results suggest that this approach may benefit other therapeutics and should be further investigated with other APIs in need of enhanced bioavailability.

PAGE 63

63 Discussion and Conclusions It was fe asible to encapsulate Class I and II APIs with this technique and showed sustained release. It was possible to achieve desired release profile by manipulating the manufacturing conditions. Percolation threshold was reached at between 10 and 20 % loading (wt./wt.), which is in line with observations in literature (Zaller) for monolithic systems. Manufacturing Process Optimization (Screening Parameters) The experimental results revealed porosity to be one of the most important parameters affecting the release profile. Drug loading directly affects the ultimate porosity and therefore is a high impact parameter. Faster release rate was observed with the spray congealing technique as compared with granulation technique with all other parameters being the same. It is suspected that the spray congealing process via air atomization, adds to the overall porosity by introducing air bubbles into the matrix enhancing the solvent penetration, dissolution of the API, and diffusion out of the matrix. The formation of the less crystalline Alpha and Beta Prime polymorphs may also contribute to the faster release. They are known to be less hydrophobic than the most stable Beta polymorph and may enhance wettability of the narrow and tortuous channels within the matrix. R esponse surface plots demonstrated the significant impact of the drug loading and encapsulation technique over the release profiles. Confocal Microscopy and Image Analysis A Confocal microscopy technique was developed that allowed visualization of the dist ribution of the drug particles within the microspheres. Nifedipine and Propranolol autofluoresce eliminating the need to utilize fluorescent tags. The results are promising and warrant further investigation. By utilizing image analysis tools, it was fe asible to

PAGE 64

64 determine the diameter and the size distribution of the drug particles within the microspheres, which in turn was a measure of the ultimate porosity for the microsphere. In vivo Rat Experiments In vivo rat experiments were very promising demonst rating sustained release over the Instant Release (IR) formulation and enhance oral bioavailability (about 9fold improvement in AUC; Area Under the Curve for plasma concentration vs. time). The rat experiment was repeated and the number of rats per group were doubled from 3 to 6. The results confirmed findings from the first experiment. Therefore, it is feasible to enhance oral bioavailability by utilizing this approach.

PAGE 65

65 CHAPTER 5 MATHEMATICAL MODELIN G Modeling Release Kinetics Mathematical modeling has been used to predict the release profile of the APIs from the elaborate systems designed to control it. These mathematical models can range from simple concentrationdriven diffusion (Ficks law) across a certain plane, to elaborate models considering to rtuous path, three dimensional release, as well as considerations for concentration, time, and distance gradients. In this chapter the basic concepts are introduced and the fitting of these models to the empirical data is analyzed. Fickian Diffusion Cr anks book on Mathematics of Diffusion is an invaluable source of information on this topic. Ficks first law measures the rate of diffusion for the material of interest across an imaginary plane. As can be seen from the equation below [Crank], this rat e is directly proportional to the concentration gradient. This is a onedimensional assessment of diffusion in the direction of x, assuming a constant diffusion coefficient. Ficks First Law: F = Flux = Rate of transfer per unit area o f the diffusion plane; g/cm2s D = Diffusion Coefficient; cm2/s C = Concentration; g/cm3 x = Distance from diffusion plane; cm

PAGE 66

66 Ficks second law measures change in concentration with respect to time. The elements of the equation are the same as described above with the addition of the term t for time in seconds. Ficks Second Law: The solution by integration yields: The equations above determine diffusion of a substrate through a media of choice. The Diffusion Coefficient will have to be determined empirically. These equations are simplified in that D is assumed constant and not changing with time. Also, it only addresses the diffusion in only one direction. When taking into consideration diffusion in three dimensions, the shape of the model will have to be considered. In rectangular coordinates, the equations will have to be adjusted to take into consideration the diffusion in x, y, and z directions: Notice that in this case D is considered to be anisotr opic (not uniform in all directions). This can be simplified by assuming a uniform media and one D can be applied to all directions. Other shapes can be modeled, but since in this effort microspheres and their release are being considered, the focus will be on spherical coordinates. In this case, x, y, and z will have to be replaced with radial coordinates:

PAGE 67

67 Q = Quantity of diffusing substance which passes through the spherical wall in time t a and b are the radius at C1 and C2, respectively. Monolithic Systems, DiffusionControlled Release Profile, and Percolation Release profiles from controlled release systems can vary widely. However, most release profiles can be categorized into one of the three common types : 1) Zero Order Release Profile: This refers to a constant API release ra te, resulting in a linear release profi le. The governing equation is as follows : a) dMt/dt = K ; where K is a constant and Mt is the mass of the API released at time t. 2) First Order Release Profile: The rate of drug release declines exponentially as the A PI source within the system approaches exhaustion. a) dMt/dt = KM0eKt; M0 refers to the initial mass of API within the matrix 3) Square Root of Time Release Profile: This is the most applicable type of release profile for our application. a) Mt = Kt1/2

PAGE 68

68 For our application, we are considering the microspheres to be constructed of an insoluble shell material with the drug uniformly distributed within (Monolithic System) as illustrated below: Figure 51. Illustration of uniform drug distribution within the shell matrix (Monolithic System) The model assumes a matrix without erosion, permeability, and/or swelling. The microsphere shell material (matrix) is not soluble in aqueous environment, but the drug distributed within dissolves at a rate proportional to its s olubility in the media. The exposed drug at the surface is dissolved first, diffusing out of the microsphere and into the bulk media, driven by the concentration gradient (diffusing from high to low concentration region). This is known as a monolithic sy stem in which the drug release is controlled by diffusion. In this case, the release of drug into the surrounding environment is governed by the solubility of the drug particles and diffusion through the porous matrix that is generated as the dissolution progresses. The drug diffusion boundary starts at the surface of the microsphere and moves towards the center. As the dissolution of the randomly distributed drug particles progresses, voids are generated that eventually form a porous network of connected channels through which the drug has to diffuse to be released into the surrounding environment. The path

PAGE 69

69 through which the dissolved drug has to diffuse is not linear and is referred to as a tortuous path. The length of the path travelled divided by the linear distance between start and finish is a measure of Tortuosity A useful model that has been applied to flow of fluids through porous matrices is percolation. These models are employed to evaluate flow of fluids through porous matrices. The model s assess the connectivity of void channels through which fluid can flow from point a to point b. In our case, this can be applied to determine if there are enough voids to form a continuous path from any drug particle to the external surface of the micros phere. Percolation threshold is a mathematical term that signifies the critical void fraction at which this connectivity is formed. Below percolation threshold the voids are isolated and there is no path from point a to b. Above percolation threshold there is a path through the void channels that connects the dissolving drug to the external surface of the microspheres allowing diffusion to occur. In our case, this threshold occurred between 10 and 20% loading (wt./wt.). As can be seen in Figure 5 2 bel ow, the continuous connectivity of the pores had not occurred at 10%, whereas at 20% it had been reached (above the peak threshold).

PAGE 70

70 0.00 20.00 40.00 60.00 80.00 100.00 120.00 0 5 10 15 20 25 30 % Released Time (Hrs) Spray Congealed M SC 10 M SC 20 M SC 30 Figure 5 2 Propranolol Release Profiles at 10, 20, and 30 % loading (wt./wt.) indicating that percolation threshold had not been reached at 10 % loading. Zaller had observed this connectivity to form a continuous network take place at 15% loading in three dimensional matrices This is in line with the above observation. Tortuosity and porosity are both captured in the Higuchi equation predicting drug release from monolithic systems. Higuchi (1961) was the first to suggest a square root of time dependence for drug release rate from an insoluble matrix based on Fickian diffusion [Gohel et al.]. In his 1963 article, Hig uchi provided an equation that accounted for the release of drug from a matrix system that would account for porosity and tortuosity. This equation is listed as follows: s)Cst]0.5=m/S0=KH t0.5 Q = Amount of drug release after time t per unit of exposed area m = Amount of drug released after time t

PAGE 71

71 S0 = Surface area of matrix exposed to the fluid D = Diffusion coefficient of the drug in the permeating fluid = Tortuosity A = Initial amount of drug in the matrix per unit volume Cs = Solubility of the drug in the permeating fluid KH = Release rate constant Yonezawa et al. have determined that this model started to fail for wax matrix tablets about the time that half the initial drug content had been released. T he Propranolol HCl release profile from the wax MSs (Figure 52), shows a very good fit to Higuchi model (Figure 61 and 62, below). However, the total release in 24 hours for the Propranolol HCl Microspheres was 57%, which is not much beyond the limit Yonezawa suggested (50% for wax matrices) and within the limit Grassi et al suggested (60%, 2005, any matrix). 0.000 10.000 20.000 30.000 40.000 50.000 60.000 0.000 5.000 10.000 15.000 20.000 25.000 30.000 Time (hours) % Released Higuchi Model Real Time Actual Propranolol Release from Wax MS Figure 5 3 Plot of time versus % released of Propranolol HCl compared with Higuchi model

PAGE 72

72 0.000 10.000 20.000 30.000 40.000 50.000 60.000 0.000 1.000 2.000 3.000 4.000 5.000 6.000 SQRT of Time (hrs) % Release Higuch Model Profile SQRT t Actual Propranolol Release from Wax MS Figure 5 4 Plot of square root of time versus % released of Propr anolol HCl compared with Higuchi model As can be seen in Figures 5 3 and 5 4 there is good correlation between the square root of time model equation proposed by Higuchi and the Propranolol HCl released from our wax matrix microspheres. Analyzing these microspheres in acidic environment (dissolution at pH 2) to simulate gastric environment, did not affect the square root of time release behavior observed at neutral pH (7). Figure 5 5 Release profile of Propranolol HCl microsphere in acidic envir onment (pH 2) y = 32.89x 30.325 R = 0.9946 20 0 20 40 60 80 100 0.0 2.0 4.0 6.0% RELEASEDSQ RT TIME (HR 1/2)SQ RT PLOT V-98

PAGE 73

73 Tortuosity is the ratio of the actual diffusion path length from point a to b, divided by the linear distance between the two points [Wu et al.]. This is generally measured empirically from dissolution release profiles. Wu et al. have suggested an optical method and algorithm for determining tortuosity within compacted particles (4575 and 212250 micron fractions). In our case, it would be beneficial to determine the pore structure within the microspheres. One potential option was to examine the distribution of the drug within the microsphere. This was accomplished by utilizing Confocal Microscopy. With this technique, it was feasible to determine the drug distribution within the microspheres (covered in the results section, Ch 5). W ith image analysis, the mean particle size within the microspheres could be determined. In order to achieve reproducible controlled release, the drug size and distribution within matrix have to be controlled. To allow for a uniform distribution of the drug within the desired 500 micrometer microspheres, improve solubility rate of drug, and reduce dose dumping, it was opted to reduce the drug particle size to below 10 microns by jet milling. It is an acceptable pharmaceutical practice with very good control of particle size and proven GMP (Good Manufacturing Processes; required for pharmaceutical products) processing methods. Figure 64 shows typical particle size distribution of Propranolol HCl achieved by jet milling process. Majority of the particles w ere below 3 microns in size. As can be seen from the size distribution chart, the apparatus had a lower limit of detection around 0.7 microns. It is evident from the chart that the distribution is cut off at this lower limit and the calculated mean parti cle size of 1.96 may

PAGE 74

74 be actually lower if the smaller particles were detected and included in the overall distribution. However, the reduction in size due to milling reduced the dose dumping effect observed with encapsulations with larger drug particles This is suspected to be due to increased tortuosity that will result due to the smaller pores that will ultimately form T here was need for a methodology to determine if the API particles remained discreet or aggregated into larger masses within the Mic rospheres. As previously mentioned, Confocal microscopy was utilized and proved to be an effective tool in answering this question. By applying image analysis it was feasible to determine the size of the particles within the microspheres utilizing the Co nfocal micrographs.

PAGE 75

75 Figure 5 6 Sample size distribution chart for jet milled Propranolol HCl showing that 95% of the particles are below 10 microns with a volume averaged mean diameter of 1.96 microns. Discussion and Conclusions In this chapter mathematical models were applied to the resulting release profiles from our diffusioncontrolled monolithic systems with very good correlation. Higuchi et al developed models for release of drug from wax matrices that required introduction of some new paramet ers (tortuosity and porosity ). This model deviates from a true Fickian release (only dependent on concentration gradient ) and are referred to as square root of time release type since a linear line is observed when the amount of drug released is graphed v s. square root of time. There are many adaptations of Higuchis mathematical model customized to specific applications (anisotropic matrices, shape effects, etc.) but the original model has stood the test of time and is generally a good fit

PAGE 76

76 if applied c orrectly The release profiles generated at low loadings (10 %) demonstrated the lack of connectivity between the pores in these matrices. This critical drug loading at which the resulting porous structure forms a continuous network is referred to as per colation threshold, which is observed to occur at about 15% void volume (Zaller). The continuous network of porous structure is clearly demonstrated in the release profile of 20% loaded Propranolol HCl microspheres.

PAGE 77

77 CHAPTER 6 CONCLUSIONS In summary, by applying the appropriate microencapsulation techniques, it was feasible to achieve the desired release profile utilizing fats and waxes as the enca psulating matrix. The most significant finding from this body of work was that applying this approach to the model poorly soluble API (Nifedipine), increased oral bioavailability in rats In vivo rat exper iments were repeated with larger n per group and confirmed the findi n g from the initial rat experiment. L ot to lot variability of the raw material and the in herent change in morphology over time, may pose challenges for APIs with narrow therapeutic window Microfluidization did not show any significant advantage in release profiles over stirrer mixing or homogenization of the molten mix. The drug release rates from granulated samples, regardless of pretreatment, are significantly slower than spray congealed samples ( see Figure 511 ). This is suspected to be mostly due to the added air porosity introduced by the spray congealing atomization process. G ranulation techniques may prove to have more utility than spray congealing for formulation development. The loading trends behaved as expected with the higher loading microspheres showing faster release profiles as evident in both Appendix C and D DSC results indicate that annealing of spray congealed samples can eliminate the lower melting point endotherm. However, the crystal structure of the final products needs to be better understood. The melting temperatures of the annealed/spray congealed and granul ated formulations, independent of pretreatment, exhibit ed slight differences. The Alpha to Beta polymorph transition must be controlled to affect a

PAGE 78

78 stable end product. The parameters that drive/control th e polymorph transition need to be characterized a nd understood in more detail. Based on these results it appears that the Alpha to Beta transition is driven by time, temperature and cooling rate during processing steps. The literature has identified polymorphs within fat matrices ( mentioned here from least to most stable) has been identified as hexagonal, orthorhombic, and triclinic, respectively. Cooling rate from a molten state has been suggested and identified as the most important parameter in the formation of the resulting polymorphs. The lowest state of energy being the X Ray diffraction of spray congealed hydrogenated vegetable oil showed 52% crystallinity with the Alpha polymor ph constituting >60% of the composition. Upon annealing the crystallinity increased to 62% with the Alpha polymorph dropping from >60% of the composition to <50%. There are inherent difficulties in controlling release profile when utilizing naturally oc curring materials such as fats and waxes. For this reason, it may be beneficial to also explore synthetic polymers as matrices. The chemistry, morphology, and critical attributes of these synthetic polymers may be easier to control, leading to more reproducible results Future Work The following is suggested for future work: 1. Apply this technique to other APIs in need of enhanced bioavailability to determine the potential for broader applicability of this approach. 2. The parameters that drive/control the polymorph transition need to be characterized and understood in more detail.

PAGE 79

79 3. Further develop the Confocal microscopy technique for determining the drug distribution within the microspheres and predicting the potential porosity of the matrix. 4. T he crystal structure of the matrices need to be explored further Understanding the critical threshold of degree of crystallinity and ratio of Beta to other crystal polymorphs will lead to more stable products for the desired shelf life. 5. Explore combining lipid matr ices for further optimization of release profiles and enhancement of oral bioavailability 6. Chemical modification of lipids should be explored to customize their properties.

PAGE 80

80 APPENDIX A: STANDARD PROTOCOLS AND PROCEDURES Encapsulation Process Figure A 1. Encapsulation Process Spray Congealing Other Additives Less than 5 Dispersion Drug Melted Wax Shell Spray Mix

PAGE 81

81 Sample Spray Congealing Procedure Equipment and Materials 1) Stainles s steel scoop 2) Wax 3) Drug 4) Glascol Va riac/temperature controller (14 A mps output) 5) 1500W Glascol heating mantle 6) Kitchen Aid stirrer + blade 7) Sta inless steel spatula 8) IR gun 9) Corning stirrer/hot plate 10) 4L stainless steel bowl that is part of kitchen aid 11) Nordson Seri es 3700 hot melt applicator Procedure (35% loading wt./wt.; 1500 g batch size) 1) Weigh 975 g of wax into a 4 L stainless steel bowl that is part of kitchen aid. 2) Heat the wax with a heating mantle controlled to 20 F above the melting point while stir ring the wax with a h ook connected to kitchen aid at 50 rpm. 3) W eigh 525 g of milled drug 4) Add drug slowly to melted wax over a period of 15 minutes 5) Let sample stir by Kitchen aid mixer for additional 10 min utes 6) Spray the mixture (speed setting= 40%, aspirator pres sure of 10 psi ) through Nordson 3700 series hot melt applicator maintained at 20 F above the melting point of the wax. 7) Spray sample into the chill zone

PAGE 82

82 8) Collect microspheres for sieving Figure A 2 Kitchenaid Mixer S ample Granulation Procedure Equipment and Materials 1) Stainless steel scoop 2) Wax

PAGE 83

83 3) Drug 4) Glascol Va riac/temperature controller (14 A mps output) 5) 1500 W Glascol heating mantle 6) Kitchen Aid stirrer + blade; or Silverson high shear lab bench homogenizer; or microfluidizer as required for the mixing step. 7) Sta inless steel spatula 8) IR gun 9) Corning stirrer/hot plate 10) 4L stainless steel bowl that is part of kitchen aid Procedure (35% loading wt./wt. ; 1500 g batch size) 1) Weigh 975 g of wax into a 4 L stainless steel bowl that is part of kitchen aid. 2) Heat the wax with a heating mantle controlled to 20 F above the melting point while stir ring the wax with a hook connected to kitchen aid at 50 rpm. 3) W eigh 525 g of milled drug 4) Add drug slowly to melted wax over a period of 15 min utes 5) Let the mixture stir by kitchen aid for an additional 10 minutes 6) Turn off the heating mantle and let stir until the bowl and its content are at room temperature. 7) Collect the granulated microparticulates for sieving

PAGE 84

84 Sieving The Table below summarizes the sieve assembly configuration and conditions for sieving the microspheres. The Mesh size of the sieves and their corresponding opening in micrometers is also listed. Table A 1. Sieving Configuration US Std Number (Mesh) Opening (Microns) Lid (top) N/A 20 841 40 420 60 250 Fines Collector (bottom) N/A Sieve Assembly Configuration Stainless Steel 3" Diameter Full-Heigh Sieves Cole Parmer Sieve Shaker Frequency: 60 taps/min for 5 minutes

PAGE 85

85 Dissolution EQUIPMENT AND MATERIALS Distek Dissolution Apparatus I I Bath Model 2100C, UV detector Agilent 8453 CN02502008 Figure A 3 Dissolution Apparatus (Model 2100C) set as Apparatus II (paddles) Microscope Olym pus BX60 Potassium Phosphate, monobasic JT Baker Lot # record Sodium Hydroxide JT Baker Lot # record Purified Water Wissahickon Balance Ohaus pH meter Orion Model 520 Propranolol HCl Standard Wyckoff #1 of 2 lot # record ( USP ) Sonicator VWR model 150T P orcelain Mortar and Pestle

PAGE 86

86 Propranolol Microspheres Amount (mg) = (0.9 900 ml 0.178 mg/ml)/(loading ; wt./wt.) PROCEDURE Preparation of dissolution media 136.03 g Potassium Phosphate and 18.01 g NaOH dissolved in 20 L of water. pH= 6.80, no adjustment needed. Preparation of Propranolol HCl Stock Standard Standard dried 105 C for 4 hours prior to use. 041002 Std wt. 89.1 mg 041202 Std wt. 88.9 mg Propranolol HCl Std transferred into a 50 ml volumetric flask. Dissolve in dissolution media, soni cated 1 min, and filled to volume, mix. [Propranolol HCl] = 1.78 mg/ml Preparation of Propranolol HCl Working Standards Std Level Vol. Stock Std (ml) Vol. Flask (ml) [Propranolol HCl] mg/ml 20% 2 100 0.0356 40% 4 1 00 0.0712 60% 6 100 0.1068 100% 10 100 0.178 120% 12 100 0.2136 Dissolution Conditions Apparatus I, baskets Speed 50 rpm Detection 289.5 nm Cell 10 mm Detection times 0.25, 0.5, 1, 2, 4, 6, 10, 14, 19, 23, 24 hr Vessel Temperature 37 0.5 C

PAGE 87

87 Propranolol HCl HPLC Procedure I. SCOPE: This method describes the procedure for determining the Propranolol HCl concentration in samples collected from the Distek dissolution apparatus over 24 hours via automated fraction collector. The results will be used to generate the cumulative release profile of Propranolol HCl from the microspheres batch being examined. II. REAGENTS: o Propranolol HCl, USP Reference standard or Inhouse Reference Standard o Potassium phosphate, monobasic (Ultra pure; bioreagent) o Phosphoric Acid (ACS grade) o Acetonitrile (HPLC grade) o Distilled Water III. EXPERIMENTAL: o Column: Symmetry C8, 4.6 x 250 mm, 5 micron, Waters Corporation, Milford, MA, USA or Equivalent o Column Temperature: Ambient o Wavelength: 220 nm o Injection Volume: 25 L o Flow Rate: 1.5 mL/min o Run Time: 15 min IV. INSTRUMENT: o Liquid Chromatograph: Shimadzu LC 10AT or LC lOATvp or equivalent o Auto Injector: Shimadzu SIL10A or equivalent o UV VIS Detector SPD 1OA or equivalent o System Controller: Shimadzu SCL1OA or SCL lOAvp or equivalent o Data Acquisition System: Class VP Data Acquisition System or Equivalent

PAGE 88

88 V. PREPARATION OF POTASSIUM PHOSPHATE SOLUTION: Dissolve 6.8 g of potassium phosphate monobasic in 1000 mL of Distilled Water. Mix well. VI. PREPARATION OF M OBILE PHASE: Transfer 750 mL of Potassium Phosphate Solution and 250 mL of Acetonitrile in to a suitable container. Adjust the pH to 3.25 0.10 using dilute phosphoric acid. VII. PREPARATION OF RINSE FLUID: Use Acetonitrile as rinse fluid. VIII. PREPARATION OF PH 6.8 PHOSPHATE BUFFER: Dissolve 6.8 g of Monobasic Potassium Phosphate and 0.9 g of Sodium Hydroxide in Distilled Water and make up the volume to 1000 mL with Distilled Water. Adjust the pH to 6.8 0.10. IX. ASSAY OF RELEASE PROFILE SAMPLES: 1) Preparation O f Propranolol HCl Stock Solution: Weigh accurately 100 mg of Propranolol HCl, USP or Inhouse reference standard and transfer it to 100 mL volumetric flask. Dissolve in pH 6.8 Phosphate buffer and make up the volume with the same.

PAGE 89

89 2) Preparation Of Working Standard Solutions: Working Standard Volume of Stock Std. to be used Q.S. to with pH 6.8 phosphate buffer Concentration of working standard Actual working std. Concentration (Linearity Level) (mL) (mL) (mcg/mL) If In house reference standard is used 1 2 100 20 100 Std. house In of Potency Std. working of Conc. 2 4 100 40 3 8 100 80 4 12 100 120 5 16 100 160 6 20 100 200 3) System Suitability Check: Obtain at least 5 replicate chromatograms of Propranolol HCl working standard (preferably with linearity level #1) 4) Reproducibility: Calculate the percentage relative standard deviation of the peak areas of the five replicate chromatograms. Specification: The relative standard deviation (RSD) < 2.0% 5) Number Of Theoretical Plates: Calculate the number of Theoretical Plates, n, using the equation: n= 16(t/w)2 Where, t = Retention time of the Propranolol HCl component peak w = Width of the Propranolol HCl component peak at the base obtained from the relatively straight sides of the peak (in time units). Specificat io n: Theoretical Plates, (n) > 4000 6) Tailing Factor:

PAGE 90

90 Determine the Tailing Factor, T, using the equation: T = W0.05/2f Where, W0.05 = Width of the peak at 5% of the peak height f = Width in (time) between the peak maximum and the front edge of the peak at 5% of the peak maximum Specification: (T) NMT 2.0 Note: If the system suitability parameters are not met, take corrective action, which includes adjusting the flow rate, mobile phase composition, pH, column temperature, flushing or replacing t he column. 7) Sample Preparation: Use samples as is collected from the fraction collector of dissolution system. 8) Sequence: Place the working standards and samples in rack 2 of the HPLC in such a way that all six working standards are evenly spaced throughout the run. Example of a sequence is given below: Vial # Sample Type # of Injections 0 Blank 2 1 Level 1 Std. 5 7 thru 18 Samples (1/2 hour & 2nd hour) 1 ea. 2 Level 2 Std. 3 19 thru 24 Samples (6th hour) 1 ea. 3 Level 3 Std. 3 25 thru 30 Samples (11th hour) 1 ea. 4 Level 4 Std. 3

PAGE 91

91 31 thru 36 Samples (14th hour) 1 ea. 5 Level 5 Std. 3 37 thru 42 Samples (19th hour) 1 ea. 6 Level 6 Std. 3 43 thru 54 Samples (23rd & 24th hours) 1 ea. 6 L evel 6 Std. 3

PAGE 92

92 DSC Procedure Required E quipment 1. Analytical balance with 0.1 mg accuracy reading. 2. TA instruments DSC Q10 instrument + software or equivalent 3. High purity nitrogen gas. 4. Spatula. 5. Hermetic Pans and Press Preparing Hermetic P ans 1. Put hermetic pan on balance set to units of mg and tare. 2. Carefully place 210 mg of sample in the pan. Do not spill or let sample touch the lip of the pan. 3. Record weight. 4. Put hermetic lid on the pan and place pan + lid on the lower die of the press. 5. Place the flat side of the performing tool against the upper die of the press and press the pan by lowering lever till stop. 6. Raise lever. Turn performing tool with flat side facing down and hold against the upper die. 7. Press pan t ill stop. Press as hard as possible. 8. Pan should have a smooth surface and there should be a complete seal around the circumference. Set up E xperimental Conditions and Run S amples I. Go to DSC, Parameter, experimental.

PAGE 93

93 2. Enter sample name, sample size, name, comments, pan type, purge gas type, flow rate. 3. At "method file", click on"..." button to select desired method. 4. Click load to load method. 5. At "save data file", check the box to have data saved while running. Enter desired file name for dat a to be saved. 6. Exit experimental parameters. Note: Recommend using flow rate of 56ml/min with a reading of 70mm at the rotameter. 7. Load sample + reference pan. Reference pan should be on the rear side of the cell. Cover cell. 8. Go to DSC, control, st art. Or press the green triangle on the tool bar or the green button on the machine.

PAGE 94

94 Confocal Microscopy Bio Rad 1024 ES Inverted Confocal Microscope equipped with MRC 1024 Confocal laser scanning system was used. The wavelength of 488 nm (FITC) was us ed to excite all samples. Samples were deposited onto Corning cover slips, No.1, 24 X 50mm and mounted with Stephens Scientific Cytoseal 60, low viscosity, mounting media. The Cytoseal 60 is an acrylic resin. The samples were prepared fresh and imaged imm ediately following the addition of the Cytoseal 60.

PAGE 95

95 X Ray Diffraction Sample experimental conditions for powder X Ray Diffr action are captured in the Figure below: Figure A 4 Powder X Ray Diffraction data conditions

PAGE 96

96 In Vivo Rat Experimental Procedures (Vaghefi et al ; 2005) Blood/Plasma Collection Blood samples (0.25 mL) are collected by jugular vein puncture. For the oral doses, blood samples are collected predose and then at 0.25, 0.5, 1, 2, 4, 6, 8 and 10 hours following dose administration. F or the intravenous dose, blood samples are collected predose, at the end of the infusion and then at 0.25, 0.5, 0.75, 1, 1.5, 2, 3 and 4 hours following the end of the infusion. The blood samples are transferred into polypropylene tubes (covered with alumi num foil and containing lithium heparin as anticoagulant). Plasma samples are separated using a refrigerated centrifuge and stored frozen at 65 C. in foil covered polypropylene tubes. Plasma Analysis The analytical method used to measure Nifedipine is ba sed on a previous method developed for Nifedipine in human heparinized plasma. A 2day revalidation is performed in the rat plasma matrix. Standards and controls are prepared in heparinized rat plasma. All samples underwent liquidliquid extraction to isolate Nifedipine, followed by reversed phase chromatography, and quantitative detection using MRM mass spectrometry. The internal standard used is a d6 stable isotope of Nifedipine. The analyt ical range of the method is 5 ng/mL to 500 ng/mL using a 50 L sample. Studies of the stability of Nifedipine in plasma samples for 4 hours at room temperature, and through 3 freezethaw cycles indicated no loss of Nifedipine.

PAGE 97

97 LIST OF REFERENCES Bernstein J., Polymorphism in Molecular Crystals, Oxford Science Publicat ions, ISBN 0 198506058, p 2728 (2002) Crank J, The Mathematics of Diffusion, Oxford Science Publications, ISBN 9780 198534112 (1979) Cusimano A. G. and Becker C. H., J. Pharm. Sci. 57(7): 1104 (1968) Carnauba wax Cussler E. L., Diffusion: Mass Trans fer in Fluid Systems (3rd edition), Cambridge University Press, ISBN 9780 52187121 1 (2009) Daniel D. R., Texas Tech University Dissertation, The Chemical and Functional Properties of Cottonseed Oil as a DeepFat Frying Medium deMan L., deMan J.M., Blac kman B., Polymorphic behavior of some fully hydrogenated oils and their mixtures with liquid oil, JAOCS, Vol. 66, no. 12 (December 1989). Eldem T, Speiser P, Hincal A. Optimization of spray dried and congealed lipid micropellets and characterization of their surface morphology by scanning electron microscopy. Pharm Res. 1991;8:4754 Eldem T, Speiser P, Hincal A. Polymorphic behavior of sprayed lipid micropellets and its evaluation by differential scanning calorimetry and scanning electron microscopy. Pharm Res. 1991;8:178184. Emas M., Nyqvist H., Methods of studying aging and stabilization of spray congealed solid dispersions with carnauba wax. 1. Microcalorimetric investigation, International Journal of Pharmaceutics, 197:117127, (2000). Ensikat H. J., B oese M., Mader W., Barthlott W., Koch K., Crystallinity of plant epicuticular waxes: electron and X ray diffraction studies, Chemistry and Physics of Lipids, 144: 4559 (2006). Frishman W H., Garofalo J. L., Rothschild A., Rohschild M., Greenberg S., So berman J., The Nifedipine gastrointestinal therapeutic system in the treatment of hypertension, The American Journal of Cardiology, 19: 6569, (1989) GhebreSellassie I., Multiparticulate oral drug delivery, Marcel Dekker Publisher, Drugs and Pharmaceuti cal Sciences, Vol 65, Pages 1734, 1994; ISBN 0824791916 Giannola L. I., Carnauba wax microspheres loaded with valproic acid: preparation and evaluation of drug release, Drug Dev. Ind. Pharm., 21: 15631572, (1995)

PAGE 98

98 Gohel M. C., Panchal M. K., Jogani, V. V., Novel mathematical method for quantitative expression of deviation from the Higuchi model, AAPS PharmSciTech, 1(4) article 31, (2000) Grassi M., Grassi G., Mathematical modeling and controlled drug delivery: Matrix Systems, Current Drug Delivery, 2: 97 116 (2005). Hamid I. S. and Becker C. H., J. Pharm. Sci. 59(4): 511 (1970) Heertje I., Leunis M., Measurement of shape and size of fat crystals by Electron Microscopy, Lebensm. Wiss. U. Technol., 30, 141146, (1997). Heertje I., Leunis M., Van Zeyl W. J M., and Berends E., Product morphology of fatty products, Food Microstructure, 6, 18, (1987). Heertje I., van Eandenburg J. M., Cornelissen J. M., Juriaanse A. C., The effect of processing on some microstructural characteristics of fat spreads, Food Mic rostructure, Vol. 7: 189193 (1988) Herrera M. L., Crystallization Behavior of Hydrogenated Sunflowerseed oil: Kinetics and Polymorphism, JAOCS, Vol. 71, no. 11, 12551260, (November 1994) Higuchi T., Rate of release of medicaments from ointment bases cont aining drugs in suspension. J. Pharm. Sci. 50:874875 (1961) Higuchi T., Mechanism of sustained action medication: theoretical analysis of rate of release of solid drugs dispersed in solid matrices. J. Pharm. Sci., 52:11451149, (1963) Hui Y. H., Bailey s Industrial Oil & Fat Products, Fifth Edition, Wiley Interscience, ISBN 0 47159424 5 (1996). Jenning V., Gohla S., Comparison of wax and glyceride solid lipid nanoparticles, International Journal of Pharmaceutics, 196:219222, (2000) John P. M. and B ecker C. H., J. Pharm. Sci. 57(4): 584 (1968). Kou S, Transport Phenomena and Materials Processing, Wiley Interscience, ISBN 047107667 8 (1996) Lantz R. J. and Robinson M. J., US Patent 3,146,167 (1964). Leeuwenkamp O. R., Visscher H. W., Mensink C. K., Jonkman J. H., A comparative study of the steady state pharmacokinetics of immediate release and controlled release Diltiazem tablets, European Journal of Clinical Pharmacology, 46:243247, (1994)

PAGE 99

99 Litwinenko J. W., Rojas A. M., Gerschenson L. N., Marangni A. G., relationship between crystallization behavior, microstructure, and mechanical properties in a palm oil based shortening, JAOCS, Vol. 79, no. 7 (2002). Mani N., Jun H. W., Microencapsulation of a hydrophilic drug into a hydrophobic matrix using a salting out procedure. I: Development and optimization of the process using factorial design, J. Microencapsulation, Vol. 21: No. 2, 125135 (March 2004) Mathiowitz E., Encyclopedia of Controlled Drug Delivery, Volume II: 736, Wiley Interscience, ISBN 0471 166634 (1999) Miyagawa Y., Okabe T., Ymaguchi Y., Miyajima M., Sato H., Sunada H., Controlled release of diclofenac sodium from wax matrix granule, Int. J., Pharm., 138:215224, (1996). Nassu R. T., Goncalves L. A. G., Determination of the melting point o f vegetable oils and fats by differential scanning calorimetry (DSC) technique, Vol. 50 Fasc. 1, 1622 (1999) Nikon Corp., Laser Scanning Confocal Microscopy (LSCM), http ://www.microscopyu.com/articles/confocal/confocalintrobasics.html Phuapradit W., Shah N. H., Lou Y., Kundu S., Infeld M. H., Critical processing factors affecting rheological behavior of a wax based formulation, European Journal of Pharmaceutics and Biopharmaceutics, 53:175179 (2002) Prasad C. M., Srivastava G.P., Study of some sustained release granulations of aspirin, Indian J. Hosp. Pharm. 1971; 8:2128. Raghunathan Y. and Becker C. H., J. Pharm. Sci. 57(10): 1748 (1968). Reza M. S., Quadir M. A., Ha ider S. S., Comparative evaluation of plastic, hydrophobic and hydrophilic polymers as matrices for controlledrelease drug delivery. J. Pharm. Sci., 6(2):282291, (2003) Riiner U., Phase behavior of hydrogenated fats III. Phase equilibria and crystal s izes, Lebensm. Wiss. U. Technol. Vol. 4 (1971) Rivarola G., Segura J.A., Anon M.C., Calvelo A., Crystallization of hydrogenated Sunflower Cottonseed oil, JAOCS, Vol. 64, no.11 (November 1987) Sahimi M., Applications of Percolation Theory, Taylor & Franci s Publishers, ISBN 0 748400761 (1994) Schwartz J.B., Simonelli A.P., and Higuchi W.I. Drug release from wax matrices I: Analysis of data with first order kinetics and with the diffusion controlled model. J. Pharm. Sci., 57:274277, 1968a.

PAGE 100

100 Schwartz J.B., Simonelli A.P. and Higuchi W.I. Drug release from wax matrices II: Application of mixture theory to the sulfanilamidewax system. J. Pharm. Sci., 57:278282, 1968b. Scott M. W., Robinson M. J., Pauls J. F., and Lantz R. J., J. Pharm. Sci. 53(6): 670 (1 964); Smith W. E., Buehler J. D., and Robinson M. J. J. Pharm. Sci. 59(6): 776 (1970). Stauffer D., Aharony A., Introduction to Percolation Theory, Second Edition, CRC Press, ISBN 978 0 748402533 ( 1994) Stoyle L., Quellette P., and Hanus E., US Patent 3,035,985 (1962) Stoyle L., Quellette P., and Hanus E., US Patent 3,037,911 (1962) Vaghefi, F.; et. al., Microcapsule matrix microspheres, absorptionenhancing pharmaceutical compositions and methods (Patent # 6,849,271 ; 2005 ) Varshosaz J., Keihanfar M., Development and evaluation of sustainedrelease Propranolol wax microspheres, J. Microencapsulation, Vol. 18: No. 3, 277284 (2001) Wade A., Weller P. J., Handbook of Pharmaceutical Excipients, 2nd Edition, American Pharmaceutical Association, ISBN 0917 30 66 8 (1994) Wu Y. S., van Vliet L. J., The determination of relative path length as a measure for tortuosity in compacts using image analysis, European Journal of Pharmaceutical Sciences, 28: 433440 (2006). Yonezawa Y., Ishida S., Sunada H., Release fr om or through a wax matrix system. VI. Analysis and prediction of the entire release process of the wax matrix tablet, Chem Pharm. Bull., 53(8) 915918 (2005) R. Zaller, in Proceedings 13th IUPAC Conference of Stat. Phys., Haifa, Israel, IUPAC p. 310 (1977)

PAGE 101

101 BIOGRAPHICAL SKETCH Farid Vaghefi attended Gainesville High School, Gainesville, Florida. As an undergraduate at University of Florida he pursued chemistry and chemical engineering degrees simultaneously, graduating with Bachelor of Science in 1984 and 1985 respectively. He graduated in 1990 with a m asters degree (with Thesis) from University of Floridas Material Science and Engineering department. Since then, he has been working in the BioPharmaceutical industry and has been granted 13 patents