Research of an Optimal Formulation for Direct Compression of a … · 2020. 7. 30. · not a compressible material, methodologies such as Design of Experiments (DOE) and Quality by

Research of an Optimal Formulation for Direct Compression of a Natural Product

(From Wet Granulation to Direct Compression)

Mémoire

Mélissa Maltais

Maîtrise en sciences pharmaceutiques

III

Résumé

La compression directe est l’une des méthodes de fabrication de comprimés les plus

communes et économiques de l’industrie pharmaceutique. Cependant, l’utilisation de ce

procédé requiert souvent un changement de formulation. Puisque le séné, un extrait naturel

de plante, n’est pas un matériel compressible, des méthodologies de conception

d’expériences et de qualité par conception ont été utilisées pour développer une formulation

convenable pour la compression directe des comprimés de séné. Le nouveau produit

formulé a démontré plusieurs bénéfices reliés au temps de fabrication, aux besoins en main

d’œuvre et aux besoins en machineries. De plus, la stabilité du produit, emballé dans des

bouteilles de polyéthylène exposées à des conditions extrêmes pendant six mois, a

démontré des résultats conformes aux limites spécifiées à l’interne. Cependant, il reste

quelques défis à relever concernant la détermination de la vitesse de fabrication à grande

échelle et de la stabilité du produit à long terme. Néanmoins, l’implémentation du procédé

de compression directe démontre suffisamment de bénéfices à long terme pour remplacer le

procédé courant de granulation humide.

V

Abstract

Direct compression is becoming one of the most common and economical method of

tablets manufacturing in the pharmaceutical industry. However, the feasibility of this

process often results in formulation change. Since senna powder, a natural plant extract, is

not a compressible material, methodologies such as Design of Experiments (DOE) and

Quality by Design (QbD) were used in this project to develop a suitable formulation for the

direct compression process of senna tablets. The newly formulated product demonstrated

benefits with regards to process time, manpower and machinery requirements. Furthermore,

product stability in high density polyethylene bottles exposed to accelerated conditions over

a six months period, demonstrated results that conforms to the limits specified in-house.

Nevertheless, a few challenges remain in terms of establishing the process speed in large

scale manufacturing and establishing the product shelf life upon completion of long term

stability studies. In spite of these challenges, implementing the direct compression process

to manufacture senna tablets instead of continuing with the wet granulation process

demonstrated sufficient benefits to adopt this new process for the long run.

VII

Table of Contents

Résumé ........................................................................................................................................ III Abstract ........................................................................................................................................ V Table of Contents ....................................................................................................................... VII List of Tables .............................................................................................................................. XI List of Figures .......................................................................................................................... XIII Brand Names & Abbreviations ................................................................................................. XV Declaration and Copyright ...................................................................................................... XVII Acknowledgment ..................................................................................................................... XIX Chapter I - Introduction ................................................................................................................ 1

Objectives & Methodologies ................................................................................................. 2

Quality by Design – Quality Target Product Profile ............................................................. 3

Appearance, Average Tablet Weight, Thickness & Diameter .............................................. 3 Tablet Hardness ..................................................................................................................... 4 Tablet Friability ..................................................................................................................... 4 Unbound Tablet Moisture ..................................................................................................... 5 Tablet Disintegration Time ................................................................................................... 6 Assay ..................................................................................................................................... 7

Quality by Design – Physicochemical Properties of Selected Excipients ............................ 7

Generally Recognised as Safe ............................................................................................... 9 Organic .................................................................................................................................. 9 Particle Size and True Density .............................................................................................. 9 Bulk Density and Tapped Density and % Compressibility ................................................. 11 Flowability .......................................................................................................................... 13 Mode of Compression (Plastic Deformation or Brittle Fracture) ....................................... 15 Water Solubility .................................................................................................................. 16 Percent Moisture in Raw Materials ..................................................................................... 17

Chapter II – First Set of Direct Compression Experiments ........................................................ 19

Objective and Experimental Design .................................................................................... 19

Master Formulae ................................................................................................................. 20

Manufacturing Process ........................................................................................................ 23

Tooling ................................................................................................................................ 25 Tablet Diameter (mm) ......................................................................................................... 26 Number of Station(s) and Rotor Speed (rpm) .................................................................... 26 Fill-o-matic Speed (rpm) ..................................................................................................... 26 Compression Force (kN) ..................................................................................................... 27 Tablet Cylindrical Height Main-pressure (mm) .................................................................. 27 Fill Cam and Fill Depth ....................................................................................................... 28

Results ................................................................................................................................. 28

Analysis of Data (Minitab analysis) .................................................................................... 30

VIII

Selection of Excipients for the Next Set of Direct Compression Experiments ................... 33

Chapter III – Second Set of Direct Compression Experiments ................................................... 35


Master Formulae .................................................................................................................. 36


Results.................................................................................................................................. 40

Analysis of Data (Minitab Analysis) ................................................................................... 42

Selection of Excipients for the Next Set of Direct Compression Experiments ................... 44

Chapter IV – Third Set of Direct Compression Experiments ..................................................... 45

Objectives and Experimental Design ................................................................................... 45

Master Formulae .................................................................................................................. 46


Results.................................................................................................................................. 52

Analysis of Data .................................................................................................................. 54

Selection of Excipients and Levels for the Next Set of Direct Compression Experiments ......................................................................................................................... 57

Chapter V – Fourth Set of Direct Compression Experiments ..................................................... 59

Objectives and Experimental Design ................................................................................... 59

Master Formulae .................................................................................................................. 60


Results.................................................................................................................................. 64

Analysis of Data .................................................................................................................. 66

Selection of Excipient Levels for the Next Set of Direct Compression Experiments ......... 68

Chapter VI – Fifth Set of Direct Compression Experiments ...................................................... 71

Objectives and Design ......................................................................................................... 71


Results.................................................................................................................................. 76

Analysis of Data .................................................................................................................. 78

Selection of Excipient Levels for the Next Set of Direct Compression Experiments ......... 81

Chapter VII – Sixth Set of Direct Compression Experiments .................................................... 83


Master Formulae .................................................................................................................. 84


IX

Results ................................................................................................................................. 90

Analysis of Data .................................................................................................................. 92

Selected Formulation ........................................................................................................... 94

Result Summary .................................................................................................................. 95

Chapter VIII – Accelerated Stability Study at 40°C and 75% Relative Humidity ..................... 97

Analysis of Data .................................................................................................................. 98

Assay, Container Closure and Shelf Life ............................................................................ 98

Chapter IX – Cost Analysis ...................................................................................................... 101 Time Analysis of the Wet Granulation Process ................................................................ 102 Time Analysis of the Direct Compression Process ........................................................... 102

Summary of Expenses ....................................................................................................... 105

Percent Yield ..................................................................................................................... 105

Additional Expenses .......................................................................................................... 106

Chapter X – Upcoming Challenges and Conclusion ................................................................ 107

Blending Time, Speed and Uniformity ............................................................................. 107

Compression Speed ........................................................................................................... 108

Tablet Hardness, Disintegration Time and Dissolution Rate with Adjusted Tablet Weight ............................................................................................................................... 108

Optimal Packaging Configuration and Long Term Stability Studies ............................... 108

Conclusion ......................................................................................................................... 109

References ................................................................................................................................. 111

XI

List of Tables Table 1: Summary of Steps Involved in Wet Granulation and Direct Compression Processes ................................................................................................................................. 1 Table 2: Quality Target Product Profile ................................................................................. 3 Table 3: Physicochemical Properties of Active Raw Material and Selected Excipients ........ 8 Table 4: Summary of Compressibility Index (%) of the Raw Materials .............................. 13 Table 5: Summary of Flowability of Raw Materials and ..................................................... 14 Table 6: The Plackett–Burman Matrix for Experimental Design With 11 factors and 2 Levels .................................................................................................................................... 20 Table 7: Senna 8.6 mg Tablets - Master Formulae – Lot # SK1301-8.6 Experiment #1 ..... 21 Table 8: Calculation of Available Weight Remaining per Tablet ........................................ 22 Table 9: Production Work Order for the First Set of Direct Compression Experiments ...... 23 Table 10: Critical Parameters Adjusted on the Fette Press 1200i ......................................... 25 Table 11: Senna Tooling Diameters ..................................................................................... 25 Table 12: Response Factors for the First Set of Direct Compression Experiments ............. 29 Table 13: Selected Excipients for the Second Set of Direct Compression Experiments ...... 34 Table 14: 5-Factorial Design with 2 Levels and 1 Center Point ........................................... 35 Table 15: Senna 8.6 mg Tablets - Master Formulae – Lot # SK1302-8.6 Experiment #13 . 36 Table 16: Calculation of Available Weight Remaining per Tablet ...................................... 37 Table 17: Production Work Order for the Second Set of Direct Compression Experiments .............................................................................................................................................. 38 Table 18: Response Factors for the Second Set of Direct Compression Experiments ......... 41 Table 19: Selected Excipients for the Third Set of Direct Compression Experiments ......... 44 Table 20: Experimental Design for the Third Set of Direct Compression Trials ................. 45 Table 21: Senna 8.6 mg Tablets - Master Formulae - Lot # SK1303-8.6 Experiment #30 .. 46 Table 22: Senna 8.6 mg Tablets - Master Formulae – Lot # SK1303-8.6 Experiment #31 . 47 Table 23: Senna 8.6 mg Tablets - Master Formulae – Lot # SK1303-8.6 Experiment #32 . 48 Table 24: Senna 8.6 mg Tablets - Master Formulae – Lot # SK1303-8.6 Experiment #33 . 49 Table 25: Production Work Order for the Third Set of Direct Compression Experiments .. 50 Table 26: Response Factors for the Third Set of Direct Compression Experiments ............ 53 Table 27: Best Experiments for Productivity ....................................................................... 54 Table 28: Best Experiments for Tablet Hardness ................................................................. 54 Table 29: Best Experiments for Tablet Friability ................................................................. 55 Table 30: Best Experiments for Disintegration Time ........................................................... 55 Table 31: Selected Excipients and Levels for the Fourth Set of Direct Compression Experiments .......................................................................................................................... 57 Table 32: Design of Experiments for the Fourth set of Direct Compression Trials. ............ 59 Table 33: Senna 8.6 mg Tablets – Master Formulae – Lot # SK1304-8.6 Experiment #34 60 Table 34: Senna 8.6 mg Tablets – Master Formulae – Lot # SK1304-8.6 Experiment #35 61 Table 35: Production Work Order for the Fourth Set of Direct Compression Experiments 62

XII

Table 36: Response Factors for the Fourth Set of Direct Compression Experiments .......... 65 Table 37: Best Hardness Responses ..................................................................................... 66 Table 38: Best and Worst Productivity Responses .............................................................. 66 Table 39: Best Friability Responses ..................................................................................... 67 Table 40: Worst Disintegration Time Response .................................................................. 67 Table 41: Selected Excipients and Levels for the Fifth Set of Direct Compression Trials .. 69 Table 42: Design of Experiments for the Fifth Set of Direct Compression Trials ............... 71 Table 43: Senna 8.6 mg Tablets – Master Formulae – Lot # SK1305-8.6 Experiment #36 72 Table 44: Senna 8.6 mg Tablets – Master Formulae – Lot # SK1305-8.6 Experiment #37 73 Table 45: Production Work Order for the Fifth Set of Direct Compression Experiments ... 74 Table 46: Response Factors for the Fifth Set of Direct Compression Experiments ............ 77 Table 47: Best and Worst Productivity Responses .............................................................. 78 Table 48: Best Hardness Responses ..................................................................................... 78 Table 49: Best Friability Responses ..................................................................................... 79 Table 50: Worst Disintegration Time Responses ................................................................. 79 Table 51: Selected Excipients and Levels for the Sixth Set of Direct Compression Trials . 81 Table 52: Design of Experiments for the Sixth Set of Direct Compression Trials .............. 83 Table 53: Senna 8.6 mg Tablets – Master Formulae – Lot # SK1306-8.6 Experiment #38 84 Table 54: Senna 8.6 mg Tablets – Master Formulae – Lot # SK1306-8.6 Experiment #39 85 Table 55: Senna 8.6 mg Tablets – Master Formulae – Lot # SK1306-8.6 Experiment #40 86 Table 56: Senna 8.6mg Tablets – Master Formulae – Lot # SK1306-8.6 Experiment #41 . 87 Table 57: Production Work Order for the Sixth set of Direct Compression Experiments ... 88 Table 58: Response Factors for the Sixth Set of Direct Compression Experiments ............ 91 Table 59: Selected Formulation for Senna 8.6 mg Tablets .................................................. 94 Table 60: Results Summary of the New Product lot# SK1306-8.6 (experiment #40). ........ 95 Table 61: Result Summary of Accelerated Stability Study at 40°C/75% RH on Optimal Formulation (SK1306-8.6 experiment #40) ......................................................................... 97 Table 62: % Assay of the New Product and Existing Product Over a Six Months Exposure Period at 40°C/75% RH in Their Respective Container Closures ....................................... 99 Table 63: Compared Formulation, Process, Labour and Machinery Expenses for Wet Granulation and Direct Compression Processes ................................................................ 105

XIII

List of Figures

Figure 1: Degradation Pathways of the Active Principle ....................................................... 5 Figure 2: Particle Size Distributions of Pre-gelatinised Starch and Senna ........................... 10 Figure 3: Pareto Chart of the Effects for Tablets Produced .................................................. 30 Figure 4: Main Effects Plot for Tablets Produced ............................................................... 30 Figure 5: Pareto Chart of the Effects for Tablet Hardness ................................................... 30 Figure 6: Main Effects Plot for Tablets Hardness ............................................................... 30 Figure 7: Main Effects Plot for Flowability ......................................................................... 31 Figure 8: Main Effects Plot for Tablets Disintegration ....................................................... 32 Figure 9: Main Effects Plot for Hardness ............................................................................. 42 Figure 10: Main Effects Plot for Friability .......................................................................... 42 Figure 11: Main Effects Plot for Disintegration Time .......................................................... 43 Figure 12: Main Effects Plot for Average Tablet Weight ................................................... 43 Figure 13: % Assay of the New Product and Existing Product Over a Six Months Exposure Period at 40°C/75% RH in Their Respective Container Closures ........................................ 99 Figure 14: Flow Chart of the Wet Granulation and Direct Compression Processes with Time Requirements ............................................................................................................. 101 Figure 15: Direct Compression Process Mapping .............................................................. 107 Figure 16: Appearance of the current senna tablet (left) and the newly formulated senna tablet (right). ....................................................................................................................... 110

XV

Brand Names & Abbreviations

Aqualon EC-N7: Ethylcellulose Avicel PH102: Brand Name for Microcrystalline Cellulose °C: Degrees Celcius DOE: Design of Experiments DT: Disintegration Time EC: Ethylcellulose Ethocel Std 4 Premium: Brand Name for Ethylcellulose FDA: Food and Drug Administration g: grams GRAS: Generally Recognized As Safe HDPE: High Density Polyethylene HEC: Hydroxyethyl Cellulose HPC: Hydroxypropyl Cellulose HPC-SSL-SFP: Brand Name for Hydroxypropyl Cellulose HPLC: High Performance Liquid Chromatography HPMC: Hydroxypropylmethyl Cellulose ICH: International Conference of Harmonisation kN: Kilonewton Kp: Kilopond L: Liters MC: Methylcellulose MCC: Microcrystalline Cellulose Methocel A15 Premium LV: Brand Name for Methylcellulose Methocel K3: Brand Name for Hydroxypropylmethyl Cellulose mg: milligrams mm: millimeters Natrosol 250: Brand Name for Hydroxyethyl Cellulose N/Ap: Not Applicable N/Av: Not Available Neosorb P150DC: Brand Name for Sorbitol NF: National Formulary NMT: Not More Than PGS: Pre-gelatinised Starch QbD: Quality by Design QTPP: Quality Target Product Profile RH: Relative Humidity RPM: Rotation per Minute RSD: Relative Standard Deviation Starch 1500: Pre-gelatinised Starch µm: Micrometers USP: United States Pharmacopeia

XVII

Declaration and Copyright

I hereby declare that no portion of the work referred to in the thesis has been submitted in

support of an application for another degree or qualification of this or any other university

or other institute of learning.

Copyright in text of this thesis rests with the author. Copies (by any process) either in full,

or of extracts, may be made only in accordance with instructions given by the author.

Details may be obtained from the author. This page must form part of such copies made.

Further copies (by any person) made in accordance with such instruction may not be made

without the written permission of the author.

The ownership of intellectual property rights, which may be described in this thesis, is

vested in the Organisation, subject to any prior agreement to the contrary, and may not be

made available for use by third parties without the written permission of the Organisation,

which will prescribe the terms and conditions of any such agreement.

Further information on the conditions under which disclosures and exploitation may take

place is available from the Organisation.

XIX

Acknowledgment

I would like to thank my directors of research Ricardo Vargas, Julianna Juhász and Thérèse

DiPaolo-Chênevert for giving me the opportunity to pursue this Master degree project

within my Organisation as part of the Master of pharmaceutical sciences program at the

Faculty of Pharmacy of Laval University.

I would also like to thank Ashok Goundalkar, Ricardo Vargas and Jangshin An for

providing extensive training on manufacturing equipment and analytical methods involved

in the manufacturing and testing of senna tablets. In addition, Shauna Callahan, Dale

Roberts, Jude Gonsalves and Ricardo Vargas contributed a great deal of their knowledge

and expertise for the cost analysis of the direct compression process. In terms of reference

documents, I would like to thank Paul Baker and Christina Lewis for providing critical

information on existing manufacturing and stability data used for comparison with the

newly formulated product.

At last, I would like to thank my family and friends for supporting me through this

challenging journey of being a full-time student and a full-time employee. Your

encouragement and believe in my potential gave me the strength to successfully complete

this Master degree project.

Thanks everyone for contributing your valuable time to the accomplishment of this project.

1

Chapter I - Introduction

The natural substance used in this research project – senna – is derived from the dried pods

of a plant originating from Alexandria named Cassia acutifolia, known for centuries for the

treatment of occasional constipation (1). The therapeutic effect of senna is produced

through irritation of the inner layer of the gastro-intestinal tract, thereby promoting

peristaltism (2). Absorption of senna through the cells demonstrates linear and positive

results shown by Waltenberger et al. (3). However, absorption is not absolutely required as

the therapeutic effect is produced by the sole presence of senna in the gastro-intestinal tract

(2).

In today’s pharmaceutical industries, senna is used to manufacture an array of products

such as syrup, coated tablets and uncoated tablets which are widely used in North America

and Europe (4) (5). The uncoated tablets discussed in this project have been manufactured

for more than fifty years by mean of wet granulation, a process that is rather tedious and

time consuming. The main benefit of changing the manufacturing process from wet

granulation to direct compression is to generate substantial savings with a significantly

shorter process. As described in Table 1, wet granulation (4 hours), drying of granules (13

hours) and assay of granules (120 hours) are the three major limiting steps of the wet

granulation process which are not required in direct compression.

Table 1: Summary of Steps Involved in Wet Granulation and Direct Compression Processes

Wet Granulation Direct Compression 1) Blend Ingredients (*) 2) Wet Granulation (4 hours) 3) Dry Granules (13 hours) 4) Sieve Granules (*) 5) Mix Granules/Lubricant (*) 6) Assay Granules (120 hours) 7) Compress Granules (**)

1) Blend Ingredients (*) 2) Sieve Blend (*) 3) Mix Blend/Lubricant (*) 4) Compress Blend (**)

* Similar time requirement for both processes. ** Half time requirement for the wet granulation process. (Refer to Chapter IX – Cost Analysis for further details).

2

Another benefit of the direct compression process is that it can potentially enhance the

stability of the finished product. Since no water and reduced amount of heat are involved in

the manufacturing process, hydrolysis and oxidation of the active principle could be

reduced as a result (6). Stability studies were performed and presented in Chapter VIII to

confirm this hypothesis.

In spite of these benefits, the direct compression process involves considerable challenges.

Senna is not compressible and occupies approximately 80% of the tablet formulation,

resulting in limited space to introduce binding excipients while maintaining the original

weight and size of the tablets. Furthermore, the current formulation which contains corn

starch, magnesium stearate and microcrystalline cellulose (2) is not compressible through

direct compression. Therefore, a new formulation must be innovated.

Objectives & Methodologies

The objective of this project is to manufacture by direct compression a natural product that

is equivalent to that of the existing product. To achieve this objective, elements of the

Quality by Design methodology such as “Quality Target Product Profile” (QTPP) and

“Design of Experiments” were used to develop a product that fulfills regulatory

requirements. The QbD methodology is a systematic approach that incorporates the use of

these tools (QTPP and DOE) to analyse the information generated at every step of

development and to ultimately create a quality product (7). In this research project, six sets

of experiments were performed using DOE methodology. Different types of DOE such as

Plackett-Burman design, five-factorial design and traditional factorial designs were used at

different stage of the project to develop an optimal formulation for the direct compression

process. The Plackett-Burman design was first used to screen the factors (excipients) that

would promote feasibility of the process. This type of DOE allows the formulator to

analyse several excipients and identify the excipients that have the most significant effects

on the desired responses (tablet properties), while performing a minimum number of

experiments (8). Thereafter, the five-factorial design was used to further analyse five

selected excipients. This type of DOE allows the formulator to specifically analyse the

3

effect of the five selected excipients on the desired responses, while performing a minimum

number of experiments (9). Once desired responses were identified with selected

excipients, traditional factorial DOE models with two levels were used to pursue

optimisation of excipients levels in the formulation and achieve responses that meet all

QTPP requirements (10).

Quality by Design – Quality Target Product Profile

The quality target product profile was established based on critical tablet parameters given

in the specifications of the existing product (11). The specifications were used as guides to

create the new product. Individual tablet parameters and methods of measurement were

further discussed below.

Table 2: Quality Target Product Profile

Test Method Specifications

Assay High Performance Liquid Chromatography (HPLC) (12) 90 – 110%

Appearance Visual “S” on one side plain on the other side Average Tablet

Weight Analytical balance Target: 260 mg

Not more than (NMT) 1 unit outside 7.5% range Thickness Thickness gauge [4.0 – 5.2] mm Diameter 8.8mm +/- 0.4 mm

Hardness Current *USP Average: [3.5 – 7.0] Kiloponds (Kp) Individuals: [3.0 – 8.0] Kp Friability Current USP NMT 1%

Tablet Moisture Moisture balance (12) NMT 6%

Shelf Life Stability Study as per International

Conference of Harmonisation (ICH) guideslines (13)

2 years

Disintegration Time Current USP NMT 45 min.

* United States Pharmacopeia (USP)

Appearance, Average Tablet Weight, Thickness & Diameter

Appearance, average tablet weight, thickness and diameter were measured in order to create

a new product with similar image and size as that of the existing product. Appearance and

4

size are extremely important for patients’ recognition and compliance. Therefore, it is ideal

to maintain these tablet parameters in order to minimize the risk of impacting the patient

compliance negatively by introducing significant product changes.

Appearance is recorded based on visual observations of the tablets. Average tablet weight is

calculated on twenty units individually weight on an analytical balance. Thickness and

diameter are measured with a Mituyoto thickness gauge.

Tablet Hardness

Adequate tablet hardness is required to ensure sufficient strength to resist mechanical

impacts during manufacturing and packaging processes. Hardness is defined by the force

required to cause the tablets to fail (i.e. break) in a specific plane. Tablets are placed

between two platens, one which moves to apply sufficient force to the tablets to cause

fracture (4).

The average tablet hardness is measured on twenty units and reported in Kilopond. The

kilopond was the selected unit for hardness which defines the force required to break the

tablet. It is important to mention that it is not necessary to achieve similar tablet hardness

between the reference and new product, as long as it maintains its integrity and meets other

critical requirements such as tablet friability and disintegration time. Nevertheless, tablets

hardness significantly lower than 3.0 Kp should be closely monitored for friability and

reversibly, tablets hardness significantly higher than 8.0 Kp should be tested for

disintegration time to ensure conformity with the specifications.

Tablet Friability

The friability test determines the ability of tablets to withstand mechanical stress.

Resistance to chipping and surface abrasion are measured by tumbling the tablets one

hundred times in a rotating cylinder. The test passes USP criteria if the percent weight loss

5

is not more than 1% after one hundred rotations of the cylinder (4). Result below 1% must

also be achieved for the new products in order to conform to the specifications given in

USP chapter . In addition, no broken tablet, capping, lamination, splitting should be

observed for the test to pass (4).

Unbound Tablet Moisture

Unbound tablet moisture is defined by the amount of unbound water molecules present in

the finished product. This parameter is important to monitor because moisture can trigger

degradation of the active principle due to hydrolysis. If significant degradation occurs due

to moisture, the product will lose potency and the shelf life can be impacted as a result. The

scheme below illustrates the theoretical impact of moisture on the finished product.

Scheme 1: Impact of Moisture on the Finished Product

↑ % Moisture ↑ Hydrolysis of Active Principle ↓ Assay Value ↓ Product Shelf Life

Figure 1 illustrates the degradation pathways of the active principle. The first pathway is a

result of oxidation and the second pathway is a result of hydrolysis.

Figure 1: Degradation Pathways of the Active Principle Legend: A. Hydrolysis. X. Further Hydrolysis. B. Oxidation. (6)

6

Based on the specifications for moisture content (NMT 6%) and assay (90 – 110%), a shelf

life of two years was established for the existing product (11). With the direct compression

process, an extended shelf life is anticipated since no water is involved in the process as

opposed to the wet granulation process. However, if the new product stability cannot be

enhanced by the direct compression process, a shelf life of two years should be achieved to

match the target product profile.

To measure tablet moisture, 5 g of powdered tablets was heated to 105°C in a Metler

moisture balance. The Metler moisture balance is equipped with a halogen lamp which

heats the powdered tablets to 105°C and evaporates all water molecules from the 5 g

sample. The instrument is set to automatically generate the moisture content value when all

water molecules have been evaporated from the powdered sample. Results equal to or

below 6% must also be achieved for the new product in order to conform to the

specification.

Tablet Disintegration Time

The disintegration test determines whether the tablets disintegrate within the prescribed

time when placed in a specified liquid media. Complete disintegration is achieved when a

soft mass having no palpable firm core is obtained in the liquid media. The apparatus used

for this test is described in current USP chapter (4). Adequate disintegration time is

extremely important to achieve in order to release the active particles from the tablet matrix

at the targeted site of the gastro-intestinal tract.

The disintegration test was performed using a disintegration tester where six tablets were

immersed in purified water at 37.0°C +/- 0.2°C to simulate in vivo conditions. The test was

performed for 45 minutes or until complete disintegration of the six tablets was achieved.

7

Assay

The assay test ensures that the potency of the finished product reflects the label claim. As

per the specification described in Table 2 : Quality Target Product Profile, the potency

of the drug product should not exceed or fall below the allowable limits (90 – 110%) to

ensure efficacy of the product and patient’s safety.

The assay test is performed on a pooled sample of tablets, dissolved in a known volume of

extraction solution. The samples are analysed by HPLC on a reverse phase system (14).

The system detects how much active substance is present in each sample and the assay

values are compared to the specifications (11).

Quality by Design – Physicochemical Properties of Selected Excipients

In order to create a product that is equivalent to the existing product, selected excipients

must have specific physicochemical properties which are listed in Table 3 below. The

number of selected excipients was based on the design of experiments used for the first set

of direct compression trials – a design with eleven factors which is further discussed in

Chapter II – First Set of Direct Compression Experiments.

Since the active raw material consists of approximately 80% of the formulation and plays a

significant role in the feasibility of the manufacturing process, it was also analysed for the

same quality attributes as the excipients. The active raw material data were used as a

reference point to define which excipients could potentially enhance the formulation and

contribute to the feasibility of the direct compression process. Individual critical quality

attributes and methods of measurement were further described below.

8

Table 3: Physicochemical Properties of Active Raw Material and Selected Excipients

Active Excipients

Physicochemical Properties Senna ** PGS **

MCC Lactose Sorbitol **

EC1 **

HEC **

HPC **MC **

HPMC **

EC2 Alginic

acid

“Genarally Recognized As Safe” (GRAS)

National Formulary (NF) grade Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Organic Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

*Particle size (µm)

****D(0.1) 8.4 13.9 34.6 19.8 86.1 41.7 39.9 6.9 26.0 23.1 33.8 11.0 ****D(0.5) 50.5 51.2 116.5 45.5 182.0 155.9 90.8 20.3 73.6 67.1 212.4 47.3 ****D(0.9) 165.1 153.0 235.7 113.5 322.8 372.0 190.5 42.0 242.5 161.0 483.0 127.4

True Density (g/ml) Not Available (N/A) 1.52 1.59 1.55 1.51 1.14 1.38 1.22 1.34 1.33 1.14 1.60

*Bulk Density (g/ml) 0.38 0.66 0.36 0.65 0.41 0.49 0.54 0.33 0.33 0.43 0.45 0.44

*Tapped Density (g/ml) 0.58 0.83 0.45 0.71 0.47 0.52 0.66 0.44 0.44 0.53 0.59 0.71

*Compressibility (%) (Flow Character) (15)

35.0 (V)

21.1 (P)

20.0 (F)

9.1 (E)

13.0 (G)

5.8 (E)

18.4 (F)

24.8 (P)

24.8 (P)

18.97 (F)

23.8 (P)

38.4 (VV)

*Flowability (mm) ***(Flow Character)

24 (F)

20 (F)

16 (G)

16 (G)

4 (E)

4 (E)

30 (P)

28 (P)

24 (F)

20 (F)

26 (P)

26 (P)

Mode of Compression (P = Plastic deformation

B = Brittle fracture) P P P B B P P P P P P P

Water Solubility Yes Yes No Yes Yes No Yes Yes Disperse Yes No No *% Moisture 7.32 4.04 2.82 0.44 0.48 1.12 4.63 1.20 2.09 3.11 0.87 6.08

*Test performed in-house. **Pre-gelatinised Starch (PGS), Microcrystalline Cellulose (MCC), Ethylcellulose (EC), Hydroxyethylcellulose (HEC), Hydroxypropylcellulose (HPC), Methylcellulose (MC), Hydroxypropylmethylcellulose (HPMC). EC1, EC2 are both ethylcellulose compounds acquired from different manufacturers. ***Flow character was established based on experimental observations; E = Excellent, G = Good, F = Fair, P = Passable, V = Very poor, VV = Very, very poor. ****D(0.1), D(0.5), D(0.9) are described as the percentages (10%, 50% and 90%, respectively) of particles that corresponds to a specific size or lower. For example, 10% of senna particles are lower or equal to 8.4 µm.

9

Generally Recognised as Safe

It was preferable to select excipients that are generally recognized as safe by the Food and

Drug Administration (FDA) (16). Therefore, all selected excipients listed in Table 3 meet

this criterion (17).

Organic

Organic excipients were selected for this formulation to favor compatibility with the active

raw material (a natural plant extract). Compatibility studies between active raw material

and excipients will not be performed in this project. However, accelerate stability studies

on the final formulation will confirm if significant degradation of the active principle

occurs at extreme conditions. Compatibility issues could be hypothesised if the drug

significantly losses potency after exposure to extreme conditions. The stability profile of

the new product was established and presented in Chapter VIII – Accelerated Stability

Study at 40°C and 75% Relative Humidity.

Particle Size and True Density

In a direct compression process, the particle size and true density of excipients and active

raw material have a great impact on the uniformity of the blend. To achieve blend

uniformity, no differential segregation should occur due to various densities and sizes of

the powder materials. Differential segregation occurs when small high density particles fall

to the bottom by gravity and large low density particles stay on top. The reverse can be

observed when large high density particles fall to the bottom and small low density

particles stay on top.

Mixing time can also significantly impact blend uniformity. In this project, adequate

mixing time was defined by visual observation of the uniformity of color in the blend. A

light brown color was obtained after sufficient mixing of the active raw material (brown in

color) with the excipients (white in color). Uniformity of data (% relative standard

10

deviation (RSD)) obtained during analytical testing of the new product was also used as an

indicator of adequate blend uniformity.

The particle size for each excipient and active raw material were measured with the

Malvern Mastersizer 2000 whereas true density values were obtained from the literature

(17). Most excipients in Table 3 did not show apparent differences of particle size in

comparison to the active raw material. The graph below is an example of particles size

distribution for active raw material and pre-gelatinised starch.

Figure 2: Particle Size Distributions of Pre-gelatinised Starch and Senna

For blend uniformity and direct compression purposes, it is ideal yet not absolutely

necessary to have excipients and active raw material with similar particle sizes and density.

However, apparent differences should be justified. For example, relatively large particle

size was observed for EC2 and relatively small particle size was observed for HPC. The

smaller particles are not much of a concern since it is often desired for a binder to have this

characteristic. In this manner, the small binder particles can adhere to and surround the

active raw material particles, creating several points of contacts to bind the particles

11

together during compression. On the other hand, uniformity of blend and finished product

should be closely monitored when using relatively larger particles size material such as

EC2 (refer to Table 3 for particle size results).

Bulk Density and Tapped Density and % Compressibility

Bulk density is determined by measuring the volume of a known mass of powder sample.

The tapped density is achieved by mechanically tapping the measuring device containing

the powder sample and measuring the final volume (4). The bulk volume compared to the

tapped volume indicates the amount of interparticular void space in a powder sample. The

compressibility index which derives from this concept is calculated as follows.

Compressibility Index = (100(Vo – Vf))/Vo

Where Vo is the volume of bulk powder sample and Vf is the tapped volume of the same

powder sample.

The compressibility index is a measure of the relative importance of interparticulate

interactions. “In a free-flowing powder, such interactions are generally less significant, and

the bulk and tapped densities will be closer in value. For poorer flowing materials, there are

frequently greater interparticle interactions, and greater difference between the bulk and

tapped densities will be observed” (4).

In large scale manufacturing, the bulk density value can be used to define the minimum and

maximum batch size that can be produced in the manufacturing equipment. As such, the

capacity of the equipment multiplied by its working volume and multiplied by the bulk

density can give an estimate of the batch size that can be produced in-house (18).

12

Smin = V x 0.5 x BD

Smax = V x 0.8 x BD

Where S is the batch size in kilograms, V is the equipment capacity in liters, 0.5 and 0.8 are

the minimum and maximum working volume, respectively, and BD is the bulk density of

the blend in g/ml (18).

Bulk density values were obtained by measuring the volume of 100g samples in a

graduated cylinder. Details of the method are described in current USP chapter (4).

Each tapped density test was performed twice for 1250 taps in a tap density tester. Not

more than 2% difference in volume was observed between the first and second sets of 1250

taps. Three of the selected excipients in Table 3 showed low differences between bulk and

tapped density values as well as low % compressibility values (i.e. lactose, sorbitol, EC2),

which indicates relative low interparticulate interactions within the intrinsic powder

sample. The inverse relationship was observed with the other excipients and active raw

material, which exhibit higher difference between bulk and tapped density and higher %

compressibility values.

However, high % compressibility values do not indicate that the powder materials have a

better ability to form tablets. It is solely an indication of its porosity. Nevertheless, powder

porosity is often used in the pharmaceutical industry to characterise the flow of powders.

As shown in Table 4, given ranges of % compressibility correspond to specific flow

characters (15). The majority of excipients selected for the first set of direct compression

experiments showed passable to excellent flow characters in order to compensate for the

“very poor (V)” flow characteristic of senna.

13

Table 4: Summary of Compressibility Index (%) of the Raw Materials

Compressibility Index (%) Flow Character Raw materials

< 15 Excellent – Good Lactose, Sorbitol, EC1

16 – 20 Fair PGS, MCC, HEC, HPMC

21 – 25 Passable HPC, MC, EC2

26 to > 38 Poor to Very, very poor Senna, Alginic Acid

Even though these % compressibility and flowability values were used as preliminary

guides in this project, many more characteristics are involved in the ability of a powder to

flow which are discussed below.

Flowability

Flowability is the property of a powder to flow evenly under the action of gravity and other

forces, which depends largely on cohesion forces, adhesion forces, particle density and

particle size and shape (19). It is important for a powder blend to flow properly in the

equipment in order to obtain adequate productivity in large scale manufacturing.

High cohesive forces in a powder material cause particles to stick together and hinder free

flow in the equipment as a result. Similarly, high adhesive forces cause particles to stick to

the hopper, resulting in poor recovery of materials. On the other hand, low cohesive and

adhesive forces in powder materials can cause particles to flow very rapidly, resulting in

overflow of the die cavities. Therefore, materials with very high or very low

cohesiveness/adhesiveness characteristics may not be optimal for direct compression (15).

The flowability of individual excipient, active raw material and blends were assessed on 50

g samples using the Flodex Hanson Research instrument (19). The Flodex is designed to

measure the smallest orifice in millimeters through which a 50g sample can freely flow

three times under the force of gravity. This occurs when the force of gravity is greater than

the forces of cohesion and adhesion in the sample. The bigger the orifice, the more likely

14

the material will flow under the force of gravity. The Flodex is a useful tool to anticipate

the flowability behaviour of the materials in large scale manufacturing.

In Table 5 below, excipients and active raw material were categorised in four ranges of

flowability. For this project, it was ideal to select excipients with diverse flowability index

since it is not yet understood whether a certain range would favorably or negatively impact

the direct compression process. It was further defined in compression trials that a mixture

of excipients that exhibits excellent to passable flow was beneficial for this formulation.

Table 5: Summary of Flowability of Raw Materials and

Blends Used in the First Set of Direct Compression Experiments

Flowability

*(Flow Character) Raw materials **Blends

4 – 10 (Excellent) Sorbitol, EC1 -

11 – 19 (Good) Lactose, MCC -

20 – 25 (Fair) Senna, PGS, MC, HPMC 1, 4, 6, 7, 10,

11, 12

26 – 30 (Passable) HEC, HPC, EC2, Alginic Acid 2, 3, 5, 8, 9

*Flow character were categorised in four ranges based on experimental observations. ** Blends 1 to 12 were prepared as per the production work order shown in Table 9 (steps 1, 2, 3) and the flowability of each blend was measured prior to compression into tablets.

From the above results, it can be concluded that the active material had great impact on

blend flowability since all blend flowability values ranged from 22 to 26 and the active raw

material (senna) flowability value was 24. From experience with direct compression, it was

previously observed that blend flowability values between 20 and 30 were ideal for this

process.

15

Mode of Compression (Plastic Deformation or Brittle Fracture)

The mode of compression of a powder material may occur through plastic deformation

and/or brittle fracture mechanisms. Plastic deformation occurs when the powder particles

bend or deform like plastic material during compression. Plastic deforming material creates

relatively weaker bonds within the tablet since bonding is achieved through intermolecular

forces (i.e. ionic bonding, hydrogen bonding, dipole-dipole forces and dispersion forces).

On the other hand, brittle fracture occurs through breakage and consolidation of bonding

within the molecules. In this mechanism, the molecules come relatively closer to one

another and covalent bonds may form, which could explain the higher bonding strength

created with this type of material (20).

The mode of compression of a powder is an important attribute that allows the formulator

to better understand the type of bonding that occurs with the tablet and anticipate its

mechanical strength as a result. The mode of compression of powders can be identified

with the Heckle relationship – by plotting the inverse natural log of tablet porosity (ln (1/ε))

versus compression pressure (21). Alternatively, the equation can be further described as

follows.

ln(1/(1−D)) = PK+A

Where P is the pressure, D is the relative density of the tablet and K and A are constants

(slope and intercept, respectively). (22)

The compression behaviour of the powder material is usually defined by the slope of the

Heckel plot, where a slight change in tablet porosity over a wide range of compression

pressures generally characterize plastic deforming materials, whereas a relatively sharp

change in porosity over a wide range of pressures generally characterize materials that

compress through brittle fracture mechanism. This sharp change in porosity may explain

the closeness of molecules within the tablet, resulting in higher breaking force (20).

16

The mode of compression of several powders is reported in the literature (16). Cellulose

compounds are usually plastic deforming when compressed whereas sugars usually

compress by brittle fracture mechanism. In this project, most selected excipients were

plastic deforming in nature (binders) except for lactose and sorbitol which deform by brittle

fractures. A variety of plastic deforming binders were selected because they could

potentially act as adhesive around the active raw material particles whereas lactose and

sorbitol could potentially provide additional tablet strength. Other sugars compounds were

avoided for this formulation due to their calorie content which can be a concern in the

elderly population who consume the product. In Chapter II – First Set of Direct

Compression Experiments and Chapter III – Second Set of Direct Compression

Experiments, it was further defined if the mode of compression of selected excipients had

a significant impact on the breaking force and integrity of the tablet.

Water Solubility

The water solubility of the active raw material and excipients plays an important role in the

disintegration and dissolution of the active principle – which directly impact on the

therapeutic effect. In terms of disintegration, the excipients and active raw material can

control the rate at which the water penetrates into the tablet in order to break it into

aggregates and release the active principle into the gastro-intestinal tract. Different types of

excipients exist on the market to serve this purpose (disintegrant), whereas some excipients

may passively assist tablet disintegration (diluent). Disintegrants have specific modes of

action by which they break down the tablet into aggregates. A few examples of these

mechanisms are capillary hydration (wicking), hydration and swelling, deformation, gas

production, exothermic heat and enzymatic action. For example, capillary hydration may

occur with hydrophilic and porous excipients, which enhance water penetration into the

hydrophilic network and weakens intermolecular bonds, resulting in breakage of the tablet

into fine particles. On the other hand, the swelling mechanism often occurs with insoluble

crosslinked excipients, which absorb water and swells to several times their original

volume. The osmotic pressure exerted on the neighboring particles eventually leads to bond

17

rupture (23). Capillary hydration and swelling mechanisms will be further discussed in this

project as some of the selected excipients have these types of disintegrating properties.

In terms of dissolution, the process initially relies on disintegration time, then on the

solubility of the active principle. Poorly soluble active principle generally shows slower

dissolution rate and incomplete extent of dissolution, which can be challenging when fast

therapeutic effect is desired. However, disintegrating and solubilising agents may be added

to the formulation to enhance disintegration time and dissolution of the active principle.

Water solubility can be measured by determination of the saturation point when adding

known quantities of powder material into a known volume of water. The powder material

can then be classified according to the solubility table given in the current USP (4). Solubility of the delivery system as a whole can also be assessed through in-

vitro disintegration and dissolution experiments. Since it is already known that senna is

sparingly soluble in water (24), disintegration and dissolution experiments were performed

to assess the solubility of the delivery system as a whole. As shown in Table 3, a mixture

of soluble and non-soluble excipients was selected since the impact of excipients solubility

on disintegration time and dissolution is not yet known for this formulation.

Percent Moisture in Raw Materials

Percent unbound moisture in the raw materials is a critical attribute that can potentially

affect the degradation of the active principle (refer to Scheme 1). However, degradation

caused by excipient’s unbound moisture was not a tremendous concern for this formulation

since all moisture values observed in Table 3 were lower than that of senna. Furthermore,

no water was involved in the direct compression process.

In terms of bonding, a positive aspect of the presence of free water molecules in the blend

could potentially be seen when the water molecules promote formation of solid bridges

within the particles during compression. Tablet integrity and breaking force could be

18

enhanced as a result. On the other hand, there was a realistic concern for relatively higher

excipient’s moisture that could cause cohesiveness in the blend and affect flowability as a

result.

Considering the above pros and cons of percent moisture in the raw materials, selected

excipients displayed a wide range of moisture values [0.44% - 6.08%], while remaining

below the percent moisture of the active raw material.

Percent moisture in raw materials was determined in a Metler moisture balance equipped

with a halogen lamp. 5g of powder sample was placed in the moisture balance and heated

to 105°C in order to evaporate all water molecules from the sample.

The following chapters will help us understand how these physicochemical properties

impact tablet integrity, breaking force and stability.

19

Chapter II – First Set of Direct Compression Experiments

Objective and Experimental Design

The objective of the first set of experiments was to identify potential excipients that would

promote feasibility of the direct compression process and produce tablets with similar

hardness and disintegration time as that of the existing product. The following

mathematical model developped by Plackett-Burman was used to identify potential

excipients for the formulation. The model constitutes of twelve experiments conducted in

the same manner using eleven excipients randomised as shown in Table 6. The presence of

excipients is identified by the “+” sign and the absence of excipients is identified by the “-”

sign (10).

The Plackett-Burman design is a powerful tool to analyse the impact of several excipients

while performing a minimum number of experiments. For example, eleven excipients

assessed at two levels would result in performing 121 experiments with the traditional

design of experiments model, where all possible combinations of excipients are attempted

(e.g. ∑(Xi )level = (11factors)2levels = 121 experiments). On the other hand, the Plackett-

Burman design only requires twelve experiments to generate sufficient data to identify

excipients that have significant effect(s) on the selected response factors. The effect of each

excipient can be assessed when combinations “-/-”, “+/-”, “-/+”, “+/+” of two excipients

appears three times in the matrix. The experiments were conducted by preparing 50g blends

containing the excipients identified by the “+” sign (8).

20

Table 6: The Plackett–Burman Matrix for Experimental Design With 11 factors and 2 Levels

Experiment # X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11

1 + − − − + + + − + + −

2 + + − + − − − + + + −

3 − + + + − + + − + − −

4 + − + − − − + + + − +

5 − + − − − + + + − + +

6 + + − + + − + − − − +

7 − − − − − − − − − − −

8 + − + + − + − − − + +

9 − − + + + − + + − + −

10 − − − + + + − + + − +

11 − + + − + − − − + + +

12 + + + − + + − + − − −

Legend: X1 = PGS, X2 = MCC, X3 = Lactose, X4 = Sorbitol, X5 = EC1, X6 = HEC, X7 = HPC, X8 = MC, X9 = HPMC, X10 = EC2, X11 = Alginic acid.

Master Formulae

Table 7 is an example of formulation for experiment #1 containing excipients X1, X5, X6,

X7, X9, X10.

21

Table 7: Senna 8.6 mg Tablets - Master Formulae – Lot # SK1301-8.6 Experiment #1

Senna 8.6 mg Tablets - Master Formulae - Experiment #1 Batch size: 52 g (200 tablets) Batch No.: SK1301-8.6 (Experiment #1)

Master Formulae - Experiment #1 Ingredients % Qty/Unit

(mg/tablet) Standard Qty

(g/batch) *Senna (40.0 mg/g,

8.6 mg/tablet) 82.6923 215.0000 43.0000

X1 = PGS 2.7180 7.0668 1.4134 X5 = EC1 2.7180 7.0668 1.4134 X6 = HEC 2.7180 7.0668 1.4134 X7 = HPC 2.7180 7.0668 1.4134

X9 = HPMC 2.7180 7.0668 1.4134 X10 = EC2 2.7180 7.0668 1.4134

Magnesium stearate 1.0000 2.6000 0.5200 Total 100.0000 260.0000 52.0000

* 40.0 mg/g is the lowest acceptable limit for assay of the active principle. This value was selected for formulating the new product since it simulates the worst case scenario for direct compression.

The same excel template was used to present the formulations for the twelve experiments.

The quantities of ingredients added to the blends were calculated to produce 260 mg tablets

containing 8.6 mg of active principle. The quantity of active material added was based on

the lowest acceptable assay value – 40 mg/g. This value allowed us to simulate the worst

case scenario for direct compression since a lower purity value lead to adding more non-

compressible material (senna) to the formulation. In terms of lubricant level (magnesium

stearate), 1% was selected based on the lowest recommended level generally used for direct

compression as per the Handbook of Pharmaceutical Excipients (17).

The function of the lubricant is to prevent the blend from adhering to the walls of the

hopper and tooling by its hydrophobic and non-polar nature. However, the lubricant can

also interfere with the bondings of particules in the blend by repelling them from one

22

another. For this reason, it is important to introduce the lubricant in moderate quantity at

the initial stage of formulation development and to mix it separately for a minimal period of

two to three minutes prior to compression in order to minimize intraparticular interactions

of the lubricant within the blend.

The remaining weight available to introduce the excipients was determined after

establishing the proper amounts of active material and lubricant to introduce in the

formulation. Table 8 illustrates the calculation used to determine the available weight

remaining per tablet.

Table 8: Calculation of Available Weight Remaining per Tablet

Target Tablet Weight 260 mg/tablet

Amount of Active Material – 215 mg/tablet

Amount of Lubricant – 2.6 mg/tablet

Available Weight Remaining 42.4 mg/tablet

Forty two point four milligrams of available weight remaining per tablet was used to

introduce the excipients in the formulation. Since there were six excipients present in each

experiment as per the Plackett-Burman design, the available weight remaining was divided

by six in order to introduce the excipients in equal portions.

23

Manufacturing Process

Table 9 describes the procedure used to manufacture the tablets for the twelve batches.

Table 9: Production Work Order for the First Set of Direct Compression Experiments

Product Name:

Senna 8.6 mg Tablets Batch Size: 50 g

Batch No.: SK1301-8.6

(Experiments #1 to #12)

Step Procedure

1 Weigh the active material and excipients as per the quantities specified in the master formulae

and transfer into a polyethylene bag.

Active Material:

- Senna, lot# 201109300009 expiry: 26/06/12.

Excipients:

- Pre-gelatised starch lot# 201111290003 expiry: 14/12/13,

- Microcrystalline cellulose lot# 201201040009 expiry: 27/02/14,

- Lactose monohydrate lot# 201109230004 expiry: 07/01/13,

- Sorbitol lot# E076N expiry: 20/09/12,

- Ethylcellulose 1 lot# 43156 Manufacturing date: 20/12/11,

- Hydroxyethylcellulose lot# W1010 Manufacturing date: 12/08/01,

- Hydroxypropyl cellulose lot# 120106-2 expiry: N/Av,

- Methylcellulose lot# ZF23012N21 expiry: 22/06/13,

- Hydroxypropylmethyl cellulose lot# YI26012N23 expiry: 25/09/12,

- Ethylcellulose 2 lot# YB26013T01 expiry: 25/02/13,

- Alginic Acid lot# H190687 expiry: 22/10/12.

2 Mix the blend in the polyethylene bag through manual shaking for 1 minute.

3 Measure blend flowability with Flodex instrument.

4 Weigh a quantity of magnesium stearate lot# 200808060001 equal to 1.0% of the net weight

of blend remaining and transfer it to the blend. Mix 10 seconds by manual shaking.

24

5 Compress the blend into tablets using the following parameters on the Fette Press 1200i.

Tooling: 0.3592 inch (9.1200 mm)

Tablet diameter (parameter #35) : 8.85 mm

Number of station(s) : 1 station

Rotor speed (parameter #2): 8 rotation per minute (rpm)

Fill-O-matic speed (parameter #3): 10 rpm

Compression force (parameter #5): Automatic setting (“K” run)

Tablet filling depth (parameter #6): 15.50 mm

Tablet cylindrical height main pressure

(parameter #18): 1.75 to 2.20 mm depending on blend density.

Fill cam No. (parameter #39): 16 mm

Tablet description: Round standard convex tablet with “S” on one

side and plain on the other side.

Target tablet weight: 260 mg

Target tablet hardness: N/Ap *Forms a tablet.

*For this first set of compresssion trials, no particular tablet hardness value was anticipated. The objective was to produce tablets that could maintain their integrity. Tablet hardness was adjusted at a later stage of formulation development (refer to chapters V, VI and VII).

The Fette Press 1200i is connected to a computer panel that allows adjustments and control

of the compression parameters. Compression parameters were programmed to obtain

tablets of desired weight, hardness and appearance (25). Most parameters remained

constant for all twelve experiments with the exception of parameter #18, discussed further

below. Note that the Fette Press 1200i has 322 adjustable parameters that were not

presented in details in this research project (26). Only critical parameters shown in Table

10 below were described.

25

Table 10: Critical Parameters Adjusted on the Fette Press 1200i

Parameter Parameter # Setting Minimum

Setting

Maximum

Setting

Tooling N/Ap 0.3592 inch

(9.1200mm)

N/Ap N/Ap

Tablet diameter (mm) 35 8.85 0.10 25.00

Number of station(s) N/Ap 1 1 6

Rotor speed (rpm) 2 8 5 120

Fill-o-matic speed (rpm) 3 10 10 120

Main compression force

kilonewton (kN)

5 Automatic setting 0.1 30.1

Tablet filling depth (mm) 6 15.00 to 15.5 11 15.5

Tablet cylindrical height main

pressure (mm)

18 1.75 - 2.20 0.50 8.50

Fill cam # 39 16 6 22

Tooling

At the time of assembly of the press, the only senna tooling size available was 0.3592 inch

(9.1200 mm). However, it is important to note that the diameter of senna tooling changed

from 0.3592 inch to 0.3437 inch. The use of either tooling size did not matter at this

developmental stage of the project since all batches were compressed with the same tooling

and the tooling diameter only differed by 0.0155 inch. Therefore, no significant impact was

anticipated due to tooling size. Table 11 describes the diameters values in inches and

millimeters for the previous and current senna tooling.

Table 11: Senna Tooling Diameters

Previous Senna Tooling Current Senna Tooling

inch mm inch mm

0.3592 9.1200 0.3437 8.7300

26

Tablet Diameter (mm)

Parameter #35 (26)

As per the settings recorded in Table 10, tablet diameter should have been set at 9.12 mm

in the control panel of the Fette press instead of 8.85 mm. The latter value was

inadvertantly selected by the computer from a previous run performed by another user.

Nevertheless, this setting discrepancy did not have a significant impact in the comparison

of experiments since parameter #35 remained constant for all twelve batches.

Number of Station(s) and Rotor Speed (rpm)

Parameter #2 (26)

The Fette Press 1200i can accommodate six stations (i.e. 6 upper punches, 6 lower punches,

6 dies). However, the tooling was installed on a single station due to the limited quantity of

blend available for each lot (50 g). When working with small quantities, it is ideal to

operate the Fette Press 1200i with one station and minimum rotor speed to ensure sufficient

dwell time (i.e. time allocated for the blend to flow into the die cavity).

Fill-o-matic Speed (rpm)

Parameter #3 (26)

The fill-o-matic is an important part of the Fette Press 1200i. It is used to contain and

dispense the blend into the die cavity with the aid of three rotating blades. The speed of the

blades can be controlled to optimise blend flow. With small amounts of blend, it is

important to consider that gravity force is reduced. Therefore, it is ideal to operate the fill-

o-matic at relatively low speed so the lateral force exerted by the blades does not exceed the

gravity force exerted on the blend. That way, the blend particles can properly fall into the

die cavity. The fill-o-matic speed in tune with the rotor speed were challenged in further

sets of experiments when larger quantities of blends were used (see chapters V, VI, VII). In

order to maximize tablets productivity, it is always ideal to determine the maximum speed

of compression for a particular formulation.

27

Compression Force (kN)

Parameter #5 (26)

The Fette Press 1200i has a caracteristic function of automatically adjusting the

compression force in kilonewton according to the quantity of material present in the die

cavity. This type of automatic adjustment occurs in “K” run and is commonly used at the

beginning of the compression process of larger batches in order to adjust tablet weight,

thickness and hardness. Once the desired tablet properties are achieved, the computer is

reverted to “L” run, where the compression force can be manually controlled throughout

the run.

For these twelve preliminary batches, the computer was set in “K” run for the entire

compression process due to limited blend quantity. Compression forces observed were used

as reference values for future compression experiments.

Tablet Cylindrical Height Main-pressure (mm)

Parameter #18 (26)

This parameter controls the distance between the lower and upper punches when the

material is compressed in the die cavity. It is used to adjust the thickness and hardness of

the tablets. In general, when the value of parameter #18 decreases, flater/harder tablets are

produced as compared to thicker/softer tablets observed as parameter #18 increases. The

value programmed in the control panel depends largely on blend density. A relatively

higher blend density requires a higher value of parameter #18 since additional space is

needed in the die cavity to accomodate for denser particles to be compressed appropriately.

The adjustment of parameter #18 was performed at the beginning of the compression

process for each batch in order to produce tablets of adequate thickness and hardness.

Values from 1.75 mm to 2.20 mm were considered adequate but since blend density was

not measured in this first set of direct compression experiments, the relationship between

28

blend density and parameter #18 will be further evaluated in the second set of compression

experiments (see Chapter III – Second Set of Direct Compression Experiments).

Fill Cam and Fill Depth

Parameter #39 and parameter #6 (26)

The fill cam and fill depth define the depth of the lower punch when the blend is introduced

in the die cavity. The adjustment of fill cam and fill depth typically controls the amount of

blend that enters the die cavity which is directly related to the tablet size and tablet weight.

The deeper the fill cam and fill depth, the higher the weight and size of the tablet.

For this set of experiments, a 16 mm fill cam was installed on the Fette press and 15.5 mm

fill depth was programmed in the computer. These settings were further evaluated in other

sets of compression experiments since desired tablet weight was not achieved in

preliminary trials. A smaller fill cam size was ordered for the sixth set of experiments to

meet the requirements for tablet weight and size (see Chapter VII – Sixth Set of Direct

Compression Experiments).

Results

In this first set of experiments, the following response factors were measured and compared

between batches:

1) Number of tablets produced,

2) Tablet hardness,

3) Blend flowability

4) Tablet disintegration time.

The number of tablets produced and hardness were the main critical response factors used

to identify the excipients that would promote feasibility of the manufacturing process.

Adequate hardness plays a major role in maintaining tablet intergrity throughout the direct

29

compression process whereas adequate productivity (percent yield) plays a critical role in

demonstrating the economical benefits of the manufacturing process. On the other hand,

blend flowability is also important for adequate productivity and consistancies in tablet

parameters. At last, proper tablet disintegration time is critical to achieve desired

therapeutic effect in vivo. Table 12 shows the response factors obtained for the twelve

experiments of the first set of direct compression trials.

Table 12: Response Factors for the First Set of Direct Compression Experiments

Experiment # Tablets Produced

Hardness (Kp)

Blend Flowability

(mm)

Disintegration Time (min.)

Excipients

1 16 1.1 22 24.57 (4 units)

PGS, EC1, HEC, HPC, HPMC, EC2

2 4 0.5 26 No test PGS, MCC, Sorbitol, MC , HPMC, EC2

3 41 1.2 26 29.34 (4 units)

MCC, Lactose, Sorbitol, HEC, HPC, HPMC

4 39 0.7 22 19.4 (3 units)

PGS, Lactose, HPC, MC, HPMC, Alginic Acid

5 31 0.9 26 23.9 (3 units)

MCC, HEC, HPC, MC, EC2, Alginic Acid

6 15 0.5 24 24.13 (3 units)

PGS, MCC, Sorbitol, EC1, HPC, Alginic acid

7 2 0.0 24 No test No excipient 8 4 0.5 26 No test PGS, lactose, Sorbitol,

HEC, EC2, Alginic acid 9 8 2.1 26 No test Lactose, Sorbitol, EC1,

HPC, MC, EC2 10 27 1.3 24 22.85

(6 units) Sorbitol, EC1, HEC, MC,

HPMC, Alginic acid 11 9 0.5 24 No test MCC, Lactose, EC1,

HPMC, EC2, Alginic acid 12 27 0.8 24 20.64

(5 units) PGS, MCC, Lactose, EC1,

HEC, MC

30

Analysis of Data (Minitab analysis)

Minitab software was used to statistically analyse the identification of excipients that

showed significant and positive effects on productivity, hardness, flowability and

disintegration time responses. The “Pareto Charts of the Effects” and the “Main Effects

Plots” were selected to illustrate the effects (10).

Figure 3: Pareto Chart of the Effects for Tablets

Produced

Figure 4: Main Effects Plot for Tablets

Produced

Figure 5: Pareto Chart of the Effects for Tablet

Hardness

Figure 6: Main Effects Plot for Tablets

Hardness

The excipients listed at the top of the the Pareto Charts of the Effects have the most

significant effects on the selected response factors. However, it is critical to couple the

Pareto Charts of the Effects with the Main Effects Plots analysis in order to determine

31

whether the effects of excipients on the response factors were positive or negative. Desired

responses were illustrated by a sharp positive slope in the Main Effects Plots which

corresponds to an increase in both response. For example, HPC-SSL-SFP showed a

significant positive effects on both tablet productivity and hardness responses. In reverse,

Ethocel Std 4 Premium (EC2) showed a significant negative effect on productivity since it

is listed at the top of the Pareto Chart of the Effects for number of tablets produced and

showed a sharp negative slope in the Main Effects Plot. This confirmed that EC2

significantly reduced productivity.

Figure 7: Main Effects Plot for Flowability

In terms of blend flowability, the results were similar between experiments (e.g. 22 – 26),

which made it difficult to select or eliminate excipients based on this response factor.

Similarity in flowability data was mainly caused by the large amount of active raw material

in the formulation (~80%). Nevertheless, lactose and sorbitol demonstrated positive

responses in the main effect plot for flowability (Figure 7), which corresponded to their

physicochemical properties of the raw materials described in Table 3. Therefore, these

excipient were considered as flowability enhancers for further sets of experiments.

32

Figure 8: Main Effects Plot for Tablets Disintegration

In terms of disintegration time, it was difficult to identify the excipients that enhanced this

response factor since insufficient data was produced in this set of experiments. No

disintegration data was obtained for experiments #2, #7, #8, #9 and #11 due to insufficient

number of tablets produced and the alternate experiments #1, #3, #4, #5, #6, #10 and #12

showed incomplete disintegration when the test was aborted after 45 minutes. Therefore,

the main effect plot for tablet disintegration (Figure 8) was not conclusive. For example,

Starch 1500 (PGS) and Avicel PH102 (MCC) did not show positive disintegration

responses as described in the litterature (17). In future sets of experiments, the

disintegration tests were conducted until complete disintegration of the tablets was

achieved (even after 45 minutes) in order to properly identify the excipients that have a

significant effect on this response factor.

In spite of insufficient and inconclusive data acquired in this set of trials, one conclusion

was drawn – a disintegrant must be added to the formulation since many units did not

disintegrate within 45 minutes. A disintegrant was not added in the first set of trials in order

to avoid the introduction of an additional variable to the experiments. In this way, the main

effects of selected excipients could be better identified.

33

Important note : Experiment #7 (no excipient) did not succeed in the direct compression

process. The two tablets produced showed hardness values of 0.0Kp which confirmed that

senna is not compressible by itself and clearly requires the presence of adequate excipients

in order to be compressed through this manufacturing process.

Selection of Excipients for the Next Set of Direct Compression Experiments

The formulator analysis using Minitab confirmed the selection of HPC, HEC and MC for

tablet productivity and tablet hardness. Sorbitol also seemed to play a positive role in tablet

hardness. In addition, sorbitol and lactose were selected for their optimal flow property and

bonding capacity originating from their mode of compression by brittle fracture mechanism

(see Chapter I Table 3: Physicochemical Properties of Active Raw Material and

Selected Excipients). The hypothesis of increasing bonding capacity within the tablet was

made since sorbitol and lactose were present in experiment #3 where the best productivity

response was observed and experiment #9 where the best hardness response was observed.

In these two experiments, bonding capacity could have been increased by the brittle nature

of sorbitol and lactose raw materials which balanced the large amount plastic deforming

material in the formulation.

In terms of disintegration time, the presence of lactose in the formulation was also desired

to achieve adequae response. Lactose has the property of creating hydrous channels via

capillary hydration within the tablet matrix which weakens the bonds between the

neighboring particles and promotes breakage of the tablet as a result. On the other hand,

croscarmellose sodium was selected for its known disintegrating efficacy as and its organic

nature. As compared to lactose, crosscarmellose sodium acts via swelling mechanism. It

has the property of absorbing water and swelling four to eight times its original volume.

The swelling exerts osmotic pressure on the neighboring particles and weakens the bonds

which eventually leads to breakage of the tablet into aggregates 23). Two percent

croscarmellose sodium was introduced in the second set of compression experiments (see

Chapter II – First Set of Direct Compression Experiments). Table 13 summarises the

34

selection of excipients for the second set of direct compression experiments based on the

desired response factors observed.

Table 13: Selected Excipients for the Second Set of Direct Compression Experiments

Selected Excipients Desired Response Factors Observed

HPC ↑ # Tablet Produced

↑ Hardness HEC

MC

Lactose monohydrate

↑ # Tablet Produced

↑ Bonding Capacity

↑ Flowability

↓ Disintegration Time

Sorbitol

↑ Hardness

↑ Bonding Capacity

↑ Flowability

Croscarmellose sodium ↓ Disintegration Time

Magnesium stearate ↓ Adhesiveness of Particles

35

Chapter III – Second Set of Direct Compression Experiments

Objective a

Research of an Optimal Formulation for Direct Compression of a … · 2020. 7. 30. · not a compressible material, methodologies such as Design of Experiments (DOE) and Quality by

Documents