design, testing, and scale-up of medical device class iib ... angelica...design, testing, and...
TRANSCRIPT
Design, testing, and scale-up of medical device class IIb:
risk analysis as a crucial tool for pharmaceutical industry
Angélica Esteves Lopes da Graça
Thesis to obtain the Master of Science Degree in Pharmaceutical Engineering
Pharmaceutical Engineering
Supervisors: Prof. Dr. Joana Marques Marto and Prof. Dr. Jorge Humberto Gomes
Leitão
Examination Committee
Chairperson: Prof. Dr. José Monteiro Cardoso de Menezes
Supervisor: Prof. Dr. Joana Marques Marto
Members of the Committee: Dr. Diogo Miguel De Sousa Manata and Prof. Dr. Lídia
Maria Diogo Gonçalves
May 2018
i
Design, testing, and scale-up of medical device class IIb:
the risk analysis as a crucial tool for pharmaceutical
industry
Angélica Esteves Lopes da Graça
Thesis to obtain the Master of Science Degree in Pharmaceutical Engineering
Pharmaceutical Engineering
Supervisors: Prof. Dr. Joana Marques Marto and Prof. Dr. Jorge Humberto Gomes
Leitão
Examination Committee
Chairperson: Prof. Dr. José Monteiro Cardoso de Menezes
Supervisor: Prof. Dr. Joana Marques Marto
Members of the Committee: Dr. Diogo Miguel De Sousa Manata and Prof. Dr. Lídia
Maria Diogo Gonçalves
May 2018
ii
Abstract Dry eye disease (DED) treatment usually consists in topical administration of eye drop solutions. This
research project was developed to create an effective and safe eye drop solution, containing Hyaluronic
Acid (HA). It was developed two medical devices (MD), an eye drop solution with 0.15% of HA (w/v) and
another with 0.30% of HA (w/v). Pre-formulation studies were performed to select suitable excipients.
After the selection the final formulation and a suitable manufacturing process, it was design a scale-up
study, producing a pilot batch manufacturing process. The pilot batch manufacturing process was
submitted to a Risk Analysis.
Two methods were used: FMEA and the ISO 14971:2007 Qualitative Analysis. After the detention of
possible failure modes, prioritization and corrective actions were applied. FMEA classified 14%
unacceptable risks and it was able to implement corrective actions, while the ISO 14971:2007 method
detected only 11% and it was not capable to implement corrective actions.
A Validation Plan was also performed to study the effectiveness of the manufacturing process. The
process controls performed showed that the results were consistent and respected the defined
specification of each parameter.
In the Product Characterization, measurements of viscosity and mucoadhesive studies were performed.
The viscosity measurements showed that HA 0.30% presented higher viscosity and the viscosity
increased when the mucin interacted with HA, suggesting the existence of interactions. In vitro tests
with ARPE-19 cells demonstrated that both products are promising MD for DED treatment since cell’s
morphology is preserved and the cell viability increased after the product administration.
Key-Words: Dry eye Disease; Hyaluronic Acid; Medical Devices; Risk Analysis; Mucoadhesion.
iii
Resumo Analítico O tratamento para o olho seco consiste na administração de soluções de conforto. Foi desenvolvido
um projeto para criar soluções eficazes e seguras contendo ácido hialurónico (AH). Decidiu-se investir
em dois produtos, um com concentração de 15% e outro com 30%. Estudos de pré-formulação foram
elaborados para definir os excipientes. Após a conclusão da formulação e do processo de fabrico foi
elaborado um estudo de scale-up onde se desenvolveu o processo de fabrico dos lotes piloto.
Submeteu-se o processo de fabrico a uma Análise de Risco onde se utilizaram duas ferramentas: FMEA
e Análise Qualitativa de acordo com a ISO 14971:2007. Após a detenção dos modos de falha foi
aplicado uma priorização e ações corretivas nas mesmas. A FMEA classificou 14% dos riscos como
inaceitáveis e implementou ações corretivas, a ISO 14971:2007 classificou 11% como inaceitáveis,
contudo não foi capaz de executar ações corretivas.
O objetivo da Validação de Processo foi estudar a eficácia do processo de fabrico. Os controlos de
processo demonstraram consistência nos resultados e conforme com os valores estabelecidos para
cada parâmetro.
Na Caracterização do Produto realizou-se um estudo de viscosidade e de muscoadesividade. Mediu-
se a viscosidade através de viscosímetros, texturómetro e potencial zeta concluindo que o AH 30%
apresenta maior viscosidade e que esta aumenta quando a mucina é adicionada, sugerindo a existência
de interações. Testes in vitro utilizando células ARPE-19 demonstraram que ambos os produtos são
promissores uma vez a morfologia celular é mantida e a viabilidade celular aumenta após a
administração do produto.
Palavras-Chave: Olho Seco; Ácido Hialurónico; Dispositivo medico; Análise de Risco;
Mucoadesividade.
iv
Agradecimentos Durante a realização desta dissertação muitos desafios foram enfrentados, tanto a nível profissional
como pessoal. Apesar do percurso não ter sido fácil e mais longo e árduo que o esperado, as
dificuldades foram ultrapassadas, na qual chega a hora de expressar os meus sentimentos a todos que
me ajudaram neste caminho e o tornaram mais simples:
À Professora Doutora Joana Marto, minha orientadora principal, pelos conselhos, pela ajuda, pela
constante preocupação e sobretudo pela amizade desenvolvida ao longo deste percurso. Considero-
me uma sortuda por ter tido uma orientadora que não só esteve sempre disponível para me guiar
durante a realização desta dissertação, como esteve sempre disponível para mim;
Ao Professor Doutor Jorge Leitão, pela disponibilidade e pelos conselhos que me facultou;
À Doutora Sara Raposo por me ter dado a oportunidade de elaborar este projeto e pelo convite de
pertencer à equipa Laboratório Edol, Produtos Farmacêuticos S.A.;
Aos Professores Doutor José Cardoso Menezes e Doutor António Almeida como coordenadores do
Mestrado e pela disponibilidade ao longo desta jornada;
Ao Laboratório de controlo microbiológico – LCM, mais precisamente à Dra. Alexandra Nogueira Silva
e à Técnica Paula Machado, pela oportunidade de entrar no mundo da indústria farmacêutica e pelos
conhecimentos desenvolvidos em microbiologia;
À Dra. Rita Carneiro pelos ensinamentos sobre a produção do produto, nomeadamente desde o seu
fabrico até ao enchimento, reforçando desta forma os meus conhecimentos;
Ao Laboratório Edol, Produtos Farmacêuticos S.A. por toda a disponibilização de todos os recursos
necessários à elaboração deste trabalho, com um especial agradecimento aos colaboradores do
departamento de Controlo de Qualidade;
À equipa “Análise de Risco” na qual fez parte o Professor Rui Loureiro e o Dr. Diogo Manata, pelos
ensinamentos aprofundados de análise de risco, pela disponibilidade, pelas ideias e conselhos em
como elaborar um dos mais complexos capítulos desta dissertação;
Às Professoras Doutora Lídia Gonçalves e Doutora Helena Margarida Ribeiro por me terem recebido
nos seus laboratórios, pela simpatia e ajuda que me forneceram;
A todos os membros do “Salão Nobre” do edifício D da Faculdade de Farmácia pela hospitalidade, pela
simpatia e pelo convívio durante a realização da parte laboratorial;
Às meninas da residência universitária Filipe Folque por todo o apoio;
Às minhas irmãs Maria Graça e Iolanda Tomás e aos meus amigos Nuno Costa e Miguel Antunes pelo
incentivo, força e ajuda;
Aos restantes familiares, mais precisamente pais e avós, pelo orgulho que sentem por mim e por o
demonstrarem.
v
Table of Contents
Abstract..................................................................................................................................................... ii
Resumo Analítico .................................................................................................................................... iii
Agradecimentos ....................................................................................................................................... iv
List of Figures ...........................................................................................................................................x
List of Tables ........................................................................................................................................... xi
Abbreviations .......................................................................................................................................... xii
Aims and Organization of the Dissertation .............................................................................................. 1
Chapter One – Literature Overview ......................................................................................................... 3
1. Theoretical Introduction ................................................................................................................... 3
1.1. Anatomy of the Eye ................................................................................................................. 3
1.2. Structure of the tear film .......................................................................................................... 4
1.2.1. The Lipid Layer ................................................................................................................ 5
1.2.2. The Aqueous Layer ......................................................................................................... 5
1.2.3. The Mucous Layer ........................................................................................................... 6
1.3. Dry Eye Disease ...................................................................................................................... 6
1.3.1. Treatment of Dry Eye Disease ........................................................................................ 8
1.3.1.1. Pharmacological Treatment ..................................................................................... 8
1.3.1.2. Food supplements ................................................................................................... 8
1.3.1.3. Medical Devices ....................................................................................................... 9
1.3.1.3.1. Ophthalmic Comfort Solutions ............................................................................ 10
1.4. Ophthalmic Comfort Solution Characterisation ..................................................................... 11
1.4.1. Mucoadhesion and mucoadhesive properties ............................................................... 11
1.4.1.1. Polymers ................................................................................................................ 12
1.4.1.1.1. Hyaluronic Acid ................................................................................................... 13
1.4.1.2. Factors Important to Mucoadhesion ...................................................................... 13
1.4.1.3. Techniques Used for the Assessment of Mucoadhesion ...................................... 14
1.4.1.4. Mucoadhesion Theories of Polymer Attachment ................................................... 15
1.4.1.4.1. Wetting Theory ................................................................................................... 15
1.4.1.4.2. Diffusion Theory ................................................................................................. 16
1.4.1.4.3. The Mechanical Interlocking Theory .................................................................. 16
1.4.1.4.4. The Electronic Theory ........................................................................................ 16
1.4.1.4.5. The Adsorption Theory ....................................................................................... 16
1.4.2. Evaluation of the Product Efficacy ................................................................................. 16
1.4.2.1. In vitro models ....................................................................................................... 16
1.4.2.1.1. 3D In Vitro Models of the Eye and Ocular Diseases .......................................... 17
1.5. Risk Analysis in Medical Devices .......................................................................................... 19
1.5.1. General Process ............................................................................................................ 20
1.5.2. Initiate Quality Risk Management Process .................................................................... 20
vi
1.5.3. Risk Assessment ........................................................................................................... 20
1.5.4. Risk Control ................................................................................................................... 21
1.5.5. Risk Communication ...................................................................................................... 21
1.5.6. Risk Review ................................................................................................................... 21
1.5.7. Risk Management Tools ................................................................................................ 21
1.5.8. The Importance of Applying Risk Analysis in Medical Devices ..................................... 22
Chapter Two – Previous Work .............................................................................................................. 23
2. Previous Work - Introduction ......................................................................................................... 23
2.1. Materials and Methods .......................................................................................................... 24
2.1.1. Materials ........................................................................................................................ 24
2.1.2. Methods ......................................................................................................................... 24
2.1.2.1. Appearance ........................................................................................................... 24
2.1.2.2. Determination of the pH values ............................................................................. 25
2.1.2.3. Determination of the Osmolality values ................................................................. 25
2.1.2.4. Determination of the Viscosity values .................................................................... 25
2.1.2.5. Efficacy of Antimicrobial Preservation ................................................................... 25
2.1.2.5.1. Strains and Preparation of Inoculum .................................................................. 25
2.1.2.5.2. Determination of the preservative efficacy of the formulation ............................ 25
2.1.2.6. Stability Testing ..................................................................................................... 26
2.1.2.7. Manufacturing Process .......................................................................................... 26
2.2. Pre-Formulation Studies ........................................................................................................ 26
2.2.1. Formulation Design ....................................................................................................... 26
2.2.2. Required specification ................................................................................................... 26
2.2.3. Selection of Materials .................................................................................................... 27
2.2.3.1. Polymer Selection .................................................................................................. 27
2.2.3.2. Preservative Selection ........................................................................................... 27
2.2.3.3. Buffer Selection ..................................................................................................... 28
2.2.4. Components of the Formulation ................................................................................ 28
2.2.5. Development of the Laboratory Batches ....................................................................... 32
2.2.6. Efficacy of Antimicrobial Preservation ........................................................................... 32
2.3. Final Formulation ................................................................................................................... 32
2.3.1. Polymer .......................................................................................................................... 32
2.3.1.1. Comfort Agents ...................................................................................................... 32
2.3.1.2. Hyaluronic Acid ...................................................................................................... 33
2.3.2. Excipients ...................................................................................................................... 34
2.3.2.1. Highly Purified Water ............................................................................................. 34
2.3.2.2. Potassium Chloride, Magnesium Chloride.6H2O and Calcium Chloride.6H2O ..... 34
2.3.2.3. Boric Acid and Sodium Tetraborate ....................................................................... 35
2.3.2.4. Suttocide (N-hydroxymethylglycinate 50%) and EDTA ......................................... 35
2.4. Manufacturing Process of the Final Formulation ................................................................... 36
vii
2.5. Stability Testing ..................................................................................................................... 36
2.6. Scale-Up ................................................................................................................................ 36
2.7. Conclusion ............................................................................................................................. 38
Chapter Three – Risk Analysis .............................................................................................................. 39
3. Risk Analysis - Introduction ........................................................................................................... 39
3.1. Materials and Methods .......................................................................................................... 40
3.1.1. Materials ........................................................................................................................ 40
3.1.2. Methods ......................................................................................................................... 41
3.1.2.1. Process Description ............................................................................................... 41
3.1.2.2. Identification of Critical Points ............................................................................... 41
3.1.2.3. Failure Cause and Effect of each Failure Mode .................................................... 41
3.1.2.4. Severity Rating Assignment .................................................................................. 42
3.1.2.5. Occurrence Rating Assignment ............................................................................. 42
3.1.2.6. Detection Rating Assignment ................................................................................ 42
3.1.2.7. Risk Priority Number Calculation ........................................................................... 42
3.1.2.8. Failure Modes Prioritization ................................................................................... 42
3.1.2.9. ISO 14971:2007 Qualitative Analysis .................................................................... 42
3.1.2.10. Corrective Actions .................................................................................................. 43
3.1.2.11. Risk Review ........................................................................................................... 43
3.2. Results ................................................................................................................................... 43
3.2.1. Ishikawa diagrams ......................................................................................................... 43
3.2.2. FMEA of Hyaluronic Acid Pilot Batch Manufacturing Process ...................................... 43
3.2.3. Failure Modes Prioritization and Corrective Actions...................................................... 48
3.2.4. Risk Review ................................................................................................................... 49
3.3. Discussion ............................................................................................................................. 50
3.4. Conclusion ............................................................................................................................. 52
Chapter Four – Pilot Scale Batches Process Validation ....................................................................... 53
4. Pilot Scale Batches Process Validation - Introduction .................................................................. 53
4.1. Materials and Methods .......................................................................................................... 53
4.1.1. Materials ........................................................................................................................ 53
4.1.2. Methods ......................................................................................................................... 53
4.1.2.1. Production of three batches of Hyaluronic Acid 0.15% and HA 0.30% ................. 53
4.1.2.2. Appearance ........................................................................................................... 53
4.1.2.3. Determination of the pH values ............................................................................. 54
4.1.2.4. Determination of the Osmolality values ................................................................. 54
4.1.2.5. Determination of the Viscosity values .................................................................... 54
4.1.2.6. Determination of the Density values ...................................................................... 54
4.1.2.7. Bioburden .............................................................................................................. 54
4.1.2.8. Sodium Hyaluronate Assay ................................................................................... 54
4.1.2.9. Sterility Test ........................................................................................................... 54
viii
4.2. Critical Steps to Control ......................................................................................................... 55
4.2.1. Acceptance Criteria ....................................................................................................... 56
4.2.2. Validation Plan ............................................................................................................... 56
4.2.2.1. Preparation of HA 0.15% and HA 0.30% Eye Drops Solution .............................. 56
4.2.2.2. In Process Control and Holding Time Validation of HA 0.15% and HA 0.30% eye
drops 56
4.2.2.3. Control of the Sterilizing Filtration .......................................................................... 56
4.2.2.4. Filling Control ......................................................................................................... 56
4.2.2.5. Finished Product Control ....................................................................................... 57
4.3. Results ................................................................................................................................... 58
4.4. Discussion ............................................................................................................................. 59
4.5. Conclusion ............................................................................................................................. 60
Chapter Five – Product Characterisation .............................................................................................. 61
5. Product Characterization - Introduction ......................................................................................... 61
5.1. Material and Methods ............................................................................................................ 61
5.1.1. Material .......................................................................................................................... 61
5.1.2. Methods ......................................................................................................................... 61
5.1.2.1. Viscosity Measurements ........................................................................................ 61
5.1.2.1.1. Brookfield viscometer ......................................................................................... 61
5.1.2.2. Mucoadhesion studies ........................................................................................... 61
5.1.2.2.1. Ostwald viscometer ............................................................................................ 62
5.1.2.2.2. Rotational Rheometer......................................................................................... 62
5.1.2.2.3. Zeta Potential ..................................................................................................... 63
5.1.2.3. In Vitro Assay ......................................................................................................... 63
5.1.2.3.1. Cell Culture Condition......................................................................................... 63
5.1.2.3.2. Cell Viability of HA 0.15% and HA 0.30% .......................................................... 63
5.1.2.3.3. 2D model - Evaluation of Cell Morphology and Cell Viability After Dehydration 63
5.1.2.3.4. 3D model - Dry Eye Model and Cell Viability...................................................... 64
5.1.2.4. Statistical Data Analysis ........................................................................................ 64
5.2. Results ................................................................................................................................... 64
5.2.1. Viscosity Measurements of HA 0.15% and HA 0.30% .................................................. 64
5.2.2. Mucoadhesive Studies .................................................................................................. 65
5.2.2.1. Viscosity Measurements ........................................................................................ 65
5.2.2.2. Rheology Measurements ....................................................................................... 66
5.2.2.2.1. Tackiness Testing ............................................................................................... 66
5.2.2.2.2. Oscillation Frequency Sweep ............................................................................. 68
5.2.2.2.4. Zeta Potential ..................................................................................................... 70
5.2.2.3. In Vitro Assay ......................................................................................................... 70
5.2.2.3.1. Cell Viability of HA 0.15% and HA 0.30% .......................................................... 70
5.2.2.3.2. 2D model - Evaluation of Cell Morphology and Cell Viability After Dehydration 71
ix
5.2.2.3.3. 3D model - Dry Eye Model and Cell Viability...................................................... 72
5.3. Discussion ............................................................................................................................. 73
5.4. Conclusion ............................................................................................................................. 77
Chapter Six – Conclusion and Future Work .......................................................................................... 79
6. Concluding Remarks and Future Work ......................................................................................... 79
6.1. Concluding Remarks ............................................................................................................. 79
6.2. Future Work ........................................................................................................................... 80
References ............................................................................................................................................ 81
Appendix ................................................................................................................................................ 89
x
List of Figures
Figure 1- Five main layers of the cornea's structure [4]. ................................................................... 3
Figure 2 - The eye anatomy diagrammatic illustration [5]. ................................................................ 4
Figure 3- Tear film multilayer composition [6]. ................................................................................... 5
Figure 4- Major etiological causes of dry eye. Adapted from [19]. ................................................... 7
Figure 5- Schematic representation of a polymer moving along the eye surface film during a
blink [47]. .............................................................................................................................................. 12
Figure 6- Contact angle between a droplet and solid surface [52]. ................................................ 15
Figure 7- Schematic representation of the in vitro cell culture model. Adapted from [65]. ......... 18
Figure 8 - Quality Risk Management Process overview (adapted from ICH Q9 and ISO 31000)
[68,71]. ................................................................................................................................................... 20
Figure 9 - Chemical repeat unit of HA [96]. ....................................................................................... 33
Figure 10 - Sodium hyaluronate chemical structure [100]. ............................................................. 34
Figure 11- Flow chart of the pilot batch manufacturing process of HA 0.15% and HA 0.30%. ... 37
Figure 12- Pareto chart of every failure mode of the manufacturing process of HA 0.15% and
0.30%. .................................................................................................................................................... 46
Figure 13- Results from the filling control test for HA 0.15% and HA 0.30%. ............................... 59
Figure 14 – Typical flow curve of shear stress as function of shear rate for HA 0.15% and HA
0.30% eye drops solution. .................................................................................................................. 65
Figure 15- Viscosity determination for HA 0.15% and HA 0.30% in absence and presence of
Mucin 5% (w/w) (mean SD, n=3). ..................................................................................................... 66
Figure 16- Three samples of pig eye used in the frequency sweep assay (A) and one of the
samples attached to the probe (B). ................................................................................................... 67
Figure 17- Frequency sweep with shear moduli as function of frequency of HA 0.15%, Mucin
and Mucin + HA 0.15% at room temperature. ................................................................................... 68
Figure 18- Frequency sweep with shear moduli as function of frequency of HA 0.30%, Mucin
and Mucin + HA 0.30% at room temperature. ................................................................................... 69
Figure 19- Time Sweep Test for HA 0.15% and HA 0.30% with and without mucin...................... 69
Figure 20- Determination of zeta potential for HA 0.15%, HA 0.30%, Mucin and both products
with mucin (Mean ± SD, n=3). ............................................................................................................. 70
Figure 21- Results of cell viability on ARPE-19 cell lines testing CR 0.30%, CR 0.15%, HA 0.30%
and HA 0.15% ate various concentrations (mean ± SD, n=8). ........................................................ 71
Figure 22-Cell viability ARPE-19 cell line after dehydration treatment exposed for 24h to HA
0.15% and HA 0.30% eye drops solution and commercial formulation CR 0.30% (mean ± SD,
n=9). ...................................................................................................................................................... 73
xi
List of Tables
Table 1- Overview of medical devices directives, categories, classes/list and examples [33][35].
............................................................................................................................................................... 10
Table 2- In vitro methods to evaluate mucoadhesion in ophthalmic solutions. ........................... 14
Table 3- Properties of selected cell culture models of ocular tissues and respective
applications. Adapted from [65]. ........................................................................................................ 18
Table 4- General case of stability testing types and respective conditions. ................................ 24
Table 5- Criteria for evaluation of antimicrobial activity in terms of log10 reduction in the
number of viable micro-organisms for eye preparations [78]......................................................... 26
Table 6- Proposed specification for the final formulations (HA 0.15% and HA 0.30%). ............... 27
Table 7- Polymer and excipients selected for the developed formulations. ................................. 29
Table 8- Main equipment used during the manufacturing process of HA 0.15% and HA 0.30%
eye drop solution. ................................................................................................................................ 41
Table 9- Qualitative Severity and Probability Levels. ...................................................................... 42
Table 10- Severity, Occurrence and Detection table with respective risk classification. ............ 44
Table 11- Weighting Factor and respective specification according to the years of work
experience and academic level. ......................................................................................................... 44
Table 12- RPN matrix. .......................................................................................................................... 47
Table 13- ISO 14971:2007 risk matrix [70]. ........................................................................................ 47
Table 14- Unacceptable risks obtained through FMEA and current and corrective action. ........ 48
Table 15- Unacceptable risks obtained through ISO 14971 qualitative analyses and current and
corrective action. ................................................................................................................................. 49
Table 16- RPN calculation after the implementation of the corrective actions. ............................ 50
Table 17 - Steps, Process Controls and Acceptance Criteria considered for the validation plan
of Sodium Hyaluronate 0.15% and 0.30% eye drops solution, 8 mL. ............................................ 55
Table 18- Results obtained in the IPC and FPC analysis with respective specifications of HA
0.15% and HA 0.30%. ........................................................................................................................... 58
Table 19- Results obtained in the Bubble Point and Bonfiglioli test for HA 0.15% and HA 0.30%.
............................................................................................................................................................... 59
Table 20- Normal force and area under force time curve results for HA 0.15% and HA 0.30% and
their interactions with Mucin and with Pig Eye. ............................................................................... 67
Table 21- Optical microscope images of ARPE-19 after dehydration in no protective conditions
(Dry Eye), after dehydration preceded by treatment with HA formulations and cells not
submitted to dehydration (Medium). ................................................................................................. 72
xii
Abbreviations Active Implantable Medical Devices (AIMD)
Active Medical Devices (AMD)
American Type Culture Collection (ATCC)
Aqueous Tear Deficiency (ATD)
Cetyltrimethylammonium broide (CTAB)
CFU (Colony-Forming Unit)
Colony Forming Units (CFU)
Commercial Reference (CR)
Detectability (DET)
Dry Eye Disease (DED)
Dulbecco's Modified Eagle Medium (DMEM)
Ethylenediamine tetraacetic acid (EDTA)
European Pharmacopeia (EP)
Evaporative dry eye (EDE)
Failure Mode Effect Analysis (FMEA)
FDA (Food and Drugs Agency)
Fourier-transform infrared spectroscopy (FTIR)
Heating Ventilation and Air Conditioning
(HVAC)
High Efficiency Particulate Arrestance (HEPA)
Hyaluronic acid (HA)
Hydroxyethyl cellulose (HEC)
In Process Control/Finished Product Control
(IPC/FPC)
In vitro Diagnostics (IVD)
Infrared (IR)
International Organization for Standardization
(ISO)
Kg (Kilogram)
LVER (Linear-Viscoelastic Region)
Medical Devices (MD)
mOsm (milliOsmol)
MW (Molecular Weight)
Negative Control (NC)
Occurrence (OCC)
Out-of-Specification (OOS)
Pa (Pascal)
PB (Pilot Batch)
Polyhexamethylene biguanide hydrochloride
(PHMB)
Polyvinyl pyrrolidone (PVP)
Polyvinylidene Difluoride (PVDF)
Positive Control (PC)
Psi (pound force per square inch)
Quality Control (QC)
R&D (Research and Development)
Risk Priority Number (RPN)
rpm (Rotation per Minute)
Sabouraud-dextrose agar (SDA)
Severity (SEV)
SFM (Serum-Free Media)
Sodium carboxymethyl cellulose (CMC)
Sodium dodecyl sulfate (SDS)
Sodium hydroxymethyl glycinate (SHMG)
Standard Derivation (SD)
The International Conference on
Harmonisation (ICH)
Tryptic soy broth (TSB)
Trypticase soy agar (TSA)
Ultraviolet–visible (UV/Vis)
V (Volt)
World Health Organization (WHO)
1
Aims and Organization of the Dissertation
Dry eye disease (DED) is one of the most common eye diseases in current times being the most frequent
diagnosed disease in ophthalmology. In order to attenuate some of the symptoms cause by this
pathology tear lubricants containing polymers in their formulation are the most common form of
treatment. For that matter the pharmaceutical company Laboratório Edol, S.A. created in 2015 a project
consisting in the development of a safe and effective eye drop comfort solution leading to this
dissertation.
There are several formulations in the marked of tear lubricant with different types of polymers, each with
their own characteristics in terms if viscosity, duration time, mechanism and mucoadhesion properties.
One polymer which has been used with success in treating patients with severe DED is Hyaluronic Acid
(HA), a natural polymer with similar properties to the mucin in terms of viscoelastic and biophysical
properties. HA eye drops available in the marked have a concentration ranging 0.10% and 0.30% of
HA. Thus, it was decided to develop two products with different concentration: one with HA 0.15% (w/v)
for less severe cases such as slight discomfort and, other with HA 0.30% (w/v) indicated for more severe
cases.
The purpose of this dissertation was to develop two medical devices (MD) suitable and efficient to treat
DED, the identification of potential risks that may be involved in the manufacturing process of the pilot
batches and if the manufacturing process was effective through a Process Validation Plan.
To determine the most suitable formulation for eye drops an intensive bibliographic search was
performed, corresponding to Chapter One – Literature Overview. Since the eye is a very sensitive
organ a formulation that mimics the eye environment was the strategy to prevent ocular damage. This
means the pH, osmolality, electrolyte composition, buffer and preservative chosen must respect certain
specifications. Once the final formulation was developed and the manufacturing process was defined,
the next step was the scale-up of final formulations. This study was performed by Edol’s collaborators
and corresponds to Chapter Two – Previous Work.
In order to evaluate possible risks that may compromise the quality of both products a Risk Analysis
was performed to the pilot batch manufacturing process, corresponding to Chapter Three – Risk
Analysis. Two methods were used, one is a general tool (Failure Mode Effect Analysis - FMEA) and
the other is specific to MD (Qualitative Analysis according to the ISO 14971:2007). The aim of this
chapter was to understand the possible risks associated with the manufacturing process and to compare
each risk analysis method in order to understand and decide which is more sensitive and more efficient
in correcting the detected risks.
HA 0.15% (w/v) and 0.30% (w/v) eye drops solution are new products, thus the pilot batch manufacturing
process must be submitted to a Process Validation to demonstrate to prove the process operates
effectively and to confirm that the resulting product obeys the product requirements. Process controls
were applied to both products to verify if both passed the required specifications and to perceive the
2
principal differences between them. This corresponds to Chapter Four - Pilot Batches Process
Validation.
The final chapter, Chapter Five – Product Characterization, had the purpose to evaluate the structural
and mucoadhesion properties of HA 0.15% (w/v) and 0.30% (w/v), eye drops solution and the efficacy
of both products by an in vitro Dry Eye model using ARPE-19 cell line. The structural and mucoadhesion
behaviour was evaluated by different viscosity and rheological measurement using mucine and pig eye.
A cell viability test as well a study of cell morphology was also performed to study cell behaviour when
the products were applied.
3
Chapter One – Literature Overview
1. Theoretical Introduction
1.1. Anatomy of the Eye
The human eye is the essential sense organ that allow us to see whose anatomy is quite complex. The
light is refracted by the eye itself and produces a focused image stimulating the nervous system, giving
the ability to see [1].
The eye consists in a fibrovascular, approximately, globular structure with 24 mm of diameter and with
a mass of about 7.5 g, corresponding less than 0.05% of the total body weight, with compact ocular
tissue whose thickness corresponds to several cell layers [2].
The structure of the eye is divided in the anterior segment and the posterior segment. The anterior
segment comprises the smaller anterior chamber, between the cornea and iris, is primarily responsible
for collecting and focusing light, and the larger posterior chamber, between the iris and the lens, is
responsible for detecting light. The anterior segment of the eye includes the lens, lachrymal system, iris,
aqueous humour, ciliary body, pupil, conjunctiva and the cornea. The posterior segment itself includes
the retina, choroid, sclera, macula, fovea, optic nerve and the vitreous humour [2].
Three concentric adjoining tissue layers comprise the eye. This organ is encircled with a collagenous
layer corresponding to the outermost layer that provides mechanical strength. The epithelial membrane
located in the interior portion is called the cornea, a clear transparent tissue with the purpose of focusing
the light to the retina. The cornea’s structure has both lipophilic and hydrophilic properties and five
distinct layers: epithelium, Bowman’s membrane, stroma, Descemet’s membrane and endothelium
(Figure 1). The epithelium is a lipophilic layer offering approximately 90% resistance to hydrophilic drugs
and 10% to hydrophilic preparations. Underneath the epithelium is the Bowman’s membrane, a
transitional acellular structure composed of protein fibres called collagen, approximately 8 to 14 µm in
thickness. The next layer is the hydrophilic stroma, a gelatinous structure composed by 80% of water,
comprising collagen, mucopolysaccharides and proteins. This layer corresponds to 90% of the cornea’s
total thickness. The Descemet’s membrane is located afterwards, with 6 µm of thickness, supporting the
endothelium and it is responsible to regulate stromal hydration [3].
Figure 1- Five main layers of the cornea's structure [4].
4
The posterior portion is represented by an opaque collagenous layer named sclera, which maintains the
shape of the eye and gives the attachment to the extrinsic muscle of the eye. The uvea is located in the
middle layer, it is a pigment layer that comprises the iris and the ciliary body in the anterior portion and
the vascular choroid in the posterior portion. The iris is the coloured part of the eye and controls the
amount of light entering the eye, serving as a biological aperture due to the presence of the pupil. The
ciliary body or muscle is a ring-shaped muscle responsible for the shape of the lens by contracting and
relaxing. It secretes aqueous humour providing nutrients to the avascular tissues in the anterior segment
and it also maintains the intraocular pressure. The choroid is a vast network of capillaries that supplies
the retina with nutrients and absorbs unused radiation. The retina corresponds to the innermost layer,
which detects and transduces light signal to the brain by the passage of light through the cornea,
aqueous humour, pupil, lens, the hyaloid and the vitreous humour, before reaching the retina. The retina
contains photosensitive elements called rods and cones, which convert the light detected into nerve
impulses and are sent onto the brain along the optic nerve. The lens is in between the anterior and
posterior segments, responsible for further refracting the light entering the eye. It is a flexible unit
consisted of layers of tissue enclosed in a tough capsule. The optic nerve contains approximately one
million fibres transmitting information from the rod and cone cells of the retina. The location where the
optic nerve leaves the retina is positioned the papilla. The anterior segment of the eye is filled with a
jelly-like substance called aqueous humour and the posterior segment is filled the gelatinous vitreous
humour. This segments are divided by a diagram called hyaloid [1,2]. All tissues and fluids are illustrated
in the Figure 2.
Figure 2 - The eye anatomy diagrammatic illustration [5].
1.2. Structure of the tear film
The tear film is a three-layered structure with the purpose to provide protection and lubrication to the
eye, reduces the risk of eye infection and keeps the surface of the eye smooth and clear. It also protects
the front of the eye from the environment and allows the eyelids to slide comfortably over the front
surface. This liquid comprises an anterior lipid layer, an aqueous layer and a deep mucin layer,
represented in Figure 3 [6][7].
5
Figure 3- Tear film multilayer composition [6].
When the eye is closed, the tear fluid is contained within a single compartment, the conjunctival sac,
and when the eye opens this fluid is re-distributed between three compartments: the sac, the pre-ocular
film and the menisci [7].
The tear film is only stable for a short period of time, thus after 20 to 40 seconds the human eye has an
urge to blink, due to an unpleasant sensation. In the short time between two blinks occurs the rupture
of the tear film due to dispersions forces and concentration gradients on the mucous layer, forming dry
spots on the cornea, creating irritation to the corneal nerve endings, which induces blinking. After
blinking and during the eyelid opening, a new tear film is created and spreads over the eye surface.
Dispersion forces, interfacial tension and viscous resistance of the mucous layer are the main factors
that define the time of rupture of the mucus and the breakup time of the tear film [8].
1.2.1. The Lipid Layer
The lipid layer, or oily layer, is an essential component to provide a smooth optical surface for the cornea,
enchasing the stability and the spreading of the tear film. This layer prevent the contamination of the
tear film by sebum, sealing the apposed lid margins during sleep, and most importantly retard water
evaporation from the surface of the open eye. The lipid layer remains stable over a series of blinks and
it is composed by meibomian lipids which are formed in the meibomian glands, a tubule-acinar holocrine
gland that discharged the entire content in the secretion process. These lipids are composed by non-
polar lipids such as cholesterol and wax esters and polar lipids such as (O-acyl-)-ω-hydroxy fatty acids
and phospholipids, which interact with the aqueous layer [7,9,10].
1.2.2. The Aqueous Layer
The aqueous layer is responsible for the formation of the tear bulk. This layer provides oxygen and
nutrients to the underlying avascular corneal tissue, flushing away epithelial debris, toxins and foreign
bodies. It is secreted from the lacrimal glands along with specific variety of proteins, electrolytes and
water. The second source of electrolytes and water in the tears is the conjunctival epithelium. This layer
also possesses anti-bodies that are critical for corneal wound repair and protects against infections
[6,10].
6
1.2.3. The Mucous Layer
The layer of the tear film in contact with the conjunctiva is mucus, which is also responsible for the
formation of the tear bulk, for enabling the tears to wet the eye by adhering to its front surface and for
the removal of unwanted fatty compounds from the corneal surface. Mucus resides at the surface of the
cornea and it is composed predominantly by sugar-rich glycosylated proteins, called mucin. The mucin
is produced by corneal and conjunctival epithelial cells, specialised goblet cells, and subsurface vesicles
found below the conjunctival cells. It is also composed by proteins, lipids, electrolytes, enzymes,
mucopolysacchrides and water. These transmembrane mucins, when anchored into the epithelial tissue
help stabilise the tear film. The gel-like structure of these mucins provide an easily wettable surface by
lower the surface tension of the tears, which helps the rapid spreading of fluid after blinks [7,8,10–12].
This spontaneous and rapid spreading is also possible due to the interaction of conjunctival mucin with
water and lipids, which contributes to the stabilization of the tear film [13].
There are up to 20 mucin genes identified in humans which produce MUC1, MUC2, MUC3A, MUC3B,
MUC4, MUC5AC, MUC5B, MUC6, MUC7, MUC8, MUC12, MUC13, MUC15, MUC16, MUC17, MUC19,
MUC 20, and MUC21. These proteins are situated in the wet-surfaced epithelia of the body such as the
ocular surface, entire airway and gastrointestinal tract. In the ocular surface occurs the expression of
MUC1, MUC4, MUC15, MUC16 and MUC20, being the last one the most expressed mucin in the human
conjunctiva. In the lacrimal gland MUC1, MUC4 and MUC16, also present in the cornea and the
conjunctiva, are present in the acinar cell membranes but only MUC4 and MUC16 are likewise present
in soluble form. All three mucins are located in the tears in soluble form [14].
1.3. Dry Eye Disease
DED, also known as keratoconjunctivitis sicca, is a pathology whose origin comes from several factors,
resulting in symptoms of discomfort, visual disturbance, tear film instability with potential damage to the
ocular surface, increased osmolality of the tear and inflammation. This disease can be divided into two
different types, aqueous tear deficiency (ATD) and excessive tear evaporation or evaporative dry eye
(EDE). DED is a commonly reported clinical problem and the most frequently diagnosed disease in
ophthalmology [11,15,16].
DED ranges from 5% to more than 30% of the population, estimated that about 3.23 million women and
1.68 million men with 50 years and older have dry eye, meaning that this disease affects mostly older
people and women, especially those who suffer from arthritis and allergies. It is also highly prevalent in
contact lens wearers [11,15,17].
Once again, DED can be divided into two main groups, the ATD and the EDE. The ATD can be
subdivided into Non-Sjögren’s syndrome and Sjögren’s syndrome, which is a generalised inflammatory
autoimmune disease associated with lacrimal and salivary gland lymphocytic infiltration. The EDE can
be divided into meibomian gland disease and exposure-related dry eye. Therefore there are four main
groups of diseases that can produce severe DED: the Sjögren’s syndrome, chronic progressive
conjunctival cicatrisation syndromes, specific ocular diseases and non-ocular diseases [9,17] (Figure
4).
7
Patients with Sjögren’s syndrome usually have dry eyes and dry mouth, but they also have or may not
have an associated rheumatological disease dependent on whether it is primary Sjögren’s syndrome or
secondary Sjögren’s syndrome, respectively [17].
The chronic progressive conjunctival cicatrisation syndromes includes several diseases such as the
Stevens-Johnson syndrome, trachoma, ocular pemphigoid, drug induced pseudopemphigoid, graft
versus host disease, chemical burns and conjunctival cicatrisation that occurs after severe membranous
conjunctivitis [17,18].
The third group is dedicated to those who have signs of DED because of specific ocular diseases, such
as dacryoadenitis, congenital absence of the lacrimal gland, Riley-Day syndrome, cholinergic blockade
due to drugs, chronic blepharoconjunctivitis, sinile atrophy of the lacrimal gland and after refractive eye
surgery [17].
The final group is devoted to those diseases that present DED symptoms but in reality have adequate
tear production, for instance trigeminal nerve paralyses with loss of corneal sensation, facial sensory
nerve paralysis, exposure keratitis and vitamin A deficiency, resulting in xerophthalmia, condition in
which the eye fails to produce tears [17].
Apart from the diseases mentioned above, other factors can cause an alteration to the evaporation rate,
including ambient conditions, hormonal regulation, blink rate, area of palpebral aperture, action of toxic
topical agents such as preservatives and complication in the tear film compartments. Contact lenses
can disrupt the stability of the tear film, since they increase the evaporation rate, causing a rupture in
the tear film about twice as fast as on the surface of the cornea. This lack of stability cause by DED is
due to the rise of the surface tension and osmolality values [9,10,19].
Figure 4- Major etiological causes of dry eye. Adapted from [19].
8
1.3.1. Treatment of Dry Eye Disease
1.3.1.1. Pharmacological Treatment
The pharmacological treatment of DED is focused on treating inflammation and tear restoration, since
DED is a symptom of various illnesses, resulting in inflammation of the cornea and conjunctiva. A widely
used approach to treat inflammation problems is the use of anti-inflammatory drugs such as topical
corticosteroid eye drops. Steroids are another treatment used in DED, however, it presents many side
effects as microbiological contamination, elevated intraocular pressure and cataract formation, therefore
it is only recommended treatments no longer than one or two weeks. Therefore, the use of non-steroidal
anti-inflammatory drugs increased due to their non-severe side effects in the treatment of DED, eye
discomfort decreases due to its analgesic effect and inflammation reduction. Cyclosporine eye drops,
the first one of the new generation immunomodulatory drugs, has an anti-inflammatory effect in the
lacrimal gland, resulting in an increase in tear production and conjunctival goblet density. Other
treatments consists in the application of antibiotics including oral doxycycline, azithromycin and
tetracycline to treat meibomian gland dysfunction [20–22].
1.3.1.2. Food supplements
According to the European Commission, food supplements are concentrated sources of nutrients (or
other substances) with a nutritional or physiological effect. Such supplements can be marketed as
tablets, capsules or even liquids in measured doses [23]. To alleviate symptoms associated with DED,
supplements available in the market are composed by fatty acids, vitamins and antioxidants such as:
Omega 3 and 6: there has been a great amount of interest generated in the area of using
essential fatty acids. Oral supplements containing omega-3 essential fatty acids, such as alpha
linoleic, docosahexaenoic acid and eicosapentaenoic acid, and omega-6 essential acids, such
as linoleic acid and gamma-linoleic acid, have shown to attenuate symptoms of dry eye by
treating chronic eye inflammation and Sjögren syndrome, due to their anti-inflammatory effect
[24];
Vitamin A: eye drops containing vitamin A have shown to be as efficient as prescription eye
drops. However, it is only indicated for patients with vitamin A deficiency since excessive
vitamin intake may trigger stomach and nerve side effects as well blurred vision [25];
Vitamin D: Patients with vitamin D deficiency present symptoms of dry eye and impaired tear
function. The administration of this vitamin is associate with the enhancement of tear film
stability and the reduction of ocular surface inflammation [26];
Lutein and Zeaxanthin: these antioxidants carotenoid pigments are presented in a high
concentration in the retina, more precisely the macula, working as a filter protecting the macula
from blue light and protect cells against oxidative stress by structurally bound antioxidants.
Studies have shown that lutein and zeaxanthin might reduce the risk of various eye diseases,
especially age-related macular degeneration, influencing cell viability and function, which may
reduce dry eye sensation [27].
9
1.3.1.3. Medical Devices
There are several approaches to treat DED, until now most treatments do not treat the cause of the
disease, but there are symptomatic treatments. This includes medical devices such as tear supplements
called artificial tears, which are artificial lubricants with hypotonic or isotonic buffers containing
electrolytes, surfactants and many types of viscosity agents. Another treatment option is tear retention
devices, which plug the lacrimal puncta preventing tear drainage. Other option is the use of implants to
permanently obstruct the lacrimal puncta, also known as punctal plug. Moisture chamber spectacles
have been supported to alleviate ocular symptoms associated with dry eye. The mechanism is by
increasing the periocular humidity and it can cause a growth of the lipid layer tear film thickness.
Because of this humidity, spectacle wears have a long inter-link interval [20,21,28].
According to the WHO, the definition of MD is an instrument, implement, machine, contrivance, implant,
in vitro reagent, or other similar or related article, including a component part, or accessory which is
intended for use in the diagnosis of diseases or other condition, or in the cure, mitigation, monitoring,
treatment, or prevention of disease, in man or animal. This products are intended to affect the structure
or any function of the body of man or other animals, and which does not achieve any of its primary
intended purposes through chemical action within or on the body of man or other animals and which is
not dependent upon being metabolized for the achievement of any of its primary intended purposes [29].
In other words, all kinds of diverse products from a simple bandage to a complex pacemaker, with no
pharmacological, metabolic or immunological activity, with the purpose of improving quality of life. Their
goal is to achieve the intended effect by other means beyond the ones given by drug consumption [30].
The European Commission defines it in a similar form, adding that MDs help improve the quality of life
of those with disabilities [31].
MDs are intended to be used by patients and consumers, which mean all products, must be submitted
to a safety assessment and performance evaluation before going to the market. Health professionals
also take advantage of the devices by giving the best information and advices on who to make the best
usage of the product [30].
There is a total of five types of MD according to the INFARMED [32]:
Active Medical Devices (AMD), dependent on an energy source non-generated by the
human body;
Active Implantable Medical Devices (AIMD), totally or partially introduced, surgically or
medically, into the human body and which is intended to remain after the procedure;
In vitro Diagnostics (IVD), devices used for examining sample material from the human body,
including excreted material, for providing information to ensure a correct patient diagnosis;
MD made by measure, manufactured specifically according to a written prescription for a
specific patient;
MD Systems and Sets for Intervention, packaged and dispensed together, which are placed
on the market with only one commercial name.
In Europe, registration of MD is subject to harmonised European directives, consisting of three
directives: Directive 90/385/EEC regarding active implantable medical devices (AIMDD 90/385/EC),
Directive 93/42/EEC regarding medical devices (MDD 93/42/EC) and the Directive 98/79/EC regarding
10
in vitro diagnostic medical devices (IVDD 98/79/EC) [33,34]. Classes and examples of each Directive
are presented in Table 1.
Table 1- Overview of medical devices directives, categories, classes/list and examples [33][35].
EU Directive Device
category Class/List Examples
MDD 93/42/EC MD
Class I (Low Risk) Bandage, wrists, compression socks,
wheelchairs, crutches
Class I Sterile (Low Risk) Examination gloves, sterile dressings
Class I Measurement (Low Risk)
Thermometer, syringes graduates without needle, blood pressure meter
Class IIa (Medium Risk) Syringes with needles, needles, lancets, surgical
gloves, resonance equipment
Class IIb (Low Risk) Blood bags, incubators, ophthalmic comfort
solutions, dressing material
Class III (High Risk) Cardiac valves, stents, hip prostheses, IUDs,
breast implants
IVDD 98/79/EC
High-risk IVD
List A - High Risk Material for detention of HIV 1 and 2, HTLV I and
II, Hepatitis B, C and D
List B - Moderate Risk
Material for detention of Rubella and toxoplasmosis, Phenylketonuria,
Cytomegalovirus infection, Chlamydia, blood groups,
Self-tests Pregnancy tests, ovulation tests, blood glucose
measurement equipment
Low-risk IVD
Other Urine or faeces collection flasks, aseptic urine
collection flask
AIMDD 90/385/EC
AIMD N/A Pacemakers, defibrillators, cochlear implants
1.3.1.3.1. Ophthalmic Comfort Solutions
Tear lubricants are the most common form of the treatment for DED and they are class IIb MD. There
are many lubricants formulation on the market that differ by their mechanism of action. It is not
completely understood the beneficial mechanism of tear lubricants, but it is related to volume
replenishment, tear film stabilization, preservation of the smooth refracting surface, reduction of tear
osmolality and reduction of the friction between the eyelids and the cornea. The mechanism depends
on the formulation used, for example, the hydrogel hydroxypropyl guar exists as monomers in a borate-
containing solution. Once installed in the patient’s eye, the contact between the patient’s tear and the
hydrogel causes the hydrogel to cross-link with the borate creating a more viscous and elastic matrix.
Another strategy is to replace or increase tear components to maintain tear stability, by producing for
example lipids [36,37].
Overall, eye drops differ in terms of composition, viscosity, duration of action, presence and type of
preservatives, osmolality and pH, being the main ingredient polymers. Polymers used in artificial tear
include hydroxypropyl methylcellulose (HPMC), carboxy methylcellulose, polyvinyl alcohol, carbopol,
polyvinylpyrrolidone, polyethylene glycol, dextran, HA and/or carbomers. These molecules may
increase the viscosity of eye drops, enhancing mucoadhesion. Increased viscosity increases the
retention time, although it also causes unwanted visual disturbance. In some cases higher viscosity eye
11
drops causes precipitation of crystals on the eyelids and lashes. For that reason it is recommended
daytime use of eye drop with low-viscosity and for highly viscous preparations use before sleep [36].
The preservative use in ophthalmic preparation can prevent contamination. Preservatives used in these
formulations include benzalkonium chloride, chlorobutanol, sodium perborate, sodium chlorite,
polyquaternium-1, cetrimonium chloride, EDTA, sodium hydroxymethyl glycinate and
polyhexamethylene biguanide hydrochloride. However, benzalkonium chloride and chlorobutanol
preservatives can cause symptoms similar to DED as they can damage the corneal epithelium, the other
preservatives mentioned may be less harmful to the ocular surface [36,38,39].
One of the characteristics of DED is increased osmolality due to corneal and conjunctival epithelial
damage. Thus, lowering tear osmolality using hypo-osmolar solutions may be a suitable strategy to treat
DED [36,40].
The variability of tear pH may be related to with carbon dioxide saturation or the meibomian lipids in the
tear film, thus the tear pH can range from 6.9 to 7.5 for normal population and patients with DED. The
most common buffer systems used in ophthalmic eye drops are citrate, phosphate, Tris-HCl and borate
buffer. Phosphate has been the buffer of choice for a long time, since it naturally occurs in the eye,
however calcific degeneration of the superficial cornea by calcium hydroxyapatite deposition has been
associated with the use of ophthalmic products containing phosphate. The borate buffer was for that
reason introduced as an ophthalmic buffer, since it was better suited than phosphate and, because of
its antimicrobial activity. There are no cases of complications after the use of citrate and Tris in literature,
though citrate buffer pH ranges between 3.0 and 6.2, which is lower than the tear film pH. Tris buffer
showed effectiveness in treating ocular acid burns, when the time of exposure to acid was short [36,41].
1.4. Ophthalmic Comfort Solution Characterisation
1.4.1. Mucoadhesion and mucoadhesive properties
Mucoadhesion is the adhesion of a material to a mucous membrane or a mucus-covered surface which
is the case of the eye surface. The problem concerning ophthalmic drugs is that due to the efficient
protective mechanisms of the eye, the bioavailability is often low. The act of blinking and lachrymation
remove rapidly foreigner substances, as well drugs from the eye surface. Formulation with
mucoadhesive properties are for that reason a topic of interest in order to enhance drug bioavailability,
prolonging the contact time between the formulation and the corneal epithelium. Due to the
mucoadhesion properties, the frequency to apply the product is decreased, reducing the toxic side
effects and the drug concentration [8,42].
The strategy to increase the residence time of the drug at the ocular surface is by changing the product
characteristics, for instance its viscosity. Although, by increasing the product viscosity the patient vison
may become blurred at the moment of administration, which may compromise the treatment. In organs
with high blood flow and considerable surface area, such as the eye, mucoadhesion drug delivery
systems give rapid absorption and high bioavailability [43].
Mucoadhesive formulations use water soluble polymers as the adhesive component. The mechanism
of adhesion starts with the wetting and swelling of the polymer, allowing an intimate contact with the
tissue. Then the polymer entanglements on the mucins chains forming weak chemical bounds and
12
allowing the adhesion on the mucosal surface. The bounds between the polymer and the mucous layer
could be physical or mechanical, secondary chemical bounds and covalent bounds. The required
characteristics to a polymer in order to obtain adhesion to the mucous layer are at least one of the
following: sufficient quantities of hydrogen-bounding chemical groups, anionic surface charges, high
molecular weight, high chain flexibility and induction of spreading by surface tensions [44,45].
1.4.1.1. Polymers
Polymers are high candidates to improve the bioavailability of ophthalmic formulations, due to their
mucoadhesive properties. High molecular weight polymers with different function groups, such as
carboxyl, hydroxyl, amino and sulphate, are capable or forming hydrogen bonds with mucous layers,
without crossing biological membranes. The interpenetration of the polymer chains in the mucous layer
cause a strengthening on itself by forming entanglements and secondary bound with the mucin
molecules or due the dehydration of the mucus caused by water movement. This strengthening is
necessary for a strong mucoadhesion. Water-soluble mucoadhesives polymers slowly dissolve in the
tear film, whereas water-insoluble would be retained until the mucin is replaced or until the shear force
of blinking dislodges the mucoadhesive system [8,43,46].
Other vantage for the use of polymers solutions is that they can increase the thickness of the normal
tear film, which is favourable on its stabilization. There are two types of mechanisms responsible for
spreading tears over the ocular surface (Figure 5). The first is by mechanical action, the upper eyelid
pulls water as it is raised, thus the amount of water moved is enhanced by the vertical spreading of
polymer molecules, dragging additional water with them. The second mechanism moves the polymer
molecules from the bulk solution to the surface, providing a surface-pressure gradient which induces
spreading [47]. Nowadays, HA is one of the most used polymer in eye drops formulations due to its
mucoadhesive properties. Examples of other mucoadhesive polymers are cellulose derivatives such as
methylcellulose, acrylates, chitosan, lectins and pectins [8,48].
Figure 5- Schematic representation of a polymer moving along the eye surface film during a blink
[47].
13
1.4.1.1.1. Hyaluronic Acid
The salt form of HA is sodium hyaluronate and its molecules have physical properties similar to the tear
glycoproteins and can easily cover the corneal epithelium. HA macromolecules are natural and are
presented in the vitreous body of the eye. First, they were used as viscosity agents, but it was found
that these polymers adsorb at the mucin/aqueous interface and extend into the adjacent aqueous layer,
forming a stabilizing thick layer. This polymer exhibits, just like the tear film, a non-Newtonian behaviour,
giving an advantage of high viscosity at rest between blinks, which decreases the dryness sensation of
DED [43,49].
Tear substitutes eye drops solution containing HA have been used with success in treating patients with
severe DED. The similar properties to mucins, such as viscoelasticity and biophysical properties, have
beneficial effects, providing a long-lasting hydration and retention time and obtaining a good lubrication
of the ocular surface. HA presents wound healing properties and promotion of corneal epithelial cell
proliferation, since HA is an important constituent of the extracellular matrix [43]. A study performed by
Johnson et al. was carried out to access the efficacy of two developed eye drops containing 0.10% and
0.30% of HA in the treatment of DED, using thirteen subjects with moderate DED. The eye drops were
administrated and the subjects’ symptoms were measured at 5, 15, 30, 45, 60 min and then hourly until
6 hours after drop installation. The authors concluded that both products reduced symptoms of ocular
irritation and lengthens tear breakup time [50]. Other study performed by Hamano et al. determined the
most effective concentration of HA in prolonging tear film stability. It was concluded that concentration
should be at least 0.10% to delay the breakup of the precorneal tear film.
1.4.1.2. Factors Important to Mucoadhesion
Several factors may affect the polymers bioadhesive properties. Polymer-related factors may be the
molecular weight, since it is necessary to have a high molecular weight in order to obtain good
bioadhesion. The concentration of the molecule also has a massive impact, highly concentrated systems
decreases significantly the adhesive strength, because the coiled molecules become solvent-poor,
making the chains for interpenetration less available. Other reason can be the lack of flexibility of
polymer chains, if the polymer chain decreases the capacity to interpenetrate and entanglement, the
mucous layer also decreases, reducing the bioadhesive strength. Besides molecular weight or chain
length, the conformation of the polymer’s molecule is also important. Exposed adhesively groups are
dependent of the molecule’s conformation, which determents the capacity of bioadhesion in the mucous
layer [44,51].
Factors related to the environment are the pH levels The mucous layer will have a different charge
density depending on the pH due to difference in dissociation of functional groups on the carbohydrate
part of the polypeptide backbone. Thus, the pH of the product is important for the degree of hydration.
The contact time is also important since it determines the extent of swelling and interpenetration of the
polymer chains. Swelling is also other factor and it is dependent of the polymer concentration, ionic
concentration and presence of water. Overhydrated molecules may result in a slippery system, which
decreases adhesion to the mucous layer [44].
14
1.4.1.3. Techniques Used for the Assessment of Mucoadhesion
The most common and convenient methods to assess the mucoadhesive properties of a potential
formulation is through in vitro tests. There are numerous studies concerning in vitro tests to measure
mucoadhesion, although there is not a standard procedure for such technique. Several approaches
were developed to rank mucoadhesion properties of polymers and formulations to understand their
adhesive behaviour [52].
Methods can be from simple analytical techniques to more sophisticated procedure. One procedure is
tensile strength measurements, also called texture analysis, were the mucoadhesive characterization is
based on the measurement of tensile force to break the interaction between the mucoadhesive platform
and the test substrate. The maximum detachment force and work of adhesion, the area under force time
curve, are used as parameters for comparison of mucoadhesive performance. This method provides
information about possible mucoadhesive strength, though it may be limited due to the dissolution of the
polymer in mucous gel and the absence of biological tissue [42,52,53].
Other methods include rheological techniques, particularly flow and oscillation methods have been
frequently used to study mucoadhesive properties of formulations. Interactions between mucin and
polymer may be measured by flow and oscillation trough viscosity and viscoelasticity properties,
respectively. Rheological methods are the most widely used techniques, providing information
concerning the deformation of the material within a broad range. However, the results obtained depend
on the concentration of the polymer, the mucin type and instrumental factors. Other disadvantage of this
method is that it cannot give information about the weakest region of the mucoadhesive junction since
only the interpenetration layer is simulated with this method [46,52].
Table 2 shows some examples given by authors to study the mucoadhesion capacity of polymers using
different methods.
Table 2- In vitro methods to evaluate mucoadhesion in ophthalmic solutions.
In vitro
methods Excipients (Polymer) Mucoadhesion results Conclusions Reference
Tensile Strength
Carbopol C971 and C974
4% (w/v) in water based
gels
Higher maximum force
in C974 (6.21 ± 2.75 N >
4.60 ± 1.71 N)
Higher elastic
component
and highly
crosslinked polymer
[54]
Tensile Strength
Polycarbophil 3% (w/w)
and poly(methylvinylether-
co-maleic anhydride 3%
(w/w) (Gantrez S97) in
vaginal gels
Gantrez S97 showed
higher mucoadhesive
strength (0.50 ± 0.04 N
< 0.37 ± 0.03 N)
Increased diffusion,
interpenetration and
entanglement with
mucin
[55]
Viscosity
Measurements
Sodium Hyaluronate
HA1100, HA800, HA500
and HA250 at several
concentrations (data not
shown)
Higher concentration
increased the viscosity
Strength of
formulation/mucin
interaction increases
with HA
concentration
[56]
15
Table 2- In vitro methods to evaluate mucoadhesion in ophthalmic solutions (Continuation).
In vitro
methods Excipients (Polymer) Mucoadhesion results Conclusions Reference
Viscosity
Measurements
Chitosan and Alginate
(concentration data not
shown)
Increase of viscosity in
both cases when
interacted with mucin
Electrostatic
interactions,
hydrogen bounding,
hydrophobic
interactions and
entanglement with
mucin
[57]
Oscillatory
Measurements
Sodium Hyaluronate
HA500 0.67% (w/w)
G’ > G’’ at high
frequencies (10 Hz)
Physical
entanglements with
mucin
[56]
1.4.1.4. Mucoadhesion Theories of Polymer Attachment
A definition of the process of mucoadhesion does not exist, since this is a complex process. Numerous
theories have been proposed in order to explain this phenomenon: the wetting theory, the diffusion
theory, the mechanical theory, the electronic theory, the adsorption theory, the cohesive theory and the
facture theory. In Table 2 is present some examples of published studies who used some of the below
theories to discuss their results.
1.4.1.4.1. Wetting Theory
The wetting theory discusses the bounding between the formulation and the surface tissue through
intermolecular interaction and interfacial tension. It is usually applied for liquid or low mucoadhesive
systems, measuring the ability of the system to spread across the biological substrate. The spreading
indicates that there are interactions and can be measured by the liquid-solid contact angle. If the contact
angle of liquids on surface tissue is lower than 90°, a greater affinity for the liquid to the layer surface is
achieve and the liquid tends to spread out to a large area, due to adhesive forces. If the contact angle
is greater than 90° the wetting of the surface is unfavourable, because of the cohesive forces within the
liquid molecules, maintaining the shape of the droplet and minimizing its contact area to the solid surface
[44,45,52] (Figure 6).
Figure 6- Contact angle between a droplet and solid surface [52].
16
The interfacial tension between the solid and the gas is represented by γSG, the interfacial tension
between solid and liquid by γLG and θ the contact angle between solid and liquid interface. The interfacial
tension may be calculated by Young’s equation:
𝛾𝑆𝐺 = 𝛾𝑆𝐿 + 𝛾𝐿𝐺 cos 𝜃 (1)
1.4.1.4.2. Diffusion Theory
The diffusion theory supports the concept that interpenetration and entanglement of bioadhesive
polymer chains and mucous chains produce semi-permanent adhesive bounds, the deeper the
penetration the stronger the bound. This penetration is dependent on the concentration gradient and
diffusion coefficient. The bound strength for a given polymer is achieved when the depth of penetration
is approximately equal to the end-to-end distance of the polymer chains. It is important that the polymer
and the mucous have a similar chemical structure, the more structurally similar a polymer is to its target,
the greater the mucoadhesive bond will be [44,45]. A study performed by Reinhart and Peppas reported
that the diffusion coefficient depended on the polymer’s molecular weight and that it decreased with the
increase of cross-linking density [58].
1.4.1.4.3. The Mechanical Interlocking Theory
This theory only considers the adhesion between the liquid and a rough surface or a surface rich in
pores forming an interlocked structure, which rise to adhesion. The adhesion between the
mucoadhesive system and the surface occurs within a diverse biological environment [44,52].
1.4.1.4.4. The Electronic Theory
The bioadhesive material and the target biological material have a different electronic structure, when
the contact occurs, it results in a transfer of electrons amongst the surface forming an electronic double
layer at the bioadhesive-biological material interface. This theory suggests that the electronic forces are
critical in generating bound adhesions [44,45,52].
1.4.1.4.5. The Adsorption Theory
The various surface interactions that results in adhesion is due to the presence of intermolecular forces,
including primary bounds and secondary bounds formation. These forces include hydrogen bonding and
Van der Waals interactions. Although these forces are individually weak, the number of interactions can
produce an intensive adhesive strength. This is the most accepted theory [44,45,52].
1.4.2. Evaluation of the Product Efficacy
1.4.2.1. In vitro models
Cell-based assays are nowadays a widely used method in drug development process providing a
simple, fast and cost-effective tool in order to avoid large-scale animal testing. The conventional 2D
culture system has helped the scientific community to study the cellular physiology and their behaviour.
This system consists in a monolayer of cells cultured on a flat and rigid substrates, in this conditions the
extracellular matrix components, cell-to-cell and cell-to-matrix interaction important to cellular
17
differentiation and proliferation and cellular functions in vivo are lost. This happens because 2D cell
cultured does not consider the natural 3D environment of cells in vivo, meaning that the results may
provide misleading and non-predictive data for in vivo responses [59,60].
3D cultured cells provides a behaviour closer to the complex in vivo conditions, giving a major advantage
over the 2D approach that the gap between cell cultured systems and the cellular physiology is
decreased. The 3D approach replicates or mimics the extracellular matrix making it a good near-to-in
vivo system, other advantages of this approach for evaluation drug candidates include oxygen and
nutrient gradients, increased cell-to-cell and cell-to-extracellular matrix interactions, non-uniform
exposure of cells within the 3D structure to the test molecule and varying cell proliferation zones [59–
62].
The reason for the near-to-in vivo behaviour in 3D cultures is the use of matrices and scaffolds. The
type of matrix or scaffold used depends on the type of morphological and physiological behaviour of the
cells as well the nature of the study. Commonly used scaffolds are agarose, collagen, fibronectin,
gelatin, laminin and vitronectin, mimicking the native extracellular matrix by porosity, fibrous,
permeability and mechanical stability. This micro environment enhances the biophysical and
biochemical interactions of the adhered cells, providing more realistic and predictive data for in vivo
studies and cost effective screening platform for drug development and testing [60,63].
1.4.2.1.1. 3D In Vitro Models of the Eye and Ocular Diseases
Ocular diseases have been investigated trough animal and cell cultured models to understand the
molecular mechanism which causes those diseases and to study potential drug candidates. The human
eye presents unique and complex features whose animal models just cannot mimic. Cell culture are
therefore an advantage, because they are experimentally controlled systems, making the results more
reproducible than those obtained from animal models [64].
To determine a drug’s efficacy in ocular diseases, in vitro 3D cultured cell studies have been widely
used, providing useful data comparable to in vivo studies. In drug discovery research, a 2D cell model
may lead to a selection of a candidate drug that cannot reach its target in vivo, making the 3D model
more efficient since it creates a microenvironment more appropriate for demonstrating long-term effects
of the drug. It only exists to date reginal parts of ocular in vitro models, like corneal, conjunctival and
retinal models, but not an in vitro ocular equivalent as an organ. Meaning that the 3D model must be
built in order to simulate the conditions and the specific region of the disease or the environment in study
(Figure 7) [64,65].
Various types of cells can be used in this model, cell-based lines systems form animals or human
resources, the second type with the advantage of providing reliable results by avoiding species-related
problems. These cells can be specified primary or immortalized cells depending of the location of
study and the application (
Table 3).
In order to test the efficacy of eye drops to treat the DED, some authors altered cell culture conditions
to mimic the conditions given by the disease. Salzillo et al. optimized hyaluronan-based eye drop
formulation using dehydrated porcine corneal epithelial cells by removing the medium and the multiwells
18
were incubated at 37 ºC and 5% CO2, without the lid for about 20 minutes [56]. Other study performed
by Meloni et al. experimentally induced in vitro dry eye on human corneal epithelium (HCE) model by
placing HCE tissues under controlled environmental conditions to mimic dryness, <40% relative humidity
at 40 ºC ± 5 ºC temperature and 5% CO2 [66].
Figure 7- Schematic representation of the in vitro cell culture model. Adapted from [65].
Table 3- Properties of selected cell culture models of ocular tissues and respective applications.
Adapted from [65].
Ocular Tissue Cell culture model Applications
Corneal epithelium Primary rabbit cells: cultured onto fibronectin/collagen/laminin coated membrane using SFM for 7-8 days
Permeability and transport studies
Immortalized human cells: HCE-T cell line, cultured on collagen-coated membranes using SFM for
6 days
Cell biology, toxicity, ocular irritancy, gene/drug delivery
Corneal endothelium Immortalized human corneal endothelial cells: IHCEn cell line, cultivated onto lyophilized human amniotic membrane
Positive expression of Na+/K+ ATPase
Conjunctival epithelium Primary rabbit cells: cultured on collagen-coated membrane using SFM for 8-10 days
Permeability and transport studies
Primary bovine cells: cultured on collagen-coated membrane, 10% serum medium for 9-11 days
Cytotoxicity screening, cytokeratin expression
Immortalized rat cells: CJ4.1A and CJ4.3C cell lines, cultured in 10% serum medium for 4 days
Investigation of antigen translocation across a mucosal barrier
Retinal pigment epithelium (RPE)
Primary isolated bovine cells: co-culture with endothelial cells for 14 days
Effect of endothelial cells on barrier function of the RPE
Primary isolated rat cells: cultured onto laminin coated filters using SFM for 5-7 days
Influence of serum on tight junction formation
Immortalized human cells: ARPE-19 cell line, cultured onto collagen-coated membrane, 10% serum medium for 9-11 days
Characterization of ARPE-19 as a human RPE cell line forming polarized epithelial monolayers
Retinal capillary
endothelium
Primary isolated bovine retinal capillary endothelial cells: cultured onto polycarbonate filters (coated with gelatin, laminin, fibronectin, and collagen)
Establishment of retinal capillary endothelial cell model
Immortalized rat retinal capillary endothelial cells: TRiBRB cell line
Functional expression of cell membrane transporters
19
1.5. Risk Analysis in Medical Devices
In the pharmaceutical industry quality systems approach was hurled by de FDA with the publication of
“Pharmaceutical cGMPs for the 21st Century—A Risk Based Approach”. The aim of this publication was
to transform regulatory approaches into a science-based and risk-based approaches with an integrated
quality systems orientation. Since that time, the use of quality systems has been a massive success in
this industry [67].
It is important to understand that the risk involving the quality of the medical product is just one
component of all the possible risks associated with the product manufacture, since the components are
also at risk. The product quality should be maintained throughout the product lifecycle, which means
that all the attributes that constitute the product quality must remain consistent with those used in clinical
studies. Risk management is therefore a powerful tool to ensure high quality of the product to the patient
by providing proactive means to identify and control potential quality issues during development and
manufacturing. By applying risk management, all kind of advantages are possible such as, make better
decision if a quality problem arises, provide regulators with grater assurance of a company’s ability to
deal with potential risks and can beneficially affect the extent and level of direct regulatory oversight
[68].
The word “risk” is defined as the combination of the probability of occurrence of harm and its severity. It
is difficult to create a standard application of risk management since every case has its own
particularities, each case might perceive different potential harms, different probabilities of occurrence
and their level of severities. In the pharmaceutical industry the main concern in the quality of the product
is its safety when administrated to the patient, all the components must be within the acceptable range
in order to guarantee patients health [68–70].
In all activities of a company there is a risk involved, they manage this risk through its analysis and the
identification, when it is detected a necessary alteration. The implementation and maintenance of risk
management allows [68]:
Increase the likelihood of achieving the pretended goals;
Encourage a proactive management;
Be aware of the necessity of identifying and manage potentials risks;
Identifies opportunities and threats;
Comply with applicable legal and regulatory obligations and international standards;
Improve mandatory and voluntary reporting;
Improve governance;
Increase confidence and credibility of the organization;
Establish a reliable basis for decision-making and planning;
Use resources in the treatment of the risk effectively;
Reinforce safety and health;
Reduces losses;
Improves organizational learning and resilience.
Quality risk management has principles, namely, the evaluation of the risk should be based on scientific
knowledge and ultimately link to the protection of the patient. However, the level of effort, formality and
20
documentation of the quality risk management process should be commensurate with the level of risk.
The company should elaborate and implement strategies to improve their risk management [68,69].
ISO 14971:2017 is an ISO standard specific for the application of risk management to MD to identify
hazards associated with MD, including in vitro diagnostic (IVD) MD, to estimate and evaluate the
associated risks, to control these risks, and to monitor the effectiveness of the controls during the
product life cycle [70].
1.5.1. General Process
Quality risk management process works as a map to help in the assessment, control, communication
and review of risks that may cause damage to the quality of the drug product (Figure 8). This process
should be included in the culture and practices of the organization and it must be built to fit perfectly in
the procedure that needs to be controlled [68,69,71].
1.5.2. Initiate Quality Risk Management Process
In order to define a problem or a potential risk, an assumption should be made of the potential risk,
collecting its background information and identify all necessary resources. These are the steps to initiate
a risk management process. Quality risk management should include systematic processes with the
propose of facilitate and improve science-based decision regarding the risk [68].
1.5.3. Risk Assessment
Risk assessment consists of identifying and analysing potential hazards, as well as identifying risks
associated with those hazards. This step of the Quality Risk Management Process should be performed
by a multidisciplinary team composed by engineering, quality assurance, validation and manufacturing
Figure 8 - Quality Risk Management Process overview (adapted from ICH Q9 and ISO 31000)
[68,71].
21
experts, resulting in a risk question which is the main focus in the risk assessment. The brainstorming
should involve the resolution of three main questions: What can go wrong, how likely is it to go wrong,
and how severe are the consequences? By answering these questions risk identification, risk analysis
and risk evaluation, which are key characteristics of a risk assessment process, are completed [67].
1.5.4. Risk Control
Once the hazards are identified and described, a plan is developed to reduce and/or accept risks. That
plan consists in determining if the risk level is acceptable, if not what can be done to reduce or eliminate
it, if there is a balance between risks, benefits and resources and if new ricks are introduced by
controlling the initial risk. The purpose of risk control is therefore to reduce the risk to an acceptable
level [67].
1.5.5. Risk Communication
Risk Communication consist in sharing information about risk and risk management between the
decision makers and other interested parties, regulators and industries, either within or outside the
company. The output/result of the quality risk management process should be appropriately
communicated and documented, the information might relate to the existence, nature, form, probability,
severity, acceptability, control, treatment, detectability and other features of risks to quality [67–69].
1.5.6. Risk Review
A mechanism to review or monitor events should be implemented as part as the ongoing quality
management process. Once a quality risk management process has been initiated, that process should
continue to be utilized for events that might impact the original quality risk management decision,
whether these events are planned or unplanned. The frequency of any review should be based upon
the level of risk and it might include reconsideration of risk acceptance decisions [67–69].
1.5.7. Risk Management Tools
Some tools are described to structure Risk Management intended to organize data and facilitate
decision-making, some methods are simple and possible to use with general information, while others
require more information and detail. No tools or set of tools are applicable to every situation, tools must
be studied and chosen according with the specific situation in order to build a specific quality risk
management procedure [68].
Simple techniques used in risk management are basically scheming to facilitate organization and
indicate, in a simple form, the progress of the situation. These includes Flowcharts, Check Sheets,
Process Mapping and Cause and Effect Diagrams, also known as Ishikawa diagram. More complex
tools used to characterize in detail possible risks and to prioritized them include: Failure Mode Effect
Analysis (FMEA), Failure Mode, Effect and Criticality Analysis (FMECA), Fault Tree Analysis (FTA),
Hazard Analysis and Critical Control Points (HACCP), Hazard Operability Analysis (HAZOP),
Preliminary Hazard Analysis (PHA) and Risk Ranking and Filtering [68].
22
The ISO 14971:2007 offer two tools to evaluate and to classify risks identified in a process, through
Qualitative analysis and through Semi-quantitative analysis.
1.5.8. The Importance of Applying Risk Analysis in Medical Devices
When manufacturing a MD it must guarantee it is safe and effective for human use. Risk analysis
involves the identification, understand, control and prevent failures that can result in hazards in the
manufacturing process of a MD. The benefits of conducting risk analysis during a MD manufacturing
process can be significant and can be used to balance some or all of the cost of implementing risk-
mitigating measures. By implementing risk analysis in a manufacturing process of a MD a higher
probability of producing a consistent product with high quality can be a possibility [67,68,70].
23
Chapter Two – Previous Work
2. Previous Work - Introduction
The International Conference on Harmonization (ICH) Q8 (R2) defines pharmaceutical development as
designing a quality product and its manufacturing process who gives a consistent delivery of the
intended performance of the product. Pharmaceutical development studies and manufacturing
experience provide important information in understanding and defining a design space, a
multidimensional combination and interaction of input variables and process parameters that
demonstrated the assurance of quality of the product [72]. The ISO 13485:2016 explains in detail the
development of a MD, which in our case is an eye drop [69].
Once a promising compound for development is defined the selection of the excipients and their
assembly is decided next, which can be a complex procedure, because the choice may not be
straightforward. A robust and reproducible product manufacturing process must result from this study
[73].
Pre-formulation testing is the first step in the rational development of dosage forms. In the formulation
of artificial tear substitutes, lubrication and nutrition with extended ocular surface time needs to be
demonstrated, meaning that some key elements need to be consider [74]:
Selection of active ingredients;
Decision on salt composition and osmolality;
Selection of viscosity agents;
Choice of pH/buffering agents/ buffering capacity;
Inclusion of other ingredients;
Exclusion or inclusion of preservatives;
Avoidance of toxicity to the ocular surface.
Osmolality and pH value parameters should be within range of normal tears in order to allow full recovery
of epithelial barrier function. The residence time of tear substitutes and its stability are two of the most
important characteristics that an eye drop should have, being viscosity, surface tension and
mucoadhesive ability to the ocular surface, important aspects to maintain the tear film intact. Eye drops
are administrated many times per day and in some cases before sleep, which means drug related
toxicity is disallow because it may induce ocular surface damage [74].
The drug substance has a shelf life and optimal storage conditions. This information is provided by
stability studies, whose purpose is to provide evidence on how the quality of a drug substance or drug
product varies with time under the influence of a variety of environmental factors for instance
temperature, humidity and light. It also establishes a re-test period for the drug substance or a shelf life
for the drug product and recommended storage conditions. Stability studies data should be provided on
at least three batches of the drug product, all tree with the same formulation and packed in the same
container closure system as proposed for marketing department [75].
The testing should include physical, chemical, biological and microbiological attributes such as
preservative content and challenge test, since these attributes of the drug are susceptible to change
during storage and are likely to influence the quality, safety and efficacy of the product. This susceptibility
24
to change is due to storage conditions such as temperature, light, air and humidity, as well as the
package components. These are the conditions under which the product shall be submitted in order to
establish a suitable shelf life [75,76].
The combination of small volume and aqueous medium for many ophthalmic products can result in
particular challenges, when applied stability test conditions [77]. The test conditions defined in the
guideline ICH QIA (R2) are represented in Table 4 [75].
Table 4- General case of stability testing types and respective conditions.
Study Storage condition Minimum time period covered by data at
submission
Long term 25 ± 2 °C / 60 ± 5% RH or
30 ± 2 °C / 65 ± 5% RH 12 months
Intermediate 30 ± 2 °C / 65 ± 5% RH 6 months
Accelerated 40 ± 2 °C / 75 ± 5% RH 6 months
*RH – relative humidity
A minimum of 12 months duration should be covered for the long term testing, using at least three
batches at the time of submission. The duration of the stability test should cover the proposed shelf life.
Results from these studies will form an integral part of the information provided to regulatory authorities
[75].
In this chapter it is described the necessary steps for the development and the respective characteristics
of two MD pilot batches. These MD are eye drops containing in their formulation HA, one with
concentration of 0.15% (w/v) (HA 0.15%) and the other 0.30% (w/v) (HA 0.30%) intended for the
treatment of DED.
2.1. Materials and Methods
2.1.1. Materials
High MW and low MW Sodium hyaluronate were a kind gift from Inquiaroma (Espanha). PHMB 20%
(Cosmocil) was a gift from DS Produtos Quimicos (Portugal) and N-hydroxymethylglycinate 50%
(Suttocide) from Ashland (EUA). Potassium chloride, sodium chloride, sodium tetraborate and EDTA
were purchased form VWR International (Portugal), calcium chloride.6H2O was purchased from José
Manuel Gomes Santos, magnesium chloride.6H2O was purchased from Sigma Aldrich Quimica and
acid boric was purchased from LaborSpirit (Portugal).
2.1.2. Methods
2.1.2.1. Appearance
The macroscopic appearance of each formulation was visually analysed and used as first stability
indicator.
25
2.1.2.2. Determination of the pH values
The pH values were measured using a pH meter with an incorporated temperature sensor (827 pH Lab,
Metrohm, Switzerland) by inserting the electrode in the sample and read directly at a temperature
between 20°C and 25 °C.
2.1.2.3. Determination of the Osmolality values
The osmolality values were measured using a osmometer (OSMOMAT 030, Gonotec, Germany) by
collecting 50 μL of the sample using a micropipette and place it in a vial, the results were read directly
in the osmometer.
2.1.2.4. Determination of the Viscosity values
The viscosity values were measured using a viscometer (BROOKFIELD DV-I +, Brookfield, USA) by
placing a 16 mL of the sample on the UL/Y adaptor, using needle nº 0, at 6 and 12 rpm for HA 0.15%
and 0.30% and 0.6 rpm for HA 0.30%.
2.1.2.5. Efficacy of Antimicrobial Preservation
2.1.2.5.1. Strains and Preparation of Inoculum
The bacterial and fungal strains tested were obtained from the American Type Culture Collection
(ATCC): Staphylococcus aureus ATCC 6538, Pseudomonas aeruginosa ATCC 9027, Candida albicans
ATCC 10231 and Aspergillus brasiliensis ATCC 16404. Bacterial strains were grown in Tryptic Soy agar
(TSA) at 30-35°C for 18-24h and C. albicans was grown in Sabouraud-Dextrose agar (SDA) agar at 20-
25°C for 24-48h. For Aspergillus brasiliensis the inoculation was performed in SDA and incubated at 20-
25°C until good sporulation was obtained. The stock solution was obtained in a test tube with 10 mL of
suspension medium adjusting the inoculum to 108 spores/mL. A suspension was prepared for each
microorganism with 108 CFU/mL by comparing with the MacFarland scale (Remel) in suspension fluid
medium (APT) and used immediately. Series of decimal dilutions were performed by transferring 1 mL
of the stock solution to 9 mL of suspension fluid medium, until final dilution of 10-8 is reached obtained.
The number of CFU presented in each dilutions was determined by spreading 100 µL of the last counting
dilutions (10-5 to 10-7) in TSA plates for bacteria and SDA plates for fungi. The plates were incubated
until possible count.
2.1.2.5.2. Determination of the preservative efficacy of the formulation
The antimicrobial effectiveness test was executed by inoculating the two eye-drops formulations, HA
0.15% and HA 0.30%, with the microbial suspension to obtain the initial concentration of 105-106
CFU/mL of each organism. The inoculated product was stored at 20-25°C protected from light and the
samples were withdrawn from the containers at specific time intervals. A sample of 1 mL of the
inoculated product was removed at 0h, 6h, 24h, 7 days, 14 days and 28 days taking into account the
neutralization method. The number of viable micro-organisms was determined by plate count in
duplicate and negatives controls of the media and the sample in study was conduct at all times.
26
Table 5- Criteria for evaluation of antimicrobial activity in terms of log10 reduction in the number
of viable micro-organisms for eye preparations [78].
NR: No recovery NI: No increase in number of viable micro-organisms compared to the previous reading
2.1.2.6. Stability Testing
Three batches of 500 mL were produce of HA 0.15% and HA 0.30% and were stored during 6 months
at room temperature (25 ± 2 °C / 60 ± 5 % RH), intermediate conditions (30 ± 2 °C / 65 ± 5 % RH) and
under accelerated conditions (40 ± 2°C / 75 ± 5 % RH).
2.1.2.7. Manufacturing Process
The manufacturing process starts with the addiction of the polymer, sodium hyaluronate, in highly
purified water. After complete dissolution of the polymer, the electrolytes potassium chloride,
magnesium chloride hexahydrated and calcium chloride hexahydrated are added and stirred until
complete dissolution. Next the sodium chloride and the buffering agents boric acid and sodium
tetraborate are added and stirred until complete dissolution. The manufacturing process finishes with
the addition of the preservative, stirred until homogenization. The pH is adjusted with hydrochloric acid
1M or sodium hydroxide 10M, if the pH level is not between 7.0-7.6, and the osmolality is adjusted with
sodium chloride if the value is outside the range of 280-320 mOsm/Kg.
2.2. Pre-Formulation Studies
2.2.1. Formulation Design
In 2015 it was developed an eye drop solution with an electrolyte composition similar to the lacrimal
fluid, with the purpose of developing an effective and safe eye drop solution for dry eye treatment. The
human eye is a very sensitive organ, which means that any type of compound that differs from its natural
environment may cause serious damages. Therefore the formulation of the saline solution must respect
all the requirements necessary to the eye comfort [79].
2.2.2. Required specification
To avoid eye irritation and provide ocular lubrication and comfort, there as some required specifications.
These specifications are the pH value, osmolality, sterility and appearance. The electrolyte composition
must be similar to the lacrimal fluid, which is mainly composed with Na+, K+, Cl- and HCO-. The tear fluid
normal pH is 7.4, consequently the formulation for an eye solution should be ideally between pH 7.0
and 7.6. The buffer should have low buffering capacity in order to allow the tears to regain the pH level
more rapidly. The osmolality of the tear film ranges between 294 and 310 mOsm/Kg [80] due to the
number of ions dissolved in the aqueous layer of the tear film. Thus, the formulation must have a similar
Log10 reduction
6 h 24 h 7 d 14 d 28 d
Bacteria A 2 3 - - NR
B - 1 3 - NI
Fungi A - - 2 - NI
B - - - 1 NI
27
osmolality value to avoid an osmotic pressure differential. The specifications for the final formulation are
represented in Table 6 [79,81].
Table 6- Proposed specification for the final formulations (HA 0.15% and HA 0.30%).
Tests Specifications
Appearance Limpid, clear and odourless solution
pH 7.0 – 7.6 at 20 - 25° C
Sodium Hyaluronate Assay 90-110 %
Viscosity Under study
Osmolality 280 – 320 mOsm/Kg
Sterility Absence of growth
2.2.3. Selection of Materials
2.2.3.1. Polymer Selection
An extensive study was carried out by Charles J. White et al. for the selection of most effective polymer
to use as a comfort agent. For the selection of comfort agents, the evaluation of water retention, apparent
flow viscosity, zero-shear viscosity, intrinsic viscosity, surface tension and comfort agent index were
performed. It was concluded that polysaccharides comfort agents are more effective than acrylic comfort
agents and that the HA was the most effective comfort agent at low molecular weight (MW) and
concentration, and should be given priority when selecting comfort agents [82].
Eye drops containing HA available in the market have a concentration ranging 0.10% to 0.30%. Topical
ophthalmic solutions should exhibit viscosity to prevent drainage from the ocular surface and to prolong
residence time. However, solutions with very high viscosity may cause blurred vision and problems in
the sterile filtration system of the manufacturing process. Precorneal tear film stability was significantly
increased in patients with dry eye when treated with 0.10% or 0.30% HA solutions, reducing the
symptoms. It was decided to develop a 0.15% HA solution for daily use and a 0.30% HA solution for
more severe cases of dry eye or before sleep application [83].
2.2.3.2. Preservative Selection
The selection of the preservative is an important assignment since microbial contamination may
represent a source of infection for the patient’s eye and modify the properties of the administrated drug.
The addition of these components in the formulation is designed to kill microorganisms or to avoid
microbial growth [39]. Two types of preservatives were tested, sodium hydroxymethyl glycinate (SHMG,
Sutoccide) with disodium ethylenediaminetetraacetic acid (EDTA) and Polyhexamethylene biguanide
hydrochloride (PHMB, Cosmocil) with and without EDTA [39,84].
28
2.2.3.3. Buffer Selection
The limited buffering capacity of the tear fluid is mainly due to the dissolved carbon dioxide and
bicarbonate, thus the buffer selected must have low buffering capacity and significant antimicrobial
activity. A borate buffer was evaluated for its antimicrobial activity in a study by R. Dennis Houlsby et
al., and the results demonstrated that the buffer exhibited significant antimicrobial activity against many
type for strains, with or without carbon source. However, it does not meet by it-self the criteria for
effectiveness required, other agents must be added to meet the specifications [85,86].
2.2.4. Components of the Formulation
The qualitative composition of the final formulations is represented in Table 7, which contains the
comfort agent and the excipients selected. During the pharmaceutical development a variety of
excipients were reviewed and used in the formulations in order to prepare and optimize the eventual
excipient mixture and the product properties. All the below excipients are recognized as safe materials
for human administration, being regarded as non-irritant and nontoxic on the amounts used [87–92].
29
Table 7- Polymer and excipients selected for the developed formulations.
Main Function Chemical Structure Molecular
Weight Solubility
Ex
cip
ien
ts
Sodium Hyaluronate Comfort Agent
*
H2O 5mg/mL at 25°C
Highly Purified Water
Solvent
18.015 g/mol
-
Potassium Chloride
Establish an electrolyte composition similar to the tear fluid
74.548 g/mol
H2O 35.5g/100mL
Magnesium Chloride Hexahydrate
Establish an electrolyte composition similar to the tear fluid
167.845 g/mol
H2O 1.670 g/L at 20°C
30
Table 7- Polymer and excipients selected for the developed formulations (Continuation).
Main Function Chemical Structure Molecular
Weight Solubility
Ex
cip
ien
ts
Magnesium chloride.6H2O
Establish an electrolyte composition similar to the
tear fluid
219.068
g/mol
H2O 81.1 g/100
mL at 25°C
Anhydrous Sodium Chloride
Adjust osmolality
58.44 g/mol
H2O 36g/100mL at 25°C
Boric Acid
Buffering agent
61.831 g/mol
H2O 50 mg/mL at 25 °C
Sodium Tetraborate
Buffering agent
201.22 g/mol
H2O 26 g/L at 20 ºC
31
Table 7- Polymer and excipients selected for the developed formulations (Continuation).
*Confidential information
Main Function Chemical Structure Molecular
Weight Solubility
Ex
cip
ien
ts
EDTA Preservative booster
292.244 g/mol
H2O 1g/mL at 25 °C
Cosmocil (PHMB 20%)
Preservative
185.275
g/mol
426 g/L at 25 °C
Suttocide (N-
hydroxymethylglycinate 50%)
Preservative
127.075
g/mol
H2O 1g/mL at 25 °C
32
2.2.5. Development of the Laboratory Batches
Initially 12 formulations were developed, 6 formulations regarding 0.15% of the HA and the other 6 with
0.30% of the HA. The formulations differed on the sodium hyaluronate MW, the addition of EDTA and
the type of preservative. After the measurements of pH, osmolality and viscosity values some
formulations suffered some adjustments in order to respect the specific ranges of each measure. It was
concluded only 4 of the initial 12 formulations passed the defined criteria proposed on Table 6 and
selected for the efficacy of antimicrobial preservation test. From the chosen formulations, 3 of them
present 0.30% concentration of HA, 2 with Sutoccide as a preservative and the other with Cosmocil.
The remaining formulation has 0.15% concentration of HA with Sutoccide as preservative.
2.2.6. Efficacy of Antimicrobial Preservation
The selected formulations were subjected to an efficacy of antimicrobial preservation test, in order to
investigate if the chosen preservative is the most adequate. The preservative properties of the product
are adequate if there is a significant fall or increased in the number of microorganisms in the inoculated
product after the times and the temperature recommended. According to the European Pharmacopeia
to ophthalmic preparations, the formulations must respect the criteria of the evaluation of antimicrobial
activity reduction on the number of vial microorganisms specifically described [93].
Once collected and analysed the results, according to Table 5 none of the formulations passed the
required criteria A, but three of them passed the criteria B.
2.3. Final Formulation
Based on the physic-chemical and microbiological results, it was decided that the final formulation has
Suttocide (N-hydroxymethylglycinate 50%) with EDTA as a preservative. Thus, excipients used in the
manufacturing process of HA 0.15% and 0.30% eye drops solution, 8 mL solution are: Sodium
Hyaluronate, Potassium chloride, Magnesium chloride.6H2O, Monosodium Phosphate monohydrate,
Disodium Phosphate dodecahydrate, Sodium Chloride, Calcium chloride.6H2O, Anhydrous Sodium
Chloride, Boric acid, Sodium tetraborate, Suttocide (N-hydroxymethylglycinate 50%), EDTA, Highly
Purified Water, Sodium hydroxide (NaOH 40%) and Hydrochloric acid (HCl 10%). All these excipients
will be further analyzed in the next section of the present report.
2.3.1. Polymer
2.3.1.1. Comfort Agents
Comfort agents are used extensively within topical eye drop formulations promoting comfort through
different mechanisms of action such as retaining tear volume by reducing drainage rates, stabilizing the
tear film, changing tear film surface tension, preventing tear evaporation, and altering tear fluid viscosity.
Frequent application of comfort agent in eye drops can greatly increase the level of comfort perceived
by consumers. Their molecules function in order to relieve ocular discomfort by enhancing
characteristics of the tear film, stabilizing and retaining tear volume and lubricating the ocular surface.
There are two broad classes of comfort agents: polysaccharide comfort agents and acrylic comfort
agents. Polysaccharide comfort agents are typically macromolecules composed of one or more types
33
of monosaccharide. Polysaccharide comfort agents are typically linear, hydrophilic, and possess high
MW. Substitution along the polymer backbone is common and can affect the overall conformation of the
macromolecule, particularly at high degrees of substitution. These substitutions can be branches, alkyl
groups, functional groups, or even salt complexes. Solution viscosity of polysaccharide comfort agents
is typically high and present a shear-thinning behavior. In general, all polysaccharides have high water
affinity and high rheological-modifying properties. This type of comfort agent was chosen for the study
formulation. Acrylic comfort agents are linear chains composed of carbon–carbon backbones with
regular repeat units, often including at least one functional group. This type of comfort agents can be
used as polyelectrolytes or in the neutral state. This category also includes polyacids, which are less
used than other agents but have slightly increased water retention properties when compared to neutral
acrylic agents [47,82,94].
2.3.1.2. Hyaluronic Acid
Since its discovery, HA, also called hyaluronan, has received great attention as a versatile and highly
functional biopolymer (Figure 9). HA is evolutionarily conserved from simple prokaryotes all the way to
complex eukaryotes, which is an undeniable testament to its biologic relevance. Structurally, HA is not
inert. The native form, high MW, can be broken down into smaller MW fragments in response to
glycosidase activity upregulated by environmental cues, such as pH and reactive oxygen species. The
MW variants have been used in a variety of biomedical applications, eliciting varying biologic responses.
For instance, low MW of HA promotes the production of inflammatory mediators. Similarly, high MW of
HA inhibits production of pro-inflammatory mediators, suggesting differential macrophage activation by
different molecular weight polymers, even dough the same molecule [56]. Clinically, HA has been used
in several applications, including ophthalmology as a drug delivery system and lubricant, in osteoarthritis
for viscosupplementation and as a dermal filler [95].
Each HA molecule consists of as many as 50000 repeats of the simple disaccharide glucuronic acid β
(1→3) N-acetylglucosamine β (1→4). There is one ionisable carboxyl group per disaccharide. Individual
segments of an HA molecule fold into a stiff rod-like conformation because of the β linkages between
the saccharides and the extensive intrachain hydrogen bonding between adjacent sugar residues. Due
to the large number of hydrophilic residues on its surface, HA binds a large amount of water and forms
a viscous hydrated gel. The first isolation of the compound was in 1934 by Meyer and Palmer extracted
from vitreous humour of 100 cattle eyes, but it can also be found in skin, umbilical cord, cartilage and
synovial fluid [96–98].
Figure 9 - Chemical repeat unit of HA [96].
34
Sodium hyaluronate is the sodium salt of HA (Figure 10) and it is a natural polymer comfort agent, class
of polysaccharides, used in many ophthalmic preparations. Its appearance is a white to beige powder,
highly soluble in water with a melting point of 241-247°C [99].
Figure 10 - Sodium hyaluronate chemical structure [100].
This salt is commonly used as a bioavailability-enhancer in eye drops. In the presence of HA, the
precorneal residence times of pilocarpine, timolol, aceclidine, tropicamide, arecoline, gentamicin, and
tobramycin were pro-longed. In addition to its viscosifying and mucoadhesive properties, HA has other
beneficial effects on the corneal epithelium, including protection against dehydration, reduction of
healing time, reduction of the inflammatory response caused by dehydration, and lubrication of the
ocular surface. Due to this clinical efficacy, HA is largely used in ophthalmology not only as an excipient
but also as the main component of the artificial tear substitutes commonly prescribed for the treatment
of DED [56].
2.3.2. Excipients
2.3.2.1. Highly Purified Water
Highly purified water is used as a vehicle and for dissolving substances, it should be prepared from
drinking water as a minimum-quality feed-water. This type of water is a unique speciation for water found
only in the European Pharmacopoeia. It must meet the same quality standard as water for injections,
including the limit for endotoxins (not more than 0.25 IU of endotoxin per mL), but the water-treatment
process used may be different. Quality standards for this type of water meet the same standards as for
water for injections but he production methods are considered less reliable than distillation and thus it is
considered unacceptable for use as water for injections. Current production methods include, for
example, double-pass reverse osmosis coupled with other suitable techniques such as ultrafiltration and
deionization [101,102]. Edol´s facility uses reverse osmosis and deionization.
2.3.2.2. Potassium Chloride, Magnesium Chloride.6H2O and Calcium Chloride.6H2O
One of the many tears function is to provide the entire ocular surface with a moist environment with the
appropriate electrolyte composition. The ocular surface has a narrow range of pH, osmolality, and ionic
concentrations necessary for optimal function. Small changes in these variables, especially osmolality
and ion concentration lead to ocular surface disease. Tear osmolality is derived from the ionic
composition of tears, which is unique when compared to plasma or other body fluids. Tears contain Na+,
K+, Cl- , HCO-, Ca2+ and Mg2+, and trace levels of other ions. Tears have higher K+ and Cl- concentration
35
and similar Na+ concentration compared to plasma [103]. The selected salts were chosen to establish
an electrolyte composition similar to the tear fluid.
2.3.2.3. Boric Acid and Sodium Tetraborate
Gram-negative rods represent some of the most common contaminants of commercially produced eye
care products. Many of these products are formulated in a phosphate buffer preserved with
benzalkonium chloride. Others, designed for use with hydrogel contact lenses, are preserved with
thimerosal, chlorhexidine, or sorbic acid. Inexpensive and widely available, these eye care products are
often subjected to widespread consumer abuse. Their preservative systems are weak in comparison to
those of topical products because of the extreme sensitivity of ocular tissues to chemicals in general.
Studies have shown the borate-buffered vehicle appears better suited for use in ophthalmic products
than phosphate-buffered formulations. The lack of proliferation and survival of microorganisms in a
borate-buffered vehicle provides an increased level of safety for ophthalmic products. However, this
preservative system does not, by itself, meet the criteria for effectiveness. Thus, the borate buffer
requires augmentation by other agents to meet the specifications [86].
2.3.2.4. Suttocide (N-hydroxymethylglycinate 50%) and EDTA
Suttocide is a colourless to yellow liquid with characteristic amine odour sold as a 50% aqueous solution.
Highly soluble in water, methanol propylene glycol, glycerine, but insoluble in most organic solvents.
This preservative releases formaldehyde when decomposed in an aqueous solution. One molecule of
formaldehyde is formed by the decomposition of each molecule of sodium hydroxymethylglycinate, as
described in the equation below [92].
(2)
It is a broad-spectrum preservative with good antimicrobial properties, active against Gram-negative,
Gram-Positive bacteria, yeast and molds. It has good anti-bacterial and anti-mold properties, but it is
weak against yeast. Multidose artificial tears containing the mixture of 0.002% sodium hydroxymethyl
glycinate (SHMG) and 0.10% EDTA as a novel preservative have been marketed in Italy recently. Since
these solutions are claimed to have a mild impact on the ocular, an evaluation was made whether
different concentrations of SHMG alone or in combination with EDTA could be an efficient preservative
for different ophthalmic preparations. It was demonstrated that SHMG shows a wide bacteriostatic
activity at concentrations ranging from 0.0025% to 0.0125% and that higher concentrations of SHMG
(0.25% or 0.5%) are needed for the fungistatic activity against C. albicans and A. brasiliensis. The
addition of 0.10% EDTA to SHMG substantially reduces the minimal concentration of SHMG required
for both bacteriostatic/fungistatic and bactericidal/ fungicidal activity in vitro. In fact, in this condition,
0.005% and 0.025% concentrations of SHMG were, respectively, sufficient to inhibit growth or to kill all
tested organisms [84]. The addition of EDTA has broaden even more the preservative antimicrobial
36
spectrum. These compost is a stabilizing and chelating agent which helps in restricting metal-catalyzed
oxidation of various drugs and has been proven to promote the action of antimicrobial preservatives
[104].
2.4. Manufacturing Process of the Final Formulation
The manufacturing process of the final formulation starts with the introduction of highly purified water in
an appropriate recipient and the addition of sodium hyaluronate stirred until complete dispersion. The
description of this manufacturing process applies to the product HA 0.15% and HA 0.30%. Afterwards
occurs the addition of potassium chloride, magnesium chloride hexahydrated and calcium chloride
hexahydrated and stirred again until complete dissolution. Sodium chloride, boric acid, sodium
tetraborate and EDTA are added and stirred at the same conditions as before or until complete
dissolution. The final step is the addition of the preservative Suttocide (N-hydroxymethylglycinate 50%),
which is stirred, once again, until complete homogenization. The pH and the osmolality levels are
adjusted with NaOH 40% and HCl 10%, and sodium chloride, respectively.
2.5. Stability Testing
Samples of HA 0.15% and HA 0.30% were analysed for macroscopic appearance, pH, osmolality,
viscosity, density and sodium hyaluronate identification before the storage period and on months 1, 3
and 6 of storage. The six batches, three of each product, remained within specifications after 6 months.
2.6. Scale-Up
Once the final formulation and its respective manufacturing process is defined, the scale-up project was
studied. Scale-up is generally defined as the process of increasing the batch size and can also be
viewed as a procedure for applying the same process to different output volumes. However, batch size
enlargement does not always translate into a size increase of the processing volume. In mixing
applications, scale-up is undeniably concerned with increasing the linear dimensions from the laboratory
to the plant size. It is then necessary and very important to understand the type of procedure applied, in
which an increase of the scale may be counterproductive and scale-down is required to improve the
overall quality of the product. The passage between R&D to production scale, it is essential to have an
intermediate batch scale called the pilot scale, which is defined as the manufacturing of product by a
procedure fully representative of the manufacturing scale. Nevertheless, inserting an intermediate step
between R&D and production scales does not in itself guarantee a successful transition. Despite of a
well-defined process that may generate a product according to the specifications in both the laboratory
and the pilot plant, it may fail quality assurance tests in production [105].
In order to avoid the repetition of lengthy and costly tests, it is necessary to gather information during
properly designed development and process optimisation studies, when scaling up from laboratory
through pilot to production scale. Such information provides the basis for justification that scale-up can
be achieved without a consequent loss in quality. Therefore, the aim of the pilot batch scale is to provide
predictive data of the production scale product. It should be notice that the pilot batch size should
correspond to no less than 10% of the production scale batch, at least [106].
37
A flow chart of HA 0.15% and HA 0.30% solutions pilot batch scale production is represented in Figure
11.
With this manufacturing process, a detailed Risk Analysis study was made in order to identify possible
risks which may compromise the final quality of the product (Chapter Three – Risk Analysis). The final
product resultant from this manufacturing process was subjected to physical-chemical studies to
determine the characteristics that the product possesses (Chapter Four – Pilot Scale Batches Process
Validation).
Figure 11- Flow chart of the pilot batch manufacturing process of HA 0.15% and HA 0.30%.
38
2.7. Conclusion
In 2015 it was developed a formulation that allowed the determination of suitable excipients to produce
an eye drop solution for dry eye treatment. During the product development a study of the possible
excipients used in eye drops formulation was performed, were three main groups where the primary
focus: the choice of the polymer, the preservative and the buffer. After some research and with the
development of the laboratory batches it was decided that the final formulation contains the polymer HA,
the preservative Suttocide and EDTA and boric acid and sodium tetraborate as a buffer. With the final
formulation decided the scale-up of the product was designed. Along the course of the production a Risk
Analysis was made and with the products resulting from the scale-up production an IPC and FPC study
was performed among with the characterization of the final product. These studies resulted in the
following chapters.
39
Chapter Three – Risk Analysis 3. Risk Analysis - Introduction
Ophthalmic products, such as eye drops, require well-defined specifications. Since the eye is a sensitive
organ it is necessary that the eye drop’s formulation must present physical-chemical characteristics
similar to the lacrimal fluid and it must be sterile, in order to avoid eye damages and infections [74].
During the manufacturing process the product may be exposed to various factors that may put at risk
its final quality. The manufacturing of sterile products requires special manipulation in order to minimize
risks of microbiological contamination, which depends on the training and skills of the personal involved
[107]. By applying risk analysis in the manufacturing process a better understanding of what may be the
primary causes that may contribute to the production of a non-conform product can be achieved,
allowing the management of potential risks [71].
The creation of risk management tools, especially tools for a specific product, encourages and
demonstrates the importance of using risk analysis in the manufacturing process of a product. ISO
14971:2017 establishes the requirements for risk management to determine the safety of a MD by the
manufacturer during the product life cycle [70].
Failure Mode Effect Analysis (FMEA)
FMEA is one of the most commonly used methods for pharmaceutical risk assessment, it is a step-by-
step approach for identifying all possible failures in design, manufacturing or assembly process, or a
product. It provides for an evaluation of potential failure modes for processes and their likely effect on
outcomes and product performance. Once the specific operation is studied, failure modes are
established, by implementing risk reduction the potential failure is may be eliminated, contained,
reduced or controlled. It is a team-based structure risk assessment method that can assign a numerical
Risk Priority Number (RPN) based on the severity of the risk, consequently this method is dependent
on the expertise of the team members. This method is more applied to processes that do not have
several sub processes. Failures should be prioritized according to how serious their consequences are,
how frequently they occur and how easily they can be detected. Action to eliminate or reduce failures
should begin with those with the highest priority [67,68,108,109].
It consists in building a table with specific columns such as:
Potential Failure Modes Colum: List of potential failures that may occur in a specific situation.
Modes can be broken down into the following categories: Total failure, partial failure, intermittent
failure, over-function and unintended function;
Effects of Failure Column: Description of the possible effects of the failure;
Severity (SEV) Column: The severance of the failure. Severity is a numeric ranking of the
seriousness of the failure. The number shall be assigned using the definition given in the FMEA
Score Sheet. The FMEA Score Sheet is a table containing the ranking criteria for the SEV, OCC
and DET;
Potential Cause of Failure Column: What are the causes or mechanisms of the failure;
40
Occurrence (OCC) Column: The frequency of the failure. Occurrence is a numeric ranking of
the probability of the cause for the failure occurring, also assign by the FMEA Score Sheet. The
occurrence is evaluated relative to the likelihood of the failure occurring when it is caused;
Control Column: Current system controls on place to pervert the failure mode, there two types
of design controls: Prevention or Detection of the cause/mechanism of failure or the failure
mode. The Prevention control prevents the failure from occurring or reduce the rate of
occurrence. The Detention control leads to a corrective action when the failure occurs.
Detection (DET) Column: The likelihood of the failure detection. Detection is a numeric ranking
of the ability of the design to detect a potential cause/mechanism and subsequent failure mode,
it is also assign by the FMEA Score Sheet.
Risk Priority Number (RPN): Number that results in the multiplication of SEV, OCC and DET
scores, the resulting number helps prioritize risks our actions for problem resolution.
𝐑𝐏𝐍 = 𝐒𝐄𝐕 × 𝐎𝐂𝐂 × 𝐃𝐄𝐓 (3)
Criticality Index (CRIT): Index that helps prioritizes risks or actions when the RPN is equal, given
greater emphasis to the SEV and frequency OCC.
𝐂𝐑𝐈𝐓 = 𝐒𝐄𝐕 × 𝐎𝐂𝐂 (4)
Analysis and Recommended Corrective Actions Column: After the calculation of all RPNs and
CRIT, recommended actions should be taken in order to reduce the overall RPN for failure
modes that era considered unacceptable or intolerable.
ISO 14971:2007 — Application of Risk Management to Medical Devices
ISO 14971:2007 specifies in identifying risks associated with MD, including in vitro diagnostic MD, to
estimate and evaluate the associated risks, to control these risks, and to monitor the effectiveness of
the controls. The requirements given in this norm are applicable to all stages of the life-cycle of a MD
and do not apply to clinical decision making [70].
Regarding Risk Management, this norm describes in detail the general requirements for risk
management process thought out the MD life cycle, which includes risk analysis, evaluation, control and
risk review in the production and post-production phases. It is given guidance in how to use risk concepts
important for managing the risks associated with the product and an example of a qualitative analyses
[70].
3.1. Materials and Methods
3.1.1. Materials
Sodium hyaluronate was a kind gift from Inquiaroma and N-hydroxymethylglycinate 50% (Suttocide)
from Ashland (EUA). Potassium chloride, sodium chloride, sodium tetraborate and EDTA were
purchased form VWR International (Portugal), calcium chloride.6H2O was purchased from José Manuel
Gomes Santos (Portugal), magnesium chloride.6H2O was purchased from Sigma Aldrich Quimica
(Portugal) and acid boric was purchased from LaborSpirit (Portugal).
41
3.1.2. Methods
The steps to design an FMEA method and a Qualitative Analysis according to the ISO 14971:2007 are
described below.
3.1.2.1. Process Description
A thorough study of the process benefits in identifying the potential risks and failures associated with
the quality of the final product. A flow diagram presented in Chapter Two, section 2.6, was used to
facilitate the identification of the potentials failure modes and to understand the whole process.
The qualitative composition of HA 0.15% and HA 0.30% eye drop solution as well the function of each
excipient are described in Table 7. The equipment used in the manufacturing process are also listed
below in Table 8.
Table 8- Main equipment used during the manufacturing process of HA 0.15% and HA 0.30% eye
drop solution.
Equipment Brand Model
Mixer SEITE-WERKE DB 110 A FW
Filling and encapsulating IMA F57
Sealing test Bonfiglioli PKV 212
HPLC Hitachi LaChrom Elite
pH meter Metrohm 827 pH LAB
Osmometer Gonotec Osmomat 030
Osmomat 3000
Viscometer Brookfield DV-II + PRO
Laminar flow cabinet ADS Laminaire Optimale 12
3.1.2.2. Identification of Critical Points
Six process steps were chosen as critical points of the process since a risk in these steps may
compromise the product quality. For each process step, numerous risks were identified.
3.1.2.3. Failure Cause and Effect of each Failure Mode
Potential causes and effects for each failure mode were described. Each step, except In Process
Control/Finished Product Control (IPC/FPC) and Storage, were subjected to a detailed study by
designing Ishikawa diagrams.
42
3.1.2.4. Severity Rating Assignment
A Severity rating was given to each effect from 1 to 10, with 10 being the most severe. This classification
was made by Edol’s collaborators from surveys design for each step. The final results correspond to the
mean, standard deviation (SD) and a weighted arithmetic mode of the collaborators' answers.
3.1.2.5. Occurrence Rating Assignment
The failure frequency was determined and rated appropriately from 1 to 10, being 10 the most likely,
after collecting data on the factors responsible for the failure. The classification was prepared the same
way as described previously in section 3.1.2.4 Severity Rating Assignment.
3.1.2.6. Detection Rating Assignment
A list of all controls currently existing in order to prevent each failure from occurring was prepared and
a Detection rating was assigned for each failure from 1 to 10, being 10 being a low likelihood of
Detection. The classification was prepared the same way as described previously in section 3.1.2.4
Severity Rating Assignment.
3.1.2.7. Risk Priority Number Calculation
RPN was calculated by multiplying the Severity rating by the Occurrence rating by the Detection rating.
3.1.2.8. Failure Modes Prioritization
The prioritization of the failure modes identified was dependent on their RPN number. The higher the
RPN number, the higher the risk. To distinguish failures which RPN is equal the CRIT number was
calculated by multiplying the Severity number by the Occurrence number.
3.1.2.9. ISO 14971:2007 Qualitative Analysis
Qualitative Analysis was performed using ISO 14971:2007 Annex D, section D.3.4.1 [70]. An N-by-M
matrix approach was used to describe the probabilities and severities of the risk associated with each
hazardous situation, defining N levels of probability and M levels of severity. To classify each failure
modes in terms of severity and probability levels, the common terms of the table below were applied
(Table 9).
Table 9- Qualitative Severity and Probability Levels.
Level Common Terms Possible description
Severity
Significant Death or loss of function or structure
Moderated Reversible or minor injury
Negligible Will not cause injury or will injure slightly
Probability
High Likely to happen, often, frequent
Medium Can happen, but not frequently
Low Unlikely to happen, rare, remote
43
3.1.2.10. Corrective Actions
For each failure mode was determined the action to be taken in order to reduce or eliminate the risk.
3.1.2.11. Risk Review
After the application of the corrective action, the classification of the Severity, Occurrence and Detection
rating was performed once more with the purpose of determining if the failures suffered a general
reduction or elimination.
3.2. Results
3.2.1. Ishikawa diagrams
Six steps of the manufacturing process were identified as potential critical points to ensure quality of the
final product. The chosen steps were Weighting, Mixing, Filtration, IPC/FPC, Filling/Packing and
Storage. It was decided to use Ishikawa diagrams to help with the brainstorming and to list potential
causes that may induce errors in each specific step. Four primary causes were used to construct the
diagrams, following the 4M’s: Manpower, Methods, Machinery and Material. An adaptation was made
to Operators, Methods, Equipment and Raw Material. However, two steps were not subjected to the
diagrams, IPC/FPC and Storage, since the four primary causes are not applicable. In the case of
IPC/FPC, the risks identified were based on various possible results from the analysis performed.
Possible risks that may occur on the Storage step were obtained from Patel et al. [107] and through
experience from Edol’s collaborators. The Ishikawa diagrams are presented in Appendix Figure A 1,
Figure A 2, Figure A 3 and Figure A 4.
3.2.2. FMEA of Hyaluronic Acid Pilot Batch Manufacturing Process
After brainstorming possible risks from each process and constructing ishikawa’s diagrams a FMEA
table was created. In order to classify each risk presented, key activities, such as, understanding the
impact of the risk, ranking the significance of risk by scoring 1 to 10 (Table 10) and the calculation the
RPN and CRIT were performed during risk analysis.
44
Table 10- Severity, Occurrence and Detection table with respective risk classification.
In Table 10 it is also present the classification regarding the ISO 14971:2007 Qualitative Analysis. It was
decided that values between 1 and 3 are considered as Negligible and Low, between 4 and 6 as
Moderate and Medium and between 7 and 10 as Significant and High in the classification of the Severity
and Occurrence aspect, respectively.
For the statistical treatment of the results obtained from the surveys it was decided to do the mean, SD
and a weighted arithmetic mode of the responses. In the weighted mode a score or weighting factor was
given to each collaborator depending on their academics level and years of work experience. In Table
11 it is shown the specification needed for each weighting factor.
Table 11- Weighting Factor and respective specification according to the years of work
experience and academic level.
Weighting Factor Years of Work Experience Academic Level
3 ≥ 3 College Degree
2 < 3 College Degree
≥ 3 Technical Course
1 < 3 Technical Course
Severity Occurrence Detection
Le
ve
l
Effect Criteria:
Severity of effect
ISO
14
97
1
Le
ve
l
Probability of failure
Criteria: Likelihood
of occurren-
ce ISO
49
71
Le
ve
l
Detection
Criteria: Likelihood
of detection
10 Dangerously
High Safety regulatory
consequences
Sig
nific
an
t
10 Almost certain Failure is
almost inevitable
Hig
h
10 Absolute
uncertainty
No design control or
no detection
9 Extremely
High 9 Extremely 9
Very remote
Very remote
detention
8 Very high
High degree of dissatisfaction
of Quality
8 Very high Repeated
failures
8 Remote Remote
detention
7 High 7 High 7 Very low Very low detention
6 Moderate
Mo
de
rate
6 Moderate to
high
Occasional failure
Me
diu
m
6 Low Low
detention
5 Low Slight
Dissatisfaction of Quality
5 Moderate 5 Moderate Moderate detention
4 Very Low 4 Low to
moderate 4
Moderately high
Moderately high
detention
3 Minor Minimum
effect
Neg
ligib
le 3 Low
Relatively few failures
Lo
w
3 High High
detention
2 Very minor 2 Very low 2 Very high Very high detention
1 None No effect 1 Remote Failure is unlikely
1 Almost certain
Almost certain
detention
45
In order to obtain the maximum score, the collaborator needs to have a college degree and work
experience equal or over 3 years. If the collaborator has a college degree but work experience under 3
years, the given score is the same as a collaborator with a technical course with work experience equal
or over 3 years. The minimum score is given to a collaborator who has a technical course and work
experience under 3 years.
The weighted mode was created as a necessity to distinguish knowledge background from the
collaborators and to give more value to those who have a college degree, since collaborators with a
higher education understood and responded to the surveys more easily and with more carefully chosen
answers then those who did not possessed that type of education. If the results were only analysed
trough simple arithmetic mean and SD the results would not make sense in cases which the SD was
too elevated. For that reason the mode is preferred over the mean when describing categorical data.
The greatest frequency of responses is important for describing categorical data because, for instance,
classifying a risk with severity 3 is very different than classifying as 4, the first one is considered to have
a minimum effect in the quality of the product as the second, demonstrating slight dissatisfaction. Twelve
collaborators responded to the surveys tacking into account their function in the manufacturing process.
The FMEA and the ISO 14971:2007 qualitative analyses of the manufacturing process of HA 0.15% and
0.30% is presented in the Appendix in Table A 1, with the calculation of the RPN and CRIT for each
failure mode detected.
After the calculation of the risk score, the results of the level or priority of the risk were estimated using
a Pareto chart (Figure 12). Using the RPN scores of the failure modes, they may be ordered from who
possesses the highest risk to the lowest risk. The Pareto chart facilitates the visualization of the different
risks degrees, which helps the identification of the highest risks and the decision making of in which
value of RPN is considered to be a high level risk.
46
Figure 12- Pareto chart of every failure mode of the manufacturing process of HA 0.15% and HA 0.30%.
0,00%
10,00%
20,00%
30,00%
40,00%
50,00%
60,00%
70,00%
80,00%
90,00%
100,00%
0
20
40
60
80
100
120
RP
N
Failure Modes
Risk Prioritization
Accumulated Percentage
47
Form the analysis of the Figure 12 it was decided that the values in which the risk is considered to be
unacceptable starts at RPN 80 and intolerable at RPN 300 (Table 12). Over this RPN value every failure
mode needs to be mitigated, thus, risks below this value are considered to be acceptable risks. From
this decision the construction of a RPN matrix was possible. The ISO 14971:2007 Qualitative Analysis
already presents a risk matrix, using the probability as rows and the severity as columns a 3x3 risk
matrix which is represented in Table 13. This classification is already present in the FMEA table.
Table 12- RPN matrix.
Seve
rity
1 1 2 3 4 5 6 7 8 9 10 1
Detectab
ility
2 4 8 12 16 20 24 28 32 36 40 2
3 9 18 27 36 45 54 63 72 81 90 3
4 16 32 48 64 80 96 112 128 144 160 4
5 25 50 75 100 125 150 175 200 225 250 5
6 36 72 108 144 180 216 252 288 324 360 6
7 49 98 147 196 245 294 343 392 441 490 7
8 64 128 192 256 320 384 448 512 576 640 8
9 81 162 243 324 405 486 567 648 729 810 9
10 100 200 300 400 500 600 700 800 900 1000 10 1 2 3 4 5 6 7 8 9 10
Occurrence
Acceptable Risk Unacceptable Risk Intolerable Risk
Table 13- ISO 14971:2007 risk matrix [70].
Severity
Negligible Moderate Significant
Occurrence
High High Risk
Medium Low Risk
Low
48
3.2.3. Failure Modes Prioritization and Corrective Actions
After the classification of risk level for each failure mode according to the RPN matrix and risk matrix
tables, prioritization of all the detected risks was made. Using the FMEA method the failure modes were
distinguished by their RPN values and CRIT values, the higher the RPN and the CRIT value the higher
the risk. In some cases the CRIT values were equal, in this case the highest risk chosen was the one
who presented a higher severity value. The risks obtained through the ISO 14971:2007 are not possible
to distinguish, since the only criteria available is the classification, which is not a numeric value. The
FMEA and the ISO 14971:2007 methods have detected two risks in common, however the FMEA
detects two more unacceptable risks comparable to the ISO 14971:2007 method. Corrective actions in
order to reduce the risk of each failure mode were applied.
Table 14- Unacceptable risks obtained through FMEA and current and corrective action.
FMEA RPN CRIT Current Process Control Corrective Action
Elevated Microbial
concentration/ non-
sterile (IPC and FPC) 108 36
The bioburden is determined/ Sterility test
Bioburden of the Validation Batch top and bottom at 0h and
24h after the manufacture
Procedure failures
(Weighting) 84 28 Required filling of the Batch
Master Record Triple check of all weights
Velocity Failure
(Mixing) 84 21 Backup shaker Visual verification of the
dissolution and registration of the same
Misuse use of uniform
(Mixing) 84 21 Uniform procedure Extra environmental controls
Product harvest in an
inappropriate container
(IPC and FPC) 81 27 The bioburden is determined
Creation of a product harvesting procedure where the type of
flask to be used is explicit
False Results (IPC and
FPC) 80 20 All methods are previously
validated Verification by the Quality
Control manager
Raw material loss
during transference
(Mixing) 80 20 Proper training
Specific training for the manufacture of the product /
Monitoring of the manufacture by a technician
Contaminated Filter
(Pre and Sterile
Filtration) 80 16
Sterilization of the filter in an autoclave with a valid load whenever there is a filling
Verification by the Production manager
IPC/FPC - In Process Control/Finished Product Control
49
Table 15- Unacceptable risks obtained through ISO 14971 qualitative analyses and current and
corrective action.
ISO 14971 Classification Current Process Control Corrective Action
Elevated Microbial
concentration/ non-
sterile (IPC and FPC) High Risk
The bioburden is determined/ Sterility test
Bioburden of the Validation Batch top and bottom at 0h and
24h after the manufacture
Procedure failures
(Weighting) High Risk Required filling of the Batch
Master Record Triple check of all weights
High/Low pH (IPC and
FPC) High Risk
pH meter is calibrated daily. pH specified in batch record and pH
readings are double verified. After all adjustments Quality Control
confirms
Identification of an OOS. An investigation must be performed
High/Low osmolaliry
(IPC and FPC) High Risk
Osmometer is calibrated monthly against osmolality patterns
solutions (purified water, 300 mOsm/kg and 150 mOsm/kg). It
is calibrated daily against a reference solution with 300
mOsm/Kg. Osmolality specified in batch record and osmolalities
readings are double verified. After all adjustments Quality Control
confirms
Identification of an OOS. An investigation must be performed
High/Low Viscosity
(IPC and FPC) High Risk
Viscometer is calibrated annually by a certified external entity.
Viscosity specified in batch record and viscosities readings are
double verified. After all adjustments Quality Control
confirms
Identification of an OOS. An investigation must be performed
Sodium Hyaluronate
UV Assay out of range
(IPC and FPC) High Risk
Sodium Hyaluronate percentage specified in batch record is
determined by a validated UV/Vis method. Assay is performed
twice.
Identification of an OOS. An investigation must be performed
IPC/FPC - In Process Control/Finished Product Control; OOS - Out of Specification.
3.2.4. Risk Review
The implementation of the corrective actions for each failure mode has the aim of reducing the SEV,
OCC or the DET parameters. All the failure modes detected by the FMEA method and the first two
detected by the ISO 14971:2007, the corrective actions effects the DET value, since they present a
detective nature. In the risk detected in the Mixing Process by the FMEA method “Raw material loss
during transference” the implemented corrective action not only effects the DET value but also the OCC,
since implementing specific training diminishes the probability of occurrence. The rest of the risks
identified by the ISO 14971:2007 are not possible to implement corrective action since they are
analytical results. When a result is an OOS it is not expected, meaning that the value obtained does not
obey the respective specification, in this case an investigation must be opened. This investigation should
result in a well-documented and exhaustive report in order to determine the causes of the result.
50
The ISO 14971:2007 quantitative analysis does not suffer alteration because it only classifies the risk
according the severity and the occurrence. In the FMEA method by reducing the DET and the OCC
parameter by one value the RPN was once again calculated, the results are presented in Table 16.
Table 16- RPN calculation after the implementation of the corrective actions.
FMEA SEV OCC DET Previous
RPN New RPN
Reduction (%)
Elevated Microbial concentration/ non-sterile (IPC and FPC)
9 4 2 108 72 33
Procedure failures (Weighting) 7 4 2 84 56 33
Velocity Failure (Mixing) 7 3 3 84 63 25
Inadequate clothing (Mixing) 7 3 3 84 63 25
Product harvest in an inappropriate container (IPC and FPC)
9 3 2 81 54 33
False Results (IPC and FPC) 10 2 3 80 60 25
Raw material loss during transference (Mixing)
5 3 3 80 45 44
Contaminated Filter (Pre and Sterile Filtration)
8 2 4 80 64 20
All the new calculated RPN are below 80 meaning that the previous unacceptable risks were reduced
to acceptable risks after the implementation of the corrective actions.
3.3. Discussion
The risk classification responses on the surveys among the collaborators presented some discrepancy
due to the fact some collaborators have shown more difficulty in answering the surveys than others.
This variance of values resulted from the fact that not all collaborators have the same background
education nor the same work experience. Other important factor was the risk analysis knowledge that
some collaborators possessed, while others did not. This discrepancy of responses in some cases led
to a very high SD, for example, the failure mode “Filling material with defects” the DET parameter present
a SD of 4.2 (Appendix Table A 1). In risk classification a 4 value difference has different meanings which
alter the type of risk. It was decided to do a weighted mode in which some collaborators responses had
more weight than others, depending on their education and work experience (Table 11). The final value
of each parameter, SEV, OCC and DET is the most frequent response of the survey with the respective
given weight. These values were used to calculate the RPN in the FMEA method and to classify the
type of risk using the ISO 14971:2007 qualitative analysis (Table 13). Thought the RPN it was possible
to order the failure modes from which represents a greater risk to the lowest risk by constituting a Pareto
chart. This exercise was necessary in order to decide the limit of RPN from which all failure modes are
considered unacceptable. In Figure 12 it is visible que maximum value of RPN and de minimum, 108
and 9 respectively. Various failure modes presented equal RPN, being the most common the value 32
with seventeen failure modes and the less common 108 with only one failure mode. The first eight at
left bars stand out from the chart in comparison with the other bars with RPN of 108, 84, 81 and 80. This
51
highlight was the decision maker in defining the limit in which a risk should be considered as
unacceptable, this decision was fundamental to determinate which failure modes should be submitted
to corrective actions. For that reason, it was decided that a risk is considered to be unacceptable at RPN
80 and intolerable at 300 (Table 12). This exercise was not necessary to do to the ISO 14971 method
since it already presents a risk matrix (Table 13).
The results have shown that the FMEA method detected eight unacceptable risks and the ISO 14971
only detected six with two risk in common with the FMEA method. The FMEA method not only detected
more risks, but detected in four different processes (Weighting, Mixing, IPC and FPC and Pre and Sterile
Filtration), while the ISO 14971 method only detected in two processes (Weighting and IPC and FPC).
The RPN and the CRIT values helps understand the difference between unacceptable risks since it
uses numeric values, while the ISO only distinguishes high from low risks. We can argue that the FMEA
is a more sensible analysis than the ISO since it detects more unacceptable risks in more processes
and the differences between them (Table 14 and Table 15).
The next step was to implement corrective actions to all the unacceptable risks detected both by the
FMEA and the ISO method (Table 14 and Table 15). Once again, the aim of the implementation is to
reduce the RPN of the detected risks by mitigate the classification given to the SEV, OCC and DET
parameters. All the corrective actions given to the FMEA method are of a detective nature except one
who also has occurrence nature, meaning all the DET and one OCC parameter suffered a reduction. In
the ISO 14971 method the risks in common with de FMEA method present the same corrective action,
since the corrective action only affects the DET parameter. The risk classification given to the failure
mode trough this method does not suffers alteration, because the ISO only studies the SEV and the
OCC parameters. The rest of the unacceptable risks detected by the ISO are all from the IPC and FPC,
being analytical results. When an analytical result out of specification is obtained and it is not expected
it is classified as an OOS and an investigation must be performed in order to determine the cause of the
result. The final report of the investigation must refer if the anomaly is due to problems during the quality
control analysis, the manufacturing process or from the quality of the raw material. Thus, implement
corrective action in order to reduce the probability of occurrence in future analysis is necessary [110].
The problem with this product is that is new, and it was never manufactured before in Edol’s facilities,
meaning that there are no documents with previous OOS and respective corrective actions since an
investigation was never performed. For that reason, it is very difficult or impossible to implement a
corrective action since the possibilities are infinite. The corrective actions in this case are dependent of
the process in which the failure occurred, examples of possible corrective actions are maintenance of
the equipment used in a specific process and increase the anticipation of their maintenance, adequate
training and sensitization of the operators, validation of the mixing time, reanalysis of the raw material
or reformulation of the product.
The FMEA method is therefore the only method possible to perform a risk review, by implementing the
corrective actions the DET and the OCC value was reduced by one value. It was decided to reduce only
one value since these corrective actions are not extreme actions. By calculating once again the RPN
the once before unacceptable risk are now considered to be acceptable risks. The reduction in all cases
were less than 50% (Table 16) due to the fact that the previous unacceptable risks presented values
52
close to the minimum limit of the Unacceptable Risk classification (Table 12), therefore it was not
necessary to do a dramatic corrective action in order to the risk become acceptable.
3.4. Conclusion
Risk Analysis is a complex and time consuming process with the necessity to understand in detail every
aspect of the process in question. In the case of the HA 0.15% and 0.30% an intensive study had to be
performed in every process, necessary to the production of the product in order to detect all the possible
risks that may occur. Two tools were chosen to detect and evaluate failure modes existent in the
manufacturing process of HA 0.15% and 0.30%, the FMEA and the ISO 14971:2007 Qualitative
Analysis. From the above evaluation of risk assessment based on FMEA and the ISO 14971:2007
Qualitative Analysis, the majority of the detected risks were considered to be acceptable risks, were the
FMEA method detected approximately 14% unacceptable risks and the ISO 14971:2007 detected 11%.
It was concluded that the FMEA method is more sensitive and precise than the ISO method, detecting
more risks in more processes and with the capacity to distinguish the differences between risks, which
makes it possible to understand the specific seriousness of each risk. Through the FMEA method is was
possible to implement corrective actions and mitigate every unacceptable risk detected, actions that the
ISO was not capable to do. By using two tools in one process it was possible to conclude that different
tools give different answers. Although the ISO 14971:2007 is an application of risk management to MD,
the FMEA demonstrated to be a most suitable risk management tool in this product.
53
Chapter Four – Pilot Scale Batches Process Validation 4. Pilot Scale Batches Process Validation - Introduction
Validation is the act of demonstrating and documenting that a procedure operates effectively. Process
validation ensures and provides documentary evidence that the concerned processes, within their
specified design parameters/specifications, is capable of consistently producing a finished product of
the required quality. Therefore, it is essential to have a deep understanding of the process in order to
perform a risk analysis and identify the critical steps that may be incurred on the product quality [111].
Pilot batch size may be used in the process development or optimization step, to support formal stability
studies and also to support pre-clinical and clinical evaluation. It should correspond to 10% of production
scale and provides the data predictive of the production scale product. After the results from the
Validation Plan, it may be necessary optimise the manufacturing process. Therefore, the pilot batch is
the link between process development and industrial production of the product. The purpose of the pilot
batch is to analyse and evaluate the difficulties and critical points of the manufacturing process and to
determine the most appropriate large-scale production, providing a high level of insurance that the
product will have the best quality and that the process will be feasible on an industrial scale [111].
The HA 0.15% and HA 0.30% were never before produced in Edol’s facilities being the validation plan
an obligatory requirement confirm if the manufacturing process is suitable or if it needs some
adjustments.
4.1. Materials and Methods
4.1.1. Materials
The materials used are described in Chapter Three – Risk Analysis, section 3.1.1.
4.1.2. Methods
4.1.2.1. Production of three batches of Hyaluronic Acid 0.15% and HA 0.30%
The manufacturing process of 3 batches of 50 L of each product starts with the introduction of highly
purified water in the mixer (DB 110 A FW, Seite-werke, Germany) and the addition of sodium
hyaluronate, stirred at 680 rpm during 120 minutes or until complete dispersion. Afterwards occurs the
addition of potassium chloride, magnesium chloride hexahydrated and calcium chloride hexahydrated
and stirred at the same velocity during 10 minutes or, again, until complete dissolution. Sodium chloride,
boric acid, sodium tetraborate and EDTA are added posteriorly and stirred at the same conditions as
before or until complete dissolution. The final step is the addition of the preservative Suttocide (N-
hydroxymethylglycinate 50%), which is stirred at 680 rpm during 15 minutes or, once again, until
complete homogenization.
4.1.2.2. Appearance
The macroscopic appearance was visually analysed by placing a sample in an appropriate container
and observed at room light.
54
4.1.2.3. Determination of the pH values
The materials used are described in Chapter Two – Previous Work, section 2.1.2.2.
4.1.2.4. Determination of the Osmolality values
The materials used are described in Chapter Two – Previous Work, section 2.1.2.3.
4.1.2.5. Determination of the Viscosity values
The materials used are described in Chapter Two – Previous Work, section 2.1.2.4.
4.1.2.6. Determination of the Density values
The pycnometer was tarred with the cap and afterwards filled with the sample. The excess was removed
when the cap was introduced. Then the pycnometer was weighted, and the value recorded, the value
of the density corresponds to the radio between the sample weight and the pycnometer calibration value.
The density value obtain is only informative and the test is made at room temperature.
4.1.2.7. Bioburden
The bioburden test was performed in a laminar-air-flux cabinet (The Baker Company SG503A-HE,
USA), using a filtration system (Milifex Plus, Merck, Germany). The cups utilized were a ready-to-use
sterilized filtration devices for microbial enumeration with a 0.45 µm PVDF membrane (EZ-Fit™ Filtration
Unit, Merck, Germany). A volume of 100 mL of sample was transferred into the cup and filtered
immediately, rinsing the membrane filter with 600 mL of fluid D (Merck, Germany). The membrane was
transferred to the surface of typticase soy agar (TSA, Merck, Germany) and incubated at 20-25°C for 5
days. After the first incubation, the plates were transferred to 30-35ºC and incubated for 2 days. The
test was performed in duplicate and with a negative control.
4.1.2.8. Sodium Hyaluronate Assay
The Sodium Hyaluronate Assay was performed using cetyltrimethylammonium bromide (CTAB)
turbidimetric method in the ultraviolet-visible spectrophotometer (Evolution 201 ÛV-Visible
Spectrophotometer, Thermo Fisher Scientific, USA). A volume of 2 mL of Acetate buffer pH 6.0 and 2
mL of standard solution or sample solution were introduced in tubes properly stoppered. The tubes were
stirred and incubated at 37°C in an incubator (Vaciotem-t, JP Selecta, Spain) in order to synchronize
the reaction temperature. CTBA was added next to each tube and incubated during 6 minutes. The
absorbance of the sample and the standard solution were read in the ultraviolet-visible
spectrophotometer at the absorption maximum of 600 nm, using the blank solution prepared previously.
4.1.2.9. Sterility Test
The sterility test was performed according to the European Pharmacopeia, Chapter 2.6.1. by Membrane
Filtration [112]. Briefly, the sterility test was performed in a laminar-air-flux cabinet (ADS Laminaire
OPTMALE 12, France) using a Steritest™ Symbio Pumps with Steritest™ EZ Canister Device for
Antibiotics (Merck, Germany), because the product is very viscous which difficult the filtration process.
This method is closed filtration method. It starts with prewetting the membranes with the rinse fluid A
55
(Merck, Germany). Next, after the decontamination of the bottles with isopropyl alcohol 70% v/v, the
Steritest needle was inserted into the first product container allowing the product to be transferred to the
canisters, this process was repeated for the rest of the product containers. After the filtration of the
product, the membrane was rinsed with fluid A. Afterwards, one of the canisters was filled with
thioglycollate media (Merck, Germany) and the other with typticase soy broth (TSB, Merck, Germany).
The canister with TSB media was incubated at 20-25°C for 14 days and the canister with thioglycollate
media at 30-35 for 14 days.
4.2. Critical Steps to Control
In the scope of the manufacturing process validation, the steps and the process controls defined as
essential for the evaluation and obtainment of a finished product with the required quality, are defined
on Table 17.
Table 17 - Steps, Process Controls and Acceptance Criteria considered for the validation plan
of HA 0.15% and HA 0.30% eye drops solution, 8 mL.
Step Process control Acceptance Criteria
Mixture Velocity and time of the mixtures
Stir until completely dissolved
In Process Control Holding time validation: 0h3
Appearance The solution must be limpid, transparent and odourless.
pH Between 7.0 and 7.6 (temperature 20-25ºC)
Osmolality Between 280 and 320 mOsm/Kg
Viscosity In study1
Sodium Hyaluronate assay Between 90% and 110% (UV-Vis)
Density Not Apply2
Bioburden < 10 CFU/10mL
Sample collected before the sterilizing filtration Holding time validation: 48h3
Appearance The solution must be limpid, transparent and odourless.
pH Between 7.0 and 7.6 (temperature 20-25ºC)
Osmolality Between 280 and 320 mOsm/Kg
Viscosity In study1
Sodium Hyaluronate assay Between 90% and 110% (HPLC)
Bioburden < 10 CFU/10mL
Sterilizing Filtration Bubble Point In study1
Filling Average weight/volume (filling quantity)
8.0 mL – 8.5 mL
Leak test Bonfiglioli PKV 212 On line leak test Conform
Control of packaging operation
Visual control Conform
Finished Product Control (analytical demand for the product release)
Appearance The solution must be limpid, transparent and odourless.
pH Between 7.0 and 7.6 (temperature 20-25ºC)
Osmolality Between 280 and 320 mOsm/Kg
Viscosity In study1
Sodium Hyaluronate assay Between 90% and 110% (UV-Vis)
Filling Volume 8.0 mL – 8.5 mL
Sterility test Absence of growth 1 These specifications will be determined during the manufacture of the pilot batches. 2 Extra analysis for the determination of the filling average. 3 These values will be determined only for the batches used in the process validation.
56
4.2.1. Acceptance Criteria
These tests aim to control the procedures considered key to the quality of the finished product, allowing
to evaluate the influence of each production step on the finished product. The results of the tests
performed in the scope of the process control should be in accordance with the acceptance criteria,
previously defined for each parameter, in each one of the manufacturing steps and should reflect the
reproducibility of the process.
In order to consider the manufacturing procedure validated, the results must be in accordance with the
defined acceptance criteria in the tests performed, in 3 pilot batches.
4.2.2. Validation Plan
The validation involves the evaluation of certain characteristics of the product in manufacturing steps
considered as potential influents. Therefore, process control test is needed to ensure the conform
specifications of the final product.
4.2.2.1. Preparation of HA 0.15% and HA 0.30% Eye Drops Solution
During the mixing process, the velocity and the mixing time of the mixer at each mixture step are
recorded. The results are evaluated and the influence on the manufacturing process and on the finished
product are defined.
4.2.2.2. In Process Control and Holding Time Validation of HA 0.15% and HA 0.30%
eye drops
After the complete homogenization of the solution in the Mixing Process, two samples of 200 mL for
microbiological control and 25 mL of physicochemical control are collected from the top and the bottom
of the mixer. The obtained results must be within the defined acceptance criteria.
Holding time study can demonstrate how much time is suitable for hold the blend or bulk stage before
processing to the next stage [113]. The defined time is 48h after the manufactured process, reflecting
the worst case scenario of waiting time between the end of the preparation and the start of the sterilizing
filtration. In this test also two sample are collected from the top and the bottom of the mixer, 200 mL for
microbiological control and also 200 mL for the physicochemical control.
These two tests have the same Process Control and the same Acceptance Criteria, represented in Table
17.
4.2.2.3. Control of the Sterilizing Filtration
The bubble point test is determined before and after the sterilizing filtration of the solution, in order to
control the suitability of sterilizing process defined for the manufacturing process. The test is made with
5L of the product and with purified water and it is made 3 times during the process. The filter must
remain intact, and the minimum pressure to rupture the membrane is 50 psi.
4.2.2.4. Filling Control
The Filling Process is controlled by determining the average weight with the determination of the density
of each batch and it must be between 8.0 mL and 8.5 mL.
57
4.2.2.5. Finished Product Control
To evaluate the influence of the manufacturing procedures on the specifications of the finished product,
on the repeatability of the filtration step and on the consistency of the analytical results for the product
release a control of the finished product is made. After the sterilizing filtration and the filling/packaging
process 61 samples of finished product will be collected: 10 from the beginning, 10 from the middle and
11 from the end of the filling process, also a pack of 30 units for placement in the sample room. For a
physicochemical analysis, 15 units were collected (5 units from the beginning, 5 from middle and 5 from
end of the filling) and for the microbiologic analysis, an average sample of 10 units (3 units from de
beginning, 3 units from the middle and 4 units from the end of the filling) for sterility testing. The Process
Control and the Acceptance Criteria are once again represented in Table 17.
58
4.3. Results
The results of the Validation Plan obtained in the IPC and FPC are presented in Table 18, Table 19 and Figure 13 for both the HA 0.15% and HA 0.30%. Three
batches of each product were produced to perform this study and the results in Table 18 correspond to the mean of the top and bottom samples.
The velocity and mixing time were recorded in all three batches from both products and it was concluded they were suitable for the manufacturing process.
Table 18- Results obtained in the IPC and FPC analysis with respective specifications of HA 0.15% and HA 0.30%.
HA 0.15 HA 0.30
Specification PB1 0.15 PB2 0.15 PB3 0.15 PB1 0.30 PB2 0.30 PB3 0.30
IPC
Aspect Limpid,
transparent and odourless
Conformed Conformed Conformed Conformed Conformed Conformed
pH (20°C - 25°C) 7.0 - 7.6 7.16 (23.2 °C ) 7.17 (24.2 °C ) 7.17 (24.2 °C ) 7.32 (20.8 °C ) 7.32 (20.9 °C ) 7.17 (24.2 °C )
Osmolality 280 - 320 mOsm/Kg
308 mOsm/Kg 302 mOsm/Kg 302 mOsm/Kg 304 mOsm/Kg 296 mOsm/Kg 297 mOsm/Kg
Viscosity (20°C - 25°C) Under study 59.4 cP (21.3 °C )
12 rpm 67.4 cP (21.4 °C )
12 rpm 42.7 cP ( 24.5°C )
12 rpm 504 cP (23.6 °C )
0.6 rpm 601.5 cP (22.2 °C )
0.6 rpm 646 cP (24.8 °C )
0.6 rpm
Sodium Hyaluronate Assay
90% - 110% 104.14% 108.45% 103.23% 100.77% 106.49% 105.90%
Bioburden < 10 CFU/10 mL 0 UFC/100 mL 0 UFC/100 mL 0 UFC/100 mL 0 UFC/100 mL 0 UFC/100 mL 0 UFC/100 mL
Density (20°C - 25°C) Not Apply 1.0049 g/mL 1.0051 g/mL 1.0045 g/mL 1.0057 g/mL 1.0059 g/mL 1.0074 g/mL
FPC
Aspect Limpid,
transparent and odourless
Conformed Conformed Conformed Conformed Conformed Conformed
pH(20°C - 25°C) 7.0 - 7.6 7.19 (24.3 °C ) 7.19 (24.3 °C ) 7.21 (23.9 °C ) 7.41 (24.2 °C ) 7.42 (2 4.2 °C ) 7.42 (24.3 °C )
Osmolality 280 - 320 mOsm/Kg
305 mOsm/Kg 298 mOsm/Kg 305 mOsm/Kg 299 mOsm/Kg 301 mOsm/Kg 296 mOsm/Kg
Viscosity (20°C - 25°C) Under study 31100 cP (21.3
°C ) 12 rpm 37933 cP (22.2 °C )
12 rpm 36483 cP (22.1 °C )
12 rpm 655 cP (22.1 °C )
0.6 rpm 537 cP (22.6 °C )
0.6 rpm 520 cP (22.6 °C )
0.6 rpm
Sodium Hyaluronate Assay
90% - 110% 101.35% 100.83% 101.03% 97.32% 99.27% 98.57%
Sterility Test Absence of
growth Absence of
growth Absence of growth Absence of growth * * *
*Results non-existent.
59
Table 19- Results obtained in the Bubble Point and Bonfiglioli test for HA 0.15% and HA 0.30%.
Batch number
Bubble Point HA 0.15%
Bubble Point HA 0.30%
Bonfiglioli HA 0.15% Bonfiglioli HA 0.30%
Before filtration process
(Psi)
After filtration process
(Psi)
Before filtration process
(Psi)
After filtration process
(Psi)
Tested Rejected Rejected
(%) Tested Rejected
Rejected (%)
PB1 56.60 56.90 56.10 56.20 4789 7 0.15 3033 8 0.26
PB 2 57.50 58.60 58.90 56.20 5035 5 0.10 4600 16 0.35
PB 3 55.30 55.30 56.70 58.40 4868 10 0.21 4362 10 0.23
Mean 56.47 56.93 57.23 56.93 4897 7 0.15 3998 11 0.28
SD 1.11 1.65 1.47 1.27 126 3 0.05 844 4 0.06
4.4. Discussion
The aim of the present study was to validate the effectiveness of the manufacturing process by analysing
the physical-chemical and microbiological aspects of the finished product. Three pilot batches of the two
products, HA 0.15% and HA 0.30% eye drops solution, were produce in order to verify if their
characteristics were persistent among each other and between their IPC and FPC, to see if the results
respected the required specifications of each analysis. Overall the results descendent from the process
controls in each pilot batch have respected the required specifications, the aspect always presented a
limpid, transparent and odourless appearance, the pH, the osmolality, the sodium hyaluronate assay,
the bioburden and the sterility test, this last aspect only applied to the HA 0.15%, all presented values
within the limit of the required specification. The density assay which was only performed for the
determination of the filling average presented identical values in all the manufactured batches.
7,9
8,0
8,1
8,2
8,3
8,4
8,5
8,6
Avera
ge s
am
ple
s v
olu
me (
mL)
Hours in Process Control
Filling Control HA 0.15% and HA 0.30%
Upper specification limit Lower specification limit
Average samples volume HA 0.15% PB1 (mL) Average samples volume PB2 HA 0.15% (mL)
Average samples volume HA 0.15% PB3 (mL) Average samples volume HA 0.30% PB1 (mL)
Average samples volume HA 0.15% PB2 (mL) Average samples volume HA 0.15% PB3 (mL)
Figure 13- Results from the filling control test for HA 0.15% and HA 0.30%.
60
Between each batch of the two products it is visible that the values of the IPC and the FPC kept similar
results for the process controls with a defined specification. The viscosity component, which is in study,
showed similar values between the three batches both in the IPC and the FPC. However, the values of
the IPC and the FPC for each batch in HA 0.15% increased significantly while in the HA 0.30% only the
PB1 presented a slight increase while the other batches diminished lightly. These unexpected values
between the IPC and the FPC of each batch are maybe due to the fact that the polymer in question is a
dry polymer and one of the most important steps in the preparation of dry polymer is polymer wetting.
The moment the polymer and the dilution water first contact it is crucial to disperse the polymer into the
water effectively wetting each individual polymer particle. In this step a high shear mixing energy is
required to avoid polymer agglomeration. The rate of hydration is dependent of the particle size and
wetting time, in other words, the smaller the polymer particle the faster it hydrates. This parameter can
be measure through viscosity, higher polymer hydration increases solution viscosity [114]. The viscosity
of the three batches of HA 0.15% had an increasement of approximately 500 times between the IPC
and the FPC. It is important to mention that the time that the IPC was performed was in July while the
FPC was in September, which means that the product was at rest during two months. The HA is a dry
powder polymer that takes some time to hydrate and for a period of two months the product had a large
amount of time to hydrate each particle individually, resulting in a viscosity increasement. This
phenomenon may have also affect the osmolality of the product. These results were not visible in the
HA 0.30% since the time between the IPC and the FCP was not so extensive. The IPC of PB1 0.30 was
performed in August and the FPC was in September, while the IPC and the FPC of both PB2 0.30 and
PB3 0.30 were performed in September. The PB1 0.30 was at rest for one month which means the
polymer had some time to hydrate, being the reasons of the slight increase. However the PB2 0.30 and
PB3 0.30 only rested for some days, not having enough time to hydrate properly and the polymer was
still in a readjustment stage, which causes the slight decrease of viscosity. Although the PB1 0.30
presented a slight increased and the PB2 0.30 and PB3 0.30 presented a slight decrease these changes
were not significant.
Regarding the sterility test, being the product an eye solution, it is demand that the product is sterile to
guarantee a safe use. The HA 0.15% three batches all contained absence of growth meaning that the
product was not contaminated. The HA 0.30% however the sterility test was not performed.
The Bubble Point and Bonfiglioli test results (Table 19) show the filter integrity records for the filters used
(before and after filtration) were in accordance with the required parameters as well as the leak tests
(Bonfiglioli). The filling control (Figure 13) shows all three batches (HA 0.15% and HA 0.30%) are within
the limited specifications
4.5. Conclusion
The results of the Validation Plan obtained in the IPC and the FPC have showed to be consistent and
respectful of the limit established for each process control, indicating homogeneity between batches, as
well as a good reproducibility of the manufacturing process. It is not possible to conclude if the
manufacturing process can be validated or if it is necessary to optimise the manufacturing process since
the sterility test of HA 0.30% is still yet to be performed.
61
Chapter Five – Product Characterisation 5. Product Characterization - Introduction
In order to develop a proper MD it is essential to verify if it is safe, effective and easy to use. To achieve
this goal usability testes are performed to scrutinize device design for potential safety issues and validate
device user interface such that errors that could occur during the use of the device are either eliminated
or reduced as far as possible [69].
When it comes to eye drops, its characteristics may determine the effectiveness and compliance of the
therapy. The solution has to assure comfort, convenience and absence of adverse effects to guarantee
treatment success [115]. Polymers are used in eye drops to regulate the surface tension and wettability
which promotes the hydration of the cornea surface [116], therefore it is important to study the strength
of the interaction between the polymer and the mucin, which is one of the layers that constitutes the tear
film.
Nowadays to study the efficacy of a product in vitro tests are quite used, for instance, though rheological
methods such as tensile strength measurements, flow and oscillation methods providing information
concerning the concentration of the polymer, the mucin and physical properties aspects of the product
[46,52]. To evaluate the products’ safety, in vitro tests cell viability assays using specific cell lines are
widely used to study cell behaviour after product administration.
5.1. Material and Methods
5.1.1. Material
The materials used are described in Chapter Three – Risk Analysis, section 3.1.1. In addition, it was
use dried mucin from porcine stomach type II (Sigma-Aldrich, USA). Human retinal pigment epithelial
cell lines ARPE-19 (ATCC® CRL-2302™) were obtained from American Type Cell Culture collection
(USA), and they were used for cell viability and dry eye assays. Cell culture medium and supplements
were from Gibco (Thermo Fisher Scientific, UK).
5.1.2. Methods
5.1.2.1. Viscosity Measurements
5.1.2.1.1. Brookfield viscometer
Shear rate against shear stress measurements were obtained using a DV-II + Pro Brookfield (Brookfield,
USA) viscometer equipped with spindle nº 21 at room temperature (20 - 25ºC). A shear rate sweep from
0.5 to 100/s and up and down for 7.5 min was assessed.
5.1.2.2. Mucoadhesion studies
The mucoadhesion was evaluated by viscosity, rheology and zeta potential (ZP) measurements. The
mucin used in this study was hydrated with water by gentle stirring until complete dissolution to yield a
dispersion of 10% (w/w) at 20-25°C.
62
5.1.2.2.1. Ostwald viscometer
The viscosity properties of the solutions were determined at room temperature by using the Ostwald
viscometer (Fisher Sientific, USA) using the following equation:
ƞ1 = ƞ2. ρ1t1/ρ2t2 (5) Where ƞ1 and ƞ2 are viscosity coefficients of the solution and water, ρ1 and ρ2 are the densities of the
solution and water, and t1 and t2 are the flow times measured in the viscometer of the solution and water,
respectively.
The viscosities of each individual component, HA 0.15%, HA 0.30% and mucin, were measure first in
triplicate, a mean of the values of each component was made. To evaluate the effect of the interaction
of the mucin with the solutions three samples were prepared: 1) 5 % w/w mucin suspension diluted in
water; 2) Mucin and HA 0.15% solution (1:1); 3) Mucin and HA 0.30% solution (1:1).
The mucoadhesion was expressed through the following equation:
Δ(%) = [ƞmuc+HA – (ƞmuc + ƞHA)] / (ƞmuc + ƞHA) x 100 (6)
Where Δ(%) is the mucoadhesion index, ƞmuc, ƞHA and ƞmuc+HA is the mucin’s, the product’s and the
solution containing mucin and product dynamic viscosity, respectively. For a mucoadhesive polymer,
which is the case of HA, the ƞmuc+Ha is higher than (ƞmuc + ƞHA) due to the interactions occurring between
the polymer and mucin. The mucoadhesive index is a measure of the mucoadhesive strength [56].
5.1.2.2.2. Rotational Rheometer
The rheological characteristics of the formulations were examined at high shear rates using continuous
shear techniques and in the viscoelastic region using oscillation techniques. These experiments were
performed with a controlled stress Malvern Kinexus Rheometer (Malvern Instruments, Malvern, UK)
using cone and plate geometry (truncated cone angle 4° and radius 40mm). The frequency sweep
method was performed between 0.1Hz and 10Hz, with a shear strain of 0.8%, at 25ᵒC, while the table
of shear rate method was performed by increasing the shear rate from 0.1s-1 to 100s-1, at 25ᵒC. The
shear stress was measured by this method and the apparent viscosity was calculated by dividing the
shear stress by the shear rate.
An oscillatory amplitude sweep and frequency testing was performed using this equipment. The
amplitude sweep conditions used were shear strain between 0.01% and 100% with the frequency of 1
Hz. It was concluded that the LVER (linear-viscoelastic region) was at shear strain of 0.25%. In the
frequency testing the frequency range used was between 0.1 - 10 Hz with a shear strain of 0.25%.
A time sweep test was also performed using this equipment with a shear strain of 0.25% and a frequency
of 1Hz during 30 minutes at room temperature (20 - 25ºC).
The adhesive strength was also measured using the same equipment and a plate and plate geometry.
It was used a toolkit with the conditions of 0.1 mm/s, 5 mm and 0.15 GAP. The same protocol was
performed using pig eyes obtained from a local slaughterhouse, instead of mucin.
63
5.1.2.2.3. Zeta Potential
The mucoadhesion interaction was determined by measuring the ZP of the mixtures of mucin and each
solution using a Zetasizer Nanoseries Nano Z (Malvern Instruments, Malvern, UK). A volume of 40 µL
of all samples were diluted in 2 mL of filtered purified water and the cell was filled verifying for the
existence of bubbles that could cause interference in the ZP measurements. All experiments were done
in triplicate.
5.1.2.3. In Vitro Assay
5.1.2.3.1. Cell Culture Condition
The ARPE-19 cell line (ATCC, CRL-2302™) was grown in DMEM/F12 culture medium (Gibco, UK)
supplemented with 10 % (w/v) fetal bovine serum (FBS, Life Technologies. Inc., UK), penicillin (100
IU/mL) and streptomycin (100 μg/mL) in a humidified 95 % O2, 5 % CO2 environment at 37°C.
5.1.2.3.2. Cell Viability of HA 0.15% and HA 0.30%
The cell viability was quantitatively evaluated in vitro using general cell viability endpoint MTT reduction
(3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl- 2H-tetrazolium bromide) assay. MTT is a yellow and water-
soluble tetrazolium dye that is converted by viable cells to a water-insoluble, purple formazan.
Cell viability was assessed after 24h of incubation of ARPE-19 cell line with different concentrations of
each sample. The negative control was the culture medium and positive control sodium dodecyl sulfate
(SDS) at 0.1 mg/mL. After the time of exposition (24 h), the culture medium was replaced by medium
containing 0.5 mg/mL MTT. The cells were further incubated for 3h. In the plates containing reduced
MTT, the media was removed, and the intracellular formazan crystals were solubilized and extracted
with dimethylsulfoxide (DMSO). After 15 min at room temperature the absorbance was measured at 570
nm in the same microplate reader.
The relative cell viability (%) compared to control cells was calculated by the following equations:
𝐶𝑒𝑙𝑙 𝑉𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑦 (%) 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑀𝑇𝑇 𝑎𝑠𝑠𝑎𝑦 = [𝐴𝑏𝑠𝑜𝑟𝑣𝑎𝑛𝑐𝑒 570 𝑛𝑚]𝑠𝑎𝑚𝑝𝑙𝑒
[𝐴𝑏𝑠𝑜𝑟𝑏𝑎𝑛𝑐𝑒 570 𝑛𝑚]𝑐𝑜𝑛𝑡𝑟𝑜𝑙 × 100 (7)
5.1.2.3.3. 2D model - Evaluation of Cell Morphology and Cell Viability After
Dehydration
The protective effect of the selected formulas against dehydration was evaluated using previously
reported protocols, with modifications [56]. Specifically, cells were seeded in 24-multiwellplates (5 x
104/well) and in DMEM/F12 until 70% confluence was reached. The medium was then replaced with
selected HA formulations (HA 0.15%, HA 0.30% and CR 0.30% wt% solutions prepared in cell culture
medium) and with the same solutions diluted 1:5. For the positive and negative controls, the medium
was replaced with fresh medium not containing HA. Cells were incubated under cell culture conditions
for 2 h. Cells treated with the HA samples and not treated (negative control, NC) were then dehydrated:
the medium was removed and the multiwells were incubated at 37°C without the lid until a stress
response (morphological change) was evident in the NC (about 20 min). The positive control (PC, not
64
treated with HA), was not dehydrated (cells were kept in the presence of the medium during all
experiments.
Cell viability was evaluated using the AlamarBlue assay (Cat. N. A13261, Invitrogen, GIBCO) according
to manufacturer’s instructions. When added to cells, the cell-permeable AlamarBlue reagent, resazurin,
is reduced by the dehydrogenases into resarufin by viable cells. The conversion is proportional to
metabolically active cells and was quantitatively determined by fluorescence measurements. Cell
viability (%) was calculated with respect to the positive control (100% viability). Results were reported
as means ± SD.
5.1.2.3.4. 3D model - Dry Eye Model and Cell Viability
For the 3D dry eye assay, cells were cultured on filters following the protocol described by Dunn et al.
with some modifications [117]. Briefly, the cells were seeded at a density of 1×105 cells/cm2 on
ThinCert™ cell culture inserts (Greiner, 3 μm, 12 wells, UK). The culture medium was supplemented
with l-ascorbic acid (50 µg/mL), β-glycerolphosphate (10 mM) and dexamethasone (10 nM) in order to
enhance the barrier properties and facilitate expression of RPE-specific genes [117]. Fresh medium with
supplements was changed twice a week.
The progress of epithelial barrier formation and polarization was followed by measuring transepithelial
electrical resistance (TEER) with a Millicell-ERS device (Millipore, Germany) and chopstick-style
electrode. The combined resistance of the filter was subtracted from the values of filter-cultured ARPE-
19 cells in order to calculate the resistance of the cell layer. The plateau in TER was reached in two
weeks and, thereafter, it remained essentially unchanged. The cells were used for experiments after
culturing them for three weeks. In the permeability experiments the resistance was determined before
and after the experiments.
After the three weeks ARPE-19 cells were placed under controlled environmental conditions to mimic
dryness (without lid, <40% relative humidity, 37 °C ± 5 °C temperature and 5% CO2). Cells were
investigated for cell viability at 48 h after establishment of dry eye conditions, using the MTT reduction
assay (see section 5.1.2.3.2). Cell viability was assessed after 24h of incubation of ARPE-19 cell line
with 20 mg/mL concentration of each sample. The negative control was the culture medium and positive
control sodium dodecyl sulfate (SDS) at 0.1 mg/mL.
5.1.2.4. Statistical Data Analysis
The data was expressed as mean and standard deviation (mean ± SD) of experiments. Tukey–Kramer
multiple comparison test (GraphPad PRISM 5 32 software, USA), was used to compare the significance
of the difference between the groups, a p <0.05 was accepted as significant.
5.2. Results
5.2.1. Viscosity Measurements of HA 0.15% and HA 0.30%
The viscosity of the three pilot batches of each product, HA 0.15% and HA 0.30%, were measured. The
results are represented in Figure 14 which shows the representative flow curves (shear stress function
of shear rate).
65
Figure 14 – Typical flow curve of shear stress as function of shear rate for HA 0.15% and HA
0.30% eye drops solution.
All batches of each product present the same rheological profile demonstrating that the relation between
shear stress and shear rate is not constant, which shows that the apparent viscosity increases with the
increase of the shear rate in a non-linear form (Figure 14). This profile shows that both products are
shear thinning fluids. It is evident that the shear stress values of HA.030% are much higher than the
shear stress values of HA 0.15% at the same rate. At the highest shear rate (122.36 sec-1) the apparent
viscosity of HA 0.30% was 83.9 mPa.s and 22.5 mPa.s for HA 0.15%.
5.2.2. Mucoadhesive Studies
5.2.2.1. Viscosity Measurements
A study using an Ostwald viscometer was also performed to evaluate the viscosity of the products and
to study the mucoadhesion, in other words, the interaction of HA 0.15% and HA.30% in solution in the
presence of mucin. By measuring the viscosity it is possible to evaluate the interaction of the product
with the mucin, because a higher interaction with the mucin is related with higher viscosity.
0
2000
4000
6000
8000
10000
12000
0 20 40 60 80 100 120 140
Sh
ear
str
ess (
Pa)
Shear rate (sec-1)
HA0.30% Batch 1
HA0.30% Batch 2
HA0.30% Batch 3
HA0.15% Batch 1
HA0.15% Batch 2
HA0.15% Batch 3
66
Figure 15- Viscosity determination for HA 0.15% and HA 0.30% in absence and presence of Mucin
5% (w/w) (mean SD, n=3).
The mucoadhesive index was calculated for the HA 0.15 + Mucin and HA 0.30% + Mucin solutions in
order to evaluate the mucoadhesive strength gain due to the interactions between the polymer and the
mucin.
The results obtained with an Ostwald viscometer presented in Figure 15 supports the results obtained
in the Brookfield viscometer, HA 0.30% presents a much higher viscosity than HA 0.15%. An increase
of viscosity of the product in the presence of mucin when compared with their individual viscosity is also
observed. This increase is more evident for the HA 0.30%, with a Mucoadhesive index of 298.07% ±
6.24% in viscosity (Equation 6) than for HA 0.15% with 67.44% ± 19.90%.
5.2.2.2. Rheology Measurements
5.2.2.2.1. Tackiness Testing
Tackiness in the context of material behaviour is associated with stickiness and may result from
adhesive forces between two materials in contact. It is measured at the maximum force needed to break
the resultant bond. In this test the peak force which is a negative normal force can be attributed to tack
and the area under the force-time curve represents the adhesive strength [118]. These parameters are
measured and the results of the prepared solutions are represented in Table 20. Two samples used are
commercial reference (CR) of HA eye drops solution, with different concentrations of HA (CR 0.15%
and CR 0.30%) available in the market.
0
50
100
150
200
250
300
350
400
450
Mucin HA 0.15% HA 0.30% HA 0.15% +Mucin
HA 0.30% +Mucin
Vis
co
sit
y (
mP
a.s
)
67
Table 20- Normal force and area under force time curve results for HA 0.15% and HA 0.30% and
their interactions with Mucin and with Pig Eye.
*1 Mean ± SD, n=6 *2 Mean ± SD, three different eyes, n=3
The results of the tack testing on the seven samples show that the Mucin + HA 0.30% appears to be
the tackiest of the seven samples analysed with a peak normal force of -0.287 N, followed by the HA
0.30% (-0.287 N), Mucin (-0.228N), CR 0.30% (-0.220 N), Mucin + HA 0.15% (-0.216 N), HA 0.15% (-
0.178 N) and CR 0.15% (-0.168 N). For the Area under force time curve the results did not show the
same profile, the CR 0.30% appears to be the strongest of all samples (1.051 N.s) and the HA 0.15%
the weakest (0.438 N.s).
A similar study was performed but instead of mucin was used pig’s eye (Figure 16), three samples were
used and attached to the probe and the adhesive force between the eye and samples of HA 0.15% and
HA 0.30% was measured. The results show there are significant differences between HA 0.15% and
HA 0.30%, were the HA 0.30% appears to be tackiest with -0.134 N and the HA 0.15% with -0.078 N,
respectively. The Area under force time curve also shows the same profile were HA 0.30% appears to
be the strongest with 1.0110 N.s and the HA 0.15% with 0.891 N.s.
Figure 16- Three samples of pig eye used in the frequency sweep assay (A) and one of the
samples attached to the probe (B).
Peak normal force - Normal Force (N)
Area under force time curve (N.s)
HA 0.15%*1 -0.178 ± 0.003 0.438 ± 0.058
HA 0.30%*1 -0.229 ± 0.013 0.775 ± 0.091
CR 0.15%*1 -0.168 ± 0.017 0.904 ± 0.069
CR 0.30%*1 -0.220 ± 0.007 1.051 ± 0.043
Mucin*1 -0.228 ± 0.004 1.012 ± 0.065
Mucin + HA 0.15%*1 -0.216 ± 0.019 0.573 ± 0.152
Mucin + HA 0.30%*1 -0.287 ± 0.030 0.747 ± 0.066
Pig Eye + HA 0.15%*2 -0.078 ± 0.029 0.891 ± 0.060
Pig Eye + HA 0.30% *2 -0.134 ± 0.034 1.010 ± 0.059
68
5.2.2.2.2. Oscillation Frequency Sweep
Before the oscillation frequency sweep, an amplitude sweep test was performed to define the fluid’s
linear viscoelastic region (LVER), and the results showed that this region was at 0.25% shear strain.
With this results the product’s structure can be further characterized using a frequency sweep proving
more information about the effect of colloidal forces, interactions among particles or droplets [119].
Figure 17- Frequency sweep with shear moduli as function of frequency of HA 0.15%, Mucin and
Mucin + HA 0.15% at room temperature.
Both Figure 17 and Figure 18 represent the frequency behaviour of the product, HA 0.15 % and HA
0.30% respectively, compared to system obtained with mucin. It is evident that with the mucin the elastic
modulus G’ and the viscoelastic modulus G’’ increased in both products. At lower frequencies both
products exhibited fluid-like mechanism spectra with G’’ modulus greater than G’, being both frequency
dependent. As the frequency increases occurs a cross-over at approximately 2 - 5 Hz, turning the G’
modules greater than the G’ indicating both products started to have a more elastic behaviour.
0,1
1
10
0,1 1 10
G' an
d G
'' (
Pa)
f(Hz)
G´HA0.15%
G´´HA0.15%
G´Mucin
G´´Mucin
G´Mucin+HA0.15%
G´´Mucin+HA0.15%
69
Figure 18- Frequency sweep with shear moduli as function of frequency of HA 0.30%, Mucin and
Mucin + HA 0.30% at room temperature.
5.2.2.2.3. Time Sweep
The test was performed to understand the changes in the sample with time at constant temperature,
stress and frequency with a time range of 30 minutes. The results show that all sample except mucin
are time independent with viscoelastic modulus (G’’) greater than the elastic modulus (G’), meaning that
their structure do not suffer changes overtime (Figure 19).
0,1
1
10
0,1 1 10
G' an
d G
'' (
Pa)
f(Hz)
G´HA0.30%
G´´HA0.30%
G´Mucin
G´´Mucin
G´Mucin+HA0.30%
G´´Mucin+HA0.30%
Figure 19- Time Sweep Test for HA 0.15% and HA 0.30% with and without mucin.
0,1
1
10
100 1000
G' an
d G
'' (
Pa)
Time (s)
G'Mucin + HA0.30% G´´Mucin + HA0.30% G'Mucin + HA0.15% G´´Mucin + HA 0.15%
G'Mucin G´´Mucin G'HA0.30% G´´HA0.30%
G'HA0.15% G´´HA0.15%
70
5.2.2.2.4. Zeta Potential
This study demonstrated that the ZP values of HA 0.15% and HA 0.30% eye drops solution are similar
to their market equivalent, CR 0.15% and CR 0.30% respectively. However, comparing the values of
the two products their values are quite different being HA 0.15% ZP values much more negative that
HA 0.30% values, which in absolute is higher (Figure 20). This negative values are in accordance to the
fact the HA presents an anionic nature due to the presence of carboxylic groups. The mucin also
presents negative charge due to the oligosaccharide chains, which confer negative charge to the mucins
through carboxyl and sulfate groups. When the mucin is added to both products an increase of the
negative charge is observed, being the ZP value more negative in Mucin + HA 0.30% than with HA
0.15%.
An overtime study was made to investigate if the interactions alter overtime or if they maintain stable.
The measurements of the samples were performed at 0, 5 10, 15 and 20 minutes and it was concluded
that the values do not suffer significant alteration overtime.
Figure 20- Determination of ZP for HA 0.15%, HA 0.30%, Mucin and both products with mucin
(Mean ± SD, n=3).
5.2.2.3. In Vitro Assay
5.2.2.3.1. Cell Viability of HA 0.15% and HA 0.30%
An initial test to evaluate the potential irritant of the HA formulations was performed in order to choose
the most suitable dilution to be used on the next assays. Four dilutions were prepared 100 µg/mL (1:1),
-40
-35
-30
-25
-20
-15
-10
-5
0
ZP
(m
V)
71
50 µg/mL (1:2), 33 µg/mL (1:3) and 20 µg/mL (1:5). In Figure 21 it is evident that the 1:5 dilution presents
a high rate of survival in all samples with cell viability above 80%. The 1:3 dilution although the CR
samples demonstrated high survival rate (approximately 100%), the HA samples did not show the same
results. For that reason it was decided to use the 1:5 dilution.
Figure 21- Results of cell viability on ARPE-19 cell lines testing CR 0.30%, CR 0.15%, HA 0.30%
and HA 0.15% at various concentrations (mean ± SD, n=8).
5.2.2.3.2. 2D model - Evaluation of Cell Morphology and Cell Viability After
Dehydration
In Table 21 it is shown optical microscope images of ARPE-19 cells stained with Crystal Violet and
exposed to desiccation under no protective conditions (Dry Eye, negative control), after being treated
with HA 0.15%, HA 0.30% and CR 0.30% and of cells that were not exposed to dehydration (Medium,
positive control). The respective cell viability determination of the samples is also present. In the Dry
eye images it is evident the cells exhibited a disintegrated and dry membrane morphology and an
increase of cell mortality. The cells treated with HA formulations, did not show the same results, the
typical morphology and high survival rate could still be observed. The results of cell viability confirmed
the microscopic observation. The dehydration was responsible for almost 50% of mortality rate in the
Dry Eye sample, while the cells pre-treated with HA formulations presented higher survival rates, 60 –
70%, confirming a protective effect displayed by the HA formulations.
.
0
20
40
60
80
100
120
CR 0.30% CR 0.15% HA 0.30% HA 0.15%
Cell v
iab
ilit
y (
%) SDS
100 µg/mL
50 µg/mL
33 µg/mL
20 µg/mL
Medium
72
Table 21- Optical microscope images of ARPE-19 after dehydration in no protective conditions
(Dry Eye), after dehydration preceded by treatment with HA formulations and cells not submitted
to dehydration (Medium). Without Die
Magnification 100x Crystal Violet
Magnification 200x Crystal Violet
Magnification 400x Cell
Viability (%)
Dry
Eye
50.7 ± 6.8
CR
0.3
0%
73.4 ± 12.6
HA
0.1
5%
62.7 ± 9.5
HA
0.3
0%
72.5 ± 6.2
Med
ium
100.0 ± 11.0
5.2.2.3.3. 3D model - Dry Eye Model and Cell Viability
The results of cell viability for the different tested samples are presented in Figure 22 showing a good
cell viability profile of ARPE-19 cell line. When applied CR 0.30%, HA 0.30% and HA 0.15% cell viability
increases when compared to the Dry Eye model. The survival rates were over 100% indicating that the
formulations were non-toxic so much so that provided a good environment for cell proliferation.
73
Figure 22-Cell viability ARPE-19 cell line after dehydration treatment exposed for 24h to HA
0.15% and HA 0.30% eye drops solution and commercial formulation CR 0.30% (mean ± SD, n=9).
5.3. Discussion
The viscosity is one of the most important parameter for an eye drop solution, since it compromises the
efficacy of the treatment. A product with low viscosity may not assure the suitable retention time to treat
or relief symptoms of DED, but a too much viscosity may cause unwanted visual disturbances or some
blurriness. Several studies were performed in order to understand the difference in viscosity of the two
products, HA 0.15% and HA 0.30%, and to understand the mucoadhesion properties of this polymer
and eye drops solution [36].
The first study was the determination of the rheological profile of the two products. Observing the Figure
14 the values of shear stress are much higher for HA 0.30% when compared to HA 0.15%, which
indicates that it is much more viscous. These results make sense since HA 0.30% has the double of HA
concentration. The viscosity of the HA is due to the fact that is it a semi-flexible polymer, the higher the
concentration the higher the viscosity [120]. The three batches of each product present similar
rheological profile, both products do not present a constant relation between shear stress and shear
rate. The shear stress increases along the increase of the shear rate in a non-linear form, meaning that
this fluid behaves as a shear thinning fluid. The viscosity was also measured using an Ostwald
viscometer. The results shown that HA 0.30% is much more viscous than HA 0.15% with 71.20 mPa.s
and 6.83 mPa.s, respectively (Figure 15). These results support the outcomes obtained with the
Brookfield viscometer. The suspension of mucin prepared for this research work (5%, w/w) presents
viscosity since it is a high glycosylated protein with high MW.
To study the mucoadhesive properties of HA, samples of mucin with HA 0.15% and mucin with HA
0.30% were prepared and the results have shown there are indeed interactions between mucin and HA,
since the viscosity increased significantly when compared with the viscosity of HA 0.15% and HA 0.30%
eye drops solution or mucin alone. The mucoadhesive index was also determined and it demonstrated
that although both product have an increase superior to 50%, HA 0.30% + Mucin presented a massively
increase of 298.07% ± 19.90%. This increase suggests that a strong interaction between mucin and HA
0
20
40
60
80
100
120
CR 0.30% HA 0.30% HA 0.15% Dry Eye Medium SDS
Cell V
iab
ilit
y (
%)
74
occurred, since HA 0.30% presents more concentration of HA, more interactions with the mucins were
possible. This increase of viscosity given by the interactions between mucin and the polymer are
possibly due to the formation of hydrogen bounding between the hydroxyl and carboxyl groups with
mucin’s amino groups. The HA is a linear molecule and can easily interpenetrate a mucin random coil
and that high molecular weight polymers can increase the probability of interfacial interactions with
mucin, creating a more stable connexion. These conclusions were given by Hassan and Gallo [121] and
their study with chitosan and mucin were the viscosity was measured at different pH levels. The
adhesive capacity was not solely due to electrostatic bonding, which showed to be ambiguous, but also
other types of bounding and interactions.
The adhesive properties of polymer are highly influenced by the viscosity as well the surface and
interfacial tensions of the polymer and substrate. In the pull away assay the adhesion was measured
throught a tack or probe testing, where by bringing a probe into contact with the surface of the polymer,
the mucin and the system polymer + mucin under a specific force and pulling the probe at a constant
velocity, the adhesive force was measured. While the probe is raised at a constant velocity it is measured
the necessary force in which the sample dissociates from the probe, resulting in a force versus distance
curve. The integral of that curve corresponding to the Area Under Force Time Curve represents the
adhesive strength of the sample [122]. The results show that are significant differences (p<0.05)
between the Normal Peak Force of HA 0.15% vs HA 0.30%, HA 0.15% vs Mucin + HA 0.15% and HA
0.30% vs Mucin + HA 0.30% (Table 20). The first case makes sense since HA 0.30% has the double
concentration of polymer, which makes it more viscous as discussed before. The significant difference
between HA 0.15% vs Mucin + HA 0.15% and HA 0.30% vs Mucin + HA 0.30% is an indication that the
addition of mucin created an interference with the polymer which formed a more viscous system,
concluding that a mucoadhesion occurred. The same conclusion was obtained with the pig eye, both
Normal Peak Force and Area Under Force Time Curve were higher with HA 0.30% than with HA 0.15%.
According to Hägerström and Edsman, who performed a similar assay, the strengthening might arise
from the entanglement of the polymer chains and the mucous glycoproteins, the formation of chemical
bonds and/or from dehydration of the mucous layer [46]. However, the Area Under Force Time Curve
values did not meet the same profile as the Normal Force, there were no significant difference between
HA 0.15% vs Mucin + HA 0.15% and HA 0.30% vs Mucin + HA 0.30%. These results may be due to the
fact that the samples variability was higher in this measurement and when performing the statistical data
analysis, it was not capable of detecting differences. Rather than having a p < 0.05, a p < 0.1 could be
sufficient to detect differences in those results. Values obtained with mucin were higher than the ones
with pig eye since the concentration of mucin is much higher in solution (5% w/w) than the concentration
existent in the eye.
In non-ideal fluids the response of the polymer will depend on frequency with both shear moduli (G’ and
G’’) increasing with frequency. In both products, HA 0.15% and HA 0.30% eye drops solution, the G’’
modulus is grater at low frequencies which indicates a fluid-like system (Figure 17 and Figure 18). With
the addition of mucin this profile still remains, but the values of both shear moduli increase. This means
that it is necessary a greater amount of force or stress to deform the sample along the plane of the
direction of the force, which indicates that some type of interaction has been established. According to
75
Ludwig [8] the interaction between a mucoadhesive polymer and the mucin may occur by the following
mechanisms: physical entanglements, Van der Walls bonds, electrostatic forces and hydrogen bonds.
At low frequencies, the products and the system HA + Mucin suffer a rearranging due to Brownian
motion, physical entanglements are created and broken quickly compared to the rate of deformation, so
they do not store elastic energy. At high frequencies the polymer/polymer + mucin system does not have
time to rearrange causing the physical entanglements to persist longer than the oscillation frequency
constraining the polymer. The elastic energy is stored and the viscous dissipates that being the reason
why at high frequencies the G’ moduli is higher than G’’. The cross-over is than the point of passage
where the formulation stops presenting fluid-like characteristics and behaves more gel-like. The shear
moduli present greater values in HA 0.30% in comparison with HA 0.15%, which indicates that the
strength of the formulation/mucin interaction increases with HA concentration [56,122].
The ZP is related to the measurement of the surface charge that a specific material possesses or
acquires when suspended in a fluid. The results have shown that HA 0.15% and CR 0.15% ZP values
are more negatively charged than HA 0.30% and CR 0.30% (Figure 20). As said before the negative
charge of HA is due to the presence of carboxyl groups who dissociate at physiological pH [123]. By this
logic, the higher the HA concentration the higher the negative charge. However, the HA in this product
is at the form of sodium hyaluronate, which means it is in salt form. Positively charged ions from the
monovalent alkali metal series such as Na+ act as counter ions on anionic structures like HA, absorbing
on the surface-dominating negative sites decreasing the absolute value of the ZP. These results are in
accordance with Romero et al. and their work on silica with salts which demonstrated that the ZP value
of silica in absolute decreased with the addition of salts. The addition of salts reduces electrostatic
repulsion which facilitates the formation of H bounds, and thus an increase in viscosity occurs [124].
With the addiction of mucin both HA 0.15% and HA 30% ZP values increased in absolute, being the
Mucin + HA 0.30% a more negatively charged system. Mucin can be described as a double-globular
protein region connected by highly glycosylated linkers containing carboxylic and sialic acid, which
confers the negative charge at physiologic pH. The overall net charge is negative, but it may also exist
positively-charge regions in the non-glycosylated globular region containing histidine, arginine and
lysine residues. A study performed by Menchicchi et al. concluded that the polymers like chitosan
presents mucoadhesive properties due to the electrostatic interactions between positively-charged
polymer and negatively-charged mucin. However negatively-charge polymers such as alginates, pectins
and acrylic acid also show mucoadhesive properties, similar to the HA case. This means that the reason
for mucoadhesion on polymers is not solely due to electrostatic interactions but also due to other types
of interactions. Mucin forms a complex macromolecular network with available functional groups such
as sialic acid, therefore it is possible to interact with the polymer by hydrogen bounding between sialic
acid and the polymer’s carboxylate residue. It is also possible to form hydrophobic interactions with
mucin’s amino acids and entanglement of the polymer. The changes in ZP value were more significant
in HA 0.30% than HA 0.15% with the addition of mucin, because the first one presents double
concentration of HA, more HA allows more interactions and thus an increase of viscosity [57].
The time sweep test (Figure 19) shows that both shear moduli of the products and HA + Mucin systems
do not suffer alteration gradually with time. This indicates that overtime it does not occur changes of
76
structure at a given oscillation frequency, concluding that stability is time independent. The same
happened in the ZP overtime study. The unchanged structure overtime and the increase of viscosity are
also indicators that the system HA + Mucin is stable. Stable interactions increase retention time, which
means that the contact time in the corneal surface remains longer. Since the product in study is a MD,
its action is of physical nature and not pharmacological, for that matter the longer the retention time the
more efficient is the treatment [36,43].
Cell viability was performed using ARPE-19 cell line in a 2D and 3D in vitro model that mimicked DED
conditions. By testing in this conditions, it is possible to study if HA 0.15% and HA 0.30% are promising
candidates for the treatment of DED. Frist, cell viability was performed using different dilutions of HA
0.15% and HA 0.30% to study which dilution is most suitable to use on the following assays, in other
words the less cytotoxic dilution (Figure 21). Four dilutions were made, 1:1, 1:2, 1:3 and 1:5. The results
concluded that the most suitable dilution was the formulation 1:5 dilution due to the high survival rate
(above 80%) in all tested samples.
A 2D in vitro assay was performed secondly to study the differences in the morphology and cell viability
of dehydrated cells who received a pre-treatment with HA formulations with cells who were not submitted
to the treatment (Table 21). It was also tested the commercial formulation CR 0.30% since this MD is
used in severer cases of DED. The results have shown that the cells who were not treated with HA
presented a disintegrated and dry membrane with a cell mortality rate close do 50% when compared to
the non-dehydrated cells. The pre-treated cells showed a morphology more similar to the hydrated cells
in both products with a high survival rate. Cells treated with HA 0.30% showed a higher cellular viability
rate of 72.5% ± 6.2%, almost equal to the CR 0.30% with 73.4% ± 12.6, when comparing to the ones
treated with HA 0.15% with 62.7% ± 9.5%. In the 3D model the results show that the application of all
three formulations obtained over 100% cell viability meaning that the application of these DM provided
a more suitable environment for cell proliferation. These results are very similar to the one performed
by Salzillo et al., were it was performed a study of cellular response of primary porcine corneal epithelial
cells when administrated different HA formulations [56]. They justified their results with the hypothesis
that the protective effect displayed by the HA formulation on the cells is related to the polymer water
retaining capacity. HA 0.30% presented higher cell viability values since more concentrated formulations
retain more water promoting higher hydration. This findings are important in the view of potential forms
of treatment for DED, the higher the HA concentration the higher the efficacy, which concludes that HA
0.30% may be more indicated for more severe cases of DED.
When comparing the 2D with the 3D in vitro assay the results of cell viability performed in both methods
were quite different. This is due to the fact that the 2D cell culture model is more sensitive than the 3D
model, this method only has a single layer which means when submitted to dehydration all cells were
exposed since all medium was removed. In the 3D model the dehydration is partial, the medium is
removed in the insert but it still exists in the well, simulating DED in vivo conditions.
77
5.4. Conclusion
The aim of product characterization in this case was to evaluate the differences between the two
products, HA 0.15% and HA 0.30%, and to study their interactions with mucin, a form to predict their
behaviour in a biological environment.
All results obtained showed that HA 0.30% presented a higher viscosity than HA 0.15%.
The mucoadhesion between the HA and the mucin was tested and all results indicated that some kind
of interaction occurred between the mucin and the HA being stronger with HA 0.30%. Physical
entanglements and hydrogen bounding are possible forms of interaction.
The increase of viscosity when the mucin is added and the unchanged structure overtime are indicators
that the system HA + Mucin is a stable interaction, which increases the retention time in the corneal
surface, improving the efficacy of the treatment.
The cell viability test was performed with a 2D and 3D in vitro Dry Eye model. In the 2D model it was
concluded that the cells pre-treated with HA preserved the cell’s morphology after the dehydration
process and maintained a high survival rate. The 3D model demonstrated that the administration of HA
0.15% and HA 0.30% increased the cell viability over 100%. These values indicate that the application
of HA favours cell proliferation by creating an optimum environment. The reason for that environment
may be due to the hydration caused by the water retention by the HA molecules. From these results the
products, HA 0.15% and HA 0.30%, are potential candidates for becoming suitable MD for DED
treatment, being HA 0.30% most suited for more severe cases.
78
79
Chapter Six – Conclusion and Future Work
6. Concluding Remarks and Future Work
6.1. Concluding Remarks
The ophthalmic administration is the most common and well-accepted form of administration for the
treatment of eye disorders, such as DED. Symptoms of discomfort and visual disturbances causes by
DED may interfere with patience quality of life, being the application of some kind of treatment who
immediately attenuates the symptoms a big necessity. In modern society were smoking, alcohol
consumption, computer-related tasks, use of contact lenses and prolonged working hours in air-
conditioned facilities are part of some people daily activities can cause dry eye, which is why DED is the
one of the most common diagnosed ocular disease. The first–line therapy are lubricants also known as
eye drop comfort solutions or artificial tears which are non-prescription eye drops, classified as a MD
[36,43].
Polymers with water binding properties and high viscosity are highly used in ophthalmic preparations.
In artificial tears, polymers sooth, protect, lubricate and possess mucoadhesive properties with mucin
presented in the ocular surface, relieving the symptoms of dry eye [83]. One of the most popular and
effective is HA, the chosen polymer of Edol’s project in developing safe and efficient eye drops solution.
It was decided to invest in two products: HA 0.15% for daily use and 0.3% for more severe cases or
before sleep application. For the selection of the excipients these had to obey some required
specification to avoid ocular irritation or even damage. The pH value, osmolality and electrolyte
composition must be similar to the lacrimal fluid and the solution must be sterile. The preservative is one
of the most important excipients because it is a multidose MD, the solution must be able to kill
microorganisms or avoid microbial growth [39]. The chosen buffer was borate, potassium, magnesium
and calcium chloride as electrolytes and sodium chloride for osmolality adjustment. The selected
preservative was Suttocide (N-hydroxymethylglycinate 50%) with EDTA. This decision led to the
conclusion of the manufacturing process which conduce the design of the scale-up pilot batch
manufacturing process.
The Risk Analysis performed on the pilot batch manufacturing process demonstrated to be a complex
and a time consuming study since an intensive study of all the process steps had to be performed. The
two methods used, FMEA and Qualitative Analysis according to the ISO 14971:2007, presented very
different results from each other. From a long list of possible failure modes, FMEA was able to detect
14% unacceptable risks, while the ISO method only classified 11% as unacceptable. This demonstrated
that the FMEA is a more sensitive tool, detecting more risks and was capable of distinguish different
types of risks trough the RPN. After the detention, corrective actions to attenuate the risk had to be
applied in both methods. In the FMEA method it was possible to implement corrective actions in all the
unacceptable risks detected which diminished their risk level from unacceptable to acceptable. The ISO
method however was not capable of implementing corrective actions meaning that it was not also
capable of mitigating the unacceptable risks detected. This concludes that the FMEA besides of being
a generic tool was a much more useful tool, detecting and correcting unacceptable risks than the ISO
method, which is an application of risk management to MD.
80
The Validation Plan was a required study since it is a new product and the manufacturing process must
produce consistently product within the specified requirements and quality. From the results obtained in
the IPC and FPC it was concluded that all values were consistent and respectful of the limit establish
for each process control. To conclude if the process can be validated the sterility test of HA 0.30% must
be performed.
The results from the Product Characterization showed that the viscosity increases with the increase of
HA concentration. This was proven by measuring viscosity through the Brookfield and Ostwald
viscometers, tack test, oscillatory frequency test and ZP measurements were in all cases the HA 0.30%
presented higher values than HA 0.15%. This proves that the viscosity increases with the increase of
concentration. The mucoadhesion was also measured trough the previous tests except the Brookfield
viscometer. The results showed that the viscosity, Normal force, both shear moduli (G’ and G’’) and ZP
value in absolute all increased in value once the mucin was in contact with HA, for both HA 0.15% and
HA 0.30%. This is a confirmation that interactions in fact occurred between the mucin and HA though
possibly entanglements and hydrogen bounding. The unchanged structure overtime time observed in
the time frequency test is an evidence that the structure is stable. Cell viability test in 2D results
demonstrated that cells pre-treated with both HA 0.15% and 0.30% maintained their morphology after
dehydration and that there was an increase of cell viability, demonstrating a protective effect. In the 3D
model the results showed that the cells treated with HA presented more than 100% cell viability
compared with the control Medium, indicating that they are in fact promising and effective MD.
6.2. Future Work
In future works besides from the cell viability test performed in ARPE-19 cell line, which are cells from
the retina, a new study of cell viability should be performed with cells from the corneal epithelium for
example HCE-T cell line, since DED effects the cornea and not so much the retina. Other in vitro test to
evaluate the effectiveness of the products could be through the quantification of the expression of
biomarkers related to different dry eye pathways such as: inflammatory mechanism at celular and matrix
level (TNF-α and MMP-19); adaptive and defence mechanism in case of lipidic film loss (MUC-4); water
channel protein that regulates water transport (AQP-3); and tight junction biomarkers for epithelial
integrity (OCLN and ZO-1), for example [125].
Mucoadhesion testing by ex vivo adhesion wash-off test method should be also interesting to perform
to compare the retention time of both HA 0.15% + Mucin and HA 0.30% + Mucin systems. This could
be performed using for example freshly excised pieces of intestinal mucosa from pig, a method used by
Veerareddy et al.[126].
Draize eye irritation in vivo test to observe changes of cornea, conjunctiva and iris in rabbit eye ball
following the exposure to the products should be also performed [127]. If the product passes this test,
human trials should be performed.
81
References
1. Kumar A, Malviya R, Sharma PK. Recent Trends in Ocular Drug Delivery : A Short Review. Eur J Appl Sci. 2011;3(3):86–92.
2. Kim Y-C, Chiang B, Wu X, Prausnitz MR. Ocular delivery of macromolecules. Natl Institutes Heal. 2014;(September):172–81.
3. Morrison PWJ, Khutoryanskiy V V. Advances in ophthalmic drug delivery. Ther Deliv. 2014;5:1297–315.
4. “Facts About the Cornea and Corneal Disease- https://nei.nih.gov/health/cornealdisease”[Online].[Accessed 02-Out-2017].
5. “Understanding The Eye: Anatomy - http://healthlifemedia.com/healthy/understanding-the-eye-anatomy/ ”[Online].[Assessed: 12-Sep-2017].
6. “Understanding the Tear Film and Dry Eeye - http://www.collinsoptometrists.com.au/dry-eye-clinic/understanding-the-tear-film-and-dry-eye/”[Online].[Assessed: 11-Out-2017].
7. Bron AJ, Tiffany JM, Gouveia SM, Yokoi N, Voon LW. Functional aspects of the tear film lipid layer. Exp Eye Res. 2004;78:347–60.
8. Ludwig A. The use of mucoadhesive polymers in ocular drug delivery. Adv Drug Deliv Rev. 2005;57(11):1595–639.
9. Foulks GN. The Correlation Between the Tear Film Lipid Layer and Dry Eye Disease. Surv Ophthalmol. 2007;52(4).
10. Dartt D, Willcox M. Complexity of the tear film: Importance in homeostasis and dysfunction during disease. Exp Eye Res. 2013;(December):1–3.
11. Cwiklik L. Tear film lipid layer: A molecular level view. BBA - Biomembr [Internet]. 2016;1858(10):2421–30. Available from: http://dx.doi.org/10.1016/j.bbamem.2016.02.020
12. George S. On the nature and the role of the subsurface vesicles in the outer epithelial cells of the conjunctiva. Br J Ophthalmol. 1985;69:477–81.
13. Holly J, Mass USA. Formation and Rupture of the Tear Film. Exp Eye Res. 1973;(1):515–25.
14. Hodges RR, Dartt DA. Tear Film Mucins: Front Line Defenders of the Ocular Surface; Comparison with Airway and Gastrointestinal Tract Mucins. Exp Eye Res. 2013;117(December):62–78.
15. Tavares FDP, Fernandes RS, Bernardes TF, Bonfioli AA, Jorge E, Soares C. Dry Eye Disease. Inf Healthc. 2010;25(3):84–93.
16. Javadi M, Feizi S. Dry Eye Syndrome. J Ophthalmic Vis Res. 2011;6(3):192–8.
17. Whitcher JP. The treatment of dry eyes. Br J Ophthalmol. 2004;603–4.
18. Bernauer W, Broadway DC, Wright P. Chronic progressive conjunctival cicatrisation. Eye. 1993;7(3):371–8.
19. The Definition and Classification of Dry Eye Disease: Report of the Definition and Classification Subcommittee of the International Dry Eye Workshop (2007). Ocul Surf [Internet]. 2007;5(2):75–92. Available from: http://linkinghub.elsevier.com/retrieve/pii/S1542012412700812
82
20. Colligris B, Alkozi HA, Pintor J. Recent developments on dry eye disease treatment compounds. Saudi J Ophthalmol [Internet]. 2014;28(1):19–30. Available from: http://dx.doi.org/10.1016/j.sjopt.2013.12.003
21. “Palovarotene for the Potential Treatment of Dry Eye Disease - http://clementiapharma.com/our-pipeline/palovarotene-for-dry-eye-disease/”[Online].[Accessed 17-Out-2017].
22. Yavuz B, Bozdağ Pehlivan S, Ünlü N. An Overview on Dry Eye Treatment: Approaches for Cyclosporin A Delivery. Sci World J [Internet]. 2012;2012:1–11. Available from: http://www.hindawi.com/journals/tswj/2012/194848/
23. “Food Supplements - https://ec.europa.eu/food/safety/labelling_nutrition/supplements_en ”[Online].[Accessed 06-Jan-2018].
24. Rand AL, York N. Current Opinion in Ophthalmology Nutritional Supplements for Dry Eye Syndrome. Natl Institutes Heal. 2012;22(4):279–82.
25. Akhtar S, Ahmed A, Randhawa MA, Atukorala S, Arlappa N, Ismail T, et al. Prevalence of vitamin A deficiency in South Asia: causes, outcomes, and possible remedies. J Health Popul Nutr. 2013;31(4):413–23.
26. Yildirim P, Garip Y, Karci AA, Guler T. Dry eye in vitamin D deficiency: More than an incidental association. Int J Rheum Dis. 2016;19(1):49–54.
27. Yang SF, Roberts JE, Liu QH, Pang J, Sarna T. Zeaxanthin and Lutein in the Management of Eye Diseases. J Ophthalmol. 2016;2016:1–2.
28. Shen G, Qi Q, Ma X. Effect of Moisture Chamber Spectacles on Tear Functions in Dry Eye Disease. Optom Vis Sci [Internet]. 2016;93(2):158–64. Available from: http://content.wkhealth.com/linkback/openurl?sid=WKPTLP:landingpage&an=00006324-201602000-00008
29. “Medical Devices - http://www.who.int/medical_devices/full_deffinition/en/”[Online].[Accessed 09-Jan-2018].
30. “European Commission: Comunicado de imprensa- http://europa.eu/rapid/press-release_IP-12-1011_pt.htm”[Online].[Accessed 20-Out-2017].
31. “Medical Devices - https://ec.europa.eu/growth/sectors/medical-devices_pt”[Online].[Accessed 09-Jan-2018].
32. INFARMED - Autoridade Nacional do Medicamento, e Produtos de Saúde IP. Dispositivos médicos. 2008; Available from: http://www.infarmed.pt/web/infarmed/entidades/dispositivos-medicos
33. Ttopstart. Certification and registration: Medical Devices on the European market. Spottlight [Internet]. 2013;(October). Available from: http://www.ttopstart.com/uploads/ttopstart_-_spottlight_report_-_Certification_and_registration_of_medical_devices.pdf
34. “Nova legislação aplicável aos Dispositivos Médicos – Infarmed - http://www.aenfermagemeasleis.pt/2017/05/05/nova-legislacao-aplicavel-aos-dispositivos-medicos-infarmed”[Online].[Accessed 06-Jan-2018].
35. British Standards Institution. A guide to the In Vitro Diagnostic Directive. 2012;1–8. Available from: http://www.bsigroup.com/meddev/LocalFiles/en-IN/Technologies/BSI-md-ivd-diagnostic-directive-guide-brochure-UK-EN.pdf
36. Tong L, Petznick A, Lee S, Tan J. Choice of Artificial Tear Formulation for Patients With Dry Eye. Cornea. 2012;31(11):S32–6.
83
37. Schmidl D, Schmetterer L, Witkowska KJ, Unterhuber A, Aranha V, Kaya S, et al. Tear Film Thickness After Treatment With Artificial Tears in Patients With Moderate Dry Eye Disease. Cornea. 2015;34(4):421–6.
38. Freeman PD, Kahook MY. Preservatives in topical ophthalmic medications: historical and clinical perspectives. Expert Rev Ophthalmol. 2009;4(1):59–64.
39. Baudouin C, Labbé A, Liang H, Pauly A. Preservatives in eyedrops : The good , the bad and the ugly q. Prog Retin Eye Res [Internet]. 2010;29(4):312–34. Available from: http://dx.doi.org/10.1016/j.preteyeres.2010.03.001
40. Dutescu RM, Panfil C, Schrage N. Osmolarity of Prevalent Eye Drops, Side Effects, and Therapeutic Approaches. Cornea [Internet]. 2015;34(5):560–6. Available from: http://content.wkhealth.com/linkback/openurl?sid=WKPTLP:landingpage&an=00003226-201505000-00015
41. Schuerer N, Stein E, Inic-kanada A, Pucher M, Hohenadl C, Bintner N, et al. Implications for Ophthalmic Formulations : Ocular Buffers Show Varied Cytotoxic Impact on Human Corneal – Limbal. Cornea. 2017;36(6):712–8.
42. Ivarsson D, Wahlgren M. Comparison of in vitro methods of measuring mucoadhesion: Ellipsometry, tensile strength and rheological measurements. Colloids Surfaces B Biointerfaces [Internet]. 2012;92:353–9. Available from: http://dx.doi.org/10.1016/j.colsurfb.2011.12.020
43. Kashikar VS. Review Article Ophthalmic Mucoadhesive Polymers - A Literature Review. Ophthalmic Res. 2011;7(1):68–73.
44. Saraswathi B, Balaji A, Umashankar MS. Polymers in mucoadhesive drug delivery system-latest updates. Int J Pharm Pharm Sci. 2013;5(SUPPL 3):423–30.
45. Chickering DE, Mathiowitz E. Definitions, Mechanisms, and Theories of Bioadhesion. In: Mathiowitz E, Chickering DE, Lehr C-M, editors. Bioadhesive Drug Delivery Systems: Fundamentals, Novel Approaches and Develpment. New York: Marcel Dekker, Inc.; 1999. p. 1–10.
46. Hägerström H, Edsman K. Interpretation of mucoadhesive properties of polymer. Jorna. 2001;1589–99.
47. Benedetto DA, Shah DO, Kaufman HE. The instilled fluid dynamics and surface chemistry of polymers in the preocular tear film. Investig Ophthalmol. 1975;(December):887–902.
48. Russo E, Selmin F, Baldassari S, Gennari CGM, Caviglioli G, Cilurzo F, et al. A focus on mucoadhesive polymers and their application in buccal dosage forms. J Drug Deliv Sci Technol [Internet]. 2016;32:113–25. Available from: http://dx.doi.org/10.1016/j.jddst.2015.06.016
49. Boddeda B, Ratna JV, Battu H. A review on mucoadhesive polymers in ophthalmics. Int J Pharm Sci Rev Res. 2014;24(1):237–45.
50. Johnson ME, Murphy PJ, Boulton M. Effectiveness of sodium hyaluronate eyedrops in the treatment of dry eye. Graefe’s Arch Clin Exp Ophthalmol. 2006;244(1):109–12.
51. Shaikh R, Raghu T, Singh R, Garland MJ, David A, Donnelly RF. Mucoadhesive drug delivery systems. J Pharm Bioallied Sci. 2011;3(1):89–100.
52. Yu T, Andrews GP, Jones DS. Mucoadhesion and Characterization of Mucoadsive Properties. In: Neves J das, Sarmento B, editors. Mucosal Delivery of Biopharmaceuticals: Biology, Challenges and Strategies. Springer; 2014. p. 35–58.
53. Thirawong N, Nunthanid J, Puttipipatkhachorn S, Sriamornsak P. Mucoadhesive properties of various pectins on gastrointestinal mucosa : An in vitro evaluation using texture analyzer. Eur J
84
Pharm Biopharm. 2007;67:132–40.
54. Bonacucina G, Martelli S, Palmieri GF. Rheological , mucoadhesive and release properties of Carbopol gels in hydrophilic cosolvents. Int J Pharm. 2004;282:115–30.
55. Andrews GP, Donnelly L, Jones DS, Curran RM, Morrow RJ, Woolfson AD, et al. Characterization of the Rheological , Mucoadhesive , and Drug Release Properties of Highly Structured Gel Platforms for Intravaginal Drug Delivery. Biomacromolecules. 2009;10:2427–35.
56. Salzillo R, Schiraldi C, Corsuto L, Agostino AD, Filosa R, Rosa M De, et al. Optimization of hyaluronan-based eye drop formulations. Carbohydr Polym [Internet]. 2016;153:275–83. Available from: http://dx.doi.org/10.1016/j.carbpol.2016.07.106
57. Menchicchi B, Fuenzalida JP, Hensel A, Swamy MJ, David L, Rochas C, et al. Biophysical Analysis of the Molecular Interactions between Polysaccharides and Mucin. Biomacromolecules. 2015;16(3):924–35.
58. V ESPB, Reinhart T, Peppas A. Solute diffusion in swollen membranes. part ii. influence of crosslinking on diffusive properties. J Memb Sci. 1984;18:227–39.
59. Edmondson R, Broglie JJ, Adcock AF, Yang L. Three-Dimensional Cell Culture Systems and Their Applications in Drug Discovery and Cell-Based Biosensors. Assay Drug Dev Technol [Internet]. 2014;12(4):207–18. Available from: http://online.liebertpub.com/doi/abs/10.1089/adt.2014.573
60. Ravi M, Paramesh V, Kaviya SR, Anuradha E, Paul Solomon FD. 3D cell culture systems: Advantages and applications. J Cell Physiol. 2015;230(1):16–26.
61. “3D Cell Culture 101 : An Introduction to 3D Cell Culture Tools and Techniques - http://www.sigmaaldrich.com/technical-documents/articles/biology/3d-biomatrix-white-paper-3d-cell-culture-101.html”[Online].[Acessed 27-Out-2017].
62. Larson B. 3D Cell Culture: A Review of Current Techniques. BioTek. 2015;6:1–10.
63. Sanyal S. Culture and Assay Systems Used for 3D Cell Culture. Corning. 2014;
64. Shafaie S, Hutter V, Cook MT, Brown MB, Chau DYS. In Vitro Cell Models for Ophthalmic Drug Development Applications. Biores Open Access [Internet]. 2016;5(1):94–108. Available from: http://online.liebertpub.com/doi/10.1089/biores.2016.0008
65. Barar J, Asadi M, Mortazavi-Tabatabaei SA, Omidi Y. Ocular Drug Delivery; Impact of in vitro Cell Culture Models. J Ophthalmic Vis Res [Internet]. 2009;4(4):238–52. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23198080%5Cnhttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC3498862
66. Meloni M, Servi B De, Marasco D, Prete S Del. Molecular mechanism of ocular surface damage : Application to an in vitro dry eye model on human corneal epithelium. Mol Vis. 2011;(January):113–26.
67. White E, White E. Risk Management for Aseptic Processing. :22–30.
68. International Conference on Harmonisation (ICH) Quality Risk Management, Q9, ICH, November 2005.
69. Norma Portuguesa ISO 13485:2016. Dispositivos médicos Sistemas de Gestão da Qualidade e Requisitos para Fins Regulamentares., Instituto Portugês da Qualidade., 2017.
70. ISO 14971:2007 Medical Devises - Application of Risk Management to Medical Devises., Internacional Standard., 2007.
85
71. Norma Portuguesa ISO 31000. Gestão de Risco. Princípios de Linhas de Orientação., Instituto Português da Qualidade., 2013. 2012;
72. International Conference on Harmonisation (ICH) of Technical Requirement for Registration of Pharmaceuticals for Human Use, Pharmaceutical Development, Q8 (R2), ICH, August 2009.
73. Walters KA, Brain KR. Dermatological Formulation and Transdermal Systems. Dermatological Transdermal Formul. 2002;
74. Lei L. Development of a New Lubricant and Nutrient Tear Substitute [Internet]. Der Universität zu Lübeck Aus der Medizinischen Fakultät. 2004. Available from: http://d-nb.info/978081064/34
75. International Conference on Harmonisation (ICH) of Technical Requirement for Registration of Pharmaceuticals for Human Use, Stability Testing of new Drugs and Products, Q1A (R2), ICH, 2003.
76. United States Phamacopeia 31/National Formulary 26. United States Pharmacopeial Convention, Inc. 2012;(c):21–3.
77. Matthews BR, Wall GM. Stability Storage and Testing of Ophthalmic Products for Global Registration. Drug Dev Ind Pharm. 2000;26(12):1227–37.
78. European Pharmacopeia 9.0 Eficacy of Antimicrobial Preservation, European Department for the Quality of Medicines within the Council of Europe, Strasbourg, section 5.1.3;2017:577-579.
79. Montés-Micó R, Cerviño A, Ferrer-Blasco T, García-Lázaro S, Madrid-Costa D. The Tear Film and the Optical Quality of the Eye. 2010;8(4):185–92.
80. Potvin R, Makari S, Rapuano CJ. Tear film osmolarity and dry eye disease: a review of the literature. Clin Ophthalmol [Internet]. 2015;9:2039–47. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26586933%5Cnhttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC4636089
81. Norn MS. Tear fluid pH in normals, contact lens wearers, and pathological cases. 1988;66:485–9.
82. White CJ, Thomas CR, Byrne ME. Bringing comfort to the masses : A novel evaluation of comfort agent solution properties. Contact Lens Anterior Eye [Internet]. 2014;37(2):81–91. Available from: http://dx.doi.org/10.1016/j.clae.2013.07.004
83. Guillaumie F, Furrer P, Felt-Baeyens O, Fuhlendorff BL, Nymand S, Westh P, et al. Comparative studies of various hyaluronic acids produced by microbial fermentation for potential topical ophthalmic applications. J Biomed Mater Res - Part A. 2010;92(4):1421–30.
84. Ghelardi E, Celandroni F, Gueye SA, Salvetti S, Campa M, Senesi S. Antimicrobial Activity of a New Preservative for Multiuse Ophthalmic Solutions. 2013;29(6):586–90.
85. Carney LG, Mauger TF, Hill RM. Buffering in Human Tears : pH Responses to Acid and Base Challenge. 1989;30(4):747–54.
86. Houlsby RD, Ghajar M, Chavez G. Antimicrobial Activity of Borate-Buffered Solutions. 1986;(May):803–6.
87. “PubChem - https://pubchem.ncbi.nlm.nih.gov/”.[Online].[Accessed: 02-Aug-2017 ].
88. Scientific Committee on Consumer Safety. The Safety of Poly(hexamethylene) biguanide hydrochloride (PHMB). 2014;
89. “Chemical Book - http://www.chemicalbook.com/ChemicalProductProperty_EN_CB6233954.htm”.[Online].[Acces
86
sed: 02-Aug-2017].
90. Scientific Committee on Consumer Safety. Safety the safety of poly ( hexamethylene ) biguanide hydrochloride ( PHMB ). 2014.
91. “sodium hydroxymethylglycinate - http://www.thegoodscentscompany.com/data/rw1303211.html”.[Online].[Accessed 02-Aug-2017].
92. The Scientific Committee On Cosmetic Products And Non-Food Products Intended For Consumers. Evaluation And Opinion on : The Determination Of Certain Formaldehyde Releasers In Cosmetic Products. 2002;(December).
93. Strasbourg: Council Of Europe : European Directorate for the Quality of Medicines and Healthcare. EP. European Pharmacopoeia 9.0. In 2017.
94. Liu L, Shaabana N, Sheardown HD. Hydrogel Materials Interacting with Polysaccharide Wetting Agents for Contact Lens Application. 2008;5(2):2008.
95. Rayahin JE, Buhrman JS, Zhang Y, Koh TJ, Richard A, Regeneration T, et al. High and low molecular weight hyaluronic acid differentially influence macrophage activation. ACS Biomater Sci Eng. 2015;1(7):481–93.
96. Potenzone R, A.J.Hopfinger. Conformational Analysis of Conformational Properties of Hyaluronic Acid and Sodium Hyaluronate. Polym J. 1978;10(November):181–99.
97. Fujii T, Sun Y, An K, Luo Z. Mechanical properties of single hyaluronan molecules. J Biomech. 2002;35:527–31.
98. Meyer K, Palmer JW. Polysaccharide of Vitreous Humor. J Biol Chem. 1934;(107):629–34.
99. “Sodium Hyalutonat, 95%- https://aksci.com/item_detail.php?cat=H926”[Online].[Accessed 12-Aug-2017].
100. “Sodium Hyaluronate - https://en.wikipedia.org/wiki/Sodium_hyaluronate”.[Online].[Accessed: 10-Aug-2017].
101. The European Agency for the Evaluation of Medical Products. Note for Guidance on Quality of Water for Pharmaceutical Use. 2002;(May):0–5.
102. WHO Expert Committee on Specifications for Pharmaceutical Preparations. Annex 2 - WHO good manufacturing practices: water for pharmaceutical use. 2005;(929):67–89.
103. Fischbarg J. The Biology of the Eye. First Edit. Fischbarg J, editor. New York: Elsevier; 2006. 394 p.
104. Furrer P, Mayer JM, Gurny R. Ocular tolerance of preservatives and alternatives. 2002;53:263–80.
105. Michael Levin, Pharmaceutical Process Scale up, Second edition, Volume 157.
106. European Medicines Agency. Guideline on process validation for finished products - information and data to be provided in regulatory submissions. 2016;44(November):1–15.
107. Patel D, Kaul R, Bhavsar A. Manufacturing Risk Assessment Study for Sterile Dry Powder Injection of Ceftriaxone Sodium. WjppsCom [Internet]. 2016;5(6):1143–53. Available from: http://www.wjpps.com/download/article/1464675961.pdf
108. “FAILURE MODE EFFECTS ANALYSIS (FMEA) - http://asq.org/learn-about-quality/process-analysis-tools/overview/fmea.html”. [Online].[Accessed: 03-Jul-2017].
87
109. Murray B, Craft S. Failure Modes and Effects Analysis Template. In: The National Ecological Observatory Network. 2013.
110. U.S. Department of Health and Human Services Food and Drug Administration Center for Drug Evaluation and Research (CDER). Guidance for Industry Guidance for Industry - Investigating Out-of-Specification (OOS) Test Results for Pharmaceutical Production. 2006;1(August):14. Available from: http://www.fda.gov/downloads/Drugs/.../Guidances/ucm070287.pdf
111. Commitee for Proprietary Medical Products and Committe for Veterinary Medical Products; EMEA. Note for Guidance on Process Validation. 2001.
112. European Pharmacopeia 9.0 Sterility, European Department for the Quality of Medicines within the Council of Europe, Strasbourg, section 2.6.1;2017:185-188. In.
113. Mallu UR, Nair AK, Bandaru S, Sankaraiah J. Hold Time Stability Studies in Pharmaceutical Industry: Review. Pharm Reg Aff. 2012;1(4):1–8.
114. Rao M. Taking the mystery out of polymer activaion. Tall Oaks Publ Inc. 1995;
115. Jampel HD. Patient Preferences for Eye Drop Characteristics. Arch Ophthalmol [Internet]. 2003;121(4):540. Available from: http://archopht.jamanetwork.com/article.aspx?doi=10.1001/archopht.121.4.540
116. Kedik SA, Yartsev EI, Levachev SM, Panov A V, Sakaeva I V, Grigor OA, et al. Approaches To Eye Drops Efficiency Evaluation Based On Physicochemical Characteristics. 2011;45(3).
117. Dunn KC, Aotaki-Keen AE, Putkey FR, Hjelmeland LM. ARPE-19, a human retinal pigment epithelial cell line with differentiated properties. Exp Eye Res. 1996;62(2):155–69.
118. Assessing tackiness and adhesion using a pull away test on a rotational rheometer. Malvern Instruments Worldw. 2015;
119. Franck A. Understanding rheology of structured fluids. B TA instruments [Internet]. 2004;1–11. Available from: http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Understanding+Rheology+of+Structured+Fluids#2%5Cnhttp://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Understanding+rheology+of+structured+fluids#2
120. Cowman MK, Schmidt TA, Raghavan P, Stecco A. Viscoelastic Properties of Hyaluronan in Physiological Conditions. F1000Research [Internet]. 2015;(0):1–13. Available from: http://f1000research.com/articles/4-622/v1
121. Hassan EE, Gallo JM. A Simple Rheological Method for the in Vitro Assessment of Mucin- Polymer Bioadhesive Bond Strength. Pharm Res. 1990;7(5):491–5.
122. Grillet AM, Wyatt NB, Gloe LM. Polymer Gel Rheology and Adhesion. J Rheol (N Y N Y) [Internet]. 2012; Available from: http://www.intechopen.com/books/rheology/rheology-and-adhesion-of-polymer-gels
123. Becker LC, Bergfeld WF, Belsito D V., Klaassen CD, Marks JG, Shank RC, et al. Final Report of the Safety Assessment of Hyaluronic Acid, Potassium Hyaluronate, and Sodium Hyaluronate. Int J Toxicol. 2009;28(4):5–67.
124. Romero CP, Jeldres RI, Quezada GR, Concha F, Toledo PG. Zeta potential and viscosity of colloidal silica suspensions: Effect of seawater salts, pH, flocculant, and shear rate. Colloids Surfaces A Physicochem Eng Asp [Internet]. 2018;538:210–8. Available from: http://dx.doi.org/10.1016/j.colsurfa.2017.10.080
125. Servi B De, Olmiere C, Meloni M, Assay T, Death C. Hyperosmolar dryness stress on 3D HCE model : a new tool for pre-clinical assessment of tear substitutes. :6031.
88
126. Veerareddy PR. Preparation and evaluation of mucoadhesive cefdinir microcapsules. Adv Pharm Technol Res. 2011;2(2).
127. Lee M, Hwang J, Lim K. Alternatives to In Vivo Draize Rabbit Eye and Skin Irritation Tests with a Focus on 3D Reconstructed Human Cornea-Like Epithelium and Epidermis Models. Toxicol Res. 2017;33(3):191–203.
89
Appendix
Figure A 1- Ishikawa diagram for possible failures in Weighting.
90
Figure A 2- Ishikawa diagram for possible failures in Mixing.
91
Figure A 3-Ishikawa diagram for possible failures in Pre and Sterile Filtration.
92
Figure A 4- Ishikawa diagram for possible failures in Filling and Packing.
93
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30%.
SEV OCC DET SxOxD SxO
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO 14971 Failure Cause
OCC (Mean ± SD)
Mode ISO 14971 Current Process Controls
DET (Mean ±
SD) Mode RPN CRIT
ISO 14971 Result
Weig
hti
ng
Raw Material Contamination during receipt
and transferring
Inappropriate quality of raw
material which directly
affects the quality of finished
dosage form
7.5 ± 0.6
7 Significant
Due to environment,
human interference
or by the vendor
2.0 ± 0.0
2 Low
Batch verification by operators and supervisors
before manufacture
3.5 ± 0.6
4 56 14 Low Risk
Equipment Wrongly cleaned
Contaminated product
8.0 ± 0.0
8 Significant Bad cleaning
practices 2.0 ± 0.0
2 Low
Room and equipment verification
by operators and
supervisors before
manufacture
2.0 ± 0.0
2 32 16 Low Risk
Raw material loss during
transference
Affects the quality of finished
dosage form
6.0 ± 1.2
5 Moderate Operating
errors 3.5 ± 0.6
3 Low Proper training
3.5 ± 0.6
4 60 15 Low Risk
Absence of Raw Material
Affects the production schedule
8.5 ± 0.6
8 Significant Failures of
supplier 1.5 ± 0.6
2 Low
Verification of raw material two weeks
before manufacture
3.5 ± 1.7
2 32 16 Low Risk
Long operator exposure
during handling
Contaminated product
4.5 ± 0.6
5 Moderate Operating
errors 3.0 ± 0.0
3 Low
Utilization of Personal protective equipment
2.5 ± 0.6
3 45 15 Low Risk
Laminar Flux failure
Contaminated product
7.5 ± 0.6
8 Significant Equipment malfunction
2.5 ± 0.6
2 Low
Semestral verification
and every six months
2.5 ± 0.6
2 32 16 Low Risk
94
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET SxOxD SxO
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
Weig
hti
ng
Uncalibrated scale
Affects the quality of finished
dosage form
6.0 ± 1.15
5 Moderate Equipment malfunction
2.5 ± 0.6
3 Low
Annual calibration and daily
verification with two
calibrated weights.
Fortnightly verification with more calibrated weights
2.5 ± 0.6
3 45 15 Low Risk
HEPA filter lack of integrity
Contaminated product
8.5 ± 0.6
8 Significant
Improper HEPA filter efficiency,
variations in velocity of air, flow
pattern of air or
leaking of the filter
2.5 ± 0.6
2 Low
Semestral verification
and every six months
2.0 ± 0.0
2 32 16 Low Risk
Procedure failures
Affects the quality of finished
dosage form
7.5 ± 0.6
7 Significant Invalidated or Unmet procedure
4.0 ± 0.0
4 Medium
Required filling of the
Batch Master Record
3.0 ± 0.0
3 84 28 High Risk
Non-verification of criteria
Inadequate weighing chamber
conditions result in poor final product
quality
6.5 ± 0.6
7 Significant Unmet
procedures 3.0 ± 0.0
3 Low
Verification of the
conditions within the
specifications
3.0 ± 0.0
3 63 21 Low Risk
Misuse use of uniform Contaminated
product 7.0 ± 0.0
7 Significant Wrong or
incomplete training
2.0 ± 0.0
2 Low Uniform
procedure 2.0 ± 0.0
2 28 14 Low Risk
95
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET SxOxD SxO
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
Weig
hti
ng
Wrong amount of raw material
Affects the quality of finished
dosage form
5.5 ± 0.6
5 Moderate
Weighing errors/ Unmet
procedures
2.0 ± 0.0
2 Low
Raw materials are checked in duplicate
and preservatives
and API´s checked by supervisors
before manufacture
2.0 ± 0.0
2 20 10 Low Risk
Untrained operators
Affects the quality of finished
dosage form
6.5 ± 1.7
8 Significant Wrong or
incomplete training
2.5 ± 0.6
3 Low Initial and
continuous training
2.5 ± 0.6
3 72 24 Low Risk
Mix
ing
Incomplete dissolution/
Insufficient mixing time
Affects the quality of finished
dosage form
7.0 ± 0.0
7 Significant
Little mixing time,
mixing blade with little force or velocity
3.0 ± 1.2
3 Low
Visual control and validation
of the manufacturing
process
2.5 ± 0.6
2 42 21 Low Risk
Excess/Low Capacity of the mixer
Equipment malfunction, affects the quality of finished
dosage form
4.5 ± 0.6
5 Moderate Wrong batch
volume
3.0 ± 0.6
3 Low
Visual control and validation
of the manufacturing
process
3.5 ± 1.7
2 30 15 Low Risk
Velocity Failure
Incomplete dissolution, effects the quality of finished
dosage form
6.0 ± 1.2
7 Significant Equipment malfunction
3.5 ± 0.6
3 Low Backup shaker
4.0 ± 0.0
4 84 21 Low Risk
Change raw material placement order
Affects the quality of finished
dosage form
4.5 ± 0.6
5 Moderate Improper working of personnel
3.5 ± 0.0
3 Low
Batch Master Record with
all the necessary
steps
3.5 ± 0.6
4 60 15 Low Risk
96
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET SxOxD SxO
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
Mix
ing
Raw material loss during transference
Affects the quality of finished
dosage form
5.5 ± 0.6
5 Moderate Operating
errors 3.3 ± 1.2
4 Medium Proper training
3.0 ± 1.2
4 80 20 Low Risk
Equipment Wrongly cleaned
Contaminated product
7.0 ± 0.0
5 Moderate Bad cleaning
practices 3.5 ± 0.6
3 Low
Room and equipment verification
by operators and
supervisors before
manufacture
3.0 ± 0.0
3 63 21 Low Risk
Misuse use of uniform Contaminated
product 7.0 ± 0.0
7 Significant Wrong or
incomplete training
3.5 ± 0.6
3 Low Uniform
procedure 2.5 ± 1.7
4 84 21 Low Risk
Procedure failures
Affects the quality of finished
dosage form
4.5 ± 0.6
5 Moderate Invalidated or
Unmet procedure
3.0 ± 0.6
3 Low
Required filling of the
Batch Master Record
5.0 ± 2.3
3 45 15 Low Risk
Incorrectly mounted equipment
Equipment malfunction, affects the quality of finished
dosage form
5.5 ± 0.6
5 Moderate Unmet
procedures 3.0 ± 0.6
3 Low Proper training
3.0 ± 0.0
3 45 15 Low Risk
IPC
/FP
C
Product harvest in an inappropriate container
Contaminated product
9.0 ± 0.0
9 Significant
Due to improper storage of containers,
environment, human
interference
3.0 ± 0.0
3 Low The
bioburden is determined
3.0 ± 0.0
3 81 27 Low Risk
97
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET SxOxD SxO
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971 Current Process
Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
IPC
/FP
C
Presence of colour and odour
Affects the quality of finished dosage
form
7.3 ± 2.1
9 Significant Bad
manufacturing practices
2.3 ± 1.3
1 Low
Samples are taken from the top and bottom for visual
check
3.3 ± 1.3
1 9 9 Low Risk
High/Low pH
Affects the quality of finished dosage
form
7.3 ± 2.1
9 Significant
Failures in weighting or loss of raw
material
2.3 ± 1.3
4 Medium
pH meter is calibrated daily. pH specified in batch
record and readings are
double verified. QC confirms
3.3 ± 1.3
1 36 36 High Risk
High/Low osmolality
Affects the quality of finished dosage
form
8.5 ± 1.0
9 Significant
Failures in weighting or loss of raw
material
3.8 ± 1.3
4 Medium
Osmometer is calibrated monthly against patterns
solutions.Calibrated daily against a
reference solution. Osmolality
specified in batch record and osmolalities readings are
double verified. After all
adjustments QC confirms
1.8 ± 1.0
2 72 36 High Risk
98
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET RPN CRIT
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
IPC
/FP
C
High/Low Viscosity
Affects the quality of finished
dosage form
7.5 ± 0.6
8 Significant
Failures in weighting or loss of raw
material
4.3 ± 1.3
4 Medium
Viscometer calibrated
annually by an external
entity. Viscosity
specified in batch record
and readings are
double verified. QC
confirms
1.8 ± 1.0
1 32 32 High Risk
Absence of Sodium Hyaluronate
Wrong formulation
8.5 ± 0.6
9 Significant
Failures in weighting or loss of raw
material
1.0 ± 0.0
1 Low
Identification of Sodium
Hyaluronate by IR
method
1.5 ± 1.0
1 9 9 Low Risk
Sodium Hyaluronate UV Assay out of range
Wrong formulation
8.5 ± 0.6
9 Significant
Failures in weighting or loss of raw
material
3.5 ± 1.3
4 Medium
Sodium Hyaluronate percentage specified in
batch record determined by UV/Vis method.
1.8 ± 0.5
2 72 36 High Risk
Elevated Microbial concentration/ non-
sterile
Contaminated product
9.0 ± 0.8
9 Significant
Operating errors in
past procedures
3.8 ± 1.3
4 Medium
The
bioburden is
determined/
Sterility test
3.3 ± 0.5
3 108 36 High Risk
False Results
Validated certificates with wrong
results
9.3 ± 1.0
10 Significant Poorly
performed analysis
1.8 ± 0.5
2 Low
All methods are
previously validated
3.5 ± 1.3
4 80 20 Low Risk
.
99
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET RPN CRIT
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO 14971 Failure Cause
OCC (Mean ± SD)
Mode ISO 14971 Current Process Controls
DET (Mean ±
SD) Mode RPN CRIT
ISO 14971 Result
Pre
an
d S
teri
le
Fil
trati
on
HEPA filter lack of integrity
Contaminated product
7.5 ± 3.0
8 Significant
Improper HEPA filter efficiency,
variations in velocity of air, flow pattern of air or leaking of the filter
4.0 ± 4.0
2 Low Verification by
external entities
3.5 ± 2.4
2 32 16 Low Risk
Laminar Flux failure
Contaminated product
6.8 ± 2.5
8 Significant Equipment malfunction
2.3 ± 0.5
2 Low
Every six months
verification and room, equipment
and operators environmental
control
3.0 ± 2.7
2 32 16 Low Risk
Membrane Integrity failure
Unable to sterilized
8.3 ± 1.3
8 Significant
Defective membrane,
misplacement of the
membrane
3.4 ± 3.5
2 Low Bubble Point
Control 3.3 ± 2.5
2 32 16 Low Risk
Product obstruction
Bubble point failure
6.8 ± 2.5
5 Moderate
Improper equipment,
elevated product viscosity
5.0 ± 3.5
4 Medium
Deposit capable of
pressurizing at high
pressures
2.3 ± 0.5
2 40 20 Low Risk
Non-sterile equipment/parts
Contaminated product
5.0 ± 3.6
8 Significant Bad cleaning
practices 1.5 ± 0.6
2 Low
Validation of autoclave
loads. Well packaged material
3.8 ± 3.0
3 48 16 Low Risk
100
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET RPN CRIT
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
Pre
an
d S
teri
le
Fil
trati
on
Contaminated Filter Contaminated
product 9.0 ± 1.2
8 Significant Poor filter
quality 3.8 ± 4.2
2 Low
Sterilization of the filter in an
autoclave with a valid
load whenever there is a
filling
7.5 ± 2.9
5 80 16 Low Risk
Broken/ Misplaced Filter
Unfiltered Product
8.0 ± 0.8
8 Significant Operating
errors/ Poor filter quality
3.8 ± 4.2
2 Low Bubble Point
Control 3.3 ± 2.5
2 32 16 Low Risk
Procedure failures
Affects the quality of finished
dosage form
7.7 ± 1.5
7 Significant Invalidated or Unmet procedure
4.3 ± 3.2
3 Low
Required filling of the
Batch Master Record
3.3 ± 0.5
3 63 21 Low Risk
Misuse use of uniform Contaminated
product 8.0 ± 0.8
8 Significant Wrong or
incomplete training
4.3 ± 3.3
2 Low
Uniform procedure
and operators environmental
control
1.5 ± 0.6
2 32 16 Low Risk
Incorrectly mounted filtration system
Equipment malfunction, affects the quality of finished
dosage form
8.0 ± 0.8
8 Significant Operating
errors 4.5 ± 3.7
3 Low Proper training
2.5 ± 1.0
2 48 24 Low Risk
Filli
ng
an
d P
ackin
g
HEPA filter lack of integrity
Contaminated product
9.0 ± 1.2
8 Significant
Improper HEPA filter efficiency,
variations in velocity of air, flow
pattern of air or
leaking of the filter
4.3 ± 3.9
2 Low Verification by
external entities
3.5 ± 1.9
2 32 16 Low Risk
101
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET RPN CRIT
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
Filli
ng
an
d P
ackin
g
Laminar Flux failure Contaminated
product 6.8 ± 2.5
8 Significant Equipment malfunction
2.5 ± 0.6
2 Low
Every six months
verification and room, equipment
and operators environmental
control
2.6 ± 2.2
2 32 16 Low Risk
Non-sterile equipment/parts
Contaminated product
5.5 ± 3.3
8 Significant Bad
cleaning practices
2.5 ± 1.0
3 Low
Validation of autoclave
loads. Well packaged material
2.8 ± 3.5
1 24 24 Low Risk
Filling area unclean Loss of sterility
8.3 ± 0.5
8 Significant
Failure in maintenance
of area classification
3.5 ± 3.7
2 Low
Cleaning procedure and room
environmental control
2.3 ± 1.3
2 32 16 Low Risk
Uniformity of Volume Failure
Final product with wrong
volume
5.5 ± 2.5
5 Moderate
Uncalibrated equipment scale, non-verification of the scale
4.0 ± 3.4
2 Low
Tare of 12 empty bottles, 10 bottles for
weight adjustment, control of
liquid weight every 30 minutes
3.4 ± 2.0
2 20 10 Low Risk
Filling material with defects
Loss of sterility, loss of integrity
6.0 ± 2.5
6 Moderate Failures of
supplier 3.8 ± 4.2
2 Low
Inline sealing in all batch, defective
material is rejected
2.8 ± 1.5
4 48 12 Low Risk
Sealing defective Loss of sterility
8.3 ± 0.5
8 Significant
Defective sealant, machine failure
4.0 ± 3.4
2 Low
Inline sealing in all batch, defective
material is rejected
1.8 ± 0.5
2 32 16 Low Risk
102
Table A 12- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation).
SEV OCC DET RPN CRIT
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
Filli
ng
an
d P
ackin
g
Incorrectly mounted equipment
Equipment malfunction, affects the quality of finished
dosage form
8.3 ± 0.5
8 Significant Unmet
procedures 4.3 ± 3.4
2 Low Proper training
2.5 ± 1.0
3 48 16 Low Risk
Procedure failures
Affects the quality of finished
dosage form
7.8 ± 1.0
7 Significant Invalidated or
Unmet procedure
4.3 ± 2.5
3 Low
Required filling of the
Batch Master Record
3.0 ± 0.0
3 63 21 Low Risk
Misuse use of uniform Contaminated
product 8.3 ± 0.5
8 Significant Wrong or
incomplete training
4.0 ± 3.4
2 Low
Uniform procedure
and operators environmental
control
1.5 ± 0.6
2 32 16 Low Risk
Labelling errors
Misleading labels and
market complains
6.8 ± 2.1
5 Moderate
Improper checking of
labels including
Names and amount of
active ingredients,
Storage requirements, Control or lot
number, Appropriate
auxiliary labelling
(including precautions)
4.5 ± 3.8
2 Low
Verification and approval by the Quality Control, batch
number confirmation,
date of manufacture and validity confirmation carried out
inline
3.5 ± 1.7
3 30 10 Low Risk
103
Table A 1- FMEA and the ISO 14971 qualitative analyses of the manufacturing process of HA 0.15% and HA 0.30% (Continuation). SEV OCC DET RPN CRIT
Pro
cess
Failure Modes Failure Effects
SEV (Mean ± SD)
Mode ISO
14971 Failure Cause
OCC (Mean ± SD)
Mode ISO
14971
Current Process Controls
DET (Mean ± SD)
Mode RPN CRIT ISO
14971 Result
Sto
rag
e
Mixing up two different types of raw material
Wrong product
storage and use
7.5 ± 0.6
7 Significant Improper working of personnel
2.0 ± 0.0
2 Low
Specific places to
avoid misplacement
2.3 ± 0.5
2 28 14 Low Risk
Mixing up two different types of packing
material
Wrong product
storage and use
7.0 ± 0.0
7 Significant Improper working of personnel
2.0 ± 0.0
2 Low
Specific places to
avoid misplacement
2.0 ± 0.0
2 28 14 Low Risk
Wrong stability test conditions
Wrong data regarding
the product stability
8.0 ± 0.0
8 Significant
Operating errors in product quantity,
storage time and type of incubator
2.3 ± 0.5
3 Low
Training of operators, instructions documented
in batch record. Storage
conditions verified in validation. Automatic email alert.
Temperature checked
twice a day by an
operator and registration on the form.
1.3 ± 0.5
2 48 24 Low Risk
Storage room conditions out of
specification
Degradation of product
7.0 ± 0.0
7 Significant
Storage room
temperature, humidity
2.0 ± 0.0
2 Low
Stability studies.
Automatic email alert.
Temperature checked
twice a day by an
operator and registration on the form.
2.5 ± 0.6
2 28 14 Low Risk